diff --git "a/checkpoint-53000/trainer_state.json" "b/checkpoint-53000/trainer_state.json" new file mode 100644--- /dev/null +++ "b/checkpoint-53000/trainer_state.json" @@ -0,0 +1,371033 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 0.53, + "eval_steps": 500, + "global_step": 53000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 1e-05, + "grad_norm": 1.085229352812366, + "learning_rate": 3e-06, + "loss": 10.849, + "step": 1 + }, + { + "epoch": 2e-05, + "grad_norm": 1.0764689186661929, + "learning_rate": 6e-06, + "loss": 10.8489, + "step": 2 + }, + { + "epoch": 3e-05, + "grad_norm": 1.0926036068515363, + "learning_rate": 9e-06, + "loss": 10.8486, + "step": 3 + }, + { + "epoch": 4e-05, + "grad_norm": 1.0859011783792423, + "learning_rate": 1.2e-05, + "loss": 10.848, + "step": 4 + }, + { + "epoch": 5e-05, + "grad_norm": 1.0906873388641662, + "learning_rate": 1.5e-05, + "loss": 10.8453, + "step": 5 + }, + { + "epoch": 6e-05, + "grad_norm": 1.0895888734627917, + "learning_rate": 1.8e-05, + "loss": 10.8447, + "step": 6 + }, + { + "epoch": 7e-05, + "grad_norm": 1.0913329404483254, + "learning_rate": 2.1000000000000002e-05, + "loss": 10.8355, + "step": 7 + }, + { + "epoch": 8e-05, + "grad_norm": 1.0766237663279077, + "learning_rate": 2.4e-05, + "loss": 10.8141, + "step": 8 + }, + { + "epoch": 9e-05, + "grad_norm": 1.0617425338278697, + "learning_rate": 2.7e-05, + "loss": 10.8099, + "step": 9 + }, + { + "epoch": 0.0001, + "grad_norm": 1.080271312928173, + "learning_rate": 3e-05, + "loss": 10.7983, + "step": 10 + }, + { + "epoch": 0.00011, + "grad_norm": 1.0663119503184246, + "learning_rate": 3.2999999999999996e-05, + "loss": 10.7844, + "step": 11 + }, + { + "epoch": 0.00012, + "grad_norm": 1.0694505502575085, + "learning_rate": 3.6e-05, + "loss": 10.7739, + "step": 12 + }, + { + "epoch": 0.00013, + "grad_norm": 1.0509206118753271, + "learning_rate": 3.9e-05, + "loss": 10.7554, + "step": 13 + }, + { + "epoch": 0.00014, + "grad_norm": 1.0427932624919014, + "learning_rate": 4.2000000000000004e-05, + "loss": 10.7452, + "step": 14 + }, + { + "epoch": 0.00015, + "grad_norm": 1.0323600599358198, + "learning_rate": 4.4999999999999996e-05, + "loss": 10.736, + "step": 15 + }, + { + "epoch": 0.00016, + "grad_norm": 1.0039727604034705, + "learning_rate": 4.8e-05, + "loss": 10.7191, + "step": 16 + }, + { + "epoch": 0.00017, + "grad_norm": 0.9736033767894897, + "learning_rate": 5.1000000000000006e-05, + "loss": 10.7049, + "step": 17 + }, + { + "epoch": 0.00018, + "grad_norm": 0.9665971933165531, + "learning_rate": 5.4e-05, + "loss": 10.69, + "step": 18 + }, + { + "epoch": 0.00019, + "grad_norm": 0.9428089164290921, + "learning_rate": 5.7e-05, + "loss": 10.6767, + "step": 19 + }, + { + "epoch": 0.0002, + "grad_norm": 0.943192779381936, + "learning_rate": 6e-05, + "loss": 10.666, + "step": 20 + }, + { + "epoch": 0.00021, + "grad_norm": 0.9246956411226879, + "learning_rate": 6.3e-05, + "loss": 10.6525, + "step": 21 + }, + { + "epoch": 0.00022, + "grad_norm": 0.9206776841693647, + "learning_rate": 6.599999999999999e-05, + "loss": 10.6378, + "step": 22 + }, + { + "epoch": 0.00023, + "grad_norm": 0.9152142474292139, + "learning_rate": 6.9e-05, + "loss": 10.626, + "step": 23 + }, + { + "epoch": 0.00024, + "grad_norm": 0.9101286615991204, + "learning_rate": 7.2e-05, + "loss": 10.613, + "step": 24 + }, + { + "epoch": 0.00025, + "grad_norm": 0.9100685942512898, + "learning_rate": 7.500000000000001e-05, + "loss": 10.5998, + "step": 25 + }, + { + "epoch": 0.00026, + "grad_norm": 0.9093157110032968, + "learning_rate": 7.8e-05, + "loss": 10.5859, + "step": 26 + }, + { + "epoch": 0.00027, + "grad_norm": 0.9036082412984342, + "learning_rate": 8.1e-05, + "loss": 10.5742, + "step": 27 + }, + { + "epoch": 0.00028, + "grad_norm": 0.9033464449252919, + "learning_rate": 8.400000000000001e-05, + "loss": 10.5612, + "step": 28 + }, + { + "epoch": 0.00029, + "grad_norm": 0.9067163029768021, + "learning_rate": 8.7e-05, + "loss": 10.5465, + "step": 29 + }, + { + "epoch": 0.0003, + "grad_norm": 0.9152916501496612, + "learning_rate": 8.999999999999999e-05, + "loss": 10.5321, + "step": 30 + }, + { + "epoch": 0.00031, + "grad_norm": 0.9122391999445313, + "learning_rate": 9.3e-05, + "loss": 10.5167, + "step": 31 + }, + { + "epoch": 0.00032, + "grad_norm": 0.9090668181794543, + "learning_rate": 9.6e-05, + "loss": 10.5034, + "step": 32 + }, + { + "epoch": 0.00033, + "grad_norm": 0.9095629132760054, + "learning_rate": 9.900000000000001e-05, + "loss": 10.4865, + "step": 33 + }, + { + "epoch": 0.00034, + "grad_norm": 0.9107692784496799, + "learning_rate": 0.00010200000000000001, + "loss": 10.4706, + "step": 34 + }, + { + "epoch": 0.00035, + "grad_norm": 0.9162740070354684, + "learning_rate": 0.00010500000000000002, + "loss": 10.4535, + "step": 35 + }, + { + "epoch": 0.00036, + "grad_norm": 0.9114324016484531, + "learning_rate": 0.000108, + "loss": 10.4367, + "step": 36 + }, + { + "epoch": 0.00037, + "grad_norm": 0.901331547883087, + "learning_rate": 0.000111, + "loss": 10.4196, + "step": 37 + }, + { + "epoch": 0.00038, + "grad_norm": 0.9056352937834914, + "learning_rate": 0.000114, + "loss": 10.4003, + "step": 38 + }, + { + "epoch": 0.00039, + "grad_norm": 0.9071824304759276, + "learning_rate": 0.000117, + "loss": 10.3815, + "step": 39 + }, + { + "epoch": 0.0004, + "grad_norm": 0.911150856035614, + "learning_rate": 0.00012, + "loss": 10.3599, + "step": 40 + }, + { + "epoch": 0.00041, + "grad_norm": 0.9001357892993758, + "learning_rate": 0.000123, + "loss": 10.3411, + "step": 41 + }, + { + "epoch": 0.00042, + "grad_norm": 0.9057215347961051, + "learning_rate": 0.000126, + "loss": 10.3208, + "step": 42 + }, + { + "epoch": 0.00043, + "grad_norm": 0.9101872556211666, + "learning_rate": 0.000129, + "loss": 10.2978, + "step": 43 + }, + { + "epoch": 0.00044, + "grad_norm": 0.9097345366629672, + "learning_rate": 0.00013199999999999998, + "loss": 10.2768, + "step": 44 + }, + { + "epoch": 0.00045, + "grad_norm": 0.9111670268341607, + "learning_rate": 0.000135, + "loss": 10.2539, + "step": 45 + }, + { + "epoch": 0.00046, + "grad_norm": 0.9030473796809102, + "learning_rate": 0.000138, + "loss": 10.2322, + "step": 46 + }, + { + "epoch": 0.00047, + "grad_norm": 0.9086055117133346, + "learning_rate": 0.000141, + "loss": 10.2064, + "step": 47 + }, + { + "epoch": 0.00048, + "grad_norm": 0.9060414917101882, + "learning_rate": 0.000144, + "loss": 10.1837, + "step": 48 + }, + { + "epoch": 0.00049, + "grad_norm": 0.9059366436676172, + "learning_rate": 0.000147, + "loss": 10.1599, + "step": 49 + }, + { + "epoch": 0.0005, + "grad_norm": 0.9106165244124662, + "learning_rate": 0.00015000000000000001, + "loss": 10.1343, + "step": 50 + }, + { + "epoch": 0.00051, + "grad_norm": 0.9110992341155927, + "learning_rate": 0.000153, + "loss": 10.1078, + "step": 51 + }, + { + "epoch": 0.00052, + "grad_norm": 0.9077294523989683, + "learning_rate": 0.000156, + "loss": 10.0815, + "step": 52 + }, + { + "epoch": 0.00053, + "grad_norm": 0.8995624264210066, + "learning_rate": 0.000159, + "loss": 10.0581, + "step": 53 + }, + { + "epoch": 0.00054, + "grad_norm": 0.9187536344258231, + "learning_rate": 0.000162, + "loss": 10.026, + "step": 54 + }, + { + "epoch": 0.00055, + "grad_norm": 0.9102867236672618, + "learning_rate": 0.000165, + "loss": 10.0019, + "step": 55 + }, + { + "epoch": 0.00056, + "grad_norm": 0.9031170239719724, + "learning_rate": 0.00016800000000000002, + "loss": 9.9743, + "step": 56 + }, + { + "epoch": 0.00057, + "grad_norm": 0.9090239107255728, + "learning_rate": 0.000171, + "loss": 9.9467, + "step": 57 + }, + { + "epoch": 0.00058, + "grad_norm": 0.9070896749665766, + "learning_rate": 0.000174, + "loss": 9.9223, + "step": 58 + }, + { + "epoch": 0.00059, + "grad_norm": 0.9116582783399498, + "learning_rate": 0.000177, + "loss": 9.8905, + "step": 59 + }, + { + "epoch": 0.0006, + "grad_norm": 0.899071209460366, + "learning_rate": 0.00017999999999999998, + "loss": 9.8642, + "step": 60 + }, + { + "epoch": 0.00061, + "grad_norm": 0.8984032767613607, + "learning_rate": 0.000183, + "loss": 9.8367, + "step": 61 + }, + { + "epoch": 0.00062, + "grad_norm": 0.9043779152804675, + "learning_rate": 0.000186, + "loss": 9.8058, + "step": 62 + }, + { + "epoch": 0.00063, + "grad_norm": 0.8969230453763369, + "learning_rate": 0.000189, + "loss": 9.7784, + "step": 63 + }, + { + "epoch": 0.00064, + "grad_norm": 0.8929896633083918, + "learning_rate": 0.000192, + "loss": 9.7485, + "step": 64 + }, + { + "epoch": 0.00065, + "grad_norm": 0.9016512171852502, + "learning_rate": 0.00019500000000000002, + "loss": 9.7165, + "step": 65 + }, + { + "epoch": 0.00066, + "grad_norm": 0.8946267084059, + "learning_rate": 0.00019800000000000002, + "loss": 9.6927, + "step": 66 + }, + { + "epoch": 0.00067, + "grad_norm": 0.8963872265737496, + "learning_rate": 0.000201, + "loss": 9.6633, + "step": 67 + }, + { + "epoch": 0.00068, + "grad_norm": 0.9035635569767697, + "learning_rate": 0.00020400000000000003, + "loss": 9.6313, + "step": 68 + }, + { + "epoch": 0.00069, + "grad_norm": 0.8884430485081615, + "learning_rate": 0.00020700000000000002, + "loss": 9.604, + "step": 69 + }, + { + "epoch": 0.0007, + "grad_norm": 0.901710386427562, + "learning_rate": 0.00021000000000000004, + "loss": 9.5675, + "step": 70 + }, + { + "epoch": 0.00071, + "grad_norm": 0.8913815692585527, + "learning_rate": 0.00021299999999999997, + "loss": 9.5403, + "step": 71 + }, + { + "epoch": 0.00072, + "grad_norm": 0.8947982003050186, + "learning_rate": 0.000216, + "loss": 9.5138, + "step": 72 + }, + { + "epoch": 0.00073, + "grad_norm": 0.8936046055705469, + "learning_rate": 0.00021899999999999998, + "loss": 9.4802, + "step": 73 + }, + { + "epoch": 0.00074, + "grad_norm": 0.894403551075387, + "learning_rate": 0.000222, + "loss": 9.4539, + "step": 74 + }, + { + "epoch": 0.00075, + "grad_norm": 0.8897139679621429, + "learning_rate": 0.000225, + "loss": 9.4187, + "step": 75 + }, + { + "epoch": 0.00076, + "grad_norm": 0.89755782108852, + "learning_rate": 0.000228, + "loss": 9.3929, + "step": 76 + }, + { + "epoch": 0.00077, + "grad_norm": 0.8917412779828411, + "learning_rate": 0.000231, + "loss": 9.3649, + "step": 77 + }, + { + "epoch": 0.00078, + "grad_norm": 0.886930008239094, + "learning_rate": 0.000234, + "loss": 9.3351, + "step": 78 + }, + { + "epoch": 0.00079, + "grad_norm": 0.8893115023720741, + "learning_rate": 0.00023700000000000001, + "loss": 9.2992, + "step": 79 + }, + { + "epoch": 0.0008, + "grad_norm": 0.8917772276535281, + "learning_rate": 0.00024, + "loss": 9.2632, + "step": 80 + }, + { + "epoch": 0.00081, + "grad_norm": 0.8971231461853929, + "learning_rate": 0.00024300000000000002, + "loss": 9.2316, + "step": 81 + }, + { + "epoch": 0.00082, + "grad_norm": 0.9071155901294882, + "learning_rate": 0.000246, + "loss": 9.206, + "step": 82 + }, + { + "epoch": 0.00083, + "grad_norm": 0.8963243651316662, + "learning_rate": 0.00024900000000000004, + "loss": 9.1708, + "step": 83 + }, + { + "epoch": 0.00084, + "grad_norm": 0.9002519637093493, + "learning_rate": 0.000252, + "loss": 9.1395, + "step": 84 + }, + { + "epoch": 0.00085, + "grad_norm": 0.8962870771351267, + "learning_rate": 0.000255, + "loss": 9.1159, + "step": 85 + }, + { + "epoch": 0.00086, + "grad_norm": 0.8973951423301171, + "learning_rate": 0.000258, + "loss": 9.0799, + "step": 86 + }, + { + "epoch": 0.00087, + "grad_norm": 0.8894399774297843, + "learning_rate": 0.000261, + "loss": 9.0571, + "step": 87 + }, + { + "epoch": 0.00088, + "grad_norm": 0.8954974565265822, + "learning_rate": 0.00026399999999999997, + "loss": 9.0211, + "step": 88 + }, + { + "epoch": 0.00089, + "grad_norm": 0.8903887799592864, + "learning_rate": 0.000267, + "loss": 8.9943, + "step": 89 + }, + { + "epoch": 0.0009, + "grad_norm": 0.8896831425922581, + "learning_rate": 0.00027, + "loss": 8.9666, + "step": 90 + }, + { + "epoch": 0.00091, + "grad_norm": 0.8841935030693385, + "learning_rate": 0.000273, + "loss": 8.9365, + "step": 91 + }, + { + "epoch": 0.00092, + "grad_norm": 0.8862593694142118, + "learning_rate": 0.000276, + "loss": 8.9052, + "step": 92 + }, + { + "epoch": 0.00093, + "grad_norm": 0.8881041222250594, + "learning_rate": 0.000279, + "loss": 8.8752, + "step": 93 + }, + { + "epoch": 0.00094, + "grad_norm": 0.8868931156100198, + "learning_rate": 0.000282, + "loss": 8.8494, + "step": 94 + }, + { + "epoch": 0.00095, + "grad_norm": 0.8826333164427848, + "learning_rate": 0.000285, + "loss": 8.8203, + "step": 95 + }, + { + "epoch": 0.00096, + "grad_norm": 0.8910955494857569, + "learning_rate": 0.000288, + "loss": 8.7905, + "step": 96 + }, + { + "epoch": 0.00097, + "grad_norm": 0.8809279978723125, + "learning_rate": 0.000291, + "loss": 8.7668, + "step": 97 + }, + { + "epoch": 0.00098, + "grad_norm": 0.8829202734417477, + "learning_rate": 0.000294, + "loss": 8.73, + "step": 98 + }, + { + "epoch": 0.00099, + "grad_norm": 0.8858154061322314, + "learning_rate": 0.000297, + "loss": 8.7072, + "step": 99 + }, + { + "epoch": 0.001, + "grad_norm": 0.8861735275161908, + "learning_rate": 0.00030000000000000003, + "loss": 8.6806, + "step": 100 + }, + { + "epoch": 0.00101, + "grad_norm": 0.8886468466177152, + "learning_rate": 0.00030300000000000005, + "loss": 8.6455, + "step": 101 + }, + { + "epoch": 0.00102, + "grad_norm": 0.8755547614914917, + "learning_rate": 0.000306, + "loss": 8.625, + "step": 102 + }, + { + "epoch": 0.00103, + "grad_norm": 0.879237266457317, + "learning_rate": 0.000309, + "loss": 8.6003, + "step": 103 + }, + { + "epoch": 0.00104, + "grad_norm": 0.8697854479956653, + "learning_rate": 0.000312, + "loss": 8.5786, + "step": 104 + }, + { + "epoch": 0.00105, + "grad_norm": 0.8733949043334459, + "learning_rate": 0.000315, + "loss": 8.552, + "step": 105 + }, + { + "epoch": 0.00106, + "grad_norm": 0.8706148334292045, + "learning_rate": 0.000318, + "loss": 8.5224, + "step": 106 + }, + { + "epoch": 0.00107, + "grad_norm": 0.8653683244965015, + "learning_rate": 0.000321, + "loss": 8.4981, + "step": 107 + }, + { + "epoch": 0.00108, + "grad_norm": 0.8737239863866451, + "learning_rate": 0.000324, + "loss": 8.4698, + "step": 108 + }, + { + "epoch": 0.00109, + "grad_norm": 0.8706512729216435, + "learning_rate": 0.000327, + "loss": 8.4501, + "step": 109 + }, + { + "epoch": 0.0011, + "grad_norm": 0.8842127152624679, + "learning_rate": 0.00033, + "loss": 8.4274, + "step": 110 + }, + { + "epoch": 0.00111, + "grad_norm": 0.9007754832304464, + "learning_rate": 0.000333, + "loss": 8.3985, + "step": 111 + }, + { + "epoch": 0.00112, + "grad_norm": 0.9159855921530741, + "learning_rate": 0.00033600000000000004, + "loss": 8.3784, + "step": 112 + }, + { + "epoch": 0.00113, + "grad_norm": 0.9063278036144603, + "learning_rate": 0.000339, + "loss": 8.3391, + "step": 113 + }, + { + "epoch": 0.00114, + "grad_norm": 0.8437820836704115, + "learning_rate": 0.000342, + "loss": 8.3286, + "step": 114 + }, + { + "epoch": 0.00115, + "grad_norm": 0.8612821674982505, + "learning_rate": 0.00034500000000000004, + "loss": 8.3022, + "step": 115 + }, + { + "epoch": 0.00116, + "grad_norm": 0.8581012057508914, + "learning_rate": 0.000348, + "loss": 8.2786, + "step": 116 + }, + { + "epoch": 0.00117, + "grad_norm": 0.834681162463853, + "learning_rate": 0.000351, + "loss": 8.2472, + "step": 117 + }, + { + "epoch": 0.00118, + "grad_norm": 0.8526508674143746, + "learning_rate": 0.000354, + "loss": 8.2312, + "step": 118 + }, + { + "epoch": 0.00119, + "grad_norm": 0.8471842679679056, + "learning_rate": 0.000357, + "loss": 8.2186, + "step": 119 + }, + { + "epoch": 0.0012, + "grad_norm": 0.8238029079166322, + "learning_rate": 0.00035999999999999997, + "loss": 8.1917, + "step": 120 + }, + { + "epoch": 0.00121, + "grad_norm": 0.832770075662114, + "learning_rate": 0.000363, + "loss": 8.1641, + "step": 121 + }, + { + "epoch": 0.00122, + "grad_norm": 0.8176689288160716, + "learning_rate": 0.000366, + "loss": 8.1471, + "step": 122 + }, + { + "epoch": 0.00123, + "grad_norm": 0.8121975389696077, + "learning_rate": 0.000369, + "loss": 8.1243, + "step": 123 + }, + { + "epoch": 0.00124, + "grad_norm": 0.8287555396618358, + "learning_rate": 0.000372, + "loss": 8.1037, + "step": 124 + }, + { + "epoch": 0.00125, + "grad_norm": 0.8387006564379252, + "learning_rate": 0.000375, + "loss": 8.0802, + "step": 125 + }, + { + "epoch": 0.00126, + "grad_norm": 0.7985855201767323, + "learning_rate": 0.000378, + "loss": 8.0672, + "step": 126 + }, + { + "epoch": 0.00127, + "grad_norm": 0.7830434817433392, + "learning_rate": 0.000381, + "loss": 8.0437, + "step": 127 + }, + { + "epoch": 0.00128, + "grad_norm": 0.8466276209824322, + "learning_rate": 0.000384, + "loss": 8.0264, + "step": 128 + }, + { + "epoch": 0.00129, + "grad_norm": 1.1259201196462498, + "learning_rate": 0.00038700000000000003, + "loss": 8.0199, + "step": 129 + }, + { + "epoch": 0.0013, + "grad_norm": 1.356061683657538, + "learning_rate": 0.00039000000000000005, + "loss": 8.0054, + "step": 130 + }, + { + "epoch": 0.00131, + "grad_norm": 0.781068809712636, + "learning_rate": 0.000393, + "loss": 7.9576, + "step": 131 + }, + { + "epoch": 0.00132, + "grad_norm": 1.0367188815921278, + "learning_rate": 0.00039600000000000003, + "loss": 7.9591, + "step": 132 + }, + { + "epoch": 0.00133, + "grad_norm": 0.7841067594214959, + "learning_rate": 0.00039900000000000005, + "loss": 7.9293, + "step": 133 + }, + { + "epoch": 0.00134, + "grad_norm": 0.8594113418007541, + "learning_rate": 0.000402, + "loss": 7.9217, + "step": 134 + }, + { + "epoch": 0.00135, + "grad_norm": 0.7468719464596628, + "learning_rate": 0.00040500000000000003, + "loss": 7.8883, + "step": 135 + }, + { + "epoch": 0.00136, + "grad_norm": 0.8415432392576198, + "learning_rate": 0.00040800000000000005, + "loss": 7.8771, + "step": 136 + }, + { + "epoch": 0.00137, + "grad_norm": 0.7386877288273068, + "learning_rate": 0.000411, + "loss": 7.8542, + "step": 137 + }, + { + "epoch": 0.00138, + "grad_norm": 0.7450176106214967, + "learning_rate": 0.00041400000000000003, + "loss": 7.8322, + "step": 138 + }, + { + "epoch": 0.00139, + "grad_norm": 0.7424698225823185, + "learning_rate": 0.00041700000000000005, + "loss": 7.8197, + "step": 139 + }, + { + "epoch": 0.0014, + "grad_norm": 0.7268306219948636, + "learning_rate": 0.00042000000000000007, + "loss": 7.8018, + "step": 140 + }, + { + "epoch": 0.00141, + "grad_norm": 0.7115037341291065, + "learning_rate": 0.000423, + "loss": 7.7905, + "step": 141 + }, + { + "epoch": 0.00142, + "grad_norm": 0.6725409058271569, + "learning_rate": 0.00042599999999999995, + "loss": 7.7772, + "step": 142 + }, + { + "epoch": 0.00143, + "grad_norm": 0.6830809074405504, + "learning_rate": 0.00042899999999999997, + "loss": 7.7496, + "step": 143 + }, + { + "epoch": 0.00144, + "grad_norm": 0.6791849969278475, + "learning_rate": 0.000432, + "loss": 7.7318, + "step": 144 + }, + { + "epoch": 0.00145, + "grad_norm": 0.6462817248800249, + "learning_rate": 0.000435, + "loss": 7.7251, + "step": 145 + }, + { + "epoch": 0.00146, + "grad_norm": 0.6695930969912641, + "learning_rate": 0.00043799999999999997, + "loss": 7.6893, + "step": 146 + }, + { + "epoch": 0.00147, + "grad_norm": 0.7012896651032599, + "learning_rate": 0.000441, + "loss": 7.6817, + "step": 147 + }, + { + "epoch": 0.00148, + "grad_norm": 0.7755580877429182, + "learning_rate": 0.000444, + "loss": 7.6698, + "step": 148 + }, + { + "epoch": 0.00149, + "grad_norm": 1.088947674236225, + "learning_rate": 0.00044699999999999997, + "loss": 7.6713, + "step": 149 + }, + { + "epoch": 0.0015, + "grad_norm": 0.9396541473912592, + "learning_rate": 0.00045, + "loss": 7.6436, + "step": 150 + }, + { + "epoch": 0.00151, + "grad_norm": 0.640217037835256, + "learning_rate": 0.000453, + "loss": 7.6133, + "step": 151 + }, + { + "epoch": 0.00152, + "grad_norm": 0.9098496631236208, + "learning_rate": 0.000456, + "loss": 7.6149, + "step": 152 + }, + { + "epoch": 0.00153, + "grad_norm": 0.6175759444520236, + "learning_rate": 0.000459, + "loss": 7.5962, + "step": 153 + }, + { + "epoch": 0.00154, + "grad_norm": 0.6884616968083866, + "learning_rate": 0.000462, + "loss": 7.5772, + "step": 154 + }, + { + "epoch": 0.00155, + "grad_norm": 0.6360938051960316, + "learning_rate": 0.000465, + "loss": 7.5598, + "step": 155 + }, + { + "epoch": 0.00156, + "grad_norm": 0.6937352973699618, + "learning_rate": 0.000468, + "loss": 7.5366, + "step": 156 + }, + { + "epoch": 0.00157, + "grad_norm": 0.6334556877551312, + "learning_rate": 0.000471, + "loss": 7.53, + "step": 157 + }, + { + "epoch": 0.00158, + "grad_norm": 0.9707170671679272, + "learning_rate": 0.00047400000000000003, + "loss": 7.5153, + "step": 158 + }, + { + "epoch": 0.00159, + "grad_norm": 1.0240750882994218, + "learning_rate": 0.000477, + "loss": 7.5076, + "step": 159 + }, + { + "epoch": 0.0016, + "grad_norm": 0.9322894026235434, + "learning_rate": 0.00048, + "loss": 7.4836, + "step": 160 + }, + { + "epoch": 0.00161, + "grad_norm": 0.5279037898518898, + "learning_rate": 0.00048300000000000003, + "loss": 7.4527, + "step": 161 + }, + { + "epoch": 0.00162, + "grad_norm": 0.687356662308957, + "learning_rate": 0.00048600000000000005, + "loss": 7.4601, + "step": 162 + }, + { + "epoch": 0.00163, + "grad_norm": 0.5623951705594973, + "learning_rate": 0.0004890000000000001, + "loss": 7.4388, + "step": 163 + }, + { + "epoch": 0.00164, + "grad_norm": 0.5581337114560441, + "learning_rate": 0.000492, + "loss": 7.4399, + "step": 164 + }, + { + "epoch": 0.00165, + "grad_norm": 0.5516159301488641, + "learning_rate": 0.000495, + "loss": 7.4126, + "step": 165 + }, + { + "epoch": 0.00166, + "grad_norm": 0.5242244578051735, + "learning_rate": 0.0004980000000000001, + "loss": 7.3876, + "step": 166 + }, + { + "epoch": 0.00167, + "grad_norm": 0.45997959649003123, + "learning_rate": 0.000501, + "loss": 7.3779, + "step": 167 + }, + { + "epoch": 0.00168, + "grad_norm": 0.5436289820614866, + "learning_rate": 0.000504, + "loss": 7.3569, + "step": 168 + }, + { + "epoch": 0.00169, + "grad_norm": 0.4983067598465849, + "learning_rate": 0.0005070000000000001, + "loss": 7.3495, + "step": 169 + }, + { + "epoch": 0.0017, + "grad_norm": 0.4402852485923817, + "learning_rate": 0.00051, + "loss": 7.3316, + "step": 170 + }, + { + "epoch": 0.00171, + "grad_norm": 0.5221521396945228, + "learning_rate": 0.000513, + "loss": 7.3138, + "step": 171 + }, + { + "epoch": 0.00172, + "grad_norm": 0.45544347662440743, + "learning_rate": 0.000516, + "loss": 7.3129, + "step": 172 + }, + { + "epoch": 0.00173, + "grad_norm": 0.4745602833877857, + "learning_rate": 0.0005189999999999999, + "loss": 7.2961, + "step": 173 + }, + { + "epoch": 0.00174, + "grad_norm": 0.5121580158942174, + "learning_rate": 0.000522, + "loss": 7.311, + "step": 174 + }, + { + "epoch": 0.00175, + "grad_norm": 0.680505499537256, + "learning_rate": 0.000525, + "loss": 7.2769, + "step": 175 + }, + { + "epoch": 0.00176, + "grad_norm": 0.7210959926983863, + "learning_rate": 0.0005279999999999999, + "loss": 7.2549, + "step": 176 + }, + { + "epoch": 0.00177, + "grad_norm": 0.7510526045152774, + "learning_rate": 0.000531, + "loss": 7.2572, + "step": 177 + }, + { + "epoch": 0.00178, + "grad_norm": 0.7343359791017195, + "learning_rate": 0.000534, + "loss": 7.2506, + "step": 178 + }, + { + "epoch": 0.00179, + "grad_norm": 0.6185909230661502, + "learning_rate": 0.000537, + "loss": 7.2371, + "step": 179 + }, + { + "epoch": 0.0018, + "grad_norm": 0.42887581745789505, + "learning_rate": 0.00054, + "loss": 7.2041, + "step": 180 + }, + { + "epoch": 0.00181, + "grad_norm": 0.5359194273155663, + "learning_rate": 0.000543, + "loss": 7.1961, + "step": 181 + }, + { + "epoch": 0.00182, + "grad_norm": 0.4369454993609359, + "learning_rate": 0.000546, + "loss": 7.188, + "step": 182 + }, + { + "epoch": 0.00183, + "grad_norm": 0.41001210068633426, + "learning_rate": 0.000549, + "loss": 7.1769, + "step": 183 + }, + { + "epoch": 0.00184, + "grad_norm": 0.4396906853721559, + "learning_rate": 0.000552, + "loss": 7.1646, + "step": 184 + }, + { + "epoch": 0.00185, + "grad_norm": 0.4311329351383538, + "learning_rate": 0.000555, + "loss": 7.1403, + "step": 185 + }, + { + "epoch": 0.00186, + "grad_norm": 0.44013221474943204, + "learning_rate": 0.000558, + "loss": 7.1495, + "step": 186 + }, + { + "epoch": 0.00187, + "grad_norm": 0.42719641993731927, + "learning_rate": 0.000561, + "loss": 7.1244, + "step": 187 + }, + { + "epoch": 0.00188, + "grad_norm": 0.39192004433941763, + "learning_rate": 0.000564, + "loss": 7.1022, + "step": 188 + }, + { + "epoch": 0.00189, + "grad_norm": 0.5474017259570552, + "learning_rate": 0.000567, + "loss": 7.1002, + "step": 189 + }, + { + "epoch": 0.0019, + "grad_norm": 0.5320208949420774, + "learning_rate": 0.00057, + "loss": 7.0824, + "step": 190 + }, + { + "epoch": 0.00191, + "grad_norm": 0.7661733453917681, + "learning_rate": 0.000573, + "loss": 7.0764, + "step": 191 + }, + { + "epoch": 0.00192, + "grad_norm": 0.9858385620454592, + "learning_rate": 0.000576, + "loss": 7.0775, + "step": 192 + }, + { + "epoch": 0.00193, + "grad_norm": 0.9591489873604585, + "learning_rate": 0.000579, + "loss": 7.0725, + "step": 193 + }, + { + "epoch": 0.00194, + "grad_norm": 0.8774582305545432, + "learning_rate": 0.000582, + "loss": 7.0715, + "step": 194 + }, + { + "epoch": 0.00195, + "grad_norm": 0.9015156284206135, + "learning_rate": 0.000585, + "loss": 7.0487, + "step": 195 + }, + { + "epoch": 0.00196, + "grad_norm": 0.8000421121121074, + "learning_rate": 0.000588, + "loss": 7.0339, + "step": 196 + }, + { + "epoch": 0.00197, + "grad_norm": 0.8468645119231468, + "learning_rate": 0.000591, + "loss": 7.0329, + "step": 197 + }, + { + "epoch": 0.00198, + "grad_norm": 0.6924161058762034, + "learning_rate": 0.000594, + "loss": 7.0197, + "step": 198 + }, + { + "epoch": 0.00199, + "grad_norm": 0.5671884633245193, + "learning_rate": 0.0005970000000000001, + "loss": 6.9992, + "step": 199 + }, + { + "epoch": 0.002, + "grad_norm": 0.5557515599131739, + "learning_rate": 0.0006000000000000001, + "loss": 6.9982, + "step": 200 + }, + { + "epoch": 0.00201, + "grad_norm": 0.4619645912916116, + "learning_rate": 0.000603, + "loss": 6.9694, + "step": 201 + }, + { + "epoch": 0.00202, + "grad_norm": 0.4929222182059183, + "learning_rate": 0.0006060000000000001, + "loss": 6.9601, + "step": 202 + }, + { + "epoch": 0.00203, + "grad_norm": 0.37702087689736097, + "learning_rate": 0.0006090000000000001, + "loss": 6.9599, + "step": 203 + }, + { + "epoch": 0.00204, + "grad_norm": 0.45518892473332057, + "learning_rate": 0.000612, + "loss": 6.9476, + "step": 204 + }, + { + "epoch": 0.00205, + "grad_norm": 0.384809834745988, + "learning_rate": 0.000615, + "loss": 6.9322, + "step": 205 + }, + { + "epoch": 0.00206, + "grad_norm": 0.38976835520829006, + "learning_rate": 0.000618, + "loss": 6.9255, + "step": 206 + }, + { + "epoch": 0.00207, + "grad_norm": 0.4375749283114229, + "learning_rate": 0.000621, + "loss": 6.9166, + "step": 207 + }, + { + "epoch": 0.00208, + "grad_norm": 0.6317927736115376, + "learning_rate": 0.000624, + "loss": 6.9157, + "step": 208 + }, + { + "epoch": 0.00209, + "grad_norm": 0.7772617938427908, + "learning_rate": 0.000627, + "loss": 6.9007, + "step": 209 + }, + { + "epoch": 0.0021, + "grad_norm": 1.2298132092555871, + "learning_rate": 0.00063, + "loss": 6.9015, + "step": 210 + }, + { + "epoch": 0.00211, + "grad_norm": 0.7435502579532725, + "learning_rate": 0.000633, + "loss": 6.8883, + "step": 211 + }, + { + "epoch": 0.00212, + "grad_norm": 0.3794792840193541, + "learning_rate": 0.000636, + "loss": 6.8693, + "step": 212 + }, + { + "epoch": 0.00213, + "grad_norm": 0.688114698891817, + "learning_rate": 0.000639, + "loss": 6.8589, + "step": 213 + }, + { + "epoch": 0.00214, + "grad_norm": 0.49868168831557785, + "learning_rate": 0.000642, + "loss": 6.8504, + "step": 214 + }, + { + "epoch": 0.00215, + "grad_norm": 0.6292678113208914, + "learning_rate": 0.000645, + "loss": 6.8506, + "step": 215 + }, + { + "epoch": 0.00216, + "grad_norm": 0.3673676300147008, + "learning_rate": 0.000648, + "loss": 6.8383, + "step": 216 + }, + { + "epoch": 0.00217, + "grad_norm": 0.4511278951279821, + "learning_rate": 0.000651, + "loss": 6.8131, + "step": 217 + }, + { + "epoch": 0.00218, + "grad_norm": 0.3783399316849399, + "learning_rate": 0.000654, + "loss": 6.8146, + "step": 218 + }, + { + "epoch": 0.00219, + "grad_norm": 0.3651267362185199, + "learning_rate": 0.000657, + "loss": 6.7867, + "step": 219 + }, + { + "epoch": 0.0022, + "grad_norm": 0.4708567379601776, + "learning_rate": 0.00066, + "loss": 6.8074, + "step": 220 + }, + { + "epoch": 0.00221, + "grad_norm": 0.5218809374007617, + "learning_rate": 0.0006630000000000001, + "loss": 6.7837, + "step": 221 + }, + { + "epoch": 0.00222, + "grad_norm": 0.7518789100021657, + "learning_rate": 0.000666, + "loss": 6.7774, + "step": 222 + }, + { + "epoch": 0.00223, + "grad_norm": 0.9964198530393009, + "learning_rate": 0.000669, + "loss": 6.7824, + "step": 223 + }, + { + "epoch": 0.00224, + "grad_norm": 0.807522133159797, + "learning_rate": 0.0006720000000000001, + "loss": 6.7748, + "step": 224 + }, + { + "epoch": 0.00225, + "grad_norm": 0.6803785939854445, + "learning_rate": 0.000675, + "loss": 6.7563, + "step": 225 + }, + { + "epoch": 0.00226, + "grad_norm": 1.0762485305388094, + "learning_rate": 0.000678, + "loss": 6.7681, + "step": 226 + }, + { + "epoch": 0.00227, + "grad_norm": 0.9129777590140102, + "learning_rate": 0.0006810000000000001, + "loss": 6.734, + "step": 227 + }, + { + "epoch": 0.00228, + "grad_norm": 1.299087365220233, + "learning_rate": 0.000684, + "loss": 6.7517, + "step": 228 + }, + { + "epoch": 0.00229, + "grad_norm": 0.7392729796935101, + "learning_rate": 0.000687, + "loss": 6.7331, + "step": 229 + }, + { + "epoch": 0.0023, + "grad_norm": 0.6158971034672628, + "learning_rate": 0.0006900000000000001, + "loss": 6.7089, + "step": 230 + }, + { + "epoch": 0.00231, + "grad_norm": 0.7016695647497186, + "learning_rate": 0.000693, + "loss": 6.6956, + "step": 231 + }, + { + "epoch": 0.00232, + "grad_norm": 0.6434185638703606, + "learning_rate": 0.000696, + "loss": 6.7082, + "step": 232 + }, + { + "epoch": 0.00233, + "grad_norm": 0.6044879418578446, + "learning_rate": 0.0006990000000000001, + "loss": 6.6921, + "step": 233 + }, + { + "epoch": 0.00234, + "grad_norm": 0.4480061456613071, + "learning_rate": 0.000702, + "loss": 6.679, + "step": 234 + }, + { + "epoch": 0.00235, + "grad_norm": 0.4492075259819563, + "learning_rate": 0.000705, + "loss": 6.6661, + "step": 235 + }, + { + "epoch": 0.00236, + "grad_norm": 0.3889271878969786, + "learning_rate": 0.000708, + "loss": 6.6458, + "step": 236 + }, + { + "epoch": 0.00237, + "grad_norm": 0.41607485698419117, + "learning_rate": 0.0007109999999999999, + "loss": 6.6575, + "step": 237 + }, + { + "epoch": 0.00238, + "grad_norm": 0.3515958046168432, + "learning_rate": 0.000714, + "loss": 6.6432, + "step": 238 + }, + { + "epoch": 0.00239, + "grad_norm": 0.4049983983930021, + "learning_rate": 0.000717, + "loss": 6.6274, + "step": 239 + }, + { + "epoch": 0.0024, + "grad_norm": 0.3247748587680522, + "learning_rate": 0.0007199999999999999, + "loss": 6.6414, + "step": 240 + }, + { + "epoch": 0.00241, + "grad_norm": 0.36262136655648425, + "learning_rate": 0.000723, + "loss": 6.6267, + "step": 241 + }, + { + "epoch": 0.00242, + "grad_norm": 0.3238608381517686, + "learning_rate": 0.000726, + "loss": 6.598, + "step": 242 + }, + { + "epoch": 0.00243, + "grad_norm": 0.31681258130419926, + "learning_rate": 0.000729, + "loss": 6.5877, + "step": 243 + }, + { + "epoch": 0.00244, + "grad_norm": 0.3616370322868285, + "learning_rate": 0.000732, + "loss": 6.5911, + "step": 244 + }, + { + "epoch": 0.00245, + "grad_norm": 0.4634076592576489, + "learning_rate": 0.000735, + "loss": 6.5815, + "step": 245 + }, + { + "epoch": 0.00246, + "grad_norm": 0.7486687333799987, + "learning_rate": 0.000738, + "loss": 6.5693, + "step": 246 + }, + { + "epoch": 0.00247, + "grad_norm": 1.662590620725261, + "learning_rate": 0.000741, + "loss": 6.5942, + "step": 247 + }, + { + "epoch": 0.00248, + "grad_norm": 0.7962452736060305, + "learning_rate": 0.000744, + "loss": 6.5799, + "step": 248 + }, + { + "epoch": 0.00249, + "grad_norm": 1.091558045553116, + "learning_rate": 0.000747, + "loss": 6.5681, + "step": 249 + }, + { + "epoch": 0.0025, + "grad_norm": 1.1566118279124307, + "learning_rate": 0.00075, + "loss": 6.5821, + "step": 250 + }, + { + "epoch": 0.00251, + "grad_norm": 0.7784764288643516, + "learning_rate": 0.000753, + "loss": 6.5311, + "step": 251 + }, + { + "epoch": 0.00252, + "grad_norm": 0.8327004278614876, + "learning_rate": 0.000756, + "loss": 6.5476, + "step": 252 + }, + { + "epoch": 0.00253, + "grad_norm": 0.8184368732684698, + "learning_rate": 0.000759, + "loss": 6.5327, + "step": 253 + }, + { + "epoch": 0.00254, + "grad_norm": 1.1486609061260633, + "learning_rate": 0.000762, + "loss": 6.541, + "step": 254 + }, + { + "epoch": 0.00255, + "grad_norm": 0.8085196503687528, + "learning_rate": 0.0007650000000000001, + "loss": 6.533, + "step": 255 + }, + { + "epoch": 0.00256, + "grad_norm": 0.6227788989369496, + "learning_rate": 0.000768, + "loss": 6.5039, + "step": 256 + }, + { + "epoch": 0.00257, + "grad_norm": 0.7880922816831363, + "learning_rate": 0.000771, + "loss": 6.5163, + "step": 257 + }, + { + "epoch": 0.00258, + "grad_norm": 0.8888953354311911, + "learning_rate": 0.0007740000000000001, + "loss": 6.497, + "step": 258 + }, + { + "epoch": 0.00259, + "grad_norm": 0.6556759438280303, + "learning_rate": 0.000777, + "loss": 6.491, + "step": 259 + }, + { + "epoch": 0.0026, + "grad_norm": 0.5556045743581063, + "learning_rate": 0.0007800000000000001, + "loss": 6.4736, + "step": 260 + }, + { + "epoch": 0.00261, + "grad_norm": 0.5757853625150707, + "learning_rate": 0.0007830000000000001, + "loss": 6.4835, + "step": 261 + }, + { + "epoch": 0.00262, + "grad_norm": 0.4702991355160636, + "learning_rate": 0.000786, + "loss": 6.4605, + "step": 262 + }, + { + "epoch": 0.00263, + "grad_norm": 0.5236270398277312, + "learning_rate": 0.0007890000000000001, + "loss": 6.4668, + "step": 263 + }, + { + "epoch": 0.00264, + "grad_norm": 0.57918798311576, + "learning_rate": 0.0007920000000000001, + "loss": 6.4494, + "step": 264 + }, + { + "epoch": 0.00265, + "grad_norm": 0.48386144230733374, + "learning_rate": 0.000795, + "loss": 6.4394, + "step": 265 + }, + { + "epoch": 0.00266, + "grad_norm": 0.44170660781498655, + "learning_rate": 0.0007980000000000001, + "loss": 6.4372, + "step": 266 + }, + { + "epoch": 0.00267, + "grad_norm": 0.35002993788441544, + "learning_rate": 0.0008010000000000001, + "loss": 6.4273, + "step": 267 + }, + { + "epoch": 0.00268, + "grad_norm": 0.49450621569184094, + "learning_rate": 0.000804, + "loss": 6.426, + "step": 268 + }, + { + "epoch": 0.00269, + "grad_norm": 0.5917094821390496, + "learning_rate": 0.0008070000000000001, + "loss": 6.4104, + "step": 269 + }, + { + "epoch": 0.0027, + "grad_norm": 0.8610573273382283, + "learning_rate": 0.0008100000000000001, + "loss": 6.4165, + "step": 270 + }, + { + "epoch": 0.00271, + "grad_norm": 0.9687752227819599, + "learning_rate": 0.000813, + "loss": 6.4342, + "step": 271 + }, + { + "epoch": 0.00272, + "grad_norm": 0.8252297813091672, + "learning_rate": 0.0008160000000000001, + "loss": 6.4109, + "step": 272 + }, + { + "epoch": 0.00273, + "grad_norm": 1.0509340869742279, + "learning_rate": 0.0008190000000000001, + "loss": 6.4091, + "step": 273 + }, + { + "epoch": 0.00274, + "grad_norm": 1.2492890299831212, + "learning_rate": 0.000822, + "loss": 6.3984, + "step": 274 + }, + { + "epoch": 0.00275, + "grad_norm": 0.7449833128353471, + "learning_rate": 0.0008250000000000001, + "loss": 6.3937, + "step": 275 + }, + { + "epoch": 0.00276, + "grad_norm": 0.650217944416595, + "learning_rate": 0.0008280000000000001, + "loss": 6.383, + "step": 276 + }, + { + "epoch": 0.00277, + "grad_norm": 0.6897831829588578, + "learning_rate": 0.0008310000000000001, + "loss": 6.3791, + "step": 277 + }, + { + "epoch": 0.00278, + "grad_norm": 0.6845751052836879, + "learning_rate": 0.0008340000000000001, + "loss": 6.3625, + "step": 278 + }, + { + "epoch": 0.00279, + "grad_norm": 0.6895680336476154, + "learning_rate": 0.0008370000000000001, + "loss": 6.3607, + "step": 279 + }, + { + "epoch": 0.0028, + "grad_norm": 0.7962805776833526, + "learning_rate": 0.0008400000000000001, + "loss": 6.3497, + "step": 280 + }, + { + "epoch": 0.00281, + "grad_norm": 0.7281805890288613, + "learning_rate": 0.0008430000000000001, + "loss": 6.3358, + "step": 281 + }, + { + "epoch": 0.00282, + "grad_norm": 0.6077351466654652, + "learning_rate": 0.000846, + "loss": 6.3343, + "step": 282 + }, + { + "epoch": 0.00283, + "grad_norm": 0.7021452115418912, + "learning_rate": 0.0008489999999999999, + "loss": 6.3268, + "step": 283 + }, + { + "epoch": 0.00284, + "grad_norm": 0.8067589645945545, + "learning_rate": 0.0008519999999999999, + "loss": 6.3217, + "step": 284 + }, + { + "epoch": 0.00285, + "grad_norm": 0.7238661685202905, + "learning_rate": 0.000855, + "loss": 6.3229, + "step": 285 + }, + { + "epoch": 0.00286, + "grad_norm": 0.6566110557704579, + "learning_rate": 0.0008579999999999999, + "loss": 6.3002, + "step": 286 + }, + { + "epoch": 0.00287, + "grad_norm": 0.5448339024770075, + "learning_rate": 0.000861, + "loss": 6.3016, + "step": 287 + }, + { + "epoch": 0.00288, + "grad_norm": 0.4516947530472597, + "learning_rate": 0.000864, + "loss": 6.279, + "step": 288 + }, + { + "epoch": 0.00289, + "grad_norm": 0.5415480649305122, + "learning_rate": 0.0008669999999999999, + "loss": 6.2929, + "step": 289 + }, + { + "epoch": 0.0029, + "grad_norm": 0.5812113854990302, + "learning_rate": 0.00087, + "loss": 6.2743, + "step": 290 + }, + { + "epoch": 0.00291, + "grad_norm": 0.5620937871349196, + "learning_rate": 0.000873, + "loss": 6.2679, + "step": 291 + }, + { + "epoch": 0.00292, + "grad_norm": 0.43362286284353735, + "learning_rate": 0.0008759999999999999, + "loss": 6.2618, + "step": 292 + }, + { + "epoch": 0.00293, + "grad_norm": 0.563953479113573, + "learning_rate": 0.000879, + "loss": 6.2491, + "step": 293 + }, + { + "epoch": 0.00294, + "grad_norm": 0.5784107789505917, + "learning_rate": 0.000882, + "loss": 6.2636, + "step": 294 + }, + { + "epoch": 0.00295, + "grad_norm": 0.5410292555611181, + "learning_rate": 0.0008849999999999999, + "loss": 6.2293, + "step": 295 + }, + { + "epoch": 0.00296, + "grad_norm": 0.5413778994680785, + "learning_rate": 0.000888, + "loss": 6.2428, + "step": 296 + }, + { + "epoch": 0.00297, + "grad_norm": 0.7587824592105622, + "learning_rate": 0.000891, + "loss": 6.2186, + "step": 297 + }, + { + "epoch": 0.00298, + "grad_norm": 1.2390223277612324, + "learning_rate": 0.0008939999999999999, + "loss": 6.2454, + "step": 298 + }, + { + "epoch": 0.00299, + "grad_norm": 1.0051162436627608, + "learning_rate": 0.000897, + "loss": 6.2195, + "step": 299 + }, + { + "epoch": 0.003, + "grad_norm": 0.9301924220055711, + "learning_rate": 0.0009, + "loss": 6.2223, + "step": 300 + }, + { + "epoch": 0.00301, + "grad_norm": 0.9092669241988458, + "learning_rate": 0.0009029999999999999, + "loss": 6.2133, + "step": 301 + }, + { + "epoch": 0.00302, + "grad_norm": 1.0280476917286576, + "learning_rate": 0.000906, + "loss": 6.2045, + "step": 302 + }, + { + "epoch": 0.00303, + "grad_norm": 0.955686909103224, + "learning_rate": 0.000909, + "loss": 6.2159, + "step": 303 + }, + { + "epoch": 0.00304, + "grad_norm": 0.924153033838841, + "learning_rate": 0.000912, + "loss": 6.2068, + "step": 304 + }, + { + "epoch": 0.00305, + "grad_norm": 1.0129569357421315, + "learning_rate": 0.000915, + "loss": 6.2162, + "step": 305 + }, + { + "epoch": 0.00306, + "grad_norm": 0.908196615580651, + "learning_rate": 0.000918, + "loss": 6.1982, + "step": 306 + }, + { + "epoch": 0.00307, + "grad_norm": 0.9412115799039344, + "learning_rate": 0.000921, + "loss": 6.1922, + "step": 307 + }, + { + "epoch": 0.00308, + "grad_norm": 1.0133210251014897, + "learning_rate": 0.000924, + "loss": 6.1801, + "step": 308 + }, + { + "epoch": 0.00309, + "grad_norm": 0.7885110954474147, + "learning_rate": 0.000927, + "loss": 6.1916, + "step": 309 + }, + { + "epoch": 0.0031, + "grad_norm": 0.8502930513548597, + "learning_rate": 0.00093, + "loss": 6.1767, + "step": 310 + }, + { + "epoch": 0.00311, + "grad_norm": 0.8830498260903274, + "learning_rate": 0.000933, + "loss": 6.1706, + "step": 311 + }, + { + "epoch": 0.00312, + "grad_norm": 0.7731682808518218, + "learning_rate": 0.000936, + "loss": 6.1483, + "step": 312 + }, + { + "epoch": 0.00313, + "grad_norm": 0.5034693382195244, + "learning_rate": 0.0009390000000000001, + "loss": 6.1541, + "step": 313 + }, + { + "epoch": 0.00314, + "grad_norm": 0.5944206771214186, + "learning_rate": 0.000942, + "loss": 6.1615, + "step": 314 + }, + { + "epoch": 0.00315, + "grad_norm": 0.6198949865082991, + "learning_rate": 0.000945, + "loss": 6.1382, + "step": 315 + }, + { + "epoch": 0.00316, + "grad_norm": 0.6322301787116348, + "learning_rate": 0.0009480000000000001, + "loss": 6.112, + "step": 316 + }, + { + "epoch": 0.00317, + "grad_norm": 0.6034799354522895, + "learning_rate": 0.000951, + "loss": 6.1294, + "step": 317 + }, + { + "epoch": 0.00318, + "grad_norm": 0.594812791501401, + "learning_rate": 0.000954, + "loss": 6.1243, + "step": 318 + }, + { + "epoch": 0.00319, + "grad_norm": 0.5121880287121718, + "learning_rate": 0.0009570000000000001, + "loss": 6.1104, + "step": 319 + }, + { + "epoch": 0.0032, + "grad_norm": 0.4383869814323231, + "learning_rate": 0.00096, + "loss": 6.1078, + "step": 320 + }, + { + "epoch": 0.00321, + "grad_norm": 0.46253996577569373, + "learning_rate": 0.000963, + "loss": 6.0957, + "step": 321 + }, + { + "epoch": 0.00322, + "grad_norm": 0.5330974668531825, + "learning_rate": 0.0009660000000000001, + "loss": 6.0852, + "step": 322 + }, + { + "epoch": 0.00323, + "grad_norm": 0.4269664457046623, + "learning_rate": 0.000969, + "loss": 6.0805, + "step": 323 + }, + { + "epoch": 0.00324, + "grad_norm": 0.4120363869720864, + "learning_rate": 0.0009720000000000001, + "loss": 6.0833, + "step": 324 + }, + { + "epoch": 0.00325, + "grad_norm": 0.43852524771556284, + "learning_rate": 0.0009750000000000001, + "loss": 6.0494, + "step": 325 + }, + { + "epoch": 0.00326, + "grad_norm": 0.5390182954459198, + "learning_rate": 0.0009780000000000001, + "loss": 6.0675, + "step": 326 + }, + { + "epoch": 0.00327, + "grad_norm": 0.7655676469366879, + "learning_rate": 0.000981, + "loss": 6.0729, + "step": 327 + }, + { + "epoch": 0.00328, + "grad_norm": 1.3095809996152918, + "learning_rate": 0.000984, + "loss": 6.0661, + "step": 328 + }, + { + "epoch": 0.00329, + "grad_norm": 0.9398963728555475, + "learning_rate": 0.000987, + "loss": 6.0593, + "step": 329 + }, + { + "epoch": 0.0033, + "grad_norm": 0.9687365266328093, + "learning_rate": 0.00099, + "loss": 6.0595, + "step": 330 + }, + { + "epoch": 0.00331, + "grad_norm": 1.0572725295338647, + "learning_rate": 0.0009930000000000002, + "loss": 6.0709, + "step": 331 + }, + { + "epoch": 0.00332, + "grad_norm": 0.7804253243852415, + "learning_rate": 0.0009960000000000001, + "loss": 6.0449, + "step": 332 + }, + { + "epoch": 0.00333, + "grad_norm": 0.9477444786898163, + "learning_rate": 0.000999, + "loss": 6.0352, + "step": 333 + }, + { + "epoch": 0.00334, + "grad_norm": 1.119636024931192, + "learning_rate": 0.001002, + "loss": 6.0509, + "step": 334 + }, + { + "epoch": 0.00335, + "grad_norm": 1.0398096390896505, + "learning_rate": 0.001005, + "loss": 6.0495, + "step": 335 + }, + { + "epoch": 0.00336, + "grad_norm": 0.8946811171098662, + "learning_rate": 0.001008, + "loss": 6.0348, + "step": 336 + }, + { + "epoch": 0.00337, + "grad_norm": 1.0240126692017184, + "learning_rate": 0.0010110000000000002, + "loss": 6.0215, + "step": 337 + }, + { + "epoch": 0.00338, + "grad_norm": 1.0245564805630587, + "learning_rate": 0.0010140000000000001, + "loss": 6.0414, + "step": 338 + }, + { + "epoch": 0.00339, + "grad_norm": 0.9606863659347135, + "learning_rate": 0.0010170000000000001, + "loss": 6.0439, + "step": 339 + }, + { + "epoch": 0.0034, + "grad_norm": 0.7549695201896788, + "learning_rate": 0.00102, + "loss": 6.0067, + "step": 340 + }, + { + "epoch": 0.00341, + "grad_norm": 0.7981881225838676, + "learning_rate": 0.001023, + "loss": 5.9935, + "step": 341 + }, + { + "epoch": 0.00342, + "grad_norm": 0.6476806075689499, + "learning_rate": 0.001026, + "loss": 5.9919, + "step": 342 + }, + { + "epoch": 0.00343, + "grad_norm": 0.6180530813345404, + "learning_rate": 0.0010290000000000002, + "loss": 5.9943, + "step": 343 + }, + { + "epoch": 0.00344, + "grad_norm": 0.5904235085071222, + "learning_rate": 0.001032, + "loss": 5.9824, + "step": 344 + }, + { + "epoch": 0.00345, + "grad_norm": 0.5662246103658737, + "learning_rate": 0.001035, + "loss": 5.9736, + "step": 345 + }, + { + "epoch": 0.00346, + "grad_norm": 0.5815647759401746, + "learning_rate": 0.0010379999999999999, + "loss": 5.9787, + "step": 346 + }, + { + "epoch": 0.00347, + "grad_norm": 0.5922585456492798, + "learning_rate": 0.001041, + "loss": 5.9505, + "step": 347 + }, + { + "epoch": 0.00348, + "grad_norm": 0.546770754957902, + "learning_rate": 0.001044, + "loss": 5.9566, + "step": 348 + }, + { + "epoch": 0.00349, + "grad_norm": 0.5259270224752429, + "learning_rate": 0.001047, + "loss": 5.9388, + "step": 349 + }, + { + "epoch": 0.0035, + "grad_norm": 0.5639453872626701, + "learning_rate": 0.00105, + "loss": 5.96, + "step": 350 + }, + { + "epoch": 0.00351, + "grad_norm": 0.562716119705, + "learning_rate": 0.001053, + "loss": 5.9575, + "step": 351 + }, + { + "epoch": 0.00352, + "grad_norm": 0.6276157353012132, + "learning_rate": 0.0010559999999999999, + "loss": 5.9357, + "step": 352 + }, + { + "epoch": 0.00353, + "grad_norm": 0.6899105072430818, + "learning_rate": 0.001059, + "loss": 5.939, + "step": 353 + }, + { + "epoch": 0.00354, + "grad_norm": 0.7414855002893123, + "learning_rate": 0.001062, + "loss": 5.9142, + "step": 354 + }, + { + "epoch": 0.00355, + "grad_norm": 0.8275245559939801, + "learning_rate": 0.001065, + "loss": 5.932, + "step": 355 + }, + { + "epoch": 0.00356, + "grad_norm": 0.8334699202442165, + "learning_rate": 0.001068, + "loss": 5.925, + "step": 356 + }, + { + "epoch": 0.00357, + "grad_norm": 0.733463396370079, + "learning_rate": 0.001071, + "loss": 5.9147, + "step": 357 + }, + { + "epoch": 0.00358, + "grad_norm": 0.5898600387972718, + "learning_rate": 0.001074, + "loss": 5.9022, + "step": 358 + }, + { + "epoch": 0.00359, + "grad_norm": 0.53807120274405, + "learning_rate": 0.001077, + "loss": 5.8933, + "step": 359 + }, + { + "epoch": 0.0036, + "grad_norm": 0.5868526923580266, + "learning_rate": 0.00108, + "loss": 5.8984, + "step": 360 + }, + { + "epoch": 0.00361, + "grad_norm": 0.6561371025439784, + "learning_rate": 0.001083, + "loss": 5.8927, + "step": 361 + }, + { + "epoch": 0.00362, + "grad_norm": 1.003392418604553, + "learning_rate": 0.001086, + "loss": 5.8918, + "step": 362 + }, + { + "epoch": 0.00363, + "grad_norm": 1.2318504963643775, + "learning_rate": 0.001089, + "loss": 5.899, + "step": 363 + }, + { + "epoch": 0.00364, + "grad_norm": 0.7073595215993886, + "learning_rate": 0.001092, + "loss": 5.8956, + "step": 364 + }, + { + "epoch": 0.00365, + "grad_norm": 0.6834408409212124, + "learning_rate": 0.001095, + "loss": 5.8924, + "step": 365 + }, + { + "epoch": 0.00366, + "grad_norm": 0.808409762735137, + "learning_rate": 0.001098, + "loss": 5.8732, + "step": 366 + }, + { + "epoch": 0.00367, + "grad_norm": 1.150363353399194, + "learning_rate": 0.001101, + "loss": 5.8871, + "step": 367 + }, + { + "epoch": 0.00368, + "grad_norm": 1.5759048240116487, + "learning_rate": 0.001104, + "loss": 5.9113, + "step": 368 + }, + { + "epoch": 0.00369, + "grad_norm": 0.6706600035773841, + "learning_rate": 0.001107, + "loss": 5.8826, + "step": 369 + }, + { + "epoch": 0.0037, + "grad_norm": 1.0165372867603926, + "learning_rate": 0.00111, + "loss": 5.8773, + "step": 370 + }, + { + "epoch": 0.00371, + "grad_norm": 1.199271508025637, + "learning_rate": 0.001113, + "loss": 5.8595, + "step": 371 + }, + { + "epoch": 0.00372, + "grad_norm": 0.8890339964058547, + "learning_rate": 0.001116, + "loss": 5.8693, + "step": 372 + }, + { + "epoch": 0.00373, + "grad_norm": 1.0853395614111072, + "learning_rate": 0.001119, + "loss": 5.8674, + "step": 373 + }, + { + "epoch": 0.00374, + "grad_norm": 0.9530307319001724, + "learning_rate": 0.001122, + "loss": 5.8703, + "step": 374 + }, + { + "epoch": 0.00375, + "grad_norm": 0.934750801302772, + "learning_rate": 0.0011250000000000001, + "loss": 5.8719, + "step": 375 + }, + { + "epoch": 0.00376, + "grad_norm": 0.7986123477309184, + "learning_rate": 0.001128, + "loss": 5.863, + "step": 376 + }, + { + "epoch": 0.00377, + "grad_norm": 1.034328210427756, + "learning_rate": 0.001131, + "loss": 5.8644, + "step": 377 + }, + { + "epoch": 0.00378, + "grad_norm": 1.1394107219722458, + "learning_rate": 0.001134, + "loss": 5.8702, + "step": 378 + }, + { + "epoch": 0.00379, + "grad_norm": 0.6778105205682995, + "learning_rate": 0.001137, + "loss": 5.8386, + "step": 379 + }, + { + "epoch": 0.0038, + "grad_norm": 0.7735792074169382, + "learning_rate": 0.00114, + "loss": 5.8619, + "step": 380 + }, + { + "epoch": 0.00381, + "grad_norm": 0.882590402307869, + "learning_rate": 0.0011430000000000001, + "loss": 5.83, + "step": 381 + }, + { + "epoch": 0.00382, + "grad_norm": 0.8860425089624719, + "learning_rate": 0.001146, + "loss": 5.8197, + "step": 382 + }, + { + "epoch": 0.00383, + "grad_norm": 1.0396684860073488, + "learning_rate": 0.001149, + "loss": 5.836, + "step": 383 + }, + { + "epoch": 0.00384, + "grad_norm": 0.9683520339794217, + "learning_rate": 0.001152, + "loss": 5.8386, + "step": 384 + }, + { + "epoch": 0.00385, + "grad_norm": 0.8210443052214895, + "learning_rate": 0.001155, + "loss": 5.8215, + "step": 385 + }, + { + "epoch": 0.00386, + "grad_norm": 1.030920073327714, + "learning_rate": 0.001158, + "loss": 5.8336, + "step": 386 + }, + { + "epoch": 0.00387, + "grad_norm": 0.8032114385497527, + "learning_rate": 0.0011610000000000001, + "loss": 5.8025, + "step": 387 + }, + { + "epoch": 0.00388, + "grad_norm": 0.6803620347459473, + "learning_rate": 0.001164, + "loss": 5.7978, + "step": 388 + }, + { + "epoch": 0.00389, + "grad_norm": 0.6752304208768743, + "learning_rate": 0.001167, + "loss": 5.8027, + "step": 389 + }, + { + "epoch": 0.0039, + "grad_norm": 0.6054825081153106, + "learning_rate": 0.00117, + "loss": 5.7828, + "step": 390 + }, + { + "epoch": 0.00391, + "grad_norm": 0.5156470046541872, + "learning_rate": 0.001173, + "loss": 5.7863, + "step": 391 + }, + { + "epoch": 0.00392, + "grad_norm": 0.49609211852516366, + "learning_rate": 0.001176, + "loss": 5.7945, + "step": 392 + }, + { + "epoch": 0.00393, + "grad_norm": 0.42817727304572534, + "learning_rate": 0.0011790000000000001, + "loss": 5.7664, + "step": 393 + }, + { + "epoch": 0.00394, + "grad_norm": 0.43894767278563757, + "learning_rate": 0.001182, + "loss": 5.7539, + "step": 394 + }, + { + "epoch": 0.00395, + "grad_norm": 0.41850660912289844, + "learning_rate": 0.001185, + "loss": 5.7366, + "step": 395 + }, + { + "epoch": 0.00396, + "grad_norm": 0.41745423249833347, + "learning_rate": 0.001188, + "loss": 5.7516, + "step": 396 + }, + { + "epoch": 0.00397, + "grad_norm": 0.40474070631964676, + "learning_rate": 0.001191, + "loss": 5.7433, + "step": 397 + }, + { + "epoch": 0.00398, + "grad_norm": 0.3677722018443306, + "learning_rate": 0.0011940000000000002, + "loss": 5.7479, + "step": 398 + }, + { + "epoch": 0.00399, + "grad_norm": 0.3480206547108819, + "learning_rate": 0.0011970000000000001, + "loss": 5.7478, + "step": 399 + }, + { + "epoch": 0.004, + "grad_norm": 0.3434828622202681, + "learning_rate": 0.0012000000000000001, + "loss": 5.7345, + "step": 400 + }, + { + "epoch": 0.00401, + "grad_norm": 0.34918136204349326, + "learning_rate": 0.001203, + "loss": 5.7155, + "step": 401 + }, + { + "epoch": 0.00402, + "grad_norm": 0.30554980038341767, + "learning_rate": 0.001206, + "loss": 5.7215, + "step": 402 + }, + { + "epoch": 0.00403, + "grad_norm": 0.38840665384838735, + "learning_rate": 0.001209, + "loss": 5.7184, + "step": 403 + }, + { + "epoch": 0.00404, + "grad_norm": 0.5409513056663879, + "learning_rate": 0.0012120000000000002, + "loss": 5.714, + "step": 404 + }, + { + "epoch": 0.00405, + "grad_norm": 0.845020924848713, + "learning_rate": 0.0012150000000000002, + "loss": 5.7181, + "step": 405 + }, + { + "epoch": 0.00406, + "grad_norm": 1.1911410915070972, + "learning_rate": 0.0012180000000000001, + "loss": 5.7205, + "step": 406 + }, + { + "epoch": 0.00407, + "grad_norm": 0.6577283880630926, + "learning_rate": 0.0012209999999999999, + "loss": 5.6994, + "step": 407 + }, + { + "epoch": 0.00408, + "grad_norm": 0.7475745975098248, + "learning_rate": 0.001224, + "loss": 5.7213, + "step": 408 + }, + { + "epoch": 0.00409, + "grad_norm": 0.9872391126413178, + "learning_rate": 0.001227, + "loss": 5.7126, + "step": 409 + }, + { + "epoch": 0.0041, + "grad_norm": 1.1359034721668335, + "learning_rate": 0.00123, + "loss": 5.7088, + "step": 410 + }, + { + "epoch": 0.00411, + "grad_norm": 1.3596329145222696, + "learning_rate": 0.001233, + "loss": 5.7402, + "step": 411 + }, + { + "epoch": 0.00412, + "grad_norm": 0.7538358474928969, + "learning_rate": 0.001236, + "loss": 5.7066, + "step": 412 + }, + { + "epoch": 0.00413, + "grad_norm": 0.9465320539051596, + "learning_rate": 0.0012389999999999999, + "loss": 5.7197, + "step": 413 + }, + { + "epoch": 0.00414, + "grad_norm": 0.9262933655624658, + "learning_rate": 0.001242, + "loss": 5.6978, + "step": 414 + }, + { + "epoch": 0.00415, + "grad_norm": 1.1564175286146172, + "learning_rate": 0.001245, + "loss": 5.7105, + "step": 415 + }, + { + "epoch": 0.00416, + "grad_norm": 1.1001247072345506, + "learning_rate": 0.001248, + "loss": 5.6929, + "step": 416 + }, + { + "epoch": 0.00417, + "grad_norm": 1.0416153435685582, + "learning_rate": 0.001251, + "loss": 5.7199, + "step": 417 + }, + { + "epoch": 0.00418, + "grad_norm": 1.0281555694116995, + "learning_rate": 0.001254, + "loss": 5.6999, + "step": 418 + }, + { + "epoch": 0.00419, + "grad_norm": 1.1154617103247704, + "learning_rate": 0.0012569999999999999, + "loss": 5.6876, + "step": 419 + }, + { + "epoch": 0.0042, + "grad_norm": 0.9999912825556322, + "learning_rate": 0.00126, + "loss": 5.7045, + "step": 420 + }, + { + "epoch": 0.00421, + "grad_norm": 0.9729638313238949, + "learning_rate": 0.001263, + "loss": 5.6933, + "step": 421 + }, + { + "epoch": 0.00422, + "grad_norm": 1.0297208241186608, + "learning_rate": 0.001266, + "loss": 5.6894, + "step": 422 + }, + { + "epoch": 0.00423, + "grad_norm": 0.8381062456476874, + "learning_rate": 0.001269, + "loss": 5.6811, + "step": 423 + }, + { + "epoch": 0.00424, + "grad_norm": 0.73825928337582, + "learning_rate": 0.001272, + "loss": 5.6773, + "step": 424 + }, + { + "epoch": 0.00425, + "grad_norm": 0.8151693610653118, + "learning_rate": 0.001275, + "loss": 5.6815, + "step": 425 + }, + { + "epoch": 0.00426, + "grad_norm": 0.9877723869544494, + "learning_rate": 0.001278, + "loss": 5.6881, + "step": 426 + }, + { + "epoch": 0.00427, + "grad_norm": 0.8875063218226354, + "learning_rate": 0.001281, + "loss": 5.6676, + "step": 427 + }, + { + "epoch": 0.00428, + "grad_norm": 0.8761493100474917, + "learning_rate": 0.001284, + "loss": 5.6437, + "step": 428 + }, + { + "epoch": 0.00429, + "grad_norm": 1.109439755296158, + "learning_rate": 0.001287, + "loss": 5.668, + "step": 429 + }, + { + "epoch": 0.0043, + "grad_norm": 0.8726361797071612, + "learning_rate": 0.00129, + "loss": 5.6611, + "step": 430 + }, + { + "epoch": 0.00431, + "grad_norm": 0.5601358924900582, + "learning_rate": 0.001293, + "loss": 5.6342, + "step": 431 + }, + { + "epoch": 0.00432, + "grad_norm": 0.5940988397687699, + "learning_rate": 0.001296, + "loss": 5.6448, + "step": 432 + }, + { + "epoch": 0.00433, + "grad_norm": 0.480538337754196, + "learning_rate": 0.001299, + "loss": 5.6336, + "step": 433 + }, + { + "epoch": 0.00434, + "grad_norm": 0.5069699406038057, + "learning_rate": 0.001302, + "loss": 5.6331, + "step": 434 + }, + { + "epoch": 0.00435, + "grad_norm": 0.417220569666801, + "learning_rate": 0.001305, + "loss": 5.6227, + "step": 435 + }, + { + "epoch": 0.00436, + "grad_norm": 0.46597565539240443, + "learning_rate": 0.001308, + "loss": 5.6039, + "step": 436 + }, + { + "epoch": 0.00437, + "grad_norm": 0.37605303006482044, + "learning_rate": 0.001311, + "loss": 5.6161, + "step": 437 + }, + { + "epoch": 0.00438, + "grad_norm": 0.3526339213940271, + "learning_rate": 0.001314, + "loss": 5.5977, + "step": 438 + }, + { + "epoch": 0.00439, + "grad_norm": 0.3618369277094543, + "learning_rate": 0.001317, + "loss": 5.6246, + "step": 439 + }, + { + "epoch": 0.0044, + "grad_norm": 0.3223158135938896, + "learning_rate": 0.00132, + "loss": 5.5938, + "step": 440 + }, + { + "epoch": 0.00441, + "grad_norm": 0.3386640445759432, + "learning_rate": 0.001323, + "loss": 5.5905, + "step": 441 + }, + { + "epoch": 0.00442, + "grad_norm": 0.3397360216396013, + "learning_rate": 0.0013260000000000001, + "loss": 5.5838, + "step": 442 + }, + { + "epoch": 0.00443, + "grad_norm": 0.34532354892574607, + "learning_rate": 0.001329, + "loss": 5.5832, + "step": 443 + }, + { + "epoch": 0.00444, + "grad_norm": 0.37928556611065656, + "learning_rate": 0.001332, + "loss": 5.5853, + "step": 444 + }, + { + "epoch": 0.00445, + "grad_norm": 0.44947894711961484, + "learning_rate": 0.001335, + "loss": 5.5723, + "step": 445 + }, + { + "epoch": 0.00446, + "grad_norm": 0.5239308309237933, + "learning_rate": 0.001338, + "loss": 5.5751, + "step": 446 + }, + { + "epoch": 0.00447, + "grad_norm": 0.60889528761374, + "learning_rate": 0.001341, + "loss": 5.5777, + "step": 447 + }, + { + "epoch": 0.00448, + "grad_norm": 0.6150556040535831, + "learning_rate": 0.0013440000000000001, + "loss": 5.561, + "step": 448 + }, + { + "epoch": 0.00449, + "grad_norm": 0.5444850551876294, + "learning_rate": 0.001347, + "loss": 5.5623, + "step": 449 + }, + { + "epoch": 0.0045, + "grad_norm": 0.6163688303714219, + "learning_rate": 0.00135, + "loss": 5.5617, + "step": 450 + }, + { + "epoch": 0.00451, + "grad_norm": 0.7972728836658292, + "learning_rate": 0.001353, + "loss": 5.5614, + "step": 451 + }, + { + "epoch": 0.00452, + "grad_norm": 0.7711764764543457, + "learning_rate": 0.001356, + "loss": 5.5454, + "step": 452 + }, + { + "epoch": 0.00453, + "grad_norm": 0.7702356986189732, + "learning_rate": 0.001359, + "loss": 5.5379, + "step": 453 + }, + { + "epoch": 0.00454, + "grad_norm": 1.0838160417982272, + "learning_rate": 0.0013620000000000001, + "loss": 5.5686, + "step": 454 + }, + { + "epoch": 0.00455, + "grad_norm": 1.152655585803101, + "learning_rate": 0.0013650000000000001, + "loss": 5.5757, + "step": 455 + }, + { + "epoch": 0.00456, + "grad_norm": 1.2776355160689266, + "learning_rate": 0.001368, + "loss": 5.5831, + "step": 456 + }, + { + "epoch": 0.00457, + "grad_norm": 1.091292995937963, + "learning_rate": 0.001371, + "loss": 5.5727, + "step": 457 + }, + { + "epoch": 0.00458, + "grad_norm": 0.8963405103823251, + "learning_rate": 0.001374, + "loss": 5.5735, + "step": 458 + }, + { + "epoch": 0.00459, + "grad_norm": 1.0168648046101516, + "learning_rate": 0.0013770000000000002, + "loss": 5.5669, + "step": 459 + }, + { + "epoch": 0.0046, + "grad_norm": 1.296131601782866, + "learning_rate": 0.0013800000000000002, + "loss": 5.5665, + "step": 460 + }, + { + "epoch": 0.00461, + "grad_norm": 0.9634420565591739, + "learning_rate": 0.0013830000000000001, + "loss": 5.556, + "step": 461 + }, + { + "epoch": 0.00462, + "grad_norm": 0.9383281224355017, + "learning_rate": 0.001386, + "loss": 5.5634, + "step": 462 + }, + { + "epoch": 0.00463, + "grad_norm": 1.2569585597421309, + "learning_rate": 0.001389, + "loss": 5.5542, + "step": 463 + }, + { + "epoch": 0.00464, + "grad_norm": 0.9874595595654581, + "learning_rate": 0.001392, + "loss": 5.5689, + "step": 464 + }, + { + "epoch": 0.00465, + "grad_norm": 1.085595749506429, + "learning_rate": 0.0013950000000000002, + "loss": 5.5385, + "step": 465 + }, + { + "epoch": 0.00466, + "grad_norm": 1.0673943770446899, + "learning_rate": 0.0013980000000000002, + "loss": 5.5603, + "step": 466 + }, + { + "epoch": 0.00467, + "grad_norm": 0.8139501043376736, + "learning_rate": 0.0014010000000000001, + "loss": 5.5432, + "step": 467 + }, + { + "epoch": 0.00468, + "grad_norm": 0.7494382793960519, + "learning_rate": 0.001404, + "loss": 5.5245, + "step": 468 + }, + { + "epoch": 0.00469, + "grad_norm": 0.7634992086588068, + "learning_rate": 0.001407, + "loss": 5.5282, + "step": 469 + }, + { + "epoch": 0.0047, + "grad_norm": 0.8018093758476836, + "learning_rate": 0.00141, + "loss": 5.5404, + "step": 470 + }, + { + "epoch": 0.00471, + "grad_norm": 0.7418690809708749, + "learning_rate": 0.001413, + "loss": 5.5115, + "step": 471 + }, + { + "epoch": 0.00472, + "grad_norm": 0.7355325431039438, + "learning_rate": 0.001416, + "loss": 5.5216, + "step": 472 + }, + { + "epoch": 0.00473, + "grad_norm": 0.709026539269664, + "learning_rate": 0.001419, + "loss": 5.5305, + "step": 473 + }, + { + "epoch": 0.00474, + "grad_norm": 0.5742329758009745, + "learning_rate": 0.0014219999999999999, + "loss": 5.5064, + "step": 474 + }, + { + "epoch": 0.00475, + "grad_norm": 0.5859758403725885, + "learning_rate": 0.001425, + "loss": 5.4971, + "step": 475 + }, + { + "epoch": 0.00476, + "grad_norm": 0.6365922795308678, + "learning_rate": 0.001428, + "loss": 5.5308, + "step": 476 + }, + { + "epoch": 0.00477, + "grad_norm": 0.6539516343537074, + "learning_rate": 0.001431, + "loss": 5.4846, + "step": 477 + }, + { + "epoch": 0.00478, + "grad_norm": 0.6446859909585969, + "learning_rate": 0.001434, + "loss": 5.5074, + "step": 478 + }, + { + "epoch": 0.00479, + "grad_norm": 0.7791938725908187, + "learning_rate": 0.001437, + "loss": 5.4897, + "step": 479 + }, + { + "epoch": 0.0048, + "grad_norm": 0.7608695568360718, + "learning_rate": 0.0014399999999999999, + "loss": 5.4855, + "step": 480 + }, + { + "epoch": 0.00481, + "grad_norm": 0.5435552532069989, + "learning_rate": 0.001443, + "loss": 5.4813, + "step": 481 + }, + { + "epoch": 0.00482, + "grad_norm": 0.4961185149512517, + "learning_rate": 0.001446, + "loss": 5.4538, + "step": 482 + }, + { + "epoch": 0.00483, + "grad_norm": 0.5120902463904886, + "learning_rate": 0.001449, + "loss": 5.4636, + "step": 483 + }, + { + "epoch": 0.00484, + "grad_norm": 0.418197369903841, + "learning_rate": 0.001452, + "loss": 5.464, + "step": 484 + }, + { + "epoch": 0.00485, + "grad_norm": 0.36311406822078424, + "learning_rate": 0.001455, + "loss": 5.4671, + "step": 485 + }, + { + "epoch": 0.00486, + "grad_norm": 0.3913777576995821, + "learning_rate": 0.001458, + "loss": 5.4393, + "step": 486 + }, + { + "epoch": 0.00487, + "grad_norm": 0.36874474197662527, + "learning_rate": 0.001461, + "loss": 5.4517, + "step": 487 + }, + { + "epoch": 0.00488, + "grad_norm": 0.41593519822402414, + "learning_rate": 0.001464, + "loss": 5.43, + "step": 488 + }, + { + "epoch": 0.00489, + "grad_norm": 0.4051383375955623, + "learning_rate": 0.001467, + "loss": 5.4204, + "step": 489 + }, + { + "epoch": 0.0049, + "grad_norm": 0.4948319932753325, + "learning_rate": 0.00147, + "loss": 5.434, + "step": 490 + }, + { + "epoch": 0.00491, + "grad_norm": 0.6887805434617323, + "learning_rate": 0.001473, + "loss": 5.4357, + "step": 491 + }, + { + "epoch": 0.00492, + "grad_norm": 0.912047432270828, + "learning_rate": 0.001476, + "loss": 5.4432, + "step": 492 + }, + { + "epoch": 0.00493, + "grad_norm": 0.8950009277905591, + "learning_rate": 0.001479, + "loss": 5.4415, + "step": 493 + }, + { + "epoch": 0.00494, + "grad_norm": 0.807683799932231, + "learning_rate": 0.001482, + "loss": 5.4427, + "step": 494 + }, + { + "epoch": 0.00495, + "grad_norm": 0.976722169908224, + "learning_rate": 0.001485, + "loss": 5.4601, + "step": 495 + }, + { + "epoch": 0.00496, + "grad_norm": 0.7224780529872387, + "learning_rate": 0.001488, + "loss": 5.4314, + "step": 496 + }, + { + "epoch": 0.00497, + "grad_norm": 0.7139613919522917, + "learning_rate": 0.001491, + "loss": 5.4172, + "step": 497 + }, + { + "epoch": 0.00498, + "grad_norm": 0.7556637282468179, + "learning_rate": 0.001494, + "loss": 5.4443, + "step": 498 + }, + { + "epoch": 0.00499, + "grad_norm": 0.8519321120342865, + "learning_rate": 0.001497, + "loss": 5.4223, + "step": 499 + }, + { + "epoch": 0.005, + "grad_norm": 0.9868888032280079, + "learning_rate": 0.0015, + "loss": 5.4308, + "step": 500 + }, + { + "epoch": 0.00501, + "grad_norm": 1.2028724714017198, + "learning_rate": 0.001503, + "loss": 5.4458, + "step": 501 + }, + { + "epoch": 0.00502, + "grad_norm": 0.9548534640519003, + "learning_rate": 0.001506, + "loss": 5.4405, + "step": 502 + }, + { + "epoch": 0.00503, + "grad_norm": 0.9423651398338494, + "learning_rate": 0.0015090000000000001, + "loss": 5.4484, + "step": 503 + }, + { + "epoch": 0.00504, + "grad_norm": 0.9672711526274779, + "learning_rate": 0.001512, + "loss": 5.4328, + "step": 504 + }, + { + "epoch": 0.00505, + "grad_norm": 0.9474889600476256, + "learning_rate": 0.001515, + "loss": 5.4283, + "step": 505 + }, + { + "epoch": 0.00506, + "grad_norm": 1.3039451405080307, + "learning_rate": 0.001518, + "loss": 5.4432, + "step": 506 + }, + { + "epoch": 0.00507, + "grad_norm": 1.2269782118632737, + "learning_rate": 0.001521, + "loss": 5.4452, + "step": 507 + }, + { + "epoch": 0.00508, + "grad_norm": 0.8597426683067237, + "learning_rate": 0.001524, + "loss": 5.4268, + "step": 508 + }, + { + "epoch": 0.00509, + "grad_norm": 0.9388657090373522, + "learning_rate": 0.0015270000000000001, + "loss": 5.4178, + "step": 509 + }, + { + "epoch": 0.0051, + "grad_norm": 0.9594427813189665, + "learning_rate": 0.0015300000000000001, + "loss": 5.4356, + "step": 510 + }, + { + "epoch": 0.00511, + "grad_norm": 1.04563577032056, + "learning_rate": 0.001533, + "loss": 5.4212, + "step": 511 + }, + { + "epoch": 0.00512, + "grad_norm": 0.733703407645156, + "learning_rate": 0.001536, + "loss": 5.4003, + "step": 512 + }, + { + "epoch": 0.00513, + "grad_norm": 0.8415210942026606, + "learning_rate": 0.001539, + "loss": 5.423, + "step": 513 + }, + { + "epoch": 0.00514, + "grad_norm": 0.8791751992621939, + "learning_rate": 0.001542, + "loss": 5.4064, + "step": 514 + }, + { + "epoch": 0.00515, + "grad_norm": 0.8161499995578689, + "learning_rate": 0.0015450000000000001, + "loss": 5.4094, + "step": 515 + }, + { + "epoch": 0.00516, + "grad_norm": 0.878383695319614, + "learning_rate": 0.0015480000000000001, + "loss": 5.4087, + "step": 516 + }, + { + "epoch": 0.00517, + "grad_norm": 0.9768725869756134, + "learning_rate": 0.001551, + "loss": 5.4055, + "step": 517 + }, + { + "epoch": 0.00518, + "grad_norm": 0.8865017723772849, + "learning_rate": 0.001554, + "loss": 5.3907, + "step": 518 + }, + { + "epoch": 0.00519, + "grad_norm": 0.8308797688973832, + "learning_rate": 0.001557, + "loss": 5.3905, + "step": 519 + }, + { + "epoch": 0.0052, + "grad_norm": 0.6978413162257922, + "learning_rate": 0.0015600000000000002, + "loss": 5.3938, + "step": 520 + }, + { + "epoch": 0.00521, + "grad_norm": 0.6562689530690187, + "learning_rate": 0.0015630000000000002, + "loss": 5.3676, + "step": 521 + }, + { + "epoch": 0.00522, + "grad_norm": 0.5577523148431155, + "learning_rate": 0.0015660000000000001, + "loss": 5.3673, + "step": 522 + }, + { + "epoch": 0.00523, + "grad_norm": 0.5298728018270966, + "learning_rate": 0.001569, + "loss": 5.3784, + "step": 523 + }, + { + "epoch": 0.00524, + "grad_norm": 0.44216623389663734, + "learning_rate": 0.001572, + "loss": 5.3811, + "step": 524 + }, + { + "epoch": 0.00525, + "grad_norm": 0.3702182111689363, + "learning_rate": 0.001575, + "loss": 5.369, + "step": 525 + }, + { + "epoch": 0.00526, + "grad_norm": 0.40289552198632295, + "learning_rate": 0.0015780000000000002, + "loss": 5.3444, + "step": 526 + }, + { + "epoch": 0.00527, + "grad_norm": 0.38284579814689895, + "learning_rate": 0.0015810000000000002, + "loss": 5.3551, + "step": 527 + }, + { + "epoch": 0.00528, + "grad_norm": 0.33950473031510653, + "learning_rate": 0.0015840000000000001, + "loss": 5.3339, + "step": 528 + }, + { + "epoch": 0.00529, + "grad_norm": 0.3777758983585419, + "learning_rate": 0.001587, + "loss": 5.343, + "step": 529 + }, + { + "epoch": 0.0053, + "grad_norm": 0.4257451161382566, + "learning_rate": 0.00159, + "loss": 5.3483, + "step": 530 + }, + { + "epoch": 0.00531, + "grad_norm": 0.5472045550610978, + "learning_rate": 0.001593, + "loss": 5.3387, + "step": 531 + }, + { + "epoch": 0.00532, + "grad_norm": 0.696856419507981, + "learning_rate": 0.0015960000000000002, + "loss": 5.3455, + "step": 532 + }, + { + "epoch": 0.00533, + "grad_norm": 0.7975941430607876, + "learning_rate": 0.0015990000000000002, + "loss": 5.3447, + "step": 533 + }, + { + "epoch": 0.00534, + "grad_norm": 0.6576622452461942, + "learning_rate": 0.0016020000000000001, + "loss": 5.3339, + "step": 534 + }, + { + "epoch": 0.00535, + "grad_norm": 0.5879483820814444, + "learning_rate": 0.001605, + "loss": 5.3347, + "step": 535 + }, + { + "epoch": 0.00536, + "grad_norm": 0.8212636469840171, + "learning_rate": 0.001608, + "loss": 5.3454, + "step": 536 + }, + { + "epoch": 0.00537, + "grad_norm": 0.7740232031924225, + "learning_rate": 0.0016110000000000002, + "loss": 5.3418, + "step": 537 + }, + { + "epoch": 0.00538, + "grad_norm": 0.683098129060214, + "learning_rate": 0.0016140000000000002, + "loss": 5.3143, + "step": 538 + }, + { + "epoch": 0.00539, + "grad_norm": 0.9170551975741953, + "learning_rate": 0.0016170000000000002, + "loss": 5.3245, + "step": 539 + }, + { + "epoch": 0.0054, + "grad_norm": 0.765080131514484, + "learning_rate": 0.0016200000000000001, + "loss": 5.3262, + "step": 540 + }, + { + "epoch": 0.00541, + "grad_norm": 0.658457026305436, + "learning_rate": 0.001623, + "loss": 5.3137, + "step": 541 + }, + { + "epoch": 0.00542, + "grad_norm": 0.601942869875084, + "learning_rate": 0.001626, + "loss": 5.3315, + "step": 542 + }, + { + "epoch": 0.00543, + "grad_norm": 0.6751097730454854, + "learning_rate": 0.0016290000000000002, + "loss": 5.2998, + "step": 543 + }, + { + "epoch": 0.00544, + "grad_norm": 0.6943293389301006, + "learning_rate": 0.0016320000000000002, + "loss": 5.3191, + "step": 544 + }, + { + "epoch": 0.00545, + "grad_norm": 0.757591621302123, + "learning_rate": 0.0016350000000000002, + "loss": 5.3293, + "step": 545 + }, + { + "epoch": 0.00546, + "grad_norm": 0.825685624372282, + "learning_rate": 0.0016380000000000001, + "loss": 5.3061, + "step": 546 + }, + { + "epoch": 0.00547, + "grad_norm": 0.8411824339962438, + "learning_rate": 0.001641, + "loss": 5.3051, + "step": 547 + }, + { + "epoch": 0.00548, + "grad_norm": 0.8225125547688507, + "learning_rate": 0.001644, + "loss": 5.303, + "step": 548 + }, + { + "epoch": 0.00549, + "grad_norm": 0.7692066087665821, + "learning_rate": 0.0016470000000000002, + "loss": 5.3016, + "step": 549 + }, + { + "epoch": 0.0055, + "grad_norm": 0.7541680263658305, + "learning_rate": 0.0016500000000000002, + "loss": 5.3099, + "step": 550 + }, + { + "epoch": 0.00551, + "grad_norm": 0.8632990593818363, + "learning_rate": 0.0016530000000000002, + "loss": 5.3123, + "step": 551 + }, + { + "epoch": 0.00552, + "grad_norm": 0.9083583396116485, + "learning_rate": 0.0016560000000000001, + "loss": 5.3063, + "step": 552 + }, + { + "epoch": 0.00553, + "grad_norm": 0.9125681148017897, + "learning_rate": 0.001659, + "loss": 5.304, + "step": 553 + }, + { + "epoch": 0.00554, + "grad_norm": 0.9360850923631763, + "learning_rate": 0.0016620000000000003, + "loss": 5.3021, + "step": 554 + }, + { + "epoch": 0.00555, + "grad_norm": 0.9519012966124738, + "learning_rate": 0.0016650000000000002, + "loss": 5.3154, + "step": 555 + }, + { + "epoch": 0.00556, + "grad_norm": 0.9297335237671114, + "learning_rate": 0.0016680000000000002, + "loss": 5.2962, + "step": 556 + }, + { + "epoch": 0.00557, + "grad_norm": 1.1187109821510643, + "learning_rate": 0.0016710000000000002, + "loss": 5.2935, + "step": 557 + }, + { + "epoch": 0.00558, + "grad_norm": 0.9868599839032521, + "learning_rate": 0.0016740000000000001, + "loss": 5.3219, + "step": 558 + }, + { + "epoch": 0.00559, + "grad_norm": 0.8786610310513777, + "learning_rate": 0.001677, + "loss": 5.2938, + "step": 559 + }, + { + "epoch": 0.0056, + "grad_norm": 0.9118163270219521, + "learning_rate": 0.0016800000000000003, + "loss": 5.2767, + "step": 560 + }, + { + "epoch": 0.00561, + "grad_norm": 0.8649386991384733, + "learning_rate": 0.0016830000000000003, + "loss": 5.3004, + "step": 561 + }, + { + "epoch": 0.00562, + "grad_norm": 0.5982058914442256, + "learning_rate": 0.0016860000000000002, + "loss": 5.279, + "step": 562 + }, + { + "epoch": 0.00563, + "grad_norm": 0.5830709316445766, + "learning_rate": 0.001689, + "loss": 5.2821, + "step": 563 + }, + { + "epoch": 0.00564, + "grad_norm": 0.5564672250442253, + "learning_rate": 0.001692, + "loss": 5.2582, + "step": 564 + }, + { + "epoch": 0.00565, + "grad_norm": 0.5903966881939692, + "learning_rate": 0.001695, + "loss": 5.2619, + "step": 565 + }, + { + "epoch": 0.00566, + "grad_norm": 0.5777761648359326, + "learning_rate": 0.0016979999999999999, + "loss": 5.2552, + "step": 566 + }, + { + "epoch": 0.00567, + "grad_norm": 0.6069030603134064, + "learning_rate": 0.0017009999999999998, + "loss": 5.2491, + "step": 567 + }, + { + "epoch": 0.00568, + "grad_norm": 0.5693132087018719, + "learning_rate": 0.0017039999999999998, + "loss": 5.2604, + "step": 568 + }, + { + "epoch": 0.00569, + "grad_norm": 0.47662406778838745, + "learning_rate": 0.001707, + "loss": 5.2359, + "step": 569 + }, + { + "epoch": 0.0057, + "grad_norm": 0.49231021705037487, + "learning_rate": 0.00171, + "loss": 5.253, + "step": 570 + }, + { + "epoch": 0.00571, + "grad_norm": 0.4167352661720621, + "learning_rate": 0.001713, + "loss": 5.2491, + "step": 571 + }, + { + "epoch": 0.00572, + "grad_norm": 0.39011854138215074, + "learning_rate": 0.0017159999999999999, + "loss": 5.2258, + "step": 572 + }, + { + "epoch": 0.00573, + "grad_norm": 0.45971389455263184, + "learning_rate": 0.0017189999999999998, + "loss": 5.2313, + "step": 573 + }, + { + "epoch": 0.00574, + "grad_norm": 0.4643694238461635, + "learning_rate": 0.001722, + "loss": 5.2399, + "step": 574 + }, + { + "epoch": 0.00575, + "grad_norm": 0.4439840100405838, + "learning_rate": 0.001725, + "loss": 5.2168, + "step": 575 + }, + { + "epoch": 0.00576, + "grad_norm": 0.4794795740566699, + "learning_rate": 0.001728, + "loss": 5.2109, + "step": 576 + }, + { + "epoch": 0.00577, + "grad_norm": 0.5445912068882687, + "learning_rate": 0.001731, + "loss": 5.2354, + "step": 577 + }, + { + "epoch": 0.00578, + "grad_norm": 0.5881619532039025, + "learning_rate": 0.0017339999999999999, + "loss": 5.2156, + "step": 578 + }, + { + "epoch": 0.00579, + "grad_norm": 0.6688964155004951, + "learning_rate": 0.0017369999999999998, + "loss": 5.2064, + "step": 579 + }, + { + "epoch": 0.0058, + "grad_norm": 0.6446644513052245, + "learning_rate": 0.00174, + "loss": 5.2223, + "step": 580 + }, + { + "epoch": 0.00581, + "grad_norm": 0.7385299343409043, + "learning_rate": 0.001743, + "loss": 5.2222, + "step": 581 + }, + { + "epoch": 0.00582, + "grad_norm": 1.0486135149054512, + "learning_rate": 0.001746, + "loss": 5.2114, + "step": 582 + }, + { + "epoch": 0.00583, + "grad_norm": 0.9184260585056472, + "learning_rate": 0.001749, + "loss": 5.2374, + "step": 583 + }, + { + "epoch": 0.00584, + "grad_norm": 0.7724931135788974, + "learning_rate": 0.0017519999999999999, + "loss": 5.2235, + "step": 584 + }, + { + "epoch": 0.00585, + "grad_norm": 0.903942587317279, + "learning_rate": 0.0017549999999999998, + "loss": 5.2218, + "step": 585 + }, + { + "epoch": 0.00586, + "grad_norm": 0.8496888678331875, + "learning_rate": 0.001758, + "loss": 5.2272, + "step": 586 + }, + { + "epoch": 0.00587, + "grad_norm": 0.8580070219006531, + "learning_rate": 0.001761, + "loss": 5.2094, + "step": 587 + }, + { + "epoch": 0.00588, + "grad_norm": 1.0100984711915582, + "learning_rate": 0.001764, + "loss": 5.233, + "step": 588 + }, + { + "epoch": 0.00589, + "grad_norm": 0.9311738464832717, + "learning_rate": 0.001767, + "loss": 5.2289, + "step": 589 + }, + { + "epoch": 0.0059, + "grad_norm": 0.9363758859588256, + "learning_rate": 0.0017699999999999999, + "loss": 5.2155, + "step": 590 + }, + { + "epoch": 0.00591, + "grad_norm": 0.9695025163620248, + "learning_rate": 0.001773, + "loss": 5.2165, + "step": 591 + }, + { + "epoch": 0.00592, + "grad_norm": 0.9805216711841677, + "learning_rate": 0.001776, + "loss": 5.2417, + "step": 592 + }, + { + "epoch": 0.00593, + "grad_norm": 0.9827241162755458, + "learning_rate": 0.001779, + "loss": 5.2392, + "step": 593 + }, + { + "epoch": 0.00594, + "grad_norm": 1.1445433178010636, + "learning_rate": 0.001782, + "loss": 5.2522, + "step": 594 + }, + { + "epoch": 0.00595, + "grad_norm": 0.950699335886284, + "learning_rate": 0.001785, + "loss": 5.2351, + "step": 595 + }, + { + "epoch": 0.00596, + "grad_norm": 0.8373647963925929, + "learning_rate": 0.0017879999999999999, + "loss": 5.2269, + "step": 596 + }, + { + "epoch": 0.00597, + "grad_norm": 0.9799263768822638, + "learning_rate": 0.001791, + "loss": 5.2118, + "step": 597 + }, + { + "epoch": 0.00598, + "grad_norm": 0.9413379648187816, + "learning_rate": 0.001794, + "loss": 5.2215, + "step": 598 + }, + { + "epoch": 0.00599, + "grad_norm": 0.8392983563516706, + "learning_rate": 0.001797, + "loss": 5.2191, + "step": 599 + }, + { + "epoch": 0.006, + "grad_norm": 0.8414286288934597, + "learning_rate": 0.0018, + "loss": 5.2134, + "step": 600 + }, + { + "epoch": 0.00601, + "grad_norm": 0.8993201270126511, + "learning_rate": 0.001803, + "loss": 5.2089, + "step": 601 + }, + { + "epoch": 0.00602, + "grad_norm": 0.8678008919316418, + "learning_rate": 0.0018059999999999999, + "loss": 5.1957, + "step": 602 + }, + { + "epoch": 0.00603, + "grad_norm": 0.6967382011990112, + "learning_rate": 0.001809, + "loss": 5.2122, + "step": 603 + }, + { + "epoch": 0.00604, + "grad_norm": 0.5312401482691863, + "learning_rate": 0.001812, + "loss": 5.1788, + "step": 604 + }, + { + "epoch": 0.00605, + "grad_norm": 0.5795052729398412, + "learning_rate": 0.001815, + "loss": 5.1915, + "step": 605 + }, + { + "epoch": 0.00606, + "grad_norm": 0.6345039250131549, + "learning_rate": 0.001818, + "loss": 5.1894, + "step": 606 + }, + { + "epoch": 0.00607, + "grad_norm": 0.6716049737606092, + "learning_rate": 0.001821, + "loss": 5.1916, + "step": 607 + }, + { + "epoch": 0.00608, + "grad_norm": 0.631863512626499, + "learning_rate": 0.001824, + "loss": 5.1762, + "step": 608 + }, + { + "epoch": 0.00609, + "grad_norm": 0.5345713642167306, + "learning_rate": 0.001827, + "loss": 5.177, + "step": 609 + }, + { + "epoch": 0.0061, + "grad_norm": 0.455626202663485, + "learning_rate": 0.00183, + "loss": 5.1434, + "step": 610 + }, + { + "epoch": 0.00611, + "grad_norm": 0.46197362206927406, + "learning_rate": 0.001833, + "loss": 5.1489, + "step": 611 + }, + { + "epoch": 0.00612, + "grad_norm": 0.3943653846553885, + "learning_rate": 0.001836, + "loss": 5.1488, + "step": 612 + }, + { + "epoch": 0.00613, + "grad_norm": 0.4053887102586243, + "learning_rate": 0.001839, + "loss": 5.1409, + "step": 613 + }, + { + "epoch": 0.00614, + "grad_norm": 0.44363743610311057, + "learning_rate": 0.001842, + "loss": 5.1611, + "step": 614 + }, + { + "epoch": 0.00615, + "grad_norm": 0.40025757691173514, + "learning_rate": 0.001845, + "loss": 5.145, + "step": 615 + }, + { + "epoch": 0.00616, + "grad_norm": 0.4067321452673224, + "learning_rate": 0.001848, + "loss": 5.1299, + "step": 616 + }, + { + "epoch": 0.00617, + "grad_norm": 0.42403873536126996, + "learning_rate": 0.001851, + "loss": 5.1397, + "step": 617 + }, + { + "epoch": 0.00618, + "grad_norm": 0.44192084147381183, + "learning_rate": 0.001854, + "loss": 5.1265, + "step": 618 + }, + { + "epoch": 0.00619, + "grad_norm": 0.5104655865853382, + "learning_rate": 0.001857, + "loss": 5.1299, + "step": 619 + }, + { + "epoch": 0.0062, + "grad_norm": 0.6226145292933671, + "learning_rate": 0.00186, + "loss": 5.118, + "step": 620 + }, + { + "epoch": 0.00621, + "grad_norm": 0.745237667035521, + "learning_rate": 0.001863, + "loss": 5.1237, + "step": 621 + }, + { + "epoch": 0.00622, + "grad_norm": 0.7986865810762169, + "learning_rate": 0.001866, + "loss": 5.1267, + "step": 622 + }, + { + "epoch": 0.00623, + "grad_norm": 0.7723208155309241, + "learning_rate": 0.001869, + "loss": 5.1084, + "step": 623 + }, + { + "epoch": 0.00624, + "grad_norm": 0.8171878766946273, + "learning_rate": 0.001872, + "loss": 5.1246, + "step": 624 + }, + { + "epoch": 0.00625, + "grad_norm": 0.7607424856276187, + "learning_rate": 0.001875, + "loss": 5.1359, + "step": 625 + }, + { + "epoch": 0.00626, + "grad_norm": 0.7005729458198662, + "learning_rate": 0.0018780000000000001, + "loss": 5.1104, + "step": 626 + }, + { + "epoch": 0.00627, + "grad_norm": 0.6735260479679158, + "learning_rate": 0.001881, + "loss": 5.1015, + "step": 627 + }, + { + "epoch": 0.00628, + "grad_norm": 0.780486998163919, + "learning_rate": 0.001884, + "loss": 5.1349, + "step": 628 + }, + { + "epoch": 0.00629, + "grad_norm": 0.8746830311438225, + "learning_rate": 0.001887, + "loss": 5.1097, + "step": 629 + }, + { + "epoch": 0.0063, + "grad_norm": 0.9536711950620466, + "learning_rate": 0.00189, + "loss": 5.1137, + "step": 630 + }, + { + "epoch": 0.00631, + "grad_norm": 0.8628106812141149, + "learning_rate": 0.0018930000000000002, + "loss": 5.1035, + "step": 631 + }, + { + "epoch": 0.00632, + "grad_norm": 0.9320890333045916, + "learning_rate": 0.0018960000000000001, + "loss": 5.1281, + "step": 632 + }, + { + "epoch": 0.00633, + "grad_norm": 1.0593968691082751, + "learning_rate": 0.001899, + "loss": 5.1321, + "step": 633 + }, + { + "epoch": 0.00634, + "grad_norm": 0.943443350801409, + "learning_rate": 0.001902, + "loss": 5.1182, + "step": 634 + }, + { + "epoch": 0.00635, + "grad_norm": 0.9628484504873114, + "learning_rate": 0.001905, + "loss": 5.1089, + "step": 635 + }, + { + "epoch": 0.00636, + "grad_norm": 1.1043383962750646, + "learning_rate": 0.001908, + "loss": 5.1289, + "step": 636 + }, + { + "epoch": 0.00637, + "grad_norm": 0.8987493866500654, + "learning_rate": 0.0019110000000000002, + "loss": 5.1351, + "step": 637 + }, + { + "epoch": 0.00638, + "grad_norm": 0.9251804377428581, + "learning_rate": 0.0019140000000000001, + "loss": 5.1288, + "step": 638 + }, + { + "epoch": 0.00639, + "grad_norm": 0.854418425044198, + "learning_rate": 0.001917, + "loss": 5.0998, + "step": 639 + }, + { + "epoch": 0.0064, + "grad_norm": 0.9324816679284724, + "learning_rate": 0.00192, + "loss": 5.1038, + "step": 640 + }, + { + "epoch": 0.00641, + "grad_norm": 0.9892889234371413, + "learning_rate": 0.001923, + "loss": 5.1163, + "step": 641 + }, + { + "epoch": 0.00642, + "grad_norm": 1.0346602459121752, + "learning_rate": 0.001926, + "loss": 5.106, + "step": 642 + }, + { + "epoch": 0.00643, + "grad_norm": 0.8661994645957561, + "learning_rate": 0.0019290000000000002, + "loss": 5.117, + "step": 643 + }, + { + "epoch": 0.00644, + "grad_norm": 0.8724056100423225, + "learning_rate": 0.0019320000000000001, + "loss": 5.0889, + "step": 644 + }, + { + "epoch": 0.00645, + "grad_norm": 0.8584186184200229, + "learning_rate": 0.001935, + "loss": 5.1004, + "step": 645 + }, + { + "epoch": 0.00646, + "grad_norm": 0.7360558672224548, + "learning_rate": 0.001938, + "loss": 5.0955, + "step": 646 + }, + { + "epoch": 0.00647, + "grad_norm": 0.7977702647925389, + "learning_rate": 0.001941, + "loss": 5.1058, + "step": 647 + }, + { + "epoch": 0.00648, + "grad_norm": 0.7872116543506851, + "learning_rate": 0.0019440000000000002, + "loss": 5.0908, + "step": 648 + }, + { + "epoch": 0.00649, + "grad_norm": 0.7104658813349117, + "learning_rate": 0.0019470000000000002, + "loss": 5.0718, + "step": 649 + }, + { + "epoch": 0.0065, + "grad_norm": 0.7453763255239747, + "learning_rate": 0.0019500000000000001, + "loss": 5.0953, + "step": 650 + }, + { + "epoch": 0.00651, + "grad_norm": 0.7781624388594444, + "learning_rate": 0.001953, + "loss": 5.0758, + "step": 651 + }, + { + "epoch": 0.00652, + "grad_norm": 0.7616046275009601, + "learning_rate": 0.0019560000000000003, + "loss": 5.0661, + "step": 652 + }, + { + "epoch": 0.00653, + "grad_norm": 0.5945469625366651, + "learning_rate": 0.0019590000000000002, + "loss": 5.0539, + "step": 653 + }, + { + "epoch": 0.00654, + "grad_norm": 0.6024408595794577, + "learning_rate": 0.001962, + "loss": 5.0374, + "step": 654 + }, + { + "epoch": 0.00655, + "grad_norm": 0.5905307565923603, + "learning_rate": 0.001965, + "loss": 5.048, + "step": 655 + }, + { + "epoch": 0.00656, + "grad_norm": 0.5236322372626927, + "learning_rate": 0.001968, + "loss": 5.04, + "step": 656 + }, + { + "epoch": 0.00657, + "grad_norm": 0.5283416618835216, + "learning_rate": 0.001971, + "loss": 5.0223, + "step": 657 + }, + { + "epoch": 0.00658, + "grad_norm": 0.5563146586062104, + "learning_rate": 0.001974, + "loss": 5.0415, + "step": 658 + }, + { + "epoch": 0.00659, + "grad_norm": 0.6297873363395704, + "learning_rate": 0.001977, + "loss": 5.0241, + "step": 659 + }, + { + "epoch": 0.0066, + "grad_norm": 0.5780538180580159, + "learning_rate": 0.00198, + "loss": 5.0197, + "step": 660 + }, + { + "epoch": 0.00661, + "grad_norm": 0.5505266007864265, + "learning_rate": 0.001983, + "loss": 4.9938, + "step": 661 + }, + { + "epoch": 0.00662, + "grad_norm": 0.6077995116545319, + "learning_rate": 0.0019860000000000004, + "loss": 4.9946, + "step": 662 + }, + { + "epoch": 0.00663, + "grad_norm": 0.693544741809416, + "learning_rate": 0.0019890000000000003, + "loss": 5.01, + "step": 663 + }, + { + "epoch": 0.00664, + "grad_norm": 0.801492715102265, + "learning_rate": 0.0019920000000000003, + "loss": 5.0017, + "step": 664 + }, + { + "epoch": 0.00665, + "grad_norm": 0.8803866066519176, + "learning_rate": 0.0019950000000000002, + "loss": 4.9845, + "step": 665 + }, + { + "epoch": 0.00666, + "grad_norm": 0.9133314823416234, + "learning_rate": 0.001998, + "loss": 5.0048, + "step": 666 + }, + { + "epoch": 0.00667, + "grad_norm": 0.9006055773427947, + "learning_rate": 0.002001, + "loss": 4.9967, + "step": 667 + }, + { + "epoch": 0.00668, + "grad_norm": 0.8268536663020751, + "learning_rate": 0.002004, + "loss": 5.0164, + "step": 668 + }, + { + "epoch": 0.00669, + "grad_norm": 1.1034218430158187, + "learning_rate": 0.002007, + "loss": 5.0358, + "step": 669 + }, + { + "epoch": 0.0067, + "grad_norm": 1.0710112933622913, + "learning_rate": 0.00201, + "loss": 5.0286, + "step": 670 + }, + { + "epoch": 0.00671, + "grad_norm": 1.0384170566197124, + "learning_rate": 0.002013, + "loss": 5.0085, + "step": 671 + }, + { + "epoch": 0.00672, + "grad_norm": 0.9247071512184438, + "learning_rate": 0.002016, + "loss": 4.9973, + "step": 672 + }, + { + "epoch": 0.00673, + "grad_norm": 0.9905174678816935, + "learning_rate": 0.002019, + "loss": 5.0151, + "step": 673 + }, + { + "epoch": 0.00674, + "grad_norm": 0.8930692065183647, + "learning_rate": 0.0020220000000000004, + "loss": 5.0014, + "step": 674 + }, + { + "epoch": 0.00675, + "grad_norm": 1.0067908289229996, + "learning_rate": 0.0020250000000000003, + "loss": 5.0182, + "step": 675 + }, + { + "epoch": 0.00676, + "grad_norm": 0.972209366764115, + "learning_rate": 0.0020280000000000003, + "loss": 5.0082, + "step": 676 + }, + { + "epoch": 0.00677, + "grad_norm": 1.0440090424594235, + "learning_rate": 0.0020310000000000003, + "loss": 5.0109, + "step": 677 + }, + { + "epoch": 0.00678, + "grad_norm": 1.1292207401563255, + "learning_rate": 0.0020340000000000002, + "loss": 5.0278, + "step": 678 + }, + { + "epoch": 0.00679, + "grad_norm": 0.9872089241398577, + "learning_rate": 0.002037, + "loss": 5.0054, + "step": 679 + }, + { + "epoch": 0.0068, + "grad_norm": 0.8748535789650627, + "learning_rate": 0.00204, + "loss": 5.0004, + "step": 680 + }, + { + "epoch": 0.00681, + "grad_norm": 1.0039141852392888, + "learning_rate": 0.002043, + "loss": 5.0018, + "step": 681 + }, + { + "epoch": 0.00682, + "grad_norm": 1.0414085118011525, + "learning_rate": 0.002046, + "loss": 5.0026, + "step": 682 + }, + { + "epoch": 0.00683, + "grad_norm": 0.8083819733286619, + "learning_rate": 0.002049, + "loss": 4.9746, + "step": 683 + }, + { + "epoch": 0.00684, + "grad_norm": 0.618682103661302, + "learning_rate": 0.002052, + "loss": 4.9803, + "step": 684 + }, + { + "epoch": 0.00685, + "grad_norm": 0.5775508276061283, + "learning_rate": 0.0020550000000000004, + "loss": 4.9792, + "step": 685 + }, + { + "epoch": 0.00686, + "grad_norm": 0.5359292821422027, + "learning_rate": 0.0020580000000000004, + "loss": 4.9508, + "step": 686 + }, + { + "epoch": 0.00687, + "grad_norm": 0.5439874489161504, + "learning_rate": 0.0020610000000000003, + "loss": 4.9456, + "step": 687 + }, + { + "epoch": 0.00688, + "grad_norm": 0.6521058673701751, + "learning_rate": 0.002064, + "loss": 4.9472, + "step": 688 + }, + { + "epoch": 0.00689, + "grad_norm": 0.7201992210148584, + "learning_rate": 0.002067, + "loss": 4.9514, + "step": 689 + }, + { + "epoch": 0.0069, + "grad_norm": 0.6329359289093167, + "learning_rate": 0.00207, + "loss": 4.9393, + "step": 690 + }, + { + "epoch": 0.00691, + "grad_norm": 0.5452341441548267, + "learning_rate": 0.0020729999999999998, + "loss": 4.9065, + "step": 691 + }, + { + "epoch": 0.00692, + "grad_norm": 0.524421270821715, + "learning_rate": 0.0020759999999999997, + "loss": 4.9264, + "step": 692 + }, + { + "epoch": 0.00693, + "grad_norm": 0.5237198583423548, + "learning_rate": 0.0020789999999999997, + "loss": 4.9275, + "step": 693 + }, + { + "epoch": 0.00694, + "grad_norm": 0.5943284166970615, + "learning_rate": 0.002082, + "loss": 4.9065, + "step": 694 + }, + { + "epoch": 0.00695, + "grad_norm": 0.6331066584123409, + "learning_rate": 0.002085, + "loss": 4.9096, + "step": 695 + }, + { + "epoch": 0.00696, + "grad_norm": 0.6552141599381052, + "learning_rate": 0.002088, + "loss": 4.9103, + "step": 696 + }, + { + "epoch": 0.00697, + "grad_norm": 0.7459258980378775, + "learning_rate": 0.002091, + "loss": 4.9188, + "step": 697 + }, + { + "epoch": 0.00698, + "grad_norm": 0.902413176094811, + "learning_rate": 0.002094, + "loss": 4.9109, + "step": 698 + }, + { + "epoch": 0.00699, + "grad_norm": 0.9488326760888935, + "learning_rate": 0.002097, + "loss": 4.9116, + "step": 699 + }, + { + "epoch": 0.007, + "grad_norm": 0.9189763967394823, + "learning_rate": 0.0021, + "loss": 4.929, + "step": 700 + }, + { + "epoch": 0.00701, + "grad_norm": 0.8924436046521577, + "learning_rate": 0.002103, + "loss": 4.9083, + "step": 701 + }, + { + "epoch": 0.00702, + "grad_norm": 0.8863614629240012, + "learning_rate": 0.002106, + "loss": 4.9092, + "step": 702 + }, + { + "epoch": 0.00703, + "grad_norm": 0.7940709541538681, + "learning_rate": 0.0021089999999999998, + "loss": 4.891, + "step": 703 + }, + { + "epoch": 0.00704, + "grad_norm": 0.7939787736751149, + "learning_rate": 0.0021119999999999997, + "loss": 4.8964, + "step": 704 + }, + { + "epoch": 0.00705, + "grad_norm": 1.0444656030359551, + "learning_rate": 0.002115, + "loss": 4.9151, + "step": 705 + }, + { + "epoch": 0.00706, + "grad_norm": 1.1431311909042268, + "learning_rate": 0.002118, + "loss": 4.9234, + "step": 706 + }, + { + "epoch": 0.00707, + "grad_norm": 0.8384635334186645, + "learning_rate": 0.002121, + "loss": 4.8902, + "step": 707 + }, + { + "epoch": 0.00708, + "grad_norm": 0.9594405079672866, + "learning_rate": 0.002124, + "loss": 4.8745, + "step": 708 + }, + { + "epoch": 0.00709, + "grad_norm": 0.8900382869322284, + "learning_rate": 0.002127, + "loss": 4.9161, + "step": 709 + }, + { + "epoch": 0.0071, + "grad_norm": 0.8868570415283396, + "learning_rate": 0.00213, + "loss": 4.9065, + "step": 710 + }, + { + "epoch": 0.00711, + "grad_norm": 0.8610490944817158, + "learning_rate": 0.002133, + "loss": 4.8804, + "step": 711 + }, + { + "epoch": 0.00712, + "grad_norm": 0.8727164938852855, + "learning_rate": 0.002136, + "loss": 4.9046, + "step": 712 + }, + { + "epoch": 0.00713, + "grad_norm": 0.8857525308493206, + "learning_rate": 0.002139, + "loss": 4.9135, + "step": 713 + }, + { + "epoch": 0.00714, + "grad_norm": 0.9495661806955594, + "learning_rate": 0.002142, + "loss": 4.918, + "step": 714 + }, + { + "epoch": 0.00715, + "grad_norm": 1.0263024097609161, + "learning_rate": 0.0021449999999999998, + "loss": 4.8857, + "step": 715 + }, + { + "epoch": 0.00716, + "grad_norm": 0.8876358680026493, + "learning_rate": 0.002148, + "loss": 4.8749, + "step": 716 + }, + { + "epoch": 0.00717, + "grad_norm": 0.8225498605776377, + "learning_rate": 0.002151, + "loss": 4.8925, + "step": 717 + }, + { + "epoch": 0.00718, + "grad_norm": 0.628552485265691, + "learning_rate": 0.002154, + "loss": 4.8659, + "step": 718 + }, + { + "epoch": 0.00719, + "grad_norm": 0.6584104654465238, + "learning_rate": 0.002157, + "loss": 4.8747, + "step": 719 + }, + { + "epoch": 0.0072, + "grad_norm": 0.6698592474865601, + "learning_rate": 0.00216, + "loss": 4.8635, + "step": 720 + }, + { + "epoch": 0.00721, + "grad_norm": 0.6673590176314685, + "learning_rate": 0.002163, + "loss": 4.8639, + "step": 721 + }, + { + "epoch": 0.00722, + "grad_norm": 0.6674098225397388, + "learning_rate": 0.002166, + "loss": 4.8386, + "step": 722 + }, + { + "epoch": 0.00723, + "grad_norm": 0.6090726175552883, + "learning_rate": 0.002169, + "loss": 4.8464, + "step": 723 + }, + { + "epoch": 0.00724, + "grad_norm": 0.6325507361418539, + "learning_rate": 0.002172, + "loss": 4.8403, + "step": 724 + }, + { + "epoch": 0.00725, + "grad_norm": 0.6927587431932604, + "learning_rate": 0.002175, + "loss": 4.8341, + "step": 725 + }, + { + "epoch": 0.00726, + "grad_norm": 0.7422551683158218, + "learning_rate": 0.002178, + "loss": 4.8448, + "step": 726 + }, + { + "epoch": 0.00727, + "grad_norm": 0.7946686392459241, + "learning_rate": 0.0021809999999999998, + "loss": 4.8264, + "step": 727 + }, + { + "epoch": 0.00728, + "grad_norm": 0.651194780867581, + "learning_rate": 0.002184, + "loss": 4.8373, + "step": 728 + }, + { + "epoch": 0.00729, + "grad_norm": 0.5507866158426874, + "learning_rate": 0.002187, + "loss": 4.8279, + "step": 729 + }, + { + "epoch": 0.0073, + "grad_norm": 0.5770531279665235, + "learning_rate": 0.00219, + "loss": 4.8256, + "step": 730 + }, + { + "epoch": 0.00731, + "grad_norm": 0.6604772562967653, + "learning_rate": 0.002193, + "loss": 4.8198, + "step": 731 + }, + { + "epoch": 0.00732, + "grad_norm": 0.7902754963422924, + "learning_rate": 0.002196, + "loss": 4.832, + "step": 732 + }, + { + "epoch": 0.00733, + "grad_norm": 0.8530754775219535, + "learning_rate": 0.002199, + "loss": 4.814, + "step": 733 + }, + { + "epoch": 0.00734, + "grad_norm": 0.808835939559151, + "learning_rate": 0.002202, + "loss": 4.8365, + "step": 734 + }, + { + "epoch": 0.00735, + "grad_norm": 0.7793455630355729, + "learning_rate": 0.002205, + "loss": 4.8484, + "step": 735 + }, + { + "epoch": 0.00736, + "grad_norm": 0.8899384636665534, + "learning_rate": 0.002208, + "loss": 4.8322, + "step": 736 + }, + { + "epoch": 0.00737, + "grad_norm": 1.0166382867407526, + "learning_rate": 0.002211, + "loss": 4.8145, + "step": 737 + }, + { + "epoch": 0.00738, + "grad_norm": 0.9857126416807526, + "learning_rate": 0.002214, + "loss": 4.8213, + "step": 738 + }, + { + "epoch": 0.00739, + "grad_norm": 0.8982016923721464, + "learning_rate": 0.0022170000000000002, + "loss": 4.809, + "step": 739 + }, + { + "epoch": 0.0074, + "grad_norm": 0.8517808259905928, + "learning_rate": 0.00222, + "loss": 4.8138, + "step": 740 + }, + { + "epoch": 0.00741, + "grad_norm": 0.6571636698582773, + "learning_rate": 0.002223, + "loss": 4.778, + "step": 741 + }, + { + "epoch": 0.00742, + "grad_norm": 0.5983530041008951, + "learning_rate": 0.002226, + "loss": 4.8043, + "step": 742 + }, + { + "epoch": 0.00743, + "grad_norm": 0.613767022264535, + "learning_rate": 0.002229, + "loss": 4.772, + "step": 743 + }, + { + "epoch": 0.00744, + "grad_norm": 0.5746332772613801, + "learning_rate": 0.002232, + "loss": 4.7755, + "step": 744 + }, + { + "epoch": 0.00745, + "grad_norm": 0.5491916195482714, + "learning_rate": 0.002235, + "loss": 4.7792, + "step": 745 + }, + { + "epoch": 0.00746, + "grad_norm": 0.6224704559098753, + "learning_rate": 0.002238, + "loss": 4.7785, + "step": 746 + }, + { + "epoch": 0.00747, + "grad_norm": 0.721423929849636, + "learning_rate": 0.002241, + "loss": 4.766, + "step": 747 + }, + { + "epoch": 0.00748, + "grad_norm": 0.8211263973312402, + "learning_rate": 0.002244, + "loss": 4.7825, + "step": 748 + }, + { + "epoch": 0.00749, + "grad_norm": 0.9630311220772746, + "learning_rate": 0.002247, + "loss": 4.7687, + "step": 749 + }, + { + "epoch": 0.0075, + "grad_norm": 1.0671208539128567, + "learning_rate": 0.0022500000000000003, + "loss": 4.8201, + "step": 750 + }, + { + "epoch": 0.00751, + "grad_norm": 1.0285377005887373, + "learning_rate": 0.0022530000000000002, + "loss": 4.8053, + "step": 751 + }, + { + "epoch": 0.00752, + "grad_norm": 0.9937387135055332, + "learning_rate": 0.002256, + "loss": 4.7635, + "step": 752 + }, + { + "epoch": 0.00753, + "grad_norm": 0.8939099354397223, + "learning_rate": 0.002259, + "loss": 4.8048, + "step": 753 + }, + { + "epoch": 0.00754, + "grad_norm": 0.958657384547811, + "learning_rate": 0.002262, + "loss": 4.816, + "step": 754 + }, + { + "epoch": 0.00755, + "grad_norm": 0.8579165829387244, + "learning_rate": 0.002265, + "loss": 4.7809, + "step": 755 + }, + { + "epoch": 0.00756, + "grad_norm": 0.8147493051985796, + "learning_rate": 0.002268, + "loss": 4.7687, + "step": 756 + }, + { + "epoch": 0.00757, + "grad_norm": 0.9899674342411924, + "learning_rate": 0.002271, + "loss": 4.7845, + "step": 757 + }, + { + "epoch": 0.00758, + "grad_norm": 1.2327208130150207, + "learning_rate": 0.002274, + "loss": 4.7967, + "step": 758 + }, + { + "epoch": 0.00759, + "grad_norm": 0.8760473410677282, + "learning_rate": 0.002277, + "loss": 4.7951, + "step": 759 + }, + { + "epoch": 0.0076, + "grad_norm": 0.9553393765090987, + "learning_rate": 0.00228, + "loss": 4.7842, + "step": 760 + }, + { + "epoch": 0.00761, + "grad_norm": 1.088855928225056, + "learning_rate": 0.002283, + "loss": 4.769, + "step": 761 + }, + { + "epoch": 0.00762, + "grad_norm": 0.9818480683263884, + "learning_rate": 0.0022860000000000003, + "loss": 4.7512, + "step": 762 + }, + { + "epoch": 0.00763, + "grad_norm": 0.9731540924632093, + "learning_rate": 0.0022890000000000002, + "loss": 4.7931, + "step": 763 + }, + { + "epoch": 0.00764, + "grad_norm": 1.0508884929557651, + "learning_rate": 0.002292, + "loss": 4.8167, + "step": 764 + }, + { + "epoch": 0.00765, + "grad_norm": 1.0020863769727308, + "learning_rate": 0.002295, + "loss": 4.7984, + "step": 765 + }, + { + "epoch": 0.00766, + "grad_norm": 1.1527463652354557, + "learning_rate": 0.002298, + "loss": 4.8085, + "step": 766 + }, + { + "epoch": 0.00767, + "grad_norm": 0.9657952239159258, + "learning_rate": 0.002301, + "loss": 4.7959, + "step": 767 + }, + { + "epoch": 0.00768, + "grad_norm": 1.0234917976922082, + "learning_rate": 0.002304, + "loss": 4.8012, + "step": 768 + }, + { + "epoch": 0.00769, + "grad_norm": 0.9850893067060651, + "learning_rate": 0.002307, + "loss": 4.8144, + "step": 769 + }, + { + "epoch": 0.0077, + "grad_norm": 0.9062134932024389, + "learning_rate": 0.00231, + "loss": 4.7653, + "step": 770 + }, + { + "epoch": 0.00771, + "grad_norm": 0.8476285286232204, + "learning_rate": 0.002313, + "loss": 4.7979, + "step": 771 + }, + { + "epoch": 0.00772, + "grad_norm": 0.9122213123018311, + "learning_rate": 0.002316, + "loss": 4.7851, + "step": 772 + }, + { + "epoch": 0.00773, + "grad_norm": 1.0718910624781612, + "learning_rate": 0.0023190000000000003, + "loss": 4.8052, + "step": 773 + }, + { + "epoch": 0.00774, + "grad_norm": 0.7792131883523417, + "learning_rate": 0.0023220000000000003, + "loss": 4.7945, + "step": 774 + }, + { + "epoch": 0.00775, + "grad_norm": 0.7995411986928386, + "learning_rate": 0.0023250000000000002, + "loss": 4.7914, + "step": 775 + }, + { + "epoch": 0.00776, + "grad_norm": 0.7054590225014301, + "learning_rate": 0.002328, + "loss": 4.7883, + "step": 776 + }, + { + "epoch": 0.00777, + "grad_norm": 0.6505869359405926, + "learning_rate": 0.002331, + "loss": 4.7585, + "step": 777 + }, + { + "epoch": 0.00778, + "grad_norm": 0.6484695284206986, + "learning_rate": 0.002334, + "loss": 4.7652, + "step": 778 + }, + { + "epoch": 0.00779, + "grad_norm": 0.6047799586124271, + "learning_rate": 0.002337, + "loss": 4.7239, + "step": 779 + }, + { + "epoch": 0.0078, + "grad_norm": 0.5436502526586032, + "learning_rate": 0.00234, + "loss": 4.7364, + "step": 780 + }, + { + "epoch": 0.00781, + "grad_norm": 0.5682167623371829, + "learning_rate": 0.002343, + "loss": 4.7355, + "step": 781 + }, + { + "epoch": 0.00782, + "grad_norm": 0.628910610744215, + "learning_rate": 0.002346, + "loss": 4.7356, + "step": 782 + }, + { + "epoch": 0.00783, + "grad_norm": 0.5845457532196663, + "learning_rate": 0.002349, + "loss": 4.7086, + "step": 783 + }, + { + "epoch": 0.00784, + "grad_norm": 0.6345965495249546, + "learning_rate": 0.002352, + "loss": 4.723, + "step": 784 + }, + { + "epoch": 0.00785, + "grad_norm": 0.4959017064429773, + "learning_rate": 0.0023550000000000003, + "loss": 4.7138, + "step": 785 + }, + { + "epoch": 0.00786, + "grad_norm": 0.4358915945164792, + "learning_rate": 0.0023580000000000003, + "loss": 4.7131, + "step": 786 + }, + { + "epoch": 0.00787, + "grad_norm": 0.4259891850743534, + "learning_rate": 0.0023610000000000003, + "loss": 4.7113, + "step": 787 + }, + { + "epoch": 0.00788, + "grad_norm": 0.4529949912379573, + "learning_rate": 0.002364, + "loss": 4.7077, + "step": 788 + }, + { + "epoch": 0.00789, + "grad_norm": 0.4944753699230628, + "learning_rate": 0.002367, + "loss": 4.7334, + "step": 789 + }, + { + "epoch": 0.0079, + "grad_norm": 0.6295031827770177, + "learning_rate": 0.00237, + "loss": 4.7144, + "step": 790 + }, + { + "epoch": 0.00791, + "grad_norm": 0.8474014146600959, + "learning_rate": 0.002373, + "loss": 4.6777, + "step": 791 + }, + { + "epoch": 0.00792, + "grad_norm": 0.8040402155973354, + "learning_rate": 0.002376, + "loss": 4.7267, + "step": 792 + }, + { + "epoch": 0.00793, + "grad_norm": 0.5568568781614048, + "learning_rate": 0.002379, + "loss": 4.7006, + "step": 793 + }, + { + "epoch": 0.00794, + "grad_norm": 0.8220014797505664, + "learning_rate": 0.002382, + "loss": 4.7246, + "step": 794 + }, + { + "epoch": 0.00795, + "grad_norm": 0.6551332917875898, + "learning_rate": 0.002385, + "loss": 4.6938, + "step": 795 + }, + { + "epoch": 0.00796, + "grad_norm": 0.5666814801389223, + "learning_rate": 0.0023880000000000004, + "loss": 4.6826, + "step": 796 + }, + { + "epoch": 0.00797, + "grad_norm": 0.654430341218369, + "learning_rate": 0.0023910000000000003, + "loss": 4.7248, + "step": 797 + }, + { + "epoch": 0.00798, + "grad_norm": 0.5511512463730408, + "learning_rate": 0.0023940000000000003, + "loss": 4.6864, + "step": 798 + }, + { + "epoch": 0.00799, + "grad_norm": 0.5084174359945534, + "learning_rate": 0.0023970000000000003, + "loss": 4.7161, + "step": 799 + }, + { + "epoch": 0.008, + "grad_norm": 0.41968398472595075, + "learning_rate": 0.0024000000000000002, + "loss": 4.6652, + "step": 800 + }, + { + "epoch": 0.00801, + "grad_norm": 0.4546309911468048, + "learning_rate": 0.002403, + "loss": 4.6967, + "step": 801 + }, + { + "epoch": 0.00802, + "grad_norm": 0.430018994115786, + "learning_rate": 0.002406, + "loss": 4.6827, + "step": 802 + }, + { + "epoch": 0.00803, + "grad_norm": 0.44614830716967085, + "learning_rate": 0.002409, + "loss": 4.6617, + "step": 803 + }, + { + "epoch": 0.00804, + "grad_norm": 0.44722400844593674, + "learning_rate": 0.002412, + "loss": 4.6598, + "step": 804 + }, + { + "epoch": 0.00805, + "grad_norm": 0.5179127215582825, + "learning_rate": 0.002415, + "loss": 4.6599, + "step": 805 + }, + { + "epoch": 0.00806, + "grad_norm": 0.5610832008078775, + "learning_rate": 0.002418, + "loss": 4.677, + "step": 806 + }, + { + "epoch": 0.00807, + "grad_norm": 0.5167453223410896, + "learning_rate": 0.0024210000000000004, + "loss": 4.6671, + "step": 807 + }, + { + "epoch": 0.00808, + "grad_norm": 0.46468933196331563, + "learning_rate": 0.0024240000000000004, + "loss": 4.6511, + "step": 808 + }, + { + "epoch": 0.00809, + "grad_norm": 0.5221883532574668, + "learning_rate": 0.0024270000000000003, + "loss": 4.6468, + "step": 809 + }, + { + "epoch": 0.0081, + "grad_norm": 0.4992566900849729, + "learning_rate": 0.0024300000000000003, + "loss": 4.6744, + "step": 810 + }, + { + "epoch": 0.00811, + "grad_norm": 0.4854147467055134, + "learning_rate": 0.0024330000000000003, + "loss": 4.646, + "step": 811 + }, + { + "epoch": 0.00812, + "grad_norm": 0.650970729431075, + "learning_rate": 0.0024360000000000002, + "loss": 4.6307, + "step": 812 + }, + { + "epoch": 0.00813, + "grad_norm": 0.8160691589494683, + "learning_rate": 0.0024389999999999998, + "loss": 4.6711, + "step": 813 + }, + { + "epoch": 0.00814, + "grad_norm": 0.9918101747931352, + "learning_rate": 0.0024419999999999997, + "loss": 4.6946, + "step": 814 + }, + { + "epoch": 0.00815, + "grad_norm": 1.247963175893729, + "learning_rate": 0.0024449999999999997, + "loss": 4.7226, + "step": 815 + }, + { + "epoch": 0.00816, + "grad_norm": 0.8376200515557375, + "learning_rate": 0.002448, + "loss": 4.6777, + "step": 816 + }, + { + "epoch": 0.00817, + "grad_norm": 0.9161032619759178, + "learning_rate": 0.002451, + "loss": 4.6939, + "step": 817 + }, + { + "epoch": 0.00818, + "grad_norm": 1.0914649908014256, + "learning_rate": 0.002454, + "loss": 4.6886, + "step": 818 + }, + { + "epoch": 0.00819, + "grad_norm": 0.9806171410774952, + "learning_rate": 0.002457, + "loss": 4.712, + "step": 819 + }, + { + "epoch": 0.0082, + "grad_norm": 0.992236077471004, + "learning_rate": 0.00246, + "loss": 4.6918, + "step": 820 + }, + { + "epoch": 0.00821, + "grad_norm": 1.0594557870263281, + "learning_rate": 0.002463, + "loss": 4.6759, + "step": 821 + }, + { + "epoch": 0.00822, + "grad_norm": 1.0346800919438124, + "learning_rate": 0.002466, + "loss": 4.6853, + "step": 822 + }, + { + "epoch": 0.00823, + "grad_norm": 0.9573573191186882, + "learning_rate": 0.002469, + "loss": 4.6833, + "step": 823 + }, + { + "epoch": 0.00824, + "grad_norm": 1.1123514933123841, + "learning_rate": 0.002472, + "loss": 4.714, + "step": 824 + }, + { + "epoch": 0.00825, + "grad_norm": 0.8463845700248506, + "learning_rate": 0.0024749999999999998, + "loss": 4.7191, + "step": 825 + }, + { + "epoch": 0.00826, + "grad_norm": 0.8444785606085857, + "learning_rate": 0.0024779999999999997, + "loss": 4.672, + "step": 826 + }, + { + "epoch": 0.00827, + "grad_norm": 0.9726341870117121, + "learning_rate": 0.002481, + "loss": 4.7078, + "step": 827 + }, + { + "epoch": 0.00828, + "grad_norm": 0.9106448417621353, + "learning_rate": 0.002484, + "loss": 4.7003, + "step": 828 + }, + { + "epoch": 0.00829, + "grad_norm": 0.7565680418878746, + "learning_rate": 0.002487, + "loss": 4.6856, + "step": 829 + }, + { + "epoch": 0.0083, + "grad_norm": 0.8537774465977133, + "learning_rate": 0.00249, + "loss": 4.7017, + "step": 830 + }, + { + "epoch": 0.00831, + "grad_norm": 0.9023323948099834, + "learning_rate": 0.002493, + "loss": 4.6871, + "step": 831 + }, + { + "epoch": 0.00832, + "grad_norm": 0.8524529451127855, + "learning_rate": 0.002496, + "loss": 4.6815, + "step": 832 + }, + { + "epoch": 0.00833, + "grad_norm": 0.9428655185832147, + "learning_rate": 0.002499, + "loss": 4.6808, + "step": 833 + }, + { + "epoch": 0.00834, + "grad_norm": 0.9597220185428569, + "learning_rate": 0.002502, + "loss": 4.6879, + "step": 834 + }, + { + "epoch": 0.00835, + "grad_norm": 0.7735101632354252, + "learning_rate": 0.002505, + "loss": 4.6724, + "step": 835 + }, + { + "epoch": 0.00836, + "grad_norm": 0.9597202731139803, + "learning_rate": 0.002508, + "loss": 4.7061, + "step": 836 + }, + { + "epoch": 0.00837, + "grad_norm": 0.9520863539431935, + "learning_rate": 0.0025109999999999998, + "loss": 4.6636, + "step": 837 + }, + { + "epoch": 0.00838, + "grad_norm": 0.7800128524395746, + "learning_rate": 0.0025139999999999997, + "loss": 4.6721, + "step": 838 + }, + { + "epoch": 0.00839, + "grad_norm": 0.8122589832425033, + "learning_rate": 0.002517, + "loss": 4.675, + "step": 839 + }, + { + "epoch": 0.0084, + "grad_norm": 0.8183344402395425, + "learning_rate": 0.00252, + "loss": 4.6669, + "step": 840 + }, + { + "epoch": 0.00841, + "grad_norm": 0.6918735110390536, + "learning_rate": 0.002523, + "loss": 4.6489, + "step": 841 + }, + { + "epoch": 0.00842, + "grad_norm": 0.6201385747244391, + "learning_rate": 0.002526, + "loss": 4.6423, + "step": 842 + }, + { + "epoch": 0.00843, + "grad_norm": 0.606127970479136, + "learning_rate": 0.002529, + "loss": 4.6465, + "step": 843 + }, + { + "epoch": 0.00844, + "grad_norm": 0.5515773209874846, + "learning_rate": 0.002532, + "loss": 4.6607, + "step": 844 + }, + { + "epoch": 0.00845, + "grad_norm": 0.6203742299859808, + "learning_rate": 0.002535, + "loss": 4.6293, + "step": 845 + }, + { + "epoch": 0.00846, + "grad_norm": 0.5875832865020281, + "learning_rate": 0.002538, + "loss": 4.6474, + "step": 846 + }, + { + "epoch": 0.00847, + "grad_norm": 0.5703256353430879, + "learning_rate": 0.002541, + "loss": 4.6282, + "step": 847 + }, + { + "epoch": 0.00848, + "grad_norm": 0.602830367643936, + "learning_rate": 0.002544, + "loss": 4.6269, + "step": 848 + }, + { + "epoch": 0.00849, + "grad_norm": 0.6741507039909044, + "learning_rate": 0.002547, + "loss": 4.6233, + "step": 849 + }, + { + "epoch": 0.0085, + "grad_norm": 0.6288739006540759, + "learning_rate": 0.00255, + "loss": 4.6341, + "step": 850 + }, + { + "epoch": 0.00851, + "grad_norm": 0.5820099008678455, + "learning_rate": 0.002553, + "loss": 4.644, + "step": 851 + }, + { + "epoch": 0.00852, + "grad_norm": 0.586123912558797, + "learning_rate": 0.002556, + "loss": 4.6367, + "step": 852 + }, + { + "epoch": 0.00853, + "grad_norm": 0.5127813487098001, + "learning_rate": 0.002559, + "loss": 4.6085, + "step": 853 + }, + { + "epoch": 0.00854, + "grad_norm": 0.4730499644759358, + "learning_rate": 0.002562, + "loss": 4.6029, + "step": 854 + }, + { + "epoch": 0.00855, + "grad_norm": 0.44708869980986227, + "learning_rate": 0.002565, + "loss": 4.5799, + "step": 855 + }, + { + "epoch": 0.00856, + "grad_norm": 0.466044480858233, + "learning_rate": 0.002568, + "loss": 4.6142, + "step": 856 + }, + { + "epoch": 0.00857, + "grad_norm": 0.5382201915945353, + "learning_rate": 0.002571, + "loss": 4.6036, + "step": 857 + }, + { + "epoch": 0.00858, + "grad_norm": 0.6780662034295477, + "learning_rate": 0.002574, + "loss": 4.609, + "step": 858 + }, + { + "epoch": 0.00859, + "grad_norm": 0.9086610382483981, + "learning_rate": 0.002577, + "loss": 4.6039, + "step": 859 + }, + { + "epoch": 0.0086, + "grad_norm": 0.8563688949272525, + "learning_rate": 0.00258, + "loss": 4.6531, + "step": 860 + }, + { + "epoch": 0.00861, + "grad_norm": 0.5965670098126366, + "learning_rate": 0.0025830000000000002, + "loss": 4.6461, + "step": 861 + }, + { + "epoch": 0.00862, + "grad_norm": 0.7975052365958228, + "learning_rate": 0.002586, + "loss": 4.611, + "step": 862 + }, + { + "epoch": 0.00863, + "grad_norm": 0.650099032572018, + "learning_rate": 0.002589, + "loss": 4.6284, + "step": 863 + }, + { + "epoch": 0.00864, + "grad_norm": 0.6277114763068243, + "learning_rate": 0.002592, + "loss": 4.5809, + "step": 864 + }, + { + "epoch": 0.00865, + "grad_norm": 0.7499269309750987, + "learning_rate": 0.002595, + "loss": 4.6024, + "step": 865 + }, + { + "epoch": 0.00866, + "grad_norm": 0.7085973518403954, + "learning_rate": 0.002598, + "loss": 4.6233, + "step": 866 + }, + { + "epoch": 0.00867, + "grad_norm": 0.605538925445329, + "learning_rate": 0.002601, + "loss": 4.6101, + "step": 867 + }, + { + "epoch": 0.00868, + "grad_norm": 0.48200776305054654, + "learning_rate": 0.002604, + "loss": 4.586, + "step": 868 + }, + { + "epoch": 0.00869, + "grad_norm": 0.5266950965425763, + "learning_rate": 0.002607, + "loss": 4.5966, + "step": 869 + }, + { + "epoch": 0.0087, + "grad_norm": 0.48953699231607295, + "learning_rate": 0.00261, + "loss": 4.5832, + "step": 870 + }, + { + "epoch": 0.00871, + "grad_norm": 0.5478274928438833, + "learning_rate": 0.002613, + "loss": 4.5992, + "step": 871 + }, + { + "epoch": 0.00872, + "grad_norm": 0.6263670684952429, + "learning_rate": 0.002616, + "loss": 4.5864, + "step": 872 + }, + { + "epoch": 0.00873, + "grad_norm": 0.6672951132542, + "learning_rate": 0.0026190000000000002, + "loss": 4.5977, + "step": 873 + }, + { + "epoch": 0.00874, + "grad_norm": 0.677096167715366, + "learning_rate": 0.002622, + "loss": 4.617, + "step": 874 + }, + { + "epoch": 0.00875, + "grad_norm": 0.6959913524482387, + "learning_rate": 0.002625, + "loss": 4.5696, + "step": 875 + }, + { + "epoch": 0.00876, + "grad_norm": 0.6682762743495083, + "learning_rate": 0.002628, + "loss": 4.5958, + "step": 876 + }, + { + "epoch": 0.00877, + "grad_norm": 0.6375266502117092, + "learning_rate": 0.002631, + "loss": 4.612, + "step": 877 + }, + { + "epoch": 0.00878, + "grad_norm": 0.7079418482290942, + "learning_rate": 0.002634, + "loss": 4.5486, + "step": 878 + }, + { + "epoch": 0.00879, + "grad_norm": 0.6282689223941402, + "learning_rate": 0.002637, + "loss": 4.578, + "step": 879 + }, + { + "epoch": 0.0088, + "grad_norm": 0.5460943252882049, + "learning_rate": 0.00264, + "loss": 4.5852, + "step": 880 + }, + { + "epoch": 0.00881, + "grad_norm": 0.5723972494402886, + "learning_rate": 0.002643, + "loss": 4.5869, + "step": 881 + }, + { + "epoch": 0.00882, + "grad_norm": 0.647818443655113, + "learning_rate": 0.002646, + "loss": 4.588, + "step": 882 + }, + { + "epoch": 0.00883, + "grad_norm": 0.8827068805337381, + "learning_rate": 0.002649, + "loss": 4.5935, + "step": 883 + }, + { + "epoch": 0.00884, + "grad_norm": 1.3000201706023533, + "learning_rate": 0.0026520000000000003, + "loss": 4.6052, + "step": 884 + }, + { + "epoch": 0.00885, + "grad_norm": 0.7527768384442359, + "learning_rate": 0.0026550000000000002, + "loss": 4.5797, + "step": 885 + }, + { + "epoch": 0.00886, + "grad_norm": 0.7313595200920677, + "learning_rate": 0.002658, + "loss": 4.6019, + "step": 886 + }, + { + "epoch": 0.00887, + "grad_norm": 0.553327654847044, + "learning_rate": 0.002661, + "loss": 4.5828, + "step": 887 + }, + { + "epoch": 0.00888, + "grad_norm": 0.6064219625843388, + "learning_rate": 0.002664, + "loss": 4.5894, + "step": 888 + }, + { + "epoch": 0.00889, + "grad_norm": 0.6392357596846293, + "learning_rate": 0.002667, + "loss": 4.5422, + "step": 889 + }, + { + "epoch": 0.0089, + "grad_norm": 0.6860123914477424, + "learning_rate": 0.00267, + "loss": 4.5989, + "step": 890 + }, + { + "epoch": 0.00891, + "grad_norm": 0.7088960904364014, + "learning_rate": 0.002673, + "loss": 4.5822, + "step": 891 + }, + { + "epoch": 0.00892, + "grad_norm": 0.7157207147763361, + "learning_rate": 0.002676, + "loss": 4.5934, + "step": 892 + }, + { + "epoch": 0.00893, + "grad_norm": 0.7412527752908875, + "learning_rate": 0.002679, + "loss": 4.5709, + "step": 893 + }, + { + "epoch": 0.00894, + "grad_norm": 0.8084836835989728, + "learning_rate": 0.002682, + "loss": 4.5639, + "step": 894 + }, + { + "epoch": 0.00895, + "grad_norm": 0.9923307111818513, + "learning_rate": 0.0026850000000000003, + "loss": 4.5864, + "step": 895 + }, + { + "epoch": 0.00896, + "grad_norm": 1.2171682577354312, + "learning_rate": 0.0026880000000000003, + "loss": 4.6057, + "step": 896 + }, + { + "epoch": 0.00897, + "grad_norm": 0.797478427208377, + "learning_rate": 0.0026910000000000002, + "loss": 4.5989, + "step": 897 + }, + { + "epoch": 0.00898, + "grad_norm": 0.7928728804117916, + "learning_rate": 0.002694, + "loss": 4.594, + "step": 898 + }, + { + "epoch": 0.00899, + "grad_norm": 0.8357403035452178, + "learning_rate": 0.002697, + "loss": 4.5983, + "step": 899 + }, + { + "epoch": 0.009, + "grad_norm": 0.8448290091163538, + "learning_rate": 0.0027, + "loss": 4.6292, + "step": 900 + }, + { + "epoch": 0.00901, + "grad_norm": 0.9488092229670547, + "learning_rate": 0.002703, + "loss": 4.5868, + "step": 901 + }, + { + "epoch": 0.00902, + "grad_norm": 0.9434404658743749, + "learning_rate": 0.002706, + "loss": 4.5999, + "step": 902 + }, + { + "epoch": 0.00903, + "grad_norm": 1.0122099567822476, + "learning_rate": 0.002709, + "loss": 4.6102, + "step": 903 + }, + { + "epoch": 0.00904, + "grad_norm": 0.9358691681287052, + "learning_rate": 0.002712, + "loss": 4.5848, + "step": 904 + }, + { + "epoch": 0.00905, + "grad_norm": 0.8321510442485943, + "learning_rate": 0.002715, + "loss": 4.5984, + "step": 905 + }, + { + "epoch": 0.00906, + "grad_norm": 0.8914473393947665, + "learning_rate": 0.002718, + "loss": 4.6112, + "step": 906 + }, + { + "epoch": 0.00907, + "grad_norm": 0.9883982303638487, + "learning_rate": 0.0027210000000000003, + "loss": 4.6386, + "step": 907 + }, + { + "epoch": 0.00908, + "grad_norm": 0.86073203349026, + "learning_rate": 0.0027240000000000003, + "loss": 4.6116, + "step": 908 + }, + { + "epoch": 0.00909, + "grad_norm": 0.7773747412069614, + "learning_rate": 0.0027270000000000003, + "loss": 4.6163, + "step": 909 + }, + { + "epoch": 0.0091, + "grad_norm": 0.7370585718531062, + "learning_rate": 0.0027300000000000002, + "loss": 4.6234, + "step": 910 + }, + { + "epoch": 0.00911, + "grad_norm": 0.6906269071273593, + "learning_rate": 0.002733, + "loss": 4.5785, + "step": 911 + }, + { + "epoch": 0.00912, + "grad_norm": 0.6578032292778252, + "learning_rate": 0.002736, + "loss": 4.5778, + "step": 912 + }, + { + "epoch": 0.00913, + "grad_norm": 0.6528626059582382, + "learning_rate": 0.002739, + "loss": 4.5704, + "step": 913 + }, + { + "epoch": 0.00914, + "grad_norm": 0.599731896856576, + "learning_rate": 0.002742, + "loss": 4.595, + "step": 914 + }, + { + "epoch": 0.00915, + "grad_norm": 0.5922054086035364, + "learning_rate": 0.002745, + "loss": 4.5555, + "step": 915 + }, + { + "epoch": 0.00916, + "grad_norm": 0.528646140228931, + "learning_rate": 0.002748, + "loss": 4.5304, + "step": 916 + }, + { + "epoch": 0.00917, + "grad_norm": 0.5305158198561161, + "learning_rate": 0.002751, + "loss": 4.5419, + "step": 917 + }, + { + "epoch": 0.00918, + "grad_norm": 0.4736382884122071, + "learning_rate": 0.0027540000000000004, + "loss": 4.5569, + "step": 918 + }, + { + "epoch": 0.00919, + "grad_norm": 0.45838817911808083, + "learning_rate": 0.0027570000000000003, + "loss": 4.5357, + "step": 919 + }, + { + "epoch": 0.0092, + "grad_norm": 0.4361472184478695, + "learning_rate": 0.0027600000000000003, + "loss": 4.5493, + "step": 920 + }, + { + "epoch": 0.00921, + "grad_norm": 0.43931126757575867, + "learning_rate": 0.0027630000000000003, + "loss": 4.5336, + "step": 921 + }, + { + "epoch": 0.00922, + "grad_norm": 0.46920285200840567, + "learning_rate": 0.0027660000000000002, + "loss": 4.5412, + "step": 922 + }, + { + "epoch": 0.00923, + "grad_norm": 0.5996209745093354, + "learning_rate": 0.002769, + "loss": 4.5218, + "step": 923 + }, + { + "epoch": 0.00924, + "grad_norm": 0.702528530852281, + "learning_rate": 0.002772, + "loss": 4.5564, + "step": 924 + }, + { + "epoch": 0.00925, + "grad_norm": 0.6651779892024297, + "learning_rate": 0.002775, + "loss": 4.5288, + "step": 925 + }, + { + "epoch": 0.00926, + "grad_norm": 0.5723855300218565, + "learning_rate": 0.002778, + "loss": 4.5426, + "step": 926 + }, + { + "epoch": 0.00927, + "grad_norm": 0.6080148945846373, + "learning_rate": 0.002781, + "loss": 4.5043, + "step": 927 + }, + { + "epoch": 0.00928, + "grad_norm": 0.6957035289314476, + "learning_rate": 0.002784, + "loss": 4.5502, + "step": 928 + }, + { + "epoch": 0.00929, + "grad_norm": 0.7366634335117053, + "learning_rate": 0.0027870000000000004, + "loss": 4.5403, + "step": 929 + }, + { + "epoch": 0.0093, + "grad_norm": 0.7796102692352119, + "learning_rate": 0.0027900000000000004, + "loss": 4.5265, + "step": 930 + }, + { + "epoch": 0.00931, + "grad_norm": 0.7116909602884058, + "learning_rate": 0.0027930000000000003, + "loss": 4.5333, + "step": 931 + }, + { + "epoch": 0.00932, + "grad_norm": 0.8319854024158545, + "learning_rate": 0.0027960000000000003, + "loss": 4.5481, + "step": 932 + }, + { + "epoch": 0.00933, + "grad_norm": 0.8001418279766108, + "learning_rate": 0.0027990000000000003, + "loss": 4.5286, + "step": 933 + }, + { + "epoch": 0.00934, + "grad_norm": 0.6519619232143173, + "learning_rate": 0.0028020000000000002, + "loss": 4.5422, + "step": 934 + }, + { + "epoch": 0.00935, + "grad_norm": 0.7580737482550882, + "learning_rate": 0.002805, + "loss": 4.5559, + "step": 935 + }, + { + "epoch": 0.00936, + "grad_norm": 0.8499038627491867, + "learning_rate": 0.002808, + "loss": 4.5679, + "step": 936 + }, + { + "epoch": 0.00937, + "grad_norm": 0.7597302495348821, + "learning_rate": 0.002811, + "loss": 4.5708, + "step": 937 + }, + { + "epoch": 0.00938, + "grad_norm": 0.9878821641788273, + "learning_rate": 0.002814, + "loss": 4.5609, + "step": 938 + }, + { + "epoch": 0.00939, + "grad_norm": 0.9691729918808772, + "learning_rate": 0.002817, + "loss": 4.563, + "step": 939 + }, + { + "epoch": 0.0094, + "grad_norm": 0.8937843559478598, + "learning_rate": 0.00282, + "loss": 4.56, + "step": 940 + }, + { + "epoch": 0.00941, + "grad_norm": 0.9477839045288606, + "learning_rate": 0.002823, + "loss": 4.5409, + "step": 941 + }, + { + "epoch": 0.00942, + "grad_norm": 1.088560613057821, + "learning_rate": 0.002826, + "loss": 4.5819, + "step": 942 + }, + { + "epoch": 0.00943, + "grad_norm": 0.8020128186220904, + "learning_rate": 0.002829, + "loss": 4.556, + "step": 943 + }, + { + "epoch": 0.00944, + "grad_norm": 0.7970499406732843, + "learning_rate": 0.002832, + "loss": 4.5652, + "step": 944 + }, + { + "epoch": 0.00945, + "grad_norm": 0.760430287307007, + "learning_rate": 0.002835, + "loss": 4.567, + "step": 945 + }, + { + "epoch": 0.00946, + "grad_norm": 0.8410168172764453, + "learning_rate": 0.002838, + "loss": 4.5808, + "step": 946 + }, + { + "epoch": 0.00947, + "grad_norm": 0.8502364092306604, + "learning_rate": 0.0028409999999999998, + "loss": 4.5581, + "step": 947 + }, + { + "epoch": 0.00948, + "grad_norm": 0.7534324730199542, + "learning_rate": 0.0028439999999999997, + "loss": 4.533, + "step": 948 + }, + { + "epoch": 0.00949, + "grad_norm": 0.8075715283027973, + "learning_rate": 0.002847, + "loss": 4.5789, + "step": 949 + }, + { + "epoch": 0.0095, + "grad_norm": 0.8790685187514339, + "learning_rate": 0.00285, + "loss": 4.5764, + "step": 950 + }, + { + "epoch": 0.00951, + "grad_norm": 0.8527621336415785, + "learning_rate": 0.002853, + "loss": 4.552, + "step": 951 + }, + { + "epoch": 0.00952, + "grad_norm": 0.793648162843131, + "learning_rate": 0.002856, + "loss": 4.5771, + "step": 952 + }, + { + "epoch": 0.00953, + "grad_norm": 0.7100823051409002, + "learning_rate": 0.002859, + "loss": 4.5151, + "step": 953 + }, + { + "epoch": 0.00954, + "grad_norm": 0.776086010454581, + "learning_rate": 0.002862, + "loss": 4.5748, + "step": 954 + }, + { + "epoch": 0.00955, + "grad_norm": 0.7357834745016256, + "learning_rate": 0.002865, + "loss": 4.5651, + "step": 955 + }, + { + "epoch": 0.00956, + "grad_norm": 0.6871788604084053, + "learning_rate": 0.002868, + "loss": 4.5378, + "step": 956 + }, + { + "epoch": 0.00957, + "grad_norm": 0.6293704920093642, + "learning_rate": 0.002871, + "loss": 4.5585, + "step": 957 + }, + { + "epoch": 0.00958, + "grad_norm": 0.6933721545151298, + "learning_rate": 0.002874, + "loss": 4.5402, + "step": 958 + }, + { + "epoch": 0.00959, + "grad_norm": 0.6216290945191316, + "learning_rate": 0.002877, + "loss": 4.5294, + "step": 959 + }, + { + "epoch": 0.0096, + "grad_norm": 0.44090482568449035, + "learning_rate": 0.0028799999999999997, + "loss": 4.5205, + "step": 960 + }, + { + "epoch": 0.00961, + "grad_norm": 0.5026549244936088, + "learning_rate": 0.002883, + "loss": 4.4973, + "step": 961 + }, + { + "epoch": 0.00962, + "grad_norm": 0.46550744372429714, + "learning_rate": 0.002886, + "loss": 4.5199, + "step": 962 + }, + { + "epoch": 0.00963, + "grad_norm": 0.4817462883709995, + "learning_rate": 0.002889, + "loss": 4.5204, + "step": 963 + }, + { + "epoch": 0.00964, + "grad_norm": 0.5021989893794454, + "learning_rate": 0.002892, + "loss": 4.5105, + "step": 964 + }, + { + "epoch": 0.00965, + "grad_norm": 0.6331237702649058, + "learning_rate": 0.002895, + "loss": 4.4888, + "step": 965 + }, + { + "epoch": 0.00966, + "grad_norm": 0.7186463225121739, + "learning_rate": 0.002898, + "loss": 4.5122, + "step": 966 + }, + { + "epoch": 0.00967, + "grad_norm": 0.835541291398658, + "learning_rate": 0.002901, + "loss": 4.5497, + "step": 967 + }, + { + "epoch": 0.00968, + "grad_norm": 0.7770950591910699, + "learning_rate": 0.002904, + "loss": 4.5246, + "step": 968 + }, + { + "epoch": 0.00969, + "grad_norm": 0.6421972738290654, + "learning_rate": 0.002907, + "loss": 4.5465, + "step": 969 + }, + { + "epoch": 0.0097, + "grad_norm": 0.6170493579190435, + "learning_rate": 0.00291, + "loss": 4.5014, + "step": 970 + }, + { + "epoch": 0.00971, + "grad_norm": 0.671893763395282, + "learning_rate": 0.002913, + "loss": 4.5134, + "step": 971 + }, + { + "epoch": 0.00972, + "grad_norm": 0.5384349268117217, + "learning_rate": 0.002916, + "loss": 4.51, + "step": 972 + }, + { + "epoch": 0.00973, + "grad_norm": 0.6570052261370841, + "learning_rate": 0.002919, + "loss": 4.5075, + "step": 973 + }, + { + "epoch": 0.00974, + "grad_norm": 0.6469437996214488, + "learning_rate": 0.002922, + "loss": 4.5042, + "step": 974 + }, + { + "epoch": 0.00975, + "grad_norm": 0.5139434995269291, + "learning_rate": 0.002925, + "loss": 4.5141, + "step": 975 + }, + { + "epoch": 0.00976, + "grad_norm": 0.540350404123188, + "learning_rate": 0.002928, + "loss": 4.4984, + "step": 976 + }, + { + "epoch": 0.00977, + "grad_norm": 0.5640158884340003, + "learning_rate": 0.002931, + "loss": 4.5359, + "step": 977 + }, + { + "epoch": 0.00978, + "grad_norm": 0.5471232322596488, + "learning_rate": 0.002934, + "loss": 4.5069, + "step": 978 + }, + { + "epoch": 0.00979, + "grad_norm": 0.6391692840311302, + "learning_rate": 0.002937, + "loss": 4.4972, + "step": 979 + }, + { + "epoch": 0.0098, + "grad_norm": 0.7065424241899814, + "learning_rate": 0.00294, + "loss": 4.4963, + "step": 980 + }, + { + "epoch": 0.00981, + "grad_norm": 0.6903724510426201, + "learning_rate": 0.002943, + "loss": 4.5078, + "step": 981 + }, + { + "epoch": 0.00982, + "grad_norm": 0.653302049548968, + "learning_rate": 0.002946, + "loss": 4.4991, + "step": 982 + }, + { + "epoch": 0.00983, + "grad_norm": 0.7786836590229197, + "learning_rate": 0.0029490000000000002, + "loss": 4.5061, + "step": 983 + }, + { + "epoch": 0.00984, + "grad_norm": 0.8944946184941699, + "learning_rate": 0.002952, + "loss": 4.5043, + "step": 984 + }, + { + "epoch": 0.00985, + "grad_norm": 0.9434362337974719, + "learning_rate": 0.002955, + "loss": 4.5335, + "step": 985 + }, + { + "epoch": 0.00986, + "grad_norm": 0.9312010498282314, + "learning_rate": 0.002958, + "loss": 4.4908, + "step": 986 + }, + { + "epoch": 0.00987, + "grad_norm": 0.7600059379317153, + "learning_rate": 0.002961, + "loss": 4.5038, + "step": 987 + }, + { + "epoch": 0.00988, + "grad_norm": 0.6949983731085813, + "learning_rate": 0.002964, + "loss": 4.533, + "step": 988 + }, + { + "epoch": 0.00989, + "grad_norm": 0.6958207764457225, + "learning_rate": 0.002967, + "loss": 4.4744, + "step": 989 + }, + { + "epoch": 0.0099, + "grad_norm": 0.8314159874496235, + "learning_rate": 0.00297, + "loss": 4.5218, + "step": 990 + }, + { + "epoch": 0.00991, + "grad_norm": 0.90109982550625, + "learning_rate": 0.002973, + "loss": 4.5224, + "step": 991 + }, + { + "epoch": 0.00992, + "grad_norm": 0.951208720196979, + "learning_rate": 0.002976, + "loss": 4.5281, + "step": 992 + }, + { + "epoch": 0.00993, + "grad_norm": 1.034654706401041, + "learning_rate": 0.002979, + "loss": 4.539, + "step": 993 + }, + { + "epoch": 0.00994, + "grad_norm": 1.0078120294806783, + "learning_rate": 0.002982, + "loss": 4.5459, + "step": 994 + }, + { + "epoch": 0.00995, + "grad_norm": 0.9905056664043064, + "learning_rate": 0.0029850000000000002, + "loss": 4.5461, + "step": 995 + }, + { + "epoch": 0.00996, + "grad_norm": 1.1603429750232241, + "learning_rate": 0.002988, + "loss": 4.5605, + "step": 996 + }, + { + "epoch": 0.00997, + "grad_norm": 0.8595677567729485, + "learning_rate": 0.002991, + "loss": 4.5366, + "step": 997 + }, + { + "epoch": 0.00998, + "grad_norm": 0.9672785170741734, + "learning_rate": 0.002994, + "loss": 4.5467, + "step": 998 + }, + { + "epoch": 0.00999, + "grad_norm": 0.8855476122040005, + "learning_rate": 0.002997, + "loss": 4.5306, + "step": 999 + }, + { + "epoch": 0.01, + "grad_norm": 0.7479993027173912, + "learning_rate": 0.003, + "loss": 4.5436, + "step": 1000 + }, + { + "epoch": 0.01001, + "grad_norm": 0.7121555496393158, + "learning_rate": 0.003, + "loss": 4.5401, + "step": 1001 + }, + { + "epoch": 0.01002, + "grad_norm": 0.72505396659877, + "learning_rate": 0.003, + "loss": 4.5099, + "step": 1002 + }, + { + "epoch": 0.01003, + "grad_norm": 0.7482409624375888, + "learning_rate": 0.003, + "loss": 4.5346, + "step": 1003 + }, + { + "epoch": 0.01004, + "grad_norm": 0.5922239695560899, + "learning_rate": 0.003, + "loss": 4.513, + "step": 1004 + }, + { + "epoch": 0.01005, + "grad_norm": 0.7163386651042714, + "learning_rate": 0.003, + "loss": 4.4897, + "step": 1005 + }, + { + "epoch": 0.01006, + "grad_norm": 0.7664341729228217, + "learning_rate": 0.003, + "loss": 4.4943, + "step": 1006 + }, + { + "epoch": 0.01007, + "grad_norm": 0.6325498197242884, + "learning_rate": 0.003, + "loss": 4.5316, + "step": 1007 + }, + { + "epoch": 0.01008, + "grad_norm": 0.5334181570982156, + "learning_rate": 0.003, + "loss": 4.5325, + "step": 1008 + }, + { + "epoch": 0.01009, + "grad_norm": 0.4756687788935663, + "learning_rate": 0.003, + "loss": 4.4998, + "step": 1009 + }, + { + "epoch": 0.0101, + "grad_norm": 0.44081077367096977, + "learning_rate": 0.003, + "loss": 4.5025, + "step": 1010 + }, + { + "epoch": 0.01011, + "grad_norm": 0.4706704982402646, + "learning_rate": 0.003, + "loss": 4.5082, + "step": 1011 + }, + { + "epoch": 0.01012, + "grad_norm": 0.47223639219621605, + "learning_rate": 0.003, + "loss": 4.492, + "step": 1012 + }, + { + "epoch": 0.01013, + "grad_norm": 0.4611333961201937, + "learning_rate": 0.003, + "loss": 4.5135, + "step": 1013 + }, + { + "epoch": 0.01014, + "grad_norm": 0.36636268932008986, + "learning_rate": 0.003, + "loss": 4.4528, + "step": 1014 + }, + { + "epoch": 0.01015, + "grad_norm": 0.4484711002206511, + "learning_rate": 0.003, + "loss": 4.4948, + "step": 1015 + }, + { + "epoch": 0.01016, + "grad_norm": 0.553209882568818, + "learning_rate": 0.003, + "loss": 4.5082, + "step": 1016 + }, + { + "epoch": 0.01017, + "grad_norm": 0.8898002239967567, + "learning_rate": 0.003, + "loss": 4.4579, + "step": 1017 + }, + { + "epoch": 0.01018, + "grad_norm": 0.9938726896514534, + "learning_rate": 0.003, + "loss": 4.5557, + "step": 1018 + }, + { + "epoch": 0.01019, + "grad_norm": 0.7008138530645361, + "learning_rate": 0.003, + "loss": 4.5003, + "step": 1019 + }, + { + "epoch": 0.0102, + "grad_norm": 1.006685411589756, + "learning_rate": 0.003, + "loss": 4.5244, + "step": 1020 + }, + { + "epoch": 0.01021, + "grad_norm": 0.6295405897908188, + "learning_rate": 0.003, + "loss": 4.4915, + "step": 1021 + }, + { + "epoch": 0.01022, + "grad_norm": 0.6613412501940705, + "learning_rate": 0.003, + "loss": 4.5063, + "step": 1022 + }, + { + "epoch": 0.01023, + "grad_norm": 0.6284055579476905, + "learning_rate": 0.003, + "loss": 4.4914, + "step": 1023 + }, + { + "epoch": 0.01024, + "grad_norm": 0.6735486877543648, + "learning_rate": 0.003, + "loss": 4.4573, + "step": 1024 + }, + { + "epoch": 0.01025, + "grad_norm": 0.6253815453908939, + "learning_rate": 0.003, + "loss": 4.4832, + "step": 1025 + }, + { + "epoch": 0.01026, + "grad_norm": 0.5932239048767913, + "learning_rate": 0.003, + "loss": 4.4735, + "step": 1026 + }, + { + "epoch": 0.01027, + "grad_norm": 0.6318859057234201, + "learning_rate": 0.003, + "loss": 4.4952, + "step": 1027 + }, + { + "epoch": 0.01028, + "grad_norm": 0.6588163663127304, + "learning_rate": 0.003, + "loss": 4.4763, + "step": 1028 + }, + { + "epoch": 0.01029, + "grad_norm": 0.55721560335989, + "learning_rate": 0.003, + "loss": 4.4524, + "step": 1029 + }, + { + "epoch": 0.0103, + "grad_norm": 0.5468199872259095, + "learning_rate": 0.003, + "loss": 4.4809, + "step": 1030 + }, + { + "epoch": 0.01031, + "grad_norm": 0.5601179932704955, + "learning_rate": 0.003, + "loss": 4.4934, + "step": 1031 + }, + { + "epoch": 0.01032, + "grad_norm": 0.5514443283118416, + "learning_rate": 0.003, + "loss": 4.464, + "step": 1032 + }, + { + "epoch": 0.01033, + "grad_norm": 0.49777794102419176, + "learning_rate": 0.003, + "loss": 4.496, + "step": 1033 + }, + { + "epoch": 0.01034, + "grad_norm": 0.5163173300909295, + "learning_rate": 0.003, + "loss": 4.4625, + "step": 1034 + }, + { + "epoch": 0.01035, + "grad_norm": 0.4798196088286999, + "learning_rate": 0.003, + "loss": 4.4573, + "step": 1035 + }, + { + "epoch": 0.01036, + "grad_norm": 0.3958499097440651, + "learning_rate": 0.003, + "loss": 4.4459, + "step": 1036 + }, + { + "epoch": 0.01037, + "grad_norm": 0.3643003447603396, + "learning_rate": 0.003, + "loss": 4.4698, + "step": 1037 + }, + { + "epoch": 0.01038, + "grad_norm": 0.40989678679265595, + "learning_rate": 0.003, + "loss": 4.4667, + "step": 1038 + }, + { + "epoch": 0.01039, + "grad_norm": 0.3544910886173235, + "learning_rate": 0.003, + "loss": 4.4556, + "step": 1039 + }, + { + "epoch": 0.0104, + "grad_norm": 0.3719879786284374, + "learning_rate": 0.003, + "loss": 4.4667, + "step": 1040 + }, + { + "epoch": 0.01041, + "grad_norm": 0.3770068025965155, + "learning_rate": 0.003, + "loss": 4.4469, + "step": 1041 + }, + { + "epoch": 0.01042, + "grad_norm": 0.4469030721524434, + "learning_rate": 0.003, + "loss": 4.4752, + "step": 1042 + }, + { + "epoch": 0.01043, + "grad_norm": 0.7272268008417163, + "learning_rate": 0.003, + "loss": 4.4274, + "step": 1043 + }, + { + "epoch": 0.01044, + "grad_norm": 0.8992907359513324, + "learning_rate": 0.003, + "loss": 4.4949, + "step": 1044 + }, + { + "epoch": 0.01045, + "grad_norm": 0.8302310255488462, + "learning_rate": 0.003, + "loss": 4.4953, + "step": 1045 + }, + { + "epoch": 0.01046, + "grad_norm": 0.9975061123002056, + "learning_rate": 0.003, + "loss": 4.4816, + "step": 1046 + }, + { + "epoch": 0.01047, + "grad_norm": 0.8067658442263902, + "learning_rate": 0.003, + "loss": 4.4954, + "step": 1047 + }, + { + "epoch": 0.01048, + "grad_norm": 0.7572637033913173, + "learning_rate": 0.003, + "loss": 4.4741, + "step": 1048 + }, + { + "epoch": 0.01049, + "grad_norm": 0.8994748129621607, + "learning_rate": 0.003, + "loss": 4.4647, + "step": 1049 + }, + { + "epoch": 0.0105, + "grad_norm": 0.7962039985766673, + "learning_rate": 0.003, + "loss": 4.5006, + "step": 1050 + }, + { + "epoch": 0.01051, + "grad_norm": 0.8207643682537104, + "learning_rate": 0.003, + "loss": 4.4811, + "step": 1051 + }, + { + "epoch": 0.01052, + "grad_norm": 0.7458558375660994, + "learning_rate": 0.003, + "loss": 4.4881, + "step": 1052 + }, + { + "epoch": 0.01053, + "grad_norm": 0.6759075645027314, + "learning_rate": 0.003, + "loss": 4.4946, + "step": 1053 + }, + { + "epoch": 0.01054, + "grad_norm": 0.6790074754239308, + "learning_rate": 0.003, + "loss": 4.4852, + "step": 1054 + }, + { + "epoch": 0.01055, + "grad_norm": 0.6920845821894167, + "learning_rate": 0.003, + "loss": 4.4644, + "step": 1055 + }, + { + "epoch": 0.01056, + "grad_norm": 0.5219544124128492, + "learning_rate": 0.003, + "loss": 4.4569, + "step": 1056 + }, + { + "epoch": 0.01057, + "grad_norm": 0.6021767240952963, + "learning_rate": 0.003, + "loss": 4.4651, + "step": 1057 + }, + { + "epoch": 0.01058, + "grad_norm": 0.5219741420101096, + "learning_rate": 0.003, + "loss": 4.4552, + "step": 1058 + }, + { + "epoch": 0.01059, + "grad_norm": 0.509217286199775, + "learning_rate": 0.003, + "loss": 4.4557, + "step": 1059 + }, + { + "epoch": 0.0106, + "grad_norm": 0.6119681984085901, + "learning_rate": 0.003, + "loss": 4.4506, + "step": 1060 + }, + { + "epoch": 0.01061, + "grad_norm": 0.6018802462486371, + "learning_rate": 0.003, + "loss": 4.4451, + "step": 1061 + }, + { + "epoch": 0.01062, + "grad_norm": 0.6072220138777784, + "learning_rate": 0.003, + "loss": 4.4398, + "step": 1062 + }, + { + "epoch": 0.01063, + "grad_norm": 0.5793152524763621, + "learning_rate": 0.003, + "loss": 4.4372, + "step": 1063 + }, + { + "epoch": 0.01064, + "grad_norm": 0.740973950389461, + "learning_rate": 0.003, + "loss": 4.4576, + "step": 1064 + }, + { + "epoch": 0.01065, + "grad_norm": 0.8998946238241571, + "learning_rate": 0.003, + "loss": 4.4756, + "step": 1065 + }, + { + "epoch": 0.01066, + "grad_norm": 0.889057289027363, + "learning_rate": 0.003, + "loss": 4.4727, + "step": 1066 + }, + { + "epoch": 0.01067, + "grad_norm": 0.9442468447502006, + "learning_rate": 0.003, + "loss": 4.4902, + "step": 1067 + }, + { + "epoch": 0.01068, + "grad_norm": 1.018364097076677, + "learning_rate": 0.003, + "loss": 4.4857, + "step": 1068 + }, + { + "epoch": 0.01069, + "grad_norm": 0.9799434684826753, + "learning_rate": 0.003, + "loss": 4.4924, + "step": 1069 + }, + { + "epoch": 0.0107, + "grad_norm": 0.8992499083099482, + "learning_rate": 0.003, + "loss": 4.5446, + "step": 1070 + }, + { + "epoch": 0.01071, + "grad_norm": 0.8735440073637682, + "learning_rate": 0.003, + "loss": 4.4651, + "step": 1071 + }, + { + "epoch": 0.01072, + "grad_norm": 0.8443267804596191, + "learning_rate": 0.003, + "loss": 4.5031, + "step": 1072 + }, + { + "epoch": 0.01073, + "grad_norm": 0.7866639405001601, + "learning_rate": 0.003, + "loss": 4.4993, + "step": 1073 + }, + { + "epoch": 0.01074, + "grad_norm": 0.7412592488158714, + "learning_rate": 0.003, + "loss": 4.4612, + "step": 1074 + }, + { + "epoch": 0.01075, + "grad_norm": 0.7055777296919327, + "learning_rate": 0.003, + "loss": 4.4949, + "step": 1075 + }, + { + "epoch": 0.01076, + "grad_norm": 0.6602103525167045, + "learning_rate": 0.003, + "loss": 4.4903, + "step": 1076 + }, + { + "epoch": 0.01077, + "grad_norm": 0.6221585201435259, + "learning_rate": 0.003, + "loss": 4.4791, + "step": 1077 + }, + { + "epoch": 0.01078, + "grad_norm": 0.6312740380808073, + "learning_rate": 0.003, + "loss": 4.4373, + "step": 1078 + }, + { + "epoch": 0.01079, + "grad_norm": 0.5465751942660528, + "learning_rate": 0.003, + "loss": 4.4641, + "step": 1079 + }, + { + "epoch": 0.0108, + "grad_norm": 0.5040258102852178, + "learning_rate": 0.003, + "loss": 4.4392, + "step": 1080 + }, + { + "epoch": 0.01081, + "grad_norm": 0.4557353176092044, + "learning_rate": 0.003, + "loss": 4.4439, + "step": 1081 + }, + { + "epoch": 0.01082, + "grad_norm": 0.43657211030445375, + "learning_rate": 0.003, + "loss": 4.4598, + "step": 1082 + }, + { + "epoch": 0.01083, + "grad_norm": 0.37468227521199704, + "learning_rate": 0.003, + "loss": 4.4419, + "step": 1083 + }, + { + "epoch": 0.01084, + "grad_norm": 0.3702365616744669, + "learning_rate": 0.003, + "loss": 4.4465, + "step": 1084 + }, + { + "epoch": 0.01085, + "grad_norm": 0.3248887668649726, + "learning_rate": 0.003, + "loss": 4.4323, + "step": 1085 + }, + { + "epoch": 0.01086, + "grad_norm": 0.28646275942866467, + "learning_rate": 0.003, + "loss": 4.4059, + "step": 1086 + }, + { + "epoch": 0.01087, + "grad_norm": 0.3109295903469056, + "learning_rate": 0.003, + "loss": 4.4215, + "step": 1087 + }, + { + "epoch": 0.01088, + "grad_norm": 0.31305777007935864, + "learning_rate": 0.003, + "loss": 4.454, + "step": 1088 + }, + { + "epoch": 0.01089, + "grad_norm": 0.3202194647994524, + "learning_rate": 0.003, + "loss": 4.423, + "step": 1089 + }, + { + "epoch": 0.0109, + "grad_norm": 0.3546167101166988, + "learning_rate": 0.003, + "loss": 4.4466, + "step": 1090 + }, + { + "epoch": 0.01091, + "grad_norm": 0.40979643182380626, + "learning_rate": 0.003, + "loss": 4.4162, + "step": 1091 + }, + { + "epoch": 0.01092, + "grad_norm": 0.44852044367674554, + "learning_rate": 0.003, + "loss": 4.4131, + "step": 1092 + }, + { + "epoch": 0.01093, + "grad_norm": 0.5395428853759895, + "learning_rate": 0.003, + "loss": 4.4351, + "step": 1093 + }, + { + "epoch": 0.01094, + "grad_norm": 0.7419472173607777, + "learning_rate": 0.003, + "loss": 4.4344, + "step": 1094 + }, + { + "epoch": 0.01095, + "grad_norm": 1.3268112301670232, + "learning_rate": 0.003, + "loss": 4.4875, + "step": 1095 + }, + { + "epoch": 0.01096, + "grad_norm": 0.8107867356113418, + "learning_rate": 0.003, + "loss": 4.4576, + "step": 1096 + }, + { + "epoch": 0.01097, + "grad_norm": 0.6635063888056452, + "learning_rate": 0.003, + "loss": 4.4388, + "step": 1097 + }, + { + "epoch": 0.01098, + "grad_norm": 0.9662363707419442, + "learning_rate": 0.003, + "loss": 4.474, + "step": 1098 + }, + { + "epoch": 0.01099, + "grad_norm": 0.7696101380141932, + "learning_rate": 0.003, + "loss": 4.4405, + "step": 1099 + }, + { + "epoch": 0.011, + "grad_norm": 0.7450236895824817, + "learning_rate": 0.003, + "loss": 4.4451, + "step": 1100 + }, + { + "epoch": 0.01101, + "grad_norm": 0.7254968126029079, + "learning_rate": 0.003, + "loss": 4.4513, + "step": 1101 + }, + { + "epoch": 0.01102, + "grad_norm": 0.7537628081039499, + "learning_rate": 0.003, + "loss": 4.4588, + "step": 1102 + }, + { + "epoch": 0.01103, + "grad_norm": 0.7123173358553468, + "learning_rate": 0.003, + "loss": 4.4551, + "step": 1103 + }, + { + "epoch": 0.01104, + "grad_norm": 0.6033148725235706, + "learning_rate": 0.003, + "loss": 4.4384, + "step": 1104 + }, + { + "epoch": 0.01105, + "grad_norm": 0.7223359575565769, + "learning_rate": 0.003, + "loss": 4.4643, + "step": 1105 + }, + { + "epoch": 0.01106, + "grad_norm": 0.6749769688136343, + "learning_rate": 0.003, + "loss": 4.4355, + "step": 1106 + }, + { + "epoch": 0.01107, + "grad_norm": 0.6519921729558845, + "learning_rate": 0.003, + "loss": 4.443, + "step": 1107 + }, + { + "epoch": 0.01108, + "grad_norm": 0.6442390747026363, + "learning_rate": 0.003, + "loss": 4.4155, + "step": 1108 + }, + { + "epoch": 0.01109, + "grad_norm": 0.5210761074063659, + "learning_rate": 0.003, + "loss": 4.4258, + "step": 1109 + }, + { + "epoch": 0.0111, + "grad_norm": 0.49112961811162964, + "learning_rate": 0.003, + "loss": 4.4439, + "step": 1110 + }, + { + "epoch": 0.01111, + "grad_norm": 0.5154689325771941, + "learning_rate": 0.003, + "loss": 4.4228, + "step": 1111 + }, + { + "epoch": 0.01112, + "grad_norm": 0.5147734027357755, + "learning_rate": 0.003, + "loss": 4.4383, + "step": 1112 + }, + { + "epoch": 0.01113, + "grad_norm": 0.4905886669705696, + "learning_rate": 0.003, + "loss": 4.4046, + "step": 1113 + }, + { + "epoch": 0.01114, + "grad_norm": 0.47555859430135783, + "learning_rate": 0.003, + "loss": 4.4166, + "step": 1114 + }, + { + "epoch": 0.01115, + "grad_norm": 0.42768329796719207, + "learning_rate": 0.003, + "loss": 4.4369, + "step": 1115 + }, + { + "epoch": 0.01116, + "grad_norm": 0.4281709726428523, + "learning_rate": 0.003, + "loss": 4.4286, + "step": 1116 + }, + { + "epoch": 0.01117, + "grad_norm": 0.4124788023594352, + "learning_rate": 0.003, + "loss": 4.414, + "step": 1117 + }, + { + "epoch": 0.01118, + "grad_norm": 0.4531459371875401, + "learning_rate": 0.003, + "loss": 4.415, + "step": 1118 + }, + { + "epoch": 0.01119, + "grad_norm": 0.5115410769007107, + "learning_rate": 0.003, + "loss": 4.4247, + "step": 1119 + }, + { + "epoch": 0.0112, + "grad_norm": 0.6598099600325713, + "learning_rate": 0.003, + "loss": 4.4115, + "step": 1120 + }, + { + "epoch": 0.01121, + "grad_norm": 0.7298362819040065, + "learning_rate": 0.003, + "loss": 4.4356, + "step": 1121 + }, + { + "epoch": 0.01122, + "grad_norm": 0.6383918217981295, + "learning_rate": 0.003, + "loss": 4.4021, + "step": 1122 + }, + { + "epoch": 0.01123, + "grad_norm": 0.7157274101955465, + "learning_rate": 0.003, + "loss": 4.3942, + "step": 1123 + }, + { + "epoch": 0.01124, + "grad_norm": 0.6763302773515354, + "learning_rate": 0.003, + "loss": 4.4444, + "step": 1124 + }, + { + "epoch": 0.01125, + "grad_norm": 0.5890253028490829, + "learning_rate": 0.003, + "loss": 4.4062, + "step": 1125 + }, + { + "epoch": 0.01126, + "grad_norm": 0.6745452077137704, + "learning_rate": 0.003, + "loss": 4.4204, + "step": 1126 + }, + { + "epoch": 0.01127, + "grad_norm": 0.6212265837956056, + "learning_rate": 0.003, + "loss": 4.446, + "step": 1127 + }, + { + "epoch": 0.01128, + "grad_norm": 0.6579375902442356, + "learning_rate": 0.003, + "loss": 4.4285, + "step": 1128 + }, + { + "epoch": 0.01129, + "grad_norm": 0.7038241088934594, + "learning_rate": 0.003, + "loss": 4.4197, + "step": 1129 + }, + { + "epoch": 0.0113, + "grad_norm": 0.8537912734504166, + "learning_rate": 0.003, + "loss": 4.4099, + "step": 1130 + }, + { + "epoch": 0.01131, + "grad_norm": 0.6676066708341881, + "learning_rate": 0.003, + "loss": 4.4184, + "step": 1131 + }, + { + "epoch": 0.01132, + "grad_norm": 0.6262944206433371, + "learning_rate": 0.003, + "loss": 4.4418, + "step": 1132 + }, + { + "epoch": 0.01133, + "grad_norm": 0.7096964978520276, + "learning_rate": 0.003, + "loss": 4.4077, + "step": 1133 + }, + { + "epoch": 0.01134, + "grad_norm": 0.7063046600590702, + "learning_rate": 0.003, + "loss": 4.4219, + "step": 1134 + }, + { + "epoch": 0.01135, + "grad_norm": 0.7389408975678705, + "learning_rate": 0.003, + "loss": 4.4511, + "step": 1135 + }, + { + "epoch": 0.01136, + "grad_norm": 0.902932943298548, + "learning_rate": 0.003, + "loss": 4.4334, + "step": 1136 + }, + { + "epoch": 0.01137, + "grad_norm": 0.8104098050855381, + "learning_rate": 0.003, + "loss": 4.4615, + "step": 1137 + }, + { + "epoch": 0.01138, + "grad_norm": 0.6304382373186352, + "learning_rate": 0.003, + "loss": 4.4061, + "step": 1138 + }, + { + "epoch": 0.01139, + "grad_norm": 0.6573985975094263, + "learning_rate": 0.003, + "loss": 4.4174, + "step": 1139 + }, + { + "epoch": 0.0114, + "grad_norm": 0.6243548176788066, + "learning_rate": 0.003, + "loss": 4.4015, + "step": 1140 + }, + { + "epoch": 0.01141, + "grad_norm": 0.6941761654842915, + "learning_rate": 0.003, + "loss": 4.4278, + "step": 1141 + }, + { + "epoch": 0.01142, + "grad_norm": 0.6422841400280502, + "learning_rate": 0.003, + "loss": 4.4042, + "step": 1142 + }, + { + "epoch": 0.01143, + "grad_norm": 0.6274623913648985, + "learning_rate": 0.003, + "loss": 4.4193, + "step": 1143 + }, + { + "epoch": 0.01144, + "grad_norm": 0.6449672978265372, + "learning_rate": 0.003, + "loss": 4.3861, + "step": 1144 + }, + { + "epoch": 0.01145, + "grad_norm": 0.8094555842673681, + "learning_rate": 0.003, + "loss": 4.4312, + "step": 1145 + }, + { + "epoch": 0.01146, + "grad_norm": 0.8440842425424007, + "learning_rate": 0.003, + "loss": 4.4247, + "step": 1146 + }, + { + "epoch": 0.01147, + "grad_norm": 0.8250289231813209, + "learning_rate": 0.003, + "loss": 4.4121, + "step": 1147 + }, + { + "epoch": 0.01148, + "grad_norm": 0.9842608879774175, + "learning_rate": 0.003, + "loss": 4.4248, + "step": 1148 + }, + { + "epoch": 0.01149, + "grad_norm": 1.140890319957312, + "learning_rate": 0.003, + "loss": 4.4515, + "step": 1149 + }, + { + "epoch": 0.0115, + "grad_norm": 1.0049560913231448, + "learning_rate": 0.003, + "loss": 4.454, + "step": 1150 + }, + { + "epoch": 0.01151, + "grad_norm": 0.9241096807179571, + "learning_rate": 0.003, + "loss": 4.4438, + "step": 1151 + }, + { + "epoch": 0.01152, + "grad_norm": 0.8197935874101315, + "learning_rate": 0.003, + "loss": 4.4496, + "step": 1152 + }, + { + "epoch": 0.01153, + "grad_norm": 0.8964418195355868, + "learning_rate": 0.003, + "loss": 4.4768, + "step": 1153 + }, + { + "epoch": 0.01154, + "grad_norm": 0.7052475574958695, + "learning_rate": 0.003, + "loss": 4.4381, + "step": 1154 + }, + { + "epoch": 0.01155, + "grad_norm": 0.6994161040159336, + "learning_rate": 0.003, + "loss": 4.425, + "step": 1155 + }, + { + "epoch": 0.01156, + "grad_norm": 0.5866991021029688, + "learning_rate": 0.003, + "loss": 4.4267, + "step": 1156 + }, + { + "epoch": 0.01157, + "grad_norm": 0.5890541796764595, + "learning_rate": 0.003, + "loss": 4.4503, + "step": 1157 + }, + { + "epoch": 0.01158, + "grad_norm": 0.5558417103751193, + "learning_rate": 0.003, + "loss": 4.4146, + "step": 1158 + }, + { + "epoch": 0.01159, + "grad_norm": 0.5494688252527891, + "learning_rate": 0.003, + "loss": 4.41, + "step": 1159 + }, + { + "epoch": 0.0116, + "grad_norm": 0.5721755744732883, + "learning_rate": 0.003, + "loss": 4.4228, + "step": 1160 + }, + { + "epoch": 0.01161, + "grad_norm": 0.6717218566295324, + "learning_rate": 0.003, + "loss": 4.4019, + "step": 1161 + }, + { + "epoch": 0.01162, + "grad_norm": 0.7268973886687413, + "learning_rate": 0.003, + "loss": 4.4138, + "step": 1162 + }, + { + "epoch": 0.01163, + "grad_norm": 0.5842283691456034, + "learning_rate": 0.003, + "loss": 4.4284, + "step": 1163 + }, + { + "epoch": 0.01164, + "grad_norm": 0.4855856513804374, + "learning_rate": 0.003, + "loss": 4.4134, + "step": 1164 + }, + { + "epoch": 0.01165, + "grad_norm": 0.4188604676147223, + "learning_rate": 0.003, + "loss": 4.4035, + "step": 1165 + }, + { + "epoch": 0.01166, + "grad_norm": 0.3540156009620338, + "learning_rate": 0.003, + "loss": 4.4019, + "step": 1166 + }, + { + "epoch": 0.01167, + "grad_norm": 0.34211903868248933, + "learning_rate": 0.003, + "loss": 4.3663, + "step": 1167 + }, + { + "epoch": 0.01168, + "grad_norm": 0.3891962127459521, + "learning_rate": 0.003, + "loss": 4.374, + "step": 1168 + }, + { + "epoch": 0.01169, + "grad_norm": 0.43062143986497, + "learning_rate": 0.003, + "loss": 4.4147, + "step": 1169 + }, + { + "epoch": 0.0117, + "grad_norm": 0.5082945055708723, + "learning_rate": 0.003, + "loss": 4.404, + "step": 1170 + }, + { + "epoch": 0.01171, + "grad_norm": 0.6470102757597544, + "learning_rate": 0.003, + "loss": 4.3885, + "step": 1171 + }, + { + "epoch": 0.01172, + "grad_norm": 0.7362820078056241, + "learning_rate": 0.003, + "loss": 4.3869, + "step": 1172 + }, + { + "epoch": 0.01173, + "grad_norm": 0.693269388291317, + "learning_rate": 0.003, + "loss": 4.4171, + "step": 1173 + }, + { + "epoch": 0.01174, + "grad_norm": 0.5646769911878592, + "learning_rate": 0.003, + "loss": 4.3891, + "step": 1174 + }, + { + "epoch": 0.01175, + "grad_norm": 0.5675605646992437, + "learning_rate": 0.003, + "loss": 4.3751, + "step": 1175 + }, + { + "epoch": 0.01176, + "grad_norm": 0.5874888204753396, + "learning_rate": 0.003, + "loss": 4.4094, + "step": 1176 + }, + { + "epoch": 0.01177, + "grad_norm": 0.4402531057426568, + "learning_rate": 0.003, + "loss": 4.3711, + "step": 1177 + }, + { + "epoch": 0.01178, + "grad_norm": 0.542910388266956, + "learning_rate": 0.003, + "loss": 4.3959, + "step": 1178 + }, + { + "epoch": 0.01179, + "grad_norm": 0.6077990883816204, + "learning_rate": 0.003, + "loss": 4.3751, + "step": 1179 + }, + { + "epoch": 0.0118, + "grad_norm": 0.5935696360755632, + "learning_rate": 0.003, + "loss": 4.4393, + "step": 1180 + }, + { + "epoch": 0.01181, + "grad_norm": 0.5077497618150401, + "learning_rate": 0.003, + "loss": 4.4227, + "step": 1181 + }, + { + "epoch": 0.01182, + "grad_norm": 0.4525800009003507, + "learning_rate": 0.003, + "loss": 4.3791, + "step": 1182 + }, + { + "epoch": 0.01183, + "grad_norm": 0.465730915302176, + "learning_rate": 0.003, + "loss": 4.3741, + "step": 1183 + }, + { + "epoch": 0.01184, + "grad_norm": 0.5691375544130988, + "learning_rate": 0.003, + "loss": 4.4172, + "step": 1184 + }, + { + "epoch": 0.01185, + "grad_norm": 0.6220094109188331, + "learning_rate": 0.003, + "loss": 4.3896, + "step": 1185 + }, + { + "epoch": 0.01186, + "grad_norm": 0.7180162404965751, + "learning_rate": 0.003, + "loss": 4.417, + "step": 1186 + }, + { + "epoch": 0.01187, + "grad_norm": 0.7918251317294935, + "learning_rate": 0.003, + "loss": 4.389, + "step": 1187 + }, + { + "epoch": 0.01188, + "grad_norm": 0.8521642061482538, + "learning_rate": 0.003, + "loss": 4.4081, + "step": 1188 + }, + { + "epoch": 0.01189, + "grad_norm": 0.884743262111852, + "learning_rate": 0.003, + "loss": 4.4173, + "step": 1189 + }, + { + "epoch": 0.0119, + "grad_norm": 0.809315268231435, + "learning_rate": 0.003, + "loss": 4.3961, + "step": 1190 + }, + { + "epoch": 0.01191, + "grad_norm": 0.8426026711506704, + "learning_rate": 0.003, + "loss": 4.3805, + "step": 1191 + }, + { + "epoch": 0.01192, + "grad_norm": 0.8039017391990436, + "learning_rate": 0.003, + "loss": 4.4134, + "step": 1192 + }, + { + "epoch": 0.01193, + "grad_norm": 1.0091767993853729, + "learning_rate": 0.003, + "loss": 4.4278, + "step": 1193 + }, + { + "epoch": 0.01194, + "grad_norm": 1.021446612913599, + "learning_rate": 0.003, + "loss": 4.4458, + "step": 1194 + }, + { + "epoch": 0.01195, + "grad_norm": 0.8197186317539255, + "learning_rate": 0.003, + "loss": 4.4136, + "step": 1195 + }, + { + "epoch": 0.01196, + "grad_norm": 1.0663075325190898, + "learning_rate": 0.003, + "loss": 4.4346, + "step": 1196 + }, + { + "epoch": 0.01197, + "grad_norm": 1.1163749964317249, + "learning_rate": 0.003, + "loss": 4.42, + "step": 1197 + }, + { + "epoch": 0.01198, + "grad_norm": 0.7911818503418244, + "learning_rate": 0.003, + "loss": 4.4756, + "step": 1198 + }, + { + "epoch": 0.01199, + "grad_norm": 0.7094954502714929, + "learning_rate": 0.003, + "loss": 4.4429, + "step": 1199 + }, + { + "epoch": 0.012, + "grad_norm": 0.6371349856097592, + "learning_rate": 0.003, + "loss": 4.4311, + "step": 1200 + }, + { + "epoch": 0.01201, + "grad_norm": 0.6670059038710348, + "learning_rate": 0.003, + "loss": 4.4586, + "step": 1201 + }, + { + "epoch": 0.01202, + "grad_norm": 0.7057344621036731, + "learning_rate": 0.003, + "loss": 4.4253, + "step": 1202 + }, + { + "epoch": 0.01203, + "grad_norm": 0.6813638781778574, + "learning_rate": 0.003, + "loss": 4.4314, + "step": 1203 + }, + { + "epoch": 0.01204, + "grad_norm": 0.6487927748495244, + "learning_rate": 0.003, + "loss": 4.4252, + "step": 1204 + }, + { + "epoch": 0.01205, + "grad_norm": 0.5702088035511471, + "learning_rate": 0.003, + "loss": 4.4134, + "step": 1205 + }, + { + "epoch": 0.01206, + "grad_norm": 0.5185412999807271, + "learning_rate": 0.003, + "loss": 4.4527, + "step": 1206 + }, + { + "epoch": 0.01207, + "grad_norm": 0.5529206635637681, + "learning_rate": 0.003, + "loss": 4.3748, + "step": 1207 + }, + { + "epoch": 0.01208, + "grad_norm": 0.6027594396835911, + "learning_rate": 0.003, + "loss": 4.4355, + "step": 1208 + }, + { + "epoch": 0.01209, + "grad_norm": 0.6294683615242583, + "learning_rate": 0.003, + "loss": 4.4139, + "step": 1209 + }, + { + "epoch": 0.0121, + "grad_norm": 0.5634425369332556, + "learning_rate": 0.003, + "loss": 4.4142, + "step": 1210 + }, + { + "epoch": 0.01211, + "grad_norm": 0.489983773731107, + "learning_rate": 0.003, + "loss": 4.3953, + "step": 1211 + }, + { + "epoch": 0.01212, + "grad_norm": 0.48590408635878257, + "learning_rate": 0.003, + "loss": 4.4163, + "step": 1212 + }, + { + "epoch": 0.01213, + "grad_norm": 0.4654031475329597, + "learning_rate": 0.003, + "loss": 4.3921, + "step": 1213 + }, + { + "epoch": 0.01214, + "grad_norm": 0.4519325610344299, + "learning_rate": 0.003, + "loss": 4.3755, + "step": 1214 + }, + { + "epoch": 0.01215, + "grad_norm": 0.4986453835908023, + "learning_rate": 0.003, + "loss": 4.3833, + "step": 1215 + }, + { + "epoch": 0.01216, + "grad_norm": 0.6836671205807361, + "learning_rate": 0.003, + "loss": 4.3816, + "step": 1216 + }, + { + "epoch": 0.01217, + "grad_norm": 0.8634560994181636, + "learning_rate": 0.003, + "loss": 4.4035, + "step": 1217 + }, + { + "epoch": 0.01218, + "grad_norm": 0.7430166358012542, + "learning_rate": 0.003, + "loss": 4.3918, + "step": 1218 + }, + { + "epoch": 0.01219, + "grad_norm": 0.5693399114743081, + "learning_rate": 0.003, + "loss": 4.3834, + "step": 1219 + }, + { + "epoch": 0.0122, + "grad_norm": 0.6203041570130704, + "learning_rate": 0.003, + "loss": 4.3803, + "step": 1220 + }, + { + "epoch": 0.01221, + "grad_norm": 0.6221216866599081, + "learning_rate": 0.003, + "loss": 4.3932, + "step": 1221 + }, + { + "epoch": 0.01222, + "grad_norm": 0.5678017035835031, + "learning_rate": 0.003, + "loss": 4.3744, + "step": 1222 + }, + { + "epoch": 0.01223, + "grad_norm": 0.4544260730538049, + "learning_rate": 0.003, + "loss": 4.3669, + "step": 1223 + }, + { + "epoch": 0.01224, + "grad_norm": 0.4498657047747612, + "learning_rate": 0.003, + "loss": 4.3528, + "step": 1224 + }, + { + "epoch": 0.01225, + "grad_norm": 0.44954702929076595, + "learning_rate": 0.003, + "loss": 4.3751, + "step": 1225 + }, + { + "epoch": 0.01226, + "grad_norm": 0.45922987677597427, + "learning_rate": 0.003, + "loss": 4.3678, + "step": 1226 + }, + { + "epoch": 0.01227, + "grad_norm": 0.444736072897216, + "learning_rate": 0.003, + "loss": 4.3955, + "step": 1227 + }, + { + "epoch": 0.01228, + "grad_norm": 0.39964234776344315, + "learning_rate": 0.003, + "loss": 4.3749, + "step": 1228 + }, + { + "epoch": 0.01229, + "grad_norm": 0.4304709469798317, + "learning_rate": 0.003, + "loss": 4.3636, + "step": 1229 + }, + { + "epoch": 0.0123, + "grad_norm": 0.45096371519587236, + "learning_rate": 0.003, + "loss": 4.3946, + "step": 1230 + }, + { + "epoch": 0.01231, + "grad_norm": 0.45260016324469293, + "learning_rate": 0.003, + "loss": 4.3662, + "step": 1231 + }, + { + "epoch": 0.01232, + "grad_norm": 0.5056071407476121, + "learning_rate": 0.003, + "loss": 4.3416, + "step": 1232 + }, + { + "epoch": 0.01233, + "grad_norm": 0.6137961234604853, + "learning_rate": 0.003, + "loss": 4.3562, + "step": 1233 + }, + { + "epoch": 0.01234, + "grad_norm": 0.7580924108092045, + "learning_rate": 0.003, + "loss": 4.3771, + "step": 1234 + }, + { + "epoch": 0.01235, + "grad_norm": 0.7682739117958468, + "learning_rate": 0.003, + "loss": 4.377, + "step": 1235 + }, + { + "epoch": 0.01236, + "grad_norm": 0.7726095308103765, + "learning_rate": 0.003, + "loss": 4.379, + "step": 1236 + }, + { + "epoch": 0.01237, + "grad_norm": 0.6515520396776402, + "learning_rate": 0.003, + "loss": 4.3868, + "step": 1237 + }, + { + "epoch": 0.01238, + "grad_norm": 0.5911401797128396, + "learning_rate": 0.003, + "loss": 4.3811, + "step": 1238 + }, + { + "epoch": 0.01239, + "grad_norm": 0.6868743136698521, + "learning_rate": 0.003, + "loss": 4.3689, + "step": 1239 + }, + { + "epoch": 0.0124, + "grad_norm": 0.7235342138686123, + "learning_rate": 0.003, + "loss": 4.3578, + "step": 1240 + }, + { + "epoch": 0.01241, + "grad_norm": 0.7086323715527123, + "learning_rate": 0.003, + "loss": 4.3714, + "step": 1241 + }, + { + "epoch": 0.01242, + "grad_norm": 0.7273279516261576, + "learning_rate": 0.003, + "loss": 4.3772, + "step": 1242 + }, + { + "epoch": 0.01243, + "grad_norm": 0.7241329186524991, + "learning_rate": 0.003, + "loss": 4.358, + "step": 1243 + }, + { + "epoch": 0.01244, + "grad_norm": 0.659589685877625, + "learning_rate": 0.003, + "loss": 4.3697, + "step": 1244 + }, + { + "epoch": 0.01245, + "grad_norm": 0.5679488489092839, + "learning_rate": 0.003, + "loss": 4.3588, + "step": 1245 + }, + { + "epoch": 0.01246, + "grad_norm": 0.5843246373122641, + "learning_rate": 0.003, + "loss": 4.3356, + "step": 1246 + }, + { + "epoch": 0.01247, + "grad_norm": 0.5386207185407424, + "learning_rate": 0.003, + "loss": 4.3884, + "step": 1247 + }, + { + "epoch": 0.01248, + "grad_norm": 0.4947062329390487, + "learning_rate": 0.003, + "loss": 4.3718, + "step": 1248 + }, + { + "epoch": 0.01249, + "grad_norm": 0.47539230004779465, + "learning_rate": 0.003, + "loss": 4.3597, + "step": 1249 + }, + { + "epoch": 0.0125, + "grad_norm": 0.4876155686681362, + "learning_rate": 0.003, + "loss": 4.3353, + "step": 1250 + }, + { + "epoch": 0.01251, + "grad_norm": 0.587984128950104, + "learning_rate": 0.003, + "loss": 4.3725, + "step": 1251 + }, + { + "epoch": 0.01252, + "grad_norm": 0.74609447185642, + "learning_rate": 0.003, + "loss": 4.3585, + "step": 1252 + }, + { + "epoch": 0.01253, + "grad_norm": 0.8297104750282515, + "learning_rate": 0.003, + "loss": 4.3426, + "step": 1253 + }, + { + "epoch": 0.01254, + "grad_norm": 0.6714602053035584, + "learning_rate": 0.003, + "loss": 4.3879, + "step": 1254 + }, + { + "epoch": 0.01255, + "grad_norm": 0.5575284697102026, + "learning_rate": 0.003, + "loss": 4.3328, + "step": 1255 + }, + { + "epoch": 0.01256, + "grad_norm": 0.6531801875316168, + "learning_rate": 0.003, + "loss": 4.3617, + "step": 1256 + }, + { + "epoch": 0.01257, + "grad_norm": 0.6725833064810157, + "learning_rate": 0.003, + "loss": 4.3523, + "step": 1257 + }, + { + "epoch": 0.01258, + "grad_norm": 0.5415379584452896, + "learning_rate": 0.003, + "loss": 4.3533, + "step": 1258 + }, + { + "epoch": 0.01259, + "grad_norm": 0.6189096611706483, + "learning_rate": 0.003, + "loss": 4.3565, + "step": 1259 + }, + { + "epoch": 0.0126, + "grad_norm": 0.7719709473721972, + "learning_rate": 0.003, + "loss": 4.3943, + "step": 1260 + }, + { + "epoch": 0.01261, + "grad_norm": 0.7141309260632757, + "learning_rate": 0.003, + "loss": 4.3828, + "step": 1261 + }, + { + "epoch": 0.01262, + "grad_norm": 0.82201580215985, + "learning_rate": 0.003, + "loss": 4.3803, + "step": 1262 + }, + { + "epoch": 0.01263, + "grad_norm": 0.7648394097891381, + "learning_rate": 0.003, + "loss": 4.3764, + "step": 1263 + }, + { + "epoch": 0.01264, + "grad_norm": 0.6534936022081808, + "learning_rate": 0.003, + "loss": 4.3615, + "step": 1264 + }, + { + "epoch": 0.01265, + "grad_norm": 0.8083064566897021, + "learning_rate": 0.003, + "loss": 4.3795, + "step": 1265 + }, + { + "epoch": 0.01266, + "grad_norm": 0.7625627444580204, + "learning_rate": 0.003, + "loss": 4.3817, + "step": 1266 + }, + { + "epoch": 0.01267, + "grad_norm": 0.7144669763786441, + "learning_rate": 0.003, + "loss": 4.3961, + "step": 1267 + }, + { + "epoch": 0.01268, + "grad_norm": 0.6176763363932604, + "learning_rate": 0.003, + "loss": 4.3815, + "step": 1268 + }, + { + "epoch": 0.01269, + "grad_norm": 0.5844219741679575, + "learning_rate": 0.003, + "loss": 4.3442, + "step": 1269 + }, + { + "epoch": 0.0127, + "grad_norm": 0.5570920646541462, + "learning_rate": 0.003, + "loss": 4.3764, + "step": 1270 + }, + { + "epoch": 0.01271, + "grad_norm": 0.665679672510672, + "learning_rate": 0.003, + "loss": 4.3628, + "step": 1271 + }, + { + "epoch": 0.01272, + "grad_norm": 0.7700987517359916, + "learning_rate": 0.003, + "loss": 4.3638, + "step": 1272 + }, + { + "epoch": 0.01273, + "grad_norm": 0.7473118276250652, + "learning_rate": 0.003, + "loss": 4.3932, + "step": 1273 + }, + { + "epoch": 0.01274, + "grad_norm": 0.6491851354934152, + "learning_rate": 0.003, + "loss": 4.3789, + "step": 1274 + }, + { + "epoch": 0.01275, + "grad_norm": 0.6744635149292406, + "learning_rate": 0.003, + "loss": 4.3494, + "step": 1275 + }, + { + "epoch": 0.01276, + "grad_norm": 0.652401206080157, + "learning_rate": 0.003, + "loss": 4.3572, + "step": 1276 + }, + { + "epoch": 0.01277, + "grad_norm": 0.8226392169345714, + "learning_rate": 0.003, + "loss": 4.3706, + "step": 1277 + }, + { + "epoch": 0.01278, + "grad_norm": 0.8325255271008436, + "learning_rate": 0.003, + "loss": 4.3906, + "step": 1278 + }, + { + "epoch": 0.01279, + "grad_norm": 0.7029576010205191, + "learning_rate": 0.003, + "loss": 4.3612, + "step": 1279 + }, + { + "epoch": 0.0128, + "grad_norm": 0.6056600803694466, + "learning_rate": 0.003, + "loss": 4.3548, + "step": 1280 + }, + { + "epoch": 0.01281, + "grad_norm": 0.6068824045406035, + "learning_rate": 0.003, + "loss": 4.3433, + "step": 1281 + }, + { + "epoch": 0.01282, + "grad_norm": 0.5449477323585487, + "learning_rate": 0.003, + "loss": 4.3801, + "step": 1282 + }, + { + "epoch": 0.01283, + "grad_norm": 0.5086376593278678, + "learning_rate": 0.003, + "loss": 4.3465, + "step": 1283 + }, + { + "epoch": 0.01284, + "grad_norm": 0.49507526795006285, + "learning_rate": 0.003, + "loss": 4.3484, + "step": 1284 + }, + { + "epoch": 0.01285, + "grad_norm": 0.4794063886412805, + "learning_rate": 0.003, + "loss": 4.3499, + "step": 1285 + }, + { + "epoch": 0.01286, + "grad_norm": 0.4314216967223394, + "learning_rate": 0.003, + "loss": 4.37, + "step": 1286 + }, + { + "epoch": 0.01287, + "grad_norm": 0.43751633400555473, + "learning_rate": 0.003, + "loss": 4.3355, + "step": 1287 + }, + { + "epoch": 0.01288, + "grad_norm": 0.4402964980769675, + "learning_rate": 0.003, + "loss": 4.3477, + "step": 1288 + }, + { + "epoch": 0.01289, + "grad_norm": 0.4330370123851316, + "learning_rate": 0.003, + "loss": 4.34, + "step": 1289 + }, + { + "epoch": 0.0129, + "grad_norm": 0.4577360505134826, + "learning_rate": 0.003, + "loss": 4.3613, + "step": 1290 + }, + { + "epoch": 0.01291, + "grad_norm": 0.4908761222350223, + "learning_rate": 0.003, + "loss": 4.3434, + "step": 1291 + }, + { + "epoch": 0.01292, + "grad_norm": 0.5197962656056466, + "learning_rate": 0.003, + "loss": 4.3769, + "step": 1292 + }, + { + "epoch": 0.01293, + "grad_norm": 0.5249407732060195, + "learning_rate": 0.003, + "loss": 4.3636, + "step": 1293 + }, + { + "epoch": 0.01294, + "grad_norm": 0.611698986416238, + "learning_rate": 0.003, + "loss": 4.3337, + "step": 1294 + }, + { + "epoch": 0.01295, + "grad_norm": 0.7156402524759599, + "learning_rate": 0.003, + "loss": 4.369, + "step": 1295 + }, + { + "epoch": 0.01296, + "grad_norm": 0.8400871327514794, + "learning_rate": 0.003, + "loss": 4.3429, + "step": 1296 + }, + { + "epoch": 0.01297, + "grad_norm": 0.9437097734689512, + "learning_rate": 0.003, + "loss": 4.3707, + "step": 1297 + }, + { + "epoch": 0.01298, + "grad_norm": 0.9584374674542437, + "learning_rate": 0.003, + "loss": 4.3505, + "step": 1298 + }, + { + "epoch": 0.01299, + "grad_norm": 0.8671660761915297, + "learning_rate": 0.003, + "loss": 4.3625, + "step": 1299 + }, + { + "epoch": 0.013, + "grad_norm": 0.7502528073695865, + "learning_rate": 0.003, + "loss": 4.3495, + "step": 1300 + }, + { + "epoch": 0.01301, + "grad_norm": 0.7063515582852367, + "learning_rate": 0.003, + "loss": 4.3619, + "step": 1301 + }, + { + "epoch": 0.01302, + "grad_norm": 0.6149436229664508, + "learning_rate": 0.003, + "loss": 4.4034, + "step": 1302 + }, + { + "epoch": 0.01303, + "grad_norm": 0.6398205689943912, + "learning_rate": 0.003, + "loss": 4.3511, + "step": 1303 + }, + { + "epoch": 0.01304, + "grad_norm": 0.5125654671319883, + "learning_rate": 0.003, + "loss": 4.3743, + "step": 1304 + }, + { + "epoch": 0.01305, + "grad_norm": 0.49151327704885306, + "learning_rate": 0.003, + "loss": 4.357, + "step": 1305 + }, + { + "epoch": 0.01306, + "grad_norm": 0.449035259984806, + "learning_rate": 0.003, + "loss": 4.3496, + "step": 1306 + }, + { + "epoch": 0.01307, + "grad_norm": 0.3786564732477656, + "learning_rate": 0.003, + "loss": 4.365, + "step": 1307 + }, + { + "epoch": 0.01308, + "grad_norm": 0.40747339154306583, + "learning_rate": 0.003, + "loss": 4.3407, + "step": 1308 + }, + { + "epoch": 0.01309, + "grad_norm": 0.3477777170477395, + "learning_rate": 0.003, + "loss": 4.3771, + "step": 1309 + }, + { + "epoch": 0.0131, + "grad_norm": 0.33985557007437, + "learning_rate": 0.003, + "loss": 4.3221, + "step": 1310 + }, + { + "epoch": 0.01311, + "grad_norm": 0.3781744690142208, + "learning_rate": 0.003, + "loss": 4.3621, + "step": 1311 + }, + { + "epoch": 0.01312, + "grad_norm": 0.390430095045798, + "learning_rate": 0.003, + "loss": 4.3295, + "step": 1312 + }, + { + "epoch": 0.01313, + "grad_norm": 0.4411339748954226, + "learning_rate": 0.003, + "loss": 4.3219, + "step": 1313 + }, + { + "epoch": 0.01314, + "grad_norm": 0.4398621851217481, + "learning_rate": 0.003, + "loss": 4.3182, + "step": 1314 + }, + { + "epoch": 0.01315, + "grad_norm": 0.4543520117536255, + "learning_rate": 0.003, + "loss": 4.3454, + "step": 1315 + }, + { + "epoch": 0.01316, + "grad_norm": 0.4647119001484631, + "learning_rate": 0.003, + "loss": 4.3315, + "step": 1316 + }, + { + "epoch": 0.01317, + "grad_norm": 0.6634726948283797, + "learning_rate": 0.003, + "loss": 4.3607, + "step": 1317 + }, + { + "epoch": 0.01318, + "grad_norm": 1.0496539085022096, + "learning_rate": 0.003, + "loss": 4.3666, + "step": 1318 + }, + { + "epoch": 0.01319, + "grad_norm": 1.3165114729988252, + "learning_rate": 0.003, + "loss": 4.392, + "step": 1319 + }, + { + "epoch": 0.0132, + "grad_norm": 0.7650097283623507, + "learning_rate": 0.003, + "loss": 4.3723, + "step": 1320 + }, + { + "epoch": 0.01321, + "grad_norm": 0.7692901736546863, + "learning_rate": 0.003, + "loss": 4.3563, + "step": 1321 + }, + { + "epoch": 0.01322, + "grad_norm": 0.8430579356733313, + "learning_rate": 0.003, + "loss": 4.3811, + "step": 1322 + }, + { + "epoch": 0.01323, + "grad_norm": 0.7787983772827455, + "learning_rate": 0.003, + "loss": 4.3749, + "step": 1323 + }, + { + "epoch": 0.01324, + "grad_norm": 0.7585640630062677, + "learning_rate": 0.003, + "loss": 4.3595, + "step": 1324 + }, + { + "epoch": 0.01325, + "grad_norm": 0.662871102926864, + "learning_rate": 0.003, + "loss": 4.3604, + "step": 1325 + }, + { + "epoch": 0.01326, + "grad_norm": 0.6552542644296041, + "learning_rate": 0.003, + "loss": 4.3471, + "step": 1326 + }, + { + "epoch": 0.01327, + "grad_norm": 0.6597442221568894, + "learning_rate": 0.003, + "loss": 4.3636, + "step": 1327 + }, + { + "epoch": 0.01328, + "grad_norm": 0.6991774516177431, + "learning_rate": 0.003, + "loss": 4.3572, + "step": 1328 + }, + { + "epoch": 0.01329, + "grad_norm": 0.7004568547673504, + "learning_rate": 0.003, + "loss": 4.323, + "step": 1329 + }, + { + "epoch": 0.0133, + "grad_norm": 0.6451796491929548, + "learning_rate": 0.003, + "loss": 4.3619, + "step": 1330 + }, + { + "epoch": 0.01331, + "grad_norm": 0.5603586741834496, + "learning_rate": 0.003, + "loss": 4.3461, + "step": 1331 + }, + { + "epoch": 0.01332, + "grad_norm": 0.6118296137650349, + "learning_rate": 0.003, + "loss": 4.3674, + "step": 1332 + }, + { + "epoch": 0.01333, + "grad_norm": 0.6556746462798643, + "learning_rate": 0.003, + "loss": 4.3385, + "step": 1333 + }, + { + "epoch": 0.01334, + "grad_norm": 0.6569171248898434, + "learning_rate": 0.003, + "loss": 4.3292, + "step": 1334 + }, + { + "epoch": 0.01335, + "grad_norm": 0.7308490257197903, + "learning_rate": 0.003, + "loss": 4.355, + "step": 1335 + }, + { + "epoch": 0.01336, + "grad_norm": 0.7694263760249205, + "learning_rate": 0.003, + "loss": 4.3562, + "step": 1336 + }, + { + "epoch": 0.01337, + "grad_norm": 0.829484027358136, + "learning_rate": 0.003, + "loss": 4.3739, + "step": 1337 + }, + { + "epoch": 0.01338, + "grad_norm": 0.7394720058707402, + "learning_rate": 0.003, + "loss": 4.3375, + "step": 1338 + }, + { + "epoch": 0.01339, + "grad_norm": 0.7663877331658486, + "learning_rate": 0.003, + "loss": 4.3614, + "step": 1339 + }, + { + "epoch": 0.0134, + "grad_norm": 0.6814802672525545, + "learning_rate": 0.003, + "loss": 4.3431, + "step": 1340 + }, + { + "epoch": 0.01341, + "grad_norm": 0.5804799698607723, + "learning_rate": 0.003, + "loss": 4.3566, + "step": 1341 + }, + { + "epoch": 0.01342, + "grad_norm": 0.5489998904660267, + "learning_rate": 0.003, + "loss": 4.3242, + "step": 1342 + }, + { + "epoch": 0.01343, + "grad_norm": 0.5528070757921824, + "learning_rate": 0.003, + "loss": 4.3308, + "step": 1343 + }, + { + "epoch": 0.01344, + "grad_norm": 0.5249427916401641, + "learning_rate": 0.003, + "loss": 4.3286, + "step": 1344 + }, + { + "epoch": 0.01345, + "grad_norm": 0.5796492875823627, + "learning_rate": 0.003, + "loss": 4.3698, + "step": 1345 + }, + { + "epoch": 0.01346, + "grad_norm": 0.7791779206443923, + "learning_rate": 0.003, + "loss": 4.349, + "step": 1346 + }, + { + "epoch": 0.01347, + "grad_norm": 0.9787862373774305, + "learning_rate": 0.003, + "loss": 4.3601, + "step": 1347 + }, + { + "epoch": 0.01348, + "grad_norm": 1.124691751963475, + "learning_rate": 0.003, + "loss": 4.3695, + "step": 1348 + }, + { + "epoch": 0.01349, + "grad_norm": 0.7239534112442599, + "learning_rate": 0.003, + "loss": 4.3703, + "step": 1349 + }, + { + "epoch": 0.0135, + "grad_norm": 0.7058824966707771, + "learning_rate": 0.003, + "loss": 4.3665, + "step": 1350 + }, + { + "epoch": 0.01351, + "grad_norm": 0.7269767228946674, + "learning_rate": 0.003, + "loss": 4.3913, + "step": 1351 + }, + { + "epoch": 0.01352, + "grad_norm": 0.629762174967035, + "learning_rate": 0.003, + "loss": 4.3652, + "step": 1352 + }, + { + "epoch": 0.01353, + "grad_norm": 0.46709847900974316, + "learning_rate": 0.003, + "loss": 4.3426, + "step": 1353 + }, + { + "epoch": 0.01354, + "grad_norm": 0.4348570974243878, + "learning_rate": 0.003, + "loss": 4.33, + "step": 1354 + }, + { + "epoch": 0.01355, + "grad_norm": 0.4058626528377844, + "learning_rate": 0.003, + "loss": 4.3397, + "step": 1355 + }, + { + "epoch": 0.01356, + "grad_norm": 0.4077904728688183, + "learning_rate": 0.003, + "loss": 4.3329, + "step": 1356 + }, + { + "epoch": 0.01357, + "grad_norm": 0.3881492908104237, + "learning_rate": 0.003, + "loss": 4.3196, + "step": 1357 + }, + { + "epoch": 0.01358, + "grad_norm": 0.37883341535998916, + "learning_rate": 0.003, + "loss": 4.3135, + "step": 1358 + }, + { + "epoch": 0.01359, + "grad_norm": 0.3673079033329598, + "learning_rate": 0.003, + "loss": 4.3515, + "step": 1359 + }, + { + "epoch": 0.0136, + "grad_norm": 0.37976733712994504, + "learning_rate": 0.003, + "loss": 4.3177, + "step": 1360 + }, + { + "epoch": 0.01361, + "grad_norm": 0.4118348378412375, + "learning_rate": 0.003, + "loss": 4.3313, + "step": 1361 + }, + { + "epoch": 0.01362, + "grad_norm": 0.4407593683397167, + "learning_rate": 0.003, + "loss": 4.345, + "step": 1362 + }, + { + "epoch": 0.01363, + "grad_norm": 0.5058671106167254, + "learning_rate": 0.003, + "loss": 4.3464, + "step": 1363 + }, + { + "epoch": 0.01364, + "grad_norm": 0.6807238620694973, + "learning_rate": 0.003, + "loss": 4.3446, + "step": 1364 + }, + { + "epoch": 0.01365, + "grad_norm": 0.8032111226209311, + "learning_rate": 0.003, + "loss": 4.3702, + "step": 1365 + }, + { + "epoch": 0.01366, + "grad_norm": 0.7976336994683142, + "learning_rate": 0.003, + "loss": 4.3233, + "step": 1366 + }, + { + "epoch": 0.01367, + "grad_norm": 0.7881501972797743, + "learning_rate": 0.003, + "loss": 4.3766, + "step": 1367 + }, + { + "epoch": 0.01368, + "grad_norm": 0.6437176795201012, + "learning_rate": 0.003, + "loss": 4.346, + "step": 1368 + }, + { + "epoch": 0.01369, + "grad_norm": 0.7544973412872852, + "learning_rate": 0.003, + "loss": 4.3393, + "step": 1369 + }, + { + "epoch": 0.0137, + "grad_norm": 0.9164166166999201, + "learning_rate": 0.003, + "loss": 4.3794, + "step": 1370 + }, + { + "epoch": 0.01371, + "grad_norm": 0.9303263432444765, + "learning_rate": 0.003, + "loss": 4.3692, + "step": 1371 + }, + { + "epoch": 0.01372, + "grad_norm": 0.7939151912920627, + "learning_rate": 0.003, + "loss": 4.3176, + "step": 1372 + }, + { + "epoch": 0.01373, + "grad_norm": 0.8088750323088586, + "learning_rate": 0.003, + "loss": 4.3626, + "step": 1373 + }, + { + "epoch": 0.01374, + "grad_norm": 0.6721224672262793, + "learning_rate": 0.003, + "loss": 4.3585, + "step": 1374 + }, + { + "epoch": 0.01375, + "grad_norm": 0.6771097427365759, + "learning_rate": 0.003, + "loss": 4.3249, + "step": 1375 + }, + { + "epoch": 0.01376, + "grad_norm": 0.6762752470113543, + "learning_rate": 0.003, + "loss": 4.347, + "step": 1376 + }, + { + "epoch": 0.01377, + "grad_norm": 0.6604690846076395, + "learning_rate": 0.003, + "loss": 4.3466, + "step": 1377 + }, + { + "epoch": 0.01378, + "grad_norm": 0.8027631418948692, + "learning_rate": 0.003, + "loss": 4.3812, + "step": 1378 + }, + { + "epoch": 0.01379, + "grad_norm": 0.8058991494927565, + "learning_rate": 0.003, + "loss": 4.3543, + "step": 1379 + }, + { + "epoch": 0.0138, + "grad_norm": 0.6522516144796072, + "learning_rate": 0.003, + "loss": 4.3459, + "step": 1380 + }, + { + "epoch": 0.01381, + "grad_norm": 0.6569912385058241, + "learning_rate": 0.003, + "loss": 4.3524, + "step": 1381 + }, + { + "epoch": 0.01382, + "grad_norm": 0.5996275495095534, + "learning_rate": 0.003, + "loss": 4.3471, + "step": 1382 + }, + { + "epoch": 0.01383, + "grad_norm": 0.7099037071072233, + "learning_rate": 0.003, + "loss": 4.3442, + "step": 1383 + }, + { + "epoch": 0.01384, + "grad_norm": 0.9143332820449893, + "learning_rate": 0.003, + "loss": 4.3854, + "step": 1384 + }, + { + "epoch": 0.01385, + "grad_norm": 1.0346171765908472, + "learning_rate": 0.003, + "loss": 4.3815, + "step": 1385 + }, + { + "epoch": 0.01386, + "grad_norm": 0.8985740978326894, + "learning_rate": 0.003, + "loss": 4.3558, + "step": 1386 + }, + { + "epoch": 0.01387, + "grad_norm": 0.6958974730086266, + "learning_rate": 0.003, + "loss": 4.335, + "step": 1387 + }, + { + "epoch": 0.01388, + "grad_norm": 0.7497022428211844, + "learning_rate": 0.003, + "loss": 4.3735, + "step": 1388 + }, + { + "epoch": 0.01389, + "grad_norm": 0.7334888360507559, + "learning_rate": 0.003, + "loss": 4.3541, + "step": 1389 + }, + { + "epoch": 0.0139, + "grad_norm": 0.6634716599193375, + "learning_rate": 0.003, + "loss": 4.3533, + "step": 1390 + }, + { + "epoch": 0.01391, + "grad_norm": 0.5884420459817105, + "learning_rate": 0.003, + "loss": 4.3483, + "step": 1391 + }, + { + "epoch": 0.01392, + "grad_norm": 0.536202889592441, + "learning_rate": 0.003, + "loss": 4.3599, + "step": 1392 + }, + { + "epoch": 0.01393, + "grad_norm": 0.5041805750347179, + "learning_rate": 0.003, + "loss": 4.327, + "step": 1393 + }, + { + "epoch": 0.01394, + "grad_norm": 0.46187226586984687, + "learning_rate": 0.003, + "loss": 4.3118, + "step": 1394 + }, + { + "epoch": 0.01395, + "grad_norm": 0.4642311356997612, + "learning_rate": 0.003, + "loss": 4.343, + "step": 1395 + }, + { + "epoch": 0.01396, + "grad_norm": 0.49219234547480073, + "learning_rate": 0.003, + "loss": 4.3333, + "step": 1396 + }, + { + "epoch": 0.01397, + "grad_norm": 0.535253115362672, + "learning_rate": 0.003, + "loss": 4.3471, + "step": 1397 + }, + { + "epoch": 0.01398, + "grad_norm": 0.6052714387106423, + "learning_rate": 0.003, + "loss": 4.3267, + "step": 1398 + }, + { + "epoch": 0.01399, + "grad_norm": 0.6277477924110956, + "learning_rate": 0.003, + "loss": 4.3504, + "step": 1399 + }, + { + "epoch": 0.014, + "grad_norm": 0.6846592906181149, + "learning_rate": 0.003, + "loss": 4.3103, + "step": 1400 + }, + { + "epoch": 0.01401, + "grad_norm": 0.8200984983857136, + "learning_rate": 0.003, + "loss": 4.3387, + "step": 1401 + }, + { + "epoch": 0.01402, + "grad_norm": 0.8742108124137586, + "learning_rate": 0.003, + "loss": 4.3557, + "step": 1402 + }, + { + "epoch": 0.01403, + "grad_norm": 0.7096844776193044, + "learning_rate": 0.003, + "loss": 4.3233, + "step": 1403 + }, + { + "epoch": 0.01404, + "grad_norm": 0.5388679123577896, + "learning_rate": 0.003, + "loss": 4.3401, + "step": 1404 + }, + { + "epoch": 0.01405, + "grad_norm": 0.4910540359683292, + "learning_rate": 0.003, + "loss": 4.3559, + "step": 1405 + }, + { + "epoch": 0.01406, + "grad_norm": 0.5213447941957257, + "learning_rate": 0.003, + "loss": 4.3366, + "step": 1406 + }, + { + "epoch": 0.01407, + "grad_norm": 0.555926299222398, + "learning_rate": 0.003, + "loss": 4.3521, + "step": 1407 + }, + { + "epoch": 0.01408, + "grad_norm": 0.692327807092523, + "learning_rate": 0.003, + "loss": 4.3137, + "step": 1408 + }, + { + "epoch": 0.01409, + "grad_norm": 0.8697281888439876, + "learning_rate": 0.003, + "loss": 4.3588, + "step": 1409 + }, + { + "epoch": 0.0141, + "grad_norm": 0.9566955718451174, + "learning_rate": 0.003, + "loss": 4.3446, + "step": 1410 + }, + { + "epoch": 0.01411, + "grad_norm": 0.7215244847418736, + "learning_rate": 0.003, + "loss": 4.3131, + "step": 1411 + }, + { + "epoch": 0.01412, + "grad_norm": 0.6315141391735041, + "learning_rate": 0.003, + "loss": 4.3535, + "step": 1412 + }, + { + "epoch": 0.01413, + "grad_norm": 0.6841381857718712, + "learning_rate": 0.003, + "loss": 4.342, + "step": 1413 + }, + { + "epoch": 0.01414, + "grad_norm": 0.5798611042820683, + "learning_rate": 0.003, + "loss": 4.3235, + "step": 1414 + }, + { + "epoch": 0.01415, + "grad_norm": 0.547398429036414, + "learning_rate": 0.003, + "loss": 4.3157, + "step": 1415 + }, + { + "epoch": 0.01416, + "grad_norm": 0.532688568072748, + "learning_rate": 0.003, + "loss": 4.3388, + "step": 1416 + }, + { + "epoch": 0.01417, + "grad_norm": 0.5474366342351658, + "learning_rate": 0.003, + "loss": 4.3317, + "step": 1417 + }, + { + "epoch": 0.01418, + "grad_norm": 0.5499547169322332, + "learning_rate": 0.003, + "loss": 4.3477, + "step": 1418 + }, + { + "epoch": 0.01419, + "grad_norm": 0.58527195779541, + "learning_rate": 0.003, + "loss": 4.3403, + "step": 1419 + }, + { + "epoch": 0.0142, + "grad_norm": 0.4513568933907034, + "learning_rate": 0.003, + "loss": 4.3401, + "step": 1420 + }, + { + "epoch": 0.01421, + "grad_norm": 0.433334289258551, + "learning_rate": 0.003, + "loss": 4.3072, + "step": 1421 + }, + { + "epoch": 0.01422, + "grad_norm": 0.39636752625840327, + "learning_rate": 0.003, + "loss": 4.3095, + "step": 1422 + }, + { + "epoch": 0.01423, + "grad_norm": 0.38308187149456013, + "learning_rate": 0.003, + "loss": 4.3141, + "step": 1423 + }, + { + "epoch": 0.01424, + "grad_norm": 0.3793175762281477, + "learning_rate": 0.003, + "loss": 4.3274, + "step": 1424 + }, + { + "epoch": 0.01425, + "grad_norm": 0.4426704955835019, + "learning_rate": 0.003, + "loss": 4.334, + "step": 1425 + }, + { + "epoch": 0.01426, + "grad_norm": 0.5006373442027768, + "learning_rate": 0.003, + "loss": 4.3106, + "step": 1426 + }, + { + "epoch": 0.01427, + "grad_norm": 0.654459732986379, + "learning_rate": 0.003, + "loss": 4.3121, + "step": 1427 + }, + { + "epoch": 0.01428, + "grad_norm": 0.8067469794918197, + "learning_rate": 0.003, + "loss": 4.3172, + "step": 1428 + }, + { + "epoch": 0.01429, + "grad_norm": 0.8729241550485854, + "learning_rate": 0.003, + "loss": 4.3306, + "step": 1429 + }, + { + "epoch": 0.0143, + "grad_norm": 0.7278485959649278, + "learning_rate": 0.003, + "loss": 4.3432, + "step": 1430 + }, + { + "epoch": 0.01431, + "grad_norm": 0.7004237859708793, + "learning_rate": 0.003, + "loss": 4.3301, + "step": 1431 + }, + { + "epoch": 0.01432, + "grad_norm": 0.8171598559675952, + "learning_rate": 0.003, + "loss": 4.3267, + "step": 1432 + }, + { + "epoch": 0.01433, + "grad_norm": 0.7196947939958487, + "learning_rate": 0.003, + "loss": 4.323, + "step": 1433 + }, + { + "epoch": 0.01434, + "grad_norm": 0.6846858190137799, + "learning_rate": 0.003, + "loss": 4.3048, + "step": 1434 + }, + { + "epoch": 0.01435, + "grad_norm": 0.8455490886566032, + "learning_rate": 0.003, + "loss": 4.3242, + "step": 1435 + }, + { + "epoch": 0.01436, + "grad_norm": 0.83940902394985, + "learning_rate": 0.003, + "loss": 4.3453, + "step": 1436 + }, + { + "epoch": 0.01437, + "grad_norm": 0.7328785180307785, + "learning_rate": 0.003, + "loss": 4.3547, + "step": 1437 + }, + { + "epoch": 0.01438, + "grad_norm": 0.7369569320140708, + "learning_rate": 0.003, + "loss": 4.349, + "step": 1438 + }, + { + "epoch": 0.01439, + "grad_norm": 0.7139149757788127, + "learning_rate": 0.003, + "loss": 4.3129, + "step": 1439 + }, + { + "epoch": 0.0144, + "grad_norm": 0.7299011882512658, + "learning_rate": 0.003, + "loss": 4.3275, + "step": 1440 + }, + { + "epoch": 0.01441, + "grad_norm": 0.670750314419178, + "learning_rate": 0.003, + "loss": 4.3309, + "step": 1441 + }, + { + "epoch": 0.01442, + "grad_norm": 0.7691195459017623, + "learning_rate": 0.003, + "loss": 4.3219, + "step": 1442 + }, + { + "epoch": 0.01443, + "grad_norm": 0.7471914299871487, + "learning_rate": 0.003, + "loss": 4.3829, + "step": 1443 + }, + { + "epoch": 0.01444, + "grad_norm": 0.7410199276775529, + "learning_rate": 0.003, + "loss": 4.3445, + "step": 1444 + }, + { + "epoch": 0.01445, + "grad_norm": 0.6515002318773166, + "learning_rate": 0.003, + "loss": 4.3041, + "step": 1445 + }, + { + "epoch": 0.01446, + "grad_norm": 0.6781090959776701, + "learning_rate": 0.003, + "loss": 4.3207, + "step": 1446 + }, + { + "epoch": 0.01447, + "grad_norm": 0.6643065072159584, + "learning_rate": 0.003, + "loss": 4.3293, + "step": 1447 + }, + { + "epoch": 0.01448, + "grad_norm": 0.6532713033801402, + "learning_rate": 0.003, + "loss": 4.3375, + "step": 1448 + }, + { + "epoch": 0.01449, + "grad_norm": 0.6344501403492658, + "learning_rate": 0.003, + "loss": 4.3302, + "step": 1449 + }, + { + "epoch": 0.0145, + "grad_norm": 0.5404283744793761, + "learning_rate": 0.003, + "loss": 4.3401, + "step": 1450 + }, + { + "epoch": 0.01451, + "grad_norm": 0.5381895515345636, + "learning_rate": 0.003, + "loss": 4.3181, + "step": 1451 + }, + { + "epoch": 0.01452, + "grad_norm": 0.5085114651371503, + "learning_rate": 0.003, + "loss": 4.3176, + "step": 1452 + }, + { + "epoch": 0.01453, + "grad_norm": 0.5297212710147206, + "learning_rate": 0.003, + "loss": 4.3302, + "step": 1453 + }, + { + "epoch": 0.01454, + "grad_norm": 0.6082304279549751, + "learning_rate": 0.003, + "loss": 4.299, + "step": 1454 + }, + { + "epoch": 0.01455, + "grad_norm": 0.799761378996593, + "learning_rate": 0.003, + "loss": 4.3379, + "step": 1455 + }, + { + "epoch": 0.01456, + "grad_norm": 0.8200199061255905, + "learning_rate": 0.003, + "loss": 4.3248, + "step": 1456 + }, + { + "epoch": 0.01457, + "grad_norm": 0.6729328173983412, + "learning_rate": 0.003, + "loss": 4.3054, + "step": 1457 + }, + { + "epoch": 0.01458, + "grad_norm": 0.5761787071333488, + "learning_rate": 0.003, + "loss": 4.2955, + "step": 1458 + }, + { + "epoch": 0.01459, + "grad_norm": 0.6468019060138231, + "learning_rate": 0.003, + "loss": 4.3289, + "step": 1459 + }, + { + "epoch": 0.0146, + "grad_norm": 0.6012651007343053, + "learning_rate": 0.003, + "loss": 4.3283, + "step": 1460 + }, + { + "epoch": 0.01461, + "grad_norm": 0.4832305210154182, + "learning_rate": 0.003, + "loss": 4.3037, + "step": 1461 + }, + { + "epoch": 0.01462, + "grad_norm": 0.4933402571331549, + "learning_rate": 0.003, + "loss": 4.3088, + "step": 1462 + }, + { + "epoch": 0.01463, + "grad_norm": 0.4774020866387011, + "learning_rate": 0.003, + "loss": 4.3072, + "step": 1463 + }, + { + "epoch": 0.01464, + "grad_norm": 0.4604171826341042, + "learning_rate": 0.003, + "loss": 4.2821, + "step": 1464 + }, + { + "epoch": 0.01465, + "grad_norm": 0.4294089797917102, + "learning_rate": 0.003, + "loss": 4.3245, + "step": 1465 + }, + { + "epoch": 0.01466, + "grad_norm": 0.41929661999999984, + "learning_rate": 0.003, + "loss": 4.2919, + "step": 1466 + }, + { + "epoch": 0.01467, + "grad_norm": 0.4391785760355703, + "learning_rate": 0.003, + "loss": 4.3283, + "step": 1467 + }, + { + "epoch": 0.01468, + "grad_norm": 0.4538846759969807, + "learning_rate": 0.003, + "loss": 4.3037, + "step": 1468 + }, + { + "epoch": 0.01469, + "grad_norm": 0.5119886030394242, + "learning_rate": 0.003, + "loss": 4.3053, + "step": 1469 + }, + { + "epoch": 0.0147, + "grad_norm": 0.716327286811815, + "learning_rate": 0.003, + "loss": 4.3041, + "step": 1470 + }, + { + "epoch": 0.01471, + "grad_norm": 0.9150052503540761, + "learning_rate": 0.003, + "loss": 4.3274, + "step": 1471 + }, + { + "epoch": 0.01472, + "grad_norm": 0.7809983181295278, + "learning_rate": 0.003, + "loss": 4.3143, + "step": 1472 + }, + { + "epoch": 0.01473, + "grad_norm": 0.8086105929105535, + "learning_rate": 0.003, + "loss": 4.3289, + "step": 1473 + }, + { + "epoch": 0.01474, + "grad_norm": 0.8656063557310713, + "learning_rate": 0.003, + "loss": 4.3258, + "step": 1474 + }, + { + "epoch": 0.01475, + "grad_norm": 0.8637306499434225, + "learning_rate": 0.003, + "loss": 4.3398, + "step": 1475 + }, + { + "epoch": 0.01476, + "grad_norm": 0.6743755524690542, + "learning_rate": 0.003, + "loss": 4.3227, + "step": 1476 + }, + { + "epoch": 0.01477, + "grad_norm": 0.6414161546100536, + "learning_rate": 0.003, + "loss": 4.3256, + "step": 1477 + }, + { + "epoch": 0.01478, + "grad_norm": 0.7067557912668699, + "learning_rate": 0.003, + "loss": 4.3113, + "step": 1478 + }, + { + "epoch": 0.01479, + "grad_norm": 0.7427717638657159, + "learning_rate": 0.003, + "loss": 4.3499, + "step": 1479 + }, + { + "epoch": 0.0148, + "grad_norm": 0.7972486147312002, + "learning_rate": 0.003, + "loss": 4.3212, + "step": 1480 + }, + { + "epoch": 0.01481, + "grad_norm": 0.8238319142257903, + "learning_rate": 0.003, + "loss": 4.3261, + "step": 1481 + }, + { + "epoch": 0.01482, + "grad_norm": 0.7010334962030825, + "learning_rate": 0.003, + "loss": 4.3089, + "step": 1482 + }, + { + "epoch": 0.01483, + "grad_norm": 0.6287754589698554, + "learning_rate": 0.003, + "loss": 4.303, + "step": 1483 + }, + { + "epoch": 0.01484, + "grad_norm": 0.5953574922658559, + "learning_rate": 0.003, + "loss": 4.3534, + "step": 1484 + }, + { + "epoch": 0.01485, + "grad_norm": 0.6859671225358986, + "learning_rate": 0.003, + "loss": 4.3172, + "step": 1485 + }, + { + "epoch": 0.01486, + "grad_norm": 0.7393843625917039, + "learning_rate": 0.003, + "loss": 4.3181, + "step": 1486 + }, + { + "epoch": 0.01487, + "grad_norm": 0.6948333501030851, + "learning_rate": 0.003, + "loss": 4.3225, + "step": 1487 + }, + { + "epoch": 0.01488, + "grad_norm": 0.5806619385051136, + "learning_rate": 0.003, + "loss": 4.3259, + "step": 1488 + }, + { + "epoch": 0.01489, + "grad_norm": 0.6342989775695094, + "learning_rate": 0.003, + "loss": 4.3221, + "step": 1489 + }, + { + "epoch": 0.0149, + "grad_norm": 0.6704129451979153, + "learning_rate": 0.003, + "loss": 4.3147, + "step": 1490 + }, + { + "epoch": 0.01491, + "grad_norm": 0.7017206728292057, + "learning_rate": 0.003, + "loss": 4.3305, + "step": 1491 + }, + { + "epoch": 0.01492, + "grad_norm": 0.6813188063148864, + "learning_rate": 0.003, + "loss": 4.2961, + "step": 1492 + }, + { + "epoch": 0.01493, + "grad_norm": 0.8525559699799811, + "learning_rate": 0.003, + "loss": 4.3324, + "step": 1493 + }, + { + "epoch": 0.01494, + "grad_norm": 0.8687834663259932, + "learning_rate": 0.003, + "loss": 4.3213, + "step": 1494 + }, + { + "epoch": 0.01495, + "grad_norm": 0.7274136348295218, + "learning_rate": 0.003, + "loss": 4.3243, + "step": 1495 + }, + { + "epoch": 0.01496, + "grad_norm": 0.5434712910501038, + "learning_rate": 0.003, + "loss": 4.2923, + "step": 1496 + }, + { + "epoch": 0.01497, + "grad_norm": 0.6148760530691155, + "learning_rate": 0.003, + "loss": 4.2701, + "step": 1497 + }, + { + "epoch": 0.01498, + "grad_norm": 0.5665585769325068, + "learning_rate": 0.003, + "loss": 4.3224, + "step": 1498 + }, + { + "epoch": 0.01499, + "grad_norm": 0.543044486095261, + "learning_rate": 0.003, + "loss": 4.3408, + "step": 1499 + }, + { + "epoch": 0.015, + "grad_norm": 0.5868245294183819, + "learning_rate": 0.003, + "loss": 4.2952, + "step": 1500 + }, + { + "epoch": 0.01501, + "grad_norm": 0.6488345831013783, + "learning_rate": 0.003, + "loss": 4.2943, + "step": 1501 + }, + { + "epoch": 0.01502, + "grad_norm": 0.5560720951051904, + "learning_rate": 0.003, + "loss": 4.3322, + "step": 1502 + }, + { + "epoch": 0.01503, + "grad_norm": 0.6170651368212409, + "learning_rate": 0.003, + "loss": 4.3142, + "step": 1503 + }, + { + "epoch": 0.01504, + "grad_norm": 0.7253173696408437, + "learning_rate": 0.003, + "loss": 4.3199, + "step": 1504 + }, + { + "epoch": 0.01505, + "grad_norm": 0.7573152828389973, + "learning_rate": 0.003, + "loss": 4.3067, + "step": 1505 + }, + { + "epoch": 0.01506, + "grad_norm": 0.7119425101662294, + "learning_rate": 0.003, + "loss": 4.3233, + "step": 1506 + }, + { + "epoch": 0.01507, + "grad_norm": 0.7408719494100054, + "learning_rate": 0.003, + "loss": 4.3394, + "step": 1507 + }, + { + "epoch": 0.01508, + "grad_norm": 0.6023053184100382, + "learning_rate": 0.003, + "loss": 4.3108, + "step": 1508 + }, + { + "epoch": 0.01509, + "grad_norm": 0.5054062616818538, + "learning_rate": 0.003, + "loss": 4.3257, + "step": 1509 + }, + { + "epoch": 0.0151, + "grad_norm": 0.4440427822667444, + "learning_rate": 0.003, + "loss": 4.2952, + "step": 1510 + }, + { + "epoch": 0.01511, + "grad_norm": 0.503043960784533, + "learning_rate": 0.003, + "loss": 4.2838, + "step": 1511 + }, + { + "epoch": 0.01512, + "grad_norm": 0.4478910644887885, + "learning_rate": 0.003, + "loss": 4.2937, + "step": 1512 + }, + { + "epoch": 0.01513, + "grad_norm": 0.46313433916934627, + "learning_rate": 0.003, + "loss": 4.2989, + "step": 1513 + }, + { + "epoch": 0.01514, + "grad_norm": 0.5237062692888148, + "learning_rate": 0.003, + "loss": 4.3066, + "step": 1514 + }, + { + "epoch": 0.01515, + "grad_norm": 0.6686824710181497, + "learning_rate": 0.003, + "loss": 4.3056, + "step": 1515 + }, + { + "epoch": 0.01516, + "grad_norm": 0.8162150436928566, + "learning_rate": 0.003, + "loss": 4.3198, + "step": 1516 + }, + { + "epoch": 0.01517, + "grad_norm": 0.8250386268904906, + "learning_rate": 0.003, + "loss": 4.274, + "step": 1517 + }, + { + "epoch": 0.01518, + "grad_norm": 0.8551846044396628, + "learning_rate": 0.003, + "loss": 4.3102, + "step": 1518 + }, + { + "epoch": 0.01519, + "grad_norm": 0.7778994522735176, + "learning_rate": 0.003, + "loss": 4.3471, + "step": 1519 + }, + { + "epoch": 0.0152, + "grad_norm": 0.6407333772962789, + "learning_rate": 0.003, + "loss": 4.2901, + "step": 1520 + }, + { + "epoch": 0.01521, + "grad_norm": 0.590099483112598, + "learning_rate": 0.003, + "loss": 4.3053, + "step": 1521 + }, + { + "epoch": 0.01522, + "grad_norm": 0.6220605799864671, + "learning_rate": 0.003, + "loss": 4.3014, + "step": 1522 + }, + { + "epoch": 0.01523, + "grad_norm": 0.6378640000910417, + "learning_rate": 0.003, + "loss": 4.2957, + "step": 1523 + }, + { + "epoch": 0.01524, + "grad_norm": 0.6800153761279158, + "learning_rate": 0.003, + "loss": 4.3332, + "step": 1524 + }, + { + "epoch": 0.01525, + "grad_norm": 0.6552309843606878, + "learning_rate": 0.003, + "loss": 4.3306, + "step": 1525 + }, + { + "epoch": 0.01526, + "grad_norm": 0.5832236683536917, + "learning_rate": 0.003, + "loss": 4.3154, + "step": 1526 + }, + { + "epoch": 0.01527, + "grad_norm": 0.5923964968536319, + "learning_rate": 0.003, + "loss": 4.2868, + "step": 1527 + }, + { + "epoch": 0.01528, + "grad_norm": 0.6438689697714796, + "learning_rate": 0.003, + "loss": 4.3067, + "step": 1528 + }, + { + "epoch": 0.01529, + "grad_norm": 0.7863523458278316, + "learning_rate": 0.003, + "loss": 4.3071, + "step": 1529 + }, + { + "epoch": 0.0153, + "grad_norm": 0.8716684688082496, + "learning_rate": 0.003, + "loss": 4.3236, + "step": 1530 + }, + { + "epoch": 0.01531, + "grad_norm": 0.7632243352103186, + "learning_rate": 0.003, + "loss": 4.314, + "step": 1531 + }, + { + "epoch": 0.01532, + "grad_norm": 0.5830764209700208, + "learning_rate": 0.003, + "loss": 4.2863, + "step": 1532 + }, + { + "epoch": 0.01533, + "grad_norm": 0.6925578088506626, + "learning_rate": 0.003, + "loss": 4.3193, + "step": 1533 + }, + { + "epoch": 0.01534, + "grad_norm": 0.6884106279584329, + "learning_rate": 0.003, + "loss": 4.3131, + "step": 1534 + }, + { + "epoch": 0.01535, + "grad_norm": 0.612086877528724, + "learning_rate": 0.003, + "loss": 4.3216, + "step": 1535 + }, + { + "epoch": 0.01536, + "grad_norm": 0.6277711721176384, + "learning_rate": 0.003, + "loss": 4.3039, + "step": 1536 + }, + { + "epoch": 0.01537, + "grad_norm": 0.7930101311629847, + "learning_rate": 0.003, + "loss": 4.3078, + "step": 1537 + }, + { + "epoch": 0.01538, + "grad_norm": 0.8987557100700295, + "learning_rate": 0.003, + "loss": 4.2971, + "step": 1538 + }, + { + "epoch": 0.01539, + "grad_norm": 0.9391352557198052, + "learning_rate": 0.003, + "loss": 4.334, + "step": 1539 + }, + { + "epoch": 0.0154, + "grad_norm": 0.9440902063787158, + "learning_rate": 0.003, + "loss": 4.3487, + "step": 1540 + }, + { + "epoch": 0.01541, + "grad_norm": 0.9127476278073733, + "learning_rate": 0.003, + "loss": 4.328, + "step": 1541 + }, + { + "epoch": 0.01542, + "grad_norm": 0.7753546344208377, + "learning_rate": 0.003, + "loss": 4.3528, + "step": 1542 + }, + { + "epoch": 0.01543, + "grad_norm": 0.7317390533493494, + "learning_rate": 0.003, + "loss": 4.3318, + "step": 1543 + }, + { + "epoch": 0.01544, + "grad_norm": 0.5946634053593371, + "learning_rate": 0.003, + "loss": 4.3162, + "step": 1544 + }, + { + "epoch": 0.01545, + "grad_norm": 0.49808121797063876, + "learning_rate": 0.003, + "loss": 4.375, + "step": 1545 + }, + { + "epoch": 0.01546, + "grad_norm": 0.5189277397467036, + "learning_rate": 0.003, + "loss": 4.3435, + "step": 1546 + }, + { + "epoch": 0.01547, + "grad_norm": 0.5075409907094708, + "learning_rate": 0.003, + "loss": 4.3431, + "step": 1547 + }, + { + "epoch": 0.01548, + "grad_norm": 0.4269782583856954, + "learning_rate": 0.003, + "loss": 4.3016, + "step": 1548 + }, + { + "epoch": 0.01549, + "grad_norm": 0.44259739876123444, + "learning_rate": 0.003, + "loss": 4.2898, + "step": 1549 + }, + { + "epoch": 0.0155, + "grad_norm": 0.39476104723697186, + "learning_rate": 0.003, + "loss": 4.3072, + "step": 1550 + }, + { + "epoch": 0.01551, + "grad_norm": 0.39719568498427116, + "learning_rate": 0.003, + "loss": 4.2836, + "step": 1551 + }, + { + "epoch": 0.01552, + "grad_norm": 0.5112856286097273, + "learning_rate": 0.003, + "loss": 4.3059, + "step": 1552 + }, + { + "epoch": 0.01553, + "grad_norm": 0.6994822828905559, + "learning_rate": 0.003, + "loss": 4.3366, + "step": 1553 + }, + { + "epoch": 0.01554, + "grad_norm": 0.9918562900778292, + "learning_rate": 0.003, + "loss": 4.3337, + "step": 1554 + }, + { + "epoch": 0.01555, + "grad_norm": 0.8597749628867734, + "learning_rate": 0.003, + "loss": 4.3154, + "step": 1555 + }, + { + "epoch": 0.01556, + "grad_norm": 0.4859845337872918, + "learning_rate": 0.003, + "loss": 4.3299, + "step": 1556 + }, + { + "epoch": 0.01557, + "grad_norm": 0.8105266239108289, + "learning_rate": 0.003, + "loss": 4.317, + "step": 1557 + }, + { + "epoch": 0.01558, + "grad_norm": 0.6797261746777549, + "learning_rate": 0.003, + "loss": 4.3032, + "step": 1558 + }, + { + "epoch": 0.01559, + "grad_norm": 0.5540991472031954, + "learning_rate": 0.003, + "loss": 4.3349, + "step": 1559 + }, + { + "epoch": 0.0156, + "grad_norm": 0.5222478848499021, + "learning_rate": 0.003, + "loss": 4.3207, + "step": 1560 + }, + { + "epoch": 0.01561, + "grad_norm": 0.4547690370956997, + "learning_rate": 0.003, + "loss": 4.3006, + "step": 1561 + }, + { + "epoch": 0.01562, + "grad_norm": 0.3960981969234329, + "learning_rate": 0.003, + "loss": 4.2916, + "step": 1562 + }, + { + "epoch": 0.01563, + "grad_norm": 0.4365008440994641, + "learning_rate": 0.003, + "loss": 4.2998, + "step": 1563 + }, + { + "epoch": 0.01564, + "grad_norm": 0.4153235041333595, + "learning_rate": 0.003, + "loss": 4.3237, + "step": 1564 + }, + { + "epoch": 0.01565, + "grad_norm": 0.4172430370227481, + "learning_rate": 0.003, + "loss": 4.2984, + "step": 1565 + }, + { + "epoch": 0.01566, + "grad_norm": 0.4030631404816366, + "learning_rate": 0.003, + "loss": 4.3153, + "step": 1566 + }, + { + "epoch": 0.01567, + "grad_norm": 0.4006086575381243, + "learning_rate": 0.003, + "loss": 4.3082, + "step": 1567 + }, + { + "epoch": 0.01568, + "grad_norm": 0.4692595605101829, + "learning_rate": 0.003, + "loss": 4.2921, + "step": 1568 + }, + { + "epoch": 0.01569, + "grad_norm": 0.5156092105425303, + "learning_rate": 0.003, + "loss": 4.2975, + "step": 1569 + }, + { + "epoch": 0.0157, + "grad_norm": 0.5574730520335043, + "learning_rate": 0.003, + "loss": 4.3166, + "step": 1570 + }, + { + "epoch": 0.01571, + "grad_norm": 0.7147942584489586, + "learning_rate": 0.003, + "loss": 4.306, + "step": 1571 + }, + { + "epoch": 0.01572, + "grad_norm": 0.8176596681794123, + "learning_rate": 0.003, + "loss": 4.3202, + "step": 1572 + }, + { + "epoch": 0.01573, + "grad_norm": 0.8859299700656846, + "learning_rate": 0.003, + "loss": 4.311, + "step": 1573 + }, + { + "epoch": 0.01574, + "grad_norm": 1.0167563301248608, + "learning_rate": 0.003, + "loss": 4.3365, + "step": 1574 + }, + { + "epoch": 0.01575, + "grad_norm": 0.9276915846212054, + "learning_rate": 0.003, + "loss": 4.3012, + "step": 1575 + }, + { + "epoch": 0.01576, + "grad_norm": 0.7519027259067711, + "learning_rate": 0.003, + "loss": 4.317, + "step": 1576 + }, + { + "epoch": 0.01577, + "grad_norm": 0.5990656751852772, + "learning_rate": 0.003, + "loss": 4.2743, + "step": 1577 + }, + { + "epoch": 0.01578, + "grad_norm": 0.652141379980981, + "learning_rate": 0.003, + "loss": 4.3047, + "step": 1578 + }, + { + "epoch": 0.01579, + "grad_norm": 0.6130941880481341, + "learning_rate": 0.003, + "loss": 4.3001, + "step": 1579 + }, + { + "epoch": 0.0158, + "grad_norm": 0.6566738181349618, + "learning_rate": 0.003, + "loss": 4.3101, + "step": 1580 + }, + { + "epoch": 0.01581, + "grad_norm": 0.5602300919728743, + "learning_rate": 0.003, + "loss": 4.3154, + "step": 1581 + }, + { + "epoch": 0.01582, + "grad_norm": 0.6614515024819504, + "learning_rate": 0.003, + "loss": 4.3018, + "step": 1582 + }, + { + "epoch": 0.01583, + "grad_norm": 0.6685116238668469, + "learning_rate": 0.003, + "loss": 4.2884, + "step": 1583 + }, + { + "epoch": 0.01584, + "grad_norm": 0.7663983884039552, + "learning_rate": 0.003, + "loss": 4.3227, + "step": 1584 + }, + { + "epoch": 0.01585, + "grad_norm": 0.8471607724061854, + "learning_rate": 0.003, + "loss": 4.3183, + "step": 1585 + }, + { + "epoch": 0.01586, + "grad_norm": 0.8029205674223586, + "learning_rate": 0.003, + "loss": 4.2999, + "step": 1586 + }, + { + "epoch": 0.01587, + "grad_norm": 0.7335558361580696, + "learning_rate": 0.003, + "loss": 4.3092, + "step": 1587 + }, + { + "epoch": 0.01588, + "grad_norm": 0.6576426322376321, + "learning_rate": 0.003, + "loss": 4.3341, + "step": 1588 + }, + { + "epoch": 0.01589, + "grad_norm": 0.7314604361721292, + "learning_rate": 0.003, + "loss": 4.328, + "step": 1589 + }, + { + "epoch": 0.0159, + "grad_norm": 0.7484813941023133, + "learning_rate": 0.003, + "loss": 4.323, + "step": 1590 + }, + { + "epoch": 0.01591, + "grad_norm": 0.6836450786052791, + "learning_rate": 0.003, + "loss": 4.3264, + "step": 1591 + }, + { + "epoch": 0.01592, + "grad_norm": 0.6522043667699231, + "learning_rate": 0.003, + "loss": 4.3127, + "step": 1592 + }, + { + "epoch": 0.01593, + "grad_norm": 0.6311012555499182, + "learning_rate": 0.003, + "loss": 4.3174, + "step": 1593 + }, + { + "epoch": 0.01594, + "grad_norm": 0.5510623120073839, + "learning_rate": 0.003, + "loss": 4.3009, + "step": 1594 + }, + { + "epoch": 0.01595, + "grad_norm": 0.48793948155833994, + "learning_rate": 0.003, + "loss": 4.3248, + "step": 1595 + }, + { + "epoch": 0.01596, + "grad_norm": 0.42717068818596066, + "learning_rate": 0.003, + "loss": 4.3142, + "step": 1596 + }, + { + "epoch": 0.01597, + "grad_norm": 0.43677376148637953, + "learning_rate": 0.003, + "loss": 4.2773, + "step": 1597 + }, + { + "epoch": 0.01598, + "grad_norm": 0.47981748858846524, + "learning_rate": 0.003, + "loss": 4.2706, + "step": 1598 + }, + { + "epoch": 0.01599, + "grad_norm": 0.5068183957178387, + "learning_rate": 0.003, + "loss": 4.2944, + "step": 1599 + }, + { + "epoch": 0.016, + "grad_norm": 0.570411039839509, + "learning_rate": 0.003, + "loss": 4.261, + "step": 1600 + }, + { + "epoch": 0.01601, + "grad_norm": 0.7428825863420937, + "learning_rate": 0.003, + "loss": 4.3164, + "step": 1601 + }, + { + "epoch": 0.01602, + "grad_norm": 1.0316614845943484, + "learning_rate": 0.003, + "loss": 4.3056, + "step": 1602 + }, + { + "epoch": 0.01603, + "grad_norm": 0.9265092099146023, + "learning_rate": 0.003, + "loss": 4.3273, + "step": 1603 + }, + { + "epoch": 0.01604, + "grad_norm": 0.6714138671323193, + "learning_rate": 0.003, + "loss": 4.3099, + "step": 1604 + }, + { + "epoch": 0.01605, + "grad_norm": 0.7449186454711473, + "learning_rate": 0.003, + "loss": 4.2937, + "step": 1605 + }, + { + "epoch": 0.01606, + "grad_norm": 0.6551508425933098, + "learning_rate": 0.003, + "loss": 4.29, + "step": 1606 + }, + { + "epoch": 0.01607, + "grad_norm": 0.6368386814210635, + "learning_rate": 0.003, + "loss": 4.3196, + "step": 1607 + }, + { + "epoch": 0.01608, + "grad_norm": 0.6007545339051799, + "learning_rate": 0.003, + "loss": 4.287, + "step": 1608 + }, + { + "epoch": 0.01609, + "grad_norm": 0.5685481620694516, + "learning_rate": 0.003, + "loss": 4.2973, + "step": 1609 + }, + { + "epoch": 0.0161, + "grad_norm": 0.5084184586189073, + "learning_rate": 0.003, + "loss": 4.2592, + "step": 1610 + }, + { + "epoch": 0.01611, + "grad_norm": 0.4840833495127669, + "learning_rate": 0.003, + "loss": 4.2802, + "step": 1611 + }, + { + "epoch": 0.01612, + "grad_norm": 0.48942028074279065, + "learning_rate": 0.003, + "loss": 4.2702, + "step": 1612 + }, + { + "epoch": 0.01613, + "grad_norm": 0.5011285237133881, + "learning_rate": 0.003, + "loss": 4.315, + "step": 1613 + }, + { + "epoch": 0.01614, + "grad_norm": 0.43995362094592044, + "learning_rate": 0.003, + "loss": 4.3364, + "step": 1614 + }, + { + "epoch": 0.01615, + "grad_norm": 0.42434808753845177, + "learning_rate": 0.003, + "loss": 4.3023, + "step": 1615 + }, + { + "epoch": 0.01616, + "grad_norm": 0.40673555464346545, + "learning_rate": 0.003, + "loss": 4.2826, + "step": 1616 + }, + { + "epoch": 0.01617, + "grad_norm": 0.3578337742407294, + "learning_rate": 0.003, + "loss": 4.2706, + "step": 1617 + }, + { + "epoch": 0.01618, + "grad_norm": 0.41081917922167766, + "learning_rate": 0.003, + "loss": 4.2753, + "step": 1618 + }, + { + "epoch": 0.01619, + "grad_norm": 0.47886760935645833, + "learning_rate": 0.003, + "loss": 4.2659, + "step": 1619 + }, + { + "epoch": 0.0162, + "grad_norm": 0.6222321239813, + "learning_rate": 0.003, + "loss": 4.2767, + "step": 1620 + }, + { + "epoch": 0.01621, + "grad_norm": 0.6903491585345826, + "learning_rate": 0.003, + "loss": 4.2786, + "step": 1621 + }, + { + "epoch": 0.01622, + "grad_norm": 0.7573622136296652, + "learning_rate": 0.003, + "loss": 4.3118, + "step": 1622 + }, + { + "epoch": 0.01623, + "grad_norm": 0.9256995453444036, + "learning_rate": 0.003, + "loss": 4.3026, + "step": 1623 + }, + { + "epoch": 0.01624, + "grad_norm": 0.9558305805696413, + "learning_rate": 0.003, + "loss": 4.3216, + "step": 1624 + }, + { + "epoch": 0.01625, + "grad_norm": 0.8114385487348762, + "learning_rate": 0.003, + "loss": 4.2898, + "step": 1625 + }, + { + "epoch": 0.01626, + "grad_norm": 0.8322826717160557, + "learning_rate": 0.003, + "loss": 4.3177, + "step": 1626 + }, + { + "epoch": 0.01627, + "grad_norm": 0.7727713367753223, + "learning_rate": 0.003, + "loss": 4.2816, + "step": 1627 + }, + { + "epoch": 0.01628, + "grad_norm": 0.7109224896248458, + "learning_rate": 0.003, + "loss": 4.2929, + "step": 1628 + }, + { + "epoch": 0.01629, + "grad_norm": 0.560897124973334, + "learning_rate": 0.003, + "loss": 4.2858, + "step": 1629 + }, + { + "epoch": 0.0163, + "grad_norm": 0.6779380557060044, + "learning_rate": 0.003, + "loss": 4.2826, + "step": 1630 + }, + { + "epoch": 0.01631, + "grad_norm": 0.6886638086492523, + "learning_rate": 0.003, + "loss": 4.3011, + "step": 1631 + }, + { + "epoch": 0.01632, + "grad_norm": 0.6589349584439033, + "learning_rate": 0.003, + "loss": 4.3124, + "step": 1632 + }, + { + "epoch": 0.01633, + "grad_norm": 0.7155276302347654, + "learning_rate": 0.003, + "loss": 4.2893, + "step": 1633 + }, + { + "epoch": 0.01634, + "grad_norm": 0.7783909364325312, + "learning_rate": 0.003, + "loss": 4.3094, + "step": 1634 + }, + { + "epoch": 0.01635, + "grad_norm": 0.8288192042951168, + "learning_rate": 0.003, + "loss": 4.2952, + "step": 1635 + }, + { + "epoch": 0.01636, + "grad_norm": 0.8309682765627266, + "learning_rate": 0.003, + "loss": 4.3157, + "step": 1636 + }, + { + "epoch": 0.01637, + "grad_norm": 0.7885672766922769, + "learning_rate": 0.003, + "loss": 4.2985, + "step": 1637 + }, + { + "epoch": 0.01638, + "grad_norm": 0.6820787883112478, + "learning_rate": 0.003, + "loss": 4.3026, + "step": 1638 + }, + { + "epoch": 0.01639, + "grad_norm": 0.5912531628997083, + "learning_rate": 0.003, + "loss": 4.3221, + "step": 1639 + }, + { + "epoch": 0.0164, + "grad_norm": 0.7081315099543188, + "learning_rate": 0.003, + "loss": 4.3062, + "step": 1640 + }, + { + "epoch": 0.01641, + "grad_norm": 0.7390837101488571, + "learning_rate": 0.003, + "loss": 4.279, + "step": 1641 + }, + { + "epoch": 0.01642, + "grad_norm": 0.8019499352702314, + "learning_rate": 0.003, + "loss": 4.3244, + "step": 1642 + }, + { + "epoch": 0.01643, + "grad_norm": 0.7922382978586373, + "learning_rate": 0.003, + "loss": 4.3126, + "step": 1643 + }, + { + "epoch": 0.01644, + "grad_norm": 0.7260441686104165, + "learning_rate": 0.003, + "loss": 4.3282, + "step": 1644 + }, + { + "epoch": 0.01645, + "grad_norm": 0.5765166219346322, + "learning_rate": 0.003, + "loss": 4.2686, + "step": 1645 + }, + { + "epoch": 0.01646, + "grad_norm": 0.4846869130781467, + "learning_rate": 0.003, + "loss": 4.298, + "step": 1646 + }, + { + "epoch": 0.01647, + "grad_norm": 0.4324516045425632, + "learning_rate": 0.003, + "loss": 4.294, + "step": 1647 + }, + { + "epoch": 0.01648, + "grad_norm": 0.45801960619516824, + "learning_rate": 0.003, + "loss": 4.2725, + "step": 1648 + }, + { + "epoch": 0.01649, + "grad_norm": 0.4519054523297515, + "learning_rate": 0.003, + "loss": 4.3034, + "step": 1649 + }, + { + "epoch": 0.0165, + "grad_norm": 0.5030069407293299, + "learning_rate": 0.003, + "loss": 4.2846, + "step": 1650 + }, + { + "epoch": 0.01651, + "grad_norm": 0.7338143580585036, + "learning_rate": 0.003, + "loss": 4.2988, + "step": 1651 + }, + { + "epoch": 0.01652, + "grad_norm": 0.9179472565177694, + "learning_rate": 0.003, + "loss": 4.2688, + "step": 1652 + }, + { + "epoch": 0.01653, + "grad_norm": 0.7974402781604697, + "learning_rate": 0.003, + "loss": 4.307, + "step": 1653 + }, + { + "epoch": 0.01654, + "grad_norm": 0.8094858670947136, + "learning_rate": 0.003, + "loss": 4.2972, + "step": 1654 + }, + { + "epoch": 0.01655, + "grad_norm": 0.8204525750124729, + "learning_rate": 0.003, + "loss": 4.313, + "step": 1655 + }, + { + "epoch": 0.01656, + "grad_norm": 0.6293302149986666, + "learning_rate": 0.003, + "loss": 4.271, + "step": 1656 + }, + { + "epoch": 0.01657, + "grad_norm": 0.5706278835378605, + "learning_rate": 0.003, + "loss": 4.293, + "step": 1657 + }, + { + "epoch": 0.01658, + "grad_norm": 0.6110254736121277, + "learning_rate": 0.003, + "loss": 4.2809, + "step": 1658 + }, + { + "epoch": 0.01659, + "grad_norm": 0.4984603935459828, + "learning_rate": 0.003, + "loss": 4.2824, + "step": 1659 + }, + { + "epoch": 0.0166, + "grad_norm": 0.46804303085612786, + "learning_rate": 0.003, + "loss": 4.2649, + "step": 1660 + }, + { + "epoch": 0.01661, + "grad_norm": 0.48879681023532395, + "learning_rate": 0.003, + "loss": 4.2671, + "step": 1661 + }, + { + "epoch": 0.01662, + "grad_norm": 0.5571765680768687, + "learning_rate": 0.003, + "loss": 4.2719, + "step": 1662 + }, + { + "epoch": 0.01663, + "grad_norm": 0.532882366154707, + "learning_rate": 0.003, + "loss": 4.294, + "step": 1663 + }, + { + "epoch": 0.01664, + "grad_norm": 0.5088338427561463, + "learning_rate": 0.003, + "loss": 4.3094, + "step": 1664 + }, + { + "epoch": 0.01665, + "grad_norm": 0.4963301596331138, + "learning_rate": 0.003, + "loss": 4.2712, + "step": 1665 + }, + { + "epoch": 0.01666, + "grad_norm": 0.5472162068898109, + "learning_rate": 0.003, + "loss": 4.2732, + "step": 1666 + }, + { + "epoch": 0.01667, + "grad_norm": 0.6863444215859392, + "learning_rate": 0.003, + "loss": 4.2951, + "step": 1667 + }, + { + "epoch": 0.01668, + "grad_norm": 0.7945747889757943, + "learning_rate": 0.003, + "loss": 4.3019, + "step": 1668 + }, + { + "epoch": 0.01669, + "grad_norm": 0.7029756676891371, + "learning_rate": 0.003, + "loss": 4.2868, + "step": 1669 + }, + { + "epoch": 0.0167, + "grad_norm": 0.564779057565769, + "learning_rate": 0.003, + "loss": 4.2701, + "step": 1670 + }, + { + "epoch": 0.01671, + "grad_norm": 0.5946383928046679, + "learning_rate": 0.003, + "loss": 4.2602, + "step": 1671 + }, + { + "epoch": 0.01672, + "grad_norm": 0.7328231067257597, + "learning_rate": 0.003, + "loss": 4.2779, + "step": 1672 + }, + { + "epoch": 0.01673, + "grad_norm": 0.7730808965686631, + "learning_rate": 0.003, + "loss": 4.2759, + "step": 1673 + }, + { + "epoch": 0.01674, + "grad_norm": 0.8250670120229329, + "learning_rate": 0.003, + "loss": 4.3013, + "step": 1674 + }, + { + "epoch": 0.01675, + "grad_norm": 0.7145691985141881, + "learning_rate": 0.003, + "loss": 4.2806, + "step": 1675 + }, + { + "epoch": 0.01676, + "grad_norm": 0.6268084931910631, + "learning_rate": 0.003, + "loss": 4.2872, + "step": 1676 + }, + { + "epoch": 0.01677, + "grad_norm": 0.6170941899124056, + "learning_rate": 0.003, + "loss": 4.2772, + "step": 1677 + }, + { + "epoch": 0.01678, + "grad_norm": 0.5951034196601164, + "learning_rate": 0.003, + "loss": 4.2879, + "step": 1678 + }, + { + "epoch": 0.01679, + "grad_norm": 0.5254502211849061, + "learning_rate": 0.003, + "loss": 4.2679, + "step": 1679 + }, + { + "epoch": 0.0168, + "grad_norm": 0.6096842133031645, + "learning_rate": 0.003, + "loss": 4.2892, + "step": 1680 + }, + { + "epoch": 0.01681, + "grad_norm": 0.6218177000062469, + "learning_rate": 0.003, + "loss": 4.2892, + "step": 1681 + }, + { + "epoch": 0.01682, + "grad_norm": 0.5832829705689185, + "learning_rate": 0.003, + "loss": 4.2817, + "step": 1682 + }, + { + "epoch": 0.01683, + "grad_norm": 0.5910768318994952, + "learning_rate": 0.003, + "loss": 4.2959, + "step": 1683 + }, + { + "epoch": 0.01684, + "grad_norm": 0.6050344929193974, + "learning_rate": 0.003, + "loss": 4.2707, + "step": 1684 + }, + { + "epoch": 0.01685, + "grad_norm": 0.6382986956510986, + "learning_rate": 0.003, + "loss": 4.2898, + "step": 1685 + }, + { + "epoch": 0.01686, + "grad_norm": 0.8077022270811443, + "learning_rate": 0.003, + "loss": 4.2942, + "step": 1686 + }, + { + "epoch": 0.01687, + "grad_norm": 0.8169553039156104, + "learning_rate": 0.003, + "loss": 4.2925, + "step": 1687 + }, + { + "epoch": 0.01688, + "grad_norm": 0.8368087154638648, + "learning_rate": 0.003, + "loss": 4.2843, + "step": 1688 + }, + { + "epoch": 0.01689, + "grad_norm": 0.869445348031783, + "learning_rate": 0.003, + "loss": 4.3002, + "step": 1689 + }, + { + "epoch": 0.0169, + "grad_norm": 0.9062293741017381, + "learning_rate": 0.003, + "loss": 4.2906, + "step": 1690 + }, + { + "epoch": 0.01691, + "grad_norm": 0.8487759964259872, + "learning_rate": 0.003, + "loss": 4.3075, + "step": 1691 + }, + { + "epoch": 0.01692, + "grad_norm": 0.8483641489848167, + "learning_rate": 0.003, + "loss": 4.3088, + "step": 1692 + }, + { + "epoch": 0.01693, + "grad_norm": 0.8279189520542338, + "learning_rate": 0.003, + "loss": 4.3369, + "step": 1693 + }, + { + "epoch": 0.01694, + "grad_norm": 0.8089873138093994, + "learning_rate": 0.003, + "loss": 4.2887, + "step": 1694 + }, + { + "epoch": 0.01695, + "grad_norm": 0.9455670720071184, + "learning_rate": 0.003, + "loss": 4.3248, + "step": 1695 + }, + { + "epoch": 0.01696, + "grad_norm": 1.064016692244757, + "learning_rate": 0.003, + "loss": 4.3315, + "step": 1696 + }, + { + "epoch": 0.01697, + "grad_norm": 0.9264587036694925, + "learning_rate": 0.003, + "loss": 4.302, + "step": 1697 + }, + { + "epoch": 0.01698, + "grad_norm": 0.6997503897611733, + "learning_rate": 0.003, + "loss": 4.289, + "step": 1698 + }, + { + "epoch": 0.01699, + "grad_norm": 0.6394457308265625, + "learning_rate": 0.003, + "loss": 4.2969, + "step": 1699 + }, + { + "epoch": 0.017, + "grad_norm": 0.5891910529153762, + "learning_rate": 0.003, + "loss": 4.3094, + "step": 1700 + }, + { + "epoch": 0.01701, + "grad_norm": 0.6435550929707158, + "learning_rate": 0.003, + "loss": 4.2939, + "step": 1701 + }, + { + "epoch": 0.01702, + "grad_norm": 0.6810306253220882, + "learning_rate": 0.003, + "loss": 4.3163, + "step": 1702 + }, + { + "epoch": 0.01703, + "grad_norm": 0.6849363850272787, + "learning_rate": 0.003, + "loss": 4.3034, + "step": 1703 + }, + { + "epoch": 0.01704, + "grad_norm": 0.6708561417744424, + "learning_rate": 0.003, + "loss": 4.2869, + "step": 1704 + }, + { + "epoch": 0.01705, + "grad_norm": 0.5968699215981799, + "learning_rate": 0.003, + "loss": 4.2879, + "step": 1705 + }, + { + "epoch": 0.01706, + "grad_norm": 0.505702133399309, + "learning_rate": 0.003, + "loss": 4.2792, + "step": 1706 + }, + { + "epoch": 0.01707, + "grad_norm": 0.5302615087387699, + "learning_rate": 0.003, + "loss": 4.3011, + "step": 1707 + }, + { + "epoch": 0.01708, + "grad_norm": 0.5210003345198112, + "learning_rate": 0.003, + "loss": 4.2762, + "step": 1708 + }, + { + "epoch": 0.01709, + "grad_norm": 0.46584995061994, + "learning_rate": 0.003, + "loss": 4.2953, + "step": 1709 + }, + { + "epoch": 0.0171, + "grad_norm": 0.42345394468243425, + "learning_rate": 0.003, + "loss": 4.2414, + "step": 1710 + }, + { + "epoch": 0.01711, + "grad_norm": 0.43754438467909607, + "learning_rate": 0.003, + "loss": 4.2717, + "step": 1711 + }, + { + "epoch": 0.01712, + "grad_norm": 0.41267034274673536, + "learning_rate": 0.003, + "loss": 4.2801, + "step": 1712 + }, + { + "epoch": 0.01713, + "grad_norm": 0.3568885732041511, + "learning_rate": 0.003, + "loss": 4.2523, + "step": 1713 + }, + { + "epoch": 0.01714, + "grad_norm": 0.3457637645495226, + "learning_rate": 0.003, + "loss": 4.2876, + "step": 1714 + }, + { + "epoch": 0.01715, + "grad_norm": 0.3722023563902025, + "learning_rate": 0.003, + "loss": 4.2958, + "step": 1715 + }, + { + "epoch": 0.01716, + "grad_norm": 0.3390050605620623, + "learning_rate": 0.003, + "loss": 4.2697, + "step": 1716 + }, + { + "epoch": 0.01717, + "grad_norm": 0.4093688383515523, + "learning_rate": 0.003, + "loss": 4.3028, + "step": 1717 + }, + { + "epoch": 0.01718, + "grad_norm": 0.5445666194239754, + "learning_rate": 0.003, + "loss": 4.2735, + "step": 1718 + }, + { + "epoch": 0.01719, + "grad_norm": 0.8818554274614839, + "learning_rate": 0.003, + "loss": 4.3037, + "step": 1719 + }, + { + "epoch": 0.0172, + "grad_norm": 1.123998698583038, + "learning_rate": 0.003, + "loss": 4.2879, + "step": 1720 + }, + { + "epoch": 0.01721, + "grad_norm": 0.6518188394707584, + "learning_rate": 0.003, + "loss": 4.2815, + "step": 1721 + }, + { + "epoch": 0.01722, + "grad_norm": 0.7473961890284897, + "learning_rate": 0.003, + "loss": 4.281, + "step": 1722 + }, + { + "epoch": 0.01723, + "grad_norm": 0.7241082541176558, + "learning_rate": 0.003, + "loss": 4.2942, + "step": 1723 + }, + { + "epoch": 0.01724, + "grad_norm": 0.5406243852034243, + "learning_rate": 0.003, + "loss": 4.2839, + "step": 1724 + }, + { + "epoch": 0.01725, + "grad_norm": 0.7661047717663526, + "learning_rate": 0.003, + "loss": 4.2758, + "step": 1725 + }, + { + "epoch": 0.01726, + "grad_norm": 0.6519733535814134, + "learning_rate": 0.003, + "loss": 4.2678, + "step": 1726 + }, + { + "epoch": 0.01727, + "grad_norm": 0.5095447102684948, + "learning_rate": 0.003, + "loss": 4.2681, + "step": 1727 + }, + { + "epoch": 0.01728, + "grad_norm": 0.5716779713985857, + "learning_rate": 0.003, + "loss": 4.2703, + "step": 1728 + }, + { + "epoch": 0.01729, + "grad_norm": 0.5506488662519434, + "learning_rate": 0.003, + "loss": 4.2777, + "step": 1729 + }, + { + "epoch": 0.0173, + "grad_norm": 0.4864830783539529, + "learning_rate": 0.003, + "loss": 4.2498, + "step": 1730 + }, + { + "epoch": 0.01731, + "grad_norm": 0.49289174915320866, + "learning_rate": 0.003, + "loss": 4.2691, + "step": 1731 + }, + { + "epoch": 0.01732, + "grad_norm": 0.5258408981010323, + "learning_rate": 0.003, + "loss": 4.2782, + "step": 1732 + }, + { + "epoch": 0.01733, + "grad_norm": 0.5689089752248919, + "learning_rate": 0.003, + "loss": 4.2578, + "step": 1733 + }, + { + "epoch": 0.01734, + "grad_norm": 0.5515499021599949, + "learning_rate": 0.003, + "loss": 4.2704, + "step": 1734 + }, + { + "epoch": 0.01735, + "grad_norm": 0.5101191368780025, + "learning_rate": 0.003, + "loss": 4.2394, + "step": 1735 + }, + { + "epoch": 0.01736, + "grad_norm": 0.5538259418110674, + "learning_rate": 0.003, + "loss": 4.2405, + "step": 1736 + }, + { + "epoch": 0.01737, + "grad_norm": 0.6368154932756357, + "learning_rate": 0.003, + "loss": 4.273, + "step": 1737 + }, + { + "epoch": 0.01738, + "grad_norm": 0.6302527957263243, + "learning_rate": 0.003, + "loss": 4.2633, + "step": 1738 + }, + { + "epoch": 0.01739, + "grad_norm": 0.5659864529447296, + "learning_rate": 0.003, + "loss": 4.2757, + "step": 1739 + }, + { + "epoch": 0.0174, + "grad_norm": 0.6034857544005698, + "learning_rate": 0.003, + "loss": 4.278, + "step": 1740 + }, + { + "epoch": 0.01741, + "grad_norm": 0.7023281656557057, + "learning_rate": 0.003, + "loss": 4.2883, + "step": 1741 + }, + { + "epoch": 0.01742, + "grad_norm": 0.7127714842932543, + "learning_rate": 0.003, + "loss": 4.2766, + "step": 1742 + }, + { + "epoch": 0.01743, + "grad_norm": 0.7178940660267726, + "learning_rate": 0.003, + "loss": 4.2889, + "step": 1743 + }, + { + "epoch": 0.01744, + "grad_norm": 0.770440209657255, + "learning_rate": 0.003, + "loss": 4.2773, + "step": 1744 + }, + { + "epoch": 0.01745, + "grad_norm": 0.7362175602563142, + "learning_rate": 0.003, + "loss": 4.2484, + "step": 1745 + }, + { + "epoch": 0.01746, + "grad_norm": 0.7588375597390518, + "learning_rate": 0.003, + "loss": 4.2719, + "step": 1746 + }, + { + "epoch": 0.01747, + "grad_norm": 0.6839956521205998, + "learning_rate": 0.003, + "loss": 4.267, + "step": 1747 + }, + { + "epoch": 0.01748, + "grad_norm": 0.6088836411369196, + "learning_rate": 0.003, + "loss": 4.2555, + "step": 1748 + }, + { + "epoch": 0.01749, + "grad_norm": 0.548971540706334, + "learning_rate": 0.003, + "loss": 4.2516, + "step": 1749 + }, + { + "epoch": 0.0175, + "grad_norm": 0.5440778124239496, + "learning_rate": 0.003, + "loss": 4.2665, + "step": 1750 + }, + { + "epoch": 0.01751, + "grad_norm": 0.5918941393242653, + "learning_rate": 0.003, + "loss": 4.2593, + "step": 1751 + }, + { + "epoch": 0.01752, + "grad_norm": 0.6455620934648258, + "learning_rate": 0.003, + "loss": 4.3014, + "step": 1752 + }, + { + "epoch": 0.01753, + "grad_norm": 0.6781800073296501, + "learning_rate": 0.003, + "loss": 4.2535, + "step": 1753 + }, + { + "epoch": 0.01754, + "grad_norm": 0.6882980076107725, + "learning_rate": 0.003, + "loss": 4.2853, + "step": 1754 + }, + { + "epoch": 0.01755, + "grad_norm": 0.7961632146248975, + "learning_rate": 0.003, + "loss": 4.233, + "step": 1755 + }, + { + "epoch": 0.01756, + "grad_norm": 0.8494780429534167, + "learning_rate": 0.003, + "loss": 4.314, + "step": 1756 + }, + { + "epoch": 0.01757, + "grad_norm": 0.9294824362906122, + "learning_rate": 0.003, + "loss": 4.2987, + "step": 1757 + }, + { + "epoch": 0.01758, + "grad_norm": 0.852175947314296, + "learning_rate": 0.003, + "loss": 4.3126, + "step": 1758 + }, + { + "epoch": 0.01759, + "grad_norm": 0.7347749740307682, + "learning_rate": 0.003, + "loss": 4.2719, + "step": 1759 + }, + { + "epoch": 0.0176, + "grad_norm": 0.832142747460363, + "learning_rate": 0.003, + "loss": 4.2994, + "step": 1760 + }, + { + "epoch": 0.01761, + "grad_norm": 0.8686282951443787, + "learning_rate": 0.003, + "loss": 4.2965, + "step": 1761 + }, + { + "epoch": 0.01762, + "grad_norm": 0.9408046252281581, + "learning_rate": 0.003, + "loss": 4.3172, + "step": 1762 + }, + { + "epoch": 0.01763, + "grad_norm": 0.8594428065536988, + "learning_rate": 0.003, + "loss": 4.3092, + "step": 1763 + }, + { + "epoch": 0.01764, + "grad_norm": 0.7749341487974934, + "learning_rate": 0.003, + "loss": 4.2754, + "step": 1764 + }, + { + "epoch": 0.01765, + "grad_norm": 0.7897730382355919, + "learning_rate": 0.003, + "loss": 4.3076, + "step": 1765 + }, + { + "epoch": 0.01766, + "grad_norm": 0.7174878978412247, + "learning_rate": 0.003, + "loss": 4.317, + "step": 1766 + }, + { + "epoch": 0.01767, + "grad_norm": 0.7171956437495152, + "learning_rate": 0.003, + "loss": 4.2757, + "step": 1767 + }, + { + "epoch": 0.01768, + "grad_norm": 0.8077487259776668, + "learning_rate": 0.003, + "loss": 4.2866, + "step": 1768 + }, + { + "epoch": 0.01769, + "grad_norm": 0.7100605985326003, + "learning_rate": 0.003, + "loss": 4.2811, + "step": 1769 + }, + { + "epoch": 0.0177, + "grad_norm": 0.578964917508603, + "learning_rate": 0.003, + "loss": 4.2635, + "step": 1770 + }, + { + "epoch": 0.01771, + "grad_norm": 0.6086271828330341, + "learning_rate": 0.003, + "loss": 4.2929, + "step": 1771 + }, + { + "epoch": 0.01772, + "grad_norm": 0.497546957954685, + "learning_rate": 0.003, + "loss": 4.262, + "step": 1772 + }, + { + "epoch": 0.01773, + "grad_norm": 0.490787842661778, + "learning_rate": 0.003, + "loss": 4.2539, + "step": 1773 + }, + { + "epoch": 0.01774, + "grad_norm": 0.40528808868462435, + "learning_rate": 0.003, + "loss": 4.272, + "step": 1774 + }, + { + "epoch": 0.01775, + "grad_norm": 0.3996997018074339, + "learning_rate": 0.003, + "loss": 4.2574, + "step": 1775 + }, + { + "epoch": 0.01776, + "grad_norm": 0.40093115740510726, + "learning_rate": 0.003, + "loss": 4.2491, + "step": 1776 + }, + { + "epoch": 0.01777, + "grad_norm": 0.4577551703461772, + "learning_rate": 0.003, + "loss": 4.27, + "step": 1777 + }, + { + "epoch": 0.01778, + "grad_norm": 0.4894775985846586, + "learning_rate": 0.003, + "loss": 4.262, + "step": 1778 + }, + { + "epoch": 0.01779, + "grad_norm": 0.619715296261353, + "learning_rate": 0.003, + "loss": 4.2821, + "step": 1779 + }, + { + "epoch": 0.0178, + "grad_norm": 0.8092821046637155, + "learning_rate": 0.003, + "loss": 4.2576, + "step": 1780 + }, + { + "epoch": 0.01781, + "grad_norm": 1.007368746628607, + "learning_rate": 0.003, + "loss": 4.2833, + "step": 1781 + }, + { + "epoch": 0.01782, + "grad_norm": 1.0605028910459124, + "learning_rate": 0.003, + "loss": 4.314, + "step": 1782 + }, + { + "epoch": 0.01783, + "grad_norm": 0.8361916169047299, + "learning_rate": 0.003, + "loss": 4.2912, + "step": 1783 + }, + { + "epoch": 0.01784, + "grad_norm": 0.9950086095977337, + "learning_rate": 0.003, + "loss": 4.2963, + "step": 1784 + }, + { + "epoch": 0.01785, + "grad_norm": 1.0228877834258285, + "learning_rate": 0.003, + "loss": 4.2673, + "step": 1785 + }, + { + "epoch": 0.01786, + "grad_norm": 0.8766088407747356, + "learning_rate": 0.003, + "loss": 4.2899, + "step": 1786 + }, + { + "epoch": 0.01787, + "grad_norm": 0.9346024274483619, + "learning_rate": 0.003, + "loss": 4.2991, + "step": 1787 + }, + { + "epoch": 0.01788, + "grad_norm": 0.8655394124739175, + "learning_rate": 0.003, + "loss": 4.3097, + "step": 1788 + }, + { + "epoch": 0.01789, + "grad_norm": 0.7320722132470949, + "learning_rate": 0.003, + "loss": 4.2742, + "step": 1789 + }, + { + "epoch": 0.0179, + "grad_norm": 0.6043898276387406, + "learning_rate": 0.003, + "loss": 4.288, + "step": 1790 + }, + { + "epoch": 0.01791, + "grad_norm": 0.551731430114501, + "learning_rate": 0.003, + "loss": 4.2984, + "step": 1791 + }, + { + "epoch": 0.01792, + "grad_norm": 0.4116411374695503, + "learning_rate": 0.003, + "loss": 4.2944, + "step": 1792 + }, + { + "epoch": 0.01793, + "grad_norm": 0.4368353784765224, + "learning_rate": 0.003, + "loss": 4.2574, + "step": 1793 + }, + { + "epoch": 0.01794, + "grad_norm": 0.4134683249639472, + "learning_rate": 0.003, + "loss": 4.2696, + "step": 1794 + }, + { + "epoch": 0.01795, + "grad_norm": 0.3898886643379656, + "learning_rate": 0.003, + "loss": 4.2661, + "step": 1795 + }, + { + "epoch": 0.01796, + "grad_norm": 0.3690232703108766, + "learning_rate": 0.003, + "loss": 4.2797, + "step": 1796 + }, + { + "epoch": 0.01797, + "grad_norm": 0.3352292075992397, + "learning_rate": 0.003, + "loss": 4.2421, + "step": 1797 + }, + { + "epoch": 0.01798, + "grad_norm": 0.3577790623133374, + "learning_rate": 0.003, + "loss": 4.2547, + "step": 1798 + }, + { + "epoch": 0.01799, + "grad_norm": 0.40125449243674877, + "learning_rate": 0.003, + "loss": 4.269, + "step": 1799 + }, + { + "epoch": 0.018, + "grad_norm": 0.4677938601833102, + "learning_rate": 0.003, + "loss": 4.26, + "step": 1800 + }, + { + "epoch": 0.01801, + "grad_norm": 0.5610552208097703, + "learning_rate": 0.003, + "loss": 4.2554, + "step": 1801 + }, + { + "epoch": 0.01802, + "grad_norm": 0.5864236478232465, + "learning_rate": 0.003, + "loss": 4.2736, + "step": 1802 + }, + { + "epoch": 0.01803, + "grad_norm": 0.540429672022444, + "learning_rate": 0.003, + "loss": 4.2776, + "step": 1803 + }, + { + "epoch": 0.01804, + "grad_norm": 0.60459180760597, + "learning_rate": 0.003, + "loss": 4.2503, + "step": 1804 + }, + { + "epoch": 0.01805, + "grad_norm": 0.6945199909530504, + "learning_rate": 0.003, + "loss": 4.2727, + "step": 1805 + }, + { + "epoch": 0.01806, + "grad_norm": 0.6888751318415884, + "learning_rate": 0.003, + "loss": 4.2338, + "step": 1806 + }, + { + "epoch": 0.01807, + "grad_norm": 0.652641222784165, + "learning_rate": 0.003, + "loss": 4.2674, + "step": 1807 + }, + { + "epoch": 0.01808, + "grad_norm": 0.7163920878946057, + "learning_rate": 0.003, + "loss": 4.2613, + "step": 1808 + }, + { + "epoch": 0.01809, + "grad_norm": 0.7231527059402414, + "learning_rate": 0.003, + "loss": 4.284, + "step": 1809 + }, + { + "epoch": 0.0181, + "grad_norm": 0.7979758654578836, + "learning_rate": 0.003, + "loss": 4.2839, + "step": 1810 + }, + { + "epoch": 0.01811, + "grad_norm": 0.7630566976684018, + "learning_rate": 0.003, + "loss": 4.2692, + "step": 1811 + }, + { + "epoch": 0.01812, + "grad_norm": 0.68601985737994, + "learning_rate": 0.003, + "loss": 4.2545, + "step": 1812 + }, + { + "epoch": 0.01813, + "grad_norm": 0.601926611394952, + "learning_rate": 0.003, + "loss": 4.2949, + "step": 1813 + }, + { + "epoch": 0.01814, + "grad_norm": 0.66852657571521, + "learning_rate": 0.003, + "loss": 4.2647, + "step": 1814 + }, + { + "epoch": 0.01815, + "grad_norm": 0.6050905896281803, + "learning_rate": 0.003, + "loss": 4.2474, + "step": 1815 + }, + { + "epoch": 0.01816, + "grad_norm": 0.531953848915138, + "learning_rate": 0.003, + "loss": 4.2731, + "step": 1816 + }, + { + "epoch": 0.01817, + "grad_norm": 0.6652005043351588, + "learning_rate": 0.003, + "loss": 4.2676, + "step": 1817 + }, + { + "epoch": 0.01818, + "grad_norm": 0.6653492631522442, + "learning_rate": 0.003, + "loss": 4.2669, + "step": 1818 + }, + { + "epoch": 0.01819, + "grad_norm": 0.6430139858317634, + "learning_rate": 0.003, + "loss": 4.2522, + "step": 1819 + }, + { + "epoch": 0.0182, + "grad_norm": 0.7307512465670702, + "learning_rate": 0.003, + "loss": 4.2731, + "step": 1820 + }, + { + "epoch": 0.01821, + "grad_norm": 0.7117919312146421, + "learning_rate": 0.003, + "loss": 4.2608, + "step": 1821 + }, + { + "epoch": 0.01822, + "grad_norm": 0.6849992334976018, + "learning_rate": 0.003, + "loss": 4.2737, + "step": 1822 + }, + { + "epoch": 0.01823, + "grad_norm": 0.6437329965546797, + "learning_rate": 0.003, + "loss": 4.2861, + "step": 1823 + }, + { + "epoch": 0.01824, + "grad_norm": 0.6298055587951418, + "learning_rate": 0.003, + "loss": 4.2663, + "step": 1824 + }, + { + "epoch": 0.01825, + "grad_norm": 0.6105471946808345, + "learning_rate": 0.003, + "loss": 4.2706, + "step": 1825 + }, + { + "epoch": 0.01826, + "grad_norm": 0.5680736127986394, + "learning_rate": 0.003, + "loss": 4.2798, + "step": 1826 + }, + { + "epoch": 0.01827, + "grad_norm": 0.6203032072106797, + "learning_rate": 0.003, + "loss": 4.2493, + "step": 1827 + }, + { + "epoch": 0.01828, + "grad_norm": 0.5646483418870301, + "learning_rate": 0.003, + "loss": 4.2595, + "step": 1828 + }, + { + "epoch": 0.01829, + "grad_norm": 0.43449345506164244, + "learning_rate": 0.003, + "loss": 4.2724, + "step": 1829 + }, + { + "epoch": 0.0183, + "grad_norm": 0.4995474978675237, + "learning_rate": 0.003, + "loss": 4.268, + "step": 1830 + }, + { + "epoch": 0.01831, + "grad_norm": 0.6731917230470822, + "learning_rate": 0.003, + "loss": 4.2625, + "step": 1831 + }, + { + "epoch": 0.01832, + "grad_norm": 1.0735909763241713, + "learning_rate": 0.003, + "loss": 4.2749, + "step": 1832 + }, + { + "epoch": 0.01833, + "grad_norm": 0.9454987640162065, + "learning_rate": 0.003, + "loss": 4.3126, + "step": 1833 + }, + { + "epoch": 0.01834, + "grad_norm": 0.6279857981331142, + "learning_rate": 0.003, + "loss": 4.265, + "step": 1834 + }, + { + "epoch": 0.01835, + "grad_norm": 0.7795333868843384, + "learning_rate": 0.003, + "loss": 4.2708, + "step": 1835 + }, + { + "epoch": 0.01836, + "grad_norm": 0.7540135177559686, + "learning_rate": 0.003, + "loss": 4.269, + "step": 1836 + }, + { + "epoch": 0.01837, + "grad_norm": 0.4873694552023737, + "learning_rate": 0.003, + "loss": 4.2594, + "step": 1837 + }, + { + "epoch": 0.01838, + "grad_norm": 0.6475671331829288, + "learning_rate": 0.003, + "loss": 4.2679, + "step": 1838 + }, + { + "epoch": 0.01839, + "grad_norm": 0.5413345448794652, + "learning_rate": 0.003, + "loss": 4.2849, + "step": 1839 + }, + { + "epoch": 0.0184, + "grad_norm": 0.505876271981352, + "learning_rate": 0.003, + "loss": 4.2815, + "step": 1840 + }, + { + "epoch": 0.01841, + "grad_norm": 0.5559518639320817, + "learning_rate": 0.003, + "loss": 4.2211, + "step": 1841 + }, + { + "epoch": 0.01842, + "grad_norm": 0.565402658078461, + "learning_rate": 0.003, + "loss": 4.2563, + "step": 1842 + }, + { + "epoch": 0.01843, + "grad_norm": 0.5176031889881334, + "learning_rate": 0.003, + "loss": 4.2347, + "step": 1843 + }, + { + "epoch": 0.01844, + "grad_norm": 0.515476660965344, + "learning_rate": 0.003, + "loss": 4.2336, + "step": 1844 + }, + { + "epoch": 0.01845, + "grad_norm": 0.537738729997812, + "learning_rate": 0.003, + "loss": 4.2366, + "step": 1845 + }, + { + "epoch": 0.01846, + "grad_norm": 0.5825872331164712, + "learning_rate": 0.003, + "loss": 4.2629, + "step": 1846 + }, + { + "epoch": 0.01847, + "grad_norm": 0.6172426924095343, + "learning_rate": 0.003, + "loss": 4.261, + "step": 1847 + }, + { + "epoch": 0.01848, + "grad_norm": 0.5675520394460739, + "learning_rate": 0.003, + "loss": 4.2733, + "step": 1848 + }, + { + "epoch": 0.01849, + "grad_norm": 0.5804129655862207, + "learning_rate": 0.003, + "loss": 4.2733, + "step": 1849 + }, + { + "epoch": 0.0185, + "grad_norm": 0.5903509862738283, + "learning_rate": 0.003, + "loss": 4.2637, + "step": 1850 + }, + { + "epoch": 0.01851, + "grad_norm": 0.5789415751735032, + "learning_rate": 0.003, + "loss": 4.2453, + "step": 1851 + }, + { + "epoch": 0.01852, + "grad_norm": 0.6073114122094332, + "learning_rate": 0.003, + "loss": 4.249, + "step": 1852 + }, + { + "epoch": 0.01853, + "grad_norm": 0.6172059367670675, + "learning_rate": 0.003, + "loss": 4.2514, + "step": 1853 + }, + { + "epoch": 0.01854, + "grad_norm": 0.5952391168819953, + "learning_rate": 0.003, + "loss": 4.282, + "step": 1854 + }, + { + "epoch": 0.01855, + "grad_norm": 0.599929499071102, + "learning_rate": 0.003, + "loss": 4.2645, + "step": 1855 + }, + { + "epoch": 0.01856, + "grad_norm": 0.7854590983227943, + "learning_rate": 0.003, + "loss": 4.2582, + "step": 1856 + }, + { + "epoch": 0.01857, + "grad_norm": 0.9479235992175757, + "learning_rate": 0.003, + "loss": 4.2521, + "step": 1857 + }, + { + "epoch": 0.01858, + "grad_norm": 1.0547185212118042, + "learning_rate": 0.003, + "loss": 4.2782, + "step": 1858 + }, + { + "epoch": 0.01859, + "grad_norm": 0.7498301751007965, + "learning_rate": 0.003, + "loss": 4.259, + "step": 1859 + }, + { + "epoch": 0.0186, + "grad_norm": 0.6772314759733408, + "learning_rate": 0.003, + "loss": 4.238, + "step": 1860 + }, + { + "epoch": 0.01861, + "grad_norm": 0.7554457171283483, + "learning_rate": 0.003, + "loss": 4.2858, + "step": 1861 + }, + { + "epoch": 0.01862, + "grad_norm": 0.6601398192741809, + "learning_rate": 0.003, + "loss": 4.2482, + "step": 1862 + }, + { + "epoch": 0.01863, + "grad_norm": 0.5632517970928785, + "learning_rate": 0.003, + "loss": 4.2548, + "step": 1863 + }, + { + "epoch": 0.01864, + "grad_norm": 0.5518520472207163, + "learning_rate": 0.003, + "loss": 4.2848, + "step": 1864 + }, + { + "epoch": 0.01865, + "grad_norm": 0.5191605026436766, + "learning_rate": 0.003, + "loss": 4.2571, + "step": 1865 + }, + { + "epoch": 0.01866, + "grad_norm": 0.487977336215745, + "learning_rate": 0.003, + "loss": 4.2583, + "step": 1866 + }, + { + "epoch": 0.01867, + "grad_norm": 0.38111750510593084, + "learning_rate": 0.003, + "loss": 4.2288, + "step": 1867 + }, + { + "epoch": 0.01868, + "grad_norm": 0.36979378128361096, + "learning_rate": 0.003, + "loss": 4.2458, + "step": 1868 + }, + { + "epoch": 0.01869, + "grad_norm": 0.4056719991155462, + "learning_rate": 0.003, + "loss": 4.259, + "step": 1869 + }, + { + "epoch": 0.0187, + "grad_norm": 0.511550197815793, + "learning_rate": 0.003, + "loss": 4.2534, + "step": 1870 + }, + { + "epoch": 0.01871, + "grad_norm": 0.6908775527772675, + "learning_rate": 0.003, + "loss": 4.2322, + "step": 1871 + }, + { + "epoch": 0.01872, + "grad_norm": 0.8655201254185445, + "learning_rate": 0.003, + "loss": 4.2892, + "step": 1872 + }, + { + "epoch": 0.01873, + "grad_norm": 0.903028038452321, + "learning_rate": 0.003, + "loss": 4.2527, + "step": 1873 + }, + { + "epoch": 0.01874, + "grad_norm": 0.7655189256325052, + "learning_rate": 0.003, + "loss": 4.2534, + "step": 1874 + }, + { + "epoch": 0.01875, + "grad_norm": 0.6845228657861865, + "learning_rate": 0.003, + "loss": 4.2697, + "step": 1875 + }, + { + "epoch": 0.01876, + "grad_norm": 0.7967028641532641, + "learning_rate": 0.003, + "loss": 4.2796, + "step": 1876 + }, + { + "epoch": 0.01877, + "grad_norm": 0.7276633971950639, + "learning_rate": 0.003, + "loss": 4.2792, + "step": 1877 + }, + { + "epoch": 0.01878, + "grad_norm": 0.6745755254962531, + "learning_rate": 0.003, + "loss": 4.2647, + "step": 1878 + }, + { + "epoch": 0.01879, + "grad_norm": 0.6794948462625433, + "learning_rate": 0.003, + "loss": 4.2421, + "step": 1879 + }, + { + "epoch": 0.0188, + "grad_norm": 0.7496570961239313, + "learning_rate": 0.003, + "loss": 4.2581, + "step": 1880 + }, + { + "epoch": 0.01881, + "grad_norm": 0.7216177452798419, + "learning_rate": 0.003, + "loss": 4.2527, + "step": 1881 + }, + { + "epoch": 0.01882, + "grad_norm": 0.6411166720639763, + "learning_rate": 0.003, + "loss": 4.2764, + "step": 1882 + }, + { + "epoch": 0.01883, + "grad_norm": 0.4976595430995637, + "learning_rate": 0.003, + "loss": 4.2463, + "step": 1883 + }, + { + "epoch": 0.01884, + "grad_norm": 0.5806395175516162, + "learning_rate": 0.003, + "loss": 4.2463, + "step": 1884 + }, + { + "epoch": 0.01885, + "grad_norm": 0.5932999040186268, + "learning_rate": 0.003, + "loss": 4.2267, + "step": 1885 + }, + { + "epoch": 0.01886, + "grad_norm": 0.6243294916603833, + "learning_rate": 0.003, + "loss": 4.2637, + "step": 1886 + }, + { + "epoch": 0.01887, + "grad_norm": 0.6594473434297762, + "learning_rate": 0.003, + "loss": 4.2583, + "step": 1887 + }, + { + "epoch": 0.01888, + "grad_norm": 0.5833874852548948, + "learning_rate": 0.003, + "loss": 4.2402, + "step": 1888 + }, + { + "epoch": 0.01889, + "grad_norm": 0.5554783687186964, + "learning_rate": 0.003, + "loss": 4.2547, + "step": 1889 + }, + { + "epoch": 0.0189, + "grad_norm": 0.5267765258339018, + "learning_rate": 0.003, + "loss": 4.2329, + "step": 1890 + }, + { + "epoch": 0.01891, + "grad_norm": 0.5533738473996538, + "learning_rate": 0.003, + "loss": 4.2819, + "step": 1891 + }, + { + "epoch": 0.01892, + "grad_norm": 0.5233002886731548, + "learning_rate": 0.003, + "loss": 4.2568, + "step": 1892 + }, + { + "epoch": 0.01893, + "grad_norm": 0.598656523235247, + "learning_rate": 0.003, + "loss": 4.2415, + "step": 1893 + }, + { + "epoch": 0.01894, + "grad_norm": 0.6329032192123055, + "learning_rate": 0.003, + "loss": 4.2498, + "step": 1894 + }, + { + "epoch": 0.01895, + "grad_norm": 0.7448049826558554, + "learning_rate": 0.003, + "loss": 4.2404, + "step": 1895 + }, + { + "epoch": 0.01896, + "grad_norm": 0.7896808634621466, + "learning_rate": 0.003, + "loss": 4.2666, + "step": 1896 + }, + { + "epoch": 0.01897, + "grad_norm": 0.8670691506268475, + "learning_rate": 0.003, + "loss": 4.267, + "step": 1897 + }, + { + "epoch": 0.01898, + "grad_norm": 0.7564766735493696, + "learning_rate": 0.003, + "loss": 4.2371, + "step": 1898 + }, + { + "epoch": 0.01899, + "grad_norm": 0.6014591021089313, + "learning_rate": 0.003, + "loss": 4.246, + "step": 1899 + }, + { + "epoch": 0.019, + "grad_norm": 0.6487238549786433, + "learning_rate": 0.003, + "loss": 4.2427, + "step": 1900 + }, + { + "epoch": 0.01901, + "grad_norm": 0.7335622998657767, + "learning_rate": 0.003, + "loss": 4.2939, + "step": 1901 + }, + { + "epoch": 0.01902, + "grad_norm": 0.727489215463111, + "learning_rate": 0.003, + "loss": 4.2706, + "step": 1902 + }, + { + "epoch": 0.01903, + "grad_norm": 0.6444511274857245, + "learning_rate": 0.003, + "loss": 4.2217, + "step": 1903 + }, + { + "epoch": 0.01904, + "grad_norm": 0.5669107570524424, + "learning_rate": 0.003, + "loss": 4.2738, + "step": 1904 + }, + { + "epoch": 0.01905, + "grad_norm": 0.47567496733321213, + "learning_rate": 0.003, + "loss": 4.2471, + "step": 1905 + }, + { + "epoch": 0.01906, + "grad_norm": 0.568871434515019, + "learning_rate": 0.003, + "loss": 4.2665, + "step": 1906 + }, + { + "epoch": 0.01907, + "grad_norm": 0.6437840413516579, + "learning_rate": 0.003, + "loss": 4.2605, + "step": 1907 + }, + { + "epoch": 0.01908, + "grad_norm": 0.8102395542085408, + "learning_rate": 0.003, + "loss": 4.2806, + "step": 1908 + }, + { + "epoch": 0.01909, + "grad_norm": 0.9019612301377007, + "learning_rate": 0.003, + "loss": 4.2522, + "step": 1909 + }, + { + "epoch": 0.0191, + "grad_norm": 0.8060528165222425, + "learning_rate": 0.003, + "loss": 4.2682, + "step": 1910 + }, + { + "epoch": 0.01911, + "grad_norm": 0.6514562643306322, + "learning_rate": 0.003, + "loss": 4.2341, + "step": 1911 + }, + { + "epoch": 0.01912, + "grad_norm": 0.7512556786074466, + "learning_rate": 0.003, + "loss": 4.2481, + "step": 1912 + }, + { + "epoch": 0.01913, + "grad_norm": 0.7206930768558686, + "learning_rate": 0.003, + "loss": 4.2688, + "step": 1913 + }, + { + "epoch": 0.01914, + "grad_norm": 0.6802559754696879, + "learning_rate": 0.003, + "loss": 4.2815, + "step": 1914 + }, + { + "epoch": 0.01915, + "grad_norm": 0.6329709283705148, + "learning_rate": 0.003, + "loss": 4.2674, + "step": 1915 + }, + { + "epoch": 0.01916, + "grad_norm": 0.6539184714683666, + "learning_rate": 0.003, + "loss": 4.2462, + "step": 1916 + }, + { + "epoch": 0.01917, + "grad_norm": 0.710899415708157, + "learning_rate": 0.003, + "loss": 4.2674, + "step": 1917 + }, + { + "epoch": 0.01918, + "grad_norm": 0.7167946113405885, + "learning_rate": 0.003, + "loss": 4.2562, + "step": 1918 + }, + { + "epoch": 0.01919, + "grad_norm": 0.7533071012807427, + "learning_rate": 0.003, + "loss": 4.2594, + "step": 1919 + }, + { + "epoch": 0.0192, + "grad_norm": 0.6703396884694851, + "learning_rate": 0.003, + "loss": 4.2612, + "step": 1920 + }, + { + "epoch": 0.01921, + "grad_norm": 0.5834644623189628, + "learning_rate": 0.003, + "loss": 4.2441, + "step": 1921 + }, + { + "epoch": 0.01922, + "grad_norm": 0.6481277683352221, + "learning_rate": 0.003, + "loss": 4.2625, + "step": 1922 + }, + { + "epoch": 0.01923, + "grad_norm": 0.6044219599617415, + "learning_rate": 0.003, + "loss": 4.2661, + "step": 1923 + }, + { + "epoch": 0.01924, + "grad_norm": 0.5337323934986105, + "learning_rate": 0.003, + "loss": 4.2643, + "step": 1924 + }, + { + "epoch": 0.01925, + "grad_norm": 0.5129044588459396, + "learning_rate": 0.003, + "loss": 4.251, + "step": 1925 + }, + { + "epoch": 0.01926, + "grad_norm": 0.48201171188935904, + "learning_rate": 0.003, + "loss": 4.2563, + "step": 1926 + }, + { + "epoch": 0.01927, + "grad_norm": 0.4295757083815179, + "learning_rate": 0.003, + "loss": 4.2456, + "step": 1927 + }, + { + "epoch": 0.01928, + "grad_norm": 0.5221994128637466, + "learning_rate": 0.003, + "loss": 4.2241, + "step": 1928 + }, + { + "epoch": 0.01929, + "grad_norm": 0.6151786259894082, + "learning_rate": 0.003, + "loss": 4.246, + "step": 1929 + }, + { + "epoch": 0.0193, + "grad_norm": 0.717185888834206, + "learning_rate": 0.003, + "loss": 4.2399, + "step": 1930 + }, + { + "epoch": 0.01931, + "grad_norm": 0.865419932280181, + "learning_rate": 0.003, + "loss": 4.2485, + "step": 1931 + }, + { + "epoch": 0.01932, + "grad_norm": 1.015142018105296, + "learning_rate": 0.003, + "loss": 4.2679, + "step": 1932 + }, + { + "epoch": 0.01933, + "grad_norm": 1.0909531598947217, + "learning_rate": 0.003, + "loss": 4.2582, + "step": 1933 + }, + { + "epoch": 0.01934, + "grad_norm": 0.7995257446738059, + "learning_rate": 0.003, + "loss": 4.2554, + "step": 1934 + }, + { + "epoch": 0.01935, + "grad_norm": 0.6434815333532291, + "learning_rate": 0.003, + "loss": 4.2669, + "step": 1935 + }, + { + "epoch": 0.01936, + "grad_norm": 0.6569925784458591, + "learning_rate": 0.003, + "loss": 4.2489, + "step": 1936 + }, + { + "epoch": 0.01937, + "grad_norm": 0.5996128457789688, + "learning_rate": 0.003, + "loss": 4.2577, + "step": 1937 + }, + { + "epoch": 0.01938, + "grad_norm": 0.6329170311784792, + "learning_rate": 0.003, + "loss": 4.2615, + "step": 1938 + }, + { + "epoch": 0.01939, + "grad_norm": 0.6354665323417178, + "learning_rate": 0.003, + "loss": 4.2808, + "step": 1939 + }, + { + "epoch": 0.0194, + "grad_norm": 0.6008328437037136, + "learning_rate": 0.003, + "loss": 4.2581, + "step": 1940 + }, + { + "epoch": 0.01941, + "grad_norm": 0.5343827591468221, + "learning_rate": 0.003, + "loss": 4.2378, + "step": 1941 + }, + { + "epoch": 0.01942, + "grad_norm": 0.5512416106954112, + "learning_rate": 0.003, + "loss": 4.2611, + "step": 1942 + }, + { + "epoch": 0.01943, + "grad_norm": 0.5866960320410021, + "learning_rate": 0.003, + "loss": 4.2341, + "step": 1943 + }, + { + "epoch": 0.01944, + "grad_norm": 0.6240489288808161, + "learning_rate": 0.003, + "loss": 4.2751, + "step": 1944 + }, + { + "epoch": 0.01945, + "grad_norm": 0.6243606216006328, + "learning_rate": 0.003, + "loss": 4.2437, + "step": 1945 + }, + { + "epoch": 0.01946, + "grad_norm": 0.5880226151273782, + "learning_rate": 0.003, + "loss": 4.2379, + "step": 1946 + }, + { + "epoch": 0.01947, + "grad_norm": 0.5934159043950513, + "learning_rate": 0.003, + "loss": 4.2471, + "step": 1947 + }, + { + "epoch": 0.01948, + "grad_norm": 0.5378813319639308, + "learning_rate": 0.003, + "loss": 4.2659, + "step": 1948 + }, + { + "epoch": 0.01949, + "grad_norm": 0.5777955568924656, + "learning_rate": 0.003, + "loss": 4.2595, + "step": 1949 + }, + { + "epoch": 0.0195, + "grad_norm": 0.6588236649280669, + "learning_rate": 0.003, + "loss": 4.2493, + "step": 1950 + }, + { + "epoch": 0.01951, + "grad_norm": 0.6662863362190039, + "learning_rate": 0.003, + "loss": 4.2457, + "step": 1951 + }, + { + "epoch": 0.01952, + "grad_norm": 0.6956758803553106, + "learning_rate": 0.003, + "loss": 4.2563, + "step": 1952 + }, + { + "epoch": 0.01953, + "grad_norm": 0.7309670323591311, + "learning_rate": 0.003, + "loss": 4.2663, + "step": 1953 + }, + { + "epoch": 0.01954, + "grad_norm": 0.7102958436956044, + "learning_rate": 0.003, + "loss": 4.2656, + "step": 1954 + }, + { + "epoch": 0.01955, + "grad_norm": 0.748927101392194, + "learning_rate": 0.003, + "loss": 4.2596, + "step": 1955 + }, + { + "epoch": 0.01956, + "grad_norm": 0.6357959051502906, + "learning_rate": 0.003, + "loss": 4.2507, + "step": 1956 + }, + { + "epoch": 0.01957, + "grad_norm": 0.5776415679220375, + "learning_rate": 0.003, + "loss": 4.2172, + "step": 1957 + }, + { + "epoch": 0.01958, + "grad_norm": 0.5010308008271186, + "learning_rate": 0.003, + "loss": 4.2108, + "step": 1958 + }, + { + "epoch": 0.01959, + "grad_norm": 0.500419502823399, + "learning_rate": 0.003, + "loss": 4.2342, + "step": 1959 + }, + { + "epoch": 0.0196, + "grad_norm": 0.5017850552651191, + "learning_rate": 0.003, + "loss": 4.229, + "step": 1960 + }, + { + "epoch": 0.01961, + "grad_norm": 0.49043553705552145, + "learning_rate": 0.003, + "loss": 4.2267, + "step": 1961 + }, + { + "epoch": 0.01962, + "grad_norm": 0.49806058358446914, + "learning_rate": 0.003, + "loss": 4.2301, + "step": 1962 + }, + { + "epoch": 0.01963, + "grad_norm": 0.5386284819858854, + "learning_rate": 0.003, + "loss": 4.2201, + "step": 1963 + }, + { + "epoch": 0.01964, + "grad_norm": 0.6462166130085853, + "learning_rate": 0.003, + "loss": 4.2492, + "step": 1964 + }, + { + "epoch": 0.01965, + "grad_norm": 0.8617584950578717, + "learning_rate": 0.003, + "loss": 4.2512, + "step": 1965 + }, + { + "epoch": 0.01966, + "grad_norm": 1.064420589029421, + "learning_rate": 0.003, + "loss": 4.2624, + "step": 1966 + }, + { + "epoch": 0.01967, + "grad_norm": 0.8064456905773513, + "learning_rate": 0.003, + "loss": 4.2378, + "step": 1967 + }, + { + "epoch": 0.01968, + "grad_norm": 0.6120027249682956, + "learning_rate": 0.003, + "loss": 4.2516, + "step": 1968 + }, + { + "epoch": 0.01969, + "grad_norm": 0.7508122971760963, + "learning_rate": 0.003, + "loss": 4.2526, + "step": 1969 + }, + { + "epoch": 0.0197, + "grad_norm": 0.7405449826366749, + "learning_rate": 0.003, + "loss": 4.2514, + "step": 1970 + }, + { + "epoch": 0.01971, + "grad_norm": 0.7056278637839583, + "learning_rate": 0.003, + "loss": 4.2377, + "step": 1971 + }, + { + "epoch": 0.01972, + "grad_norm": 0.7105683738164339, + "learning_rate": 0.003, + "loss": 4.2407, + "step": 1972 + }, + { + "epoch": 0.01973, + "grad_norm": 0.7508003521914575, + "learning_rate": 0.003, + "loss": 4.2264, + "step": 1973 + }, + { + "epoch": 0.01974, + "grad_norm": 0.7061088132196116, + "learning_rate": 0.003, + "loss": 4.2641, + "step": 1974 + }, + { + "epoch": 0.01975, + "grad_norm": 0.6140071978685933, + "learning_rate": 0.003, + "loss": 4.2539, + "step": 1975 + }, + { + "epoch": 0.01976, + "grad_norm": 0.6214170582947839, + "learning_rate": 0.003, + "loss": 4.243, + "step": 1976 + }, + { + "epoch": 0.01977, + "grad_norm": 0.6032091701282252, + "learning_rate": 0.003, + "loss": 4.2759, + "step": 1977 + }, + { + "epoch": 0.01978, + "grad_norm": 0.6890579485446906, + "learning_rate": 0.003, + "loss": 4.2782, + "step": 1978 + }, + { + "epoch": 0.01979, + "grad_norm": 0.7078287139411981, + "learning_rate": 0.003, + "loss": 4.269, + "step": 1979 + }, + { + "epoch": 0.0198, + "grad_norm": 0.726915367409026, + "learning_rate": 0.003, + "loss": 4.2503, + "step": 1980 + }, + { + "epoch": 0.01981, + "grad_norm": 0.7277727772925232, + "learning_rate": 0.003, + "loss": 4.2394, + "step": 1981 + }, + { + "epoch": 0.01982, + "grad_norm": 0.6618110470758236, + "learning_rate": 0.003, + "loss": 4.262, + "step": 1982 + }, + { + "epoch": 0.01983, + "grad_norm": 0.6282900232413579, + "learning_rate": 0.003, + "loss": 4.2497, + "step": 1983 + }, + { + "epoch": 0.01984, + "grad_norm": 0.6651788332059134, + "learning_rate": 0.003, + "loss": 4.2635, + "step": 1984 + }, + { + "epoch": 0.01985, + "grad_norm": 0.6195263920647736, + "learning_rate": 0.003, + "loss": 4.2484, + "step": 1985 + }, + { + "epoch": 0.01986, + "grad_norm": 0.6562857171469267, + "learning_rate": 0.003, + "loss": 4.2792, + "step": 1986 + }, + { + "epoch": 0.01987, + "grad_norm": 0.6274806569391715, + "learning_rate": 0.003, + "loss": 4.2331, + "step": 1987 + }, + { + "epoch": 0.01988, + "grad_norm": 0.6627977107268054, + "learning_rate": 0.003, + "loss": 4.2466, + "step": 1988 + }, + { + "epoch": 0.01989, + "grad_norm": 0.6286439849240877, + "learning_rate": 0.003, + "loss": 4.2344, + "step": 1989 + }, + { + "epoch": 0.0199, + "grad_norm": 0.5786457654364954, + "learning_rate": 0.003, + "loss": 4.246, + "step": 1990 + }, + { + "epoch": 0.01991, + "grad_norm": 0.5167049375276446, + "learning_rate": 0.003, + "loss": 4.2238, + "step": 1991 + }, + { + "epoch": 0.01992, + "grad_norm": 0.50421095170271, + "learning_rate": 0.003, + "loss": 4.261, + "step": 1992 + }, + { + "epoch": 0.01993, + "grad_norm": 0.4699320041008658, + "learning_rate": 0.003, + "loss": 4.2447, + "step": 1993 + }, + { + "epoch": 0.01994, + "grad_norm": 0.5186369446769569, + "learning_rate": 0.003, + "loss": 4.2629, + "step": 1994 + }, + { + "epoch": 0.01995, + "grad_norm": 0.5867398105885127, + "learning_rate": 0.003, + "loss": 4.2547, + "step": 1995 + }, + { + "epoch": 0.01996, + "grad_norm": 0.8031924386202848, + "learning_rate": 0.003, + "loss": 4.2326, + "step": 1996 + }, + { + "epoch": 0.01997, + "grad_norm": 0.9756949043315064, + "learning_rate": 0.003, + "loss": 4.2746, + "step": 1997 + }, + { + "epoch": 0.01998, + "grad_norm": 0.8340147112264455, + "learning_rate": 0.003, + "loss": 4.2546, + "step": 1998 + }, + { + "epoch": 0.01999, + "grad_norm": 0.6652469206537822, + "learning_rate": 0.003, + "loss": 4.2393, + "step": 1999 + }, + { + "epoch": 0.02, + "grad_norm": 0.7695645348180246, + "learning_rate": 0.003, + "loss": 4.2318, + "step": 2000 + }, + { + "epoch": 0.02001, + "grad_norm": 0.6865633733578148, + "learning_rate": 0.003, + "loss": 4.262, + "step": 2001 + }, + { + "epoch": 0.02002, + "grad_norm": 0.7345949774088514, + "learning_rate": 0.003, + "loss": 4.2618, + "step": 2002 + }, + { + "epoch": 0.02003, + "grad_norm": 0.5984396930779281, + "learning_rate": 0.003, + "loss": 4.2543, + "step": 2003 + }, + { + "epoch": 0.02004, + "grad_norm": 0.5305961077422008, + "learning_rate": 0.003, + "loss": 4.2408, + "step": 2004 + }, + { + "epoch": 0.02005, + "grad_norm": 0.48452673426344656, + "learning_rate": 0.003, + "loss": 4.2431, + "step": 2005 + }, + { + "epoch": 0.02006, + "grad_norm": 0.5689487878863901, + "learning_rate": 0.003, + "loss": 4.2442, + "step": 2006 + }, + { + "epoch": 0.02007, + "grad_norm": 0.5838792869042901, + "learning_rate": 0.003, + "loss": 4.2584, + "step": 2007 + }, + { + "epoch": 0.02008, + "grad_norm": 0.6291452839633271, + "learning_rate": 0.003, + "loss": 4.2506, + "step": 2008 + }, + { + "epoch": 0.02009, + "grad_norm": 0.6903355213664546, + "learning_rate": 0.003, + "loss": 4.2495, + "step": 2009 + }, + { + "epoch": 0.0201, + "grad_norm": 0.6289287565263567, + "learning_rate": 0.003, + "loss": 4.2607, + "step": 2010 + }, + { + "epoch": 0.02011, + "grad_norm": 0.5984818046752132, + "learning_rate": 0.003, + "loss": 4.2452, + "step": 2011 + }, + { + "epoch": 0.02012, + "grad_norm": 0.6682118694640922, + "learning_rate": 0.003, + "loss": 4.2452, + "step": 2012 + }, + { + "epoch": 0.02013, + "grad_norm": 0.7050046827190253, + "learning_rate": 0.003, + "loss": 4.2532, + "step": 2013 + }, + { + "epoch": 0.02014, + "grad_norm": 0.6761323120848991, + "learning_rate": 0.003, + "loss": 4.2415, + "step": 2014 + }, + { + "epoch": 0.02015, + "grad_norm": 0.6372907183843184, + "learning_rate": 0.003, + "loss": 4.2336, + "step": 2015 + }, + { + "epoch": 0.02016, + "grad_norm": 0.5972921997645952, + "learning_rate": 0.003, + "loss": 4.2404, + "step": 2016 + }, + { + "epoch": 0.02017, + "grad_norm": 0.49149348634303763, + "learning_rate": 0.003, + "loss": 4.2312, + "step": 2017 + }, + { + "epoch": 0.02018, + "grad_norm": 0.6030678218331625, + "learning_rate": 0.003, + "loss": 4.2371, + "step": 2018 + }, + { + "epoch": 0.02019, + "grad_norm": 0.6395667829535132, + "learning_rate": 0.003, + "loss": 4.2431, + "step": 2019 + }, + { + "epoch": 0.0202, + "grad_norm": 0.6210152394675877, + "learning_rate": 0.003, + "loss": 4.2512, + "step": 2020 + }, + { + "epoch": 0.02021, + "grad_norm": 0.6950283205655802, + "learning_rate": 0.003, + "loss": 4.268, + "step": 2021 + }, + { + "epoch": 0.02022, + "grad_norm": 0.8101179933347831, + "learning_rate": 0.003, + "loss": 4.2328, + "step": 2022 + }, + { + "epoch": 0.02023, + "grad_norm": 0.8379556185876997, + "learning_rate": 0.003, + "loss": 4.2128, + "step": 2023 + }, + { + "epoch": 0.02024, + "grad_norm": 0.8247606718710992, + "learning_rate": 0.003, + "loss": 4.2373, + "step": 2024 + }, + { + "epoch": 0.02025, + "grad_norm": 0.8736161526776162, + "learning_rate": 0.003, + "loss": 4.291, + "step": 2025 + }, + { + "epoch": 0.02026, + "grad_norm": 0.8472496904448235, + "learning_rate": 0.003, + "loss": 4.2727, + "step": 2026 + }, + { + "epoch": 0.02027, + "grad_norm": 0.8056170331271387, + "learning_rate": 0.003, + "loss": 4.271, + "step": 2027 + }, + { + "epoch": 0.02028, + "grad_norm": 0.8498369070927115, + "learning_rate": 0.003, + "loss": 4.2569, + "step": 2028 + }, + { + "epoch": 0.02029, + "grad_norm": 0.9047394863120564, + "learning_rate": 0.003, + "loss": 4.2658, + "step": 2029 + }, + { + "epoch": 0.0203, + "grad_norm": 1.0823713408101607, + "learning_rate": 0.003, + "loss": 4.2684, + "step": 2030 + }, + { + "epoch": 0.02031, + "grad_norm": 0.9309326946376139, + "learning_rate": 0.003, + "loss": 4.2732, + "step": 2031 + }, + { + "epoch": 0.02032, + "grad_norm": 0.9146514324992777, + "learning_rate": 0.003, + "loss": 4.2447, + "step": 2032 + }, + { + "epoch": 0.02033, + "grad_norm": 0.8591399303672403, + "learning_rate": 0.003, + "loss": 4.2848, + "step": 2033 + }, + { + "epoch": 0.02034, + "grad_norm": 0.8501591519538507, + "learning_rate": 0.003, + "loss": 4.2786, + "step": 2034 + }, + { + "epoch": 0.02035, + "grad_norm": 0.8000776065354223, + "learning_rate": 0.003, + "loss": 4.2613, + "step": 2035 + }, + { + "epoch": 0.02036, + "grad_norm": 0.7789847842987989, + "learning_rate": 0.003, + "loss": 4.2629, + "step": 2036 + }, + { + "epoch": 0.02037, + "grad_norm": 0.8301107777503408, + "learning_rate": 0.003, + "loss": 4.2828, + "step": 2037 + }, + { + "epoch": 0.02038, + "grad_norm": 0.8246182480374772, + "learning_rate": 0.003, + "loss": 4.2649, + "step": 2038 + }, + { + "epoch": 0.02039, + "grad_norm": 0.8433738197183226, + "learning_rate": 0.003, + "loss": 4.2745, + "step": 2039 + }, + { + "epoch": 0.0204, + "grad_norm": 0.7922524759276299, + "learning_rate": 0.003, + "loss": 4.2766, + "step": 2040 + }, + { + "epoch": 0.02041, + "grad_norm": 0.7214169410748629, + "learning_rate": 0.003, + "loss": 4.2381, + "step": 2041 + }, + { + "epoch": 0.02042, + "grad_norm": 0.5980473954078982, + "learning_rate": 0.003, + "loss": 4.2311, + "step": 2042 + }, + { + "epoch": 0.02043, + "grad_norm": 0.6544292652761489, + "learning_rate": 0.003, + "loss": 4.2765, + "step": 2043 + }, + { + "epoch": 0.02044, + "grad_norm": 0.6815263020074647, + "learning_rate": 0.003, + "loss": 4.2533, + "step": 2044 + }, + { + "epoch": 0.02045, + "grad_norm": 0.6716783756563022, + "learning_rate": 0.003, + "loss": 4.266, + "step": 2045 + }, + { + "epoch": 0.02046, + "grad_norm": 0.5685869111952888, + "learning_rate": 0.003, + "loss": 4.2364, + "step": 2046 + }, + { + "epoch": 0.02047, + "grad_norm": 0.539068698157252, + "learning_rate": 0.003, + "loss": 4.2693, + "step": 2047 + }, + { + "epoch": 0.02048, + "grad_norm": 0.5462552506791943, + "learning_rate": 0.003, + "loss": 4.2521, + "step": 2048 + }, + { + "epoch": 0.02049, + "grad_norm": 0.5078072105462605, + "learning_rate": 0.003, + "loss": 4.2715, + "step": 2049 + }, + { + "epoch": 0.0205, + "grad_norm": 0.5740623442922046, + "learning_rate": 0.003, + "loss": 4.2205, + "step": 2050 + }, + { + "epoch": 0.02051, + "grad_norm": 0.631335957620747, + "learning_rate": 0.003, + "loss": 4.2224, + "step": 2051 + }, + { + "epoch": 0.02052, + "grad_norm": 0.6486110208729147, + "learning_rate": 0.003, + "loss": 4.2258, + "step": 2052 + }, + { + "epoch": 0.02053, + "grad_norm": 0.5888200401710874, + "learning_rate": 0.003, + "loss": 4.2235, + "step": 2053 + }, + { + "epoch": 0.02054, + "grad_norm": 0.5515462434036196, + "learning_rate": 0.003, + "loss": 4.2611, + "step": 2054 + }, + { + "epoch": 0.02055, + "grad_norm": 0.4722653472823476, + "learning_rate": 0.003, + "loss": 4.2178, + "step": 2055 + }, + { + "epoch": 0.02056, + "grad_norm": 0.44396394612077134, + "learning_rate": 0.003, + "loss": 4.2149, + "step": 2056 + }, + { + "epoch": 0.02057, + "grad_norm": 0.4548184991267254, + "learning_rate": 0.003, + "loss": 4.2514, + "step": 2057 + }, + { + "epoch": 0.02058, + "grad_norm": 0.5196145685763705, + "learning_rate": 0.003, + "loss": 4.2286, + "step": 2058 + }, + { + "epoch": 0.02059, + "grad_norm": 0.5150404447757911, + "learning_rate": 0.003, + "loss": 4.2089, + "step": 2059 + }, + { + "epoch": 0.0206, + "grad_norm": 0.44383213075680966, + "learning_rate": 0.003, + "loss": 4.251, + "step": 2060 + }, + { + "epoch": 0.02061, + "grad_norm": 0.37383563921339574, + "learning_rate": 0.003, + "loss": 4.2222, + "step": 2061 + }, + { + "epoch": 0.02062, + "grad_norm": 0.35393934819548667, + "learning_rate": 0.003, + "loss": 4.221, + "step": 2062 + }, + { + "epoch": 0.02063, + "grad_norm": 0.3267679809833493, + "learning_rate": 0.003, + "loss": 4.2165, + "step": 2063 + }, + { + "epoch": 0.02064, + "grad_norm": 0.34033928546810105, + "learning_rate": 0.003, + "loss": 4.2259, + "step": 2064 + }, + { + "epoch": 0.02065, + "grad_norm": 0.35466259038416126, + "learning_rate": 0.003, + "loss": 4.2059, + "step": 2065 + }, + { + "epoch": 0.02066, + "grad_norm": 0.3016586113305969, + "learning_rate": 0.003, + "loss": 4.2224, + "step": 2066 + }, + { + "epoch": 0.02067, + "grad_norm": 0.3498296991074057, + "learning_rate": 0.003, + "loss": 4.22, + "step": 2067 + }, + { + "epoch": 0.02068, + "grad_norm": 0.3676321022781276, + "learning_rate": 0.003, + "loss": 4.2017, + "step": 2068 + }, + { + "epoch": 0.02069, + "grad_norm": 0.40962533052122363, + "learning_rate": 0.003, + "loss": 4.2244, + "step": 2069 + }, + { + "epoch": 0.0207, + "grad_norm": 0.5995962395694695, + "learning_rate": 0.003, + "loss": 4.2023, + "step": 2070 + }, + { + "epoch": 0.02071, + "grad_norm": 1.0281554874982604, + "learning_rate": 0.003, + "loss": 4.2499, + "step": 2071 + }, + { + "epoch": 0.02072, + "grad_norm": 1.2174002009617175, + "learning_rate": 0.003, + "loss": 4.2817, + "step": 2072 + }, + { + "epoch": 0.02073, + "grad_norm": 0.5766208529721838, + "learning_rate": 0.003, + "loss": 4.2339, + "step": 2073 + }, + { + "epoch": 0.02074, + "grad_norm": 0.8482120839475051, + "learning_rate": 0.003, + "loss": 4.2516, + "step": 2074 + }, + { + "epoch": 0.02075, + "grad_norm": 0.914249694296833, + "learning_rate": 0.003, + "loss": 4.2873, + "step": 2075 + }, + { + "epoch": 0.02076, + "grad_norm": 1.0073872264521357, + "learning_rate": 0.003, + "loss": 4.2898, + "step": 2076 + }, + { + "epoch": 0.02077, + "grad_norm": 0.9613925913617914, + "learning_rate": 0.003, + "loss": 4.2576, + "step": 2077 + }, + { + "epoch": 0.02078, + "grad_norm": 1.0251025412787642, + "learning_rate": 0.003, + "loss": 4.2665, + "step": 2078 + }, + { + "epoch": 0.02079, + "grad_norm": 0.8112389212571772, + "learning_rate": 0.003, + "loss": 4.243, + "step": 2079 + }, + { + "epoch": 0.0208, + "grad_norm": 0.9069389526132752, + "learning_rate": 0.003, + "loss": 4.2693, + "step": 2080 + }, + { + "epoch": 0.02081, + "grad_norm": 1.1627691288871134, + "learning_rate": 0.003, + "loss": 4.3161, + "step": 2081 + }, + { + "epoch": 0.02082, + "grad_norm": 1.0285610491992647, + "learning_rate": 0.003, + "loss": 4.3126, + "step": 2082 + }, + { + "epoch": 0.02083, + "grad_norm": 1.0850701155180529, + "learning_rate": 0.003, + "loss": 4.3091, + "step": 2083 + }, + { + "epoch": 0.02084, + "grad_norm": 0.9461237384614024, + "learning_rate": 0.003, + "loss": 4.2987, + "step": 2084 + }, + { + "epoch": 0.02085, + "grad_norm": 0.976394782581613, + "learning_rate": 0.003, + "loss": 4.284, + "step": 2085 + }, + { + "epoch": 0.02086, + "grad_norm": 0.82502607345478, + "learning_rate": 0.003, + "loss": 4.2825, + "step": 2086 + }, + { + "epoch": 0.02087, + "grad_norm": 0.6641978639980478, + "learning_rate": 0.003, + "loss": 4.2762, + "step": 2087 + }, + { + "epoch": 0.02088, + "grad_norm": 0.5823960320780495, + "learning_rate": 0.003, + "loss": 4.3019, + "step": 2088 + }, + { + "epoch": 0.02089, + "grad_norm": 0.6056623050213609, + "learning_rate": 0.003, + "loss": 4.2696, + "step": 2089 + }, + { + "epoch": 0.0209, + "grad_norm": 0.6222015176775844, + "learning_rate": 0.003, + "loss": 4.2845, + "step": 2090 + }, + { + "epoch": 0.02091, + "grad_norm": 0.8107138314533802, + "learning_rate": 0.003, + "loss": 4.275, + "step": 2091 + }, + { + "epoch": 0.02092, + "grad_norm": 0.9481339301239347, + "learning_rate": 0.003, + "loss": 4.2799, + "step": 2092 + }, + { + "epoch": 0.02093, + "grad_norm": 0.8740047196660798, + "learning_rate": 0.003, + "loss": 4.2456, + "step": 2093 + }, + { + "epoch": 0.02094, + "grad_norm": 0.6305842273976254, + "learning_rate": 0.003, + "loss": 4.2436, + "step": 2094 + }, + { + "epoch": 0.02095, + "grad_norm": 0.5189901063960811, + "learning_rate": 0.003, + "loss": 4.2773, + "step": 2095 + }, + { + "epoch": 0.02096, + "grad_norm": 0.7024275821596554, + "learning_rate": 0.003, + "loss": 4.2741, + "step": 2096 + }, + { + "epoch": 0.02097, + "grad_norm": 0.6837847162193671, + "learning_rate": 0.003, + "loss": 4.2518, + "step": 2097 + }, + { + "epoch": 0.02098, + "grad_norm": 0.6707932407749881, + "learning_rate": 0.003, + "loss": 4.2705, + "step": 2098 + }, + { + "epoch": 0.02099, + "grad_norm": 0.5110579033218214, + "learning_rate": 0.003, + "loss": 4.2724, + "step": 2099 + }, + { + "epoch": 0.021, + "grad_norm": 0.4612182402663896, + "learning_rate": 0.003, + "loss": 4.2819, + "step": 2100 + }, + { + "epoch": 0.02101, + "grad_norm": 0.45331179352856693, + "learning_rate": 0.003, + "loss": 4.2684, + "step": 2101 + }, + { + "epoch": 0.02102, + "grad_norm": 0.42546683266841523, + "learning_rate": 0.003, + "loss": 4.271, + "step": 2102 + }, + { + "epoch": 0.02103, + "grad_norm": 0.3547519868664197, + "learning_rate": 0.003, + "loss": 4.2273, + "step": 2103 + }, + { + "epoch": 0.02104, + "grad_norm": 0.3900500431869311, + "learning_rate": 0.003, + "loss": 4.2306, + "step": 2104 + }, + { + "epoch": 0.02105, + "grad_norm": 0.4122345047469997, + "learning_rate": 0.003, + "loss": 4.2329, + "step": 2105 + }, + { + "epoch": 0.02106, + "grad_norm": 0.4185113160695593, + "learning_rate": 0.003, + "loss": 4.2247, + "step": 2106 + }, + { + "epoch": 0.02107, + "grad_norm": 0.3925543469529075, + "learning_rate": 0.003, + "loss": 4.2367, + "step": 2107 + }, + { + "epoch": 0.02108, + "grad_norm": 0.33327126218953085, + "learning_rate": 0.003, + "loss": 4.2153, + "step": 2108 + }, + { + "epoch": 0.02109, + "grad_norm": 0.3153436836891661, + "learning_rate": 0.003, + "loss": 4.2035, + "step": 2109 + }, + { + "epoch": 0.0211, + "grad_norm": 0.35294607433058894, + "learning_rate": 0.003, + "loss": 4.2193, + "step": 2110 + }, + { + "epoch": 0.02111, + "grad_norm": 0.37843249197178286, + "learning_rate": 0.003, + "loss": 4.2278, + "step": 2111 + }, + { + "epoch": 0.02112, + "grad_norm": 0.38829107411949254, + "learning_rate": 0.003, + "loss": 4.226, + "step": 2112 + }, + { + "epoch": 0.02113, + "grad_norm": 0.36727891938554497, + "learning_rate": 0.003, + "loss": 4.2209, + "step": 2113 + }, + { + "epoch": 0.02114, + "grad_norm": 0.3796589384288701, + "learning_rate": 0.003, + "loss": 4.2115, + "step": 2114 + }, + { + "epoch": 0.02115, + "grad_norm": 0.49933195549505976, + "learning_rate": 0.003, + "loss": 4.2045, + "step": 2115 + }, + { + "epoch": 0.02116, + "grad_norm": 0.7078427672787017, + "learning_rate": 0.003, + "loss": 4.2392, + "step": 2116 + }, + { + "epoch": 0.02117, + "grad_norm": 0.9422169918121244, + "learning_rate": 0.003, + "loss": 4.2414, + "step": 2117 + }, + { + "epoch": 0.02118, + "grad_norm": 0.8924259465708047, + "learning_rate": 0.003, + "loss": 4.2414, + "step": 2118 + }, + { + "epoch": 0.02119, + "grad_norm": 0.5681560145798641, + "learning_rate": 0.003, + "loss": 4.2397, + "step": 2119 + }, + { + "epoch": 0.0212, + "grad_norm": 0.5070503337124914, + "learning_rate": 0.003, + "loss": 4.213, + "step": 2120 + }, + { + "epoch": 0.02121, + "grad_norm": 0.5980827300624205, + "learning_rate": 0.003, + "loss": 4.2118, + "step": 2121 + }, + { + "epoch": 0.02122, + "grad_norm": 0.48859129709793475, + "learning_rate": 0.003, + "loss": 4.2167, + "step": 2122 + }, + { + "epoch": 0.02123, + "grad_norm": 0.463982076261229, + "learning_rate": 0.003, + "loss": 4.1963, + "step": 2123 + }, + { + "epoch": 0.02124, + "grad_norm": 0.44089509619617673, + "learning_rate": 0.003, + "loss": 4.2174, + "step": 2124 + }, + { + "epoch": 0.02125, + "grad_norm": 0.48261858431331583, + "learning_rate": 0.003, + "loss": 4.2332, + "step": 2125 + }, + { + "epoch": 0.02126, + "grad_norm": 0.5099262151069514, + "learning_rate": 0.003, + "loss": 4.2302, + "step": 2126 + }, + { + "epoch": 0.02127, + "grad_norm": 0.5196898928719708, + "learning_rate": 0.003, + "loss": 4.2379, + "step": 2127 + }, + { + "epoch": 0.02128, + "grad_norm": 0.5656439270875484, + "learning_rate": 0.003, + "loss": 4.2314, + "step": 2128 + }, + { + "epoch": 0.02129, + "grad_norm": 0.5221861829481255, + "learning_rate": 0.003, + "loss": 4.2297, + "step": 2129 + }, + { + "epoch": 0.0213, + "grad_norm": 0.4521057397115596, + "learning_rate": 0.003, + "loss": 4.2027, + "step": 2130 + }, + { + "epoch": 0.02131, + "grad_norm": 0.45364673608644235, + "learning_rate": 0.003, + "loss": 4.2268, + "step": 2131 + }, + { + "epoch": 0.02132, + "grad_norm": 0.519443445123456, + "learning_rate": 0.003, + "loss": 4.2417, + "step": 2132 + }, + { + "epoch": 0.02133, + "grad_norm": 0.7304407810627219, + "learning_rate": 0.003, + "loss": 4.2391, + "step": 2133 + }, + { + "epoch": 0.02134, + "grad_norm": 0.9541768215835498, + "learning_rate": 0.003, + "loss": 4.2522, + "step": 2134 + }, + { + "epoch": 0.02135, + "grad_norm": 1.0307643830076532, + "learning_rate": 0.003, + "loss": 4.2588, + "step": 2135 + }, + { + "epoch": 0.02136, + "grad_norm": 0.997681208751024, + "learning_rate": 0.003, + "loss": 4.2363, + "step": 2136 + }, + { + "epoch": 0.02137, + "grad_norm": 0.8370170695212052, + "learning_rate": 0.003, + "loss": 4.2696, + "step": 2137 + }, + { + "epoch": 0.02138, + "grad_norm": 0.9176192838894591, + "learning_rate": 0.003, + "loss": 4.2858, + "step": 2138 + }, + { + "epoch": 0.02139, + "grad_norm": 0.9058155406944488, + "learning_rate": 0.003, + "loss": 4.2833, + "step": 2139 + }, + { + "epoch": 0.0214, + "grad_norm": 0.8679543052073141, + "learning_rate": 0.003, + "loss": 4.2707, + "step": 2140 + }, + { + "epoch": 0.02141, + "grad_norm": 0.7817830154220525, + "learning_rate": 0.003, + "loss": 4.2527, + "step": 2141 + }, + { + "epoch": 0.02142, + "grad_norm": 0.798437758428762, + "learning_rate": 0.003, + "loss": 4.2453, + "step": 2142 + }, + { + "epoch": 0.02143, + "grad_norm": 0.7871322126205197, + "learning_rate": 0.003, + "loss": 4.2612, + "step": 2143 + }, + { + "epoch": 0.02144, + "grad_norm": 0.8115888487824465, + "learning_rate": 0.003, + "loss": 4.2653, + "step": 2144 + }, + { + "epoch": 0.02145, + "grad_norm": 0.7440847388519684, + "learning_rate": 0.003, + "loss": 4.2448, + "step": 2145 + }, + { + "epoch": 0.02146, + "grad_norm": 0.6836186393980395, + "learning_rate": 0.003, + "loss": 4.2558, + "step": 2146 + }, + { + "epoch": 0.02147, + "grad_norm": 0.6617245709810007, + "learning_rate": 0.003, + "loss": 4.2727, + "step": 2147 + }, + { + "epoch": 0.02148, + "grad_norm": 0.7076496122050561, + "learning_rate": 0.003, + "loss": 4.2832, + "step": 2148 + }, + { + "epoch": 0.02149, + "grad_norm": 0.729078859235357, + "learning_rate": 0.003, + "loss": 4.2694, + "step": 2149 + }, + { + "epoch": 0.0215, + "grad_norm": 0.6732791642286783, + "learning_rate": 0.003, + "loss": 4.2532, + "step": 2150 + }, + { + "epoch": 0.02151, + "grad_norm": 0.6647883181229363, + "learning_rate": 0.003, + "loss": 4.2571, + "step": 2151 + }, + { + "epoch": 0.02152, + "grad_norm": 0.539629911002914, + "learning_rate": 0.003, + "loss": 4.2298, + "step": 2152 + }, + { + "epoch": 0.02153, + "grad_norm": 0.4979893391256132, + "learning_rate": 0.003, + "loss": 4.2372, + "step": 2153 + }, + { + "epoch": 0.02154, + "grad_norm": 0.43644785385536694, + "learning_rate": 0.003, + "loss": 4.2266, + "step": 2154 + }, + { + "epoch": 0.02155, + "grad_norm": 0.3633150618094754, + "learning_rate": 0.003, + "loss": 4.2532, + "step": 2155 + }, + { + "epoch": 0.02156, + "grad_norm": 0.39513834073913734, + "learning_rate": 0.003, + "loss": 4.239, + "step": 2156 + }, + { + "epoch": 0.02157, + "grad_norm": 0.3860719279707547, + "learning_rate": 0.003, + "loss": 4.2034, + "step": 2157 + }, + { + "epoch": 0.02158, + "grad_norm": 0.4137178291066844, + "learning_rate": 0.003, + "loss": 4.2241, + "step": 2158 + }, + { + "epoch": 0.02159, + "grad_norm": 0.5163329954248042, + "learning_rate": 0.003, + "loss": 4.2392, + "step": 2159 + }, + { + "epoch": 0.0216, + "grad_norm": 0.6337653160295451, + "learning_rate": 0.003, + "loss": 4.2443, + "step": 2160 + }, + { + "epoch": 0.02161, + "grad_norm": 0.7844529335341662, + "learning_rate": 0.003, + "loss": 4.2371, + "step": 2161 + }, + { + "epoch": 0.02162, + "grad_norm": 0.8327279956908368, + "learning_rate": 0.003, + "loss": 4.2454, + "step": 2162 + }, + { + "epoch": 0.02163, + "grad_norm": 0.6324138237584291, + "learning_rate": 0.003, + "loss": 4.2546, + "step": 2163 + }, + { + "epoch": 0.02164, + "grad_norm": 0.4272642114449097, + "learning_rate": 0.003, + "loss": 4.2393, + "step": 2164 + }, + { + "epoch": 0.02165, + "grad_norm": 0.5755156335143842, + "learning_rate": 0.003, + "loss": 4.2107, + "step": 2165 + }, + { + "epoch": 0.02166, + "grad_norm": 0.5179351914703093, + "learning_rate": 0.003, + "loss": 4.2452, + "step": 2166 + }, + { + "epoch": 0.02167, + "grad_norm": 0.47795702712318133, + "learning_rate": 0.003, + "loss": 4.2215, + "step": 2167 + }, + { + "epoch": 0.02168, + "grad_norm": 0.39306911706804365, + "learning_rate": 0.003, + "loss": 4.2181, + "step": 2168 + }, + { + "epoch": 0.02169, + "grad_norm": 0.43640956278445237, + "learning_rate": 0.003, + "loss": 4.2172, + "step": 2169 + }, + { + "epoch": 0.0217, + "grad_norm": 0.44151372609355966, + "learning_rate": 0.003, + "loss": 4.2252, + "step": 2170 + }, + { + "epoch": 0.02171, + "grad_norm": 0.4398868813684568, + "learning_rate": 0.003, + "loss": 4.238, + "step": 2171 + }, + { + "epoch": 0.02172, + "grad_norm": 0.453127984813779, + "learning_rate": 0.003, + "loss": 4.2099, + "step": 2172 + }, + { + "epoch": 0.02173, + "grad_norm": 0.4200137719224221, + "learning_rate": 0.003, + "loss": 4.2112, + "step": 2173 + }, + { + "epoch": 0.02174, + "grad_norm": 0.40927864912838796, + "learning_rate": 0.003, + "loss": 4.2074, + "step": 2174 + }, + { + "epoch": 0.02175, + "grad_norm": 0.37273652046629646, + "learning_rate": 0.003, + "loss": 4.2305, + "step": 2175 + }, + { + "epoch": 0.02176, + "grad_norm": 0.3722266811459993, + "learning_rate": 0.003, + "loss": 4.2097, + "step": 2176 + }, + { + "epoch": 0.02177, + "grad_norm": 0.43921830418385466, + "learning_rate": 0.003, + "loss": 4.2187, + "step": 2177 + }, + { + "epoch": 0.02178, + "grad_norm": 0.5214363509579703, + "learning_rate": 0.003, + "loss": 4.1985, + "step": 2178 + }, + { + "epoch": 0.02179, + "grad_norm": 0.8402711445332305, + "learning_rate": 0.003, + "loss": 4.2235, + "step": 2179 + }, + { + "epoch": 0.0218, + "grad_norm": 1.247371610474746, + "learning_rate": 0.003, + "loss": 4.2635, + "step": 2180 + }, + { + "epoch": 0.02181, + "grad_norm": 0.8116435611354782, + "learning_rate": 0.003, + "loss": 4.2484, + "step": 2181 + }, + { + "epoch": 0.02182, + "grad_norm": 0.6837903180839037, + "learning_rate": 0.003, + "loss": 4.2355, + "step": 2182 + }, + { + "epoch": 0.02183, + "grad_norm": 0.7005151877778275, + "learning_rate": 0.003, + "loss": 4.2476, + "step": 2183 + }, + { + "epoch": 0.02184, + "grad_norm": 0.8236745566528523, + "learning_rate": 0.003, + "loss": 4.2265, + "step": 2184 + }, + { + "epoch": 0.02185, + "grad_norm": 0.91418771570573, + "learning_rate": 0.003, + "loss": 4.2327, + "step": 2185 + }, + { + "epoch": 0.02186, + "grad_norm": 0.8323760045804593, + "learning_rate": 0.003, + "loss": 4.2482, + "step": 2186 + }, + { + "epoch": 0.02187, + "grad_norm": 0.8176331489387878, + "learning_rate": 0.003, + "loss": 4.245, + "step": 2187 + }, + { + "epoch": 0.02188, + "grad_norm": 0.8776011343747697, + "learning_rate": 0.003, + "loss": 4.2325, + "step": 2188 + }, + { + "epoch": 0.02189, + "grad_norm": 0.8972484956693035, + "learning_rate": 0.003, + "loss": 4.2667, + "step": 2189 + }, + { + "epoch": 0.0219, + "grad_norm": 0.8431917439790871, + "learning_rate": 0.003, + "loss": 4.2584, + "step": 2190 + }, + { + "epoch": 0.02191, + "grad_norm": 0.7416940909972495, + "learning_rate": 0.003, + "loss": 4.2439, + "step": 2191 + }, + { + "epoch": 0.02192, + "grad_norm": 0.6213980069901923, + "learning_rate": 0.003, + "loss": 4.2676, + "step": 2192 + }, + { + "epoch": 0.02193, + "grad_norm": 0.6597970415430513, + "learning_rate": 0.003, + "loss": 4.2343, + "step": 2193 + }, + { + "epoch": 0.02194, + "grad_norm": 0.7061865144419822, + "learning_rate": 0.003, + "loss": 4.2832, + "step": 2194 + }, + { + "epoch": 0.02195, + "grad_norm": 0.7006382568298751, + "learning_rate": 0.003, + "loss": 4.251, + "step": 2195 + }, + { + "epoch": 0.02196, + "grad_norm": 0.6722854605002367, + "learning_rate": 0.003, + "loss": 4.2565, + "step": 2196 + }, + { + "epoch": 0.02197, + "grad_norm": 0.6383221093399062, + "learning_rate": 0.003, + "loss": 4.2472, + "step": 2197 + }, + { + "epoch": 0.02198, + "grad_norm": 0.6598256711875262, + "learning_rate": 0.003, + "loss": 4.2361, + "step": 2198 + }, + { + "epoch": 0.02199, + "grad_norm": 0.6310822502739497, + "learning_rate": 0.003, + "loss": 4.2038, + "step": 2199 + }, + { + "epoch": 0.022, + "grad_norm": 0.5807092706295028, + "learning_rate": 0.003, + "loss": 4.2307, + "step": 2200 + }, + { + "epoch": 0.02201, + "grad_norm": 0.5164997329736423, + "learning_rate": 0.003, + "loss": 4.2429, + "step": 2201 + }, + { + "epoch": 0.02202, + "grad_norm": 0.5150399093991901, + "learning_rate": 0.003, + "loss": 4.2262, + "step": 2202 + }, + { + "epoch": 0.02203, + "grad_norm": 0.5301942341385608, + "learning_rate": 0.003, + "loss": 4.2546, + "step": 2203 + }, + { + "epoch": 0.02204, + "grad_norm": 0.5633853956311704, + "learning_rate": 0.003, + "loss": 4.2484, + "step": 2204 + }, + { + "epoch": 0.02205, + "grad_norm": 0.5924154634458846, + "learning_rate": 0.003, + "loss": 4.2516, + "step": 2205 + }, + { + "epoch": 0.02206, + "grad_norm": 0.6443822475335407, + "learning_rate": 0.003, + "loss": 4.2165, + "step": 2206 + }, + { + "epoch": 0.02207, + "grad_norm": 0.6602238711518633, + "learning_rate": 0.003, + "loss": 4.2293, + "step": 2207 + }, + { + "epoch": 0.02208, + "grad_norm": 0.6586527270486826, + "learning_rate": 0.003, + "loss": 4.2124, + "step": 2208 + }, + { + "epoch": 0.02209, + "grad_norm": 0.5991461642461526, + "learning_rate": 0.003, + "loss": 4.2276, + "step": 2209 + }, + { + "epoch": 0.0221, + "grad_norm": 0.5268358370568649, + "learning_rate": 0.003, + "loss": 4.2071, + "step": 2210 + }, + { + "epoch": 0.02211, + "grad_norm": 0.6149216818507237, + "learning_rate": 0.003, + "loss": 4.2195, + "step": 2211 + }, + { + "epoch": 0.02212, + "grad_norm": 0.7463716016406938, + "learning_rate": 0.003, + "loss": 4.2226, + "step": 2212 + }, + { + "epoch": 0.02213, + "grad_norm": 0.7963090983602238, + "learning_rate": 0.003, + "loss": 4.2341, + "step": 2213 + }, + { + "epoch": 0.02214, + "grad_norm": 0.7746780288539415, + "learning_rate": 0.003, + "loss": 4.2283, + "step": 2214 + }, + { + "epoch": 0.02215, + "grad_norm": 0.7305597710999601, + "learning_rate": 0.003, + "loss": 4.2338, + "step": 2215 + }, + { + "epoch": 0.02216, + "grad_norm": 0.6802860521431111, + "learning_rate": 0.003, + "loss": 4.2097, + "step": 2216 + }, + { + "epoch": 0.02217, + "grad_norm": 0.814645356813129, + "learning_rate": 0.003, + "loss": 4.2149, + "step": 2217 + }, + { + "epoch": 0.02218, + "grad_norm": 0.8644777161103991, + "learning_rate": 0.003, + "loss": 4.2502, + "step": 2218 + }, + { + "epoch": 0.02219, + "grad_norm": 0.8895266477572761, + "learning_rate": 0.003, + "loss": 4.2436, + "step": 2219 + }, + { + "epoch": 0.0222, + "grad_norm": 0.8379487765766322, + "learning_rate": 0.003, + "loss": 4.241, + "step": 2220 + }, + { + "epoch": 0.02221, + "grad_norm": 0.8832652718422841, + "learning_rate": 0.003, + "loss": 4.23, + "step": 2221 + }, + { + "epoch": 0.02222, + "grad_norm": 0.7112453634526715, + "learning_rate": 0.003, + "loss": 4.2245, + "step": 2222 + }, + { + "epoch": 0.02223, + "grad_norm": 0.5578145775957193, + "learning_rate": 0.003, + "loss": 4.2345, + "step": 2223 + }, + { + "epoch": 0.02224, + "grad_norm": 0.5281644921526645, + "learning_rate": 0.003, + "loss": 4.216, + "step": 2224 + }, + { + "epoch": 0.02225, + "grad_norm": 0.499018726060841, + "learning_rate": 0.003, + "loss": 4.216, + "step": 2225 + }, + { + "epoch": 0.02226, + "grad_norm": 0.5772234482094156, + "learning_rate": 0.003, + "loss": 4.2362, + "step": 2226 + }, + { + "epoch": 0.02227, + "grad_norm": 0.6481086517118151, + "learning_rate": 0.003, + "loss": 4.2311, + "step": 2227 + }, + { + "epoch": 0.02228, + "grad_norm": 0.669454673605761, + "learning_rate": 0.003, + "loss": 4.2187, + "step": 2228 + }, + { + "epoch": 0.02229, + "grad_norm": 0.677093243217859, + "learning_rate": 0.003, + "loss": 4.1992, + "step": 2229 + }, + { + "epoch": 0.0223, + "grad_norm": 0.6785810316864852, + "learning_rate": 0.003, + "loss": 4.2386, + "step": 2230 + }, + { + "epoch": 0.02231, + "grad_norm": 0.697493374925287, + "learning_rate": 0.003, + "loss": 4.2447, + "step": 2231 + }, + { + "epoch": 0.02232, + "grad_norm": 0.8916380613416046, + "learning_rate": 0.003, + "loss": 4.2575, + "step": 2232 + }, + { + "epoch": 0.02233, + "grad_norm": 1.0887336160767505, + "learning_rate": 0.003, + "loss": 4.2371, + "step": 2233 + }, + { + "epoch": 0.02234, + "grad_norm": 0.654547980354846, + "learning_rate": 0.003, + "loss": 4.2454, + "step": 2234 + }, + { + "epoch": 0.02235, + "grad_norm": 0.5125878831412668, + "learning_rate": 0.003, + "loss": 4.2654, + "step": 2235 + }, + { + "epoch": 0.02236, + "grad_norm": 0.7208588554620924, + "learning_rate": 0.003, + "loss": 4.2213, + "step": 2236 + }, + { + "epoch": 0.02237, + "grad_norm": 0.7459329466933863, + "learning_rate": 0.003, + "loss": 4.2569, + "step": 2237 + }, + { + "epoch": 0.02238, + "grad_norm": 0.7047944961318278, + "learning_rate": 0.003, + "loss": 4.2359, + "step": 2238 + }, + { + "epoch": 0.02239, + "grad_norm": 0.5989118371963864, + "learning_rate": 0.003, + "loss": 4.2173, + "step": 2239 + }, + { + "epoch": 0.0224, + "grad_norm": 0.4980068818057585, + "learning_rate": 0.003, + "loss": 4.1972, + "step": 2240 + }, + { + "epoch": 0.02241, + "grad_norm": 0.4765997364648245, + "learning_rate": 0.003, + "loss": 4.2278, + "step": 2241 + }, + { + "epoch": 0.02242, + "grad_norm": 0.46816719473424767, + "learning_rate": 0.003, + "loss": 4.2111, + "step": 2242 + }, + { + "epoch": 0.02243, + "grad_norm": 0.4356611881103323, + "learning_rate": 0.003, + "loss": 4.2383, + "step": 2243 + }, + { + "epoch": 0.02244, + "grad_norm": 0.5122363302058777, + "learning_rate": 0.003, + "loss": 4.2149, + "step": 2244 + }, + { + "epoch": 0.02245, + "grad_norm": 0.48883381472091325, + "learning_rate": 0.003, + "loss": 4.2031, + "step": 2245 + }, + { + "epoch": 0.02246, + "grad_norm": 0.46925512406079145, + "learning_rate": 0.003, + "loss": 4.2303, + "step": 2246 + }, + { + "epoch": 0.02247, + "grad_norm": 0.4305805801856768, + "learning_rate": 0.003, + "loss": 4.2025, + "step": 2247 + }, + { + "epoch": 0.02248, + "grad_norm": 0.49466299772622346, + "learning_rate": 0.003, + "loss": 4.2113, + "step": 2248 + }, + { + "epoch": 0.02249, + "grad_norm": 0.5526228295993102, + "learning_rate": 0.003, + "loss": 4.204, + "step": 2249 + }, + { + "epoch": 0.0225, + "grad_norm": 0.5428701465777451, + "learning_rate": 0.003, + "loss": 4.219, + "step": 2250 + }, + { + "epoch": 0.02251, + "grad_norm": 0.5748138491222456, + "learning_rate": 0.003, + "loss": 4.242, + "step": 2251 + }, + { + "epoch": 0.02252, + "grad_norm": 0.5627410118560184, + "learning_rate": 0.003, + "loss": 4.2308, + "step": 2252 + }, + { + "epoch": 0.02253, + "grad_norm": 0.5464235773202091, + "learning_rate": 0.003, + "loss": 4.2107, + "step": 2253 + }, + { + "epoch": 0.02254, + "grad_norm": 0.5460056774038026, + "learning_rate": 0.003, + "loss": 4.2195, + "step": 2254 + }, + { + "epoch": 0.02255, + "grad_norm": 0.5905579639478153, + "learning_rate": 0.003, + "loss": 4.2038, + "step": 2255 + }, + { + "epoch": 0.02256, + "grad_norm": 0.5996630986155933, + "learning_rate": 0.003, + "loss": 4.1955, + "step": 2256 + }, + { + "epoch": 0.02257, + "grad_norm": 0.5786721112589904, + "learning_rate": 0.003, + "loss": 4.2091, + "step": 2257 + }, + { + "epoch": 0.02258, + "grad_norm": 0.6117786281389065, + "learning_rate": 0.003, + "loss": 4.2219, + "step": 2258 + }, + { + "epoch": 0.02259, + "grad_norm": 0.6324462778570742, + "learning_rate": 0.003, + "loss": 4.1884, + "step": 2259 + }, + { + "epoch": 0.0226, + "grad_norm": 0.5884786495082438, + "learning_rate": 0.003, + "loss": 4.2409, + "step": 2260 + }, + { + "epoch": 0.02261, + "grad_norm": 0.57968484048356, + "learning_rate": 0.003, + "loss": 4.2168, + "step": 2261 + }, + { + "epoch": 0.02262, + "grad_norm": 0.6074577050747278, + "learning_rate": 0.003, + "loss": 4.2183, + "step": 2262 + }, + { + "epoch": 0.02263, + "grad_norm": 0.7175518494233851, + "learning_rate": 0.003, + "loss": 4.237, + "step": 2263 + }, + { + "epoch": 0.02264, + "grad_norm": 0.7379078827688124, + "learning_rate": 0.003, + "loss": 4.2337, + "step": 2264 + }, + { + "epoch": 0.02265, + "grad_norm": 0.8347809834095073, + "learning_rate": 0.003, + "loss": 4.194, + "step": 2265 + }, + { + "epoch": 0.02266, + "grad_norm": 0.9870112085605878, + "learning_rate": 0.003, + "loss": 4.2201, + "step": 2266 + }, + { + "epoch": 0.02267, + "grad_norm": 1.1653841189257785, + "learning_rate": 0.003, + "loss": 4.2605, + "step": 2267 + }, + { + "epoch": 0.02268, + "grad_norm": 0.7766778764734964, + "learning_rate": 0.003, + "loss": 4.2198, + "step": 2268 + }, + { + "epoch": 0.02269, + "grad_norm": 0.7319600539673422, + "learning_rate": 0.003, + "loss": 4.2435, + "step": 2269 + }, + { + "epoch": 0.0227, + "grad_norm": 0.8988269964618911, + "learning_rate": 0.003, + "loss": 4.2441, + "step": 2270 + }, + { + "epoch": 0.02271, + "grad_norm": 1.0408382709602264, + "learning_rate": 0.003, + "loss": 4.2637, + "step": 2271 + }, + { + "epoch": 0.02272, + "grad_norm": 1.0101614929632108, + "learning_rate": 0.003, + "loss": 4.2626, + "step": 2272 + }, + { + "epoch": 0.02273, + "grad_norm": 0.8989352397704572, + "learning_rate": 0.003, + "loss": 4.2643, + "step": 2273 + }, + { + "epoch": 0.02274, + "grad_norm": 0.68535867805162, + "learning_rate": 0.003, + "loss": 4.2412, + "step": 2274 + }, + { + "epoch": 0.02275, + "grad_norm": 0.6683944403644445, + "learning_rate": 0.003, + "loss": 4.2325, + "step": 2275 + }, + { + "epoch": 0.02276, + "grad_norm": 0.6494300645141144, + "learning_rate": 0.003, + "loss": 4.2263, + "step": 2276 + }, + { + "epoch": 0.02277, + "grad_norm": 0.6973670182968851, + "learning_rate": 0.003, + "loss": 4.2382, + "step": 2277 + }, + { + "epoch": 0.02278, + "grad_norm": 0.7902320444222205, + "learning_rate": 0.003, + "loss": 4.2463, + "step": 2278 + }, + { + "epoch": 0.02279, + "grad_norm": 0.8146336114310585, + "learning_rate": 0.003, + "loss": 4.2598, + "step": 2279 + }, + { + "epoch": 0.0228, + "grad_norm": 0.8796339021642547, + "learning_rate": 0.003, + "loss": 4.2507, + "step": 2280 + }, + { + "epoch": 0.02281, + "grad_norm": 0.8184015106359132, + "learning_rate": 0.003, + "loss": 4.2484, + "step": 2281 + }, + { + "epoch": 0.02282, + "grad_norm": 0.6892747359251186, + "learning_rate": 0.003, + "loss": 4.2524, + "step": 2282 + }, + { + "epoch": 0.02283, + "grad_norm": 0.6565456722535347, + "learning_rate": 0.003, + "loss": 4.2659, + "step": 2283 + }, + { + "epoch": 0.02284, + "grad_norm": 0.5726981092319336, + "learning_rate": 0.003, + "loss": 4.2571, + "step": 2284 + }, + { + "epoch": 0.02285, + "grad_norm": 0.5302500755986935, + "learning_rate": 0.003, + "loss": 4.1961, + "step": 2285 + }, + { + "epoch": 0.02286, + "grad_norm": 0.5726425860790162, + "learning_rate": 0.003, + "loss": 4.2575, + "step": 2286 + }, + { + "epoch": 0.02287, + "grad_norm": 0.5919298156896359, + "learning_rate": 0.003, + "loss": 4.2367, + "step": 2287 + }, + { + "epoch": 0.02288, + "grad_norm": 0.5989015507067659, + "learning_rate": 0.003, + "loss": 4.2599, + "step": 2288 + }, + { + "epoch": 0.02289, + "grad_norm": 0.6577684390532245, + "learning_rate": 0.003, + "loss": 4.2586, + "step": 2289 + }, + { + "epoch": 0.0229, + "grad_norm": 0.8096348684056334, + "learning_rate": 0.003, + "loss": 4.2424, + "step": 2290 + }, + { + "epoch": 0.02291, + "grad_norm": 0.8395348686742803, + "learning_rate": 0.003, + "loss": 4.2207, + "step": 2291 + }, + { + "epoch": 0.02292, + "grad_norm": 0.6824435433690341, + "learning_rate": 0.003, + "loss": 4.2133, + "step": 2292 + }, + { + "epoch": 0.02293, + "grad_norm": 0.5182235922586542, + "learning_rate": 0.003, + "loss": 4.2116, + "step": 2293 + }, + { + "epoch": 0.02294, + "grad_norm": 0.5893445423849215, + "learning_rate": 0.003, + "loss": 4.2139, + "step": 2294 + }, + { + "epoch": 0.02295, + "grad_norm": 0.5909258163581438, + "learning_rate": 0.003, + "loss": 4.2083, + "step": 2295 + }, + { + "epoch": 0.02296, + "grad_norm": 0.49380609648697416, + "learning_rate": 0.003, + "loss": 4.2381, + "step": 2296 + }, + { + "epoch": 0.02297, + "grad_norm": 0.45509453612267814, + "learning_rate": 0.003, + "loss": 4.2063, + "step": 2297 + }, + { + "epoch": 0.02298, + "grad_norm": 0.39795367258729464, + "learning_rate": 0.003, + "loss": 4.2154, + "step": 2298 + }, + { + "epoch": 0.02299, + "grad_norm": 0.41864078809034155, + "learning_rate": 0.003, + "loss": 4.2206, + "step": 2299 + }, + { + "epoch": 0.023, + "grad_norm": 0.45486028161368913, + "learning_rate": 0.003, + "loss": 4.2385, + "step": 2300 + }, + { + "epoch": 0.02301, + "grad_norm": 0.4383928009279485, + "learning_rate": 0.003, + "loss": 4.2022, + "step": 2301 + }, + { + "epoch": 0.02302, + "grad_norm": 0.4499222406460722, + "learning_rate": 0.003, + "loss": 4.2089, + "step": 2302 + }, + { + "epoch": 0.02303, + "grad_norm": 0.4500683500939326, + "learning_rate": 0.003, + "loss": 4.1978, + "step": 2303 + }, + { + "epoch": 0.02304, + "grad_norm": 0.4684736001381324, + "learning_rate": 0.003, + "loss": 4.2294, + "step": 2304 + }, + { + "epoch": 0.02305, + "grad_norm": 0.4868306442044598, + "learning_rate": 0.003, + "loss": 4.2051, + "step": 2305 + }, + { + "epoch": 0.02306, + "grad_norm": 0.5334764519084825, + "learning_rate": 0.003, + "loss": 4.1822, + "step": 2306 + }, + { + "epoch": 0.02307, + "grad_norm": 0.6492467780995748, + "learning_rate": 0.003, + "loss": 4.2166, + "step": 2307 + }, + { + "epoch": 0.02308, + "grad_norm": 0.7095262909438931, + "learning_rate": 0.003, + "loss": 4.2125, + "step": 2308 + }, + { + "epoch": 0.02309, + "grad_norm": 0.54863336292105, + "learning_rate": 0.003, + "loss": 4.2349, + "step": 2309 + }, + { + "epoch": 0.0231, + "grad_norm": 0.5340829751385735, + "learning_rate": 0.003, + "loss": 4.2136, + "step": 2310 + }, + { + "epoch": 0.02311, + "grad_norm": 0.6450529599308674, + "learning_rate": 0.003, + "loss": 4.2222, + "step": 2311 + }, + { + "epoch": 0.02312, + "grad_norm": 0.820530552930021, + "learning_rate": 0.003, + "loss": 4.2023, + "step": 2312 + }, + { + "epoch": 0.02313, + "grad_norm": 0.8724275505071452, + "learning_rate": 0.003, + "loss": 4.2036, + "step": 2313 + }, + { + "epoch": 0.02314, + "grad_norm": 0.7571797533207066, + "learning_rate": 0.003, + "loss": 4.2199, + "step": 2314 + }, + { + "epoch": 0.02315, + "grad_norm": 0.6953439594266967, + "learning_rate": 0.003, + "loss": 4.221, + "step": 2315 + }, + { + "epoch": 0.02316, + "grad_norm": 0.7680069713239076, + "learning_rate": 0.003, + "loss": 4.2379, + "step": 2316 + }, + { + "epoch": 0.02317, + "grad_norm": 0.8161604949475788, + "learning_rate": 0.003, + "loss": 4.223, + "step": 2317 + }, + { + "epoch": 0.02318, + "grad_norm": 0.7516223217475189, + "learning_rate": 0.003, + "loss": 4.2136, + "step": 2318 + }, + { + "epoch": 0.02319, + "grad_norm": 0.7000350308630269, + "learning_rate": 0.003, + "loss": 4.2358, + "step": 2319 + }, + { + "epoch": 0.0232, + "grad_norm": 0.7134745769645445, + "learning_rate": 0.003, + "loss": 4.2452, + "step": 2320 + }, + { + "epoch": 0.02321, + "grad_norm": 0.771931890706088, + "learning_rate": 0.003, + "loss": 4.2279, + "step": 2321 + }, + { + "epoch": 0.02322, + "grad_norm": 0.9854722178210035, + "learning_rate": 0.003, + "loss": 4.2409, + "step": 2322 + }, + { + "epoch": 0.02323, + "grad_norm": 0.9767405875240949, + "learning_rate": 0.003, + "loss": 4.2246, + "step": 2323 + }, + { + "epoch": 0.02324, + "grad_norm": 0.8048742564770486, + "learning_rate": 0.003, + "loss": 4.2207, + "step": 2324 + }, + { + "epoch": 0.02325, + "grad_norm": 0.7747178630515028, + "learning_rate": 0.003, + "loss": 4.242, + "step": 2325 + }, + { + "epoch": 0.02326, + "grad_norm": 0.7969057026631506, + "learning_rate": 0.003, + "loss": 4.2598, + "step": 2326 + }, + { + "epoch": 0.02327, + "grad_norm": 0.8966295433578796, + "learning_rate": 0.003, + "loss": 4.2563, + "step": 2327 + }, + { + "epoch": 0.02328, + "grad_norm": 0.7936452621630464, + "learning_rate": 0.003, + "loss": 4.2524, + "step": 2328 + }, + { + "epoch": 0.02329, + "grad_norm": 0.6876802756045969, + "learning_rate": 0.003, + "loss": 4.2386, + "step": 2329 + }, + { + "epoch": 0.0233, + "grad_norm": 0.7923806233690935, + "learning_rate": 0.003, + "loss": 4.2331, + "step": 2330 + }, + { + "epoch": 0.02331, + "grad_norm": 0.8629660125432402, + "learning_rate": 0.003, + "loss": 4.278, + "step": 2331 + }, + { + "epoch": 0.02332, + "grad_norm": 0.8224989442560882, + "learning_rate": 0.003, + "loss": 4.2382, + "step": 2332 + }, + { + "epoch": 0.02333, + "grad_norm": 0.7543416736322919, + "learning_rate": 0.003, + "loss": 4.2433, + "step": 2333 + }, + { + "epoch": 0.02334, + "grad_norm": 0.6971129018109938, + "learning_rate": 0.003, + "loss": 4.209, + "step": 2334 + }, + { + "epoch": 0.02335, + "grad_norm": 0.7296913197019357, + "learning_rate": 0.003, + "loss": 4.2519, + "step": 2335 + }, + { + "epoch": 0.02336, + "grad_norm": 0.7163909054433683, + "learning_rate": 0.003, + "loss": 4.2383, + "step": 2336 + }, + { + "epoch": 0.02337, + "grad_norm": 0.6354188850134015, + "learning_rate": 0.003, + "loss": 4.2384, + "step": 2337 + }, + { + "epoch": 0.02338, + "grad_norm": 0.7380334344563654, + "learning_rate": 0.003, + "loss": 4.2287, + "step": 2338 + }, + { + "epoch": 0.02339, + "grad_norm": 0.6731414857527438, + "learning_rate": 0.003, + "loss": 4.2221, + "step": 2339 + }, + { + "epoch": 0.0234, + "grad_norm": 0.5402269466528953, + "learning_rate": 0.003, + "loss": 4.2314, + "step": 2340 + }, + { + "epoch": 0.02341, + "grad_norm": 0.442581637819893, + "learning_rate": 0.003, + "loss": 4.2031, + "step": 2341 + }, + { + "epoch": 0.02342, + "grad_norm": 0.37485389657180546, + "learning_rate": 0.003, + "loss": 4.2138, + "step": 2342 + }, + { + "epoch": 0.02343, + "grad_norm": 0.3841947544396341, + "learning_rate": 0.003, + "loss": 4.213, + "step": 2343 + }, + { + "epoch": 0.02344, + "grad_norm": 0.37024132956414024, + "learning_rate": 0.003, + "loss": 4.2104, + "step": 2344 + }, + { + "epoch": 0.02345, + "grad_norm": 0.3532869549675326, + "learning_rate": 0.003, + "loss": 4.2035, + "step": 2345 + }, + { + "epoch": 0.02346, + "grad_norm": 0.3470711196570803, + "learning_rate": 0.003, + "loss": 4.1998, + "step": 2346 + }, + { + "epoch": 0.02347, + "grad_norm": 0.4249636779242577, + "learning_rate": 0.003, + "loss": 4.2374, + "step": 2347 + }, + { + "epoch": 0.02348, + "grad_norm": 0.5336097138298947, + "learning_rate": 0.003, + "loss": 4.2032, + "step": 2348 + }, + { + "epoch": 0.02349, + "grad_norm": 0.7560598476745078, + "learning_rate": 0.003, + "loss": 4.2252, + "step": 2349 + }, + { + "epoch": 0.0235, + "grad_norm": 0.9366290665402001, + "learning_rate": 0.003, + "loss": 4.2316, + "step": 2350 + }, + { + "epoch": 0.02351, + "grad_norm": 0.8241230319323669, + "learning_rate": 0.003, + "loss": 4.2392, + "step": 2351 + }, + { + "epoch": 0.02352, + "grad_norm": 0.4838312119944649, + "learning_rate": 0.003, + "loss": 4.1838, + "step": 2352 + }, + { + "epoch": 0.02353, + "grad_norm": 0.5962852237534145, + "learning_rate": 0.003, + "loss": 4.2218, + "step": 2353 + }, + { + "epoch": 0.02354, + "grad_norm": 0.6656896770625591, + "learning_rate": 0.003, + "loss": 4.2262, + "step": 2354 + }, + { + "epoch": 0.02355, + "grad_norm": 0.5663365054891429, + "learning_rate": 0.003, + "loss": 4.2216, + "step": 2355 + }, + { + "epoch": 0.02356, + "grad_norm": 0.4266142835453117, + "learning_rate": 0.003, + "loss": 4.1971, + "step": 2356 + }, + { + "epoch": 0.02357, + "grad_norm": 0.4433400684484211, + "learning_rate": 0.003, + "loss": 4.1998, + "step": 2357 + }, + { + "epoch": 0.02358, + "grad_norm": 0.5052976773226068, + "learning_rate": 0.003, + "loss": 4.2226, + "step": 2358 + }, + { + "epoch": 0.02359, + "grad_norm": 0.45316233898993824, + "learning_rate": 0.003, + "loss": 4.1954, + "step": 2359 + }, + { + "epoch": 0.0236, + "grad_norm": 0.4014671488900521, + "learning_rate": 0.003, + "loss": 4.2325, + "step": 2360 + }, + { + "epoch": 0.02361, + "grad_norm": 0.36698656753000536, + "learning_rate": 0.003, + "loss": 4.212, + "step": 2361 + }, + { + "epoch": 0.02362, + "grad_norm": 0.42314800926533436, + "learning_rate": 0.003, + "loss": 4.2053, + "step": 2362 + }, + { + "epoch": 0.02363, + "grad_norm": 0.4217103245805847, + "learning_rate": 0.003, + "loss": 4.1792, + "step": 2363 + }, + { + "epoch": 0.02364, + "grad_norm": 0.4210664479788169, + "learning_rate": 0.003, + "loss": 4.2013, + "step": 2364 + }, + { + "epoch": 0.02365, + "grad_norm": 0.4933666084308532, + "learning_rate": 0.003, + "loss": 4.2208, + "step": 2365 + }, + { + "epoch": 0.02366, + "grad_norm": 0.5419476181345154, + "learning_rate": 0.003, + "loss": 4.1974, + "step": 2366 + }, + { + "epoch": 0.02367, + "grad_norm": 0.5315884871641553, + "learning_rate": 0.003, + "loss": 4.2086, + "step": 2367 + }, + { + "epoch": 0.02368, + "grad_norm": 0.5513031505324636, + "learning_rate": 0.003, + "loss": 4.1821, + "step": 2368 + }, + { + "epoch": 0.02369, + "grad_norm": 0.6237648160785263, + "learning_rate": 0.003, + "loss": 4.1978, + "step": 2369 + }, + { + "epoch": 0.0237, + "grad_norm": 0.7305017234045693, + "learning_rate": 0.003, + "loss": 4.2311, + "step": 2370 + }, + { + "epoch": 0.02371, + "grad_norm": 0.7544509646548179, + "learning_rate": 0.003, + "loss": 4.2181, + "step": 2371 + }, + { + "epoch": 0.02372, + "grad_norm": 0.8927671954437898, + "learning_rate": 0.003, + "loss": 4.2125, + "step": 2372 + }, + { + "epoch": 0.02373, + "grad_norm": 0.8832047313566855, + "learning_rate": 0.003, + "loss": 4.2061, + "step": 2373 + }, + { + "epoch": 0.02374, + "grad_norm": 0.9298955170309446, + "learning_rate": 0.003, + "loss": 4.2122, + "step": 2374 + }, + { + "epoch": 0.02375, + "grad_norm": 0.8640645189121525, + "learning_rate": 0.003, + "loss": 4.2188, + "step": 2375 + }, + { + "epoch": 0.02376, + "grad_norm": 0.8134531891441159, + "learning_rate": 0.003, + "loss": 4.2414, + "step": 2376 + }, + { + "epoch": 0.02377, + "grad_norm": 0.8309736476293886, + "learning_rate": 0.003, + "loss": 4.2122, + "step": 2377 + }, + { + "epoch": 0.02378, + "grad_norm": 0.8951590746635169, + "learning_rate": 0.003, + "loss": 4.2274, + "step": 2378 + }, + { + "epoch": 0.02379, + "grad_norm": 0.8950538453176826, + "learning_rate": 0.003, + "loss": 4.2181, + "step": 2379 + }, + { + "epoch": 0.0238, + "grad_norm": 0.8724035600141309, + "learning_rate": 0.003, + "loss": 4.1913, + "step": 2380 + }, + { + "epoch": 0.02381, + "grad_norm": 0.7988879477893653, + "learning_rate": 0.003, + "loss": 4.2027, + "step": 2381 + }, + { + "epoch": 0.02382, + "grad_norm": 0.7723567017862097, + "learning_rate": 0.003, + "loss": 4.2578, + "step": 2382 + }, + { + "epoch": 0.02383, + "grad_norm": 0.671665890468045, + "learning_rate": 0.003, + "loss": 4.2171, + "step": 2383 + }, + { + "epoch": 0.02384, + "grad_norm": 0.6389101019379912, + "learning_rate": 0.003, + "loss": 4.2056, + "step": 2384 + }, + { + "epoch": 0.02385, + "grad_norm": 0.580627279569301, + "learning_rate": 0.003, + "loss": 4.2294, + "step": 2385 + }, + { + "epoch": 0.02386, + "grad_norm": 0.6112201616439135, + "learning_rate": 0.003, + "loss": 4.2052, + "step": 2386 + }, + { + "epoch": 0.02387, + "grad_norm": 0.609528097348241, + "learning_rate": 0.003, + "loss": 4.2032, + "step": 2387 + }, + { + "epoch": 0.02388, + "grad_norm": 0.6686638362844006, + "learning_rate": 0.003, + "loss": 4.2564, + "step": 2388 + }, + { + "epoch": 0.02389, + "grad_norm": 0.740840757442375, + "learning_rate": 0.003, + "loss": 4.2206, + "step": 2389 + }, + { + "epoch": 0.0239, + "grad_norm": 0.795817868194079, + "learning_rate": 0.003, + "loss": 4.2173, + "step": 2390 + }, + { + "epoch": 0.02391, + "grad_norm": 0.6885254297129437, + "learning_rate": 0.003, + "loss": 4.2366, + "step": 2391 + }, + { + "epoch": 0.02392, + "grad_norm": 0.57208190101736, + "learning_rate": 0.003, + "loss": 4.2339, + "step": 2392 + }, + { + "epoch": 0.02393, + "grad_norm": 0.6766423034488698, + "learning_rate": 0.003, + "loss": 4.2001, + "step": 2393 + }, + { + "epoch": 0.02394, + "grad_norm": 0.7708917314903707, + "learning_rate": 0.003, + "loss": 4.219, + "step": 2394 + }, + { + "epoch": 0.02395, + "grad_norm": 0.8419139355655553, + "learning_rate": 0.003, + "loss": 4.234, + "step": 2395 + }, + { + "epoch": 0.02396, + "grad_norm": 0.7491203962086143, + "learning_rate": 0.003, + "loss": 4.2487, + "step": 2396 + }, + { + "epoch": 0.02397, + "grad_norm": 0.6046499284946664, + "learning_rate": 0.003, + "loss": 4.2249, + "step": 2397 + }, + { + "epoch": 0.02398, + "grad_norm": 0.654682549696845, + "learning_rate": 0.003, + "loss": 4.2309, + "step": 2398 + }, + { + "epoch": 0.02399, + "grad_norm": 0.600390647492162, + "learning_rate": 0.003, + "loss": 4.209, + "step": 2399 + }, + { + "epoch": 0.024, + "grad_norm": 0.5155181636394677, + "learning_rate": 0.003, + "loss": 4.2198, + "step": 2400 + }, + { + "epoch": 0.02401, + "grad_norm": 0.48968723675185105, + "learning_rate": 0.003, + "loss": 4.1921, + "step": 2401 + }, + { + "epoch": 0.02402, + "grad_norm": 0.5172416496163427, + "learning_rate": 0.003, + "loss": 4.2245, + "step": 2402 + }, + { + "epoch": 0.02403, + "grad_norm": 0.5647625943348626, + "learning_rate": 0.003, + "loss": 4.1949, + "step": 2403 + }, + { + "epoch": 0.02404, + "grad_norm": 0.5032714506791921, + "learning_rate": 0.003, + "loss": 4.233, + "step": 2404 + }, + { + "epoch": 0.02405, + "grad_norm": 0.4675987670630769, + "learning_rate": 0.003, + "loss": 4.2334, + "step": 2405 + }, + { + "epoch": 0.02406, + "grad_norm": 0.5178645102430114, + "learning_rate": 0.003, + "loss": 4.2192, + "step": 2406 + }, + { + "epoch": 0.02407, + "grad_norm": 0.5559303502201767, + "learning_rate": 0.003, + "loss": 4.1976, + "step": 2407 + }, + { + "epoch": 0.02408, + "grad_norm": 0.551443134128045, + "learning_rate": 0.003, + "loss": 4.2064, + "step": 2408 + }, + { + "epoch": 0.02409, + "grad_norm": 0.5057926630793051, + "learning_rate": 0.003, + "loss": 4.2079, + "step": 2409 + }, + { + "epoch": 0.0241, + "grad_norm": 0.6021866868934532, + "learning_rate": 0.003, + "loss": 4.2249, + "step": 2410 + }, + { + "epoch": 0.02411, + "grad_norm": 0.7613042549642509, + "learning_rate": 0.003, + "loss": 4.1979, + "step": 2411 + }, + { + "epoch": 0.02412, + "grad_norm": 0.7823591519195993, + "learning_rate": 0.003, + "loss": 4.2054, + "step": 2412 + }, + { + "epoch": 0.02413, + "grad_norm": 0.6762814430739601, + "learning_rate": 0.003, + "loss": 4.2005, + "step": 2413 + }, + { + "epoch": 0.02414, + "grad_norm": 0.6515291647943621, + "learning_rate": 0.003, + "loss": 4.2148, + "step": 2414 + }, + { + "epoch": 0.02415, + "grad_norm": 0.5905595191694923, + "learning_rate": 0.003, + "loss": 4.2125, + "step": 2415 + }, + { + "epoch": 0.02416, + "grad_norm": 0.5267459284750431, + "learning_rate": 0.003, + "loss": 4.1927, + "step": 2416 + }, + { + "epoch": 0.02417, + "grad_norm": 0.6612447088265828, + "learning_rate": 0.003, + "loss": 4.2248, + "step": 2417 + }, + { + "epoch": 0.02418, + "grad_norm": 0.7551414883340356, + "learning_rate": 0.003, + "loss": 4.2234, + "step": 2418 + }, + { + "epoch": 0.02419, + "grad_norm": 0.8205697648408953, + "learning_rate": 0.003, + "loss": 4.2244, + "step": 2419 + }, + { + "epoch": 0.0242, + "grad_norm": 0.8899430556748202, + "learning_rate": 0.003, + "loss": 4.2436, + "step": 2420 + }, + { + "epoch": 0.02421, + "grad_norm": 0.7830186936394068, + "learning_rate": 0.003, + "loss": 4.2007, + "step": 2421 + }, + { + "epoch": 0.02422, + "grad_norm": 0.707149933198692, + "learning_rate": 0.003, + "loss": 4.2528, + "step": 2422 + }, + { + "epoch": 0.02423, + "grad_norm": 0.6635737571843986, + "learning_rate": 0.003, + "loss": 4.2424, + "step": 2423 + }, + { + "epoch": 0.02424, + "grad_norm": 0.6518995770283964, + "learning_rate": 0.003, + "loss": 4.2374, + "step": 2424 + }, + { + "epoch": 0.02425, + "grad_norm": 0.6795937334698865, + "learning_rate": 0.003, + "loss": 4.2233, + "step": 2425 + }, + { + "epoch": 0.02426, + "grad_norm": 0.7581646116983769, + "learning_rate": 0.003, + "loss": 4.2315, + "step": 2426 + }, + { + "epoch": 0.02427, + "grad_norm": 0.9154720690249456, + "learning_rate": 0.003, + "loss": 4.2647, + "step": 2427 + }, + { + "epoch": 0.02428, + "grad_norm": 0.9816907611239863, + "learning_rate": 0.003, + "loss": 4.2433, + "step": 2428 + }, + { + "epoch": 0.02429, + "grad_norm": 0.9289957548466957, + "learning_rate": 0.003, + "loss": 4.2347, + "step": 2429 + }, + { + "epoch": 0.0243, + "grad_norm": 0.6787755121793266, + "learning_rate": 0.003, + "loss": 4.2464, + "step": 2430 + }, + { + "epoch": 0.02431, + "grad_norm": 0.7835751457530344, + "learning_rate": 0.003, + "loss": 4.2265, + "step": 2431 + }, + { + "epoch": 0.02432, + "grad_norm": 0.7857412874952409, + "learning_rate": 0.003, + "loss": 4.2512, + "step": 2432 + }, + { + "epoch": 0.02433, + "grad_norm": 0.7902925766285801, + "learning_rate": 0.003, + "loss": 4.263, + "step": 2433 + }, + { + "epoch": 0.02434, + "grad_norm": 0.6888559592034768, + "learning_rate": 0.003, + "loss": 4.2265, + "step": 2434 + }, + { + "epoch": 0.02435, + "grad_norm": 0.7021573732138202, + "learning_rate": 0.003, + "loss": 4.2272, + "step": 2435 + }, + { + "epoch": 0.02436, + "grad_norm": 0.7401895262850188, + "learning_rate": 0.003, + "loss": 4.235, + "step": 2436 + }, + { + "epoch": 0.02437, + "grad_norm": 0.6162969571454292, + "learning_rate": 0.003, + "loss": 4.2159, + "step": 2437 + }, + { + "epoch": 0.02438, + "grad_norm": 0.5086196755044915, + "learning_rate": 0.003, + "loss": 4.2268, + "step": 2438 + }, + { + "epoch": 0.02439, + "grad_norm": 0.47970313870881265, + "learning_rate": 0.003, + "loss": 4.1957, + "step": 2439 + }, + { + "epoch": 0.0244, + "grad_norm": 0.4288371919581958, + "learning_rate": 0.003, + "loss": 4.2421, + "step": 2440 + }, + { + "epoch": 0.02441, + "grad_norm": 0.38882426885049926, + "learning_rate": 0.003, + "loss": 4.2052, + "step": 2441 + }, + { + "epoch": 0.02442, + "grad_norm": 0.4241944568677858, + "learning_rate": 0.003, + "loss": 4.2386, + "step": 2442 + }, + { + "epoch": 0.02443, + "grad_norm": 0.4820755060057781, + "learning_rate": 0.003, + "loss": 4.1743, + "step": 2443 + }, + { + "epoch": 0.02444, + "grad_norm": 0.620820199570667, + "learning_rate": 0.003, + "loss": 4.2049, + "step": 2444 + }, + { + "epoch": 0.02445, + "grad_norm": 0.8695985731954037, + "learning_rate": 0.003, + "loss": 4.2178, + "step": 2445 + }, + { + "epoch": 0.02446, + "grad_norm": 0.93656467702838, + "learning_rate": 0.003, + "loss": 4.2301, + "step": 2446 + }, + { + "epoch": 0.02447, + "grad_norm": 0.7193216075423335, + "learning_rate": 0.003, + "loss": 4.2297, + "step": 2447 + }, + { + "epoch": 0.02448, + "grad_norm": 0.77368716084308, + "learning_rate": 0.003, + "loss": 4.2355, + "step": 2448 + }, + { + "epoch": 0.02449, + "grad_norm": 0.93482214454163, + "learning_rate": 0.003, + "loss": 4.2378, + "step": 2449 + }, + { + "epoch": 0.0245, + "grad_norm": 0.7890224806153067, + "learning_rate": 0.003, + "loss": 4.2333, + "step": 2450 + }, + { + "epoch": 0.02451, + "grad_norm": 0.6715615912076842, + "learning_rate": 0.003, + "loss": 4.2085, + "step": 2451 + }, + { + "epoch": 0.02452, + "grad_norm": 0.6401183815147788, + "learning_rate": 0.003, + "loss": 4.2138, + "step": 2452 + }, + { + "epoch": 0.02453, + "grad_norm": 0.5995798001531515, + "learning_rate": 0.003, + "loss": 4.1908, + "step": 2453 + }, + { + "epoch": 0.02454, + "grad_norm": 0.5267826981306184, + "learning_rate": 0.003, + "loss": 4.1933, + "step": 2454 + }, + { + "epoch": 0.02455, + "grad_norm": 0.4821819525696827, + "learning_rate": 0.003, + "loss": 4.2032, + "step": 2455 + }, + { + "epoch": 0.02456, + "grad_norm": 0.5089117196246842, + "learning_rate": 0.003, + "loss": 4.1693, + "step": 2456 + }, + { + "epoch": 0.02457, + "grad_norm": 0.5510451022091881, + "learning_rate": 0.003, + "loss": 4.2212, + "step": 2457 + }, + { + "epoch": 0.02458, + "grad_norm": 0.48930227370319684, + "learning_rate": 0.003, + "loss": 4.1938, + "step": 2458 + }, + { + "epoch": 0.02459, + "grad_norm": 0.4331220174779214, + "learning_rate": 0.003, + "loss": 4.2193, + "step": 2459 + }, + { + "epoch": 0.0246, + "grad_norm": 0.41158812351788215, + "learning_rate": 0.003, + "loss": 4.2215, + "step": 2460 + }, + { + "epoch": 0.02461, + "grad_norm": 0.39108806608008795, + "learning_rate": 0.003, + "loss": 4.1964, + "step": 2461 + }, + { + "epoch": 0.02462, + "grad_norm": 0.3754314162918144, + "learning_rate": 0.003, + "loss": 4.2216, + "step": 2462 + }, + { + "epoch": 0.02463, + "grad_norm": 0.4122698662760153, + "learning_rate": 0.003, + "loss": 4.2201, + "step": 2463 + }, + { + "epoch": 0.02464, + "grad_norm": 0.4350495849540257, + "learning_rate": 0.003, + "loss": 4.1992, + "step": 2464 + }, + { + "epoch": 0.02465, + "grad_norm": 0.49445669847439655, + "learning_rate": 0.003, + "loss": 4.2175, + "step": 2465 + }, + { + "epoch": 0.02466, + "grad_norm": 0.5818409584444987, + "learning_rate": 0.003, + "loss": 4.1996, + "step": 2466 + }, + { + "epoch": 0.02467, + "grad_norm": 0.6235372640193676, + "learning_rate": 0.003, + "loss": 4.1803, + "step": 2467 + }, + { + "epoch": 0.02468, + "grad_norm": 0.5927407776473445, + "learning_rate": 0.003, + "loss": 4.223, + "step": 2468 + }, + { + "epoch": 0.02469, + "grad_norm": 0.6223619407097589, + "learning_rate": 0.003, + "loss": 4.2028, + "step": 2469 + }, + { + "epoch": 0.0247, + "grad_norm": 0.583979309128103, + "learning_rate": 0.003, + "loss": 4.2, + "step": 2470 + }, + { + "epoch": 0.02471, + "grad_norm": 0.5491918199641306, + "learning_rate": 0.003, + "loss": 4.1836, + "step": 2471 + }, + { + "epoch": 0.02472, + "grad_norm": 0.635376629969627, + "learning_rate": 0.003, + "loss": 4.1909, + "step": 2472 + }, + { + "epoch": 0.02473, + "grad_norm": 0.9426912757457121, + "learning_rate": 0.003, + "loss": 4.1862, + "step": 2473 + }, + { + "epoch": 0.02474, + "grad_norm": 1.1494984995182613, + "learning_rate": 0.003, + "loss": 4.2384, + "step": 2474 + }, + { + "epoch": 0.02475, + "grad_norm": 0.6832524660748021, + "learning_rate": 0.003, + "loss": 4.1958, + "step": 2475 + }, + { + "epoch": 0.02476, + "grad_norm": 0.6659288683302277, + "learning_rate": 0.003, + "loss": 4.2064, + "step": 2476 + }, + { + "epoch": 0.02477, + "grad_norm": 0.7078551330691552, + "learning_rate": 0.003, + "loss": 4.229, + "step": 2477 + }, + { + "epoch": 0.02478, + "grad_norm": 0.682879510763541, + "learning_rate": 0.003, + "loss": 4.2039, + "step": 2478 + }, + { + "epoch": 0.02479, + "grad_norm": 0.6453701850218885, + "learning_rate": 0.003, + "loss": 4.1922, + "step": 2479 + }, + { + "epoch": 0.0248, + "grad_norm": 0.6214835257281583, + "learning_rate": 0.003, + "loss": 4.2323, + "step": 2480 + }, + { + "epoch": 0.02481, + "grad_norm": 0.5814705370549468, + "learning_rate": 0.003, + "loss": 4.2104, + "step": 2481 + }, + { + "epoch": 0.02482, + "grad_norm": 0.5717194029886628, + "learning_rate": 0.003, + "loss": 4.2206, + "step": 2482 + }, + { + "epoch": 0.02483, + "grad_norm": 0.5507096505833119, + "learning_rate": 0.003, + "loss": 4.1741, + "step": 2483 + }, + { + "epoch": 0.02484, + "grad_norm": 0.48818940634137703, + "learning_rate": 0.003, + "loss": 4.2, + "step": 2484 + }, + { + "epoch": 0.02485, + "grad_norm": 0.5078768442645458, + "learning_rate": 0.003, + "loss": 4.2218, + "step": 2485 + }, + { + "epoch": 0.02486, + "grad_norm": 0.6158842388358957, + "learning_rate": 0.003, + "loss": 4.2483, + "step": 2486 + }, + { + "epoch": 0.02487, + "grad_norm": 0.7113530958465855, + "learning_rate": 0.003, + "loss": 4.2057, + "step": 2487 + }, + { + "epoch": 0.02488, + "grad_norm": 0.8543927391668017, + "learning_rate": 0.003, + "loss": 4.1886, + "step": 2488 + }, + { + "epoch": 0.02489, + "grad_norm": 0.9446890974964458, + "learning_rate": 0.003, + "loss": 4.1911, + "step": 2489 + }, + { + "epoch": 0.0249, + "grad_norm": 0.7778085412828454, + "learning_rate": 0.003, + "loss": 4.1871, + "step": 2490 + }, + { + "epoch": 0.02491, + "grad_norm": 0.6481694251808287, + "learning_rate": 0.003, + "loss": 4.2111, + "step": 2491 + }, + { + "epoch": 0.02492, + "grad_norm": 0.6221254984926949, + "learning_rate": 0.003, + "loss": 4.2219, + "step": 2492 + }, + { + "epoch": 0.02493, + "grad_norm": 0.6659856187784612, + "learning_rate": 0.003, + "loss": 4.1927, + "step": 2493 + }, + { + "epoch": 0.02494, + "grad_norm": 0.6928875164931092, + "learning_rate": 0.003, + "loss": 4.206, + "step": 2494 + }, + { + "epoch": 0.02495, + "grad_norm": 0.6666259038056425, + "learning_rate": 0.003, + "loss": 4.2171, + "step": 2495 + }, + { + "epoch": 0.02496, + "grad_norm": 0.6388189581951018, + "learning_rate": 0.003, + "loss": 4.2077, + "step": 2496 + }, + { + "epoch": 0.02497, + "grad_norm": 0.6011544202060304, + "learning_rate": 0.003, + "loss": 4.2109, + "step": 2497 + }, + { + "epoch": 0.02498, + "grad_norm": 0.6445306947945534, + "learning_rate": 0.003, + "loss": 4.1989, + "step": 2498 + }, + { + "epoch": 0.02499, + "grad_norm": 0.6336412786289282, + "learning_rate": 0.003, + "loss": 4.2101, + "step": 2499 + }, + { + "epoch": 0.025, + "grad_norm": 0.7751683320425504, + "learning_rate": 0.003, + "loss": 4.1983, + "step": 2500 + }, + { + "epoch": 0.02501, + "grad_norm": 0.8226642072050464, + "learning_rate": 0.003, + "loss": 4.2348, + "step": 2501 + }, + { + "epoch": 0.02502, + "grad_norm": 0.7525954653781657, + "learning_rate": 0.003, + "loss": 4.2151, + "step": 2502 + }, + { + "epoch": 0.02503, + "grad_norm": 0.6443708527662554, + "learning_rate": 0.003, + "loss": 4.22, + "step": 2503 + }, + { + "epoch": 0.02504, + "grad_norm": 0.7431879561442126, + "learning_rate": 0.003, + "loss": 4.2035, + "step": 2504 + }, + { + "epoch": 0.02505, + "grad_norm": 0.8444858068379042, + "learning_rate": 0.003, + "loss": 4.206, + "step": 2505 + }, + { + "epoch": 0.02506, + "grad_norm": 0.8357688056745337, + "learning_rate": 0.003, + "loss": 4.2242, + "step": 2506 + }, + { + "epoch": 0.02507, + "grad_norm": 0.8015204150666452, + "learning_rate": 0.003, + "loss": 4.21, + "step": 2507 + }, + { + "epoch": 0.02508, + "grad_norm": 0.7988221880335916, + "learning_rate": 0.003, + "loss": 4.2096, + "step": 2508 + }, + { + "epoch": 0.02509, + "grad_norm": 0.8648254923532768, + "learning_rate": 0.003, + "loss": 4.2197, + "step": 2509 + }, + { + "epoch": 0.0251, + "grad_norm": 0.9052649096369385, + "learning_rate": 0.003, + "loss": 4.2352, + "step": 2510 + }, + { + "epoch": 0.02511, + "grad_norm": 0.8287055915677601, + "learning_rate": 0.003, + "loss": 4.243, + "step": 2511 + }, + { + "epoch": 0.02512, + "grad_norm": 0.6322731338734815, + "learning_rate": 0.003, + "loss": 4.2191, + "step": 2512 + }, + { + "epoch": 0.02513, + "grad_norm": 0.6027563249296022, + "learning_rate": 0.003, + "loss": 4.2132, + "step": 2513 + }, + { + "epoch": 0.02514, + "grad_norm": 0.5639092637703635, + "learning_rate": 0.003, + "loss": 4.1993, + "step": 2514 + }, + { + "epoch": 0.02515, + "grad_norm": 0.5109276454863091, + "learning_rate": 0.003, + "loss": 4.1846, + "step": 2515 + }, + { + "epoch": 0.02516, + "grad_norm": 0.49613462202561587, + "learning_rate": 0.003, + "loss": 4.2057, + "step": 2516 + }, + { + "epoch": 0.02517, + "grad_norm": 0.4638933308319014, + "learning_rate": 0.003, + "loss": 4.2043, + "step": 2517 + }, + { + "epoch": 0.02518, + "grad_norm": 0.47895707517955144, + "learning_rate": 0.003, + "loss": 4.2156, + "step": 2518 + }, + { + "epoch": 0.02519, + "grad_norm": 0.540563379395588, + "learning_rate": 0.003, + "loss": 4.191, + "step": 2519 + }, + { + "epoch": 0.0252, + "grad_norm": 0.593588962667406, + "learning_rate": 0.003, + "loss": 4.2039, + "step": 2520 + }, + { + "epoch": 0.02521, + "grad_norm": 0.5859708147834043, + "learning_rate": 0.003, + "loss": 4.2007, + "step": 2521 + }, + { + "epoch": 0.02522, + "grad_norm": 0.6750984334319393, + "learning_rate": 0.003, + "loss": 4.1967, + "step": 2522 + }, + { + "epoch": 0.02523, + "grad_norm": 0.8294919931793535, + "learning_rate": 0.003, + "loss": 4.2041, + "step": 2523 + }, + { + "epoch": 0.02524, + "grad_norm": 0.8897442405025344, + "learning_rate": 0.003, + "loss": 4.2231, + "step": 2524 + }, + { + "epoch": 0.02525, + "grad_norm": 0.8626054901448138, + "learning_rate": 0.003, + "loss": 4.2361, + "step": 2525 + }, + { + "epoch": 0.02526, + "grad_norm": 0.6609622839443053, + "learning_rate": 0.003, + "loss": 4.1771, + "step": 2526 + }, + { + "epoch": 0.02527, + "grad_norm": 0.5788832246003424, + "learning_rate": 0.003, + "loss": 4.2082, + "step": 2527 + }, + { + "epoch": 0.02528, + "grad_norm": 0.6673373775723666, + "learning_rate": 0.003, + "loss": 4.2334, + "step": 2528 + }, + { + "epoch": 0.02529, + "grad_norm": 0.6949939363717212, + "learning_rate": 0.003, + "loss": 4.2031, + "step": 2529 + }, + { + "epoch": 0.0253, + "grad_norm": 0.6926157434556843, + "learning_rate": 0.003, + "loss": 4.2244, + "step": 2530 + }, + { + "epoch": 0.02531, + "grad_norm": 0.6799840691277568, + "learning_rate": 0.003, + "loss": 4.2264, + "step": 2531 + }, + { + "epoch": 0.02532, + "grad_norm": 0.8293024556709889, + "learning_rate": 0.003, + "loss": 4.1997, + "step": 2532 + }, + { + "epoch": 0.02533, + "grad_norm": 0.9626324272072302, + "learning_rate": 0.003, + "loss": 4.2444, + "step": 2533 + }, + { + "epoch": 0.02534, + "grad_norm": 0.9642180813694794, + "learning_rate": 0.003, + "loss": 4.2316, + "step": 2534 + }, + { + "epoch": 0.02535, + "grad_norm": 0.8934366837294853, + "learning_rate": 0.003, + "loss": 4.2235, + "step": 2535 + }, + { + "epoch": 0.02536, + "grad_norm": 0.8077546891906852, + "learning_rate": 0.003, + "loss": 4.2184, + "step": 2536 + }, + { + "epoch": 0.02537, + "grad_norm": 0.6532019172502265, + "learning_rate": 0.003, + "loss": 4.2093, + "step": 2537 + }, + { + "epoch": 0.02538, + "grad_norm": 0.6905511076952489, + "learning_rate": 0.003, + "loss": 4.184, + "step": 2538 + }, + { + "epoch": 0.02539, + "grad_norm": 0.7499705806639466, + "learning_rate": 0.003, + "loss": 4.2378, + "step": 2539 + }, + { + "epoch": 0.0254, + "grad_norm": 0.6787160415395472, + "learning_rate": 0.003, + "loss": 4.2251, + "step": 2540 + }, + { + "epoch": 0.02541, + "grad_norm": 0.6164160513405081, + "learning_rate": 0.003, + "loss": 4.2157, + "step": 2541 + }, + { + "epoch": 0.02542, + "grad_norm": 0.6175874501025589, + "learning_rate": 0.003, + "loss": 4.1908, + "step": 2542 + }, + { + "epoch": 0.02543, + "grad_norm": 0.5640476735952517, + "learning_rate": 0.003, + "loss": 4.2185, + "step": 2543 + }, + { + "epoch": 0.02544, + "grad_norm": 0.6088525590071204, + "learning_rate": 0.003, + "loss": 4.2062, + "step": 2544 + }, + { + "epoch": 0.02545, + "grad_norm": 0.5774041309556538, + "learning_rate": 0.003, + "loss": 4.1853, + "step": 2545 + }, + { + "epoch": 0.02546, + "grad_norm": 0.5080999994690095, + "learning_rate": 0.003, + "loss": 4.194, + "step": 2546 + }, + { + "epoch": 0.02547, + "grad_norm": 0.491841642905056, + "learning_rate": 0.003, + "loss": 4.2303, + "step": 2547 + }, + { + "epoch": 0.02548, + "grad_norm": 0.5507165108887961, + "learning_rate": 0.003, + "loss": 4.2283, + "step": 2548 + }, + { + "epoch": 0.02549, + "grad_norm": 0.6923243100094933, + "learning_rate": 0.003, + "loss": 4.2227, + "step": 2549 + }, + { + "epoch": 0.0255, + "grad_norm": 0.7817347310990386, + "learning_rate": 0.003, + "loss": 4.1886, + "step": 2550 + }, + { + "epoch": 0.02551, + "grad_norm": 0.7447240542500363, + "learning_rate": 0.003, + "loss": 4.2189, + "step": 2551 + }, + { + "epoch": 0.02552, + "grad_norm": 0.6307913070627245, + "learning_rate": 0.003, + "loss": 4.1874, + "step": 2552 + }, + { + "epoch": 0.02553, + "grad_norm": 0.6438248127745729, + "learning_rate": 0.003, + "loss": 4.2512, + "step": 2553 + }, + { + "epoch": 0.02554, + "grad_norm": 0.5917787379710305, + "learning_rate": 0.003, + "loss": 4.1983, + "step": 2554 + }, + { + "epoch": 0.02555, + "grad_norm": 0.6080534723714437, + "learning_rate": 0.003, + "loss": 4.1896, + "step": 2555 + }, + { + "epoch": 0.02556, + "grad_norm": 0.6279427379491616, + "learning_rate": 0.003, + "loss": 4.2002, + "step": 2556 + }, + { + "epoch": 0.02557, + "grad_norm": 0.5766106602168388, + "learning_rate": 0.003, + "loss": 4.2029, + "step": 2557 + }, + { + "epoch": 0.02558, + "grad_norm": 0.5614919439663087, + "learning_rate": 0.003, + "loss": 4.2176, + "step": 2558 + }, + { + "epoch": 0.02559, + "grad_norm": 0.5870187050191585, + "learning_rate": 0.003, + "loss": 4.2144, + "step": 2559 + }, + { + "epoch": 0.0256, + "grad_norm": 0.5866194992858084, + "learning_rate": 0.003, + "loss": 4.1883, + "step": 2560 + }, + { + "epoch": 0.02561, + "grad_norm": 0.56223387782931, + "learning_rate": 0.003, + "loss": 4.1893, + "step": 2561 + }, + { + "epoch": 0.02562, + "grad_norm": 0.4873971454432025, + "learning_rate": 0.003, + "loss": 4.203, + "step": 2562 + }, + { + "epoch": 0.02563, + "grad_norm": 0.6113475215270017, + "learning_rate": 0.003, + "loss": 4.1898, + "step": 2563 + }, + { + "epoch": 0.02564, + "grad_norm": 0.7598159388308933, + "learning_rate": 0.003, + "loss": 4.2128, + "step": 2564 + }, + { + "epoch": 0.02565, + "grad_norm": 0.9584592936312263, + "learning_rate": 0.003, + "loss": 4.2023, + "step": 2565 + }, + { + "epoch": 0.02566, + "grad_norm": 1.059905791929432, + "learning_rate": 0.003, + "loss": 4.231, + "step": 2566 + }, + { + "epoch": 0.02567, + "grad_norm": 0.6761577058111206, + "learning_rate": 0.003, + "loss": 4.219, + "step": 2567 + }, + { + "epoch": 0.02568, + "grad_norm": 0.582893785111441, + "learning_rate": 0.003, + "loss": 4.2104, + "step": 2568 + }, + { + "epoch": 0.02569, + "grad_norm": 0.7274372672775176, + "learning_rate": 0.003, + "loss": 4.1941, + "step": 2569 + }, + { + "epoch": 0.0257, + "grad_norm": 0.6382797975462534, + "learning_rate": 0.003, + "loss": 4.1988, + "step": 2570 + }, + { + "epoch": 0.02571, + "grad_norm": 0.5842618410643182, + "learning_rate": 0.003, + "loss": 4.2168, + "step": 2571 + }, + { + "epoch": 0.02572, + "grad_norm": 0.5733806386596177, + "learning_rate": 0.003, + "loss": 4.1929, + "step": 2572 + }, + { + "epoch": 0.02573, + "grad_norm": 0.6098335904085562, + "learning_rate": 0.003, + "loss": 4.2012, + "step": 2573 + }, + { + "epoch": 0.02574, + "grad_norm": 0.7180717115741516, + "learning_rate": 0.003, + "loss": 4.233, + "step": 2574 + }, + { + "epoch": 0.02575, + "grad_norm": 0.7732503452131877, + "learning_rate": 0.003, + "loss": 4.1693, + "step": 2575 + }, + { + "epoch": 0.02576, + "grad_norm": 0.8039085737347494, + "learning_rate": 0.003, + "loss": 4.1792, + "step": 2576 + }, + { + "epoch": 0.02577, + "grad_norm": 0.7708873473006739, + "learning_rate": 0.003, + "loss": 4.183, + "step": 2577 + }, + { + "epoch": 0.02578, + "grad_norm": 0.6416087124695665, + "learning_rate": 0.003, + "loss": 4.2007, + "step": 2578 + }, + { + "epoch": 0.02579, + "grad_norm": 0.6401520636312507, + "learning_rate": 0.003, + "loss": 4.1503, + "step": 2579 + }, + { + "epoch": 0.0258, + "grad_norm": 0.609541740323575, + "learning_rate": 0.003, + "loss": 4.2065, + "step": 2580 + }, + { + "epoch": 0.02581, + "grad_norm": 0.6036763081432622, + "learning_rate": 0.003, + "loss": 4.2268, + "step": 2581 + }, + { + "epoch": 0.02582, + "grad_norm": 0.5959601922482883, + "learning_rate": 0.003, + "loss": 4.2238, + "step": 2582 + }, + { + "epoch": 0.02583, + "grad_norm": 0.6183189840573183, + "learning_rate": 0.003, + "loss": 4.1962, + "step": 2583 + }, + { + "epoch": 0.02584, + "grad_norm": 0.7087135376074408, + "learning_rate": 0.003, + "loss": 4.2138, + "step": 2584 + }, + { + "epoch": 0.02585, + "grad_norm": 0.6812145006805832, + "learning_rate": 0.003, + "loss": 4.1952, + "step": 2585 + }, + { + "epoch": 0.02586, + "grad_norm": 0.7034361159454041, + "learning_rate": 0.003, + "loss": 4.2173, + "step": 2586 + }, + { + "epoch": 0.02587, + "grad_norm": 0.7458480631965212, + "learning_rate": 0.003, + "loss": 4.2079, + "step": 2587 + }, + { + "epoch": 0.02588, + "grad_norm": 0.6631960688574444, + "learning_rate": 0.003, + "loss": 4.2037, + "step": 2588 + }, + { + "epoch": 0.02589, + "grad_norm": 0.6894974904796883, + "learning_rate": 0.003, + "loss": 4.2122, + "step": 2589 + }, + { + "epoch": 0.0259, + "grad_norm": 0.6129883156712106, + "learning_rate": 0.003, + "loss": 4.2205, + "step": 2590 + }, + { + "epoch": 0.02591, + "grad_norm": 0.52233908649411, + "learning_rate": 0.003, + "loss": 4.1499, + "step": 2591 + }, + { + "epoch": 0.02592, + "grad_norm": 0.5621629848199292, + "learning_rate": 0.003, + "loss": 4.1958, + "step": 2592 + }, + { + "epoch": 0.02593, + "grad_norm": 0.579018019134519, + "learning_rate": 0.003, + "loss": 4.1981, + "step": 2593 + }, + { + "epoch": 0.02594, + "grad_norm": 0.7055147915095833, + "learning_rate": 0.003, + "loss": 4.2202, + "step": 2594 + }, + { + "epoch": 0.02595, + "grad_norm": 0.9044844479304169, + "learning_rate": 0.003, + "loss": 4.182, + "step": 2595 + }, + { + "epoch": 0.02596, + "grad_norm": 0.9468656207805218, + "learning_rate": 0.003, + "loss": 4.2266, + "step": 2596 + }, + { + "epoch": 0.02597, + "grad_norm": 0.6963050525881186, + "learning_rate": 0.003, + "loss": 4.1873, + "step": 2597 + }, + { + "epoch": 0.02598, + "grad_norm": 0.6825938651999739, + "learning_rate": 0.003, + "loss": 4.2289, + "step": 2598 + }, + { + "epoch": 0.02599, + "grad_norm": 0.7880468862794131, + "learning_rate": 0.003, + "loss": 4.225, + "step": 2599 + }, + { + "epoch": 0.026, + "grad_norm": 0.8906440360951334, + "learning_rate": 0.003, + "loss": 4.2307, + "step": 2600 + }, + { + "epoch": 0.02601, + "grad_norm": 0.8491652376595475, + "learning_rate": 0.003, + "loss": 4.2355, + "step": 2601 + }, + { + "epoch": 0.02602, + "grad_norm": 0.7504592913224991, + "learning_rate": 0.003, + "loss": 4.2094, + "step": 2602 + }, + { + "epoch": 0.02603, + "grad_norm": 0.6893393420157942, + "learning_rate": 0.003, + "loss": 4.2031, + "step": 2603 + }, + { + "epoch": 0.02604, + "grad_norm": 0.8097986318712986, + "learning_rate": 0.003, + "loss": 4.1981, + "step": 2604 + }, + { + "epoch": 0.02605, + "grad_norm": 0.9104130090907385, + "learning_rate": 0.003, + "loss": 4.2234, + "step": 2605 + }, + { + "epoch": 0.02606, + "grad_norm": 1.0310289052047719, + "learning_rate": 0.003, + "loss": 4.2436, + "step": 2606 + }, + { + "epoch": 0.02607, + "grad_norm": 0.9582413171497328, + "learning_rate": 0.003, + "loss": 4.2397, + "step": 2607 + }, + { + "epoch": 0.02608, + "grad_norm": 0.8048575866472404, + "learning_rate": 0.003, + "loss": 4.2123, + "step": 2608 + }, + { + "epoch": 0.02609, + "grad_norm": 0.8595866026929856, + "learning_rate": 0.003, + "loss": 4.2586, + "step": 2609 + }, + { + "epoch": 0.0261, + "grad_norm": 0.823900371529962, + "learning_rate": 0.003, + "loss": 4.2068, + "step": 2610 + }, + { + "epoch": 0.02611, + "grad_norm": 0.7141217984189516, + "learning_rate": 0.003, + "loss": 4.1945, + "step": 2611 + }, + { + "epoch": 0.02612, + "grad_norm": 0.716568582474741, + "learning_rate": 0.003, + "loss": 4.2324, + "step": 2612 + }, + { + "epoch": 0.02613, + "grad_norm": 0.6977866203993734, + "learning_rate": 0.003, + "loss": 4.2227, + "step": 2613 + }, + { + "epoch": 0.02614, + "grad_norm": 0.6120459222412155, + "learning_rate": 0.003, + "loss": 4.1988, + "step": 2614 + }, + { + "epoch": 0.02615, + "grad_norm": 0.48978414596737385, + "learning_rate": 0.003, + "loss": 4.185, + "step": 2615 + }, + { + "epoch": 0.02616, + "grad_norm": 0.4111484941092938, + "learning_rate": 0.003, + "loss": 4.193, + "step": 2616 + }, + { + "epoch": 0.02617, + "grad_norm": 0.4314798871804994, + "learning_rate": 0.003, + "loss": 4.1808, + "step": 2617 + }, + { + "epoch": 0.02618, + "grad_norm": 0.45257700858495165, + "learning_rate": 0.003, + "loss": 4.2091, + "step": 2618 + }, + { + "epoch": 0.02619, + "grad_norm": 0.43820325303953606, + "learning_rate": 0.003, + "loss": 4.2202, + "step": 2619 + }, + { + "epoch": 0.0262, + "grad_norm": 0.48135056397229203, + "learning_rate": 0.003, + "loss": 4.1827, + "step": 2620 + }, + { + "epoch": 0.02621, + "grad_norm": 0.5541701604207973, + "learning_rate": 0.003, + "loss": 4.2081, + "step": 2621 + }, + { + "epoch": 0.02622, + "grad_norm": 0.6217167828478687, + "learning_rate": 0.003, + "loss": 4.2095, + "step": 2622 + }, + { + "epoch": 0.02623, + "grad_norm": 0.5922780087971841, + "learning_rate": 0.003, + "loss": 4.2099, + "step": 2623 + }, + { + "epoch": 0.02624, + "grad_norm": 0.6090506282888395, + "learning_rate": 0.003, + "loss": 4.246, + "step": 2624 + }, + { + "epoch": 0.02625, + "grad_norm": 0.5994929489144406, + "learning_rate": 0.003, + "loss": 4.2398, + "step": 2625 + }, + { + "epoch": 0.02626, + "grad_norm": 0.6828660510269309, + "learning_rate": 0.003, + "loss": 4.1745, + "step": 2626 + }, + { + "epoch": 0.02627, + "grad_norm": 0.7689467228473569, + "learning_rate": 0.003, + "loss": 4.1978, + "step": 2627 + }, + { + "epoch": 0.02628, + "grad_norm": 0.772945921723307, + "learning_rate": 0.003, + "loss": 4.2279, + "step": 2628 + }, + { + "epoch": 0.02629, + "grad_norm": 0.7998878360923921, + "learning_rate": 0.003, + "loss": 4.1718, + "step": 2629 + }, + { + "epoch": 0.0263, + "grad_norm": 0.8071005154576086, + "learning_rate": 0.003, + "loss": 4.2029, + "step": 2630 + }, + { + "epoch": 0.02631, + "grad_norm": 0.7063928050043093, + "learning_rate": 0.003, + "loss": 4.2214, + "step": 2631 + }, + { + "epoch": 0.02632, + "grad_norm": 0.5358237009716528, + "learning_rate": 0.003, + "loss": 4.1777, + "step": 2632 + }, + { + "epoch": 0.02633, + "grad_norm": 0.5777990078306479, + "learning_rate": 0.003, + "loss": 4.1915, + "step": 2633 + }, + { + "epoch": 0.02634, + "grad_norm": 0.5758302081275357, + "learning_rate": 0.003, + "loss": 4.2102, + "step": 2634 + }, + { + "epoch": 0.02635, + "grad_norm": 0.5582382698428011, + "learning_rate": 0.003, + "loss": 4.2059, + "step": 2635 + }, + { + "epoch": 0.02636, + "grad_norm": 0.5270921433631331, + "learning_rate": 0.003, + "loss": 4.1886, + "step": 2636 + }, + { + "epoch": 0.02637, + "grad_norm": 0.4952392838330578, + "learning_rate": 0.003, + "loss": 4.2209, + "step": 2637 + }, + { + "epoch": 0.02638, + "grad_norm": 0.42063663579214783, + "learning_rate": 0.003, + "loss": 4.1814, + "step": 2638 + }, + { + "epoch": 0.02639, + "grad_norm": 0.38387035795548136, + "learning_rate": 0.003, + "loss": 4.1685, + "step": 2639 + }, + { + "epoch": 0.0264, + "grad_norm": 0.39356773800785, + "learning_rate": 0.003, + "loss": 4.1664, + "step": 2640 + }, + { + "epoch": 0.02641, + "grad_norm": 0.4027429798510006, + "learning_rate": 0.003, + "loss": 4.1789, + "step": 2641 + }, + { + "epoch": 0.02642, + "grad_norm": 0.4954271688101395, + "learning_rate": 0.003, + "loss": 4.1892, + "step": 2642 + }, + { + "epoch": 0.02643, + "grad_norm": 0.544589775918662, + "learning_rate": 0.003, + "loss": 4.176, + "step": 2643 + }, + { + "epoch": 0.02644, + "grad_norm": 0.5577274451115164, + "learning_rate": 0.003, + "loss": 4.1927, + "step": 2644 + }, + { + "epoch": 0.02645, + "grad_norm": 0.7183426191932024, + "learning_rate": 0.003, + "loss": 4.2101, + "step": 2645 + }, + { + "epoch": 0.02646, + "grad_norm": 0.9615412387915055, + "learning_rate": 0.003, + "loss": 4.1833, + "step": 2646 + }, + { + "epoch": 0.02647, + "grad_norm": 1.0740028901393692, + "learning_rate": 0.003, + "loss": 4.2196, + "step": 2647 + }, + { + "epoch": 0.02648, + "grad_norm": 0.7345014390620105, + "learning_rate": 0.003, + "loss": 4.2256, + "step": 2648 + }, + { + "epoch": 0.02649, + "grad_norm": 0.6618619800661987, + "learning_rate": 0.003, + "loss": 4.2112, + "step": 2649 + }, + { + "epoch": 0.0265, + "grad_norm": 0.7347893947046258, + "learning_rate": 0.003, + "loss": 4.1759, + "step": 2650 + }, + { + "epoch": 0.02651, + "grad_norm": 0.6484062784740668, + "learning_rate": 0.003, + "loss": 4.1885, + "step": 2651 + }, + { + "epoch": 0.02652, + "grad_norm": 0.7667553931843348, + "learning_rate": 0.003, + "loss": 4.2255, + "step": 2652 + }, + { + "epoch": 0.02653, + "grad_norm": 0.7289832175113301, + "learning_rate": 0.003, + "loss": 4.1797, + "step": 2653 + }, + { + "epoch": 0.02654, + "grad_norm": 0.7103736472197925, + "learning_rate": 0.003, + "loss": 4.1924, + "step": 2654 + }, + { + "epoch": 0.02655, + "grad_norm": 0.7580675808718691, + "learning_rate": 0.003, + "loss": 4.2053, + "step": 2655 + }, + { + "epoch": 0.02656, + "grad_norm": 0.7969812553544864, + "learning_rate": 0.003, + "loss": 4.1802, + "step": 2656 + }, + { + "epoch": 0.02657, + "grad_norm": 0.6347600401053554, + "learning_rate": 0.003, + "loss": 4.2057, + "step": 2657 + }, + { + "epoch": 0.02658, + "grad_norm": 0.5896722976882536, + "learning_rate": 0.003, + "loss": 4.2116, + "step": 2658 + }, + { + "epoch": 0.02659, + "grad_norm": 0.651082411912306, + "learning_rate": 0.003, + "loss": 4.2054, + "step": 2659 + }, + { + "epoch": 0.0266, + "grad_norm": 0.6950508733673917, + "learning_rate": 0.003, + "loss": 4.1947, + "step": 2660 + }, + { + "epoch": 0.02661, + "grad_norm": 0.7309821342606786, + "learning_rate": 0.003, + "loss": 4.2168, + "step": 2661 + }, + { + "epoch": 0.02662, + "grad_norm": 0.8443517188841829, + "learning_rate": 0.003, + "loss": 4.2102, + "step": 2662 + }, + { + "epoch": 0.02663, + "grad_norm": 0.9107871566375996, + "learning_rate": 0.003, + "loss": 4.1952, + "step": 2663 + }, + { + "epoch": 0.02664, + "grad_norm": 0.8948636609916674, + "learning_rate": 0.003, + "loss": 4.2015, + "step": 2664 + }, + { + "epoch": 0.02665, + "grad_norm": 0.7895030310092762, + "learning_rate": 0.003, + "loss": 4.2447, + "step": 2665 + }, + { + "epoch": 0.02666, + "grad_norm": 0.7544169624146947, + "learning_rate": 0.003, + "loss": 4.2351, + "step": 2666 + }, + { + "epoch": 0.02667, + "grad_norm": 0.7917512909838754, + "learning_rate": 0.003, + "loss": 4.2156, + "step": 2667 + }, + { + "epoch": 0.02668, + "grad_norm": 0.8660435271921453, + "learning_rate": 0.003, + "loss": 4.1906, + "step": 2668 + }, + { + "epoch": 0.02669, + "grad_norm": 0.8211806952379437, + "learning_rate": 0.003, + "loss": 4.2092, + "step": 2669 + }, + { + "epoch": 0.0267, + "grad_norm": 0.7376659322382964, + "learning_rate": 0.003, + "loss": 4.2095, + "step": 2670 + }, + { + "epoch": 0.02671, + "grad_norm": 0.7319080249702333, + "learning_rate": 0.003, + "loss": 4.2252, + "step": 2671 + }, + { + "epoch": 0.02672, + "grad_norm": 0.674274482359015, + "learning_rate": 0.003, + "loss": 4.2161, + "step": 2672 + }, + { + "epoch": 0.02673, + "grad_norm": 0.7262456950172295, + "learning_rate": 0.003, + "loss": 4.2152, + "step": 2673 + }, + { + "epoch": 0.02674, + "grad_norm": 0.7478436146050179, + "learning_rate": 0.003, + "loss": 4.1824, + "step": 2674 + }, + { + "epoch": 0.02675, + "grad_norm": 0.6974424835283173, + "learning_rate": 0.003, + "loss": 4.1794, + "step": 2675 + }, + { + "epoch": 0.02676, + "grad_norm": 0.7773148175185325, + "learning_rate": 0.003, + "loss": 4.2023, + "step": 2676 + }, + { + "epoch": 0.02677, + "grad_norm": 0.9696205190703804, + "learning_rate": 0.003, + "loss": 4.2371, + "step": 2677 + }, + { + "epoch": 0.02678, + "grad_norm": 0.9527960429936483, + "learning_rate": 0.003, + "loss": 4.237, + "step": 2678 + }, + { + "epoch": 0.02679, + "grad_norm": 0.686114258952289, + "learning_rate": 0.003, + "loss": 4.1997, + "step": 2679 + }, + { + "epoch": 0.0268, + "grad_norm": 0.587978184638772, + "learning_rate": 0.003, + "loss": 4.2094, + "step": 2680 + }, + { + "epoch": 0.02681, + "grad_norm": 0.5919245996634437, + "learning_rate": 0.003, + "loss": 4.2288, + "step": 2681 + }, + { + "epoch": 0.02682, + "grad_norm": 0.5747684944993391, + "learning_rate": 0.003, + "loss": 4.2039, + "step": 2682 + }, + { + "epoch": 0.02683, + "grad_norm": 0.49692411048438434, + "learning_rate": 0.003, + "loss": 4.2231, + "step": 2683 + }, + { + "epoch": 0.02684, + "grad_norm": 0.42371991981886814, + "learning_rate": 0.003, + "loss": 4.1939, + "step": 2684 + }, + { + "epoch": 0.02685, + "grad_norm": 0.43270831648127167, + "learning_rate": 0.003, + "loss": 4.1987, + "step": 2685 + }, + { + "epoch": 0.02686, + "grad_norm": 0.4373309691769235, + "learning_rate": 0.003, + "loss": 4.1811, + "step": 2686 + }, + { + "epoch": 0.02687, + "grad_norm": 0.4372836741597173, + "learning_rate": 0.003, + "loss": 4.189, + "step": 2687 + }, + { + "epoch": 0.02688, + "grad_norm": 0.49637578837338964, + "learning_rate": 0.003, + "loss": 4.1657, + "step": 2688 + }, + { + "epoch": 0.02689, + "grad_norm": 0.42425016692512946, + "learning_rate": 0.003, + "loss": 4.197, + "step": 2689 + }, + { + "epoch": 0.0269, + "grad_norm": 0.3687359570047172, + "learning_rate": 0.003, + "loss": 4.199, + "step": 2690 + }, + { + "epoch": 0.02691, + "grad_norm": 0.422533215672721, + "learning_rate": 0.003, + "loss": 4.1917, + "step": 2691 + }, + { + "epoch": 0.02692, + "grad_norm": 0.36876790177094304, + "learning_rate": 0.003, + "loss": 4.1944, + "step": 2692 + }, + { + "epoch": 0.02693, + "grad_norm": 0.33344364158726264, + "learning_rate": 0.003, + "loss": 4.2054, + "step": 2693 + }, + { + "epoch": 0.02694, + "grad_norm": 0.35193571774116406, + "learning_rate": 0.003, + "loss": 4.1673, + "step": 2694 + }, + { + "epoch": 0.02695, + "grad_norm": 0.37618933218650874, + "learning_rate": 0.003, + "loss": 4.1828, + "step": 2695 + }, + { + "epoch": 0.02696, + "grad_norm": 0.41130959986438415, + "learning_rate": 0.003, + "loss": 4.185, + "step": 2696 + }, + { + "epoch": 0.02697, + "grad_norm": 0.5181896124624418, + "learning_rate": 0.003, + "loss": 4.198, + "step": 2697 + }, + { + "epoch": 0.02698, + "grad_norm": 0.8288766752985389, + "learning_rate": 0.003, + "loss": 4.1898, + "step": 2698 + }, + { + "epoch": 0.02699, + "grad_norm": 1.1773125298816605, + "learning_rate": 0.003, + "loss": 4.23, + "step": 2699 + }, + { + "epoch": 0.027, + "grad_norm": 0.7553245674542797, + "learning_rate": 0.003, + "loss": 4.1855, + "step": 2700 + }, + { + "epoch": 0.02701, + "grad_norm": 0.8921172592412265, + "learning_rate": 0.003, + "loss": 4.1838, + "step": 2701 + }, + { + "epoch": 0.02702, + "grad_norm": 1.0764387127470998, + "learning_rate": 0.003, + "loss": 4.2606, + "step": 2702 + }, + { + "epoch": 0.02703, + "grad_norm": 0.9044993098249169, + "learning_rate": 0.003, + "loss": 4.2449, + "step": 2703 + }, + { + "epoch": 0.02704, + "grad_norm": 0.8806690266182866, + "learning_rate": 0.003, + "loss": 4.1921, + "step": 2704 + }, + { + "epoch": 0.02705, + "grad_norm": 0.8651796636346343, + "learning_rate": 0.003, + "loss": 4.2521, + "step": 2705 + }, + { + "epoch": 0.02706, + "grad_norm": 0.8193327778471594, + "learning_rate": 0.003, + "loss": 4.2075, + "step": 2706 + }, + { + "epoch": 0.02707, + "grad_norm": 0.8383738553676314, + "learning_rate": 0.003, + "loss": 4.2115, + "step": 2707 + }, + { + "epoch": 0.02708, + "grad_norm": 0.7927971133508214, + "learning_rate": 0.003, + "loss": 4.2392, + "step": 2708 + }, + { + "epoch": 0.02709, + "grad_norm": 0.702011577126357, + "learning_rate": 0.003, + "loss": 4.2044, + "step": 2709 + }, + { + "epoch": 0.0271, + "grad_norm": 0.7610112656844941, + "learning_rate": 0.003, + "loss": 4.2165, + "step": 2710 + }, + { + "epoch": 0.02711, + "grad_norm": 0.7565309559289577, + "learning_rate": 0.003, + "loss": 4.1983, + "step": 2711 + }, + { + "epoch": 0.02712, + "grad_norm": 0.7177786599726909, + "learning_rate": 0.003, + "loss": 4.2091, + "step": 2712 + }, + { + "epoch": 0.02713, + "grad_norm": 0.6187976568863948, + "learning_rate": 0.003, + "loss": 4.1995, + "step": 2713 + }, + { + "epoch": 0.02714, + "grad_norm": 0.5823945630189209, + "learning_rate": 0.003, + "loss": 4.1941, + "step": 2714 + }, + { + "epoch": 0.02715, + "grad_norm": 0.6590346464153963, + "learning_rate": 0.003, + "loss": 4.1907, + "step": 2715 + }, + { + "epoch": 0.02716, + "grad_norm": 0.8913115921228741, + "learning_rate": 0.003, + "loss": 4.2092, + "step": 2716 + }, + { + "epoch": 0.02717, + "grad_norm": 0.9991228875759847, + "learning_rate": 0.003, + "loss": 4.219, + "step": 2717 + }, + { + "epoch": 0.02718, + "grad_norm": 0.9312424963750244, + "learning_rate": 0.003, + "loss": 4.2354, + "step": 2718 + }, + { + "epoch": 0.02719, + "grad_norm": 0.6956685262676899, + "learning_rate": 0.003, + "loss": 4.2005, + "step": 2719 + }, + { + "epoch": 0.0272, + "grad_norm": 0.6830262491518131, + "learning_rate": 0.003, + "loss": 4.192, + "step": 2720 + }, + { + "epoch": 0.02721, + "grad_norm": 0.5790742761201393, + "learning_rate": 0.003, + "loss": 4.2022, + "step": 2721 + }, + { + "epoch": 0.02722, + "grad_norm": 0.6475692617370225, + "learning_rate": 0.003, + "loss": 4.2016, + "step": 2722 + }, + { + "epoch": 0.02723, + "grad_norm": 0.6175956588274222, + "learning_rate": 0.003, + "loss": 4.2065, + "step": 2723 + }, + { + "epoch": 0.02724, + "grad_norm": 0.5457779888221875, + "learning_rate": 0.003, + "loss": 4.2041, + "step": 2724 + }, + { + "epoch": 0.02725, + "grad_norm": 0.5026905482131955, + "learning_rate": 0.003, + "loss": 4.1969, + "step": 2725 + }, + { + "epoch": 0.02726, + "grad_norm": 0.42536659188846526, + "learning_rate": 0.003, + "loss": 4.1647, + "step": 2726 + }, + { + "epoch": 0.02727, + "grad_norm": 0.5003680404401011, + "learning_rate": 0.003, + "loss": 4.201, + "step": 2727 + }, + { + "epoch": 0.02728, + "grad_norm": 0.48640002941271926, + "learning_rate": 0.003, + "loss": 4.1833, + "step": 2728 + }, + { + "epoch": 0.02729, + "grad_norm": 0.41280683340266905, + "learning_rate": 0.003, + "loss": 4.2125, + "step": 2729 + }, + { + "epoch": 0.0273, + "grad_norm": 0.43446915053933527, + "learning_rate": 0.003, + "loss": 4.1856, + "step": 2730 + }, + { + "epoch": 0.02731, + "grad_norm": 0.51314133885114, + "learning_rate": 0.003, + "loss": 4.1847, + "step": 2731 + }, + { + "epoch": 0.02732, + "grad_norm": 0.5875155983292144, + "learning_rate": 0.003, + "loss": 4.1792, + "step": 2732 + }, + { + "epoch": 0.02733, + "grad_norm": 0.7435494959956005, + "learning_rate": 0.003, + "loss": 4.2133, + "step": 2733 + }, + { + "epoch": 0.02734, + "grad_norm": 0.8318623947885522, + "learning_rate": 0.003, + "loss": 4.1703, + "step": 2734 + }, + { + "epoch": 0.02735, + "grad_norm": 0.7270117460217965, + "learning_rate": 0.003, + "loss": 4.2139, + "step": 2735 + }, + { + "epoch": 0.02736, + "grad_norm": 0.5421807988773258, + "learning_rate": 0.003, + "loss": 4.1837, + "step": 2736 + }, + { + "epoch": 0.02737, + "grad_norm": 0.5828799726606484, + "learning_rate": 0.003, + "loss": 4.1756, + "step": 2737 + }, + { + "epoch": 0.02738, + "grad_norm": 0.6293605543493668, + "learning_rate": 0.003, + "loss": 4.2025, + "step": 2738 + }, + { + "epoch": 0.02739, + "grad_norm": 0.5663805709619927, + "learning_rate": 0.003, + "loss": 4.1907, + "step": 2739 + }, + { + "epoch": 0.0274, + "grad_norm": 0.5687899337220772, + "learning_rate": 0.003, + "loss": 4.1804, + "step": 2740 + }, + { + "epoch": 0.02741, + "grad_norm": 0.6075448196131129, + "learning_rate": 0.003, + "loss": 4.1946, + "step": 2741 + }, + { + "epoch": 0.02742, + "grad_norm": 0.550237016515108, + "learning_rate": 0.003, + "loss": 4.1804, + "step": 2742 + }, + { + "epoch": 0.02743, + "grad_norm": 0.6055901207765589, + "learning_rate": 0.003, + "loss": 4.2013, + "step": 2743 + }, + { + "epoch": 0.02744, + "grad_norm": 0.7300871101889295, + "learning_rate": 0.003, + "loss": 4.1884, + "step": 2744 + }, + { + "epoch": 0.02745, + "grad_norm": 0.8259955139853259, + "learning_rate": 0.003, + "loss": 4.2039, + "step": 2745 + }, + { + "epoch": 0.02746, + "grad_norm": 0.83396545170588, + "learning_rate": 0.003, + "loss": 4.1856, + "step": 2746 + }, + { + "epoch": 0.02747, + "grad_norm": 0.729485971791691, + "learning_rate": 0.003, + "loss": 4.2152, + "step": 2747 + }, + { + "epoch": 0.02748, + "grad_norm": 0.695156151246541, + "learning_rate": 0.003, + "loss": 4.1664, + "step": 2748 + }, + { + "epoch": 0.02749, + "grad_norm": 0.6271520188600127, + "learning_rate": 0.003, + "loss": 4.1715, + "step": 2749 + }, + { + "epoch": 0.0275, + "grad_norm": 0.6580176171244524, + "learning_rate": 0.003, + "loss": 4.2067, + "step": 2750 + }, + { + "epoch": 0.02751, + "grad_norm": 0.672675352925045, + "learning_rate": 0.003, + "loss": 4.2192, + "step": 2751 + }, + { + "epoch": 0.02752, + "grad_norm": 0.6249956014069801, + "learning_rate": 0.003, + "loss": 4.2068, + "step": 2752 + }, + { + "epoch": 0.02753, + "grad_norm": 0.6422052590772455, + "learning_rate": 0.003, + "loss": 4.1858, + "step": 2753 + }, + { + "epoch": 0.02754, + "grad_norm": 0.6380940449715866, + "learning_rate": 0.003, + "loss": 4.1719, + "step": 2754 + }, + { + "epoch": 0.02755, + "grad_norm": 0.5972581452735276, + "learning_rate": 0.003, + "loss": 4.1688, + "step": 2755 + }, + { + "epoch": 0.02756, + "grad_norm": 0.6077948327385506, + "learning_rate": 0.003, + "loss": 4.1967, + "step": 2756 + }, + { + "epoch": 0.02757, + "grad_norm": 0.7032781486647851, + "learning_rate": 0.003, + "loss": 4.2127, + "step": 2757 + }, + { + "epoch": 0.02758, + "grad_norm": 0.793287683394734, + "learning_rate": 0.003, + "loss": 4.1637, + "step": 2758 + }, + { + "epoch": 0.02759, + "grad_norm": 0.9418904888299408, + "learning_rate": 0.003, + "loss": 4.2024, + "step": 2759 + }, + { + "epoch": 0.0276, + "grad_norm": 0.9351479329737886, + "learning_rate": 0.003, + "loss": 4.215, + "step": 2760 + }, + { + "epoch": 0.02761, + "grad_norm": 0.8139578363148168, + "learning_rate": 0.003, + "loss": 4.2282, + "step": 2761 + }, + { + "epoch": 0.02762, + "grad_norm": 0.6757269439646322, + "learning_rate": 0.003, + "loss": 4.1918, + "step": 2762 + }, + { + "epoch": 0.02763, + "grad_norm": 0.673672124363711, + "learning_rate": 0.003, + "loss": 4.2006, + "step": 2763 + }, + { + "epoch": 0.02764, + "grad_norm": 0.8103484124048353, + "learning_rate": 0.003, + "loss": 4.199, + "step": 2764 + }, + { + "epoch": 0.02765, + "grad_norm": 0.861735499978179, + "learning_rate": 0.003, + "loss": 4.2084, + "step": 2765 + }, + { + "epoch": 0.02766, + "grad_norm": 0.7923213801421585, + "learning_rate": 0.003, + "loss": 4.1751, + "step": 2766 + }, + { + "epoch": 0.02767, + "grad_norm": 0.7631227515735746, + "learning_rate": 0.003, + "loss": 4.2302, + "step": 2767 + }, + { + "epoch": 0.02768, + "grad_norm": 0.8386215859490249, + "learning_rate": 0.003, + "loss": 4.1676, + "step": 2768 + }, + { + "epoch": 0.02769, + "grad_norm": 0.8716391723421844, + "learning_rate": 0.003, + "loss": 4.2312, + "step": 2769 + }, + { + "epoch": 0.0277, + "grad_norm": 0.79734786858364, + "learning_rate": 0.003, + "loss": 4.1984, + "step": 2770 + }, + { + "epoch": 0.02771, + "grad_norm": 0.7636865958720973, + "learning_rate": 0.003, + "loss": 4.187, + "step": 2771 + }, + { + "epoch": 0.02772, + "grad_norm": 0.7711684731670819, + "learning_rate": 0.003, + "loss": 4.1786, + "step": 2772 + }, + { + "epoch": 0.02773, + "grad_norm": 0.6855679524292124, + "learning_rate": 0.003, + "loss": 4.2122, + "step": 2773 + }, + { + "epoch": 0.02774, + "grad_norm": 0.6373232969577011, + "learning_rate": 0.003, + "loss": 4.2297, + "step": 2774 + }, + { + "epoch": 0.02775, + "grad_norm": 0.5808981715198482, + "learning_rate": 0.003, + "loss": 4.1967, + "step": 2775 + }, + { + "epoch": 0.02776, + "grad_norm": 0.551216477733259, + "learning_rate": 0.003, + "loss": 4.1912, + "step": 2776 + }, + { + "epoch": 0.02777, + "grad_norm": 0.5502623559066758, + "learning_rate": 0.003, + "loss": 4.1695, + "step": 2777 + }, + { + "epoch": 0.02778, + "grad_norm": 0.5471574038460382, + "learning_rate": 0.003, + "loss": 4.209, + "step": 2778 + }, + { + "epoch": 0.02779, + "grad_norm": 0.5740086450201776, + "learning_rate": 0.003, + "loss": 4.1802, + "step": 2779 + }, + { + "epoch": 0.0278, + "grad_norm": 0.6603722539886849, + "learning_rate": 0.003, + "loss": 4.1784, + "step": 2780 + }, + { + "epoch": 0.02781, + "grad_norm": 0.7045319397684864, + "learning_rate": 0.003, + "loss": 4.1629, + "step": 2781 + }, + { + "epoch": 0.02782, + "grad_norm": 0.7384112585585232, + "learning_rate": 0.003, + "loss": 4.189, + "step": 2782 + }, + { + "epoch": 0.02783, + "grad_norm": 0.7302997601128179, + "learning_rate": 0.003, + "loss": 4.1951, + "step": 2783 + }, + { + "epoch": 0.02784, + "grad_norm": 0.623059472233842, + "learning_rate": 0.003, + "loss": 4.1896, + "step": 2784 + }, + { + "epoch": 0.02785, + "grad_norm": 0.6333396786278916, + "learning_rate": 0.003, + "loss": 4.2038, + "step": 2785 + }, + { + "epoch": 0.02786, + "grad_norm": 0.7323878741557076, + "learning_rate": 0.003, + "loss": 4.2023, + "step": 2786 + }, + { + "epoch": 0.02787, + "grad_norm": 0.7244716928344251, + "learning_rate": 0.003, + "loss": 4.2035, + "step": 2787 + }, + { + "epoch": 0.02788, + "grad_norm": 0.6542769748417147, + "learning_rate": 0.003, + "loss": 4.1892, + "step": 2788 + }, + { + "epoch": 0.02789, + "grad_norm": 0.6574058601306716, + "learning_rate": 0.003, + "loss": 4.1939, + "step": 2789 + }, + { + "epoch": 0.0279, + "grad_norm": 0.7081427842141735, + "learning_rate": 0.003, + "loss": 4.1695, + "step": 2790 + }, + { + "epoch": 0.02791, + "grad_norm": 0.6942117835439664, + "learning_rate": 0.003, + "loss": 4.181, + "step": 2791 + }, + { + "epoch": 0.02792, + "grad_norm": 0.6067557855062717, + "learning_rate": 0.003, + "loss": 4.1954, + "step": 2792 + }, + { + "epoch": 0.02793, + "grad_norm": 0.5947324428169551, + "learning_rate": 0.003, + "loss": 4.1906, + "step": 2793 + }, + { + "epoch": 0.02794, + "grad_norm": 0.5401502710295608, + "learning_rate": 0.003, + "loss": 4.1858, + "step": 2794 + }, + { + "epoch": 0.02795, + "grad_norm": 0.5028686206702352, + "learning_rate": 0.003, + "loss": 4.2, + "step": 2795 + }, + { + "epoch": 0.02796, + "grad_norm": 0.49627952868559183, + "learning_rate": 0.003, + "loss": 4.191, + "step": 2796 + }, + { + "epoch": 0.02797, + "grad_norm": 0.4876291995281765, + "learning_rate": 0.003, + "loss": 4.2019, + "step": 2797 + }, + { + "epoch": 0.02798, + "grad_norm": 0.5294169569466303, + "learning_rate": 0.003, + "loss": 4.1979, + "step": 2798 + }, + { + "epoch": 0.02799, + "grad_norm": 0.5927669437619445, + "learning_rate": 0.003, + "loss": 4.1858, + "step": 2799 + }, + { + "epoch": 0.028, + "grad_norm": 0.6682805931240281, + "learning_rate": 0.003, + "loss": 4.1988, + "step": 2800 + }, + { + "epoch": 0.02801, + "grad_norm": 0.9068693661645921, + "learning_rate": 0.003, + "loss": 4.1815, + "step": 2801 + }, + { + "epoch": 0.02802, + "grad_norm": 1.0254407756501984, + "learning_rate": 0.003, + "loss": 4.1982, + "step": 2802 + }, + { + "epoch": 0.02803, + "grad_norm": 0.8128875626073906, + "learning_rate": 0.003, + "loss": 4.2114, + "step": 2803 + }, + { + "epoch": 0.02804, + "grad_norm": 0.7353680218352796, + "learning_rate": 0.003, + "loss": 4.18, + "step": 2804 + }, + { + "epoch": 0.02805, + "grad_norm": 0.7856799021536293, + "learning_rate": 0.003, + "loss": 4.222, + "step": 2805 + }, + { + "epoch": 0.02806, + "grad_norm": 0.7319188472007512, + "learning_rate": 0.003, + "loss": 4.1915, + "step": 2806 + }, + { + "epoch": 0.02807, + "grad_norm": 0.6281145778193726, + "learning_rate": 0.003, + "loss": 4.1913, + "step": 2807 + }, + { + "epoch": 0.02808, + "grad_norm": 0.6007996487222561, + "learning_rate": 0.003, + "loss": 4.1745, + "step": 2808 + }, + { + "epoch": 0.02809, + "grad_norm": 0.6107598223212796, + "learning_rate": 0.003, + "loss": 4.2139, + "step": 2809 + }, + { + "epoch": 0.0281, + "grad_norm": 0.5658938842173376, + "learning_rate": 0.003, + "loss": 4.1671, + "step": 2810 + }, + { + "epoch": 0.02811, + "grad_norm": 0.5188959758106707, + "learning_rate": 0.003, + "loss": 4.1753, + "step": 2811 + }, + { + "epoch": 0.02812, + "grad_norm": 0.47017604842162736, + "learning_rate": 0.003, + "loss": 4.19, + "step": 2812 + }, + { + "epoch": 0.02813, + "grad_norm": 0.43817425390848724, + "learning_rate": 0.003, + "loss": 4.1784, + "step": 2813 + }, + { + "epoch": 0.02814, + "grad_norm": 0.49665633422440225, + "learning_rate": 0.003, + "loss": 4.1789, + "step": 2814 + }, + { + "epoch": 0.02815, + "grad_norm": 0.5737919498303556, + "learning_rate": 0.003, + "loss": 4.2179, + "step": 2815 + }, + { + "epoch": 0.02816, + "grad_norm": 0.5958113314856148, + "learning_rate": 0.003, + "loss": 4.1978, + "step": 2816 + }, + { + "epoch": 0.02817, + "grad_norm": 0.7178958177645314, + "learning_rate": 0.003, + "loss": 4.1563, + "step": 2817 + }, + { + "epoch": 0.02818, + "grad_norm": 0.856700432217493, + "learning_rate": 0.003, + "loss": 4.1797, + "step": 2818 + }, + { + "epoch": 0.02819, + "grad_norm": 0.8381122457158061, + "learning_rate": 0.003, + "loss": 4.2052, + "step": 2819 + }, + { + "epoch": 0.0282, + "grad_norm": 0.9145463072400436, + "learning_rate": 0.003, + "loss": 4.2085, + "step": 2820 + }, + { + "epoch": 0.02821, + "grad_norm": 0.8180716884994599, + "learning_rate": 0.003, + "loss": 4.2204, + "step": 2821 + }, + { + "epoch": 0.02822, + "grad_norm": 0.749853073972982, + "learning_rate": 0.003, + "loss": 4.1962, + "step": 2822 + }, + { + "epoch": 0.02823, + "grad_norm": 0.8042960763019706, + "learning_rate": 0.003, + "loss": 4.2228, + "step": 2823 + }, + { + "epoch": 0.02824, + "grad_norm": 0.9373801354350708, + "learning_rate": 0.003, + "loss": 4.2156, + "step": 2824 + }, + { + "epoch": 0.02825, + "grad_norm": 1.0578826545081004, + "learning_rate": 0.003, + "loss": 4.2331, + "step": 2825 + }, + { + "epoch": 0.02826, + "grad_norm": 0.7996056170859099, + "learning_rate": 0.003, + "loss": 4.1963, + "step": 2826 + }, + { + "epoch": 0.02827, + "grad_norm": 0.8051765401350005, + "learning_rate": 0.003, + "loss": 4.1832, + "step": 2827 + }, + { + "epoch": 0.02828, + "grad_norm": 0.8027875497458861, + "learning_rate": 0.003, + "loss": 4.2252, + "step": 2828 + }, + { + "epoch": 0.02829, + "grad_norm": 0.7733380033017664, + "learning_rate": 0.003, + "loss": 4.2487, + "step": 2829 + }, + { + "epoch": 0.0283, + "grad_norm": 0.7838154488025357, + "learning_rate": 0.003, + "loss": 4.215, + "step": 2830 + }, + { + "epoch": 0.02831, + "grad_norm": 0.7961302139800379, + "learning_rate": 0.003, + "loss": 4.2125, + "step": 2831 + }, + { + "epoch": 0.02832, + "grad_norm": 0.8562794848672607, + "learning_rate": 0.003, + "loss": 4.2283, + "step": 2832 + }, + { + "epoch": 0.02833, + "grad_norm": 0.9219375196224617, + "learning_rate": 0.003, + "loss": 4.2033, + "step": 2833 + }, + { + "epoch": 0.02834, + "grad_norm": 1.066449364140732, + "learning_rate": 0.003, + "loss": 4.2018, + "step": 2834 + }, + { + "epoch": 0.02835, + "grad_norm": 0.9222685550708334, + "learning_rate": 0.003, + "loss": 4.2111, + "step": 2835 + }, + { + "epoch": 0.02836, + "grad_norm": 0.7684380143787258, + "learning_rate": 0.003, + "loss": 4.2263, + "step": 2836 + }, + { + "epoch": 0.02837, + "grad_norm": 0.6557155817543934, + "learning_rate": 0.003, + "loss": 4.1966, + "step": 2837 + }, + { + "epoch": 0.02838, + "grad_norm": 0.635865016407609, + "learning_rate": 0.003, + "loss": 4.1597, + "step": 2838 + }, + { + "epoch": 0.02839, + "grad_norm": 0.6046255360999631, + "learning_rate": 0.003, + "loss": 4.159, + "step": 2839 + }, + { + "epoch": 0.0284, + "grad_norm": 0.5605750331386489, + "learning_rate": 0.003, + "loss": 4.1687, + "step": 2840 + }, + { + "epoch": 0.02841, + "grad_norm": 0.5616642948887489, + "learning_rate": 0.003, + "loss": 4.206, + "step": 2841 + }, + { + "epoch": 0.02842, + "grad_norm": 0.57068171158362, + "learning_rate": 0.003, + "loss": 4.1872, + "step": 2842 + }, + { + "epoch": 0.02843, + "grad_norm": 0.5405942011469851, + "learning_rate": 0.003, + "loss": 4.2077, + "step": 2843 + }, + { + "epoch": 0.02844, + "grad_norm": 0.5611073031217337, + "learning_rate": 0.003, + "loss": 4.2038, + "step": 2844 + }, + { + "epoch": 0.02845, + "grad_norm": 0.6289027848917312, + "learning_rate": 0.003, + "loss": 4.1835, + "step": 2845 + }, + { + "epoch": 0.02846, + "grad_norm": 0.6920775632747969, + "learning_rate": 0.003, + "loss": 4.178, + "step": 2846 + }, + { + "epoch": 0.02847, + "grad_norm": 0.767579826338002, + "learning_rate": 0.003, + "loss": 4.2364, + "step": 2847 + }, + { + "epoch": 0.02848, + "grad_norm": 0.7249701206375546, + "learning_rate": 0.003, + "loss": 4.1955, + "step": 2848 + }, + { + "epoch": 0.02849, + "grad_norm": 0.516455337675884, + "learning_rate": 0.003, + "loss": 4.173, + "step": 2849 + }, + { + "epoch": 0.0285, + "grad_norm": 0.5059786145152202, + "learning_rate": 0.003, + "loss": 4.1762, + "step": 2850 + }, + { + "epoch": 0.02851, + "grad_norm": 0.5163885986759773, + "learning_rate": 0.003, + "loss": 4.2134, + "step": 2851 + }, + { + "epoch": 0.02852, + "grad_norm": 0.49258101706112184, + "learning_rate": 0.003, + "loss": 4.1891, + "step": 2852 + }, + { + "epoch": 0.02853, + "grad_norm": 0.500381435832321, + "learning_rate": 0.003, + "loss": 4.1572, + "step": 2853 + }, + { + "epoch": 0.02854, + "grad_norm": 0.4710049420804533, + "learning_rate": 0.003, + "loss": 4.1946, + "step": 2854 + }, + { + "epoch": 0.02855, + "grad_norm": 0.4796815944787949, + "learning_rate": 0.003, + "loss": 4.1909, + "step": 2855 + }, + { + "epoch": 0.02856, + "grad_norm": 0.5101632070884339, + "learning_rate": 0.003, + "loss": 4.1943, + "step": 2856 + }, + { + "epoch": 0.02857, + "grad_norm": 0.4913306093184678, + "learning_rate": 0.003, + "loss": 4.1728, + "step": 2857 + }, + { + "epoch": 0.02858, + "grad_norm": 0.4711311880423705, + "learning_rate": 0.003, + "loss": 4.1626, + "step": 2858 + }, + { + "epoch": 0.02859, + "grad_norm": 0.5082622329894141, + "learning_rate": 0.003, + "loss": 4.1713, + "step": 2859 + }, + { + "epoch": 0.0286, + "grad_norm": 0.5184604450005492, + "learning_rate": 0.003, + "loss": 4.1843, + "step": 2860 + }, + { + "epoch": 0.02861, + "grad_norm": 0.48267289618449316, + "learning_rate": 0.003, + "loss": 4.1541, + "step": 2861 + }, + { + "epoch": 0.02862, + "grad_norm": 0.4761246257076003, + "learning_rate": 0.003, + "loss": 4.1721, + "step": 2862 + }, + { + "epoch": 0.02863, + "grad_norm": 0.555990003004704, + "learning_rate": 0.003, + "loss": 4.1996, + "step": 2863 + }, + { + "epoch": 0.02864, + "grad_norm": 0.6398184414941633, + "learning_rate": 0.003, + "loss": 4.1725, + "step": 2864 + }, + { + "epoch": 0.02865, + "grad_norm": 0.7105267215046387, + "learning_rate": 0.003, + "loss": 4.2076, + "step": 2865 + }, + { + "epoch": 0.02866, + "grad_norm": 0.7313418455236842, + "learning_rate": 0.003, + "loss": 4.1911, + "step": 2866 + }, + { + "epoch": 0.02867, + "grad_norm": 0.6641528598178345, + "learning_rate": 0.003, + "loss": 4.1753, + "step": 2867 + }, + { + "epoch": 0.02868, + "grad_norm": 0.5930404639850378, + "learning_rate": 0.003, + "loss": 4.1816, + "step": 2868 + }, + { + "epoch": 0.02869, + "grad_norm": 0.6719622014213373, + "learning_rate": 0.003, + "loss": 4.1744, + "step": 2869 + }, + { + "epoch": 0.0287, + "grad_norm": 0.6550610400030609, + "learning_rate": 0.003, + "loss": 4.1971, + "step": 2870 + }, + { + "epoch": 0.02871, + "grad_norm": 0.6686050845776436, + "learning_rate": 0.003, + "loss": 4.1824, + "step": 2871 + }, + { + "epoch": 0.02872, + "grad_norm": 0.7301947656997149, + "learning_rate": 0.003, + "loss": 4.2172, + "step": 2872 + }, + { + "epoch": 0.02873, + "grad_norm": 0.6807495486455808, + "learning_rate": 0.003, + "loss": 4.1867, + "step": 2873 + }, + { + "epoch": 0.02874, + "grad_norm": 0.625144556477654, + "learning_rate": 0.003, + "loss": 4.1995, + "step": 2874 + }, + { + "epoch": 0.02875, + "grad_norm": 0.6566904903408397, + "learning_rate": 0.003, + "loss": 4.2051, + "step": 2875 + }, + { + "epoch": 0.02876, + "grad_norm": 0.759141268501963, + "learning_rate": 0.003, + "loss": 4.1921, + "step": 2876 + }, + { + "epoch": 0.02877, + "grad_norm": 0.8807634127058958, + "learning_rate": 0.003, + "loss": 4.1842, + "step": 2877 + }, + { + "epoch": 0.02878, + "grad_norm": 1.0261973428455773, + "learning_rate": 0.003, + "loss": 4.1965, + "step": 2878 + }, + { + "epoch": 0.02879, + "grad_norm": 0.888483399285398, + "learning_rate": 0.003, + "loss": 4.1815, + "step": 2879 + }, + { + "epoch": 0.0288, + "grad_norm": 0.8435453856609076, + "learning_rate": 0.003, + "loss": 4.1882, + "step": 2880 + }, + { + "epoch": 0.02881, + "grad_norm": 0.8865486451512147, + "learning_rate": 0.003, + "loss": 4.204, + "step": 2881 + }, + { + "epoch": 0.02882, + "grad_norm": 0.8928993496723746, + "learning_rate": 0.003, + "loss": 4.2242, + "step": 2882 + }, + { + "epoch": 0.02883, + "grad_norm": 0.8226859467424321, + "learning_rate": 0.003, + "loss": 4.2032, + "step": 2883 + }, + { + "epoch": 0.02884, + "grad_norm": 0.77429031635637, + "learning_rate": 0.003, + "loss": 4.2401, + "step": 2884 + }, + { + "epoch": 0.02885, + "grad_norm": 0.8922281925752906, + "learning_rate": 0.003, + "loss": 4.1845, + "step": 2885 + }, + { + "epoch": 0.02886, + "grad_norm": 0.9214758547731667, + "learning_rate": 0.003, + "loss": 4.2251, + "step": 2886 + }, + { + "epoch": 0.02887, + "grad_norm": 0.7646945576422838, + "learning_rate": 0.003, + "loss": 4.205, + "step": 2887 + }, + { + "epoch": 0.02888, + "grad_norm": 0.5765970376943947, + "learning_rate": 0.003, + "loss": 4.1923, + "step": 2888 + }, + { + "epoch": 0.02889, + "grad_norm": 0.6149544262580506, + "learning_rate": 0.003, + "loss": 4.1823, + "step": 2889 + }, + { + "epoch": 0.0289, + "grad_norm": 0.5856358962779576, + "learning_rate": 0.003, + "loss": 4.2099, + "step": 2890 + }, + { + "epoch": 0.02891, + "grad_norm": 0.5989364430008874, + "learning_rate": 0.003, + "loss": 4.2274, + "step": 2891 + }, + { + "epoch": 0.02892, + "grad_norm": 0.6055146901336932, + "learning_rate": 0.003, + "loss": 4.1785, + "step": 2892 + }, + { + "epoch": 0.02893, + "grad_norm": 0.6302785437079876, + "learning_rate": 0.003, + "loss": 4.185, + "step": 2893 + }, + { + "epoch": 0.02894, + "grad_norm": 0.5540105467551689, + "learning_rate": 0.003, + "loss": 4.1878, + "step": 2894 + }, + { + "epoch": 0.02895, + "grad_norm": 0.4023612128343426, + "learning_rate": 0.003, + "loss": 4.2266, + "step": 2895 + }, + { + "epoch": 0.02896, + "grad_norm": 0.46314453746126943, + "learning_rate": 0.003, + "loss": 4.1815, + "step": 2896 + }, + { + "epoch": 0.02897, + "grad_norm": 0.5689975115372741, + "learning_rate": 0.003, + "loss": 4.1638, + "step": 2897 + }, + { + "epoch": 0.02898, + "grad_norm": 0.5759099877999215, + "learning_rate": 0.003, + "loss": 4.2076, + "step": 2898 + }, + { + "epoch": 0.02899, + "grad_norm": 0.5841372181240183, + "learning_rate": 0.003, + "loss": 4.1911, + "step": 2899 + }, + { + "epoch": 0.029, + "grad_norm": 0.5742774291659906, + "learning_rate": 0.003, + "loss": 4.1791, + "step": 2900 + }, + { + "epoch": 0.02901, + "grad_norm": 0.505503655503479, + "learning_rate": 0.003, + "loss": 4.1922, + "step": 2901 + }, + { + "epoch": 0.02902, + "grad_norm": 0.47503490148304944, + "learning_rate": 0.003, + "loss": 4.2092, + "step": 2902 + }, + { + "epoch": 0.02903, + "grad_norm": 0.5474247569951972, + "learning_rate": 0.003, + "loss": 4.1689, + "step": 2903 + }, + { + "epoch": 0.02904, + "grad_norm": 0.6382457425375883, + "learning_rate": 0.003, + "loss": 4.1897, + "step": 2904 + }, + { + "epoch": 0.02905, + "grad_norm": 0.707934254197877, + "learning_rate": 0.003, + "loss": 4.2073, + "step": 2905 + }, + { + "epoch": 0.02906, + "grad_norm": 0.8403476091404251, + "learning_rate": 0.003, + "loss": 4.192, + "step": 2906 + }, + { + "epoch": 0.02907, + "grad_norm": 0.8318613643365537, + "learning_rate": 0.003, + "loss": 4.1704, + "step": 2907 + }, + { + "epoch": 0.02908, + "grad_norm": 0.7272031628626594, + "learning_rate": 0.003, + "loss": 4.1912, + "step": 2908 + }, + { + "epoch": 0.02909, + "grad_norm": 0.669946483884344, + "learning_rate": 0.003, + "loss": 4.1853, + "step": 2909 + }, + { + "epoch": 0.0291, + "grad_norm": 0.6105283375602145, + "learning_rate": 0.003, + "loss": 4.1946, + "step": 2910 + }, + { + "epoch": 0.02911, + "grad_norm": 0.6965941358145008, + "learning_rate": 0.003, + "loss": 4.1939, + "step": 2911 + }, + { + "epoch": 0.02912, + "grad_norm": 0.6975491884532862, + "learning_rate": 0.003, + "loss": 4.1654, + "step": 2912 + }, + { + "epoch": 0.02913, + "grad_norm": 0.7857229509367818, + "learning_rate": 0.003, + "loss": 4.1992, + "step": 2913 + }, + { + "epoch": 0.02914, + "grad_norm": 0.811281083177593, + "learning_rate": 0.003, + "loss": 4.1892, + "step": 2914 + }, + { + "epoch": 0.02915, + "grad_norm": 0.6690033742013554, + "learning_rate": 0.003, + "loss": 4.1796, + "step": 2915 + }, + { + "epoch": 0.02916, + "grad_norm": 0.5795750985593426, + "learning_rate": 0.003, + "loss": 4.1847, + "step": 2916 + }, + { + "epoch": 0.02917, + "grad_norm": 0.6521162065161483, + "learning_rate": 0.003, + "loss": 4.2172, + "step": 2917 + }, + { + "epoch": 0.02918, + "grad_norm": 0.6434845720798231, + "learning_rate": 0.003, + "loss": 4.1719, + "step": 2918 + }, + { + "epoch": 0.02919, + "grad_norm": 0.6415437095317365, + "learning_rate": 0.003, + "loss": 4.2133, + "step": 2919 + }, + { + "epoch": 0.0292, + "grad_norm": 0.6445863883052798, + "learning_rate": 0.003, + "loss": 4.1534, + "step": 2920 + }, + { + "epoch": 0.02921, + "grad_norm": 0.5711168990862754, + "learning_rate": 0.003, + "loss": 4.1919, + "step": 2921 + }, + { + "epoch": 0.02922, + "grad_norm": 0.6255743522411465, + "learning_rate": 0.003, + "loss": 4.1926, + "step": 2922 + }, + { + "epoch": 0.02923, + "grad_norm": 0.7756468163112304, + "learning_rate": 0.003, + "loss": 4.1675, + "step": 2923 + }, + { + "epoch": 0.02924, + "grad_norm": 0.8580521016504452, + "learning_rate": 0.003, + "loss": 4.2067, + "step": 2924 + }, + { + "epoch": 0.02925, + "grad_norm": 0.8062805285298449, + "learning_rate": 0.003, + "loss": 4.2084, + "step": 2925 + }, + { + "epoch": 0.02926, + "grad_norm": 0.6123939436837456, + "learning_rate": 0.003, + "loss": 4.1817, + "step": 2926 + }, + { + "epoch": 0.02927, + "grad_norm": 0.5880261599317353, + "learning_rate": 0.003, + "loss": 4.1862, + "step": 2927 + }, + { + "epoch": 0.02928, + "grad_norm": 0.6052041112444663, + "learning_rate": 0.003, + "loss": 4.2074, + "step": 2928 + }, + { + "epoch": 0.02929, + "grad_norm": 0.6235842112912531, + "learning_rate": 0.003, + "loss": 4.192, + "step": 2929 + }, + { + "epoch": 0.0293, + "grad_norm": 0.660214912657757, + "learning_rate": 0.003, + "loss": 4.1869, + "step": 2930 + }, + { + "epoch": 0.02931, + "grad_norm": 0.5466629516534421, + "learning_rate": 0.003, + "loss": 4.1832, + "step": 2931 + }, + { + "epoch": 0.02932, + "grad_norm": 0.5338608341273119, + "learning_rate": 0.003, + "loss": 4.1686, + "step": 2932 + }, + { + "epoch": 0.02933, + "grad_norm": 0.4969939045829935, + "learning_rate": 0.003, + "loss": 4.1951, + "step": 2933 + }, + { + "epoch": 0.02934, + "grad_norm": 0.49477597477846896, + "learning_rate": 0.003, + "loss": 4.1818, + "step": 2934 + }, + { + "epoch": 0.02935, + "grad_norm": 0.4464607167642475, + "learning_rate": 0.003, + "loss": 4.1672, + "step": 2935 + }, + { + "epoch": 0.02936, + "grad_norm": 0.49350443190661963, + "learning_rate": 0.003, + "loss": 4.1877, + "step": 2936 + }, + { + "epoch": 0.02937, + "grad_norm": 0.6198926034158319, + "learning_rate": 0.003, + "loss": 4.1699, + "step": 2937 + }, + { + "epoch": 0.02938, + "grad_norm": 0.7272182520738907, + "learning_rate": 0.003, + "loss": 4.1923, + "step": 2938 + }, + { + "epoch": 0.02939, + "grad_norm": 0.8486086179503688, + "learning_rate": 0.003, + "loss": 4.1848, + "step": 2939 + }, + { + "epoch": 0.0294, + "grad_norm": 0.8446986466812662, + "learning_rate": 0.003, + "loss": 4.2102, + "step": 2940 + }, + { + "epoch": 0.02941, + "grad_norm": 0.7723043017297002, + "learning_rate": 0.003, + "loss": 4.208, + "step": 2941 + }, + { + "epoch": 0.02942, + "grad_norm": 0.7697250754745237, + "learning_rate": 0.003, + "loss": 4.197, + "step": 2942 + }, + { + "epoch": 0.02943, + "grad_norm": 0.741727859654241, + "learning_rate": 0.003, + "loss": 4.181, + "step": 2943 + }, + { + "epoch": 0.02944, + "grad_norm": 0.691917422319892, + "learning_rate": 0.003, + "loss": 4.2047, + "step": 2944 + }, + { + "epoch": 0.02945, + "grad_norm": 0.6550566422880618, + "learning_rate": 0.003, + "loss": 4.1868, + "step": 2945 + }, + { + "epoch": 0.02946, + "grad_norm": 0.7775286126916453, + "learning_rate": 0.003, + "loss": 4.1841, + "step": 2946 + }, + { + "epoch": 0.02947, + "grad_norm": 0.989103701566936, + "learning_rate": 0.003, + "loss": 4.2271, + "step": 2947 + }, + { + "epoch": 0.02948, + "grad_norm": 1.1031413486335138, + "learning_rate": 0.003, + "loss": 4.2196, + "step": 2948 + }, + { + "epoch": 0.02949, + "grad_norm": 0.9071172160919259, + "learning_rate": 0.003, + "loss": 4.1919, + "step": 2949 + }, + { + "epoch": 0.0295, + "grad_norm": 0.9640749695276918, + "learning_rate": 0.003, + "loss": 4.2161, + "step": 2950 + }, + { + "epoch": 0.02951, + "grad_norm": 1.0552794520418902, + "learning_rate": 0.003, + "loss": 4.1768, + "step": 2951 + }, + { + "epoch": 0.02952, + "grad_norm": 1.0245244401252407, + "learning_rate": 0.003, + "loss": 4.2154, + "step": 2952 + }, + { + "epoch": 0.02953, + "grad_norm": 1.0404102150119146, + "learning_rate": 0.003, + "loss": 4.2329, + "step": 2953 + }, + { + "epoch": 0.02954, + "grad_norm": 1.200181394447266, + "learning_rate": 0.003, + "loss": 4.2778, + "step": 2954 + }, + { + "epoch": 0.02955, + "grad_norm": 0.8161704211969852, + "learning_rate": 0.003, + "loss": 4.2394, + "step": 2955 + }, + { + "epoch": 0.02956, + "grad_norm": 0.8295432503492015, + "learning_rate": 0.003, + "loss": 4.2255, + "step": 2956 + }, + { + "epoch": 0.02957, + "grad_norm": 0.8024603706062242, + "learning_rate": 0.003, + "loss": 4.2343, + "step": 2957 + }, + { + "epoch": 0.02958, + "grad_norm": 0.9622635969093861, + "learning_rate": 0.003, + "loss": 4.2527, + "step": 2958 + }, + { + "epoch": 0.02959, + "grad_norm": 1.144055914507281, + "learning_rate": 0.003, + "loss": 4.2509, + "step": 2959 + }, + { + "epoch": 0.0296, + "grad_norm": 0.8230875124438284, + "learning_rate": 0.003, + "loss": 4.2477, + "step": 2960 + }, + { + "epoch": 0.02961, + "grad_norm": 0.7761645102479322, + "learning_rate": 0.003, + "loss": 4.2177, + "step": 2961 + }, + { + "epoch": 0.02962, + "grad_norm": 0.5798322996384551, + "learning_rate": 0.003, + "loss": 4.2146, + "step": 2962 + }, + { + "epoch": 0.02963, + "grad_norm": 0.6082007449676078, + "learning_rate": 0.003, + "loss": 4.1885, + "step": 2963 + }, + { + "epoch": 0.02964, + "grad_norm": 0.5952539583915469, + "learning_rate": 0.003, + "loss": 4.1722, + "step": 2964 + }, + { + "epoch": 0.02965, + "grad_norm": 0.6067339945313683, + "learning_rate": 0.003, + "loss": 4.2192, + "step": 2965 + }, + { + "epoch": 0.02966, + "grad_norm": 0.6300423766569044, + "learning_rate": 0.003, + "loss": 4.2275, + "step": 2966 + }, + { + "epoch": 0.02967, + "grad_norm": 0.6210602868633931, + "learning_rate": 0.003, + "loss": 4.2121, + "step": 2967 + }, + { + "epoch": 0.02968, + "grad_norm": 0.5447153342269296, + "learning_rate": 0.003, + "loss": 4.208, + "step": 2968 + }, + { + "epoch": 0.02969, + "grad_norm": 0.5492051288557623, + "learning_rate": 0.003, + "loss": 4.2026, + "step": 2969 + }, + { + "epoch": 0.0297, + "grad_norm": 0.5667658969213387, + "learning_rate": 0.003, + "loss": 4.1857, + "step": 2970 + }, + { + "epoch": 0.02971, + "grad_norm": 0.5386297699076645, + "learning_rate": 0.003, + "loss": 4.2025, + "step": 2971 + }, + { + "epoch": 0.02972, + "grad_norm": 0.4946208196601275, + "learning_rate": 0.003, + "loss": 4.1702, + "step": 2972 + }, + { + "epoch": 0.02973, + "grad_norm": 0.5841767565220707, + "learning_rate": 0.003, + "loss": 4.1584, + "step": 2973 + }, + { + "epoch": 0.02974, + "grad_norm": 0.7076979032412727, + "learning_rate": 0.003, + "loss": 4.1953, + "step": 2974 + }, + { + "epoch": 0.02975, + "grad_norm": 0.9835377294492083, + "learning_rate": 0.003, + "loss": 4.1927, + "step": 2975 + }, + { + "epoch": 0.02976, + "grad_norm": 1.1071095377980806, + "learning_rate": 0.003, + "loss": 4.2143, + "step": 2976 + }, + { + "epoch": 0.02977, + "grad_norm": 0.7259690969712574, + "learning_rate": 0.003, + "loss": 4.1859, + "step": 2977 + }, + { + "epoch": 0.02978, + "grad_norm": 0.7734023943128602, + "learning_rate": 0.003, + "loss": 4.2094, + "step": 2978 + }, + { + "epoch": 0.02979, + "grad_norm": 0.7278346414839765, + "learning_rate": 0.003, + "loss": 4.1886, + "step": 2979 + }, + { + "epoch": 0.0298, + "grad_norm": 0.660497169411274, + "learning_rate": 0.003, + "loss": 4.1931, + "step": 2980 + }, + { + "epoch": 0.02981, + "grad_norm": 0.6667577508097533, + "learning_rate": 0.003, + "loss": 4.2009, + "step": 2981 + }, + { + "epoch": 0.02982, + "grad_norm": 0.6386554072904321, + "learning_rate": 0.003, + "loss": 4.1958, + "step": 2982 + }, + { + "epoch": 0.02983, + "grad_norm": 0.5418480426752083, + "learning_rate": 0.003, + "loss": 4.2068, + "step": 2983 + }, + { + "epoch": 0.02984, + "grad_norm": 0.5015711298760291, + "learning_rate": 0.003, + "loss": 4.1815, + "step": 2984 + }, + { + "epoch": 0.02985, + "grad_norm": 0.47335120613727616, + "learning_rate": 0.003, + "loss": 4.2107, + "step": 2985 + }, + { + "epoch": 0.02986, + "grad_norm": 0.45169284672546406, + "learning_rate": 0.003, + "loss": 4.1679, + "step": 2986 + }, + { + "epoch": 0.02987, + "grad_norm": 0.3872178111815347, + "learning_rate": 0.003, + "loss": 4.1603, + "step": 2987 + }, + { + "epoch": 0.02988, + "grad_norm": 0.5156766146850885, + "learning_rate": 0.003, + "loss": 4.176, + "step": 2988 + }, + { + "epoch": 0.02989, + "grad_norm": 0.5734795816734266, + "learning_rate": 0.003, + "loss": 4.1903, + "step": 2989 + }, + { + "epoch": 0.0299, + "grad_norm": 0.6108581797072585, + "learning_rate": 0.003, + "loss": 4.1491, + "step": 2990 + }, + { + "epoch": 0.02991, + "grad_norm": 0.6465308189507616, + "learning_rate": 0.003, + "loss": 4.2032, + "step": 2991 + }, + { + "epoch": 0.02992, + "grad_norm": 0.6024426991759066, + "learning_rate": 0.003, + "loss": 4.1963, + "step": 2992 + }, + { + "epoch": 0.02993, + "grad_norm": 0.5450535416942648, + "learning_rate": 0.003, + "loss": 4.1677, + "step": 2993 + }, + { + "epoch": 0.02994, + "grad_norm": 0.5271060232078253, + "learning_rate": 0.003, + "loss": 4.1764, + "step": 2994 + }, + { + "epoch": 0.02995, + "grad_norm": 0.45562729812821884, + "learning_rate": 0.003, + "loss": 4.2083, + "step": 2995 + }, + { + "epoch": 0.02996, + "grad_norm": 0.460767745641914, + "learning_rate": 0.003, + "loss": 4.1977, + "step": 2996 + }, + { + "epoch": 0.02997, + "grad_norm": 0.49883686701327856, + "learning_rate": 0.003, + "loss": 4.1594, + "step": 2997 + }, + { + "epoch": 0.02998, + "grad_norm": 0.5495557644619213, + "learning_rate": 0.003, + "loss": 4.1591, + "step": 2998 + }, + { + "epoch": 0.02999, + "grad_norm": 0.6523752894410377, + "learning_rate": 0.003, + "loss": 4.1565, + "step": 2999 + }, + { + "epoch": 0.03, + "grad_norm": 0.7129180858865645, + "learning_rate": 0.003, + "loss": 4.2061, + "step": 3000 + }, + { + "epoch": 0.03001, + "grad_norm": 0.7274677458257305, + "learning_rate": 0.003, + "loss": 4.1741, + "step": 3001 + }, + { + "epoch": 0.03002, + "grad_norm": 0.6467948184705592, + "learning_rate": 0.003, + "loss": 4.1853, + "step": 3002 + }, + { + "epoch": 0.03003, + "grad_norm": 0.6600036116283045, + "learning_rate": 0.003, + "loss": 4.1654, + "step": 3003 + }, + { + "epoch": 0.03004, + "grad_norm": 0.7176122662961554, + "learning_rate": 0.003, + "loss": 4.1486, + "step": 3004 + }, + { + "epoch": 0.03005, + "grad_norm": 0.6686551022549297, + "learning_rate": 0.003, + "loss": 4.1965, + "step": 3005 + }, + { + "epoch": 0.03006, + "grad_norm": 0.7091833087612831, + "learning_rate": 0.003, + "loss": 4.1949, + "step": 3006 + }, + { + "epoch": 0.03007, + "grad_norm": 0.6742803685976605, + "learning_rate": 0.003, + "loss": 4.1664, + "step": 3007 + }, + { + "epoch": 0.03008, + "grad_norm": 0.6530504414551399, + "learning_rate": 0.003, + "loss": 4.1926, + "step": 3008 + }, + { + "epoch": 0.03009, + "grad_norm": 0.7147134856842983, + "learning_rate": 0.003, + "loss": 4.1681, + "step": 3009 + }, + { + "epoch": 0.0301, + "grad_norm": 0.803723268068343, + "learning_rate": 0.003, + "loss": 4.1972, + "step": 3010 + }, + { + "epoch": 0.03011, + "grad_norm": 0.8381383897634889, + "learning_rate": 0.003, + "loss": 4.1805, + "step": 3011 + }, + { + "epoch": 0.03012, + "grad_norm": 0.6693037897815421, + "learning_rate": 0.003, + "loss": 4.1812, + "step": 3012 + }, + { + "epoch": 0.03013, + "grad_norm": 0.5695069096544523, + "learning_rate": 0.003, + "loss": 4.14, + "step": 3013 + }, + { + "epoch": 0.03014, + "grad_norm": 0.5914912051432484, + "learning_rate": 0.003, + "loss": 4.1697, + "step": 3014 + }, + { + "epoch": 0.03015, + "grad_norm": 0.6342676904398465, + "learning_rate": 0.003, + "loss": 4.1663, + "step": 3015 + }, + { + "epoch": 0.03016, + "grad_norm": 0.643666546817278, + "learning_rate": 0.003, + "loss": 4.1758, + "step": 3016 + }, + { + "epoch": 0.03017, + "grad_norm": 0.6248622988385808, + "learning_rate": 0.003, + "loss": 4.1711, + "step": 3017 + }, + { + "epoch": 0.03018, + "grad_norm": 0.5659743586845699, + "learning_rate": 0.003, + "loss": 4.1652, + "step": 3018 + }, + { + "epoch": 0.03019, + "grad_norm": 0.543953415462997, + "learning_rate": 0.003, + "loss": 4.1825, + "step": 3019 + }, + { + "epoch": 0.0302, + "grad_norm": 0.4979549215791465, + "learning_rate": 0.003, + "loss": 4.171, + "step": 3020 + }, + { + "epoch": 0.03021, + "grad_norm": 0.5988839209053217, + "learning_rate": 0.003, + "loss": 4.1986, + "step": 3021 + }, + { + "epoch": 0.03022, + "grad_norm": 0.6917001050231241, + "learning_rate": 0.003, + "loss": 4.1995, + "step": 3022 + }, + { + "epoch": 0.03023, + "grad_norm": 0.767246802866098, + "learning_rate": 0.003, + "loss": 4.215, + "step": 3023 + }, + { + "epoch": 0.03024, + "grad_norm": 0.7171193879296623, + "learning_rate": 0.003, + "loss": 4.2081, + "step": 3024 + }, + { + "epoch": 0.03025, + "grad_norm": 0.8041231345280121, + "learning_rate": 0.003, + "loss": 4.1898, + "step": 3025 + }, + { + "epoch": 0.03026, + "grad_norm": 0.8792691727950304, + "learning_rate": 0.003, + "loss": 4.169, + "step": 3026 + }, + { + "epoch": 0.03027, + "grad_norm": 1.0175395228935675, + "learning_rate": 0.003, + "loss": 4.1955, + "step": 3027 + }, + { + "epoch": 0.03028, + "grad_norm": 0.923391249014174, + "learning_rate": 0.003, + "loss": 4.1605, + "step": 3028 + }, + { + "epoch": 0.03029, + "grad_norm": 0.7571617694978044, + "learning_rate": 0.003, + "loss": 4.1972, + "step": 3029 + }, + { + "epoch": 0.0303, + "grad_norm": 0.8910919036216922, + "learning_rate": 0.003, + "loss": 4.1858, + "step": 3030 + }, + { + "epoch": 0.03031, + "grad_norm": 0.9641578814377423, + "learning_rate": 0.003, + "loss": 4.1735, + "step": 3031 + }, + { + "epoch": 0.03032, + "grad_norm": 0.9409295179057755, + "learning_rate": 0.003, + "loss": 4.2025, + "step": 3032 + }, + { + "epoch": 0.03033, + "grad_norm": 0.8317283746521951, + "learning_rate": 0.003, + "loss": 4.1969, + "step": 3033 + }, + { + "epoch": 0.03034, + "grad_norm": 0.7313668328730979, + "learning_rate": 0.003, + "loss": 4.1976, + "step": 3034 + }, + { + "epoch": 0.03035, + "grad_norm": 0.7132789456809936, + "learning_rate": 0.003, + "loss": 4.2248, + "step": 3035 + }, + { + "epoch": 0.03036, + "grad_norm": 0.674714029441459, + "learning_rate": 0.003, + "loss": 4.1962, + "step": 3036 + }, + { + "epoch": 0.03037, + "grad_norm": 0.7009049198448731, + "learning_rate": 0.003, + "loss": 4.1813, + "step": 3037 + }, + { + "epoch": 0.03038, + "grad_norm": 0.6254153988955986, + "learning_rate": 0.003, + "loss": 4.1778, + "step": 3038 + }, + { + "epoch": 0.03039, + "grad_norm": 0.6327013070531727, + "learning_rate": 0.003, + "loss": 4.1827, + "step": 3039 + }, + { + "epoch": 0.0304, + "grad_norm": 0.6070726773833675, + "learning_rate": 0.003, + "loss": 4.1868, + "step": 3040 + }, + { + "epoch": 0.03041, + "grad_norm": 0.5945854551887333, + "learning_rate": 0.003, + "loss": 4.1928, + "step": 3041 + }, + { + "epoch": 0.03042, + "grad_norm": 0.5653702360336539, + "learning_rate": 0.003, + "loss": 4.2138, + "step": 3042 + }, + { + "epoch": 0.03043, + "grad_norm": 0.48339928862065246, + "learning_rate": 0.003, + "loss": 4.1665, + "step": 3043 + }, + { + "epoch": 0.03044, + "grad_norm": 0.4882765542102775, + "learning_rate": 0.003, + "loss": 4.1658, + "step": 3044 + }, + { + "epoch": 0.03045, + "grad_norm": 0.49906211495716746, + "learning_rate": 0.003, + "loss": 4.1682, + "step": 3045 + }, + { + "epoch": 0.03046, + "grad_norm": 0.6073778663358335, + "learning_rate": 0.003, + "loss": 4.1808, + "step": 3046 + }, + { + "epoch": 0.03047, + "grad_norm": 0.767480218845333, + "learning_rate": 0.003, + "loss": 4.1898, + "step": 3047 + }, + { + "epoch": 0.03048, + "grad_norm": 0.8141142563495186, + "learning_rate": 0.003, + "loss": 4.1741, + "step": 3048 + }, + { + "epoch": 0.03049, + "grad_norm": 0.882596084462819, + "learning_rate": 0.003, + "loss": 4.164, + "step": 3049 + }, + { + "epoch": 0.0305, + "grad_norm": 0.872061418696622, + "learning_rate": 0.003, + "loss": 4.1847, + "step": 3050 + }, + { + "epoch": 0.03051, + "grad_norm": 0.8096508372990807, + "learning_rate": 0.003, + "loss": 4.1923, + "step": 3051 + }, + { + "epoch": 0.03052, + "grad_norm": 0.7731212144212846, + "learning_rate": 0.003, + "loss": 4.1995, + "step": 3052 + }, + { + "epoch": 0.03053, + "grad_norm": 0.8121040926587938, + "learning_rate": 0.003, + "loss": 4.2108, + "step": 3053 + }, + { + "epoch": 0.03054, + "grad_norm": 0.8031490148625648, + "learning_rate": 0.003, + "loss": 4.19, + "step": 3054 + }, + { + "epoch": 0.03055, + "grad_norm": 0.7123837580650032, + "learning_rate": 0.003, + "loss": 4.1921, + "step": 3055 + }, + { + "epoch": 0.03056, + "grad_norm": 0.6851652643314644, + "learning_rate": 0.003, + "loss": 4.2013, + "step": 3056 + }, + { + "epoch": 0.03057, + "grad_norm": 0.7381209914846429, + "learning_rate": 0.003, + "loss": 4.205, + "step": 3057 + }, + { + "epoch": 0.03058, + "grad_norm": 0.8318560657251862, + "learning_rate": 0.003, + "loss": 4.1792, + "step": 3058 + }, + { + "epoch": 0.03059, + "grad_norm": 0.8791895081718305, + "learning_rate": 0.003, + "loss": 4.1917, + "step": 3059 + }, + { + "epoch": 0.0306, + "grad_norm": 0.814947872922507, + "learning_rate": 0.003, + "loss": 4.1796, + "step": 3060 + }, + { + "epoch": 0.03061, + "grad_norm": 0.6938901592564853, + "learning_rate": 0.003, + "loss": 4.1987, + "step": 3061 + }, + { + "epoch": 0.03062, + "grad_norm": 0.7477036973476255, + "learning_rate": 0.003, + "loss": 4.225, + "step": 3062 + }, + { + "epoch": 0.03063, + "grad_norm": 0.6602150387540897, + "learning_rate": 0.003, + "loss": 4.1652, + "step": 3063 + }, + { + "epoch": 0.03064, + "grad_norm": 0.588757114074522, + "learning_rate": 0.003, + "loss": 4.1973, + "step": 3064 + }, + { + "epoch": 0.03065, + "grad_norm": 0.5129768984617096, + "learning_rate": 0.003, + "loss": 4.1599, + "step": 3065 + }, + { + "epoch": 0.03066, + "grad_norm": 0.5610278353764739, + "learning_rate": 0.003, + "loss": 4.2055, + "step": 3066 + }, + { + "epoch": 0.03067, + "grad_norm": 0.5604919558505147, + "learning_rate": 0.003, + "loss": 4.209, + "step": 3067 + }, + { + "epoch": 0.03068, + "grad_norm": 0.6135705525861784, + "learning_rate": 0.003, + "loss": 4.1804, + "step": 3068 + }, + { + "epoch": 0.03069, + "grad_norm": 0.6668599988517637, + "learning_rate": 0.003, + "loss": 4.1668, + "step": 3069 + }, + { + "epoch": 0.0307, + "grad_norm": 0.7555455276952089, + "learning_rate": 0.003, + "loss": 4.1799, + "step": 3070 + }, + { + "epoch": 0.03071, + "grad_norm": 0.6246930367493289, + "learning_rate": 0.003, + "loss": 4.1743, + "step": 3071 + }, + { + "epoch": 0.03072, + "grad_norm": 0.5918561396926248, + "learning_rate": 0.003, + "loss": 4.1774, + "step": 3072 + }, + { + "epoch": 0.03073, + "grad_norm": 0.543534573866796, + "learning_rate": 0.003, + "loss": 4.1741, + "step": 3073 + }, + { + "epoch": 0.03074, + "grad_norm": 0.5906611145957643, + "learning_rate": 0.003, + "loss": 4.1755, + "step": 3074 + }, + { + "epoch": 0.03075, + "grad_norm": 0.6400240036041446, + "learning_rate": 0.003, + "loss": 4.2013, + "step": 3075 + }, + { + "epoch": 0.03076, + "grad_norm": 0.5704867962206898, + "learning_rate": 0.003, + "loss": 4.1943, + "step": 3076 + }, + { + "epoch": 0.03077, + "grad_norm": 0.5207074704544816, + "learning_rate": 0.003, + "loss": 4.1858, + "step": 3077 + }, + { + "epoch": 0.03078, + "grad_norm": 0.49735495203144253, + "learning_rate": 0.003, + "loss": 4.1718, + "step": 3078 + }, + { + "epoch": 0.03079, + "grad_norm": 0.49371618460089295, + "learning_rate": 0.003, + "loss": 4.159, + "step": 3079 + }, + { + "epoch": 0.0308, + "grad_norm": 0.5286026033951621, + "learning_rate": 0.003, + "loss": 4.1869, + "step": 3080 + }, + { + "epoch": 0.03081, + "grad_norm": 0.597818150474852, + "learning_rate": 0.003, + "loss": 4.1589, + "step": 3081 + }, + { + "epoch": 0.03082, + "grad_norm": 0.673672308591032, + "learning_rate": 0.003, + "loss": 4.1645, + "step": 3082 + }, + { + "epoch": 0.03083, + "grad_norm": 0.742812911307247, + "learning_rate": 0.003, + "loss": 4.1694, + "step": 3083 + }, + { + "epoch": 0.03084, + "grad_norm": 0.7219826879276637, + "learning_rate": 0.003, + "loss": 4.1823, + "step": 3084 + }, + { + "epoch": 0.03085, + "grad_norm": 0.6340360256960204, + "learning_rate": 0.003, + "loss": 4.1754, + "step": 3085 + }, + { + "epoch": 0.03086, + "grad_norm": 0.5692756028815859, + "learning_rate": 0.003, + "loss": 4.1965, + "step": 3086 + }, + { + "epoch": 0.03087, + "grad_norm": 0.6389893757240679, + "learning_rate": 0.003, + "loss": 4.1728, + "step": 3087 + }, + { + "epoch": 0.03088, + "grad_norm": 0.6896729441240512, + "learning_rate": 0.003, + "loss": 4.1909, + "step": 3088 + }, + { + "epoch": 0.03089, + "grad_norm": 0.8131458592240662, + "learning_rate": 0.003, + "loss": 4.1601, + "step": 3089 + }, + { + "epoch": 0.0309, + "grad_norm": 0.8512834574988407, + "learning_rate": 0.003, + "loss": 4.1699, + "step": 3090 + }, + { + "epoch": 0.03091, + "grad_norm": 0.8422191636508738, + "learning_rate": 0.003, + "loss": 4.1335, + "step": 3091 + }, + { + "epoch": 0.03092, + "grad_norm": 0.7541794006345667, + "learning_rate": 0.003, + "loss": 4.1613, + "step": 3092 + }, + { + "epoch": 0.03093, + "grad_norm": 0.6725158454989346, + "learning_rate": 0.003, + "loss": 4.1914, + "step": 3093 + }, + { + "epoch": 0.03094, + "grad_norm": 0.7201284858059102, + "learning_rate": 0.003, + "loss": 4.1757, + "step": 3094 + }, + { + "epoch": 0.03095, + "grad_norm": 0.7621210210503538, + "learning_rate": 0.003, + "loss": 4.1919, + "step": 3095 + }, + { + "epoch": 0.03096, + "grad_norm": 0.6578929196198599, + "learning_rate": 0.003, + "loss": 4.1684, + "step": 3096 + }, + { + "epoch": 0.03097, + "grad_norm": 0.695067390761494, + "learning_rate": 0.003, + "loss": 4.1922, + "step": 3097 + }, + { + "epoch": 0.03098, + "grad_norm": 0.740033596710499, + "learning_rate": 0.003, + "loss": 4.1912, + "step": 3098 + }, + { + "epoch": 0.03099, + "grad_norm": 0.7403836940802626, + "learning_rate": 0.003, + "loss": 4.185, + "step": 3099 + }, + { + "epoch": 0.031, + "grad_norm": 0.6979751976411313, + "learning_rate": 0.003, + "loss": 4.179, + "step": 3100 + }, + { + "epoch": 0.03101, + "grad_norm": 0.7670376808263826, + "learning_rate": 0.003, + "loss": 4.1742, + "step": 3101 + }, + { + "epoch": 0.03102, + "grad_norm": 0.8040700995473709, + "learning_rate": 0.003, + "loss": 4.1768, + "step": 3102 + }, + { + "epoch": 0.03103, + "grad_norm": 0.7605055478614147, + "learning_rate": 0.003, + "loss": 4.1649, + "step": 3103 + }, + { + "epoch": 0.03104, + "grad_norm": 0.7299989579866449, + "learning_rate": 0.003, + "loss": 4.1722, + "step": 3104 + }, + { + "epoch": 0.03105, + "grad_norm": 0.8293979417721875, + "learning_rate": 0.003, + "loss": 4.1542, + "step": 3105 + }, + { + "epoch": 0.03106, + "grad_norm": 0.7907107069059506, + "learning_rate": 0.003, + "loss": 4.1755, + "step": 3106 + }, + { + "epoch": 0.03107, + "grad_norm": 0.7368559192868307, + "learning_rate": 0.003, + "loss": 4.1551, + "step": 3107 + }, + { + "epoch": 0.03108, + "grad_norm": 0.6702525929923716, + "learning_rate": 0.003, + "loss": 4.1641, + "step": 3108 + }, + { + "epoch": 0.03109, + "grad_norm": 0.7199266096748919, + "learning_rate": 0.003, + "loss": 4.1721, + "step": 3109 + }, + { + "epoch": 0.0311, + "grad_norm": 0.7526329246324333, + "learning_rate": 0.003, + "loss": 4.1926, + "step": 3110 + }, + { + "epoch": 0.03111, + "grad_norm": 0.7398905284889996, + "learning_rate": 0.003, + "loss": 4.173, + "step": 3111 + }, + { + "epoch": 0.03112, + "grad_norm": 0.8085890464082365, + "learning_rate": 0.003, + "loss": 4.1956, + "step": 3112 + }, + { + "epoch": 0.03113, + "grad_norm": 0.8975031706869222, + "learning_rate": 0.003, + "loss": 4.1852, + "step": 3113 + }, + { + "epoch": 0.03114, + "grad_norm": 0.9143315414333046, + "learning_rate": 0.003, + "loss": 4.1926, + "step": 3114 + }, + { + "epoch": 0.03115, + "grad_norm": 0.7984002320222364, + "learning_rate": 0.003, + "loss": 4.1932, + "step": 3115 + }, + { + "epoch": 0.03116, + "grad_norm": 0.7251259112559764, + "learning_rate": 0.003, + "loss": 4.1766, + "step": 3116 + }, + { + "epoch": 0.03117, + "grad_norm": 0.6765418867772506, + "learning_rate": 0.003, + "loss": 4.1803, + "step": 3117 + }, + { + "epoch": 0.03118, + "grad_norm": 0.6564914593167945, + "learning_rate": 0.003, + "loss": 4.1805, + "step": 3118 + }, + { + "epoch": 0.03119, + "grad_norm": 0.6469214981930856, + "learning_rate": 0.003, + "loss": 4.1951, + "step": 3119 + }, + { + "epoch": 0.0312, + "grad_norm": 0.6004553747682227, + "learning_rate": 0.003, + "loss": 4.1991, + "step": 3120 + }, + { + "epoch": 0.03121, + "grad_norm": 0.5447439216302157, + "learning_rate": 0.003, + "loss": 4.1713, + "step": 3121 + }, + { + "epoch": 0.03122, + "grad_norm": 0.472628289298214, + "learning_rate": 0.003, + "loss": 4.1883, + "step": 3122 + }, + { + "epoch": 0.03123, + "grad_norm": 0.4637134521023551, + "learning_rate": 0.003, + "loss": 4.1843, + "step": 3123 + }, + { + "epoch": 0.03124, + "grad_norm": 0.5031937627853148, + "learning_rate": 0.003, + "loss": 4.1606, + "step": 3124 + }, + { + "epoch": 0.03125, + "grad_norm": 0.49290767185737955, + "learning_rate": 0.003, + "loss": 4.1619, + "step": 3125 + }, + { + "epoch": 0.03126, + "grad_norm": 0.5046421633962421, + "learning_rate": 0.003, + "loss": 4.166, + "step": 3126 + }, + { + "epoch": 0.03127, + "grad_norm": 0.5295145475165267, + "learning_rate": 0.003, + "loss": 4.1725, + "step": 3127 + }, + { + "epoch": 0.03128, + "grad_norm": 0.5768647951144659, + "learning_rate": 0.003, + "loss": 4.1631, + "step": 3128 + }, + { + "epoch": 0.03129, + "grad_norm": 0.5526301834618427, + "learning_rate": 0.003, + "loss": 4.1801, + "step": 3129 + }, + { + "epoch": 0.0313, + "grad_norm": 0.6096664293295351, + "learning_rate": 0.003, + "loss": 4.187, + "step": 3130 + }, + { + "epoch": 0.03131, + "grad_norm": 0.6262440538187416, + "learning_rate": 0.003, + "loss": 4.1696, + "step": 3131 + }, + { + "epoch": 0.03132, + "grad_norm": 0.6399356853738972, + "learning_rate": 0.003, + "loss": 4.1889, + "step": 3132 + }, + { + "epoch": 0.03133, + "grad_norm": 0.6986536472969243, + "learning_rate": 0.003, + "loss": 4.182, + "step": 3133 + }, + { + "epoch": 0.03134, + "grad_norm": 0.8533836264241755, + "learning_rate": 0.003, + "loss": 4.1841, + "step": 3134 + }, + { + "epoch": 0.03135, + "grad_norm": 0.9857590467203682, + "learning_rate": 0.003, + "loss": 4.1591, + "step": 3135 + }, + { + "epoch": 0.03136, + "grad_norm": 0.985916813895558, + "learning_rate": 0.003, + "loss": 4.1745, + "step": 3136 + }, + { + "epoch": 0.03137, + "grad_norm": 0.7996019468603507, + "learning_rate": 0.003, + "loss": 4.177, + "step": 3137 + }, + { + "epoch": 0.03138, + "grad_norm": 0.7409168937664911, + "learning_rate": 0.003, + "loss": 4.1757, + "step": 3138 + }, + { + "epoch": 0.03139, + "grad_norm": 0.6767699843822652, + "learning_rate": 0.003, + "loss": 4.1828, + "step": 3139 + }, + { + "epoch": 0.0314, + "grad_norm": 0.7304920419704491, + "learning_rate": 0.003, + "loss": 4.1836, + "step": 3140 + }, + { + "epoch": 0.03141, + "grad_norm": 0.8085325981037152, + "learning_rate": 0.003, + "loss": 4.2179, + "step": 3141 + }, + { + "epoch": 0.03142, + "grad_norm": 0.7280300882309108, + "learning_rate": 0.003, + "loss": 4.1945, + "step": 3142 + }, + { + "epoch": 0.03143, + "grad_norm": 0.723146974389316, + "learning_rate": 0.003, + "loss": 4.1826, + "step": 3143 + }, + { + "epoch": 0.03144, + "grad_norm": 0.6065206122686737, + "learning_rate": 0.003, + "loss": 4.1484, + "step": 3144 + }, + { + "epoch": 0.03145, + "grad_norm": 0.644934935501266, + "learning_rate": 0.003, + "loss": 4.1764, + "step": 3145 + }, + { + "epoch": 0.03146, + "grad_norm": 0.6592280348022013, + "learning_rate": 0.003, + "loss": 4.1869, + "step": 3146 + }, + { + "epoch": 0.03147, + "grad_norm": 0.6530269899474325, + "learning_rate": 0.003, + "loss": 4.1855, + "step": 3147 + }, + { + "epoch": 0.03148, + "grad_norm": 0.5745843025691852, + "learning_rate": 0.003, + "loss": 4.1852, + "step": 3148 + }, + { + "epoch": 0.03149, + "grad_norm": 0.6065668720311537, + "learning_rate": 0.003, + "loss": 4.1911, + "step": 3149 + }, + { + "epoch": 0.0315, + "grad_norm": 0.6922007846823143, + "learning_rate": 0.003, + "loss": 4.1787, + "step": 3150 + }, + { + "epoch": 0.03151, + "grad_norm": 0.7094328965622133, + "learning_rate": 0.003, + "loss": 4.1911, + "step": 3151 + }, + { + "epoch": 0.03152, + "grad_norm": 0.6960957361399359, + "learning_rate": 0.003, + "loss": 4.1913, + "step": 3152 + }, + { + "epoch": 0.03153, + "grad_norm": 0.7486079877329721, + "learning_rate": 0.003, + "loss": 4.1997, + "step": 3153 + }, + { + "epoch": 0.03154, + "grad_norm": 0.7456103075414062, + "learning_rate": 0.003, + "loss": 4.1959, + "step": 3154 + }, + { + "epoch": 0.03155, + "grad_norm": 0.8177011172560863, + "learning_rate": 0.003, + "loss": 4.1844, + "step": 3155 + }, + { + "epoch": 0.03156, + "grad_norm": 0.8793941608963775, + "learning_rate": 0.003, + "loss": 4.1749, + "step": 3156 + }, + { + "epoch": 0.03157, + "grad_norm": 0.704857810319287, + "learning_rate": 0.003, + "loss": 4.1956, + "step": 3157 + }, + { + "epoch": 0.03158, + "grad_norm": 0.6113967004695787, + "learning_rate": 0.003, + "loss": 4.1799, + "step": 3158 + }, + { + "epoch": 0.03159, + "grad_norm": 0.66211174258146, + "learning_rate": 0.003, + "loss": 4.1801, + "step": 3159 + }, + { + "epoch": 0.0316, + "grad_norm": 0.5523265028957371, + "learning_rate": 0.003, + "loss": 4.1916, + "step": 3160 + }, + { + "epoch": 0.03161, + "grad_norm": 0.608325644395274, + "learning_rate": 0.003, + "loss": 4.1998, + "step": 3161 + }, + { + "epoch": 0.03162, + "grad_norm": 0.6269366928420552, + "learning_rate": 0.003, + "loss": 4.2023, + "step": 3162 + }, + { + "epoch": 0.03163, + "grad_norm": 0.6819002848904914, + "learning_rate": 0.003, + "loss": 4.188, + "step": 3163 + }, + { + "epoch": 0.03164, + "grad_norm": 0.7746466933382178, + "learning_rate": 0.003, + "loss": 4.1862, + "step": 3164 + }, + { + "epoch": 0.03165, + "grad_norm": 0.8342216139344243, + "learning_rate": 0.003, + "loss": 4.1658, + "step": 3165 + }, + { + "epoch": 0.03166, + "grad_norm": 0.854621355905751, + "learning_rate": 0.003, + "loss": 4.1812, + "step": 3166 + }, + { + "epoch": 0.03167, + "grad_norm": 0.9512159018248347, + "learning_rate": 0.003, + "loss": 4.1756, + "step": 3167 + }, + { + "epoch": 0.03168, + "grad_norm": 0.9746218610383118, + "learning_rate": 0.003, + "loss": 4.2065, + "step": 3168 + }, + { + "epoch": 0.03169, + "grad_norm": 0.9858598226166518, + "learning_rate": 0.003, + "loss": 4.224, + "step": 3169 + }, + { + "epoch": 0.0317, + "grad_norm": 0.7498416277122048, + "learning_rate": 0.003, + "loss": 4.1691, + "step": 3170 + }, + { + "epoch": 0.03171, + "grad_norm": 0.6044881164445568, + "learning_rate": 0.003, + "loss": 4.2016, + "step": 3171 + }, + { + "epoch": 0.03172, + "grad_norm": 0.6815636176834087, + "learning_rate": 0.003, + "loss": 4.191, + "step": 3172 + }, + { + "epoch": 0.03173, + "grad_norm": 0.6654525131928357, + "learning_rate": 0.003, + "loss": 4.1647, + "step": 3173 + }, + { + "epoch": 0.03174, + "grad_norm": 0.6474897594742762, + "learning_rate": 0.003, + "loss": 4.1744, + "step": 3174 + }, + { + "epoch": 0.03175, + "grad_norm": 0.6418698031488217, + "learning_rate": 0.003, + "loss": 4.201, + "step": 3175 + }, + { + "epoch": 0.03176, + "grad_norm": 0.6191051450233946, + "learning_rate": 0.003, + "loss": 4.1609, + "step": 3176 + }, + { + "epoch": 0.03177, + "grad_norm": 0.5618627528537171, + "learning_rate": 0.003, + "loss": 4.1676, + "step": 3177 + }, + { + "epoch": 0.03178, + "grad_norm": 0.574624051159771, + "learning_rate": 0.003, + "loss": 4.2133, + "step": 3178 + }, + { + "epoch": 0.03179, + "grad_norm": 0.5170487183975983, + "learning_rate": 0.003, + "loss": 4.1718, + "step": 3179 + }, + { + "epoch": 0.0318, + "grad_norm": 0.5443225922897362, + "learning_rate": 0.003, + "loss": 4.1734, + "step": 3180 + }, + { + "epoch": 0.03181, + "grad_norm": 0.48857658279023347, + "learning_rate": 0.003, + "loss": 4.1797, + "step": 3181 + }, + { + "epoch": 0.03182, + "grad_norm": 0.5003587458379335, + "learning_rate": 0.003, + "loss": 4.1768, + "step": 3182 + }, + { + "epoch": 0.03183, + "grad_norm": 0.47427786201564304, + "learning_rate": 0.003, + "loss": 4.1791, + "step": 3183 + }, + { + "epoch": 0.03184, + "grad_norm": 0.5056283615164667, + "learning_rate": 0.003, + "loss": 4.1534, + "step": 3184 + }, + { + "epoch": 0.03185, + "grad_norm": 0.5400490284012653, + "learning_rate": 0.003, + "loss": 4.1741, + "step": 3185 + }, + { + "epoch": 0.03186, + "grad_norm": 0.6985528632077707, + "learning_rate": 0.003, + "loss": 4.1767, + "step": 3186 + }, + { + "epoch": 0.03187, + "grad_norm": 0.8150902063241722, + "learning_rate": 0.003, + "loss": 4.1448, + "step": 3187 + }, + { + "epoch": 0.03188, + "grad_norm": 0.9325360176643979, + "learning_rate": 0.003, + "loss": 4.1849, + "step": 3188 + }, + { + "epoch": 0.03189, + "grad_norm": 0.8937715656838259, + "learning_rate": 0.003, + "loss": 4.1792, + "step": 3189 + }, + { + "epoch": 0.0319, + "grad_norm": 0.744447678504765, + "learning_rate": 0.003, + "loss": 4.1667, + "step": 3190 + }, + { + "epoch": 0.03191, + "grad_norm": 0.6776373550333729, + "learning_rate": 0.003, + "loss": 4.1971, + "step": 3191 + }, + { + "epoch": 0.03192, + "grad_norm": 0.6440396042092965, + "learning_rate": 0.003, + "loss": 4.1698, + "step": 3192 + }, + { + "epoch": 0.03193, + "grad_norm": 0.6355371339608338, + "learning_rate": 0.003, + "loss": 4.17, + "step": 3193 + }, + { + "epoch": 0.03194, + "grad_norm": 0.6540570284226682, + "learning_rate": 0.003, + "loss": 4.1966, + "step": 3194 + }, + { + "epoch": 0.03195, + "grad_norm": 0.5776431167458079, + "learning_rate": 0.003, + "loss": 4.1621, + "step": 3195 + }, + { + "epoch": 0.03196, + "grad_norm": 0.5567486580718531, + "learning_rate": 0.003, + "loss": 4.1736, + "step": 3196 + }, + { + "epoch": 0.03197, + "grad_norm": 0.6211522115218998, + "learning_rate": 0.003, + "loss": 4.1433, + "step": 3197 + }, + { + "epoch": 0.03198, + "grad_norm": 0.6754236762792072, + "learning_rate": 0.003, + "loss": 4.1783, + "step": 3198 + }, + { + "epoch": 0.03199, + "grad_norm": 0.6909028789552336, + "learning_rate": 0.003, + "loss": 4.1534, + "step": 3199 + }, + { + "epoch": 0.032, + "grad_norm": 0.6851130691205759, + "learning_rate": 0.003, + "loss": 4.1726, + "step": 3200 + }, + { + "epoch": 0.03201, + "grad_norm": 0.8236929229091875, + "learning_rate": 0.003, + "loss": 4.1411, + "step": 3201 + }, + { + "epoch": 0.03202, + "grad_norm": 0.951709446477759, + "learning_rate": 0.003, + "loss": 4.1883, + "step": 3202 + }, + { + "epoch": 0.03203, + "grad_norm": 0.9116506118668389, + "learning_rate": 0.003, + "loss": 4.1713, + "step": 3203 + }, + { + "epoch": 0.03204, + "grad_norm": 0.737364092313657, + "learning_rate": 0.003, + "loss": 4.182, + "step": 3204 + }, + { + "epoch": 0.03205, + "grad_norm": 0.5733762258500861, + "learning_rate": 0.003, + "loss": 4.1751, + "step": 3205 + }, + { + "epoch": 0.03206, + "grad_norm": 0.6520837940224375, + "learning_rate": 0.003, + "loss": 4.1707, + "step": 3206 + }, + { + "epoch": 0.03207, + "grad_norm": 0.6713973023408082, + "learning_rate": 0.003, + "loss": 4.152, + "step": 3207 + }, + { + "epoch": 0.03208, + "grad_norm": 0.6638942407046395, + "learning_rate": 0.003, + "loss": 4.1749, + "step": 3208 + }, + { + "epoch": 0.03209, + "grad_norm": 0.6926100039707507, + "learning_rate": 0.003, + "loss": 4.2122, + "step": 3209 + }, + { + "epoch": 0.0321, + "grad_norm": 0.7197536710037978, + "learning_rate": 0.003, + "loss": 4.1931, + "step": 3210 + }, + { + "epoch": 0.03211, + "grad_norm": 0.7299656936668171, + "learning_rate": 0.003, + "loss": 4.1681, + "step": 3211 + }, + { + "epoch": 0.03212, + "grad_norm": 0.6674134844073308, + "learning_rate": 0.003, + "loss": 4.1736, + "step": 3212 + }, + { + "epoch": 0.03213, + "grad_norm": 0.602090480831783, + "learning_rate": 0.003, + "loss": 4.176, + "step": 3213 + }, + { + "epoch": 0.03214, + "grad_norm": 0.5632210286186823, + "learning_rate": 0.003, + "loss": 4.1402, + "step": 3214 + }, + { + "epoch": 0.03215, + "grad_norm": 0.6424317594884178, + "learning_rate": 0.003, + "loss": 4.1555, + "step": 3215 + }, + { + "epoch": 0.03216, + "grad_norm": 0.6946877458250511, + "learning_rate": 0.003, + "loss": 4.1642, + "step": 3216 + }, + { + "epoch": 0.03217, + "grad_norm": 0.7162006314166186, + "learning_rate": 0.003, + "loss": 4.1667, + "step": 3217 + }, + { + "epoch": 0.03218, + "grad_norm": 0.6688710116110083, + "learning_rate": 0.003, + "loss": 4.1573, + "step": 3218 + }, + { + "epoch": 0.03219, + "grad_norm": 0.6849969993386218, + "learning_rate": 0.003, + "loss": 4.1667, + "step": 3219 + }, + { + "epoch": 0.0322, + "grad_norm": 0.6568426784434412, + "learning_rate": 0.003, + "loss": 4.1774, + "step": 3220 + }, + { + "epoch": 0.03221, + "grad_norm": 0.7253483658822476, + "learning_rate": 0.003, + "loss": 4.1857, + "step": 3221 + }, + { + "epoch": 0.03222, + "grad_norm": 0.793341013671458, + "learning_rate": 0.003, + "loss": 4.183, + "step": 3222 + }, + { + "epoch": 0.03223, + "grad_norm": 0.824497677720901, + "learning_rate": 0.003, + "loss": 4.1862, + "step": 3223 + }, + { + "epoch": 0.03224, + "grad_norm": 0.7552119583336456, + "learning_rate": 0.003, + "loss": 4.1463, + "step": 3224 + }, + { + "epoch": 0.03225, + "grad_norm": 0.5734236212732641, + "learning_rate": 0.003, + "loss": 4.1756, + "step": 3225 + }, + { + "epoch": 0.03226, + "grad_norm": 0.6220562562626234, + "learning_rate": 0.003, + "loss": 4.1558, + "step": 3226 + }, + { + "epoch": 0.03227, + "grad_norm": 0.6820845746631659, + "learning_rate": 0.003, + "loss": 4.1783, + "step": 3227 + }, + { + "epoch": 0.03228, + "grad_norm": 0.5933306652689198, + "learning_rate": 0.003, + "loss": 4.1516, + "step": 3228 + }, + { + "epoch": 0.03229, + "grad_norm": 0.5119085736725962, + "learning_rate": 0.003, + "loss": 4.1681, + "step": 3229 + }, + { + "epoch": 0.0323, + "grad_norm": 0.5310047730371823, + "learning_rate": 0.003, + "loss": 4.1594, + "step": 3230 + }, + { + "epoch": 0.03231, + "grad_norm": 0.444458596160572, + "learning_rate": 0.003, + "loss": 4.1741, + "step": 3231 + }, + { + "epoch": 0.03232, + "grad_norm": 0.48863148670767037, + "learning_rate": 0.003, + "loss": 4.1621, + "step": 3232 + }, + { + "epoch": 0.03233, + "grad_norm": 0.5890179829718989, + "learning_rate": 0.003, + "loss": 4.1823, + "step": 3233 + }, + { + "epoch": 0.03234, + "grad_norm": 0.6507977897435386, + "learning_rate": 0.003, + "loss": 4.1635, + "step": 3234 + }, + { + "epoch": 0.03235, + "grad_norm": 0.6735430538160202, + "learning_rate": 0.003, + "loss": 4.1755, + "step": 3235 + }, + { + "epoch": 0.03236, + "grad_norm": 0.6224436805872925, + "learning_rate": 0.003, + "loss": 4.1638, + "step": 3236 + }, + { + "epoch": 0.03237, + "grad_norm": 0.6153286251420005, + "learning_rate": 0.003, + "loss": 4.1832, + "step": 3237 + }, + { + "epoch": 0.03238, + "grad_norm": 0.5653843948515308, + "learning_rate": 0.003, + "loss": 4.1875, + "step": 3238 + }, + { + "epoch": 0.03239, + "grad_norm": 0.5579474828935078, + "learning_rate": 0.003, + "loss": 4.1662, + "step": 3239 + }, + { + "epoch": 0.0324, + "grad_norm": 0.5628889716808217, + "learning_rate": 0.003, + "loss": 4.1826, + "step": 3240 + }, + { + "epoch": 0.03241, + "grad_norm": 0.618198116761749, + "learning_rate": 0.003, + "loss": 4.1781, + "step": 3241 + }, + { + "epoch": 0.03242, + "grad_norm": 0.7766046798616605, + "learning_rate": 0.003, + "loss": 4.1586, + "step": 3242 + }, + { + "epoch": 0.03243, + "grad_norm": 0.9785268652346966, + "learning_rate": 0.003, + "loss": 4.1547, + "step": 3243 + }, + { + "epoch": 0.03244, + "grad_norm": 0.9580973048940927, + "learning_rate": 0.003, + "loss": 4.1852, + "step": 3244 + }, + { + "epoch": 0.03245, + "grad_norm": 0.8576334375089313, + "learning_rate": 0.003, + "loss": 4.1916, + "step": 3245 + }, + { + "epoch": 0.03246, + "grad_norm": 0.8423468581019681, + "learning_rate": 0.003, + "loss": 4.1821, + "step": 3246 + }, + { + "epoch": 0.03247, + "grad_norm": 0.9544112800192335, + "learning_rate": 0.003, + "loss": 4.1972, + "step": 3247 + }, + { + "epoch": 0.03248, + "grad_norm": 1.108974194257081, + "learning_rate": 0.003, + "loss": 4.2245, + "step": 3248 + }, + { + "epoch": 0.03249, + "grad_norm": 0.922677655020294, + "learning_rate": 0.003, + "loss": 4.1848, + "step": 3249 + }, + { + "epoch": 0.0325, + "grad_norm": 0.87283028555209, + "learning_rate": 0.003, + "loss": 4.2167, + "step": 3250 + }, + { + "epoch": 0.03251, + "grad_norm": 0.9163748774315567, + "learning_rate": 0.003, + "loss": 4.2201, + "step": 3251 + }, + { + "epoch": 0.03252, + "grad_norm": 0.81456143100221, + "learning_rate": 0.003, + "loss": 4.1827, + "step": 3252 + }, + { + "epoch": 0.03253, + "grad_norm": 0.7968908258913222, + "learning_rate": 0.003, + "loss": 4.1861, + "step": 3253 + }, + { + "epoch": 0.03254, + "grad_norm": 0.7235050634300025, + "learning_rate": 0.003, + "loss": 4.1768, + "step": 3254 + }, + { + "epoch": 0.03255, + "grad_norm": 0.6266440712092289, + "learning_rate": 0.003, + "loss": 4.1867, + "step": 3255 + }, + { + "epoch": 0.03256, + "grad_norm": 0.6449674375686789, + "learning_rate": 0.003, + "loss": 4.1852, + "step": 3256 + }, + { + "epoch": 0.03257, + "grad_norm": 0.5622485206306796, + "learning_rate": 0.003, + "loss": 4.1703, + "step": 3257 + }, + { + "epoch": 0.03258, + "grad_norm": 0.5825851281947257, + "learning_rate": 0.003, + "loss": 4.1799, + "step": 3258 + }, + { + "epoch": 0.03259, + "grad_norm": 0.6444343388033511, + "learning_rate": 0.003, + "loss": 4.1826, + "step": 3259 + }, + { + "epoch": 0.0326, + "grad_norm": 0.6989193151493163, + "learning_rate": 0.003, + "loss": 4.1815, + "step": 3260 + }, + { + "epoch": 0.03261, + "grad_norm": 0.8302855362092207, + "learning_rate": 0.003, + "loss": 4.1597, + "step": 3261 + }, + { + "epoch": 0.03262, + "grad_norm": 0.9498523114829631, + "learning_rate": 0.003, + "loss": 4.1887, + "step": 3262 + }, + { + "epoch": 0.03263, + "grad_norm": 0.983071934935445, + "learning_rate": 0.003, + "loss": 4.2007, + "step": 3263 + }, + { + "epoch": 0.03264, + "grad_norm": 0.8331337315585593, + "learning_rate": 0.003, + "loss": 4.208, + "step": 3264 + }, + { + "epoch": 0.03265, + "grad_norm": 0.7498656994034634, + "learning_rate": 0.003, + "loss": 4.2003, + "step": 3265 + }, + { + "epoch": 0.03266, + "grad_norm": 0.7807040232610372, + "learning_rate": 0.003, + "loss": 4.1549, + "step": 3266 + }, + { + "epoch": 0.03267, + "grad_norm": 0.8713471135209633, + "learning_rate": 0.003, + "loss": 4.1715, + "step": 3267 + }, + { + "epoch": 0.03268, + "grad_norm": 0.8995394972894192, + "learning_rate": 0.003, + "loss": 4.1612, + "step": 3268 + }, + { + "epoch": 0.03269, + "grad_norm": 0.7799010442197561, + "learning_rate": 0.003, + "loss": 4.1875, + "step": 3269 + }, + { + "epoch": 0.0327, + "grad_norm": 0.6333137664038297, + "learning_rate": 0.003, + "loss": 4.2075, + "step": 3270 + }, + { + "epoch": 0.03271, + "grad_norm": 0.5431600237872565, + "learning_rate": 0.003, + "loss": 4.1673, + "step": 3271 + }, + { + "epoch": 0.03272, + "grad_norm": 0.5431026489790501, + "learning_rate": 0.003, + "loss": 4.1751, + "step": 3272 + }, + { + "epoch": 0.03273, + "grad_norm": 0.45854609240948824, + "learning_rate": 0.003, + "loss": 4.1729, + "step": 3273 + }, + { + "epoch": 0.03274, + "grad_norm": 0.44688508931158466, + "learning_rate": 0.003, + "loss": 4.1868, + "step": 3274 + }, + { + "epoch": 0.03275, + "grad_norm": 0.45373894096093864, + "learning_rate": 0.003, + "loss": 4.1852, + "step": 3275 + }, + { + "epoch": 0.03276, + "grad_norm": 0.41006589936140064, + "learning_rate": 0.003, + "loss": 4.1613, + "step": 3276 + }, + { + "epoch": 0.03277, + "grad_norm": 0.38524925786232783, + "learning_rate": 0.003, + "loss": 4.1604, + "step": 3277 + }, + { + "epoch": 0.03278, + "grad_norm": 0.35950981826812556, + "learning_rate": 0.003, + "loss": 4.1675, + "step": 3278 + }, + { + "epoch": 0.03279, + "grad_norm": 0.38292545999031286, + "learning_rate": 0.003, + "loss": 4.155, + "step": 3279 + }, + { + "epoch": 0.0328, + "grad_norm": 0.4242164622322683, + "learning_rate": 0.003, + "loss": 4.1405, + "step": 3280 + }, + { + "epoch": 0.03281, + "grad_norm": 0.488691641648266, + "learning_rate": 0.003, + "loss": 4.1888, + "step": 3281 + }, + { + "epoch": 0.03282, + "grad_norm": 0.5739177249611346, + "learning_rate": 0.003, + "loss": 4.1447, + "step": 3282 + }, + { + "epoch": 0.03283, + "grad_norm": 0.554048169102402, + "learning_rate": 0.003, + "loss": 4.1403, + "step": 3283 + }, + { + "epoch": 0.03284, + "grad_norm": 0.4915028252102749, + "learning_rate": 0.003, + "loss": 4.1816, + "step": 3284 + }, + { + "epoch": 0.03285, + "grad_norm": 0.5580359822634506, + "learning_rate": 0.003, + "loss": 4.1293, + "step": 3285 + }, + { + "epoch": 0.03286, + "grad_norm": 0.6079781410505047, + "learning_rate": 0.003, + "loss": 4.1723, + "step": 3286 + }, + { + "epoch": 0.03287, + "grad_norm": 0.5821835778978477, + "learning_rate": 0.003, + "loss": 4.176, + "step": 3287 + }, + { + "epoch": 0.03288, + "grad_norm": 0.5438338131339984, + "learning_rate": 0.003, + "loss": 4.1565, + "step": 3288 + }, + { + "epoch": 0.03289, + "grad_norm": 0.570468604964685, + "learning_rate": 0.003, + "loss": 4.1609, + "step": 3289 + }, + { + "epoch": 0.0329, + "grad_norm": 0.6096925395272201, + "learning_rate": 0.003, + "loss": 4.1931, + "step": 3290 + }, + { + "epoch": 0.03291, + "grad_norm": 0.7214109941856076, + "learning_rate": 0.003, + "loss": 4.1807, + "step": 3291 + }, + { + "epoch": 0.03292, + "grad_norm": 0.9305508295977749, + "learning_rate": 0.003, + "loss": 4.1568, + "step": 3292 + }, + { + "epoch": 0.03293, + "grad_norm": 0.9205488988397595, + "learning_rate": 0.003, + "loss": 4.1782, + "step": 3293 + }, + { + "epoch": 0.03294, + "grad_norm": 0.8983261374723139, + "learning_rate": 0.003, + "loss": 4.1833, + "step": 3294 + }, + { + "epoch": 0.03295, + "grad_norm": 0.9704372414132879, + "learning_rate": 0.003, + "loss": 4.1739, + "step": 3295 + }, + { + "epoch": 0.03296, + "grad_norm": 1.4039738619774527, + "learning_rate": 0.003, + "loss": 4.1989, + "step": 3296 + }, + { + "epoch": 0.03297, + "grad_norm": 0.8478482918673438, + "learning_rate": 0.003, + "loss": 4.188, + "step": 3297 + }, + { + "epoch": 0.03298, + "grad_norm": 0.8625233369451641, + "learning_rate": 0.003, + "loss": 4.2085, + "step": 3298 + }, + { + "epoch": 0.03299, + "grad_norm": 0.946434504765274, + "learning_rate": 0.003, + "loss": 4.201, + "step": 3299 + }, + { + "epoch": 0.033, + "grad_norm": 0.9408833966323229, + "learning_rate": 0.003, + "loss": 4.1893, + "step": 3300 + }, + { + "epoch": 0.03301, + "grad_norm": 1.0399735000004064, + "learning_rate": 0.003, + "loss": 4.2331, + "step": 3301 + }, + { + "epoch": 0.03302, + "grad_norm": 1.060889182219052, + "learning_rate": 0.003, + "loss": 4.2111, + "step": 3302 + }, + { + "epoch": 0.03303, + "grad_norm": 0.910480011898513, + "learning_rate": 0.003, + "loss": 4.2117, + "step": 3303 + }, + { + "epoch": 0.03304, + "grad_norm": 0.9051761770071215, + "learning_rate": 0.003, + "loss": 4.2061, + "step": 3304 + }, + { + "epoch": 0.03305, + "grad_norm": 0.8593496427676475, + "learning_rate": 0.003, + "loss": 4.2046, + "step": 3305 + }, + { + "epoch": 0.03306, + "grad_norm": 0.788406787032699, + "learning_rate": 0.003, + "loss": 4.2208, + "step": 3306 + }, + { + "epoch": 0.03307, + "grad_norm": 0.6264143830894844, + "learning_rate": 0.003, + "loss": 4.2132, + "step": 3307 + }, + { + "epoch": 0.03308, + "grad_norm": 0.5647145644735359, + "learning_rate": 0.003, + "loss": 4.1897, + "step": 3308 + }, + { + "epoch": 0.03309, + "grad_norm": 0.5464103710288922, + "learning_rate": 0.003, + "loss": 4.1758, + "step": 3309 + }, + { + "epoch": 0.0331, + "grad_norm": 0.4958683830756014, + "learning_rate": 0.003, + "loss": 4.1767, + "step": 3310 + }, + { + "epoch": 0.03311, + "grad_norm": 0.4522879146976332, + "learning_rate": 0.003, + "loss": 4.1941, + "step": 3311 + }, + { + "epoch": 0.03312, + "grad_norm": 0.4414886343825775, + "learning_rate": 0.003, + "loss": 4.1614, + "step": 3312 + }, + { + "epoch": 0.03313, + "grad_norm": 0.4255185332999733, + "learning_rate": 0.003, + "loss": 4.1887, + "step": 3313 + }, + { + "epoch": 0.03314, + "grad_norm": 0.4381607158366914, + "learning_rate": 0.003, + "loss": 4.1299, + "step": 3314 + }, + { + "epoch": 0.03315, + "grad_norm": 0.42262937541001916, + "learning_rate": 0.003, + "loss": 4.1816, + "step": 3315 + }, + { + "epoch": 0.03316, + "grad_norm": 0.4970426554871505, + "learning_rate": 0.003, + "loss": 4.1826, + "step": 3316 + }, + { + "epoch": 0.03317, + "grad_norm": 0.5627776265011807, + "learning_rate": 0.003, + "loss": 4.1717, + "step": 3317 + }, + { + "epoch": 0.03318, + "grad_norm": 0.6893835679497516, + "learning_rate": 0.003, + "loss": 4.1564, + "step": 3318 + }, + { + "epoch": 0.03319, + "grad_norm": 0.7468347673987352, + "learning_rate": 0.003, + "loss": 4.1531, + "step": 3319 + }, + { + "epoch": 0.0332, + "grad_norm": 0.8688918354050132, + "learning_rate": 0.003, + "loss": 4.1819, + "step": 3320 + }, + { + "epoch": 0.03321, + "grad_norm": 0.9441155117884158, + "learning_rate": 0.003, + "loss": 4.2001, + "step": 3321 + }, + { + "epoch": 0.03322, + "grad_norm": 0.8347403031059345, + "learning_rate": 0.003, + "loss": 4.1519, + "step": 3322 + }, + { + "epoch": 0.03323, + "grad_norm": 0.7804052129413228, + "learning_rate": 0.003, + "loss": 4.1955, + "step": 3323 + }, + { + "epoch": 0.03324, + "grad_norm": 0.7276937121862084, + "learning_rate": 0.003, + "loss": 4.1636, + "step": 3324 + }, + { + "epoch": 0.03325, + "grad_norm": 0.7364714952754682, + "learning_rate": 0.003, + "loss": 4.1961, + "step": 3325 + }, + { + "epoch": 0.03326, + "grad_norm": 0.6138475460467973, + "learning_rate": 0.003, + "loss": 4.1513, + "step": 3326 + }, + { + "epoch": 0.03327, + "grad_norm": 0.5484242765422025, + "learning_rate": 0.003, + "loss": 4.1774, + "step": 3327 + }, + { + "epoch": 0.03328, + "grad_norm": 0.5021334597809965, + "learning_rate": 0.003, + "loss": 4.1949, + "step": 3328 + }, + { + "epoch": 0.03329, + "grad_norm": 0.531052392624268, + "learning_rate": 0.003, + "loss": 4.1523, + "step": 3329 + }, + { + "epoch": 0.0333, + "grad_norm": 0.5172636277516343, + "learning_rate": 0.003, + "loss": 4.1562, + "step": 3330 + }, + { + "epoch": 0.03331, + "grad_norm": 0.5347276948931869, + "learning_rate": 0.003, + "loss": 4.1547, + "step": 3331 + }, + { + "epoch": 0.03332, + "grad_norm": 0.5266769298617034, + "learning_rate": 0.003, + "loss": 4.1514, + "step": 3332 + }, + { + "epoch": 0.03333, + "grad_norm": 0.5119624308137339, + "learning_rate": 0.003, + "loss": 4.1366, + "step": 3333 + }, + { + "epoch": 0.03334, + "grad_norm": 0.5227592678561254, + "learning_rate": 0.003, + "loss": 4.1672, + "step": 3334 + }, + { + "epoch": 0.03335, + "grad_norm": 0.5475154089666096, + "learning_rate": 0.003, + "loss": 4.1732, + "step": 3335 + }, + { + "epoch": 0.03336, + "grad_norm": 0.6301047632403133, + "learning_rate": 0.003, + "loss": 4.1763, + "step": 3336 + }, + { + "epoch": 0.03337, + "grad_norm": 0.8194265063356702, + "learning_rate": 0.003, + "loss": 4.1621, + "step": 3337 + }, + { + "epoch": 0.03338, + "grad_norm": 1.1445574176004276, + "learning_rate": 0.003, + "loss": 4.1827, + "step": 3338 + }, + { + "epoch": 0.03339, + "grad_norm": 0.8189097928363258, + "learning_rate": 0.003, + "loss": 4.1787, + "step": 3339 + }, + { + "epoch": 0.0334, + "grad_norm": 0.660449401420429, + "learning_rate": 0.003, + "loss": 4.1521, + "step": 3340 + }, + { + "epoch": 0.03341, + "grad_norm": 0.7754430989891157, + "learning_rate": 0.003, + "loss": 4.1764, + "step": 3341 + }, + { + "epoch": 0.03342, + "grad_norm": 0.8792442379704091, + "learning_rate": 0.003, + "loss": 4.1663, + "step": 3342 + }, + { + "epoch": 0.03343, + "grad_norm": 0.8167670859904461, + "learning_rate": 0.003, + "loss": 4.1402, + "step": 3343 + }, + { + "epoch": 0.03344, + "grad_norm": 0.7480631734050127, + "learning_rate": 0.003, + "loss": 4.1936, + "step": 3344 + }, + { + "epoch": 0.03345, + "grad_norm": 0.6496750703892737, + "learning_rate": 0.003, + "loss": 4.1695, + "step": 3345 + }, + { + "epoch": 0.03346, + "grad_norm": 0.6081719260800688, + "learning_rate": 0.003, + "loss": 4.1825, + "step": 3346 + }, + { + "epoch": 0.03347, + "grad_norm": 0.6721063598583025, + "learning_rate": 0.003, + "loss": 4.168, + "step": 3347 + }, + { + "epoch": 0.03348, + "grad_norm": 0.6283159116392718, + "learning_rate": 0.003, + "loss": 4.1561, + "step": 3348 + }, + { + "epoch": 0.03349, + "grad_norm": 0.5798454168131916, + "learning_rate": 0.003, + "loss": 4.2006, + "step": 3349 + }, + { + "epoch": 0.0335, + "grad_norm": 0.588557434396112, + "learning_rate": 0.003, + "loss": 4.1558, + "step": 3350 + }, + { + "epoch": 0.03351, + "grad_norm": 0.572435380329449, + "learning_rate": 0.003, + "loss": 4.1639, + "step": 3351 + }, + { + "epoch": 0.03352, + "grad_norm": 0.5946129998027531, + "learning_rate": 0.003, + "loss": 4.1756, + "step": 3352 + }, + { + "epoch": 0.03353, + "grad_norm": 0.6650155372083263, + "learning_rate": 0.003, + "loss": 4.1783, + "step": 3353 + }, + { + "epoch": 0.03354, + "grad_norm": 0.6819620362483191, + "learning_rate": 0.003, + "loss": 4.1426, + "step": 3354 + }, + { + "epoch": 0.03355, + "grad_norm": 0.7239397479894265, + "learning_rate": 0.003, + "loss": 4.1701, + "step": 3355 + }, + { + "epoch": 0.03356, + "grad_norm": 0.7696928848130231, + "learning_rate": 0.003, + "loss": 4.197, + "step": 3356 + }, + { + "epoch": 0.03357, + "grad_norm": 0.6661695560767795, + "learning_rate": 0.003, + "loss": 4.1658, + "step": 3357 + }, + { + "epoch": 0.03358, + "grad_norm": 0.6123082718814552, + "learning_rate": 0.003, + "loss": 4.1823, + "step": 3358 + }, + { + "epoch": 0.03359, + "grad_norm": 0.770393140263048, + "learning_rate": 0.003, + "loss": 4.1771, + "step": 3359 + }, + { + "epoch": 0.0336, + "grad_norm": 0.8303005904862942, + "learning_rate": 0.003, + "loss": 4.1911, + "step": 3360 + }, + { + "epoch": 0.03361, + "grad_norm": 0.9034060736336073, + "learning_rate": 0.003, + "loss": 4.1953, + "step": 3361 + }, + { + "epoch": 0.03362, + "grad_norm": 0.8488627855760285, + "learning_rate": 0.003, + "loss": 4.1936, + "step": 3362 + }, + { + "epoch": 0.03363, + "grad_norm": 0.8500592898597722, + "learning_rate": 0.003, + "loss": 4.1482, + "step": 3363 + }, + { + "epoch": 0.03364, + "grad_norm": 0.8053495290807956, + "learning_rate": 0.003, + "loss": 4.167, + "step": 3364 + }, + { + "epoch": 0.03365, + "grad_norm": 0.6918565332600101, + "learning_rate": 0.003, + "loss": 4.1581, + "step": 3365 + }, + { + "epoch": 0.03366, + "grad_norm": 0.6631827092506345, + "learning_rate": 0.003, + "loss": 4.1325, + "step": 3366 + }, + { + "epoch": 0.03367, + "grad_norm": 0.5898333237323294, + "learning_rate": 0.003, + "loss": 4.1467, + "step": 3367 + }, + { + "epoch": 0.03368, + "grad_norm": 0.5105692997315376, + "learning_rate": 0.003, + "loss": 4.1944, + "step": 3368 + }, + { + "epoch": 0.03369, + "grad_norm": 0.4833179204197231, + "learning_rate": 0.003, + "loss": 4.1504, + "step": 3369 + }, + { + "epoch": 0.0337, + "grad_norm": 0.4434936266579207, + "learning_rate": 0.003, + "loss": 4.1827, + "step": 3370 + }, + { + "epoch": 0.03371, + "grad_norm": 0.43752326668000696, + "learning_rate": 0.003, + "loss": 4.1182, + "step": 3371 + }, + { + "epoch": 0.03372, + "grad_norm": 0.5212167164579171, + "learning_rate": 0.003, + "loss": 4.1472, + "step": 3372 + }, + { + "epoch": 0.03373, + "grad_norm": 0.5782660861068077, + "learning_rate": 0.003, + "loss": 4.1812, + "step": 3373 + }, + { + "epoch": 0.03374, + "grad_norm": 0.6248054334937597, + "learning_rate": 0.003, + "loss": 4.1659, + "step": 3374 + }, + { + "epoch": 0.03375, + "grad_norm": 0.5656089592222998, + "learning_rate": 0.003, + "loss": 4.1486, + "step": 3375 + }, + { + "epoch": 0.03376, + "grad_norm": 0.533250430480493, + "learning_rate": 0.003, + "loss": 4.1263, + "step": 3376 + }, + { + "epoch": 0.03377, + "grad_norm": 0.6893576110724466, + "learning_rate": 0.003, + "loss": 4.1726, + "step": 3377 + }, + { + "epoch": 0.03378, + "grad_norm": 0.8724259510840375, + "learning_rate": 0.003, + "loss": 4.1825, + "step": 3378 + }, + { + "epoch": 0.03379, + "grad_norm": 1.0969642859125113, + "learning_rate": 0.003, + "loss": 4.1661, + "step": 3379 + }, + { + "epoch": 0.0338, + "grad_norm": 0.9222105012112505, + "learning_rate": 0.003, + "loss": 4.1929, + "step": 3380 + }, + { + "epoch": 0.03381, + "grad_norm": 0.6636117483340874, + "learning_rate": 0.003, + "loss": 4.1858, + "step": 3381 + }, + { + "epoch": 0.03382, + "grad_norm": 0.6448818843218406, + "learning_rate": 0.003, + "loss": 4.1756, + "step": 3382 + }, + { + "epoch": 0.03383, + "grad_norm": 0.7558871316698539, + "learning_rate": 0.003, + "loss": 4.1454, + "step": 3383 + }, + { + "epoch": 0.03384, + "grad_norm": 0.784895312607238, + "learning_rate": 0.003, + "loss": 4.17, + "step": 3384 + }, + { + "epoch": 0.03385, + "grad_norm": 0.7068746569705847, + "learning_rate": 0.003, + "loss": 4.1685, + "step": 3385 + }, + { + "epoch": 0.03386, + "grad_norm": 0.5763966935533144, + "learning_rate": 0.003, + "loss": 4.1617, + "step": 3386 + }, + { + "epoch": 0.03387, + "grad_norm": 0.5637260656795919, + "learning_rate": 0.003, + "loss": 4.1642, + "step": 3387 + }, + { + "epoch": 0.03388, + "grad_norm": 0.5764694412310111, + "learning_rate": 0.003, + "loss": 4.1757, + "step": 3388 + }, + { + "epoch": 0.03389, + "grad_norm": 0.5063857748030218, + "learning_rate": 0.003, + "loss": 4.1917, + "step": 3389 + }, + { + "epoch": 0.0339, + "grad_norm": 0.48268849043201856, + "learning_rate": 0.003, + "loss": 4.1773, + "step": 3390 + }, + { + "epoch": 0.03391, + "grad_norm": 0.46369864578553305, + "learning_rate": 0.003, + "loss": 4.1547, + "step": 3391 + }, + { + "epoch": 0.03392, + "grad_norm": 0.44960224345626854, + "learning_rate": 0.003, + "loss": 4.1611, + "step": 3392 + }, + { + "epoch": 0.03393, + "grad_norm": 0.4909883361811901, + "learning_rate": 0.003, + "loss": 4.1707, + "step": 3393 + }, + { + "epoch": 0.03394, + "grad_norm": 0.6259272860691563, + "learning_rate": 0.003, + "loss": 4.1742, + "step": 3394 + }, + { + "epoch": 0.03395, + "grad_norm": 0.7506927184483131, + "learning_rate": 0.003, + "loss": 4.1835, + "step": 3395 + }, + { + "epoch": 0.03396, + "grad_norm": 0.9370743209078207, + "learning_rate": 0.003, + "loss": 4.1713, + "step": 3396 + }, + { + "epoch": 0.03397, + "grad_norm": 1.1783116967146359, + "learning_rate": 0.003, + "loss": 4.1545, + "step": 3397 + }, + { + "epoch": 0.03398, + "grad_norm": 0.8089632724306486, + "learning_rate": 0.003, + "loss": 4.1941, + "step": 3398 + }, + { + "epoch": 0.03399, + "grad_norm": 0.7596909526208954, + "learning_rate": 0.003, + "loss": 4.2163, + "step": 3399 + }, + { + "epoch": 0.034, + "grad_norm": 0.8562517296822207, + "learning_rate": 0.003, + "loss": 4.2009, + "step": 3400 + }, + { + "epoch": 0.03401, + "grad_norm": 1.0068688368658953, + "learning_rate": 0.003, + "loss": 4.188, + "step": 3401 + }, + { + "epoch": 0.03402, + "grad_norm": 0.99343979847827, + "learning_rate": 0.003, + "loss": 4.2136, + "step": 3402 + }, + { + "epoch": 0.03403, + "grad_norm": 1.1467610890884863, + "learning_rate": 0.003, + "loss": 4.1901, + "step": 3403 + }, + { + "epoch": 0.03404, + "grad_norm": 1.0385468087744694, + "learning_rate": 0.003, + "loss": 4.2248, + "step": 3404 + }, + { + "epoch": 0.03405, + "grad_norm": 0.9754588896396058, + "learning_rate": 0.003, + "loss": 4.2094, + "step": 3405 + }, + { + "epoch": 0.03406, + "grad_norm": 1.0318386962627828, + "learning_rate": 0.003, + "loss": 4.193, + "step": 3406 + }, + { + "epoch": 0.03407, + "grad_norm": 1.0407604591460764, + "learning_rate": 0.003, + "loss": 4.2114, + "step": 3407 + }, + { + "epoch": 0.03408, + "grad_norm": 0.9991368525042554, + "learning_rate": 0.003, + "loss": 4.1916, + "step": 3408 + }, + { + "epoch": 0.03409, + "grad_norm": 0.9201374434882152, + "learning_rate": 0.003, + "loss": 4.2017, + "step": 3409 + }, + { + "epoch": 0.0341, + "grad_norm": 0.7897132061669699, + "learning_rate": 0.003, + "loss": 4.1919, + "step": 3410 + }, + { + "epoch": 0.03411, + "grad_norm": 0.8111203199333381, + "learning_rate": 0.003, + "loss": 4.2024, + "step": 3411 + }, + { + "epoch": 0.03412, + "grad_norm": 0.7752208914216027, + "learning_rate": 0.003, + "loss": 4.219, + "step": 3412 + }, + { + "epoch": 0.03413, + "grad_norm": 0.8018027400540946, + "learning_rate": 0.003, + "loss": 4.2281, + "step": 3413 + }, + { + "epoch": 0.03414, + "grad_norm": 0.7248771456875713, + "learning_rate": 0.003, + "loss": 4.2139, + "step": 3414 + }, + { + "epoch": 0.03415, + "grad_norm": 0.6901233244425752, + "learning_rate": 0.003, + "loss": 4.1927, + "step": 3415 + }, + { + "epoch": 0.03416, + "grad_norm": 0.6847318566109114, + "learning_rate": 0.003, + "loss": 4.1943, + "step": 3416 + }, + { + "epoch": 0.03417, + "grad_norm": 0.6799194082415111, + "learning_rate": 0.003, + "loss": 4.1883, + "step": 3417 + }, + { + "epoch": 0.03418, + "grad_norm": 0.6775692861547594, + "learning_rate": 0.003, + "loss": 4.1576, + "step": 3418 + }, + { + "epoch": 0.03419, + "grad_norm": 0.6946056576603092, + "learning_rate": 0.003, + "loss": 4.1802, + "step": 3419 + }, + { + "epoch": 0.0342, + "grad_norm": 0.7015136907830581, + "learning_rate": 0.003, + "loss": 4.1882, + "step": 3420 + }, + { + "epoch": 0.03421, + "grad_norm": 0.8322618875343657, + "learning_rate": 0.003, + "loss": 4.184, + "step": 3421 + }, + { + "epoch": 0.03422, + "grad_norm": 0.877410615683893, + "learning_rate": 0.003, + "loss": 4.2019, + "step": 3422 + }, + { + "epoch": 0.03423, + "grad_norm": 0.7815901156339347, + "learning_rate": 0.003, + "loss": 4.18, + "step": 3423 + }, + { + "epoch": 0.03424, + "grad_norm": 0.7461859372495792, + "learning_rate": 0.003, + "loss": 4.2038, + "step": 3424 + }, + { + "epoch": 0.03425, + "grad_norm": 0.6612154502175626, + "learning_rate": 0.003, + "loss": 4.1654, + "step": 3425 + }, + { + "epoch": 0.03426, + "grad_norm": 0.5350775915801592, + "learning_rate": 0.003, + "loss": 4.182, + "step": 3426 + }, + { + "epoch": 0.03427, + "grad_norm": 0.5758126205992435, + "learning_rate": 0.003, + "loss": 4.1525, + "step": 3427 + }, + { + "epoch": 0.03428, + "grad_norm": 0.57747553400605, + "learning_rate": 0.003, + "loss": 4.151, + "step": 3428 + }, + { + "epoch": 0.03429, + "grad_norm": 0.6495727211029925, + "learning_rate": 0.003, + "loss": 4.1682, + "step": 3429 + }, + { + "epoch": 0.0343, + "grad_norm": 0.6970333948744754, + "learning_rate": 0.003, + "loss": 4.1638, + "step": 3430 + }, + { + "epoch": 0.03431, + "grad_norm": 0.7157526291407539, + "learning_rate": 0.003, + "loss": 4.1584, + "step": 3431 + }, + { + "epoch": 0.03432, + "grad_norm": 0.6010910249067187, + "learning_rate": 0.003, + "loss": 4.163, + "step": 3432 + }, + { + "epoch": 0.03433, + "grad_norm": 0.42007320422820055, + "learning_rate": 0.003, + "loss": 4.13, + "step": 3433 + }, + { + "epoch": 0.03434, + "grad_norm": 0.47622504855038245, + "learning_rate": 0.003, + "loss": 4.153, + "step": 3434 + }, + { + "epoch": 0.03435, + "grad_norm": 0.5574984023909559, + "learning_rate": 0.003, + "loss": 4.1379, + "step": 3435 + }, + { + "epoch": 0.03436, + "grad_norm": 0.6117380821653216, + "learning_rate": 0.003, + "loss": 4.1654, + "step": 3436 + }, + { + "epoch": 0.03437, + "grad_norm": 0.5586731673929375, + "learning_rate": 0.003, + "loss": 4.1549, + "step": 3437 + }, + { + "epoch": 0.03438, + "grad_norm": 0.41091865924021914, + "learning_rate": 0.003, + "loss": 4.1663, + "step": 3438 + }, + { + "epoch": 0.03439, + "grad_norm": 0.3844506479299855, + "learning_rate": 0.003, + "loss": 4.1632, + "step": 3439 + }, + { + "epoch": 0.0344, + "grad_norm": 0.47026248688155026, + "learning_rate": 0.003, + "loss": 4.1395, + "step": 3440 + }, + { + "epoch": 0.03441, + "grad_norm": 0.49960141515204526, + "learning_rate": 0.003, + "loss": 4.1602, + "step": 3441 + }, + { + "epoch": 0.03442, + "grad_norm": 0.48160504106231705, + "learning_rate": 0.003, + "loss": 4.1263, + "step": 3442 + }, + { + "epoch": 0.03443, + "grad_norm": 0.4503151278381613, + "learning_rate": 0.003, + "loss": 4.1555, + "step": 3443 + }, + { + "epoch": 0.03444, + "grad_norm": 0.4618551136735495, + "learning_rate": 0.003, + "loss": 4.1663, + "step": 3444 + }, + { + "epoch": 0.03445, + "grad_norm": 0.5002612465674857, + "learning_rate": 0.003, + "loss": 4.1323, + "step": 3445 + }, + { + "epoch": 0.03446, + "grad_norm": 0.49371947051027193, + "learning_rate": 0.003, + "loss": 4.1213, + "step": 3446 + }, + { + "epoch": 0.03447, + "grad_norm": 0.48167624788075264, + "learning_rate": 0.003, + "loss": 4.1697, + "step": 3447 + }, + { + "epoch": 0.03448, + "grad_norm": 0.48645338009833317, + "learning_rate": 0.003, + "loss": 4.1398, + "step": 3448 + }, + { + "epoch": 0.03449, + "grad_norm": 0.42185936569719606, + "learning_rate": 0.003, + "loss": 4.1255, + "step": 3449 + }, + { + "epoch": 0.0345, + "grad_norm": 0.39079332947086687, + "learning_rate": 0.003, + "loss": 4.1609, + "step": 3450 + }, + { + "epoch": 0.03451, + "grad_norm": 0.42129819422128534, + "learning_rate": 0.003, + "loss": 4.1342, + "step": 3451 + }, + { + "epoch": 0.03452, + "grad_norm": 0.5355871931302184, + "learning_rate": 0.003, + "loss": 4.1655, + "step": 3452 + }, + { + "epoch": 0.03453, + "grad_norm": 0.7320917006665237, + "learning_rate": 0.003, + "loss": 4.1474, + "step": 3453 + }, + { + "epoch": 0.03454, + "grad_norm": 1.1401386064666243, + "learning_rate": 0.003, + "loss": 4.1442, + "step": 3454 + }, + { + "epoch": 0.03455, + "grad_norm": 1.05569726012347, + "learning_rate": 0.003, + "loss": 4.1531, + "step": 3455 + }, + { + "epoch": 0.03456, + "grad_norm": 0.7753833366744568, + "learning_rate": 0.003, + "loss": 4.163, + "step": 3456 + }, + { + "epoch": 0.03457, + "grad_norm": 0.8398952536178134, + "learning_rate": 0.003, + "loss": 4.1307, + "step": 3457 + }, + { + "epoch": 0.03458, + "grad_norm": 0.9216186608796643, + "learning_rate": 0.003, + "loss": 4.1748, + "step": 3458 + }, + { + "epoch": 0.03459, + "grad_norm": 0.8741651793984747, + "learning_rate": 0.003, + "loss": 4.1982, + "step": 3459 + }, + { + "epoch": 0.0346, + "grad_norm": 0.7072954882447885, + "learning_rate": 0.003, + "loss": 4.1995, + "step": 3460 + }, + { + "epoch": 0.03461, + "grad_norm": 0.6974866321122205, + "learning_rate": 0.003, + "loss": 4.1749, + "step": 3461 + }, + { + "epoch": 0.03462, + "grad_norm": 0.6828331664556785, + "learning_rate": 0.003, + "loss": 4.1676, + "step": 3462 + }, + { + "epoch": 0.03463, + "grad_norm": 0.8097332341464493, + "learning_rate": 0.003, + "loss": 4.1557, + "step": 3463 + }, + { + "epoch": 0.03464, + "grad_norm": 0.8057895102942476, + "learning_rate": 0.003, + "loss": 4.1731, + "step": 3464 + }, + { + "epoch": 0.03465, + "grad_norm": 0.754598199709423, + "learning_rate": 0.003, + "loss": 4.1797, + "step": 3465 + }, + { + "epoch": 0.03466, + "grad_norm": 0.6771478949940347, + "learning_rate": 0.003, + "loss": 4.1778, + "step": 3466 + }, + { + "epoch": 0.03467, + "grad_norm": 0.6186375156162432, + "learning_rate": 0.003, + "loss": 4.1574, + "step": 3467 + }, + { + "epoch": 0.03468, + "grad_norm": 0.6769160235929496, + "learning_rate": 0.003, + "loss": 4.1742, + "step": 3468 + }, + { + "epoch": 0.03469, + "grad_norm": 0.6809204256661774, + "learning_rate": 0.003, + "loss": 4.1707, + "step": 3469 + }, + { + "epoch": 0.0347, + "grad_norm": 0.7756647140071914, + "learning_rate": 0.003, + "loss": 4.186, + "step": 3470 + }, + { + "epoch": 0.03471, + "grad_norm": 0.8803633516066148, + "learning_rate": 0.003, + "loss": 4.1775, + "step": 3471 + }, + { + "epoch": 0.03472, + "grad_norm": 0.8377841449731822, + "learning_rate": 0.003, + "loss": 4.177, + "step": 3472 + }, + { + "epoch": 0.03473, + "grad_norm": 0.8583404258765982, + "learning_rate": 0.003, + "loss": 4.1747, + "step": 3473 + }, + { + "epoch": 0.03474, + "grad_norm": 0.821362643848504, + "learning_rate": 0.003, + "loss": 4.2036, + "step": 3474 + }, + { + "epoch": 0.03475, + "grad_norm": 0.7366543444137226, + "learning_rate": 0.003, + "loss": 4.171, + "step": 3475 + }, + { + "epoch": 0.03476, + "grad_norm": 0.8990414035833436, + "learning_rate": 0.003, + "loss": 4.1854, + "step": 3476 + }, + { + "epoch": 0.03477, + "grad_norm": 0.9251586453057641, + "learning_rate": 0.003, + "loss": 4.1836, + "step": 3477 + }, + { + "epoch": 0.03478, + "grad_norm": 0.9081762714287535, + "learning_rate": 0.003, + "loss": 4.1933, + "step": 3478 + }, + { + "epoch": 0.03479, + "grad_norm": 0.95014653125921, + "learning_rate": 0.003, + "loss": 4.1698, + "step": 3479 + }, + { + "epoch": 0.0348, + "grad_norm": 0.8211598773714466, + "learning_rate": 0.003, + "loss": 4.1704, + "step": 3480 + }, + { + "epoch": 0.03481, + "grad_norm": 0.6714433574061417, + "learning_rate": 0.003, + "loss": 4.182, + "step": 3481 + }, + { + "epoch": 0.03482, + "grad_norm": 0.6624685031516976, + "learning_rate": 0.003, + "loss": 4.1874, + "step": 3482 + }, + { + "epoch": 0.03483, + "grad_norm": 0.6364069563539299, + "learning_rate": 0.003, + "loss": 4.163, + "step": 3483 + }, + { + "epoch": 0.03484, + "grad_norm": 0.7252638284821149, + "learning_rate": 0.003, + "loss": 4.2109, + "step": 3484 + }, + { + "epoch": 0.03485, + "grad_norm": 0.6599203755955035, + "learning_rate": 0.003, + "loss": 4.2011, + "step": 3485 + }, + { + "epoch": 0.03486, + "grad_norm": 0.6351455036938864, + "learning_rate": 0.003, + "loss": 4.1709, + "step": 3486 + }, + { + "epoch": 0.03487, + "grad_norm": 0.7047689274973644, + "learning_rate": 0.003, + "loss": 4.1641, + "step": 3487 + }, + { + "epoch": 0.03488, + "grad_norm": 0.5998971286625794, + "learning_rate": 0.003, + "loss": 4.1568, + "step": 3488 + }, + { + "epoch": 0.03489, + "grad_norm": 0.5928637335909946, + "learning_rate": 0.003, + "loss": 4.1727, + "step": 3489 + }, + { + "epoch": 0.0349, + "grad_norm": 0.7224065582654237, + "learning_rate": 0.003, + "loss": 4.1727, + "step": 3490 + }, + { + "epoch": 0.03491, + "grad_norm": 0.8948866147367468, + "learning_rate": 0.003, + "loss": 4.1927, + "step": 3491 + }, + { + "epoch": 0.03492, + "grad_norm": 1.0698897809029222, + "learning_rate": 0.003, + "loss": 4.1887, + "step": 3492 + }, + { + "epoch": 0.03493, + "grad_norm": 0.879044396872005, + "learning_rate": 0.003, + "loss": 4.1775, + "step": 3493 + }, + { + "epoch": 0.03494, + "grad_norm": 0.7308396541826835, + "learning_rate": 0.003, + "loss": 4.1738, + "step": 3494 + }, + { + "epoch": 0.03495, + "grad_norm": 0.7338448975678681, + "learning_rate": 0.003, + "loss": 4.1823, + "step": 3495 + }, + { + "epoch": 0.03496, + "grad_norm": 0.8206218724499419, + "learning_rate": 0.003, + "loss": 4.1564, + "step": 3496 + }, + { + "epoch": 0.03497, + "grad_norm": 0.8420218799600944, + "learning_rate": 0.003, + "loss": 4.1943, + "step": 3497 + }, + { + "epoch": 0.03498, + "grad_norm": 0.7276352170340858, + "learning_rate": 0.003, + "loss": 4.1624, + "step": 3498 + }, + { + "epoch": 0.03499, + "grad_norm": 0.6164613363129237, + "learning_rate": 0.003, + "loss": 4.1637, + "step": 3499 + }, + { + "epoch": 0.035, + "grad_norm": 0.5533625720370539, + "learning_rate": 0.003, + "loss": 4.1821, + "step": 3500 + }, + { + "epoch": 0.03501, + "grad_norm": 0.5440318172984431, + "learning_rate": 0.003, + "loss": 4.1716, + "step": 3501 + }, + { + "epoch": 0.03502, + "grad_norm": 0.570449376318866, + "learning_rate": 0.003, + "loss": 4.1707, + "step": 3502 + }, + { + "epoch": 0.03503, + "grad_norm": 0.5938098529331512, + "learning_rate": 0.003, + "loss": 4.1865, + "step": 3503 + }, + { + "epoch": 0.03504, + "grad_norm": 0.5807097638534694, + "learning_rate": 0.003, + "loss": 4.1574, + "step": 3504 + }, + { + "epoch": 0.03505, + "grad_norm": 0.5346263597287685, + "learning_rate": 0.003, + "loss": 4.1718, + "step": 3505 + }, + { + "epoch": 0.03506, + "grad_norm": 0.5906950615820623, + "learning_rate": 0.003, + "loss": 4.182, + "step": 3506 + }, + { + "epoch": 0.03507, + "grad_norm": 0.5854291049171028, + "learning_rate": 0.003, + "loss": 4.1601, + "step": 3507 + }, + { + "epoch": 0.03508, + "grad_norm": 0.5919853377246803, + "learning_rate": 0.003, + "loss": 4.1373, + "step": 3508 + }, + { + "epoch": 0.03509, + "grad_norm": 0.6611088078675262, + "learning_rate": 0.003, + "loss": 4.1537, + "step": 3509 + }, + { + "epoch": 0.0351, + "grad_norm": 0.694428361139572, + "learning_rate": 0.003, + "loss": 4.1446, + "step": 3510 + }, + { + "epoch": 0.03511, + "grad_norm": 0.5309620036695369, + "learning_rate": 0.003, + "loss": 4.1557, + "step": 3511 + }, + { + "epoch": 0.03512, + "grad_norm": 0.4298187490647193, + "learning_rate": 0.003, + "loss": 4.1416, + "step": 3512 + }, + { + "epoch": 0.03513, + "grad_norm": 0.4111172909761725, + "learning_rate": 0.003, + "loss": 4.1448, + "step": 3513 + }, + { + "epoch": 0.03514, + "grad_norm": 0.45911964934305066, + "learning_rate": 0.003, + "loss": 4.1562, + "step": 3514 + }, + { + "epoch": 0.03515, + "grad_norm": 0.5301625577665102, + "learning_rate": 0.003, + "loss": 4.1654, + "step": 3515 + }, + { + "epoch": 0.03516, + "grad_norm": 0.6537735331219108, + "learning_rate": 0.003, + "loss": 4.1947, + "step": 3516 + }, + { + "epoch": 0.03517, + "grad_norm": 0.8603631139968849, + "learning_rate": 0.003, + "loss": 4.153, + "step": 3517 + }, + { + "epoch": 0.03518, + "grad_norm": 0.9381334187285012, + "learning_rate": 0.003, + "loss": 4.1471, + "step": 3518 + }, + { + "epoch": 0.03519, + "grad_norm": 0.7499517998636877, + "learning_rate": 0.003, + "loss": 4.1858, + "step": 3519 + }, + { + "epoch": 0.0352, + "grad_norm": 0.6063052992661483, + "learning_rate": 0.003, + "loss": 4.1619, + "step": 3520 + }, + { + "epoch": 0.03521, + "grad_norm": 0.6898167453059805, + "learning_rate": 0.003, + "loss": 4.1613, + "step": 3521 + }, + { + "epoch": 0.03522, + "grad_norm": 0.7285294145828302, + "learning_rate": 0.003, + "loss": 4.1629, + "step": 3522 + }, + { + "epoch": 0.03523, + "grad_norm": 0.786382262009744, + "learning_rate": 0.003, + "loss": 4.1349, + "step": 3523 + }, + { + "epoch": 0.03524, + "grad_norm": 0.8485073208145252, + "learning_rate": 0.003, + "loss": 4.1811, + "step": 3524 + }, + { + "epoch": 0.03525, + "grad_norm": 0.7884181271432541, + "learning_rate": 0.003, + "loss": 4.1487, + "step": 3525 + }, + { + "epoch": 0.03526, + "grad_norm": 0.6895666981950603, + "learning_rate": 0.003, + "loss": 4.1533, + "step": 3526 + }, + { + "epoch": 0.03527, + "grad_norm": 0.6609448619741584, + "learning_rate": 0.003, + "loss": 4.1841, + "step": 3527 + }, + { + "epoch": 0.03528, + "grad_norm": 0.6552099898208482, + "learning_rate": 0.003, + "loss": 4.1683, + "step": 3528 + }, + { + "epoch": 0.03529, + "grad_norm": 0.5657486402844111, + "learning_rate": 0.003, + "loss": 4.1582, + "step": 3529 + }, + { + "epoch": 0.0353, + "grad_norm": 0.6271377691303136, + "learning_rate": 0.003, + "loss": 4.1603, + "step": 3530 + }, + { + "epoch": 0.03531, + "grad_norm": 0.6215946372415877, + "learning_rate": 0.003, + "loss": 4.1691, + "step": 3531 + }, + { + "epoch": 0.03532, + "grad_norm": 0.6387289961881946, + "learning_rate": 0.003, + "loss": 4.1723, + "step": 3532 + }, + { + "epoch": 0.03533, + "grad_norm": 0.6872453476711173, + "learning_rate": 0.003, + "loss": 4.1733, + "step": 3533 + }, + { + "epoch": 0.03534, + "grad_norm": 0.7783886281484835, + "learning_rate": 0.003, + "loss": 4.1711, + "step": 3534 + }, + { + "epoch": 0.03535, + "grad_norm": 0.8725146995031193, + "learning_rate": 0.003, + "loss": 4.1506, + "step": 3535 + }, + { + "epoch": 0.03536, + "grad_norm": 0.9246122970504058, + "learning_rate": 0.003, + "loss": 4.1551, + "step": 3536 + }, + { + "epoch": 0.03537, + "grad_norm": 1.065477970890386, + "learning_rate": 0.003, + "loss": 4.1712, + "step": 3537 + }, + { + "epoch": 0.03538, + "grad_norm": 0.9594169294264835, + "learning_rate": 0.003, + "loss": 4.1706, + "step": 3538 + }, + { + "epoch": 0.03539, + "grad_norm": 0.8330662665517189, + "learning_rate": 0.003, + "loss": 4.1586, + "step": 3539 + }, + { + "epoch": 0.0354, + "grad_norm": 0.8526080452896662, + "learning_rate": 0.003, + "loss": 4.2005, + "step": 3540 + }, + { + "epoch": 0.03541, + "grad_norm": 0.9085329983346553, + "learning_rate": 0.003, + "loss": 4.1886, + "step": 3541 + }, + { + "epoch": 0.03542, + "grad_norm": 0.833985994066796, + "learning_rate": 0.003, + "loss": 4.1702, + "step": 3542 + }, + { + "epoch": 0.03543, + "grad_norm": 0.7626821688858162, + "learning_rate": 0.003, + "loss": 4.1763, + "step": 3543 + }, + { + "epoch": 0.03544, + "grad_norm": 0.806738250706975, + "learning_rate": 0.003, + "loss": 4.1843, + "step": 3544 + }, + { + "epoch": 0.03545, + "grad_norm": 0.9242341338381844, + "learning_rate": 0.003, + "loss": 4.1613, + "step": 3545 + }, + { + "epoch": 0.03546, + "grad_norm": 0.8293267592237652, + "learning_rate": 0.003, + "loss": 4.1759, + "step": 3546 + }, + { + "epoch": 0.03547, + "grad_norm": 0.7605341067555178, + "learning_rate": 0.003, + "loss": 4.1823, + "step": 3547 + }, + { + "epoch": 0.03548, + "grad_norm": 0.697933140723856, + "learning_rate": 0.003, + "loss": 4.1877, + "step": 3548 + }, + { + "epoch": 0.03549, + "grad_norm": 0.5804050326776297, + "learning_rate": 0.003, + "loss": 4.1739, + "step": 3549 + }, + { + "epoch": 0.0355, + "grad_norm": 0.603076289815789, + "learning_rate": 0.003, + "loss": 4.1292, + "step": 3550 + }, + { + "epoch": 0.03551, + "grad_norm": 0.5646969274817457, + "learning_rate": 0.003, + "loss": 4.1847, + "step": 3551 + }, + { + "epoch": 0.03552, + "grad_norm": 0.5353048682624879, + "learning_rate": 0.003, + "loss": 4.1445, + "step": 3552 + }, + { + "epoch": 0.03553, + "grad_norm": 0.43948675430004575, + "learning_rate": 0.003, + "loss": 4.1551, + "step": 3553 + }, + { + "epoch": 0.03554, + "grad_norm": 0.4404992722313449, + "learning_rate": 0.003, + "loss": 4.1584, + "step": 3554 + }, + { + "epoch": 0.03555, + "grad_norm": 0.4408782434286909, + "learning_rate": 0.003, + "loss": 4.1708, + "step": 3555 + }, + { + "epoch": 0.03556, + "grad_norm": 0.4089405885332322, + "learning_rate": 0.003, + "loss": 4.1496, + "step": 3556 + }, + { + "epoch": 0.03557, + "grad_norm": 0.4886396097558162, + "learning_rate": 0.003, + "loss": 4.1516, + "step": 3557 + }, + { + "epoch": 0.03558, + "grad_norm": 0.5683554684886454, + "learning_rate": 0.003, + "loss": 4.1794, + "step": 3558 + }, + { + "epoch": 0.03559, + "grad_norm": 0.7367360745457323, + "learning_rate": 0.003, + "loss": 4.1684, + "step": 3559 + }, + { + "epoch": 0.0356, + "grad_norm": 0.8520355513833522, + "learning_rate": 0.003, + "loss": 4.1751, + "step": 3560 + }, + { + "epoch": 0.03561, + "grad_norm": 0.8538012995310599, + "learning_rate": 0.003, + "loss": 4.1983, + "step": 3561 + }, + { + "epoch": 0.03562, + "grad_norm": 0.6793279389206137, + "learning_rate": 0.003, + "loss": 4.1602, + "step": 3562 + }, + { + "epoch": 0.03563, + "grad_norm": 0.5212190292484681, + "learning_rate": 0.003, + "loss": 4.1876, + "step": 3563 + }, + { + "epoch": 0.03564, + "grad_norm": 0.6272603340476319, + "learning_rate": 0.003, + "loss": 4.1651, + "step": 3564 + }, + { + "epoch": 0.03565, + "grad_norm": 0.7385189206717337, + "learning_rate": 0.003, + "loss": 4.1607, + "step": 3565 + }, + { + "epoch": 0.03566, + "grad_norm": 0.6717765632340529, + "learning_rate": 0.003, + "loss": 4.1622, + "step": 3566 + }, + { + "epoch": 0.03567, + "grad_norm": 0.6137260235822769, + "learning_rate": 0.003, + "loss": 4.154, + "step": 3567 + }, + { + "epoch": 0.03568, + "grad_norm": 0.5473859937693163, + "learning_rate": 0.003, + "loss": 4.1497, + "step": 3568 + }, + { + "epoch": 0.03569, + "grad_norm": 0.5260257267109332, + "learning_rate": 0.003, + "loss": 4.1459, + "step": 3569 + }, + { + "epoch": 0.0357, + "grad_norm": 0.5464547294802432, + "learning_rate": 0.003, + "loss": 4.1462, + "step": 3570 + }, + { + "epoch": 0.03571, + "grad_norm": 0.5170616946027875, + "learning_rate": 0.003, + "loss": 4.1547, + "step": 3571 + }, + { + "epoch": 0.03572, + "grad_norm": 0.5245302678289541, + "learning_rate": 0.003, + "loss": 4.1721, + "step": 3572 + }, + { + "epoch": 0.03573, + "grad_norm": 0.5539674018003774, + "learning_rate": 0.003, + "loss": 4.1421, + "step": 3573 + }, + { + "epoch": 0.03574, + "grad_norm": 0.5717457987893073, + "learning_rate": 0.003, + "loss": 4.1694, + "step": 3574 + }, + { + "epoch": 0.03575, + "grad_norm": 0.6597241159959504, + "learning_rate": 0.003, + "loss": 4.1498, + "step": 3575 + }, + { + "epoch": 0.03576, + "grad_norm": 0.7987769449923495, + "learning_rate": 0.003, + "loss": 4.1377, + "step": 3576 + }, + { + "epoch": 0.03577, + "grad_norm": 0.9166920267526417, + "learning_rate": 0.003, + "loss": 4.1801, + "step": 3577 + }, + { + "epoch": 0.03578, + "grad_norm": 0.8840919078143189, + "learning_rate": 0.003, + "loss": 4.1728, + "step": 3578 + }, + { + "epoch": 0.03579, + "grad_norm": 0.8228224911003803, + "learning_rate": 0.003, + "loss": 4.1888, + "step": 3579 + }, + { + "epoch": 0.0358, + "grad_norm": 0.9419353853796917, + "learning_rate": 0.003, + "loss": 4.1488, + "step": 3580 + }, + { + "epoch": 0.03581, + "grad_norm": 0.8980601226690328, + "learning_rate": 0.003, + "loss": 4.1607, + "step": 3581 + }, + { + "epoch": 0.03582, + "grad_norm": 0.8072733688617605, + "learning_rate": 0.003, + "loss": 4.1621, + "step": 3582 + }, + { + "epoch": 0.03583, + "grad_norm": 0.8230842831411875, + "learning_rate": 0.003, + "loss": 4.1635, + "step": 3583 + }, + { + "epoch": 0.03584, + "grad_norm": 0.6285337030399905, + "learning_rate": 0.003, + "loss": 4.1474, + "step": 3584 + }, + { + "epoch": 0.03585, + "grad_norm": 0.6220326404370078, + "learning_rate": 0.003, + "loss": 4.1449, + "step": 3585 + }, + { + "epoch": 0.03586, + "grad_norm": 0.6861320623389772, + "learning_rate": 0.003, + "loss": 4.1677, + "step": 3586 + }, + { + "epoch": 0.03587, + "grad_norm": 0.7015780277654053, + "learning_rate": 0.003, + "loss": 4.1689, + "step": 3587 + }, + { + "epoch": 0.03588, + "grad_norm": 0.6838138540479132, + "learning_rate": 0.003, + "loss": 4.1653, + "step": 3588 + }, + { + "epoch": 0.03589, + "grad_norm": 0.6735354715174068, + "learning_rate": 0.003, + "loss": 4.1806, + "step": 3589 + }, + { + "epoch": 0.0359, + "grad_norm": 0.6330341634771458, + "learning_rate": 0.003, + "loss": 4.1448, + "step": 3590 + }, + { + "epoch": 0.03591, + "grad_norm": 0.5910317698548738, + "learning_rate": 0.003, + "loss": 4.1771, + "step": 3591 + }, + { + "epoch": 0.03592, + "grad_norm": 0.5618523931442762, + "learning_rate": 0.003, + "loss": 4.1551, + "step": 3592 + }, + { + "epoch": 0.03593, + "grad_norm": 0.8151654550358379, + "learning_rate": 0.003, + "loss": 4.1589, + "step": 3593 + }, + { + "epoch": 0.03594, + "grad_norm": 1.0948812647746626, + "learning_rate": 0.003, + "loss": 4.1768, + "step": 3594 + }, + { + "epoch": 0.03595, + "grad_norm": 0.8469599816388995, + "learning_rate": 0.003, + "loss": 4.1615, + "step": 3595 + }, + { + "epoch": 0.03596, + "grad_norm": 0.5860459737545368, + "learning_rate": 0.003, + "loss": 4.1904, + "step": 3596 + }, + { + "epoch": 0.03597, + "grad_norm": 0.6216520906256218, + "learning_rate": 0.003, + "loss": 4.1663, + "step": 3597 + }, + { + "epoch": 0.03598, + "grad_norm": 0.716220967428177, + "learning_rate": 0.003, + "loss": 4.1706, + "step": 3598 + }, + { + "epoch": 0.03599, + "grad_norm": 0.7599051671476444, + "learning_rate": 0.003, + "loss": 4.1682, + "step": 3599 + }, + { + "epoch": 0.036, + "grad_norm": 0.6520846019848653, + "learning_rate": 0.003, + "loss": 4.1898, + "step": 3600 + }, + { + "epoch": 0.03601, + "grad_norm": 0.5364426998203575, + "learning_rate": 0.003, + "loss": 4.1587, + "step": 3601 + }, + { + "epoch": 0.03602, + "grad_norm": 0.5192638145820389, + "learning_rate": 0.003, + "loss": 4.1402, + "step": 3602 + }, + { + "epoch": 0.03603, + "grad_norm": 0.48444783985893647, + "learning_rate": 0.003, + "loss": 4.131, + "step": 3603 + }, + { + "epoch": 0.03604, + "grad_norm": 0.44019612697059707, + "learning_rate": 0.003, + "loss": 4.1673, + "step": 3604 + }, + { + "epoch": 0.03605, + "grad_norm": 0.46867930201266456, + "learning_rate": 0.003, + "loss": 4.1559, + "step": 3605 + }, + { + "epoch": 0.03606, + "grad_norm": 0.4990137003463985, + "learning_rate": 0.003, + "loss": 4.1437, + "step": 3606 + }, + { + "epoch": 0.03607, + "grad_norm": 0.6032395506240065, + "learning_rate": 0.003, + "loss": 4.1394, + "step": 3607 + }, + { + "epoch": 0.03608, + "grad_norm": 0.6248433459385002, + "learning_rate": 0.003, + "loss": 4.1451, + "step": 3608 + }, + { + "epoch": 0.03609, + "grad_norm": 0.6134188004656983, + "learning_rate": 0.003, + "loss": 4.1716, + "step": 3609 + }, + { + "epoch": 0.0361, + "grad_norm": 0.6290690472912259, + "learning_rate": 0.003, + "loss": 4.1749, + "step": 3610 + }, + { + "epoch": 0.03611, + "grad_norm": 0.72557440929179, + "learning_rate": 0.003, + "loss": 4.1283, + "step": 3611 + }, + { + "epoch": 0.03612, + "grad_norm": 0.8893798610273862, + "learning_rate": 0.003, + "loss": 4.1323, + "step": 3612 + }, + { + "epoch": 0.03613, + "grad_norm": 0.9668223551113271, + "learning_rate": 0.003, + "loss": 4.1826, + "step": 3613 + }, + { + "epoch": 0.03614, + "grad_norm": 0.734407911853949, + "learning_rate": 0.003, + "loss": 4.1531, + "step": 3614 + }, + { + "epoch": 0.03615, + "grad_norm": 0.6729100967664783, + "learning_rate": 0.003, + "loss": 4.1648, + "step": 3615 + }, + { + "epoch": 0.03616, + "grad_norm": 0.8657549358453647, + "learning_rate": 0.003, + "loss": 4.177, + "step": 3616 + }, + { + "epoch": 0.03617, + "grad_norm": 0.9260684291150802, + "learning_rate": 0.003, + "loss": 4.1798, + "step": 3617 + }, + { + "epoch": 0.03618, + "grad_norm": 0.8570047461014029, + "learning_rate": 0.003, + "loss": 4.17, + "step": 3618 + }, + { + "epoch": 0.03619, + "grad_norm": 0.866479505486592, + "learning_rate": 0.003, + "loss": 4.175, + "step": 3619 + }, + { + "epoch": 0.0362, + "grad_norm": 0.8958378530573641, + "learning_rate": 0.003, + "loss": 4.1828, + "step": 3620 + }, + { + "epoch": 0.03621, + "grad_norm": 0.8754496119000232, + "learning_rate": 0.003, + "loss": 4.1836, + "step": 3621 + }, + { + "epoch": 0.03622, + "grad_norm": 0.7829921953128642, + "learning_rate": 0.003, + "loss": 4.1499, + "step": 3622 + }, + { + "epoch": 0.03623, + "grad_norm": 0.7264263724470046, + "learning_rate": 0.003, + "loss": 4.1427, + "step": 3623 + }, + { + "epoch": 0.03624, + "grad_norm": 0.7103073858138196, + "learning_rate": 0.003, + "loss": 4.1467, + "step": 3624 + }, + { + "epoch": 0.03625, + "grad_norm": 0.7362483226450557, + "learning_rate": 0.003, + "loss": 4.1852, + "step": 3625 + }, + { + "epoch": 0.03626, + "grad_norm": 0.6700204258570956, + "learning_rate": 0.003, + "loss": 4.1584, + "step": 3626 + }, + { + "epoch": 0.03627, + "grad_norm": 0.8160965768748923, + "learning_rate": 0.003, + "loss": 4.1658, + "step": 3627 + }, + { + "epoch": 0.03628, + "grad_norm": 0.7835131802445204, + "learning_rate": 0.003, + "loss": 4.1739, + "step": 3628 + }, + { + "epoch": 0.03629, + "grad_norm": 0.6985149126811981, + "learning_rate": 0.003, + "loss": 4.1895, + "step": 3629 + }, + { + "epoch": 0.0363, + "grad_norm": 0.6039041743694569, + "learning_rate": 0.003, + "loss": 4.1703, + "step": 3630 + }, + { + "epoch": 0.03631, + "grad_norm": 0.5592412769390824, + "learning_rate": 0.003, + "loss": 4.165, + "step": 3631 + }, + { + "epoch": 0.03632, + "grad_norm": 0.49011079854965545, + "learning_rate": 0.003, + "loss": 4.1538, + "step": 3632 + }, + { + "epoch": 0.03633, + "grad_norm": 0.5382307200734877, + "learning_rate": 0.003, + "loss": 4.1853, + "step": 3633 + }, + { + "epoch": 0.03634, + "grad_norm": 0.5367696632611142, + "learning_rate": 0.003, + "loss": 4.1594, + "step": 3634 + }, + { + "epoch": 0.03635, + "grad_norm": 0.5142486579138349, + "learning_rate": 0.003, + "loss": 4.1415, + "step": 3635 + }, + { + "epoch": 0.03636, + "grad_norm": 0.5652066346276485, + "learning_rate": 0.003, + "loss": 4.1613, + "step": 3636 + }, + { + "epoch": 0.03637, + "grad_norm": 0.59668194695564, + "learning_rate": 0.003, + "loss": 4.1502, + "step": 3637 + }, + { + "epoch": 0.03638, + "grad_norm": 0.6925042816123054, + "learning_rate": 0.003, + "loss": 4.1624, + "step": 3638 + }, + { + "epoch": 0.03639, + "grad_norm": 0.8037659165365143, + "learning_rate": 0.003, + "loss": 4.1869, + "step": 3639 + }, + { + "epoch": 0.0364, + "grad_norm": 0.8148685837655183, + "learning_rate": 0.003, + "loss": 4.1832, + "step": 3640 + }, + { + "epoch": 0.03641, + "grad_norm": 0.7919945416520671, + "learning_rate": 0.003, + "loss": 4.1676, + "step": 3641 + }, + { + "epoch": 0.03642, + "grad_norm": 0.8387878217923936, + "learning_rate": 0.003, + "loss": 4.1883, + "step": 3642 + }, + { + "epoch": 0.03643, + "grad_norm": 0.7524748876313961, + "learning_rate": 0.003, + "loss": 4.1494, + "step": 3643 + }, + { + "epoch": 0.03644, + "grad_norm": 0.7273454421844323, + "learning_rate": 0.003, + "loss": 4.1748, + "step": 3644 + }, + { + "epoch": 0.03645, + "grad_norm": 0.7955485790365344, + "learning_rate": 0.003, + "loss": 4.1475, + "step": 3645 + }, + { + "epoch": 0.03646, + "grad_norm": 0.7720894887238786, + "learning_rate": 0.003, + "loss": 4.1682, + "step": 3646 + }, + { + "epoch": 0.03647, + "grad_norm": 0.7086978218067327, + "learning_rate": 0.003, + "loss": 4.1616, + "step": 3647 + }, + { + "epoch": 0.03648, + "grad_norm": 0.6770260559623669, + "learning_rate": 0.003, + "loss": 4.1506, + "step": 3648 + }, + { + "epoch": 0.03649, + "grad_norm": 0.5881027107624417, + "learning_rate": 0.003, + "loss": 4.1738, + "step": 3649 + }, + { + "epoch": 0.0365, + "grad_norm": 0.6131807187241587, + "learning_rate": 0.003, + "loss": 4.1529, + "step": 3650 + }, + { + "epoch": 0.03651, + "grad_norm": 0.594204825708316, + "learning_rate": 0.003, + "loss": 4.1729, + "step": 3651 + }, + { + "epoch": 0.03652, + "grad_norm": 0.6399255321695406, + "learning_rate": 0.003, + "loss": 4.1571, + "step": 3652 + }, + { + "epoch": 0.03653, + "grad_norm": 0.6119522794582246, + "learning_rate": 0.003, + "loss": 4.1419, + "step": 3653 + }, + { + "epoch": 0.03654, + "grad_norm": 0.5566016631522781, + "learning_rate": 0.003, + "loss": 4.1386, + "step": 3654 + }, + { + "epoch": 0.03655, + "grad_norm": 0.5193773149688491, + "learning_rate": 0.003, + "loss": 4.1618, + "step": 3655 + }, + { + "epoch": 0.03656, + "grad_norm": 0.6264013089429903, + "learning_rate": 0.003, + "loss": 4.1523, + "step": 3656 + }, + { + "epoch": 0.03657, + "grad_norm": 0.6979744938653328, + "learning_rate": 0.003, + "loss": 4.1269, + "step": 3657 + }, + { + "epoch": 0.03658, + "grad_norm": 0.8558728547971467, + "learning_rate": 0.003, + "loss": 4.1964, + "step": 3658 + }, + { + "epoch": 0.03659, + "grad_norm": 0.8839768938776124, + "learning_rate": 0.003, + "loss": 4.1477, + "step": 3659 + }, + { + "epoch": 0.0366, + "grad_norm": 0.738686352865498, + "learning_rate": 0.003, + "loss": 4.1749, + "step": 3660 + }, + { + "epoch": 0.03661, + "grad_norm": 0.7093891716169941, + "learning_rate": 0.003, + "loss": 4.1523, + "step": 3661 + }, + { + "epoch": 0.03662, + "grad_norm": 0.7587081351482545, + "learning_rate": 0.003, + "loss": 4.1525, + "step": 3662 + }, + { + "epoch": 0.03663, + "grad_norm": 0.7076929569924106, + "learning_rate": 0.003, + "loss": 4.169, + "step": 3663 + }, + { + "epoch": 0.03664, + "grad_norm": 0.7127846105810646, + "learning_rate": 0.003, + "loss": 4.1482, + "step": 3664 + }, + { + "epoch": 0.03665, + "grad_norm": 0.7078459190405836, + "learning_rate": 0.003, + "loss": 4.1741, + "step": 3665 + }, + { + "epoch": 0.03666, + "grad_norm": 0.6805684369907511, + "learning_rate": 0.003, + "loss": 4.1297, + "step": 3666 + }, + { + "epoch": 0.03667, + "grad_norm": 0.7150672278612442, + "learning_rate": 0.003, + "loss": 4.1398, + "step": 3667 + }, + { + "epoch": 0.03668, + "grad_norm": 0.7699785342164407, + "learning_rate": 0.003, + "loss": 4.159, + "step": 3668 + }, + { + "epoch": 0.03669, + "grad_norm": 0.7952877417083865, + "learning_rate": 0.003, + "loss": 4.1361, + "step": 3669 + }, + { + "epoch": 0.0367, + "grad_norm": 0.7569891220037553, + "learning_rate": 0.003, + "loss": 4.1731, + "step": 3670 + }, + { + "epoch": 0.03671, + "grad_norm": 0.7098731435624128, + "learning_rate": 0.003, + "loss": 4.1585, + "step": 3671 + }, + { + "epoch": 0.03672, + "grad_norm": 0.6352485396363875, + "learning_rate": 0.003, + "loss": 4.1619, + "step": 3672 + }, + { + "epoch": 0.03673, + "grad_norm": 0.6885305441098153, + "learning_rate": 0.003, + "loss": 4.172, + "step": 3673 + }, + { + "epoch": 0.03674, + "grad_norm": 0.7059131456483599, + "learning_rate": 0.003, + "loss": 4.1377, + "step": 3674 + }, + { + "epoch": 0.03675, + "grad_norm": 0.6038823828670427, + "learning_rate": 0.003, + "loss": 4.1744, + "step": 3675 + }, + { + "epoch": 0.03676, + "grad_norm": 0.57222994957781, + "learning_rate": 0.003, + "loss": 4.1319, + "step": 3676 + }, + { + "epoch": 0.03677, + "grad_norm": 0.6802347986469087, + "learning_rate": 0.003, + "loss": 4.1492, + "step": 3677 + }, + { + "epoch": 0.03678, + "grad_norm": 0.7740620961238671, + "learning_rate": 0.003, + "loss": 4.1812, + "step": 3678 + }, + { + "epoch": 0.03679, + "grad_norm": 0.8635840591832115, + "learning_rate": 0.003, + "loss": 4.1775, + "step": 3679 + }, + { + "epoch": 0.0368, + "grad_norm": 0.9019802875931285, + "learning_rate": 0.003, + "loss": 4.173, + "step": 3680 + }, + { + "epoch": 0.03681, + "grad_norm": 0.9372442716730703, + "learning_rate": 0.003, + "loss": 4.1511, + "step": 3681 + }, + { + "epoch": 0.03682, + "grad_norm": 0.9892266156518691, + "learning_rate": 0.003, + "loss": 4.1916, + "step": 3682 + }, + { + "epoch": 0.03683, + "grad_norm": 1.1018562984300873, + "learning_rate": 0.003, + "loss": 4.1632, + "step": 3683 + }, + { + "epoch": 0.03684, + "grad_norm": 0.9775119392439015, + "learning_rate": 0.003, + "loss": 4.1875, + "step": 3684 + }, + { + "epoch": 0.03685, + "grad_norm": 0.9024302815079079, + "learning_rate": 0.003, + "loss": 4.1723, + "step": 3685 + }, + { + "epoch": 0.03686, + "grad_norm": 0.9419765899805256, + "learning_rate": 0.003, + "loss": 4.1901, + "step": 3686 + }, + { + "epoch": 0.03687, + "grad_norm": 1.1488204202771457, + "learning_rate": 0.003, + "loss": 4.2226, + "step": 3687 + }, + { + "epoch": 0.03688, + "grad_norm": 0.9052265907645448, + "learning_rate": 0.003, + "loss": 4.2006, + "step": 3688 + }, + { + "epoch": 0.03689, + "grad_norm": 0.8161052846618367, + "learning_rate": 0.003, + "loss": 4.1667, + "step": 3689 + }, + { + "epoch": 0.0369, + "grad_norm": 0.8630363350313656, + "learning_rate": 0.003, + "loss": 4.1995, + "step": 3690 + }, + { + "epoch": 0.03691, + "grad_norm": 1.000347066652698, + "learning_rate": 0.003, + "loss": 4.1833, + "step": 3691 + }, + { + "epoch": 0.03692, + "grad_norm": 1.0125314193391763, + "learning_rate": 0.003, + "loss": 4.2237, + "step": 3692 + }, + { + "epoch": 0.03693, + "grad_norm": 0.7792710807207066, + "learning_rate": 0.003, + "loss": 4.187, + "step": 3693 + }, + { + "epoch": 0.03694, + "grad_norm": 0.6567620211689106, + "learning_rate": 0.003, + "loss": 4.1772, + "step": 3694 + }, + { + "epoch": 0.03695, + "grad_norm": 0.6069200650214434, + "learning_rate": 0.003, + "loss": 4.1749, + "step": 3695 + }, + { + "epoch": 0.03696, + "grad_norm": 0.6245228527312463, + "learning_rate": 0.003, + "loss": 4.169, + "step": 3696 + }, + { + "epoch": 0.03697, + "grad_norm": 0.5956476153449444, + "learning_rate": 0.003, + "loss": 4.1568, + "step": 3697 + }, + { + "epoch": 0.03698, + "grad_norm": 0.6172146239896532, + "learning_rate": 0.003, + "loss": 4.1911, + "step": 3698 + }, + { + "epoch": 0.03699, + "grad_norm": 0.6369440678819488, + "learning_rate": 0.003, + "loss": 4.1784, + "step": 3699 + }, + { + "epoch": 0.037, + "grad_norm": 0.7397659340274994, + "learning_rate": 0.003, + "loss": 4.1703, + "step": 3700 + }, + { + "epoch": 0.03701, + "grad_norm": 0.7400755144798878, + "learning_rate": 0.003, + "loss": 4.1708, + "step": 3701 + }, + { + "epoch": 0.03702, + "grad_norm": 0.6011624195612609, + "learning_rate": 0.003, + "loss": 4.1568, + "step": 3702 + }, + { + "epoch": 0.03703, + "grad_norm": 0.45688971739502526, + "learning_rate": 0.003, + "loss": 4.1658, + "step": 3703 + }, + { + "epoch": 0.03704, + "grad_norm": 0.4625467744385196, + "learning_rate": 0.003, + "loss": 4.1661, + "step": 3704 + }, + { + "epoch": 0.03705, + "grad_norm": 0.4601609886250813, + "learning_rate": 0.003, + "loss": 4.1799, + "step": 3705 + }, + { + "epoch": 0.03706, + "grad_norm": 0.4966406347936221, + "learning_rate": 0.003, + "loss": 4.1635, + "step": 3706 + }, + { + "epoch": 0.03707, + "grad_norm": 0.42477832568061835, + "learning_rate": 0.003, + "loss": 4.142, + "step": 3707 + }, + { + "epoch": 0.03708, + "grad_norm": 0.40682526312634587, + "learning_rate": 0.003, + "loss": 4.1616, + "step": 3708 + }, + { + "epoch": 0.03709, + "grad_norm": 0.4133271539515952, + "learning_rate": 0.003, + "loss": 4.1372, + "step": 3709 + }, + { + "epoch": 0.0371, + "grad_norm": 0.43081117839287136, + "learning_rate": 0.003, + "loss": 4.1728, + "step": 3710 + }, + { + "epoch": 0.03711, + "grad_norm": 0.46229609767440477, + "learning_rate": 0.003, + "loss": 4.1425, + "step": 3711 + }, + { + "epoch": 0.03712, + "grad_norm": 0.4820292193047058, + "learning_rate": 0.003, + "loss": 4.1421, + "step": 3712 + }, + { + "epoch": 0.03713, + "grad_norm": 0.5404513625982291, + "learning_rate": 0.003, + "loss": 4.1328, + "step": 3713 + }, + { + "epoch": 0.03714, + "grad_norm": 0.6436389631284867, + "learning_rate": 0.003, + "loss": 4.1474, + "step": 3714 + }, + { + "epoch": 0.03715, + "grad_norm": 0.8290141406987748, + "learning_rate": 0.003, + "loss": 4.1603, + "step": 3715 + }, + { + "epoch": 0.03716, + "grad_norm": 0.9048610214356393, + "learning_rate": 0.003, + "loss": 4.1556, + "step": 3716 + }, + { + "epoch": 0.03717, + "grad_norm": 0.7058124491704675, + "learning_rate": 0.003, + "loss": 4.1532, + "step": 3717 + }, + { + "epoch": 0.03718, + "grad_norm": 0.4918162555153287, + "learning_rate": 0.003, + "loss": 4.117, + "step": 3718 + }, + { + "epoch": 0.03719, + "grad_norm": 0.6622934600969125, + "learning_rate": 0.003, + "loss": 4.1532, + "step": 3719 + }, + { + "epoch": 0.0372, + "grad_norm": 0.7298879766213411, + "learning_rate": 0.003, + "loss": 4.1559, + "step": 3720 + }, + { + "epoch": 0.03721, + "grad_norm": 0.7368836467114779, + "learning_rate": 0.003, + "loss": 4.1451, + "step": 3721 + }, + { + "epoch": 0.03722, + "grad_norm": 0.6530001576353514, + "learning_rate": 0.003, + "loss": 4.1484, + "step": 3722 + }, + { + "epoch": 0.03723, + "grad_norm": 0.6828551463380831, + "learning_rate": 0.003, + "loss": 4.1778, + "step": 3723 + }, + { + "epoch": 0.03724, + "grad_norm": 0.8098638547098356, + "learning_rate": 0.003, + "loss": 4.1595, + "step": 3724 + }, + { + "epoch": 0.03725, + "grad_norm": 0.787968577179379, + "learning_rate": 0.003, + "loss": 4.1709, + "step": 3725 + }, + { + "epoch": 0.03726, + "grad_norm": 0.790721512863178, + "learning_rate": 0.003, + "loss": 4.1451, + "step": 3726 + }, + { + "epoch": 0.03727, + "grad_norm": 0.7875255929753493, + "learning_rate": 0.003, + "loss": 4.1842, + "step": 3727 + }, + { + "epoch": 0.03728, + "grad_norm": 0.7766724710074185, + "learning_rate": 0.003, + "loss": 4.1313, + "step": 3728 + }, + { + "epoch": 0.03729, + "grad_norm": 0.7446811620432703, + "learning_rate": 0.003, + "loss": 4.1678, + "step": 3729 + }, + { + "epoch": 0.0373, + "grad_norm": 0.7631969023181137, + "learning_rate": 0.003, + "loss": 4.1535, + "step": 3730 + }, + { + "epoch": 0.03731, + "grad_norm": 0.8301810283690936, + "learning_rate": 0.003, + "loss": 4.1798, + "step": 3731 + }, + { + "epoch": 0.03732, + "grad_norm": 0.795728492246486, + "learning_rate": 0.003, + "loss": 4.1797, + "step": 3732 + }, + { + "epoch": 0.03733, + "grad_norm": 0.6502399531894205, + "learning_rate": 0.003, + "loss": 4.152, + "step": 3733 + }, + { + "epoch": 0.03734, + "grad_norm": 0.6112985055672381, + "learning_rate": 0.003, + "loss": 4.1521, + "step": 3734 + }, + { + "epoch": 0.03735, + "grad_norm": 0.7852832594186502, + "learning_rate": 0.003, + "loss": 4.1517, + "step": 3735 + }, + { + "epoch": 0.03736, + "grad_norm": 0.9467543719299969, + "learning_rate": 0.003, + "loss": 4.1902, + "step": 3736 + }, + { + "epoch": 0.03737, + "grad_norm": 0.9659184343775771, + "learning_rate": 0.003, + "loss": 4.1847, + "step": 3737 + }, + { + "epoch": 0.03738, + "grad_norm": 0.7573591347533927, + "learning_rate": 0.003, + "loss": 4.1903, + "step": 3738 + }, + { + "epoch": 0.03739, + "grad_norm": 0.7992106244448375, + "learning_rate": 0.003, + "loss": 4.1561, + "step": 3739 + }, + { + "epoch": 0.0374, + "grad_norm": 0.7486649819121083, + "learning_rate": 0.003, + "loss": 4.1738, + "step": 3740 + }, + { + "epoch": 0.03741, + "grad_norm": 0.5976608138563914, + "learning_rate": 0.003, + "loss": 4.1544, + "step": 3741 + }, + { + "epoch": 0.03742, + "grad_norm": 0.5995036064200935, + "learning_rate": 0.003, + "loss": 4.1461, + "step": 3742 + }, + { + "epoch": 0.03743, + "grad_norm": 0.6901682901857126, + "learning_rate": 0.003, + "loss": 4.1447, + "step": 3743 + }, + { + "epoch": 0.03744, + "grad_norm": 0.6080608979861319, + "learning_rate": 0.003, + "loss": 4.1734, + "step": 3744 + }, + { + "epoch": 0.03745, + "grad_norm": 0.6576923197541351, + "learning_rate": 0.003, + "loss": 4.1593, + "step": 3745 + }, + { + "epoch": 0.03746, + "grad_norm": 0.6727287733203032, + "learning_rate": 0.003, + "loss": 4.1675, + "step": 3746 + }, + { + "epoch": 0.03747, + "grad_norm": 0.7572729409130596, + "learning_rate": 0.003, + "loss": 4.175, + "step": 3747 + }, + { + "epoch": 0.03748, + "grad_norm": 0.7027720695663551, + "learning_rate": 0.003, + "loss": 4.1438, + "step": 3748 + }, + { + "epoch": 0.03749, + "grad_norm": 0.7669711680997199, + "learning_rate": 0.003, + "loss": 4.1588, + "step": 3749 + }, + { + "epoch": 0.0375, + "grad_norm": 0.7408458574924264, + "learning_rate": 0.003, + "loss": 4.1786, + "step": 3750 + }, + { + "epoch": 0.03751, + "grad_norm": 0.6282480274827662, + "learning_rate": 0.003, + "loss": 4.1276, + "step": 3751 + }, + { + "epoch": 0.03752, + "grad_norm": 0.6086039975418626, + "learning_rate": 0.003, + "loss": 4.1587, + "step": 3752 + }, + { + "epoch": 0.03753, + "grad_norm": 0.6045335662772724, + "learning_rate": 0.003, + "loss": 4.1601, + "step": 3753 + }, + { + "epoch": 0.03754, + "grad_norm": 0.6756444191450836, + "learning_rate": 0.003, + "loss": 4.172, + "step": 3754 + }, + { + "epoch": 0.03755, + "grad_norm": 0.6814576187053439, + "learning_rate": 0.003, + "loss": 4.1638, + "step": 3755 + }, + { + "epoch": 0.03756, + "grad_norm": 0.6979671812601237, + "learning_rate": 0.003, + "loss": 4.1331, + "step": 3756 + }, + { + "epoch": 0.03757, + "grad_norm": 0.7762754751971193, + "learning_rate": 0.003, + "loss": 4.1703, + "step": 3757 + }, + { + "epoch": 0.03758, + "grad_norm": 0.9320275397343457, + "learning_rate": 0.003, + "loss": 4.1158, + "step": 3758 + }, + { + "epoch": 0.03759, + "grad_norm": 1.029430729208351, + "learning_rate": 0.003, + "loss": 4.1823, + "step": 3759 + }, + { + "epoch": 0.0376, + "grad_norm": 0.6884103372684509, + "learning_rate": 0.003, + "loss": 4.1603, + "step": 3760 + }, + { + "epoch": 0.03761, + "grad_norm": 0.5726368511742514, + "learning_rate": 0.003, + "loss": 4.1248, + "step": 3761 + }, + { + "epoch": 0.03762, + "grad_norm": 0.7431155489308362, + "learning_rate": 0.003, + "loss": 4.1857, + "step": 3762 + }, + { + "epoch": 0.03763, + "grad_norm": 0.7334829914103421, + "learning_rate": 0.003, + "loss": 4.1677, + "step": 3763 + }, + { + "epoch": 0.03764, + "grad_norm": 0.7755402436276377, + "learning_rate": 0.003, + "loss": 4.182, + "step": 3764 + }, + { + "epoch": 0.03765, + "grad_norm": 0.8071077893148694, + "learning_rate": 0.003, + "loss": 4.2095, + "step": 3765 + }, + { + "epoch": 0.03766, + "grad_norm": 0.8062094613718822, + "learning_rate": 0.003, + "loss": 4.1459, + "step": 3766 + }, + { + "epoch": 0.03767, + "grad_norm": 0.7450385148464626, + "learning_rate": 0.003, + "loss": 4.1846, + "step": 3767 + }, + { + "epoch": 0.03768, + "grad_norm": 0.8501525775153304, + "learning_rate": 0.003, + "loss": 4.1688, + "step": 3768 + }, + { + "epoch": 0.03769, + "grad_norm": 0.8490270851264778, + "learning_rate": 0.003, + "loss": 4.1598, + "step": 3769 + }, + { + "epoch": 0.0377, + "grad_norm": 0.6528754246262384, + "learning_rate": 0.003, + "loss": 4.1366, + "step": 3770 + }, + { + "epoch": 0.03771, + "grad_norm": 0.514456285674655, + "learning_rate": 0.003, + "loss": 4.1472, + "step": 3771 + }, + { + "epoch": 0.03772, + "grad_norm": 0.5778069788565984, + "learning_rate": 0.003, + "loss": 4.1704, + "step": 3772 + }, + { + "epoch": 0.03773, + "grad_norm": 0.590491520248524, + "learning_rate": 0.003, + "loss": 4.1726, + "step": 3773 + }, + { + "epoch": 0.03774, + "grad_norm": 0.6156498757993644, + "learning_rate": 0.003, + "loss": 4.16, + "step": 3774 + }, + { + "epoch": 0.03775, + "grad_norm": 0.5373760491110208, + "learning_rate": 0.003, + "loss": 4.1714, + "step": 3775 + }, + { + "epoch": 0.03776, + "grad_norm": 0.58872450634308, + "learning_rate": 0.003, + "loss": 4.1508, + "step": 3776 + }, + { + "epoch": 0.03777, + "grad_norm": 0.57682380030575, + "learning_rate": 0.003, + "loss": 4.1632, + "step": 3777 + }, + { + "epoch": 0.03778, + "grad_norm": 0.5253188081583163, + "learning_rate": 0.003, + "loss": 4.1383, + "step": 3778 + }, + { + "epoch": 0.03779, + "grad_norm": 0.47209668216344913, + "learning_rate": 0.003, + "loss": 4.1517, + "step": 3779 + }, + { + "epoch": 0.0378, + "grad_norm": 0.5332773200600602, + "learning_rate": 0.003, + "loss": 4.1514, + "step": 3780 + }, + { + "epoch": 0.03781, + "grad_norm": 0.6825162371448047, + "learning_rate": 0.003, + "loss": 4.1838, + "step": 3781 + }, + { + "epoch": 0.03782, + "grad_norm": 0.8418563836091014, + "learning_rate": 0.003, + "loss": 4.1608, + "step": 3782 + }, + { + "epoch": 0.03783, + "grad_norm": 0.8512086413504435, + "learning_rate": 0.003, + "loss": 4.1648, + "step": 3783 + }, + { + "epoch": 0.03784, + "grad_norm": 0.6671577794115535, + "learning_rate": 0.003, + "loss": 4.1611, + "step": 3784 + }, + { + "epoch": 0.03785, + "grad_norm": 0.5730225846531266, + "learning_rate": 0.003, + "loss": 4.1395, + "step": 3785 + }, + { + "epoch": 0.03786, + "grad_norm": 0.6939236048367627, + "learning_rate": 0.003, + "loss": 4.1621, + "step": 3786 + }, + { + "epoch": 0.03787, + "grad_norm": 0.7368644792806699, + "learning_rate": 0.003, + "loss": 4.1509, + "step": 3787 + }, + { + "epoch": 0.03788, + "grad_norm": 0.6874505246320723, + "learning_rate": 0.003, + "loss": 4.1821, + "step": 3788 + }, + { + "epoch": 0.03789, + "grad_norm": 0.6323049623986694, + "learning_rate": 0.003, + "loss": 4.1126, + "step": 3789 + }, + { + "epoch": 0.0379, + "grad_norm": 0.6673306922274641, + "learning_rate": 0.003, + "loss": 4.1402, + "step": 3790 + }, + { + "epoch": 0.03791, + "grad_norm": 0.6943632593992325, + "learning_rate": 0.003, + "loss": 4.1536, + "step": 3791 + }, + { + "epoch": 0.03792, + "grad_norm": 0.6791072764782452, + "learning_rate": 0.003, + "loss": 4.1837, + "step": 3792 + }, + { + "epoch": 0.03793, + "grad_norm": 0.6789029894480806, + "learning_rate": 0.003, + "loss": 4.1784, + "step": 3793 + }, + { + "epoch": 0.03794, + "grad_norm": 0.7369905694175998, + "learning_rate": 0.003, + "loss": 4.181, + "step": 3794 + }, + { + "epoch": 0.03795, + "grad_norm": 0.7808489227373923, + "learning_rate": 0.003, + "loss": 4.1682, + "step": 3795 + }, + { + "epoch": 0.03796, + "grad_norm": 0.8037760915310143, + "learning_rate": 0.003, + "loss": 4.1535, + "step": 3796 + }, + { + "epoch": 0.03797, + "grad_norm": 0.7083927011043929, + "learning_rate": 0.003, + "loss": 4.1407, + "step": 3797 + }, + { + "epoch": 0.03798, + "grad_norm": 0.706481627859496, + "learning_rate": 0.003, + "loss": 4.1545, + "step": 3798 + }, + { + "epoch": 0.03799, + "grad_norm": 0.6536139692203335, + "learning_rate": 0.003, + "loss": 4.1627, + "step": 3799 + }, + { + "epoch": 0.038, + "grad_norm": 0.667914966075993, + "learning_rate": 0.003, + "loss": 4.1697, + "step": 3800 + }, + { + "epoch": 0.03801, + "grad_norm": 0.6933040831212794, + "learning_rate": 0.003, + "loss": 4.1758, + "step": 3801 + }, + { + "epoch": 0.03802, + "grad_norm": 0.6580821376084047, + "learning_rate": 0.003, + "loss": 4.1491, + "step": 3802 + }, + { + "epoch": 0.03803, + "grad_norm": 0.7779130392910976, + "learning_rate": 0.003, + "loss": 4.1782, + "step": 3803 + }, + { + "epoch": 0.03804, + "grad_norm": 0.8777193399244893, + "learning_rate": 0.003, + "loss": 4.1529, + "step": 3804 + }, + { + "epoch": 0.03805, + "grad_norm": 1.0020099764204873, + "learning_rate": 0.003, + "loss": 4.1747, + "step": 3805 + }, + { + "epoch": 0.03806, + "grad_norm": 0.9634820397091458, + "learning_rate": 0.003, + "loss": 4.1743, + "step": 3806 + }, + { + "epoch": 0.03807, + "grad_norm": 0.7477715351876884, + "learning_rate": 0.003, + "loss": 4.161, + "step": 3807 + }, + { + "epoch": 0.03808, + "grad_norm": 0.6392070927979916, + "learning_rate": 0.003, + "loss": 4.1744, + "step": 3808 + }, + { + "epoch": 0.03809, + "grad_norm": 0.6899923101840185, + "learning_rate": 0.003, + "loss": 4.1548, + "step": 3809 + }, + { + "epoch": 0.0381, + "grad_norm": 0.6421320844146812, + "learning_rate": 0.003, + "loss": 4.1399, + "step": 3810 + }, + { + "epoch": 0.03811, + "grad_norm": 0.5707513677972008, + "learning_rate": 0.003, + "loss": 4.1678, + "step": 3811 + }, + { + "epoch": 0.03812, + "grad_norm": 0.6104281618032581, + "learning_rate": 0.003, + "loss": 4.1598, + "step": 3812 + }, + { + "epoch": 0.03813, + "grad_norm": 0.7409188794471754, + "learning_rate": 0.003, + "loss": 4.1356, + "step": 3813 + }, + { + "epoch": 0.03814, + "grad_norm": 0.9399869172219258, + "learning_rate": 0.003, + "loss": 4.175, + "step": 3814 + }, + { + "epoch": 0.03815, + "grad_norm": 1.0585281835423583, + "learning_rate": 0.003, + "loss": 4.1727, + "step": 3815 + }, + { + "epoch": 0.03816, + "grad_norm": 0.9799276300500098, + "learning_rate": 0.003, + "loss": 4.1748, + "step": 3816 + }, + { + "epoch": 0.03817, + "grad_norm": 0.9941910477954538, + "learning_rate": 0.003, + "loss": 4.1578, + "step": 3817 + }, + { + "epoch": 0.03818, + "grad_norm": 0.9313720636167057, + "learning_rate": 0.003, + "loss": 4.2098, + "step": 3818 + }, + { + "epoch": 0.03819, + "grad_norm": 0.8314238810075018, + "learning_rate": 0.003, + "loss": 4.1583, + "step": 3819 + }, + { + "epoch": 0.0382, + "grad_norm": 0.7056825175881773, + "learning_rate": 0.003, + "loss": 4.2022, + "step": 3820 + }, + { + "epoch": 0.03821, + "grad_norm": 0.6762157784042842, + "learning_rate": 0.003, + "loss": 4.148, + "step": 3821 + }, + { + "epoch": 0.03822, + "grad_norm": 0.6810353799131027, + "learning_rate": 0.003, + "loss": 4.1704, + "step": 3822 + }, + { + "epoch": 0.03823, + "grad_norm": 0.693219699873982, + "learning_rate": 0.003, + "loss": 4.1655, + "step": 3823 + }, + { + "epoch": 0.03824, + "grad_norm": 0.7816987127966681, + "learning_rate": 0.003, + "loss": 4.1624, + "step": 3824 + }, + { + "epoch": 0.03825, + "grad_norm": 0.8580364555019204, + "learning_rate": 0.003, + "loss": 4.1796, + "step": 3825 + }, + { + "epoch": 0.03826, + "grad_norm": 0.737317401518091, + "learning_rate": 0.003, + "loss": 4.1636, + "step": 3826 + }, + { + "epoch": 0.03827, + "grad_norm": 0.6028910295160049, + "learning_rate": 0.003, + "loss": 4.1852, + "step": 3827 + }, + { + "epoch": 0.03828, + "grad_norm": 0.6588227313580575, + "learning_rate": 0.003, + "loss": 4.1624, + "step": 3828 + }, + { + "epoch": 0.03829, + "grad_norm": 0.6720290422366224, + "learning_rate": 0.003, + "loss": 4.1669, + "step": 3829 + }, + { + "epoch": 0.0383, + "grad_norm": 0.5935234244676734, + "learning_rate": 0.003, + "loss": 4.1537, + "step": 3830 + }, + { + "epoch": 0.03831, + "grad_norm": 0.5027973321474926, + "learning_rate": 0.003, + "loss": 4.151, + "step": 3831 + }, + { + "epoch": 0.03832, + "grad_norm": 0.4793372599756564, + "learning_rate": 0.003, + "loss": 4.1472, + "step": 3832 + }, + { + "epoch": 0.03833, + "grad_norm": 0.46250153234520147, + "learning_rate": 0.003, + "loss": 4.1459, + "step": 3833 + }, + { + "epoch": 0.03834, + "grad_norm": 0.5005957325640198, + "learning_rate": 0.003, + "loss": 4.1758, + "step": 3834 + }, + { + "epoch": 0.03835, + "grad_norm": 0.5638040713097137, + "learning_rate": 0.003, + "loss": 4.1646, + "step": 3835 + }, + { + "epoch": 0.03836, + "grad_norm": 0.7454000939697492, + "learning_rate": 0.003, + "loss": 4.1705, + "step": 3836 + }, + { + "epoch": 0.03837, + "grad_norm": 0.84129158678745, + "learning_rate": 0.003, + "loss": 4.1598, + "step": 3837 + }, + { + "epoch": 0.03838, + "grad_norm": 0.8657748961672681, + "learning_rate": 0.003, + "loss": 4.1511, + "step": 3838 + }, + { + "epoch": 0.03839, + "grad_norm": 0.7593243937614058, + "learning_rate": 0.003, + "loss": 4.1537, + "step": 3839 + }, + { + "epoch": 0.0384, + "grad_norm": 0.6950210484679211, + "learning_rate": 0.003, + "loss": 4.1578, + "step": 3840 + }, + { + "epoch": 0.03841, + "grad_norm": 0.6626542758150601, + "learning_rate": 0.003, + "loss": 4.1611, + "step": 3841 + }, + { + "epoch": 0.03842, + "grad_norm": 0.6047316460014022, + "learning_rate": 0.003, + "loss": 4.159, + "step": 3842 + }, + { + "epoch": 0.03843, + "grad_norm": 0.662178693915001, + "learning_rate": 0.003, + "loss": 4.1461, + "step": 3843 + }, + { + "epoch": 0.03844, + "grad_norm": 0.706679379810732, + "learning_rate": 0.003, + "loss": 4.1568, + "step": 3844 + }, + { + "epoch": 0.03845, + "grad_norm": 0.7114665877564104, + "learning_rate": 0.003, + "loss": 4.1576, + "step": 3845 + }, + { + "epoch": 0.03846, + "grad_norm": 0.6429293496562273, + "learning_rate": 0.003, + "loss": 4.1681, + "step": 3846 + }, + { + "epoch": 0.03847, + "grad_norm": 0.6208196779861823, + "learning_rate": 0.003, + "loss": 4.1565, + "step": 3847 + }, + { + "epoch": 0.03848, + "grad_norm": 0.5991718704769494, + "learning_rate": 0.003, + "loss": 4.1418, + "step": 3848 + }, + { + "epoch": 0.03849, + "grad_norm": 0.6327221679341399, + "learning_rate": 0.003, + "loss": 4.1542, + "step": 3849 + }, + { + "epoch": 0.0385, + "grad_norm": 0.6393009934280616, + "learning_rate": 0.003, + "loss": 4.1654, + "step": 3850 + }, + { + "epoch": 0.03851, + "grad_norm": 0.7219030496541264, + "learning_rate": 0.003, + "loss": 4.1303, + "step": 3851 + }, + { + "epoch": 0.03852, + "grad_norm": 0.741478675399483, + "learning_rate": 0.003, + "loss": 4.1667, + "step": 3852 + }, + { + "epoch": 0.03853, + "grad_norm": 0.6467861453700433, + "learning_rate": 0.003, + "loss": 4.1618, + "step": 3853 + }, + { + "epoch": 0.03854, + "grad_norm": 0.698278301052957, + "learning_rate": 0.003, + "loss": 4.1565, + "step": 3854 + }, + { + "epoch": 0.03855, + "grad_norm": 0.7121845392242947, + "learning_rate": 0.003, + "loss": 4.1765, + "step": 3855 + }, + { + "epoch": 0.03856, + "grad_norm": 0.6723670160555435, + "learning_rate": 0.003, + "loss": 4.1461, + "step": 3856 + }, + { + "epoch": 0.03857, + "grad_norm": 0.6243401732081879, + "learning_rate": 0.003, + "loss": 4.1494, + "step": 3857 + }, + { + "epoch": 0.03858, + "grad_norm": 0.5301294562723087, + "learning_rate": 0.003, + "loss": 4.1414, + "step": 3858 + }, + { + "epoch": 0.03859, + "grad_norm": 0.50901761681548, + "learning_rate": 0.003, + "loss": 4.1651, + "step": 3859 + }, + { + "epoch": 0.0386, + "grad_norm": 0.5147249879881411, + "learning_rate": 0.003, + "loss": 4.1236, + "step": 3860 + }, + { + "epoch": 0.03861, + "grad_norm": 0.4939497869282373, + "learning_rate": 0.003, + "loss": 4.1638, + "step": 3861 + }, + { + "epoch": 0.03862, + "grad_norm": 0.5565978694443094, + "learning_rate": 0.003, + "loss": 4.1425, + "step": 3862 + }, + { + "epoch": 0.03863, + "grad_norm": 0.6930312131073376, + "learning_rate": 0.003, + "loss": 4.1403, + "step": 3863 + }, + { + "epoch": 0.03864, + "grad_norm": 0.7338788764609174, + "learning_rate": 0.003, + "loss": 4.1481, + "step": 3864 + }, + { + "epoch": 0.03865, + "grad_norm": 0.716452591015328, + "learning_rate": 0.003, + "loss": 4.1614, + "step": 3865 + }, + { + "epoch": 0.03866, + "grad_norm": 0.8035841485064186, + "learning_rate": 0.003, + "loss": 4.1395, + "step": 3866 + }, + { + "epoch": 0.03867, + "grad_norm": 0.9790747131837212, + "learning_rate": 0.003, + "loss": 4.1305, + "step": 3867 + }, + { + "epoch": 0.03868, + "grad_norm": 0.9557607039135098, + "learning_rate": 0.003, + "loss": 4.1685, + "step": 3868 + }, + { + "epoch": 0.03869, + "grad_norm": 0.8269471619194556, + "learning_rate": 0.003, + "loss": 4.1558, + "step": 3869 + }, + { + "epoch": 0.0387, + "grad_norm": 0.736530154455655, + "learning_rate": 0.003, + "loss": 4.1642, + "step": 3870 + }, + { + "epoch": 0.03871, + "grad_norm": 0.5662896465927079, + "learning_rate": 0.003, + "loss": 4.1341, + "step": 3871 + }, + { + "epoch": 0.03872, + "grad_norm": 0.7064773963599414, + "learning_rate": 0.003, + "loss": 4.1325, + "step": 3872 + }, + { + "epoch": 0.03873, + "grad_norm": 0.7511440909675522, + "learning_rate": 0.003, + "loss": 4.1762, + "step": 3873 + }, + { + "epoch": 0.03874, + "grad_norm": 0.8087487769850815, + "learning_rate": 0.003, + "loss": 4.1496, + "step": 3874 + }, + { + "epoch": 0.03875, + "grad_norm": 0.840451180426337, + "learning_rate": 0.003, + "loss": 4.1553, + "step": 3875 + }, + { + "epoch": 0.03876, + "grad_norm": 0.8711005630005544, + "learning_rate": 0.003, + "loss": 4.1888, + "step": 3876 + }, + { + "epoch": 0.03877, + "grad_norm": 0.7489732856457578, + "learning_rate": 0.003, + "loss": 4.1645, + "step": 3877 + }, + { + "epoch": 0.03878, + "grad_norm": 0.6479494485701219, + "learning_rate": 0.003, + "loss": 4.1753, + "step": 3878 + }, + { + "epoch": 0.03879, + "grad_norm": 0.5974171547568563, + "learning_rate": 0.003, + "loss": 4.1755, + "step": 3879 + }, + { + "epoch": 0.0388, + "grad_norm": 0.6504260745071084, + "learning_rate": 0.003, + "loss": 4.141, + "step": 3880 + }, + { + "epoch": 0.03881, + "grad_norm": 0.6733083067264833, + "learning_rate": 0.003, + "loss": 4.1685, + "step": 3881 + }, + { + "epoch": 0.03882, + "grad_norm": 0.660030960264416, + "learning_rate": 0.003, + "loss": 4.1768, + "step": 3882 + }, + { + "epoch": 0.03883, + "grad_norm": 0.7215073977898953, + "learning_rate": 0.003, + "loss": 4.1491, + "step": 3883 + }, + { + "epoch": 0.03884, + "grad_norm": 0.6784371169912143, + "learning_rate": 0.003, + "loss": 4.1572, + "step": 3884 + }, + { + "epoch": 0.03885, + "grad_norm": 0.6961417852606488, + "learning_rate": 0.003, + "loss": 4.1747, + "step": 3885 + }, + { + "epoch": 0.03886, + "grad_norm": 0.6846762234353307, + "learning_rate": 0.003, + "loss": 4.1479, + "step": 3886 + }, + { + "epoch": 0.03887, + "grad_norm": 0.6018705177852628, + "learning_rate": 0.003, + "loss": 4.1575, + "step": 3887 + }, + { + "epoch": 0.03888, + "grad_norm": 0.7545299954550155, + "learning_rate": 0.003, + "loss": 4.138, + "step": 3888 + }, + { + "epoch": 0.03889, + "grad_norm": 0.9698514203515268, + "learning_rate": 0.003, + "loss": 4.1627, + "step": 3889 + }, + { + "epoch": 0.0389, + "grad_norm": 1.1649616934398974, + "learning_rate": 0.003, + "loss": 4.1644, + "step": 3890 + }, + { + "epoch": 0.03891, + "grad_norm": 0.7396396017755505, + "learning_rate": 0.003, + "loss": 4.1579, + "step": 3891 + }, + { + "epoch": 0.03892, + "grad_norm": 0.7127971514741257, + "learning_rate": 0.003, + "loss": 4.1754, + "step": 3892 + }, + { + "epoch": 0.03893, + "grad_norm": 0.8147903386909522, + "learning_rate": 0.003, + "loss": 4.1498, + "step": 3893 + }, + { + "epoch": 0.03894, + "grad_norm": 0.8553199932435573, + "learning_rate": 0.003, + "loss": 4.1639, + "step": 3894 + }, + { + "epoch": 0.03895, + "grad_norm": 0.8064783210870279, + "learning_rate": 0.003, + "loss": 4.1385, + "step": 3895 + }, + { + "epoch": 0.03896, + "grad_norm": 0.7327396202925885, + "learning_rate": 0.003, + "loss": 4.1635, + "step": 3896 + }, + { + "epoch": 0.03897, + "grad_norm": 0.6832165978250844, + "learning_rate": 0.003, + "loss": 4.1362, + "step": 3897 + }, + { + "epoch": 0.03898, + "grad_norm": 0.7381820210282796, + "learning_rate": 0.003, + "loss": 4.1722, + "step": 3898 + }, + { + "epoch": 0.03899, + "grad_norm": 0.7120798145498066, + "learning_rate": 0.003, + "loss": 4.163, + "step": 3899 + }, + { + "epoch": 0.039, + "grad_norm": 0.706935890187933, + "learning_rate": 0.003, + "loss": 4.1611, + "step": 3900 + }, + { + "epoch": 0.03901, + "grad_norm": 0.7172271819943744, + "learning_rate": 0.003, + "loss": 4.1754, + "step": 3901 + }, + { + "epoch": 0.03902, + "grad_norm": 0.6436464145708447, + "learning_rate": 0.003, + "loss": 4.1447, + "step": 3902 + }, + { + "epoch": 0.03903, + "grad_norm": 0.7120130287904197, + "learning_rate": 0.003, + "loss": 4.1839, + "step": 3903 + }, + { + "epoch": 0.03904, + "grad_norm": 0.7363979573377551, + "learning_rate": 0.003, + "loss": 4.1613, + "step": 3904 + }, + { + "epoch": 0.03905, + "grad_norm": 0.6989523475801707, + "learning_rate": 0.003, + "loss": 4.1777, + "step": 3905 + }, + { + "epoch": 0.03906, + "grad_norm": 0.7664511025672687, + "learning_rate": 0.003, + "loss": 4.1524, + "step": 3906 + }, + { + "epoch": 0.03907, + "grad_norm": 0.6973849941812954, + "learning_rate": 0.003, + "loss": 4.1633, + "step": 3907 + }, + { + "epoch": 0.03908, + "grad_norm": 0.6968835116178974, + "learning_rate": 0.003, + "loss": 4.1467, + "step": 3908 + }, + { + "epoch": 0.03909, + "grad_norm": 0.6214766339187717, + "learning_rate": 0.003, + "loss": 4.1479, + "step": 3909 + }, + { + "epoch": 0.0391, + "grad_norm": 0.606928879147604, + "learning_rate": 0.003, + "loss": 4.1783, + "step": 3910 + }, + { + "epoch": 0.03911, + "grad_norm": 0.6150917723982117, + "learning_rate": 0.003, + "loss": 4.1618, + "step": 3911 + }, + { + "epoch": 0.03912, + "grad_norm": 0.5315811614956929, + "learning_rate": 0.003, + "loss": 4.1576, + "step": 3912 + }, + { + "epoch": 0.03913, + "grad_norm": 0.49513392829810154, + "learning_rate": 0.003, + "loss": 4.1298, + "step": 3913 + }, + { + "epoch": 0.03914, + "grad_norm": 0.5316802449083329, + "learning_rate": 0.003, + "loss": 4.1655, + "step": 3914 + }, + { + "epoch": 0.03915, + "grad_norm": 0.5092855956655066, + "learning_rate": 0.003, + "loss": 4.1352, + "step": 3915 + }, + { + "epoch": 0.03916, + "grad_norm": 0.45924359376251017, + "learning_rate": 0.003, + "loss": 4.1598, + "step": 3916 + }, + { + "epoch": 0.03917, + "grad_norm": 0.45484869852430065, + "learning_rate": 0.003, + "loss": 4.1506, + "step": 3917 + }, + { + "epoch": 0.03918, + "grad_norm": 0.4501493787308128, + "learning_rate": 0.003, + "loss": 4.1358, + "step": 3918 + }, + { + "epoch": 0.03919, + "grad_norm": 0.5428853651303225, + "learning_rate": 0.003, + "loss": 4.1072, + "step": 3919 + }, + { + "epoch": 0.0392, + "grad_norm": 0.7284545482558801, + "learning_rate": 0.003, + "loss": 4.1372, + "step": 3920 + }, + { + "epoch": 0.03921, + "grad_norm": 0.9552348159276692, + "learning_rate": 0.003, + "loss": 4.1389, + "step": 3921 + }, + { + "epoch": 0.03922, + "grad_norm": 1.132432474415542, + "learning_rate": 0.003, + "loss": 4.1914, + "step": 3922 + }, + { + "epoch": 0.03923, + "grad_norm": 0.6932654094565733, + "learning_rate": 0.003, + "loss": 4.1356, + "step": 3923 + }, + { + "epoch": 0.03924, + "grad_norm": 0.6532999077661216, + "learning_rate": 0.003, + "loss": 4.1431, + "step": 3924 + }, + { + "epoch": 0.03925, + "grad_norm": 0.9135572051042363, + "learning_rate": 0.003, + "loss": 4.155, + "step": 3925 + }, + { + "epoch": 0.03926, + "grad_norm": 0.9923603322118285, + "learning_rate": 0.003, + "loss": 4.1508, + "step": 3926 + }, + { + "epoch": 0.03927, + "grad_norm": 0.9789133497360833, + "learning_rate": 0.003, + "loss": 4.168, + "step": 3927 + }, + { + "epoch": 0.03928, + "grad_norm": 0.8942775586441536, + "learning_rate": 0.003, + "loss": 4.1834, + "step": 3928 + }, + { + "epoch": 0.03929, + "grad_norm": 1.0319586305334827, + "learning_rate": 0.003, + "loss": 4.2001, + "step": 3929 + }, + { + "epoch": 0.0393, + "grad_norm": 1.0750467044406122, + "learning_rate": 0.003, + "loss": 4.1904, + "step": 3930 + }, + { + "epoch": 0.03931, + "grad_norm": 0.9184595530968939, + "learning_rate": 0.003, + "loss": 4.1919, + "step": 3931 + }, + { + "epoch": 0.03932, + "grad_norm": 0.7648894274462913, + "learning_rate": 0.003, + "loss": 4.1855, + "step": 3932 + }, + { + "epoch": 0.03933, + "grad_norm": 0.724112317823659, + "learning_rate": 0.003, + "loss": 4.172, + "step": 3933 + }, + { + "epoch": 0.03934, + "grad_norm": 0.6449540845441923, + "learning_rate": 0.003, + "loss": 4.1627, + "step": 3934 + }, + { + "epoch": 0.03935, + "grad_norm": 0.7490619555451586, + "learning_rate": 0.003, + "loss": 4.1618, + "step": 3935 + }, + { + "epoch": 0.03936, + "grad_norm": 0.7177217614636706, + "learning_rate": 0.003, + "loss": 4.1226, + "step": 3936 + }, + { + "epoch": 0.03937, + "grad_norm": 0.653893710648041, + "learning_rate": 0.003, + "loss": 4.1584, + "step": 3937 + }, + { + "epoch": 0.03938, + "grad_norm": 0.689030208208419, + "learning_rate": 0.003, + "loss": 4.1605, + "step": 3938 + }, + { + "epoch": 0.03939, + "grad_norm": 0.7098991679111472, + "learning_rate": 0.003, + "loss": 4.1446, + "step": 3939 + }, + { + "epoch": 0.0394, + "grad_norm": 0.6955232506989616, + "learning_rate": 0.003, + "loss": 4.1709, + "step": 3940 + }, + { + "epoch": 0.03941, + "grad_norm": 0.6598355292317952, + "learning_rate": 0.003, + "loss": 4.1607, + "step": 3941 + }, + { + "epoch": 0.03942, + "grad_norm": 0.738032728669511, + "learning_rate": 0.003, + "loss": 4.1785, + "step": 3942 + }, + { + "epoch": 0.03943, + "grad_norm": 0.8081766812397826, + "learning_rate": 0.003, + "loss": 4.1681, + "step": 3943 + }, + { + "epoch": 0.03944, + "grad_norm": 0.8593029564204311, + "learning_rate": 0.003, + "loss": 4.1857, + "step": 3944 + }, + { + "epoch": 0.03945, + "grad_norm": 0.8814910808518954, + "learning_rate": 0.003, + "loss": 4.1561, + "step": 3945 + }, + { + "epoch": 0.03946, + "grad_norm": 0.8816608783626361, + "learning_rate": 0.003, + "loss": 4.1852, + "step": 3946 + }, + { + "epoch": 0.03947, + "grad_norm": 0.8064135711097247, + "learning_rate": 0.003, + "loss": 4.1834, + "step": 3947 + }, + { + "epoch": 0.03948, + "grad_norm": 0.8804914835705606, + "learning_rate": 0.003, + "loss": 4.2055, + "step": 3948 + }, + { + "epoch": 0.03949, + "grad_norm": 0.8976280856029432, + "learning_rate": 0.003, + "loss": 4.1479, + "step": 3949 + }, + { + "epoch": 0.0395, + "grad_norm": 0.7915245308890937, + "learning_rate": 0.003, + "loss": 4.1755, + "step": 3950 + }, + { + "epoch": 0.03951, + "grad_norm": 0.7023597435152198, + "learning_rate": 0.003, + "loss": 4.2015, + "step": 3951 + }, + { + "epoch": 0.03952, + "grad_norm": 0.5970926719547551, + "learning_rate": 0.003, + "loss": 4.1776, + "step": 3952 + }, + { + "epoch": 0.03953, + "grad_norm": 0.5127193759525883, + "learning_rate": 0.003, + "loss": 4.1524, + "step": 3953 + }, + { + "epoch": 0.03954, + "grad_norm": 0.5083765022250437, + "learning_rate": 0.003, + "loss": 4.1525, + "step": 3954 + }, + { + "epoch": 0.03955, + "grad_norm": 0.5581234697293194, + "learning_rate": 0.003, + "loss": 4.1532, + "step": 3955 + }, + { + "epoch": 0.03956, + "grad_norm": 0.5875448440882535, + "learning_rate": 0.003, + "loss": 4.1566, + "step": 3956 + }, + { + "epoch": 0.03957, + "grad_norm": 0.597994191226493, + "learning_rate": 0.003, + "loss": 4.1624, + "step": 3957 + }, + { + "epoch": 0.03958, + "grad_norm": 0.5717890299511298, + "learning_rate": 0.003, + "loss": 4.1378, + "step": 3958 + }, + { + "epoch": 0.03959, + "grad_norm": 0.5436034530009164, + "learning_rate": 0.003, + "loss": 4.1361, + "step": 3959 + }, + { + "epoch": 0.0396, + "grad_norm": 0.49029266691650636, + "learning_rate": 0.003, + "loss": 4.1519, + "step": 3960 + }, + { + "epoch": 0.03961, + "grad_norm": 0.517401344634264, + "learning_rate": 0.003, + "loss": 4.1482, + "step": 3961 + }, + { + "epoch": 0.03962, + "grad_norm": 0.5330594146947688, + "learning_rate": 0.003, + "loss": 4.1598, + "step": 3962 + }, + { + "epoch": 0.03963, + "grad_norm": 0.5905667821025009, + "learning_rate": 0.003, + "loss": 4.1463, + "step": 3963 + }, + { + "epoch": 0.03964, + "grad_norm": 0.558494601973704, + "learning_rate": 0.003, + "loss": 4.1469, + "step": 3964 + }, + { + "epoch": 0.03965, + "grad_norm": 0.5647463812091732, + "learning_rate": 0.003, + "loss": 4.1433, + "step": 3965 + }, + { + "epoch": 0.03966, + "grad_norm": 0.5672438671895076, + "learning_rate": 0.003, + "loss": 4.1833, + "step": 3966 + }, + { + "epoch": 0.03967, + "grad_norm": 0.5990092369456635, + "learning_rate": 0.003, + "loss": 4.1392, + "step": 3967 + }, + { + "epoch": 0.03968, + "grad_norm": 0.5306380158511645, + "learning_rate": 0.003, + "loss": 4.1653, + "step": 3968 + }, + { + "epoch": 0.03969, + "grad_norm": 0.5335941140441545, + "learning_rate": 0.003, + "loss": 4.1448, + "step": 3969 + }, + { + "epoch": 0.0397, + "grad_norm": 0.5101263285016783, + "learning_rate": 0.003, + "loss": 4.1497, + "step": 3970 + }, + { + "epoch": 0.03971, + "grad_norm": 0.4861456325644423, + "learning_rate": 0.003, + "loss": 4.1253, + "step": 3971 + }, + { + "epoch": 0.03972, + "grad_norm": 0.5082221340172262, + "learning_rate": 0.003, + "loss": 4.1461, + "step": 3972 + }, + { + "epoch": 0.03973, + "grad_norm": 0.6064259836747726, + "learning_rate": 0.003, + "loss": 4.1523, + "step": 3973 + }, + { + "epoch": 0.03974, + "grad_norm": 0.8810228923861501, + "learning_rate": 0.003, + "loss": 4.1552, + "step": 3974 + }, + { + "epoch": 0.03975, + "grad_norm": 1.1594440991062704, + "learning_rate": 0.003, + "loss": 4.1581, + "step": 3975 + }, + { + "epoch": 0.03976, + "grad_norm": 0.6242516132147851, + "learning_rate": 0.003, + "loss": 4.1795, + "step": 3976 + }, + { + "epoch": 0.03977, + "grad_norm": 0.6310756758939645, + "learning_rate": 0.003, + "loss": 4.1651, + "step": 3977 + }, + { + "epoch": 0.03978, + "grad_norm": 0.9800843159523053, + "learning_rate": 0.003, + "loss": 4.1417, + "step": 3978 + }, + { + "epoch": 0.03979, + "grad_norm": 0.8894332385272683, + "learning_rate": 0.003, + "loss": 4.1646, + "step": 3979 + }, + { + "epoch": 0.0398, + "grad_norm": 0.5906456865918563, + "learning_rate": 0.003, + "loss": 4.1475, + "step": 3980 + }, + { + "epoch": 0.03981, + "grad_norm": 0.6582973485295706, + "learning_rate": 0.003, + "loss": 4.1444, + "step": 3981 + }, + { + "epoch": 0.03982, + "grad_norm": 0.7206583498121457, + "learning_rate": 0.003, + "loss": 4.1283, + "step": 3982 + }, + { + "epoch": 0.03983, + "grad_norm": 0.8701732888100308, + "learning_rate": 0.003, + "loss": 4.155, + "step": 3983 + }, + { + "epoch": 0.03984, + "grad_norm": 0.8661641192026968, + "learning_rate": 0.003, + "loss": 4.1668, + "step": 3984 + }, + { + "epoch": 0.03985, + "grad_norm": 0.9243378871553372, + "learning_rate": 0.003, + "loss": 4.1681, + "step": 3985 + }, + { + "epoch": 0.03986, + "grad_norm": 0.9956369167150667, + "learning_rate": 0.003, + "loss": 4.1705, + "step": 3986 + }, + { + "epoch": 0.03987, + "grad_norm": 0.8887499652438708, + "learning_rate": 0.003, + "loss": 4.1771, + "step": 3987 + }, + { + "epoch": 0.03988, + "grad_norm": 0.7725996597052387, + "learning_rate": 0.003, + "loss": 4.1764, + "step": 3988 + }, + { + "epoch": 0.03989, + "grad_norm": 0.9540752093113514, + "learning_rate": 0.003, + "loss": 4.1732, + "step": 3989 + }, + { + "epoch": 0.0399, + "grad_norm": 0.9866842358730207, + "learning_rate": 0.003, + "loss": 4.1507, + "step": 3990 + }, + { + "epoch": 0.03991, + "grad_norm": 1.0032653576960733, + "learning_rate": 0.003, + "loss": 4.1979, + "step": 3991 + }, + { + "epoch": 0.03992, + "grad_norm": 0.9799832060702593, + "learning_rate": 0.003, + "loss": 4.1809, + "step": 3992 + }, + { + "epoch": 0.03993, + "grad_norm": 0.9143397893823952, + "learning_rate": 0.003, + "loss": 4.1772, + "step": 3993 + }, + { + "epoch": 0.03994, + "grad_norm": 0.8867228330129578, + "learning_rate": 0.003, + "loss": 4.1785, + "step": 3994 + }, + { + "epoch": 0.03995, + "grad_norm": 0.9777127443180489, + "learning_rate": 0.003, + "loss": 4.1672, + "step": 3995 + }, + { + "epoch": 0.03996, + "grad_norm": 1.121916971358567, + "learning_rate": 0.003, + "loss": 4.2091, + "step": 3996 + }, + { + "epoch": 0.03997, + "grad_norm": 1.0260835587044206, + "learning_rate": 0.003, + "loss": 4.21, + "step": 3997 + }, + { + "epoch": 0.03998, + "grad_norm": 0.9733483073411772, + "learning_rate": 0.003, + "loss": 4.2091, + "step": 3998 + }, + { + "epoch": 0.03999, + "grad_norm": 0.9253787235141333, + "learning_rate": 0.003, + "loss": 4.2098, + "step": 3999 + }, + { + "epoch": 0.04, + "grad_norm": 0.860383401913097, + "learning_rate": 0.003, + "loss": 4.2101, + "step": 4000 + }, + { + "epoch": 0.04001, + "grad_norm": 0.8527166028367188, + "learning_rate": 0.003, + "loss": 4.2089, + "step": 4001 + }, + { + "epoch": 0.04002, + "grad_norm": 0.7955822227821913, + "learning_rate": 0.003, + "loss": 4.1762, + "step": 4002 + }, + { + "epoch": 0.04003, + "grad_norm": 0.8159059659987256, + "learning_rate": 0.003, + "loss": 4.1966, + "step": 4003 + }, + { + "epoch": 0.04004, + "grad_norm": 0.8242608605578613, + "learning_rate": 0.003, + "loss": 4.2014, + "step": 4004 + }, + { + "epoch": 0.04005, + "grad_norm": 0.7442104756929572, + "learning_rate": 0.003, + "loss": 4.2052, + "step": 4005 + }, + { + "epoch": 0.04006, + "grad_norm": 0.6862974810950646, + "learning_rate": 0.003, + "loss": 4.1579, + "step": 4006 + }, + { + "epoch": 0.04007, + "grad_norm": 0.6713665895022044, + "learning_rate": 0.003, + "loss": 4.1851, + "step": 4007 + }, + { + "epoch": 0.04008, + "grad_norm": 0.6515792038366235, + "learning_rate": 0.003, + "loss": 4.1589, + "step": 4008 + }, + { + "epoch": 0.04009, + "grad_norm": 0.6670829454786272, + "learning_rate": 0.003, + "loss": 4.1561, + "step": 4009 + }, + { + "epoch": 0.0401, + "grad_norm": 0.5890722285246915, + "learning_rate": 0.003, + "loss": 4.1713, + "step": 4010 + }, + { + "epoch": 0.04011, + "grad_norm": 0.5154042218197626, + "learning_rate": 0.003, + "loss": 4.1727, + "step": 4011 + }, + { + "epoch": 0.04012, + "grad_norm": 0.511904267386738, + "learning_rate": 0.003, + "loss": 4.1666, + "step": 4012 + }, + { + "epoch": 0.04013, + "grad_norm": 0.4301130431551674, + "learning_rate": 0.003, + "loss": 4.1595, + "step": 4013 + }, + { + "epoch": 0.04014, + "grad_norm": 0.4755384019802336, + "learning_rate": 0.003, + "loss": 4.145, + "step": 4014 + }, + { + "epoch": 0.04015, + "grad_norm": 0.4526642613482027, + "learning_rate": 0.003, + "loss": 4.1259, + "step": 4015 + }, + { + "epoch": 0.04016, + "grad_norm": 0.5025871602100715, + "learning_rate": 0.003, + "loss": 4.1428, + "step": 4016 + }, + { + "epoch": 0.04017, + "grad_norm": 0.613279269884007, + "learning_rate": 0.003, + "loss": 4.1847, + "step": 4017 + }, + { + "epoch": 0.04018, + "grad_norm": 0.8281359691209577, + "learning_rate": 0.003, + "loss": 4.1594, + "step": 4018 + }, + { + "epoch": 0.04019, + "grad_norm": 0.941035832708172, + "learning_rate": 0.003, + "loss": 4.1434, + "step": 4019 + }, + { + "epoch": 0.0402, + "grad_norm": 0.8795192610150039, + "learning_rate": 0.003, + "loss": 4.1785, + "step": 4020 + }, + { + "epoch": 0.04021, + "grad_norm": 0.6515100017542129, + "learning_rate": 0.003, + "loss": 4.1464, + "step": 4021 + }, + { + "epoch": 0.04022, + "grad_norm": 0.5063361011739121, + "learning_rate": 0.003, + "loss": 4.1404, + "step": 4022 + }, + { + "epoch": 0.04023, + "grad_norm": 0.6481572855287395, + "learning_rate": 0.003, + "loss": 4.1413, + "step": 4023 + }, + { + "epoch": 0.04024, + "grad_norm": 0.7010903711503769, + "learning_rate": 0.003, + "loss": 4.131, + "step": 4024 + }, + { + "epoch": 0.04025, + "grad_norm": 0.7840447772680712, + "learning_rate": 0.003, + "loss": 4.1399, + "step": 4025 + }, + { + "epoch": 0.04026, + "grad_norm": 0.7341068822639932, + "learning_rate": 0.003, + "loss": 4.1655, + "step": 4026 + }, + { + "epoch": 0.04027, + "grad_norm": 0.6238472142782437, + "learning_rate": 0.003, + "loss": 4.1663, + "step": 4027 + }, + { + "epoch": 0.04028, + "grad_norm": 0.4935095791712522, + "learning_rate": 0.003, + "loss": 4.1428, + "step": 4028 + }, + { + "epoch": 0.04029, + "grad_norm": 0.4905704871714714, + "learning_rate": 0.003, + "loss": 4.1713, + "step": 4029 + }, + { + "epoch": 0.0403, + "grad_norm": 0.5059882465087885, + "learning_rate": 0.003, + "loss": 4.1199, + "step": 4030 + }, + { + "epoch": 0.04031, + "grad_norm": 0.49185506270358387, + "learning_rate": 0.003, + "loss": 4.1374, + "step": 4031 + }, + { + "epoch": 0.04032, + "grad_norm": 0.5505039809907691, + "learning_rate": 0.003, + "loss": 4.1625, + "step": 4032 + }, + { + "epoch": 0.04033, + "grad_norm": 0.5349055922421693, + "learning_rate": 0.003, + "loss": 4.1577, + "step": 4033 + }, + { + "epoch": 0.04034, + "grad_norm": 0.4887336563792077, + "learning_rate": 0.003, + "loss": 4.1256, + "step": 4034 + }, + { + "epoch": 0.04035, + "grad_norm": 0.5026476582372887, + "learning_rate": 0.003, + "loss": 4.1498, + "step": 4035 + }, + { + "epoch": 0.04036, + "grad_norm": 0.5741733182678186, + "learning_rate": 0.003, + "loss": 4.1638, + "step": 4036 + }, + { + "epoch": 0.04037, + "grad_norm": 0.6338724243857007, + "learning_rate": 0.003, + "loss": 4.1187, + "step": 4037 + }, + { + "epoch": 0.04038, + "grad_norm": 0.5866801909982847, + "learning_rate": 0.003, + "loss": 4.1345, + "step": 4038 + }, + { + "epoch": 0.04039, + "grad_norm": 0.4469263176133214, + "learning_rate": 0.003, + "loss": 4.1291, + "step": 4039 + }, + { + "epoch": 0.0404, + "grad_norm": 0.5203352096732194, + "learning_rate": 0.003, + "loss": 4.1298, + "step": 4040 + }, + { + "epoch": 0.04041, + "grad_norm": 0.5903209349905957, + "learning_rate": 0.003, + "loss": 4.1276, + "step": 4041 + }, + { + "epoch": 0.04042, + "grad_norm": 0.7850173938632014, + "learning_rate": 0.003, + "loss": 4.1729, + "step": 4042 + }, + { + "epoch": 0.04043, + "grad_norm": 1.0374432013836312, + "learning_rate": 0.003, + "loss": 4.1727, + "step": 4043 + }, + { + "epoch": 0.04044, + "grad_norm": 1.0008411828594295, + "learning_rate": 0.003, + "loss": 4.1662, + "step": 4044 + }, + { + "epoch": 0.04045, + "grad_norm": 0.721767328260429, + "learning_rate": 0.003, + "loss": 4.1048, + "step": 4045 + }, + { + "epoch": 0.04046, + "grad_norm": 0.6027053735164074, + "learning_rate": 0.003, + "loss": 4.1496, + "step": 4046 + }, + { + "epoch": 0.04047, + "grad_norm": 0.8146571762133998, + "learning_rate": 0.003, + "loss": 4.1466, + "step": 4047 + }, + { + "epoch": 0.04048, + "grad_norm": 0.8947220522958645, + "learning_rate": 0.003, + "loss": 4.1436, + "step": 4048 + }, + { + "epoch": 0.04049, + "grad_norm": 0.9307304863945178, + "learning_rate": 0.003, + "loss": 4.1891, + "step": 4049 + }, + { + "epoch": 0.0405, + "grad_norm": 0.7045766049752963, + "learning_rate": 0.003, + "loss": 4.178, + "step": 4050 + }, + { + "epoch": 0.04051, + "grad_norm": 0.6846375832757631, + "learning_rate": 0.003, + "loss": 4.1564, + "step": 4051 + }, + { + "epoch": 0.04052, + "grad_norm": 0.8615665313113385, + "learning_rate": 0.003, + "loss": 4.1722, + "step": 4052 + }, + { + "epoch": 0.04053, + "grad_norm": 0.9373878770279569, + "learning_rate": 0.003, + "loss": 4.125, + "step": 4053 + }, + { + "epoch": 0.04054, + "grad_norm": 0.8371593982203549, + "learning_rate": 0.003, + "loss": 4.1443, + "step": 4054 + }, + { + "epoch": 0.04055, + "grad_norm": 0.6967664819995452, + "learning_rate": 0.003, + "loss": 4.138, + "step": 4055 + }, + { + "epoch": 0.04056, + "grad_norm": 0.6741494833756686, + "learning_rate": 0.003, + "loss": 4.1595, + "step": 4056 + }, + { + "epoch": 0.04057, + "grad_norm": 0.8030676782045835, + "learning_rate": 0.003, + "loss": 4.1632, + "step": 4057 + }, + { + "epoch": 0.04058, + "grad_norm": 0.774339384731219, + "learning_rate": 0.003, + "loss": 4.1338, + "step": 4058 + }, + { + "epoch": 0.04059, + "grad_norm": 0.8430830883517224, + "learning_rate": 0.003, + "loss": 4.1192, + "step": 4059 + }, + { + "epoch": 0.0406, + "grad_norm": 0.6742657640683123, + "learning_rate": 0.003, + "loss": 4.1739, + "step": 4060 + }, + { + "epoch": 0.04061, + "grad_norm": 0.6572926238853734, + "learning_rate": 0.003, + "loss": 4.1598, + "step": 4061 + }, + { + "epoch": 0.04062, + "grad_norm": 0.5997992092489303, + "learning_rate": 0.003, + "loss": 4.1368, + "step": 4062 + }, + { + "epoch": 0.04063, + "grad_norm": 0.5451542397429993, + "learning_rate": 0.003, + "loss": 4.1422, + "step": 4063 + }, + { + "epoch": 0.04064, + "grad_norm": 0.5706051829517365, + "learning_rate": 0.003, + "loss": 4.1561, + "step": 4064 + }, + { + "epoch": 0.04065, + "grad_norm": 0.6573749206271802, + "learning_rate": 0.003, + "loss": 4.1936, + "step": 4065 + }, + { + "epoch": 0.04066, + "grad_norm": 0.6711377067040638, + "learning_rate": 0.003, + "loss": 4.1533, + "step": 4066 + }, + { + "epoch": 0.04067, + "grad_norm": 0.6151905993857413, + "learning_rate": 0.003, + "loss": 4.1541, + "step": 4067 + }, + { + "epoch": 0.04068, + "grad_norm": 0.610139511076282, + "learning_rate": 0.003, + "loss": 4.1627, + "step": 4068 + }, + { + "epoch": 0.04069, + "grad_norm": 0.6435562548550586, + "learning_rate": 0.003, + "loss": 4.153, + "step": 4069 + }, + { + "epoch": 0.0407, + "grad_norm": 0.5473010440748541, + "learning_rate": 0.003, + "loss": 4.1547, + "step": 4070 + }, + { + "epoch": 0.04071, + "grad_norm": 0.601566675296323, + "learning_rate": 0.003, + "loss": 4.1223, + "step": 4071 + }, + { + "epoch": 0.04072, + "grad_norm": 0.6509579069490173, + "learning_rate": 0.003, + "loss": 4.146, + "step": 4072 + }, + { + "epoch": 0.04073, + "grad_norm": 0.5973245201602951, + "learning_rate": 0.003, + "loss": 4.1302, + "step": 4073 + }, + { + "epoch": 0.04074, + "grad_norm": 0.607520676438575, + "learning_rate": 0.003, + "loss": 4.1574, + "step": 4074 + }, + { + "epoch": 0.04075, + "grad_norm": 0.8288720279764606, + "learning_rate": 0.003, + "loss": 4.1612, + "step": 4075 + }, + { + "epoch": 0.04076, + "grad_norm": 0.9232154672269041, + "learning_rate": 0.003, + "loss": 4.1605, + "step": 4076 + }, + { + "epoch": 0.04077, + "grad_norm": 0.9604554082892389, + "learning_rate": 0.003, + "loss": 4.1752, + "step": 4077 + }, + { + "epoch": 0.04078, + "grad_norm": 0.9176998486040583, + "learning_rate": 0.003, + "loss": 4.1501, + "step": 4078 + }, + { + "epoch": 0.04079, + "grad_norm": 0.7781092418368329, + "learning_rate": 0.003, + "loss": 4.1586, + "step": 4079 + }, + { + "epoch": 0.0408, + "grad_norm": 0.7144390148261879, + "learning_rate": 0.003, + "loss": 4.1488, + "step": 4080 + }, + { + "epoch": 0.04081, + "grad_norm": 0.7985914865590669, + "learning_rate": 0.003, + "loss": 4.1329, + "step": 4081 + }, + { + "epoch": 0.04082, + "grad_norm": 0.8262338793037239, + "learning_rate": 0.003, + "loss": 4.1699, + "step": 4082 + }, + { + "epoch": 0.04083, + "grad_norm": 0.7465376916541703, + "learning_rate": 0.003, + "loss": 4.1459, + "step": 4083 + }, + { + "epoch": 0.04084, + "grad_norm": 0.6728219238709465, + "learning_rate": 0.003, + "loss": 4.1811, + "step": 4084 + }, + { + "epoch": 0.04085, + "grad_norm": 0.755821930925202, + "learning_rate": 0.003, + "loss": 4.163, + "step": 4085 + }, + { + "epoch": 0.04086, + "grad_norm": 0.637261349390268, + "learning_rate": 0.003, + "loss": 4.1791, + "step": 4086 + }, + { + "epoch": 0.04087, + "grad_norm": 0.5906659081612129, + "learning_rate": 0.003, + "loss": 4.1597, + "step": 4087 + }, + { + "epoch": 0.04088, + "grad_norm": 0.547995647785304, + "learning_rate": 0.003, + "loss": 4.1366, + "step": 4088 + }, + { + "epoch": 0.04089, + "grad_norm": 0.5708520101184565, + "learning_rate": 0.003, + "loss": 4.1354, + "step": 4089 + }, + { + "epoch": 0.0409, + "grad_norm": 0.6099350559443698, + "learning_rate": 0.003, + "loss": 4.1539, + "step": 4090 + }, + { + "epoch": 0.04091, + "grad_norm": 0.696520536628286, + "learning_rate": 0.003, + "loss": 4.1662, + "step": 4091 + }, + { + "epoch": 0.04092, + "grad_norm": 0.8551793129611549, + "learning_rate": 0.003, + "loss": 4.129, + "step": 4092 + }, + { + "epoch": 0.04093, + "grad_norm": 1.0511160284337668, + "learning_rate": 0.003, + "loss": 4.1606, + "step": 4093 + }, + { + "epoch": 0.04094, + "grad_norm": 0.9303098870853981, + "learning_rate": 0.003, + "loss": 4.151, + "step": 4094 + }, + { + "epoch": 0.04095, + "grad_norm": 0.7394320217083529, + "learning_rate": 0.003, + "loss": 4.1258, + "step": 4095 + }, + { + "epoch": 0.04096, + "grad_norm": 0.7856690666821323, + "learning_rate": 0.003, + "loss": 4.1325, + "step": 4096 + }, + { + "epoch": 0.04097, + "grad_norm": 0.7724952999564255, + "learning_rate": 0.003, + "loss": 4.1507, + "step": 4097 + }, + { + "epoch": 0.04098, + "grad_norm": 0.6418338461235239, + "learning_rate": 0.003, + "loss": 4.1525, + "step": 4098 + }, + { + "epoch": 0.04099, + "grad_norm": 0.6141872194967258, + "learning_rate": 0.003, + "loss": 4.1751, + "step": 4099 + }, + { + "epoch": 0.041, + "grad_norm": 0.6337061381438701, + "learning_rate": 0.003, + "loss": 4.1314, + "step": 4100 + }, + { + "epoch": 0.04101, + "grad_norm": 0.7005605990414404, + "learning_rate": 0.003, + "loss": 4.129, + "step": 4101 + }, + { + "epoch": 0.04102, + "grad_norm": 0.7781518273799513, + "learning_rate": 0.003, + "loss": 4.151, + "step": 4102 + }, + { + "epoch": 0.04103, + "grad_norm": 0.7717092605525707, + "learning_rate": 0.003, + "loss": 4.1769, + "step": 4103 + }, + { + "epoch": 0.04104, + "grad_norm": 0.7998045960483534, + "learning_rate": 0.003, + "loss": 4.1621, + "step": 4104 + }, + { + "epoch": 0.04105, + "grad_norm": 0.7266397106851961, + "learning_rate": 0.003, + "loss": 4.1275, + "step": 4105 + }, + { + "epoch": 0.04106, + "grad_norm": 0.7375729801536005, + "learning_rate": 0.003, + "loss": 4.1665, + "step": 4106 + }, + { + "epoch": 0.04107, + "grad_norm": 0.8517602436934472, + "learning_rate": 0.003, + "loss": 4.1672, + "step": 4107 + }, + { + "epoch": 0.04108, + "grad_norm": 0.7155354553416942, + "learning_rate": 0.003, + "loss": 4.1487, + "step": 4108 + }, + { + "epoch": 0.04109, + "grad_norm": 0.8472831379025649, + "learning_rate": 0.003, + "loss": 4.1412, + "step": 4109 + }, + { + "epoch": 0.0411, + "grad_norm": 0.9994412956163683, + "learning_rate": 0.003, + "loss": 4.1607, + "step": 4110 + }, + { + "epoch": 0.04111, + "grad_norm": 1.0865359108420907, + "learning_rate": 0.003, + "loss": 4.1499, + "step": 4111 + }, + { + "epoch": 0.04112, + "grad_norm": 0.7944687116955357, + "learning_rate": 0.003, + "loss": 4.1621, + "step": 4112 + }, + { + "epoch": 0.04113, + "grad_norm": 0.6695248184455779, + "learning_rate": 0.003, + "loss": 4.2009, + "step": 4113 + }, + { + "epoch": 0.04114, + "grad_norm": 0.8304852025142981, + "learning_rate": 0.003, + "loss": 4.1456, + "step": 4114 + }, + { + "epoch": 0.04115, + "grad_norm": 0.9696455727697199, + "learning_rate": 0.003, + "loss": 4.164, + "step": 4115 + }, + { + "epoch": 0.04116, + "grad_norm": 1.0300022050612345, + "learning_rate": 0.003, + "loss": 4.1831, + "step": 4116 + }, + { + "epoch": 0.04117, + "grad_norm": 0.7267792671268358, + "learning_rate": 0.003, + "loss": 4.1373, + "step": 4117 + }, + { + "epoch": 0.04118, + "grad_norm": 0.6123016813019814, + "learning_rate": 0.003, + "loss": 4.1925, + "step": 4118 + }, + { + "epoch": 0.04119, + "grad_norm": 0.6465840606735817, + "learning_rate": 0.003, + "loss": 4.1697, + "step": 4119 + }, + { + "epoch": 0.0412, + "grad_norm": 0.7384604480296841, + "learning_rate": 0.003, + "loss": 4.1608, + "step": 4120 + }, + { + "epoch": 0.04121, + "grad_norm": 0.6577587889661327, + "learning_rate": 0.003, + "loss": 4.1798, + "step": 4121 + }, + { + "epoch": 0.04122, + "grad_norm": 0.5559068248334778, + "learning_rate": 0.003, + "loss": 4.1519, + "step": 4122 + }, + { + "epoch": 0.04123, + "grad_norm": 0.596362393082762, + "learning_rate": 0.003, + "loss": 4.1613, + "step": 4123 + }, + { + "epoch": 0.04124, + "grad_norm": 0.5443723484819623, + "learning_rate": 0.003, + "loss": 4.1386, + "step": 4124 + }, + { + "epoch": 0.04125, + "grad_norm": 0.5563286329797662, + "learning_rate": 0.003, + "loss": 4.1617, + "step": 4125 + }, + { + "epoch": 0.04126, + "grad_norm": 0.523384653689282, + "learning_rate": 0.003, + "loss": 4.1506, + "step": 4126 + }, + { + "epoch": 0.04127, + "grad_norm": 0.4495176170519381, + "learning_rate": 0.003, + "loss": 4.1384, + "step": 4127 + }, + { + "epoch": 0.04128, + "grad_norm": 0.5072318783524735, + "learning_rate": 0.003, + "loss": 4.1231, + "step": 4128 + }, + { + "epoch": 0.04129, + "grad_norm": 0.5052405740035331, + "learning_rate": 0.003, + "loss": 4.1627, + "step": 4129 + }, + { + "epoch": 0.0413, + "grad_norm": 0.575809172371195, + "learning_rate": 0.003, + "loss": 4.1293, + "step": 4130 + }, + { + "epoch": 0.04131, + "grad_norm": 0.6143169354531615, + "learning_rate": 0.003, + "loss": 4.155, + "step": 4131 + }, + { + "epoch": 0.04132, + "grad_norm": 0.6223256630595382, + "learning_rate": 0.003, + "loss": 4.1196, + "step": 4132 + }, + { + "epoch": 0.04133, + "grad_norm": 0.5894212387409717, + "learning_rate": 0.003, + "loss": 4.161, + "step": 4133 + }, + { + "epoch": 0.04134, + "grad_norm": 0.5069566465455504, + "learning_rate": 0.003, + "loss": 4.1487, + "step": 4134 + }, + { + "epoch": 0.04135, + "grad_norm": 0.47827820618941536, + "learning_rate": 0.003, + "loss": 4.1296, + "step": 4135 + }, + { + "epoch": 0.04136, + "grad_norm": 0.4653725022496872, + "learning_rate": 0.003, + "loss": 4.1252, + "step": 4136 + }, + { + "epoch": 0.04137, + "grad_norm": 0.47773052230780755, + "learning_rate": 0.003, + "loss": 4.1658, + "step": 4137 + }, + { + "epoch": 0.04138, + "grad_norm": 0.5380882786268274, + "learning_rate": 0.003, + "loss": 4.15, + "step": 4138 + }, + { + "epoch": 0.04139, + "grad_norm": 0.6469827605874696, + "learning_rate": 0.003, + "loss": 4.1496, + "step": 4139 + }, + { + "epoch": 0.0414, + "grad_norm": 0.6947344870174545, + "learning_rate": 0.003, + "loss": 4.1555, + "step": 4140 + }, + { + "epoch": 0.04141, + "grad_norm": 0.748345923041137, + "learning_rate": 0.003, + "loss": 4.1318, + "step": 4141 + }, + { + "epoch": 0.04142, + "grad_norm": 0.7319349091665185, + "learning_rate": 0.003, + "loss": 4.157, + "step": 4142 + }, + { + "epoch": 0.04143, + "grad_norm": 0.628358425706577, + "learning_rate": 0.003, + "loss": 4.1389, + "step": 4143 + }, + { + "epoch": 0.04144, + "grad_norm": 0.6834841445843723, + "learning_rate": 0.003, + "loss": 4.1412, + "step": 4144 + }, + { + "epoch": 0.04145, + "grad_norm": 0.7376058073772749, + "learning_rate": 0.003, + "loss": 4.1295, + "step": 4145 + }, + { + "epoch": 0.04146, + "grad_norm": 0.7587407207350676, + "learning_rate": 0.003, + "loss": 4.1526, + "step": 4146 + }, + { + "epoch": 0.04147, + "grad_norm": 0.7947181963356904, + "learning_rate": 0.003, + "loss": 4.1641, + "step": 4147 + }, + { + "epoch": 0.04148, + "grad_norm": 0.7182381427435356, + "learning_rate": 0.003, + "loss": 4.1204, + "step": 4148 + }, + { + "epoch": 0.04149, + "grad_norm": 0.7323007766800972, + "learning_rate": 0.003, + "loss": 4.1348, + "step": 4149 + }, + { + "epoch": 0.0415, + "grad_norm": 0.857281738554918, + "learning_rate": 0.003, + "loss": 4.1422, + "step": 4150 + }, + { + "epoch": 0.04151, + "grad_norm": 0.9753058360121778, + "learning_rate": 0.003, + "loss": 4.1696, + "step": 4151 + }, + { + "epoch": 0.04152, + "grad_norm": 1.0237271601242721, + "learning_rate": 0.003, + "loss": 4.1707, + "step": 4152 + }, + { + "epoch": 0.04153, + "grad_norm": 0.9306677581318379, + "learning_rate": 0.003, + "loss": 4.135, + "step": 4153 + }, + { + "epoch": 0.04154, + "grad_norm": 0.8382038016807482, + "learning_rate": 0.003, + "loss": 4.1686, + "step": 4154 + }, + { + "epoch": 0.04155, + "grad_norm": 0.7006562175213897, + "learning_rate": 0.003, + "loss": 4.1385, + "step": 4155 + }, + { + "epoch": 0.04156, + "grad_norm": 0.6295351230747147, + "learning_rate": 0.003, + "loss": 4.1592, + "step": 4156 + }, + { + "epoch": 0.04157, + "grad_norm": 0.7233308933121653, + "learning_rate": 0.003, + "loss": 4.1755, + "step": 4157 + }, + { + "epoch": 0.04158, + "grad_norm": 0.7870559063957332, + "learning_rate": 0.003, + "loss": 4.1784, + "step": 4158 + }, + { + "epoch": 0.04159, + "grad_norm": 0.9866760938792348, + "learning_rate": 0.003, + "loss": 4.1559, + "step": 4159 + }, + { + "epoch": 0.0416, + "grad_norm": 1.0470738285698136, + "learning_rate": 0.003, + "loss": 4.1774, + "step": 4160 + }, + { + "epoch": 0.04161, + "grad_norm": 0.9307304193810675, + "learning_rate": 0.003, + "loss": 4.1786, + "step": 4161 + }, + { + "epoch": 0.04162, + "grad_norm": 0.8721482244521436, + "learning_rate": 0.003, + "loss": 4.1406, + "step": 4162 + }, + { + "epoch": 0.04163, + "grad_norm": 0.8678721020792026, + "learning_rate": 0.003, + "loss": 4.1682, + "step": 4163 + }, + { + "epoch": 0.04164, + "grad_norm": 0.8226440605014521, + "learning_rate": 0.003, + "loss": 4.1918, + "step": 4164 + }, + { + "epoch": 0.04165, + "grad_norm": 0.825975020523602, + "learning_rate": 0.003, + "loss": 4.1651, + "step": 4165 + }, + { + "epoch": 0.04166, + "grad_norm": 0.8820368404280395, + "learning_rate": 0.003, + "loss": 4.1594, + "step": 4166 + }, + { + "epoch": 0.04167, + "grad_norm": 0.8857197236082065, + "learning_rate": 0.003, + "loss": 4.1834, + "step": 4167 + }, + { + "epoch": 0.04168, + "grad_norm": 0.9094188409333465, + "learning_rate": 0.003, + "loss": 4.1886, + "step": 4168 + }, + { + "epoch": 0.04169, + "grad_norm": 0.8608115262399164, + "learning_rate": 0.003, + "loss": 4.1926, + "step": 4169 + }, + { + "epoch": 0.0417, + "grad_norm": 0.8339384653513617, + "learning_rate": 0.003, + "loss": 4.1527, + "step": 4170 + }, + { + "epoch": 0.04171, + "grad_norm": 0.8053429340063298, + "learning_rate": 0.003, + "loss": 4.1752, + "step": 4171 + }, + { + "epoch": 0.04172, + "grad_norm": 0.8778996219659019, + "learning_rate": 0.003, + "loss": 4.1968, + "step": 4172 + }, + { + "epoch": 0.04173, + "grad_norm": 0.9642055884954371, + "learning_rate": 0.003, + "loss": 4.1939, + "step": 4173 + }, + { + "epoch": 0.04174, + "grad_norm": 1.0380118480990967, + "learning_rate": 0.003, + "loss": 4.1636, + "step": 4174 + }, + { + "epoch": 0.04175, + "grad_norm": 0.9536264797990276, + "learning_rate": 0.003, + "loss": 4.2107, + "step": 4175 + }, + { + "epoch": 0.04176, + "grad_norm": 0.8997005452165456, + "learning_rate": 0.003, + "loss": 4.1724, + "step": 4176 + }, + { + "epoch": 0.04177, + "grad_norm": 0.9315086547062951, + "learning_rate": 0.003, + "loss": 4.151, + "step": 4177 + }, + { + "epoch": 0.04178, + "grad_norm": 0.7486982272609372, + "learning_rate": 0.003, + "loss": 4.1756, + "step": 4178 + }, + { + "epoch": 0.04179, + "grad_norm": 0.7322985384737938, + "learning_rate": 0.003, + "loss": 4.1539, + "step": 4179 + }, + { + "epoch": 0.0418, + "grad_norm": 0.6634118485342885, + "learning_rate": 0.003, + "loss": 4.1634, + "step": 4180 + }, + { + "epoch": 0.04181, + "grad_norm": 0.6257131964317331, + "learning_rate": 0.003, + "loss": 4.1609, + "step": 4181 + }, + { + "epoch": 0.04182, + "grad_norm": 0.6155966650745514, + "learning_rate": 0.003, + "loss": 4.1442, + "step": 4182 + }, + { + "epoch": 0.04183, + "grad_norm": 0.5888122942739422, + "learning_rate": 0.003, + "loss": 4.147, + "step": 4183 + }, + { + "epoch": 0.04184, + "grad_norm": 0.48396597336103814, + "learning_rate": 0.003, + "loss": 4.1975, + "step": 4184 + }, + { + "epoch": 0.04185, + "grad_norm": 0.46085090618271063, + "learning_rate": 0.003, + "loss": 4.1762, + "step": 4185 + }, + { + "epoch": 0.04186, + "grad_norm": 0.40263981014725103, + "learning_rate": 0.003, + "loss": 4.1172, + "step": 4186 + }, + { + "epoch": 0.04187, + "grad_norm": 0.4072854325073614, + "learning_rate": 0.003, + "loss": 4.1673, + "step": 4187 + }, + { + "epoch": 0.04188, + "grad_norm": 0.3949031488342765, + "learning_rate": 0.003, + "loss": 4.129, + "step": 4188 + }, + { + "epoch": 0.04189, + "grad_norm": 0.4614400695553244, + "learning_rate": 0.003, + "loss": 4.1471, + "step": 4189 + }, + { + "epoch": 0.0419, + "grad_norm": 0.591341946932479, + "learning_rate": 0.003, + "loss": 4.1422, + "step": 4190 + }, + { + "epoch": 0.04191, + "grad_norm": 0.6794599625720497, + "learning_rate": 0.003, + "loss": 4.1502, + "step": 4191 + }, + { + "epoch": 0.04192, + "grad_norm": 0.8639843192836074, + "learning_rate": 0.003, + "loss": 4.1353, + "step": 4192 + }, + { + "epoch": 0.04193, + "grad_norm": 0.9261826031541774, + "learning_rate": 0.003, + "loss": 4.1201, + "step": 4193 + }, + { + "epoch": 0.04194, + "grad_norm": 0.6816042869205202, + "learning_rate": 0.003, + "loss": 4.1131, + "step": 4194 + }, + { + "epoch": 0.04195, + "grad_norm": 0.4846458487192246, + "learning_rate": 0.003, + "loss": 4.1594, + "step": 4195 + }, + { + "epoch": 0.04196, + "grad_norm": 0.6733457250112305, + "learning_rate": 0.003, + "loss": 4.1205, + "step": 4196 + }, + { + "epoch": 0.04197, + "grad_norm": 0.8628147041804719, + "learning_rate": 0.003, + "loss": 4.1368, + "step": 4197 + }, + { + "epoch": 0.04198, + "grad_norm": 0.8069270994807441, + "learning_rate": 0.003, + "loss": 4.1332, + "step": 4198 + }, + { + "epoch": 0.04199, + "grad_norm": 0.5843819059317946, + "learning_rate": 0.003, + "loss": 4.1649, + "step": 4199 + }, + { + "epoch": 0.042, + "grad_norm": 0.6524001973236165, + "learning_rate": 0.003, + "loss": 4.1484, + "step": 4200 + }, + { + "epoch": 0.04201, + "grad_norm": 0.7455750389366074, + "learning_rate": 0.003, + "loss": 4.1283, + "step": 4201 + }, + { + "epoch": 0.04202, + "grad_norm": 0.8073311014615949, + "learning_rate": 0.003, + "loss": 4.1482, + "step": 4202 + }, + { + "epoch": 0.04203, + "grad_norm": 0.6797266710402026, + "learning_rate": 0.003, + "loss": 4.165, + "step": 4203 + }, + { + "epoch": 0.04204, + "grad_norm": 0.6818305150769544, + "learning_rate": 0.003, + "loss": 4.1288, + "step": 4204 + }, + { + "epoch": 0.04205, + "grad_norm": 0.5915119548922819, + "learning_rate": 0.003, + "loss": 4.1526, + "step": 4205 + }, + { + "epoch": 0.04206, + "grad_norm": 0.5385651548116344, + "learning_rate": 0.003, + "loss": 4.1466, + "step": 4206 + }, + { + "epoch": 0.04207, + "grad_norm": 0.48635296552930585, + "learning_rate": 0.003, + "loss": 4.1349, + "step": 4207 + }, + { + "epoch": 0.04208, + "grad_norm": 0.501480728550346, + "learning_rate": 0.003, + "loss": 4.1426, + "step": 4208 + }, + { + "epoch": 0.04209, + "grad_norm": 0.5039540680425979, + "learning_rate": 0.003, + "loss": 4.1448, + "step": 4209 + }, + { + "epoch": 0.0421, + "grad_norm": 0.5190305179039434, + "learning_rate": 0.003, + "loss": 4.1031, + "step": 4210 + }, + { + "epoch": 0.04211, + "grad_norm": 0.5620311787567622, + "learning_rate": 0.003, + "loss": 4.1438, + "step": 4211 + }, + { + "epoch": 0.04212, + "grad_norm": 0.5635968224621314, + "learning_rate": 0.003, + "loss": 4.1366, + "step": 4212 + }, + { + "epoch": 0.04213, + "grad_norm": 0.5728113060887386, + "learning_rate": 0.003, + "loss": 4.1176, + "step": 4213 + }, + { + "epoch": 0.04214, + "grad_norm": 0.548258707730465, + "learning_rate": 0.003, + "loss": 4.1223, + "step": 4214 + }, + { + "epoch": 0.04215, + "grad_norm": 0.599545067228449, + "learning_rate": 0.003, + "loss": 4.1464, + "step": 4215 + }, + { + "epoch": 0.04216, + "grad_norm": 0.5748773014990753, + "learning_rate": 0.003, + "loss": 4.1096, + "step": 4216 + }, + { + "epoch": 0.04217, + "grad_norm": 0.48220599086889576, + "learning_rate": 0.003, + "loss": 4.1305, + "step": 4217 + }, + { + "epoch": 0.04218, + "grad_norm": 0.5634734523353527, + "learning_rate": 0.003, + "loss": 4.1642, + "step": 4218 + }, + { + "epoch": 0.04219, + "grad_norm": 0.6074224940320497, + "learning_rate": 0.003, + "loss": 4.0856, + "step": 4219 + }, + { + "epoch": 0.0422, + "grad_norm": 0.6971835957336069, + "learning_rate": 0.003, + "loss": 4.155, + "step": 4220 + }, + { + "epoch": 0.04221, + "grad_norm": 0.8744022212709924, + "learning_rate": 0.003, + "loss": 4.1343, + "step": 4221 + }, + { + "epoch": 0.04222, + "grad_norm": 1.091644449624445, + "learning_rate": 0.003, + "loss": 4.1115, + "step": 4222 + }, + { + "epoch": 0.04223, + "grad_norm": 0.9670759098939631, + "learning_rate": 0.003, + "loss": 4.1714, + "step": 4223 + }, + { + "epoch": 0.04224, + "grad_norm": 1.0117671228316643, + "learning_rate": 0.003, + "loss": 4.1837, + "step": 4224 + }, + { + "epoch": 0.04225, + "grad_norm": 0.9113076215312503, + "learning_rate": 0.003, + "loss": 4.1713, + "step": 4225 + }, + { + "epoch": 0.04226, + "grad_norm": 0.8960930319481113, + "learning_rate": 0.003, + "loss": 4.1523, + "step": 4226 + }, + { + "epoch": 0.04227, + "grad_norm": 0.9167424550159433, + "learning_rate": 0.003, + "loss": 4.156, + "step": 4227 + }, + { + "epoch": 0.04228, + "grad_norm": 0.967478343035116, + "learning_rate": 0.003, + "loss": 4.1417, + "step": 4228 + }, + { + "epoch": 0.04229, + "grad_norm": 0.8656039877374752, + "learning_rate": 0.003, + "loss": 4.1711, + "step": 4229 + }, + { + "epoch": 0.0423, + "grad_norm": 0.8626263746774161, + "learning_rate": 0.003, + "loss": 4.1599, + "step": 4230 + }, + { + "epoch": 0.04231, + "grad_norm": 0.7823486614059999, + "learning_rate": 0.003, + "loss": 4.1578, + "step": 4231 + }, + { + "epoch": 0.04232, + "grad_norm": 0.6738192762506559, + "learning_rate": 0.003, + "loss": 4.1671, + "step": 4232 + }, + { + "epoch": 0.04233, + "grad_norm": 0.6857804646466646, + "learning_rate": 0.003, + "loss": 4.119, + "step": 4233 + }, + { + "epoch": 0.04234, + "grad_norm": 0.7651833364067767, + "learning_rate": 0.003, + "loss": 4.1545, + "step": 4234 + }, + { + "epoch": 0.04235, + "grad_norm": 0.8557695250207406, + "learning_rate": 0.003, + "loss": 4.1383, + "step": 4235 + }, + { + "epoch": 0.04236, + "grad_norm": 0.8879388664685244, + "learning_rate": 0.003, + "loss": 4.1514, + "step": 4236 + }, + { + "epoch": 0.04237, + "grad_norm": 0.913885936433401, + "learning_rate": 0.003, + "loss": 4.1255, + "step": 4237 + }, + { + "epoch": 0.04238, + "grad_norm": 0.8386396394298068, + "learning_rate": 0.003, + "loss": 4.1852, + "step": 4238 + }, + { + "epoch": 0.04239, + "grad_norm": 0.8690235809658854, + "learning_rate": 0.003, + "loss": 4.1661, + "step": 4239 + }, + { + "epoch": 0.0424, + "grad_norm": 0.8289113096147429, + "learning_rate": 0.003, + "loss": 4.1637, + "step": 4240 + }, + { + "epoch": 0.04241, + "grad_norm": 0.7332475305569869, + "learning_rate": 0.003, + "loss": 4.1777, + "step": 4241 + }, + { + "epoch": 0.04242, + "grad_norm": 0.6138420318092145, + "learning_rate": 0.003, + "loss": 4.1602, + "step": 4242 + }, + { + "epoch": 0.04243, + "grad_norm": 0.6214280439461072, + "learning_rate": 0.003, + "loss": 4.1436, + "step": 4243 + }, + { + "epoch": 0.04244, + "grad_norm": 0.618625833924493, + "learning_rate": 0.003, + "loss": 4.1493, + "step": 4244 + }, + { + "epoch": 0.04245, + "grad_norm": 0.5938287908990628, + "learning_rate": 0.003, + "loss": 4.1435, + "step": 4245 + }, + { + "epoch": 0.04246, + "grad_norm": 0.5233831349524802, + "learning_rate": 0.003, + "loss": 4.1675, + "step": 4246 + }, + { + "epoch": 0.04247, + "grad_norm": 0.4617503229970775, + "learning_rate": 0.003, + "loss": 4.138, + "step": 4247 + }, + { + "epoch": 0.04248, + "grad_norm": 0.4580443757090383, + "learning_rate": 0.003, + "loss": 4.1323, + "step": 4248 + }, + { + "epoch": 0.04249, + "grad_norm": 0.4036456069229838, + "learning_rate": 0.003, + "loss": 4.1278, + "step": 4249 + }, + { + "epoch": 0.0425, + "grad_norm": 0.4380751494962389, + "learning_rate": 0.003, + "loss": 4.1444, + "step": 4250 + }, + { + "epoch": 0.04251, + "grad_norm": 0.5168262583557149, + "learning_rate": 0.003, + "loss": 4.1054, + "step": 4251 + }, + { + "epoch": 0.04252, + "grad_norm": 0.6808353915492127, + "learning_rate": 0.003, + "loss": 4.1541, + "step": 4252 + }, + { + "epoch": 0.04253, + "grad_norm": 0.8432737064416014, + "learning_rate": 0.003, + "loss": 4.1227, + "step": 4253 + }, + { + "epoch": 0.04254, + "grad_norm": 0.852211211018834, + "learning_rate": 0.003, + "loss": 4.1472, + "step": 4254 + }, + { + "epoch": 0.04255, + "grad_norm": 0.6240445618140698, + "learning_rate": 0.003, + "loss": 4.1184, + "step": 4255 + }, + { + "epoch": 0.04256, + "grad_norm": 0.4411537713654708, + "learning_rate": 0.003, + "loss": 4.1255, + "step": 4256 + }, + { + "epoch": 0.04257, + "grad_norm": 0.5991668514983637, + "learning_rate": 0.003, + "loss": 4.1492, + "step": 4257 + }, + { + "epoch": 0.04258, + "grad_norm": 0.7037912926105352, + "learning_rate": 0.003, + "loss": 4.1538, + "step": 4258 + }, + { + "epoch": 0.04259, + "grad_norm": 0.6762366578254241, + "learning_rate": 0.003, + "loss": 4.1198, + "step": 4259 + }, + { + "epoch": 0.0426, + "grad_norm": 0.6629398663628234, + "learning_rate": 0.003, + "loss": 4.1354, + "step": 4260 + }, + { + "epoch": 0.04261, + "grad_norm": 0.7122714093545782, + "learning_rate": 0.003, + "loss": 4.1495, + "step": 4261 + }, + { + "epoch": 0.04262, + "grad_norm": 0.6594037833976022, + "learning_rate": 0.003, + "loss": 4.1673, + "step": 4262 + }, + { + "epoch": 0.04263, + "grad_norm": 0.6354881188170681, + "learning_rate": 0.003, + "loss": 4.1474, + "step": 4263 + }, + { + "epoch": 0.04264, + "grad_norm": 0.6654557887834439, + "learning_rate": 0.003, + "loss": 4.1598, + "step": 4264 + }, + { + "epoch": 0.04265, + "grad_norm": 0.6938449312414295, + "learning_rate": 0.003, + "loss": 4.1586, + "step": 4265 + }, + { + "epoch": 0.04266, + "grad_norm": 0.7221897294325286, + "learning_rate": 0.003, + "loss": 4.1403, + "step": 4266 + }, + { + "epoch": 0.04267, + "grad_norm": 0.7436634345125996, + "learning_rate": 0.003, + "loss": 4.1158, + "step": 4267 + }, + { + "epoch": 0.04268, + "grad_norm": 0.8458877078713634, + "learning_rate": 0.003, + "loss": 4.1612, + "step": 4268 + }, + { + "epoch": 0.04269, + "grad_norm": 0.8904618229753511, + "learning_rate": 0.003, + "loss": 4.13, + "step": 4269 + }, + { + "epoch": 0.0427, + "grad_norm": 0.7112249953312418, + "learning_rate": 0.003, + "loss": 4.1401, + "step": 4270 + }, + { + "epoch": 0.04271, + "grad_norm": 0.7216953907198649, + "learning_rate": 0.003, + "loss": 4.1547, + "step": 4271 + }, + { + "epoch": 0.04272, + "grad_norm": 0.7847449594647588, + "learning_rate": 0.003, + "loss": 4.1462, + "step": 4272 + }, + { + "epoch": 0.04273, + "grad_norm": 1.004517275551493, + "learning_rate": 0.003, + "loss": 4.1318, + "step": 4273 + }, + { + "epoch": 0.04274, + "grad_norm": 1.255471199279273, + "learning_rate": 0.003, + "loss": 4.1681, + "step": 4274 + }, + { + "epoch": 0.04275, + "grad_norm": 0.6280388986370492, + "learning_rate": 0.003, + "loss": 4.1565, + "step": 4275 + }, + { + "epoch": 0.04276, + "grad_norm": 0.6763844718810441, + "learning_rate": 0.003, + "loss": 4.1743, + "step": 4276 + }, + { + "epoch": 0.04277, + "grad_norm": 0.9807377268843779, + "learning_rate": 0.003, + "loss": 4.1684, + "step": 4277 + }, + { + "epoch": 0.04278, + "grad_norm": 1.0197483171024961, + "learning_rate": 0.003, + "loss": 4.1775, + "step": 4278 + }, + { + "epoch": 0.04279, + "grad_norm": 0.9328879956310648, + "learning_rate": 0.003, + "loss": 4.1556, + "step": 4279 + }, + { + "epoch": 0.0428, + "grad_norm": 0.7779948668786022, + "learning_rate": 0.003, + "loss": 4.1299, + "step": 4280 + }, + { + "epoch": 0.04281, + "grad_norm": 0.7431305171276716, + "learning_rate": 0.003, + "loss": 4.1851, + "step": 4281 + }, + { + "epoch": 0.04282, + "grad_norm": 0.6283341114449361, + "learning_rate": 0.003, + "loss": 4.1564, + "step": 4282 + }, + { + "epoch": 0.04283, + "grad_norm": 0.6394182134064753, + "learning_rate": 0.003, + "loss": 4.1901, + "step": 4283 + }, + { + "epoch": 0.04284, + "grad_norm": 0.6758957287997666, + "learning_rate": 0.003, + "loss": 4.1437, + "step": 4284 + }, + { + "epoch": 0.04285, + "grad_norm": 0.6131823431058723, + "learning_rate": 0.003, + "loss": 4.1543, + "step": 4285 + }, + { + "epoch": 0.04286, + "grad_norm": 0.6284313302797474, + "learning_rate": 0.003, + "loss": 4.1751, + "step": 4286 + }, + { + "epoch": 0.04287, + "grad_norm": 0.6980874101445044, + "learning_rate": 0.003, + "loss": 4.1541, + "step": 4287 + }, + { + "epoch": 0.04288, + "grad_norm": 0.696667601658298, + "learning_rate": 0.003, + "loss": 4.1361, + "step": 4288 + }, + { + "epoch": 0.04289, + "grad_norm": 0.5891879790472746, + "learning_rate": 0.003, + "loss": 4.1405, + "step": 4289 + }, + { + "epoch": 0.0429, + "grad_norm": 0.6679196724071268, + "learning_rate": 0.003, + "loss": 4.127, + "step": 4290 + }, + { + "epoch": 0.04291, + "grad_norm": 0.7132152030025544, + "learning_rate": 0.003, + "loss": 4.1321, + "step": 4291 + }, + { + "epoch": 0.04292, + "grad_norm": 0.7831211413634629, + "learning_rate": 0.003, + "loss": 4.141, + "step": 4292 + }, + { + "epoch": 0.04293, + "grad_norm": 0.7420395140404202, + "learning_rate": 0.003, + "loss": 4.1725, + "step": 4293 + }, + { + "epoch": 0.04294, + "grad_norm": 0.664819817807019, + "learning_rate": 0.003, + "loss": 4.1508, + "step": 4294 + }, + { + "epoch": 0.04295, + "grad_norm": 0.6533174489648946, + "learning_rate": 0.003, + "loss": 4.1223, + "step": 4295 + }, + { + "epoch": 0.04296, + "grad_norm": 0.6539485681842259, + "learning_rate": 0.003, + "loss": 4.1205, + "step": 4296 + }, + { + "epoch": 0.04297, + "grad_norm": 0.5445244387293794, + "learning_rate": 0.003, + "loss": 4.1064, + "step": 4297 + }, + { + "epoch": 0.04298, + "grad_norm": 0.509532772235553, + "learning_rate": 0.003, + "loss": 4.1518, + "step": 4298 + }, + { + "epoch": 0.04299, + "grad_norm": 0.45809506613975526, + "learning_rate": 0.003, + "loss": 4.1361, + "step": 4299 + }, + { + "epoch": 0.043, + "grad_norm": 0.4552028711707873, + "learning_rate": 0.003, + "loss": 4.1165, + "step": 4300 + }, + { + "epoch": 0.04301, + "grad_norm": 0.5511852682855607, + "learning_rate": 0.003, + "loss": 4.1668, + "step": 4301 + }, + { + "epoch": 0.04302, + "grad_norm": 0.7137895720159015, + "learning_rate": 0.003, + "loss": 4.1605, + "step": 4302 + }, + { + "epoch": 0.04303, + "grad_norm": 0.968773809309932, + "learning_rate": 0.003, + "loss": 4.1573, + "step": 4303 + }, + { + "epoch": 0.04304, + "grad_norm": 1.0008705380109228, + "learning_rate": 0.003, + "loss": 4.1581, + "step": 4304 + }, + { + "epoch": 0.04305, + "grad_norm": 0.7364014697863502, + "learning_rate": 0.003, + "loss": 4.1497, + "step": 4305 + }, + { + "epoch": 0.04306, + "grad_norm": 0.8422192934651608, + "learning_rate": 0.003, + "loss": 4.1268, + "step": 4306 + }, + { + "epoch": 0.04307, + "grad_norm": 0.8960004698755968, + "learning_rate": 0.003, + "loss": 4.1409, + "step": 4307 + }, + { + "epoch": 0.04308, + "grad_norm": 0.8900888278943823, + "learning_rate": 0.003, + "loss": 4.1516, + "step": 4308 + }, + { + "epoch": 0.04309, + "grad_norm": 0.7851447078463367, + "learning_rate": 0.003, + "loss": 4.1418, + "step": 4309 + }, + { + "epoch": 0.0431, + "grad_norm": 0.8003446758283859, + "learning_rate": 0.003, + "loss": 4.1648, + "step": 4310 + }, + { + "epoch": 0.04311, + "grad_norm": 0.9468061968202307, + "learning_rate": 0.003, + "loss": 4.1711, + "step": 4311 + }, + { + "epoch": 0.04312, + "grad_norm": 0.8521671702866287, + "learning_rate": 0.003, + "loss": 4.149, + "step": 4312 + }, + { + "epoch": 0.04313, + "grad_norm": 0.7923961445552284, + "learning_rate": 0.003, + "loss": 4.1585, + "step": 4313 + }, + { + "epoch": 0.04314, + "grad_norm": 0.739461542428857, + "learning_rate": 0.003, + "loss": 4.1432, + "step": 4314 + }, + { + "epoch": 0.04315, + "grad_norm": 0.8509444047900186, + "learning_rate": 0.003, + "loss": 4.1727, + "step": 4315 + }, + { + "epoch": 0.04316, + "grad_norm": 0.8601049555561022, + "learning_rate": 0.003, + "loss": 4.1754, + "step": 4316 + }, + { + "epoch": 0.04317, + "grad_norm": 0.7623314110787873, + "learning_rate": 0.003, + "loss": 4.1871, + "step": 4317 + }, + { + "epoch": 0.04318, + "grad_norm": 0.6587491558124051, + "learning_rate": 0.003, + "loss": 4.1227, + "step": 4318 + }, + { + "epoch": 0.04319, + "grad_norm": 0.5849951298753531, + "learning_rate": 0.003, + "loss": 4.1174, + "step": 4319 + }, + { + "epoch": 0.0432, + "grad_norm": 0.57401029054835, + "learning_rate": 0.003, + "loss": 4.1395, + "step": 4320 + }, + { + "epoch": 0.04321, + "grad_norm": 0.6311486059735384, + "learning_rate": 0.003, + "loss": 4.1439, + "step": 4321 + }, + { + "epoch": 0.04322, + "grad_norm": 0.6281833102961318, + "learning_rate": 0.003, + "loss": 4.1751, + "step": 4322 + }, + { + "epoch": 0.04323, + "grad_norm": 0.7033374538455368, + "learning_rate": 0.003, + "loss": 4.1459, + "step": 4323 + }, + { + "epoch": 0.04324, + "grad_norm": 0.6437486136063232, + "learning_rate": 0.003, + "loss": 4.1378, + "step": 4324 + }, + { + "epoch": 0.04325, + "grad_norm": 0.6232679102675521, + "learning_rate": 0.003, + "loss": 4.1663, + "step": 4325 + }, + { + "epoch": 0.04326, + "grad_norm": 0.6515618951702632, + "learning_rate": 0.003, + "loss": 4.1454, + "step": 4326 + }, + { + "epoch": 0.04327, + "grad_norm": 0.6736681660679901, + "learning_rate": 0.003, + "loss": 4.1287, + "step": 4327 + }, + { + "epoch": 0.04328, + "grad_norm": 0.719714363226295, + "learning_rate": 0.003, + "loss": 4.1434, + "step": 4328 + }, + { + "epoch": 0.04329, + "grad_norm": 0.8360323655381567, + "learning_rate": 0.003, + "loss": 4.164, + "step": 4329 + }, + { + "epoch": 0.0433, + "grad_norm": 0.8102704190603504, + "learning_rate": 0.003, + "loss": 4.1354, + "step": 4330 + }, + { + "epoch": 0.04331, + "grad_norm": 0.7636451395661018, + "learning_rate": 0.003, + "loss": 4.1319, + "step": 4331 + }, + { + "epoch": 0.04332, + "grad_norm": 0.7789172902676315, + "learning_rate": 0.003, + "loss": 4.179, + "step": 4332 + }, + { + "epoch": 0.04333, + "grad_norm": 0.6881019378708992, + "learning_rate": 0.003, + "loss": 4.1475, + "step": 4333 + }, + { + "epoch": 0.04334, + "grad_norm": 0.5741071284174022, + "learning_rate": 0.003, + "loss": 4.1328, + "step": 4334 + }, + { + "epoch": 0.04335, + "grad_norm": 0.6016578373290047, + "learning_rate": 0.003, + "loss": 4.1203, + "step": 4335 + }, + { + "epoch": 0.04336, + "grad_norm": 0.6309338575418205, + "learning_rate": 0.003, + "loss": 4.1558, + "step": 4336 + }, + { + "epoch": 0.04337, + "grad_norm": 0.5725441440390032, + "learning_rate": 0.003, + "loss": 4.1395, + "step": 4337 + }, + { + "epoch": 0.04338, + "grad_norm": 0.5493515413414659, + "learning_rate": 0.003, + "loss": 4.1403, + "step": 4338 + }, + { + "epoch": 0.04339, + "grad_norm": 0.5145808044001364, + "learning_rate": 0.003, + "loss": 4.1442, + "step": 4339 + }, + { + "epoch": 0.0434, + "grad_norm": 0.5155260956336473, + "learning_rate": 0.003, + "loss": 4.1247, + "step": 4340 + }, + { + "epoch": 0.04341, + "grad_norm": 0.49209801146871524, + "learning_rate": 0.003, + "loss": 4.1362, + "step": 4341 + }, + { + "epoch": 0.04342, + "grad_norm": 0.5776994696106369, + "learning_rate": 0.003, + "loss": 4.1231, + "step": 4342 + }, + { + "epoch": 0.04343, + "grad_norm": 0.8551512196078731, + "learning_rate": 0.003, + "loss": 4.1689, + "step": 4343 + }, + { + "epoch": 0.04344, + "grad_norm": 1.0783725488561569, + "learning_rate": 0.003, + "loss": 4.124, + "step": 4344 + }, + { + "epoch": 0.04345, + "grad_norm": 0.9655180256016967, + "learning_rate": 0.003, + "loss": 4.1629, + "step": 4345 + }, + { + "epoch": 0.04346, + "grad_norm": 0.7582683532329971, + "learning_rate": 0.003, + "loss": 4.1515, + "step": 4346 + }, + { + "epoch": 0.04347, + "grad_norm": 0.6941505792622187, + "learning_rate": 0.003, + "loss": 4.1535, + "step": 4347 + }, + { + "epoch": 0.04348, + "grad_norm": 0.8405157089264003, + "learning_rate": 0.003, + "loss": 4.1369, + "step": 4348 + }, + { + "epoch": 0.04349, + "grad_norm": 0.8523333890856811, + "learning_rate": 0.003, + "loss": 4.1558, + "step": 4349 + }, + { + "epoch": 0.0435, + "grad_norm": 0.7396539502996479, + "learning_rate": 0.003, + "loss": 4.1379, + "step": 4350 + }, + { + "epoch": 0.04351, + "grad_norm": 0.6266986999030948, + "learning_rate": 0.003, + "loss": 4.15, + "step": 4351 + }, + { + "epoch": 0.04352, + "grad_norm": 0.5282839844073566, + "learning_rate": 0.003, + "loss": 4.1624, + "step": 4352 + }, + { + "epoch": 0.04353, + "grad_norm": 0.5005795044157166, + "learning_rate": 0.003, + "loss": 4.1304, + "step": 4353 + }, + { + "epoch": 0.04354, + "grad_norm": 0.4851237319282445, + "learning_rate": 0.003, + "loss": 4.134, + "step": 4354 + }, + { + "epoch": 0.04355, + "grad_norm": 0.5595788754843884, + "learning_rate": 0.003, + "loss": 4.1409, + "step": 4355 + }, + { + "epoch": 0.04356, + "grad_norm": 0.5667796441106236, + "learning_rate": 0.003, + "loss": 4.1467, + "step": 4356 + }, + { + "epoch": 0.04357, + "grad_norm": 0.5983610407473247, + "learning_rate": 0.003, + "loss": 4.1307, + "step": 4357 + }, + { + "epoch": 0.04358, + "grad_norm": 0.6456602317217889, + "learning_rate": 0.003, + "loss": 4.1221, + "step": 4358 + }, + { + "epoch": 0.04359, + "grad_norm": 0.6069596406446627, + "learning_rate": 0.003, + "loss": 4.1417, + "step": 4359 + }, + { + "epoch": 0.0436, + "grad_norm": 0.5908374036987593, + "learning_rate": 0.003, + "loss": 4.144, + "step": 4360 + }, + { + "epoch": 0.04361, + "grad_norm": 0.6735288968429441, + "learning_rate": 0.003, + "loss": 4.1113, + "step": 4361 + }, + { + "epoch": 0.04362, + "grad_norm": 0.7257053819990018, + "learning_rate": 0.003, + "loss": 4.1256, + "step": 4362 + }, + { + "epoch": 0.04363, + "grad_norm": 0.6914693554965775, + "learning_rate": 0.003, + "loss": 4.1294, + "step": 4363 + }, + { + "epoch": 0.04364, + "grad_norm": 0.7451580269204352, + "learning_rate": 0.003, + "loss": 4.1478, + "step": 4364 + }, + { + "epoch": 0.04365, + "grad_norm": 0.7858248696802116, + "learning_rate": 0.003, + "loss": 4.1514, + "step": 4365 + }, + { + "epoch": 0.04366, + "grad_norm": 0.663121974241136, + "learning_rate": 0.003, + "loss": 4.1013, + "step": 4366 + }, + { + "epoch": 0.04367, + "grad_norm": 0.7175426458581545, + "learning_rate": 0.003, + "loss": 4.1542, + "step": 4367 + }, + { + "epoch": 0.04368, + "grad_norm": 0.7351488980711208, + "learning_rate": 0.003, + "loss": 4.1422, + "step": 4368 + }, + { + "epoch": 0.04369, + "grad_norm": 0.8536889689371039, + "learning_rate": 0.003, + "loss": 4.1453, + "step": 4369 + }, + { + "epoch": 0.0437, + "grad_norm": 0.854111892194005, + "learning_rate": 0.003, + "loss": 4.1678, + "step": 4370 + }, + { + "epoch": 0.04371, + "grad_norm": 1.046284154850743, + "learning_rate": 0.003, + "loss": 4.1575, + "step": 4371 + }, + { + "epoch": 0.04372, + "grad_norm": 1.2262372791586333, + "learning_rate": 0.003, + "loss": 4.1517, + "step": 4372 + }, + { + "epoch": 0.04373, + "grad_norm": 0.8968734983342949, + "learning_rate": 0.003, + "loss": 4.1673, + "step": 4373 + }, + { + "epoch": 0.04374, + "grad_norm": 0.9753259228506409, + "learning_rate": 0.003, + "loss": 4.2155, + "step": 4374 + }, + { + "epoch": 0.04375, + "grad_norm": 0.9920981237100678, + "learning_rate": 0.003, + "loss": 4.1324, + "step": 4375 + }, + { + "epoch": 0.04376, + "grad_norm": 1.0294125384070931, + "learning_rate": 0.003, + "loss": 4.1801, + "step": 4376 + }, + { + "epoch": 0.04377, + "grad_norm": 1.0068899366919466, + "learning_rate": 0.003, + "loss": 4.2068, + "step": 4377 + }, + { + "epoch": 0.04378, + "grad_norm": 1.0341191509965633, + "learning_rate": 0.003, + "loss": 4.191, + "step": 4378 + }, + { + "epoch": 0.04379, + "grad_norm": 0.9717889790676081, + "learning_rate": 0.003, + "loss": 4.2061, + "step": 4379 + }, + { + "epoch": 0.0438, + "grad_norm": 1.0471882312294167, + "learning_rate": 0.003, + "loss": 4.1807, + "step": 4380 + }, + { + "epoch": 0.04381, + "grad_norm": 0.8790592687859848, + "learning_rate": 0.003, + "loss": 4.1637, + "step": 4381 + }, + { + "epoch": 0.04382, + "grad_norm": 0.8968294394751086, + "learning_rate": 0.003, + "loss": 4.1774, + "step": 4382 + }, + { + "epoch": 0.04383, + "grad_norm": 0.8820678241101328, + "learning_rate": 0.003, + "loss": 4.157, + "step": 4383 + }, + { + "epoch": 0.04384, + "grad_norm": 1.0214298134194522, + "learning_rate": 0.003, + "loss": 4.1664, + "step": 4384 + }, + { + "epoch": 0.04385, + "grad_norm": 0.900876137883741, + "learning_rate": 0.003, + "loss": 4.2023, + "step": 4385 + }, + { + "epoch": 0.04386, + "grad_norm": 0.7692038806596491, + "learning_rate": 0.003, + "loss": 4.1737, + "step": 4386 + }, + { + "epoch": 0.04387, + "grad_norm": 0.653702371460772, + "learning_rate": 0.003, + "loss": 4.1914, + "step": 4387 + }, + { + "epoch": 0.04388, + "grad_norm": 0.6492229267746281, + "learning_rate": 0.003, + "loss": 4.1506, + "step": 4388 + }, + { + "epoch": 0.04389, + "grad_norm": 0.5786656683989158, + "learning_rate": 0.003, + "loss": 4.179, + "step": 4389 + }, + { + "epoch": 0.0439, + "grad_norm": 0.5012080835531559, + "learning_rate": 0.003, + "loss": 4.1702, + "step": 4390 + }, + { + "epoch": 0.04391, + "grad_norm": 0.48775107179113014, + "learning_rate": 0.003, + "loss": 4.1402, + "step": 4391 + }, + { + "epoch": 0.04392, + "grad_norm": 0.46374848591785856, + "learning_rate": 0.003, + "loss": 4.1462, + "step": 4392 + }, + { + "epoch": 0.04393, + "grad_norm": 0.450356925230617, + "learning_rate": 0.003, + "loss": 4.1327, + "step": 4393 + }, + { + "epoch": 0.04394, + "grad_norm": 0.4674972403264732, + "learning_rate": 0.003, + "loss": 4.1498, + "step": 4394 + }, + { + "epoch": 0.04395, + "grad_norm": 0.5893115115334077, + "learning_rate": 0.003, + "loss": 4.1342, + "step": 4395 + }, + { + "epoch": 0.04396, + "grad_norm": 0.7658769835507094, + "learning_rate": 0.003, + "loss": 4.1118, + "step": 4396 + }, + { + "epoch": 0.04397, + "grad_norm": 0.9383339109696239, + "learning_rate": 0.003, + "loss": 4.1344, + "step": 4397 + }, + { + "epoch": 0.04398, + "grad_norm": 0.8777679801246571, + "learning_rate": 0.003, + "loss": 4.1501, + "step": 4398 + }, + { + "epoch": 0.04399, + "grad_norm": 0.7720983136183994, + "learning_rate": 0.003, + "loss": 4.1556, + "step": 4399 + }, + { + "epoch": 0.044, + "grad_norm": 0.7229759596282145, + "learning_rate": 0.003, + "loss": 4.1478, + "step": 4400 + }, + { + "epoch": 0.04401, + "grad_norm": 0.8501411177864325, + "learning_rate": 0.003, + "loss": 4.1629, + "step": 4401 + }, + { + "epoch": 0.04402, + "grad_norm": 0.8319303801252986, + "learning_rate": 0.003, + "loss": 4.172, + "step": 4402 + }, + { + "epoch": 0.04403, + "grad_norm": 0.6805684530525856, + "learning_rate": 0.003, + "loss": 4.1317, + "step": 4403 + }, + { + "epoch": 0.04404, + "grad_norm": 0.5579714343751891, + "learning_rate": 0.003, + "loss": 4.1573, + "step": 4404 + }, + { + "epoch": 0.04405, + "grad_norm": 0.6272316247141385, + "learning_rate": 0.003, + "loss": 4.151, + "step": 4405 + }, + { + "epoch": 0.04406, + "grad_norm": 0.7189270603588521, + "learning_rate": 0.003, + "loss": 4.1491, + "step": 4406 + }, + { + "epoch": 0.04407, + "grad_norm": 0.762459715751194, + "learning_rate": 0.003, + "loss": 4.1227, + "step": 4407 + }, + { + "epoch": 0.04408, + "grad_norm": 0.7250097639288197, + "learning_rate": 0.003, + "loss": 4.1272, + "step": 4408 + }, + { + "epoch": 0.04409, + "grad_norm": 0.7106703945924645, + "learning_rate": 0.003, + "loss": 4.1606, + "step": 4409 + }, + { + "epoch": 0.0441, + "grad_norm": 0.5937514209420578, + "learning_rate": 0.003, + "loss": 4.1417, + "step": 4410 + }, + { + "epoch": 0.04411, + "grad_norm": 0.6265606257146901, + "learning_rate": 0.003, + "loss": 4.124, + "step": 4411 + }, + { + "epoch": 0.04412, + "grad_norm": 0.684509287657226, + "learning_rate": 0.003, + "loss": 4.114, + "step": 4412 + }, + { + "epoch": 0.04413, + "grad_norm": 0.7827609729511189, + "learning_rate": 0.003, + "loss": 4.1251, + "step": 4413 + }, + { + "epoch": 0.04414, + "grad_norm": 0.8264979154885277, + "learning_rate": 0.003, + "loss": 4.1806, + "step": 4414 + }, + { + "epoch": 0.04415, + "grad_norm": 0.7365639173006763, + "learning_rate": 0.003, + "loss": 4.1413, + "step": 4415 + }, + { + "epoch": 0.04416, + "grad_norm": 0.5988319699730835, + "learning_rate": 0.003, + "loss": 4.1256, + "step": 4416 + }, + { + "epoch": 0.04417, + "grad_norm": 0.6185388120705024, + "learning_rate": 0.003, + "loss": 4.1033, + "step": 4417 + }, + { + "epoch": 0.04418, + "grad_norm": 0.7024277083040185, + "learning_rate": 0.003, + "loss": 4.145, + "step": 4418 + }, + { + "epoch": 0.04419, + "grad_norm": 0.6522881055559766, + "learning_rate": 0.003, + "loss": 4.1651, + "step": 4419 + }, + { + "epoch": 0.0442, + "grad_norm": 0.6345542061505771, + "learning_rate": 0.003, + "loss": 4.1337, + "step": 4420 + }, + { + "epoch": 0.04421, + "grad_norm": 0.6652846957230868, + "learning_rate": 0.003, + "loss": 4.1668, + "step": 4421 + }, + { + "epoch": 0.04422, + "grad_norm": 0.7005501321538639, + "learning_rate": 0.003, + "loss": 4.1322, + "step": 4422 + }, + { + "epoch": 0.04423, + "grad_norm": 0.7044884290461375, + "learning_rate": 0.003, + "loss": 4.1513, + "step": 4423 + }, + { + "epoch": 0.04424, + "grad_norm": 0.6741401349701589, + "learning_rate": 0.003, + "loss": 4.1007, + "step": 4424 + }, + { + "epoch": 0.04425, + "grad_norm": 0.6934002596368363, + "learning_rate": 0.003, + "loss": 4.1257, + "step": 4425 + }, + { + "epoch": 0.04426, + "grad_norm": 0.6003050918470045, + "learning_rate": 0.003, + "loss": 4.1458, + "step": 4426 + }, + { + "epoch": 0.04427, + "grad_norm": 0.6672810210726654, + "learning_rate": 0.003, + "loss": 4.1355, + "step": 4427 + }, + { + "epoch": 0.04428, + "grad_norm": 0.6318153966455053, + "learning_rate": 0.003, + "loss": 4.1305, + "step": 4428 + }, + { + "epoch": 0.04429, + "grad_norm": 0.623912292564508, + "learning_rate": 0.003, + "loss": 4.1717, + "step": 4429 + }, + { + "epoch": 0.0443, + "grad_norm": 0.7110054552912546, + "learning_rate": 0.003, + "loss": 4.132, + "step": 4430 + }, + { + "epoch": 0.04431, + "grad_norm": 0.7246952184785321, + "learning_rate": 0.003, + "loss": 4.157, + "step": 4431 + }, + { + "epoch": 0.04432, + "grad_norm": 0.7298502791104271, + "learning_rate": 0.003, + "loss": 4.1607, + "step": 4432 + }, + { + "epoch": 0.04433, + "grad_norm": 0.8250488623045142, + "learning_rate": 0.003, + "loss": 4.1389, + "step": 4433 + }, + { + "epoch": 0.04434, + "grad_norm": 0.9624884861403035, + "learning_rate": 0.003, + "loss": 4.1442, + "step": 4434 + }, + { + "epoch": 0.04435, + "grad_norm": 0.9889176292713632, + "learning_rate": 0.003, + "loss": 4.1624, + "step": 4435 + }, + { + "epoch": 0.04436, + "grad_norm": 0.9593070985375517, + "learning_rate": 0.003, + "loss": 4.1373, + "step": 4436 + }, + { + "epoch": 0.04437, + "grad_norm": 0.754019474989507, + "learning_rate": 0.003, + "loss": 4.1545, + "step": 4437 + }, + { + "epoch": 0.04438, + "grad_norm": 0.6112092782747661, + "learning_rate": 0.003, + "loss": 4.1893, + "step": 4438 + }, + { + "epoch": 0.04439, + "grad_norm": 0.8008555881950424, + "learning_rate": 0.003, + "loss": 4.1584, + "step": 4439 + }, + { + "epoch": 0.0444, + "grad_norm": 0.8796518350119269, + "learning_rate": 0.003, + "loss": 4.1203, + "step": 4440 + }, + { + "epoch": 0.04441, + "grad_norm": 0.881979649964416, + "learning_rate": 0.003, + "loss": 4.155, + "step": 4441 + }, + { + "epoch": 0.04442, + "grad_norm": 0.8372483887218065, + "learning_rate": 0.003, + "loss": 4.1465, + "step": 4442 + }, + { + "epoch": 0.04443, + "grad_norm": 0.6759359878775629, + "learning_rate": 0.003, + "loss": 4.1188, + "step": 4443 + }, + { + "epoch": 0.04444, + "grad_norm": 0.6554603066254844, + "learning_rate": 0.003, + "loss": 4.12, + "step": 4444 + }, + { + "epoch": 0.04445, + "grad_norm": 0.5717002727777339, + "learning_rate": 0.003, + "loss": 4.1477, + "step": 4445 + }, + { + "epoch": 0.04446, + "grad_norm": 0.49898785075563756, + "learning_rate": 0.003, + "loss": 4.1144, + "step": 4446 + }, + { + "epoch": 0.04447, + "grad_norm": 0.5124508174653755, + "learning_rate": 0.003, + "loss": 4.1181, + "step": 4447 + }, + { + "epoch": 0.04448, + "grad_norm": 0.46026046835883233, + "learning_rate": 0.003, + "loss": 4.1155, + "step": 4448 + }, + { + "epoch": 0.04449, + "grad_norm": 0.40191804438918904, + "learning_rate": 0.003, + "loss": 4.1268, + "step": 4449 + }, + { + "epoch": 0.0445, + "grad_norm": 0.4056694001793259, + "learning_rate": 0.003, + "loss": 4.1311, + "step": 4450 + }, + { + "epoch": 0.04451, + "grad_norm": 0.4140029705225631, + "learning_rate": 0.003, + "loss": 4.1277, + "step": 4451 + }, + { + "epoch": 0.04452, + "grad_norm": 0.4961606990959985, + "learning_rate": 0.003, + "loss": 4.1432, + "step": 4452 + }, + { + "epoch": 0.04453, + "grad_norm": 0.7225243712391776, + "learning_rate": 0.003, + "loss": 4.1556, + "step": 4453 + }, + { + "epoch": 0.04454, + "grad_norm": 0.9541612332110946, + "learning_rate": 0.003, + "loss": 4.1428, + "step": 4454 + }, + { + "epoch": 0.04455, + "grad_norm": 1.004523212032084, + "learning_rate": 0.003, + "loss": 4.1579, + "step": 4455 + }, + { + "epoch": 0.04456, + "grad_norm": 0.7366078757936858, + "learning_rate": 0.003, + "loss": 4.1199, + "step": 4456 + }, + { + "epoch": 0.04457, + "grad_norm": 0.7291092880614075, + "learning_rate": 0.003, + "loss": 4.1359, + "step": 4457 + }, + { + "epoch": 0.04458, + "grad_norm": 0.8886507308195684, + "learning_rate": 0.003, + "loss": 4.1658, + "step": 4458 + }, + { + "epoch": 0.04459, + "grad_norm": 1.023956211330971, + "learning_rate": 0.003, + "loss": 4.1783, + "step": 4459 + }, + { + "epoch": 0.0446, + "grad_norm": 0.9783050212355968, + "learning_rate": 0.003, + "loss": 4.1765, + "step": 4460 + }, + { + "epoch": 0.04461, + "grad_norm": 0.8910569073793239, + "learning_rate": 0.003, + "loss": 4.1774, + "step": 4461 + }, + { + "epoch": 0.04462, + "grad_norm": 1.0114250150322548, + "learning_rate": 0.003, + "loss": 4.1382, + "step": 4462 + }, + { + "epoch": 0.04463, + "grad_norm": 1.109635261633442, + "learning_rate": 0.003, + "loss": 4.159, + "step": 4463 + }, + { + "epoch": 0.04464, + "grad_norm": 0.9892179336121286, + "learning_rate": 0.003, + "loss": 4.1722, + "step": 4464 + }, + { + "epoch": 0.04465, + "grad_norm": 0.9496628250922594, + "learning_rate": 0.003, + "loss": 4.1619, + "step": 4465 + }, + { + "epoch": 0.04466, + "grad_norm": 0.9698557150392482, + "learning_rate": 0.003, + "loss": 4.1735, + "step": 4466 + }, + { + "epoch": 0.04467, + "grad_norm": 0.9285468685707471, + "learning_rate": 0.003, + "loss": 4.1725, + "step": 4467 + }, + { + "epoch": 0.04468, + "grad_norm": 0.9397451172133555, + "learning_rate": 0.003, + "loss": 4.1519, + "step": 4468 + }, + { + "epoch": 0.04469, + "grad_norm": 0.9565209269681922, + "learning_rate": 0.003, + "loss": 4.1882, + "step": 4469 + }, + { + "epoch": 0.0447, + "grad_norm": 0.8769720683928874, + "learning_rate": 0.003, + "loss": 4.1844, + "step": 4470 + }, + { + "epoch": 0.04471, + "grad_norm": 0.795338760730945, + "learning_rate": 0.003, + "loss": 4.2008, + "step": 4471 + }, + { + "epoch": 0.04472, + "grad_norm": 0.7843283232673075, + "learning_rate": 0.003, + "loss": 4.1549, + "step": 4472 + }, + { + "epoch": 0.04473, + "grad_norm": 0.8909484017306056, + "learning_rate": 0.003, + "loss": 4.1593, + "step": 4473 + }, + { + "epoch": 0.04474, + "grad_norm": 0.7784392006819544, + "learning_rate": 0.003, + "loss": 4.152, + "step": 4474 + }, + { + "epoch": 0.04475, + "grad_norm": 0.7042700663520809, + "learning_rate": 0.003, + "loss": 4.1398, + "step": 4475 + }, + { + "epoch": 0.04476, + "grad_norm": 0.6389974316195782, + "learning_rate": 0.003, + "loss": 4.1499, + "step": 4476 + }, + { + "epoch": 0.04477, + "grad_norm": 0.6071760811715112, + "learning_rate": 0.003, + "loss": 4.1616, + "step": 4477 + }, + { + "epoch": 0.04478, + "grad_norm": 0.5219307858429795, + "learning_rate": 0.003, + "loss": 4.1544, + "step": 4478 + }, + { + "epoch": 0.04479, + "grad_norm": 0.52342819477722, + "learning_rate": 0.003, + "loss": 4.1329, + "step": 4479 + }, + { + "epoch": 0.0448, + "grad_norm": 0.5268154790194748, + "learning_rate": 0.003, + "loss": 4.1425, + "step": 4480 + }, + { + "epoch": 0.04481, + "grad_norm": 0.5399282796391196, + "learning_rate": 0.003, + "loss": 4.1341, + "step": 4481 + }, + { + "epoch": 0.04482, + "grad_norm": 0.6507902172065512, + "learning_rate": 0.003, + "loss": 4.1572, + "step": 4482 + }, + { + "epoch": 0.04483, + "grad_norm": 0.867713110982925, + "learning_rate": 0.003, + "loss": 4.1524, + "step": 4483 + }, + { + "epoch": 0.04484, + "grad_norm": 0.9801856447775387, + "learning_rate": 0.003, + "loss": 4.1697, + "step": 4484 + }, + { + "epoch": 0.04485, + "grad_norm": 0.8358981195735171, + "learning_rate": 0.003, + "loss": 4.176, + "step": 4485 + }, + { + "epoch": 0.04486, + "grad_norm": 0.726373516850574, + "learning_rate": 0.003, + "loss": 4.1324, + "step": 4486 + }, + { + "epoch": 0.04487, + "grad_norm": 0.7123593368535847, + "learning_rate": 0.003, + "loss": 4.1419, + "step": 4487 + }, + { + "epoch": 0.04488, + "grad_norm": 0.726522221164894, + "learning_rate": 0.003, + "loss": 4.1587, + "step": 4488 + }, + { + "epoch": 0.04489, + "grad_norm": 0.7095321623068442, + "learning_rate": 0.003, + "loss": 4.163, + "step": 4489 + }, + { + "epoch": 0.0449, + "grad_norm": 0.6252082433565278, + "learning_rate": 0.003, + "loss": 4.1567, + "step": 4490 + }, + { + "epoch": 0.04491, + "grad_norm": 0.6640250192773183, + "learning_rate": 0.003, + "loss": 4.1429, + "step": 4491 + }, + { + "epoch": 0.04492, + "grad_norm": 0.6961055886888501, + "learning_rate": 0.003, + "loss": 4.1552, + "step": 4492 + }, + { + "epoch": 0.04493, + "grad_norm": 0.8458020064467313, + "learning_rate": 0.003, + "loss": 4.1623, + "step": 4493 + }, + { + "epoch": 0.04494, + "grad_norm": 0.7474034453269607, + "learning_rate": 0.003, + "loss": 4.1407, + "step": 4494 + }, + { + "epoch": 0.04495, + "grad_norm": 0.6451994546687867, + "learning_rate": 0.003, + "loss": 4.1365, + "step": 4495 + }, + { + "epoch": 0.04496, + "grad_norm": 0.540011005371107, + "learning_rate": 0.003, + "loss": 4.1741, + "step": 4496 + }, + { + "epoch": 0.04497, + "grad_norm": 0.4950562236330396, + "learning_rate": 0.003, + "loss": 4.1537, + "step": 4497 + }, + { + "epoch": 0.04498, + "grad_norm": 0.4943734925550196, + "learning_rate": 0.003, + "loss": 4.1237, + "step": 4498 + }, + { + "epoch": 0.04499, + "grad_norm": 0.47393717260525386, + "learning_rate": 0.003, + "loss": 4.0968, + "step": 4499 + }, + { + "epoch": 0.045, + "grad_norm": 0.41885472582226696, + "learning_rate": 0.003, + "loss": 4.1243, + "step": 4500 + }, + { + "epoch": 0.04501, + "grad_norm": 0.4834752022098693, + "learning_rate": 0.003, + "loss": 4.1175, + "step": 4501 + }, + { + "epoch": 0.04502, + "grad_norm": 0.45694364065485266, + "learning_rate": 0.003, + "loss": 4.1588, + "step": 4502 + }, + { + "epoch": 0.04503, + "grad_norm": 0.48865003766237863, + "learning_rate": 0.003, + "loss": 4.1582, + "step": 4503 + }, + { + "epoch": 0.04504, + "grad_norm": 0.5438297922402094, + "learning_rate": 0.003, + "loss": 4.1576, + "step": 4504 + }, + { + "epoch": 0.04505, + "grad_norm": 0.5871203728299376, + "learning_rate": 0.003, + "loss": 4.154, + "step": 4505 + }, + { + "epoch": 0.04506, + "grad_norm": 0.5917294880351803, + "learning_rate": 0.003, + "loss": 4.1318, + "step": 4506 + }, + { + "epoch": 0.04507, + "grad_norm": 0.7183532404676503, + "learning_rate": 0.003, + "loss": 4.1091, + "step": 4507 + }, + { + "epoch": 0.04508, + "grad_norm": 0.8049591048914756, + "learning_rate": 0.003, + "loss": 4.1329, + "step": 4508 + }, + { + "epoch": 0.04509, + "grad_norm": 0.9156955919586133, + "learning_rate": 0.003, + "loss": 4.1602, + "step": 4509 + }, + { + "epoch": 0.0451, + "grad_norm": 0.8889056537954558, + "learning_rate": 0.003, + "loss": 4.1618, + "step": 4510 + }, + { + "epoch": 0.04511, + "grad_norm": 0.693800418390815, + "learning_rate": 0.003, + "loss": 4.1419, + "step": 4511 + }, + { + "epoch": 0.04512, + "grad_norm": 0.7135831824345485, + "learning_rate": 0.003, + "loss": 4.0971, + "step": 4512 + }, + { + "epoch": 0.04513, + "grad_norm": 0.6688562196054199, + "learning_rate": 0.003, + "loss": 4.1268, + "step": 4513 + }, + { + "epoch": 0.04514, + "grad_norm": 0.6423991582569938, + "learning_rate": 0.003, + "loss": 4.1345, + "step": 4514 + }, + { + "epoch": 0.04515, + "grad_norm": 0.7711378950211699, + "learning_rate": 0.003, + "loss": 4.1345, + "step": 4515 + }, + { + "epoch": 0.04516, + "grad_norm": 0.8939759348625197, + "learning_rate": 0.003, + "loss": 4.1535, + "step": 4516 + }, + { + "epoch": 0.04517, + "grad_norm": 0.7606381004553531, + "learning_rate": 0.003, + "loss": 4.1665, + "step": 4517 + }, + { + "epoch": 0.04518, + "grad_norm": 0.6218485906194273, + "learning_rate": 0.003, + "loss": 4.1359, + "step": 4518 + }, + { + "epoch": 0.04519, + "grad_norm": 0.7964516800865484, + "learning_rate": 0.003, + "loss": 4.1353, + "step": 4519 + }, + { + "epoch": 0.0452, + "grad_norm": 0.8973981028823347, + "learning_rate": 0.003, + "loss": 4.1488, + "step": 4520 + }, + { + "epoch": 0.04521, + "grad_norm": 0.807736716714071, + "learning_rate": 0.003, + "loss": 4.1345, + "step": 4521 + }, + { + "epoch": 0.04522, + "grad_norm": 0.8117844738071615, + "learning_rate": 0.003, + "loss": 4.1378, + "step": 4522 + }, + { + "epoch": 0.04523, + "grad_norm": 0.7829719745475495, + "learning_rate": 0.003, + "loss": 4.1416, + "step": 4523 + }, + { + "epoch": 0.04524, + "grad_norm": 0.7479885268590871, + "learning_rate": 0.003, + "loss": 4.1194, + "step": 4524 + }, + { + "epoch": 0.04525, + "grad_norm": 0.7243373752912751, + "learning_rate": 0.003, + "loss": 4.127, + "step": 4525 + }, + { + "epoch": 0.04526, + "grad_norm": 0.7384870219800298, + "learning_rate": 0.003, + "loss": 4.157, + "step": 4526 + }, + { + "epoch": 0.04527, + "grad_norm": 0.8786326776352553, + "learning_rate": 0.003, + "loss": 4.1433, + "step": 4527 + }, + { + "epoch": 0.04528, + "grad_norm": 0.9239987067801139, + "learning_rate": 0.003, + "loss": 4.1245, + "step": 4528 + }, + { + "epoch": 0.04529, + "grad_norm": 0.8389910760544824, + "learning_rate": 0.003, + "loss": 4.1271, + "step": 4529 + }, + { + "epoch": 0.0453, + "grad_norm": 0.748168700266448, + "learning_rate": 0.003, + "loss": 4.1231, + "step": 4530 + }, + { + "epoch": 0.04531, + "grad_norm": 0.7202788367913475, + "learning_rate": 0.003, + "loss": 4.1225, + "step": 4531 + }, + { + "epoch": 0.04532, + "grad_norm": 0.6571475637254836, + "learning_rate": 0.003, + "loss": 4.1461, + "step": 4532 + }, + { + "epoch": 0.04533, + "grad_norm": 0.7185379271005126, + "learning_rate": 0.003, + "loss": 4.1112, + "step": 4533 + }, + { + "epoch": 0.04534, + "grad_norm": 0.6309360237693905, + "learning_rate": 0.003, + "loss": 4.1522, + "step": 4534 + }, + { + "epoch": 0.04535, + "grad_norm": 0.6779837587183748, + "learning_rate": 0.003, + "loss": 4.1737, + "step": 4535 + }, + { + "epoch": 0.04536, + "grad_norm": 0.6922814218735922, + "learning_rate": 0.003, + "loss": 4.1687, + "step": 4536 + }, + { + "epoch": 0.04537, + "grad_norm": 0.6707461416618695, + "learning_rate": 0.003, + "loss": 4.1466, + "step": 4537 + }, + { + "epoch": 0.04538, + "grad_norm": 0.6533648131571776, + "learning_rate": 0.003, + "loss": 4.1468, + "step": 4538 + }, + { + "epoch": 0.04539, + "grad_norm": 0.7226394008940196, + "learning_rate": 0.003, + "loss": 4.1106, + "step": 4539 + }, + { + "epoch": 0.0454, + "grad_norm": 0.7809591559130085, + "learning_rate": 0.003, + "loss": 4.1225, + "step": 4540 + }, + { + "epoch": 0.04541, + "grad_norm": 0.8598716052287259, + "learning_rate": 0.003, + "loss": 4.1744, + "step": 4541 + }, + { + "epoch": 0.04542, + "grad_norm": 0.7887863556115788, + "learning_rate": 0.003, + "loss": 4.1194, + "step": 4542 + }, + { + "epoch": 0.04543, + "grad_norm": 0.5682149072319508, + "learning_rate": 0.003, + "loss": 4.146, + "step": 4543 + }, + { + "epoch": 0.04544, + "grad_norm": 0.5505936637479015, + "learning_rate": 0.003, + "loss": 4.1237, + "step": 4544 + }, + { + "epoch": 0.04545, + "grad_norm": 0.6999578034297054, + "learning_rate": 0.003, + "loss": 4.1378, + "step": 4545 + }, + { + "epoch": 0.04546, + "grad_norm": 0.6684727088847572, + "learning_rate": 0.003, + "loss": 4.1187, + "step": 4546 + }, + { + "epoch": 0.04547, + "grad_norm": 0.6131696984235054, + "learning_rate": 0.003, + "loss": 4.1242, + "step": 4547 + }, + { + "epoch": 0.04548, + "grad_norm": 0.6233401797280433, + "learning_rate": 0.003, + "loss": 4.1371, + "step": 4548 + }, + { + "epoch": 0.04549, + "grad_norm": 0.7411565175033444, + "learning_rate": 0.003, + "loss": 4.1303, + "step": 4549 + }, + { + "epoch": 0.0455, + "grad_norm": 0.7508874744904848, + "learning_rate": 0.003, + "loss": 4.1646, + "step": 4550 + }, + { + "epoch": 0.04551, + "grad_norm": 0.7534956856948848, + "learning_rate": 0.003, + "loss": 4.1738, + "step": 4551 + }, + { + "epoch": 0.04552, + "grad_norm": 0.689574501986947, + "learning_rate": 0.003, + "loss": 4.1544, + "step": 4552 + }, + { + "epoch": 0.04553, + "grad_norm": 0.6536678327481642, + "learning_rate": 0.003, + "loss": 4.1419, + "step": 4553 + }, + { + "epoch": 0.04554, + "grad_norm": 0.6521633276844322, + "learning_rate": 0.003, + "loss": 4.1285, + "step": 4554 + }, + { + "epoch": 0.04555, + "grad_norm": 0.7438153849355749, + "learning_rate": 0.003, + "loss": 4.1448, + "step": 4555 + }, + { + "epoch": 0.04556, + "grad_norm": 0.813199349395972, + "learning_rate": 0.003, + "loss": 4.1481, + "step": 4556 + }, + { + "epoch": 0.04557, + "grad_norm": 0.8670448776825898, + "learning_rate": 0.003, + "loss": 4.1339, + "step": 4557 + }, + { + "epoch": 0.04558, + "grad_norm": 0.9173442370434104, + "learning_rate": 0.003, + "loss": 4.165, + "step": 4558 + }, + { + "epoch": 0.04559, + "grad_norm": 0.9293838780402448, + "learning_rate": 0.003, + "loss": 4.1785, + "step": 4559 + }, + { + "epoch": 0.0456, + "grad_norm": 0.9721860045634062, + "learning_rate": 0.003, + "loss": 4.142, + "step": 4560 + }, + { + "epoch": 0.04561, + "grad_norm": 0.9130374577828355, + "learning_rate": 0.003, + "loss": 4.1572, + "step": 4561 + }, + { + "epoch": 0.04562, + "grad_norm": 0.8204531300346364, + "learning_rate": 0.003, + "loss": 4.1735, + "step": 4562 + }, + { + "epoch": 0.04563, + "grad_norm": 0.941182990714371, + "learning_rate": 0.003, + "loss": 4.145, + "step": 4563 + }, + { + "epoch": 0.04564, + "grad_norm": 0.9950132484736411, + "learning_rate": 0.003, + "loss": 4.1607, + "step": 4564 + }, + { + "epoch": 0.04565, + "grad_norm": 0.8713983224408589, + "learning_rate": 0.003, + "loss": 4.1308, + "step": 4565 + }, + { + "epoch": 0.04566, + "grad_norm": 0.7742242108390238, + "learning_rate": 0.003, + "loss": 4.158, + "step": 4566 + }, + { + "epoch": 0.04567, + "grad_norm": 0.7732353202308767, + "learning_rate": 0.003, + "loss": 4.1557, + "step": 4567 + }, + { + "epoch": 0.04568, + "grad_norm": 0.7949276600918046, + "learning_rate": 0.003, + "loss": 4.1324, + "step": 4568 + }, + { + "epoch": 0.04569, + "grad_norm": 0.76216115917809, + "learning_rate": 0.003, + "loss": 4.1586, + "step": 4569 + }, + { + "epoch": 0.0457, + "grad_norm": 0.665803751777818, + "learning_rate": 0.003, + "loss": 4.1443, + "step": 4570 + }, + { + "epoch": 0.04571, + "grad_norm": 0.61898217206735, + "learning_rate": 0.003, + "loss": 4.1709, + "step": 4571 + }, + { + "epoch": 0.04572, + "grad_norm": 0.6630102965712841, + "learning_rate": 0.003, + "loss": 4.1438, + "step": 4572 + }, + { + "epoch": 0.04573, + "grad_norm": 0.6545388007004492, + "learning_rate": 0.003, + "loss": 4.1147, + "step": 4573 + }, + { + "epoch": 0.04574, + "grad_norm": 0.6434199338973368, + "learning_rate": 0.003, + "loss": 4.1269, + "step": 4574 + }, + { + "epoch": 0.04575, + "grad_norm": 0.6084613496329172, + "learning_rate": 0.003, + "loss": 4.1507, + "step": 4575 + }, + { + "epoch": 0.04576, + "grad_norm": 0.5926985992517696, + "learning_rate": 0.003, + "loss": 4.1399, + "step": 4576 + }, + { + "epoch": 0.04577, + "grad_norm": 0.7068872849239232, + "learning_rate": 0.003, + "loss": 4.161, + "step": 4577 + }, + { + "epoch": 0.04578, + "grad_norm": 0.7815668152529673, + "learning_rate": 0.003, + "loss": 4.1309, + "step": 4578 + }, + { + "epoch": 0.04579, + "grad_norm": 0.705130762977042, + "learning_rate": 0.003, + "loss": 4.1148, + "step": 4579 + }, + { + "epoch": 0.0458, + "grad_norm": 0.706838854182526, + "learning_rate": 0.003, + "loss": 4.1481, + "step": 4580 + }, + { + "epoch": 0.04581, + "grad_norm": 0.7719932048413563, + "learning_rate": 0.003, + "loss": 4.1624, + "step": 4581 + }, + { + "epoch": 0.04582, + "grad_norm": 0.8196871294195923, + "learning_rate": 0.003, + "loss": 4.1644, + "step": 4582 + }, + { + "epoch": 0.04583, + "grad_norm": 0.9747741235878243, + "learning_rate": 0.003, + "loss": 4.1405, + "step": 4583 + }, + { + "epoch": 0.04584, + "grad_norm": 1.0614405255220143, + "learning_rate": 0.003, + "loss": 4.148, + "step": 4584 + }, + { + "epoch": 0.04585, + "grad_norm": 0.8692609981568404, + "learning_rate": 0.003, + "loss": 4.1458, + "step": 4585 + }, + { + "epoch": 0.04586, + "grad_norm": 0.8017360688038584, + "learning_rate": 0.003, + "loss": 4.1494, + "step": 4586 + }, + { + "epoch": 0.04587, + "grad_norm": 0.737110938579538, + "learning_rate": 0.003, + "loss": 4.1647, + "step": 4587 + }, + { + "epoch": 0.04588, + "grad_norm": 0.6638294236901521, + "learning_rate": 0.003, + "loss": 4.1602, + "step": 4588 + }, + { + "epoch": 0.04589, + "grad_norm": 0.6411690157984203, + "learning_rate": 0.003, + "loss": 4.1251, + "step": 4589 + }, + { + "epoch": 0.0459, + "grad_norm": 0.6088730568376585, + "learning_rate": 0.003, + "loss": 4.1499, + "step": 4590 + }, + { + "epoch": 0.04591, + "grad_norm": 0.6362826036737852, + "learning_rate": 0.003, + "loss": 4.1154, + "step": 4591 + }, + { + "epoch": 0.04592, + "grad_norm": 0.6389570140066912, + "learning_rate": 0.003, + "loss": 4.1392, + "step": 4592 + }, + { + "epoch": 0.04593, + "grad_norm": 0.7097656474131322, + "learning_rate": 0.003, + "loss": 4.1641, + "step": 4593 + }, + { + "epoch": 0.04594, + "grad_norm": 0.6175568409115185, + "learning_rate": 0.003, + "loss": 4.1529, + "step": 4594 + }, + { + "epoch": 0.04595, + "grad_norm": 0.620699543326453, + "learning_rate": 0.003, + "loss": 4.1404, + "step": 4595 + }, + { + "epoch": 0.04596, + "grad_norm": 0.5962342876114253, + "learning_rate": 0.003, + "loss": 4.1429, + "step": 4596 + }, + { + "epoch": 0.04597, + "grad_norm": 0.6324593221701938, + "learning_rate": 0.003, + "loss": 4.1494, + "step": 4597 + }, + { + "epoch": 0.04598, + "grad_norm": 0.6611397602971393, + "learning_rate": 0.003, + "loss": 4.1806, + "step": 4598 + }, + { + "epoch": 0.04599, + "grad_norm": 0.6234361489034259, + "learning_rate": 0.003, + "loss": 4.1279, + "step": 4599 + }, + { + "epoch": 0.046, + "grad_norm": 0.6075516815675354, + "learning_rate": 0.003, + "loss": 4.1403, + "step": 4600 + }, + { + "epoch": 0.04601, + "grad_norm": 0.6042104496947458, + "learning_rate": 0.003, + "loss": 4.1434, + "step": 4601 + }, + { + "epoch": 0.04602, + "grad_norm": 0.6164999421939368, + "learning_rate": 0.003, + "loss": 4.1389, + "step": 4602 + }, + { + "epoch": 0.04603, + "grad_norm": 0.6490308990358886, + "learning_rate": 0.003, + "loss": 4.1506, + "step": 4603 + }, + { + "epoch": 0.04604, + "grad_norm": 0.7661434840060077, + "learning_rate": 0.003, + "loss": 4.1474, + "step": 4604 + }, + { + "epoch": 0.04605, + "grad_norm": 1.0569916498623362, + "learning_rate": 0.003, + "loss": 4.1659, + "step": 4605 + }, + { + "epoch": 0.04606, + "grad_norm": 1.027715504643133, + "learning_rate": 0.003, + "loss": 4.1333, + "step": 4606 + }, + { + "epoch": 0.04607, + "grad_norm": 0.8342746434973768, + "learning_rate": 0.003, + "loss": 4.1373, + "step": 4607 + }, + { + "epoch": 0.04608, + "grad_norm": 0.7850330384795643, + "learning_rate": 0.003, + "loss": 4.121, + "step": 4608 + }, + { + "epoch": 0.04609, + "grad_norm": 0.7443196821550939, + "learning_rate": 0.003, + "loss": 4.1433, + "step": 4609 + }, + { + "epoch": 0.0461, + "grad_norm": 0.7785237638327334, + "learning_rate": 0.003, + "loss": 4.1237, + "step": 4610 + }, + { + "epoch": 0.04611, + "grad_norm": 0.8838166156331966, + "learning_rate": 0.003, + "loss": 4.1784, + "step": 4611 + }, + { + "epoch": 0.04612, + "grad_norm": 0.8874166663110993, + "learning_rate": 0.003, + "loss": 4.1625, + "step": 4612 + }, + { + "epoch": 0.04613, + "grad_norm": 0.8148264887171576, + "learning_rate": 0.003, + "loss": 4.1424, + "step": 4613 + }, + { + "epoch": 0.04614, + "grad_norm": 0.7820861967247282, + "learning_rate": 0.003, + "loss": 4.1074, + "step": 4614 + }, + { + "epoch": 0.04615, + "grad_norm": 0.7280484691047371, + "learning_rate": 0.003, + "loss": 4.1473, + "step": 4615 + }, + { + "epoch": 0.04616, + "grad_norm": 0.7490023925524317, + "learning_rate": 0.003, + "loss": 4.1301, + "step": 4616 + }, + { + "epoch": 0.04617, + "grad_norm": 0.7518558085096564, + "learning_rate": 0.003, + "loss": 4.1549, + "step": 4617 + }, + { + "epoch": 0.04618, + "grad_norm": 0.625986793466225, + "learning_rate": 0.003, + "loss": 4.134, + "step": 4618 + }, + { + "epoch": 0.04619, + "grad_norm": 0.6472629361360867, + "learning_rate": 0.003, + "loss": 4.1346, + "step": 4619 + }, + { + "epoch": 0.0462, + "grad_norm": 0.6557801977850327, + "learning_rate": 0.003, + "loss": 4.1625, + "step": 4620 + }, + { + "epoch": 0.04621, + "grad_norm": 0.73229721178544, + "learning_rate": 0.003, + "loss": 4.1693, + "step": 4621 + }, + { + "epoch": 0.04622, + "grad_norm": 0.9280201932819085, + "learning_rate": 0.003, + "loss": 4.1688, + "step": 4622 + }, + { + "epoch": 0.04623, + "grad_norm": 1.042833156417488, + "learning_rate": 0.003, + "loss": 4.1581, + "step": 4623 + }, + { + "epoch": 0.04624, + "grad_norm": 0.8256463072634586, + "learning_rate": 0.003, + "loss": 4.1292, + "step": 4624 + }, + { + "epoch": 0.04625, + "grad_norm": 0.7149715088227584, + "learning_rate": 0.003, + "loss": 4.1471, + "step": 4625 + }, + { + "epoch": 0.04626, + "grad_norm": 0.6877823576969472, + "learning_rate": 0.003, + "loss": 4.1522, + "step": 4626 + }, + { + "epoch": 0.04627, + "grad_norm": 0.6742469710012651, + "learning_rate": 0.003, + "loss": 4.1327, + "step": 4627 + }, + { + "epoch": 0.04628, + "grad_norm": 0.62532013220358, + "learning_rate": 0.003, + "loss": 4.1618, + "step": 4628 + }, + { + "epoch": 0.04629, + "grad_norm": 0.5130605726615276, + "learning_rate": 0.003, + "loss": 4.1316, + "step": 4629 + }, + { + "epoch": 0.0463, + "grad_norm": 0.5654280379865769, + "learning_rate": 0.003, + "loss": 4.1481, + "step": 4630 + }, + { + "epoch": 0.04631, + "grad_norm": 0.6014781265202847, + "learning_rate": 0.003, + "loss": 4.1292, + "step": 4631 + }, + { + "epoch": 0.04632, + "grad_norm": 0.6307983761305929, + "learning_rate": 0.003, + "loss": 4.1198, + "step": 4632 + }, + { + "epoch": 0.04633, + "grad_norm": 0.6189113146917462, + "learning_rate": 0.003, + "loss": 4.1255, + "step": 4633 + }, + { + "epoch": 0.04634, + "grad_norm": 0.6421436906797112, + "learning_rate": 0.003, + "loss": 4.1275, + "step": 4634 + }, + { + "epoch": 0.04635, + "grad_norm": 0.5590583262211648, + "learning_rate": 0.003, + "loss": 4.1297, + "step": 4635 + }, + { + "epoch": 0.04636, + "grad_norm": 0.503140076293387, + "learning_rate": 0.003, + "loss": 4.1643, + "step": 4636 + }, + { + "epoch": 0.04637, + "grad_norm": 0.5299879986072056, + "learning_rate": 0.003, + "loss": 4.1066, + "step": 4637 + }, + { + "epoch": 0.04638, + "grad_norm": 0.4214634549110216, + "learning_rate": 0.003, + "loss": 4.1395, + "step": 4638 + }, + { + "epoch": 0.04639, + "grad_norm": 0.397393509216712, + "learning_rate": 0.003, + "loss": 4.1094, + "step": 4639 + }, + { + "epoch": 0.0464, + "grad_norm": 0.43447214142466917, + "learning_rate": 0.003, + "loss": 4.134, + "step": 4640 + }, + { + "epoch": 0.04641, + "grad_norm": 0.5062252032170508, + "learning_rate": 0.003, + "loss": 4.1499, + "step": 4641 + }, + { + "epoch": 0.04642, + "grad_norm": 0.5711310545610656, + "learning_rate": 0.003, + "loss": 4.0962, + "step": 4642 + }, + { + "epoch": 0.04643, + "grad_norm": 0.6248454986616232, + "learning_rate": 0.003, + "loss": 4.1224, + "step": 4643 + }, + { + "epoch": 0.04644, + "grad_norm": 0.7210053036878883, + "learning_rate": 0.003, + "loss": 4.1317, + "step": 4644 + }, + { + "epoch": 0.04645, + "grad_norm": 0.90249716060301, + "learning_rate": 0.003, + "loss": 4.1299, + "step": 4645 + }, + { + "epoch": 0.04646, + "grad_norm": 1.2409050533897181, + "learning_rate": 0.003, + "loss": 4.1424, + "step": 4646 + }, + { + "epoch": 0.04647, + "grad_norm": 0.8393901607525605, + "learning_rate": 0.003, + "loss": 4.1312, + "step": 4647 + }, + { + "epoch": 0.04648, + "grad_norm": 0.7746500580810028, + "learning_rate": 0.003, + "loss": 4.1499, + "step": 4648 + }, + { + "epoch": 0.04649, + "grad_norm": 0.8739807831426791, + "learning_rate": 0.003, + "loss": 4.1372, + "step": 4649 + }, + { + "epoch": 0.0465, + "grad_norm": 0.9522286444793494, + "learning_rate": 0.003, + "loss": 4.1127, + "step": 4650 + }, + { + "epoch": 0.04651, + "grad_norm": 1.1296690683804316, + "learning_rate": 0.003, + "loss": 4.1538, + "step": 4651 + }, + { + "epoch": 0.04652, + "grad_norm": 0.9265569184990113, + "learning_rate": 0.003, + "loss": 4.1575, + "step": 4652 + }, + { + "epoch": 0.04653, + "grad_norm": 0.900473448775536, + "learning_rate": 0.003, + "loss": 4.1588, + "step": 4653 + }, + { + "epoch": 0.04654, + "grad_norm": 1.0602801619166438, + "learning_rate": 0.003, + "loss": 4.1627, + "step": 4654 + }, + { + "epoch": 0.04655, + "grad_norm": 0.9413459563751835, + "learning_rate": 0.003, + "loss": 4.1634, + "step": 4655 + }, + { + "epoch": 0.04656, + "grad_norm": 0.7433440151400471, + "learning_rate": 0.003, + "loss": 4.135, + "step": 4656 + }, + { + "epoch": 0.04657, + "grad_norm": 0.7855838722575367, + "learning_rate": 0.003, + "loss": 4.1731, + "step": 4657 + }, + { + "epoch": 0.04658, + "grad_norm": 0.7026074720403149, + "learning_rate": 0.003, + "loss": 4.1442, + "step": 4658 + }, + { + "epoch": 0.04659, + "grad_norm": 0.7054872170953301, + "learning_rate": 0.003, + "loss": 4.1422, + "step": 4659 + }, + { + "epoch": 0.0466, + "grad_norm": 0.7253225489512056, + "learning_rate": 0.003, + "loss": 4.162, + "step": 4660 + }, + { + "epoch": 0.04661, + "grad_norm": 0.6938113299882246, + "learning_rate": 0.003, + "loss": 4.1462, + "step": 4661 + }, + { + "epoch": 0.04662, + "grad_norm": 0.6661030601372132, + "learning_rate": 0.003, + "loss": 4.1305, + "step": 4662 + }, + { + "epoch": 0.04663, + "grad_norm": 0.7495766554410687, + "learning_rate": 0.003, + "loss": 4.152, + "step": 4663 + }, + { + "epoch": 0.04664, + "grad_norm": 0.7803399105366602, + "learning_rate": 0.003, + "loss": 4.129, + "step": 4664 + }, + { + "epoch": 0.04665, + "grad_norm": 0.7715228893902641, + "learning_rate": 0.003, + "loss": 4.1305, + "step": 4665 + }, + { + "epoch": 0.04666, + "grad_norm": 0.8189022468897561, + "learning_rate": 0.003, + "loss": 4.1281, + "step": 4666 + }, + { + "epoch": 0.04667, + "grad_norm": 0.9073148500434436, + "learning_rate": 0.003, + "loss": 4.1352, + "step": 4667 + }, + { + "epoch": 0.04668, + "grad_norm": 0.8663369305932774, + "learning_rate": 0.003, + "loss": 4.1091, + "step": 4668 + }, + { + "epoch": 0.04669, + "grad_norm": 0.7981760342295242, + "learning_rate": 0.003, + "loss": 4.1642, + "step": 4669 + }, + { + "epoch": 0.0467, + "grad_norm": 0.7399100042776996, + "learning_rate": 0.003, + "loss": 4.1113, + "step": 4670 + }, + { + "epoch": 0.04671, + "grad_norm": 0.79400390511152, + "learning_rate": 0.003, + "loss": 4.1397, + "step": 4671 + }, + { + "epoch": 0.04672, + "grad_norm": 0.7886276399898601, + "learning_rate": 0.003, + "loss": 4.1526, + "step": 4672 + }, + { + "epoch": 0.04673, + "grad_norm": 0.7587981117840258, + "learning_rate": 0.003, + "loss": 4.1559, + "step": 4673 + }, + { + "epoch": 0.04674, + "grad_norm": 0.6833849503950475, + "learning_rate": 0.003, + "loss": 4.1246, + "step": 4674 + }, + { + "epoch": 0.04675, + "grad_norm": 0.5521830886006064, + "learning_rate": 0.003, + "loss": 4.1453, + "step": 4675 + }, + { + "epoch": 0.04676, + "grad_norm": 0.5551519809403406, + "learning_rate": 0.003, + "loss": 4.1502, + "step": 4676 + }, + { + "epoch": 0.04677, + "grad_norm": 0.4956612676914781, + "learning_rate": 0.003, + "loss": 4.1512, + "step": 4677 + }, + { + "epoch": 0.04678, + "grad_norm": 0.5248385967176291, + "learning_rate": 0.003, + "loss": 4.1077, + "step": 4678 + }, + { + "epoch": 0.04679, + "grad_norm": 0.5169108616408382, + "learning_rate": 0.003, + "loss": 4.1375, + "step": 4679 + }, + { + "epoch": 0.0468, + "grad_norm": 0.7044973029382834, + "learning_rate": 0.003, + "loss": 4.111, + "step": 4680 + }, + { + "epoch": 0.04681, + "grad_norm": 0.845934209130146, + "learning_rate": 0.003, + "loss": 4.1661, + "step": 4681 + }, + { + "epoch": 0.04682, + "grad_norm": 0.8277130051517226, + "learning_rate": 0.003, + "loss": 4.1349, + "step": 4682 + }, + { + "epoch": 0.04683, + "grad_norm": 0.7576471069505033, + "learning_rate": 0.003, + "loss": 4.1267, + "step": 4683 + }, + { + "epoch": 0.04684, + "grad_norm": 0.6931368300408345, + "learning_rate": 0.003, + "loss": 4.1455, + "step": 4684 + }, + { + "epoch": 0.04685, + "grad_norm": 0.6709891251557847, + "learning_rate": 0.003, + "loss": 4.1292, + "step": 4685 + }, + { + "epoch": 0.04686, + "grad_norm": 0.5581781513115139, + "learning_rate": 0.003, + "loss": 4.1336, + "step": 4686 + }, + { + "epoch": 0.04687, + "grad_norm": 0.6261887980349435, + "learning_rate": 0.003, + "loss": 4.1643, + "step": 4687 + }, + { + "epoch": 0.04688, + "grad_norm": 0.7985962391701531, + "learning_rate": 0.003, + "loss": 4.1465, + "step": 4688 + }, + { + "epoch": 0.04689, + "grad_norm": 1.0550803093277015, + "learning_rate": 0.003, + "loss": 4.1722, + "step": 4689 + }, + { + "epoch": 0.0469, + "grad_norm": 0.9618900206251452, + "learning_rate": 0.003, + "loss": 4.1247, + "step": 4690 + }, + { + "epoch": 0.04691, + "grad_norm": 0.6523829450195608, + "learning_rate": 0.003, + "loss": 4.1317, + "step": 4691 + }, + { + "epoch": 0.04692, + "grad_norm": 0.6896621979107347, + "learning_rate": 0.003, + "loss": 4.1375, + "step": 4692 + }, + { + "epoch": 0.04693, + "grad_norm": 0.9401038472008131, + "learning_rate": 0.003, + "loss": 4.1706, + "step": 4693 + }, + { + "epoch": 0.04694, + "grad_norm": 0.9720475262714349, + "learning_rate": 0.003, + "loss": 4.1363, + "step": 4694 + }, + { + "epoch": 0.04695, + "grad_norm": 0.9719251158591558, + "learning_rate": 0.003, + "loss": 4.1313, + "step": 4695 + }, + { + "epoch": 0.04696, + "grad_norm": 0.8427853123728982, + "learning_rate": 0.003, + "loss": 4.1711, + "step": 4696 + }, + { + "epoch": 0.04697, + "grad_norm": 0.8499034656864161, + "learning_rate": 0.003, + "loss": 4.1544, + "step": 4697 + }, + { + "epoch": 0.04698, + "grad_norm": 0.7021228867021835, + "learning_rate": 0.003, + "loss": 4.1502, + "step": 4698 + }, + { + "epoch": 0.04699, + "grad_norm": 0.7398294832914439, + "learning_rate": 0.003, + "loss": 4.144, + "step": 4699 + }, + { + "epoch": 0.047, + "grad_norm": 0.6942643738589905, + "learning_rate": 0.003, + "loss": 4.1544, + "step": 4700 + }, + { + "epoch": 0.04701, + "grad_norm": 0.6683525277935326, + "learning_rate": 0.003, + "loss": 4.1603, + "step": 4701 + }, + { + "epoch": 0.04702, + "grad_norm": 0.5837305852992254, + "learning_rate": 0.003, + "loss": 4.151, + "step": 4702 + }, + { + "epoch": 0.04703, + "grad_norm": 0.5004181371776402, + "learning_rate": 0.003, + "loss": 4.1344, + "step": 4703 + }, + { + "epoch": 0.04704, + "grad_norm": 0.500112229497055, + "learning_rate": 0.003, + "loss": 4.1276, + "step": 4704 + }, + { + "epoch": 0.04705, + "grad_norm": 0.46870504628905124, + "learning_rate": 0.003, + "loss": 4.1571, + "step": 4705 + }, + { + "epoch": 0.04706, + "grad_norm": 0.5084124708467841, + "learning_rate": 0.003, + "loss": 4.0994, + "step": 4706 + }, + { + "epoch": 0.04707, + "grad_norm": 0.5592830079070097, + "learning_rate": 0.003, + "loss": 4.1412, + "step": 4707 + }, + { + "epoch": 0.04708, + "grad_norm": 0.6734818421768685, + "learning_rate": 0.003, + "loss": 4.0984, + "step": 4708 + }, + { + "epoch": 0.04709, + "grad_norm": 0.7820705785181057, + "learning_rate": 0.003, + "loss": 4.1325, + "step": 4709 + }, + { + "epoch": 0.0471, + "grad_norm": 0.8173042330246184, + "learning_rate": 0.003, + "loss": 4.1472, + "step": 4710 + }, + { + "epoch": 0.04711, + "grad_norm": 0.7323489999342726, + "learning_rate": 0.003, + "loss": 4.1382, + "step": 4711 + }, + { + "epoch": 0.04712, + "grad_norm": 0.6858752455645303, + "learning_rate": 0.003, + "loss": 4.1261, + "step": 4712 + }, + { + "epoch": 0.04713, + "grad_norm": 0.5541474886267923, + "learning_rate": 0.003, + "loss": 4.1189, + "step": 4713 + }, + { + "epoch": 0.04714, + "grad_norm": 0.5810454942772322, + "learning_rate": 0.003, + "loss": 4.1281, + "step": 4714 + }, + { + "epoch": 0.04715, + "grad_norm": 0.6021618978261117, + "learning_rate": 0.003, + "loss": 4.1564, + "step": 4715 + }, + { + "epoch": 0.04716, + "grad_norm": 0.6343272321037665, + "learning_rate": 0.003, + "loss": 4.1271, + "step": 4716 + }, + { + "epoch": 0.04717, + "grad_norm": 0.7013052653860089, + "learning_rate": 0.003, + "loss": 4.1476, + "step": 4717 + }, + { + "epoch": 0.04718, + "grad_norm": 0.66349485254104, + "learning_rate": 0.003, + "loss": 4.1109, + "step": 4718 + }, + { + "epoch": 0.04719, + "grad_norm": 0.8713918643231059, + "learning_rate": 0.003, + "loss": 4.127, + "step": 4719 + }, + { + "epoch": 0.0472, + "grad_norm": 1.0008190365871503, + "learning_rate": 0.003, + "loss": 4.1663, + "step": 4720 + }, + { + "epoch": 0.04721, + "grad_norm": 0.9716425512049767, + "learning_rate": 0.003, + "loss": 4.1264, + "step": 4721 + }, + { + "epoch": 0.04722, + "grad_norm": 1.1120781580345849, + "learning_rate": 0.003, + "loss": 4.1498, + "step": 4722 + }, + { + "epoch": 0.04723, + "grad_norm": 0.8161047263055594, + "learning_rate": 0.003, + "loss": 4.1529, + "step": 4723 + }, + { + "epoch": 0.04724, + "grad_norm": 0.8833523326572259, + "learning_rate": 0.003, + "loss": 4.1324, + "step": 4724 + }, + { + "epoch": 0.04725, + "grad_norm": 0.8898592287073411, + "learning_rate": 0.003, + "loss": 4.1569, + "step": 4725 + }, + { + "epoch": 0.04726, + "grad_norm": 0.9859327457193534, + "learning_rate": 0.003, + "loss": 4.1487, + "step": 4726 + }, + { + "epoch": 0.04727, + "grad_norm": 0.9165866060948171, + "learning_rate": 0.003, + "loss": 4.1321, + "step": 4727 + }, + { + "epoch": 0.04728, + "grad_norm": 0.9068684412672875, + "learning_rate": 0.003, + "loss": 4.1421, + "step": 4728 + }, + { + "epoch": 0.04729, + "grad_norm": 0.8318177657045663, + "learning_rate": 0.003, + "loss": 4.1603, + "step": 4729 + }, + { + "epoch": 0.0473, + "grad_norm": 0.8027800490874897, + "learning_rate": 0.003, + "loss": 4.1585, + "step": 4730 + }, + { + "epoch": 0.04731, + "grad_norm": 0.7978896274361092, + "learning_rate": 0.003, + "loss": 4.1507, + "step": 4731 + }, + { + "epoch": 0.04732, + "grad_norm": 0.8802600387002112, + "learning_rate": 0.003, + "loss": 4.164, + "step": 4732 + }, + { + "epoch": 0.04733, + "grad_norm": 0.8234843622598483, + "learning_rate": 0.003, + "loss": 4.148, + "step": 4733 + }, + { + "epoch": 0.04734, + "grad_norm": 0.6599234947722757, + "learning_rate": 0.003, + "loss": 4.1501, + "step": 4734 + }, + { + "epoch": 0.04735, + "grad_norm": 0.8208592352781166, + "learning_rate": 0.003, + "loss": 4.1681, + "step": 4735 + }, + { + "epoch": 0.04736, + "grad_norm": 1.1729409563655009, + "learning_rate": 0.003, + "loss": 4.1639, + "step": 4736 + }, + { + "epoch": 0.04737, + "grad_norm": 0.9659463338122776, + "learning_rate": 0.003, + "loss": 4.1274, + "step": 4737 + }, + { + "epoch": 0.04738, + "grad_norm": 0.7651058866068673, + "learning_rate": 0.003, + "loss": 4.1646, + "step": 4738 + }, + { + "epoch": 0.04739, + "grad_norm": 0.717869067195337, + "learning_rate": 0.003, + "loss": 4.0994, + "step": 4739 + }, + { + "epoch": 0.0474, + "grad_norm": 0.6851887051422195, + "learning_rate": 0.003, + "loss": 4.1663, + "step": 4740 + }, + { + "epoch": 0.04741, + "grad_norm": 0.6717713840804643, + "learning_rate": 0.003, + "loss": 4.1559, + "step": 4741 + }, + { + "epoch": 0.04742, + "grad_norm": 0.6107738091433065, + "learning_rate": 0.003, + "loss": 4.1372, + "step": 4742 + }, + { + "epoch": 0.04743, + "grad_norm": 0.5441883526059924, + "learning_rate": 0.003, + "loss": 4.1554, + "step": 4743 + }, + { + "epoch": 0.04744, + "grad_norm": 0.5198168694209138, + "learning_rate": 0.003, + "loss": 4.1156, + "step": 4744 + }, + { + "epoch": 0.04745, + "grad_norm": 0.43524115400010815, + "learning_rate": 0.003, + "loss": 4.117, + "step": 4745 + }, + { + "epoch": 0.04746, + "grad_norm": 0.48653308583746263, + "learning_rate": 0.003, + "loss": 4.1322, + "step": 4746 + }, + { + "epoch": 0.04747, + "grad_norm": 0.5339120307603674, + "learning_rate": 0.003, + "loss": 4.1062, + "step": 4747 + }, + { + "epoch": 0.04748, + "grad_norm": 0.5479071965461091, + "learning_rate": 0.003, + "loss": 4.1554, + "step": 4748 + }, + { + "epoch": 0.04749, + "grad_norm": 0.5417055693516375, + "learning_rate": 0.003, + "loss": 4.1617, + "step": 4749 + }, + { + "epoch": 0.0475, + "grad_norm": 0.5280951958997921, + "learning_rate": 0.003, + "loss": 4.1257, + "step": 4750 + }, + { + "epoch": 0.04751, + "grad_norm": 0.5350329833617672, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 4751 + }, + { + "epoch": 0.04752, + "grad_norm": 0.57851596805048, + "learning_rate": 0.003, + "loss": 4.1221, + "step": 4752 + }, + { + "epoch": 0.04753, + "grad_norm": 0.72902319198213, + "learning_rate": 0.003, + "loss": 4.1305, + "step": 4753 + }, + { + "epoch": 0.04754, + "grad_norm": 0.871184049188565, + "learning_rate": 0.003, + "loss": 4.1544, + "step": 4754 + }, + { + "epoch": 0.04755, + "grad_norm": 0.9103037208186765, + "learning_rate": 0.003, + "loss": 4.1375, + "step": 4755 + }, + { + "epoch": 0.04756, + "grad_norm": 0.859550535967343, + "learning_rate": 0.003, + "loss": 4.1417, + "step": 4756 + }, + { + "epoch": 0.04757, + "grad_norm": 0.7466205077193813, + "learning_rate": 0.003, + "loss": 4.1385, + "step": 4757 + }, + { + "epoch": 0.04758, + "grad_norm": 0.7507245421944043, + "learning_rate": 0.003, + "loss": 4.1287, + "step": 4758 + }, + { + "epoch": 0.04759, + "grad_norm": 0.8294869880303999, + "learning_rate": 0.003, + "loss": 4.1452, + "step": 4759 + }, + { + "epoch": 0.0476, + "grad_norm": 0.8251883621259142, + "learning_rate": 0.003, + "loss": 4.1243, + "step": 4760 + }, + { + "epoch": 0.04761, + "grad_norm": 0.7765732752545424, + "learning_rate": 0.003, + "loss": 4.1329, + "step": 4761 + }, + { + "epoch": 0.04762, + "grad_norm": 0.6280062378308506, + "learning_rate": 0.003, + "loss": 4.1228, + "step": 4762 + }, + { + "epoch": 0.04763, + "grad_norm": 0.5686058703500096, + "learning_rate": 0.003, + "loss": 4.1247, + "step": 4763 + }, + { + "epoch": 0.04764, + "grad_norm": 0.46584482494791357, + "learning_rate": 0.003, + "loss": 4.1095, + "step": 4764 + }, + { + "epoch": 0.04765, + "grad_norm": 0.45316919798397814, + "learning_rate": 0.003, + "loss": 4.1591, + "step": 4765 + }, + { + "epoch": 0.04766, + "grad_norm": 0.5554760863440187, + "learning_rate": 0.003, + "loss": 4.1327, + "step": 4766 + }, + { + "epoch": 0.04767, + "grad_norm": 0.6418170833104156, + "learning_rate": 0.003, + "loss": 4.1325, + "step": 4767 + }, + { + "epoch": 0.04768, + "grad_norm": 0.5645471402502725, + "learning_rate": 0.003, + "loss": 4.1237, + "step": 4768 + }, + { + "epoch": 0.04769, + "grad_norm": 0.5927099803110258, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 4769 + }, + { + "epoch": 0.0477, + "grad_norm": 0.6595637919437056, + "learning_rate": 0.003, + "loss": 4.1055, + "step": 4770 + }, + { + "epoch": 0.04771, + "grad_norm": 0.5520312176680346, + "learning_rate": 0.003, + "loss": 4.1292, + "step": 4771 + }, + { + "epoch": 0.04772, + "grad_norm": 0.5616366446629109, + "learning_rate": 0.003, + "loss": 4.1263, + "step": 4772 + }, + { + "epoch": 0.04773, + "grad_norm": 0.7069869424163182, + "learning_rate": 0.003, + "loss": 4.1512, + "step": 4773 + }, + { + "epoch": 0.04774, + "grad_norm": 0.8943049067297537, + "learning_rate": 0.003, + "loss": 4.1426, + "step": 4774 + }, + { + "epoch": 0.04775, + "grad_norm": 1.0379517887129035, + "learning_rate": 0.003, + "loss": 4.1448, + "step": 4775 + }, + { + "epoch": 0.04776, + "grad_norm": 1.0601938367245354, + "learning_rate": 0.003, + "loss": 4.1369, + "step": 4776 + }, + { + "epoch": 0.04777, + "grad_norm": 0.8123830997039955, + "learning_rate": 0.003, + "loss": 4.13, + "step": 4777 + }, + { + "epoch": 0.04778, + "grad_norm": 0.6466306116567022, + "learning_rate": 0.003, + "loss": 4.1035, + "step": 4778 + }, + { + "epoch": 0.04779, + "grad_norm": 0.8059965327611258, + "learning_rate": 0.003, + "loss": 4.1478, + "step": 4779 + }, + { + "epoch": 0.0478, + "grad_norm": 1.005661350204286, + "learning_rate": 0.003, + "loss": 4.1578, + "step": 4780 + }, + { + "epoch": 0.04781, + "grad_norm": 0.9783065445314736, + "learning_rate": 0.003, + "loss": 4.1382, + "step": 4781 + }, + { + "epoch": 0.04782, + "grad_norm": 0.9777333193008619, + "learning_rate": 0.003, + "loss": 4.1427, + "step": 4782 + }, + { + "epoch": 0.04783, + "grad_norm": 1.0138916385667516, + "learning_rate": 0.003, + "loss": 4.1442, + "step": 4783 + }, + { + "epoch": 0.04784, + "grad_norm": 1.0210591580782744, + "learning_rate": 0.003, + "loss": 4.1688, + "step": 4784 + }, + { + "epoch": 0.04785, + "grad_norm": 1.0867592349380752, + "learning_rate": 0.003, + "loss": 4.1783, + "step": 4785 + }, + { + "epoch": 0.04786, + "grad_norm": 0.9270426683995183, + "learning_rate": 0.003, + "loss": 4.1726, + "step": 4786 + }, + { + "epoch": 0.04787, + "grad_norm": 0.9581278828591302, + "learning_rate": 0.003, + "loss": 4.1717, + "step": 4787 + }, + { + "epoch": 0.04788, + "grad_norm": 1.0463519085350252, + "learning_rate": 0.003, + "loss": 4.1652, + "step": 4788 + }, + { + "epoch": 0.04789, + "grad_norm": 1.048457678252623, + "learning_rate": 0.003, + "loss": 4.149, + "step": 4789 + }, + { + "epoch": 0.0479, + "grad_norm": 1.0802030566717422, + "learning_rate": 0.003, + "loss": 4.1693, + "step": 4790 + }, + { + "epoch": 0.04791, + "grad_norm": 0.9188967498168278, + "learning_rate": 0.003, + "loss": 4.1796, + "step": 4791 + }, + { + "epoch": 0.04792, + "grad_norm": 1.0152593647469241, + "learning_rate": 0.003, + "loss": 4.1927, + "step": 4792 + }, + { + "epoch": 0.04793, + "grad_norm": 1.018584486769021, + "learning_rate": 0.003, + "loss": 4.1531, + "step": 4793 + }, + { + "epoch": 0.04794, + "grad_norm": 0.8402721654087, + "learning_rate": 0.003, + "loss": 4.1722, + "step": 4794 + }, + { + "epoch": 0.04795, + "grad_norm": 0.7643727384922673, + "learning_rate": 0.003, + "loss": 4.1655, + "step": 4795 + }, + { + "epoch": 0.04796, + "grad_norm": 0.813774811366339, + "learning_rate": 0.003, + "loss": 4.137, + "step": 4796 + }, + { + "epoch": 0.04797, + "grad_norm": 0.6494127001881433, + "learning_rate": 0.003, + "loss": 4.1645, + "step": 4797 + }, + { + "epoch": 0.04798, + "grad_norm": 0.664089910476502, + "learning_rate": 0.003, + "loss": 4.1242, + "step": 4798 + }, + { + "epoch": 0.04799, + "grad_norm": 0.6031144588619443, + "learning_rate": 0.003, + "loss": 4.1338, + "step": 4799 + }, + { + "epoch": 0.048, + "grad_norm": 0.5598439245281802, + "learning_rate": 0.003, + "loss": 4.1362, + "step": 4800 + }, + { + "epoch": 0.04801, + "grad_norm": 0.6477000775289928, + "learning_rate": 0.003, + "loss": 4.1467, + "step": 4801 + }, + { + "epoch": 0.04802, + "grad_norm": 0.706840356865147, + "learning_rate": 0.003, + "loss": 4.1209, + "step": 4802 + }, + { + "epoch": 0.04803, + "grad_norm": 0.687443212408336, + "learning_rate": 0.003, + "loss": 4.1292, + "step": 4803 + }, + { + "epoch": 0.04804, + "grad_norm": 0.6817235142027526, + "learning_rate": 0.003, + "loss": 4.1578, + "step": 4804 + }, + { + "epoch": 0.04805, + "grad_norm": 0.7014528353258211, + "learning_rate": 0.003, + "loss": 4.1588, + "step": 4805 + }, + { + "epoch": 0.04806, + "grad_norm": 0.6063087225759103, + "learning_rate": 0.003, + "loss": 4.1019, + "step": 4806 + }, + { + "epoch": 0.04807, + "grad_norm": 0.4870524793552668, + "learning_rate": 0.003, + "loss": 4.1518, + "step": 4807 + }, + { + "epoch": 0.04808, + "grad_norm": 0.4304737736314177, + "learning_rate": 0.003, + "loss": 4.1263, + "step": 4808 + }, + { + "epoch": 0.04809, + "grad_norm": 0.444058010799397, + "learning_rate": 0.003, + "loss": 4.1187, + "step": 4809 + }, + { + "epoch": 0.0481, + "grad_norm": 0.44159249278167406, + "learning_rate": 0.003, + "loss": 4.1384, + "step": 4810 + }, + { + "epoch": 0.04811, + "grad_norm": 0.46942470854293217, + "learning_rate": 0.003, + "loss": 4.1346, + "step": 4811 + }, + { + "epoch": 0.04812, + "grad_norm": 0.4477933921171643, + "learning_rate": 0.003, + "loss": 4.1339, + "step": 4812 + }, + { + "epoch": 0.04813, + "grad_norm": 0.43354928602746334, + "learning_rate": 0.003, + "loss": 4.1197, + "step": 4813 + }, + { + "epoch": 0.04814, + "grad_norm": 0.45498153005997544, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 4814 + }, + { + "epoch": 0.04815, + "grad_norm": 0.48463241046460065, + "learning_rate": 0.003, + "loss": 4.1186, + "step": 4815 + }, + { + "epoch": 0.04816, + "grad_norm": 0.5311818047677171, + "learning_rate": 0.003, + "loss": 4.1276, + "step": 4816 + }, + { + "epoch": 0.04817, + "grad_norm": 0.631506579388757, + "learning_rate": 0.003, + "loss": 4.1052, + "step": 4817 + }, + { + "epoch": 0.04818, + "grad_norm": 0.8170443890052518, + "learning_rate": 0.003, + "loss": 4.123, + "step": 4818 + }, + { + "epoch": 0.04819, + "grad_norm": 1.172132734592305, + "learning_rate": 0.003, + "loss": 4.1381, + "step": 4819 + }, + { + "epoch": 0.0482, + "grad_norm": 0.838017513510012, + "learning_rate": 0.003, + "loss": 4.121, + "step": 4820 + }, + { + "epoch": 0.04821, + "grad_norm": 0.6975770749182754, + "learning_rate": 0.003, + "loss": 4.1459, + "step": 4821 + }, + { + "epoch": 0.04822, + "grad_norm": 0.7527718267712026, + "learning_rate": 0.003, + "loss": 4.126, + "step": 4822 + }, + { + "epoch": 0.04823, + "grad_norm": 0.7414326559856069, + "learning_rate": 0.003, + "loss": 4.1219, + "step": 4823 + }, + { + "epoch": 0.04824, + "grad_norm": 0.858775362232865, + "learning_rate": 0.003, + "loss": 4.1143, + "step": 4824 + }, + { + "epoch": 0.04825, + "grad_norm": 0.695583690590602, + "learning_rate": 0.003, + "loss": 4.1259, + "step": 4825 + }, + { + "epoch": 0.04826, + "grad_norm": 0.5781760330827292, + "learning_rate": 0.003, + "loss": 4.126, + "step": 4826 + }, + { + "epoch": 0.04827, + "grad_norm": 0.5806936123154903, + "learning_rate": 0.003, + "loss": 4.1311, + "step": 4827 + }, + { + "epoch": 0.04828, + "grad_norm": 0.7721046134763737, + "learning_rate": 0.003, + "loss": 4.1447, + "step": 4828 + }, + { + "epoch": 0.04829, + "grad_norm": 0.7885233902013562, + "learning_rate": 0.003, + "loss": 4.1266, + "step": 4829 + }, + { + "epoch": 0.0483, + "grad_norm": 0.7091409575187497, + "learning_rate": 0.003, + "loss": 4.1361, + "step": 4830 + }, + { + "epoch": 0.04831, + "grad_norm": 0.7110846809520083, + "learning_rate": 0.003, + "loss": 4.1021, + "step": 4831 + }, + { + "epoch": 0.04832, + "grad_norm": 0.680243507107158, + "learning_rate": 0.003, + "loss": 4.1282, + "step": 4832 + }, + { + "epoch": 0.04833, + "grad_norm": 0.7284136950991514, + "learning_rate": 0.003, + "loss": 4.1327, + "step": 4833 + }, + { + "epoch": 0.04834, + "grad_norm": 0.7137894246322048, + "learning_rate": 0.003, + "loss": 4.1037, + "step": 4834 + }, + { + "epoch": 0.04835, + "grad_norm": 0.636140630835025, + "learning_rate": 0.003, + "loss": 4.1411, + "step": 4835 + }, + { + "epoch": 0.04836, + "grad_norm": 0.75652748414691, + "learning_rate": 0.003, + "loss": 4.1105, + "step": 4836 + }, + { + "epoch": 0.04837, + "grad_norm": 0.785548659376736, + "learning_rate": 0.003, + "loss": 4.1377, + "step": 4837 + }, + { + "epoch": 0.04838, + "grad_norm": 0.8897930937393894, + "learning_rate": 0.003, + "loss": 4.1521, + "step": 4838 + }, + { + "epoch": 0.04839, + "grad_norm": 0.9660346509912735, + "learning_rate": 0.003, + "loss": 4.1398, + "step": 4839 + }, + { + "epoch": 0.0484, + "grad_norm": 0.9300911895678211, + "learning_rate": 0.003, + "loss": 4.1437, + "step": 4840 + }, + { + "epoch": 0.04841, + "grad_norm": 0.7818946985346545, + "learning_rate": 0.003, + "loss": 4.1377, + "step": 4841 + }, + { + "epoch": 0.04842, + "grad_norm": 0.7032209957018698, + "learning_rate": 0.003, + "loss": 4.1538, + "step": 4842 + }, + { + "epoch": 0.04843, + "grad_norm": 0.6722771974917844, + "learning_rate": 0.003, + "loss": 4.1053, + "step": 4843 + }, + { + "epoch": 0.04844, + "grad_norm": 0.6970990299983332, + "learning_rate": 0.003, + "loss": 4.1379, + "step": 4844 + }, + { + "epoch": 0.04845, + "grad_norm": 0.6367868288526063, + "learning_rate": 0.003, + "loss": 4.1214, + "step": 4845 + }, + { + "epoch": 0.04846, + "grad_norm": 0.6829062431633811, + "learning_rate": 0.003, + "loss": 4.146, + "step": 4846 + }, + { + "epoch": 0.04847, + "grad_norm": 0.6460341296632274, + "learning_rate": 0.003, + "loss": 4.1086, + "step": 4847 + }, + { + "epoch": 0.04848, + "grad_norm": 0.6713362647372938, + "learning_rate": 0.003, + "loss": 4.1515, + "step": 4848 + }, + { + "epoch": 0.04849, + "grad_norm": 0.6714682036160667, + "learning_rate": 0.003, + "loss": 4.1426, + "step": 4849 + }, + { + "epoch": 0.0485, + "grad_norm": 0.7057113852906085, + "learning_rate": 0.003, + "loss": 4.132, + "step": 4850 + }, + { + "epoch": 0.04851, + "grad_norm": 0.6646770579142436, + "learning_rate": 0.003, + "loss": 4.1472, + "step": 4851 + }, + { + "epoch": 0.04852, + "grad_norm": 0.6577931410879579, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 4852 + }, + { + "epoch": 0.04853, + "grad_norm": 0.6521737566217662, + "learning_rate": 0.003, + "loss": 4.1384, + "step": 4853 + }, + { + "epoch": 0.04854, + "grad_norm": 0.7638656389712092, + "learning_rate": 0.003, + "loss": 4.1383, + "step": 4854 + }, + { + "epoch": 0.04855, + "grad_norm": 0.9457310244447173, + "learning_rate": 0.003, + "loss": 4.1514, + "step": 4855 + }, + { + "epoch": 0.04856, + "grad_norm": 1.1517585285172265, + "learning_rate": 0.003, + "loss": 4.1588, + "step": 4856 + }, + { + "epoch": 0.04857, + "grad_norm": 0.9746651389714764, + "learning_rate": 0.003, + "loss": 4.1672, + "step": 4857 + }, + { + "epoch": 0.04858, + "grad_norm": 0.8324446739497157, + "learning_rate": 0.003, + "loss": 4.135, + "step": 4858 + }, + { + "epoch": 0.04859, + "grad_norm": 0.7506712881500571, + "learning_rate": 0.003, + "loss": 4.1678, + "step": 4859 + }, + { + "epoch": 0.0486, + "grad_norm": 0.7957902928578074, + "learning_rate": 0.003, + "loss": 4.1483, + "step": 4860 + }, + { + "epoch": 0.04861, + "grad_norm": 1.0534393033014007, + "learning_rate": 0.003, + "loss": 4.1484, + "step": 4861 + }, + { + "epoch": 0.04862, + "grad_norm": 1.072043293213455, + "learning_rate": 0.003, + "loss": 4.1568, + "step": 4862 + }, + { + "epoch": 0.04863, + "grad_norm": 0.8542078395141963, + "learning_rate": 0.003, + "loss": 4.1674, + "step": 4863 + }, + { + "epoch": 0.04864, + "grad_norm": 0.979149461429483, + "learning_rate": 0.003, + "loss": 4.1489, + "step": 4864 + }, + { + "epoch": 0.04865, + "grad_norm": 0.9889727360280101, + "learning_rate": 0.003, + "loss": 4.1728, + "step": 4865 + }, + { + "epoch": 0.04866, + "grad_norm": 0.87668738837822, + "learning_rate": 0.003, + "loss": 4.176, + "step": 4866 + }, + { + "epoch": 0.04867, + "grad_norm": 0.9480918411172308, + "learning_rate": 0.003, + "loss": 4.1508, + "step": 4867 + }, + { + "epoch": 0.04868, + "grad_norm": 0.8900699953728738, + "learning_rate": 0.003, + "loss": 4.1571, + "step": 4868 + }, + { + "epoch": 0.04869, + "grad_norm": 0.8410323236753658, + "learning_rate": 0.003, + "loss": 4.169, + "step": 4869 + }, + { + "epoch": 0.0487, + "grad_norm": 0.8052365609623746, + "learning_rate": 0.003, + "loss": 4.1563, + "step": 4870 + }, + { + "epoch": 0.04871, + "grad_norm": 0.732052294025559, + "learning_rate": 0.003, + "loss": 4.1727, + "step": 4871 + }, + { + "epoch": 0.04872, + "grad_norm": 0.6093794224074204, + "learning_rate": 0.003, + "loss": 4.1332, + "step": 4872 + }, + { + "epoch": 0.04873, + "grad_norm": 0.5550033492784485, + "learning_rate": 0.003, + "loss": 4.1554, + "step": 4873 + }, + { + "epoch": 0.04874, + "grad_norm": 0.6407687802372767, + "learning_rate": 0.003, + "loss": 4.1462, + "step": 4874 + }, + { + "epoch": 0.04875, + "grad_norm": 0.6445083276518792, + "learning_rate": 0.003, + "loss": 4.1398, + "step": 4875 + }, + { + "epoch": 0.04876, + "grad_norm": 0.598683403431915, + "learning_rate": 0.003, + "loss": 4.1391, + "step": 4876 + }, + { + "epoch": 0.04877, + "grad_norm": 0.5879420193521339, + "learning_rate": 0.003, + "loss": 4.1392, + "step": 4877 + }, + { + "epoch": 0.04878, + "grad_norm": 0.6015687919635526, + "learning_rate": 0.003, + "loss": 4.1193, + "step": 4878 + }, + { + "epoch": 0.04879, + "grad_norm": 0.6406016359581286, + "learning_rate": 0.003, + "loss": 4.1157, + "step": 4879 + }, + { + "epoch": 0.0488, + "grad_norm": 0.6520274839816634, + "learning_rate": 0.003, + "loss": 4.1271, + "step": 4880 + }, + { + "epoch": 0.04881, + "grad_norm": 0.6746577474599795, + "learning_rate": 0.003, + "loss": 4.1215, + "step": 4881 + }, + { + "epoch": 0.04882, + "grad_norm": 0.6988276300361237, + "learning_rate": 0.003, + "loss": 4.1281, + "step": 4882 + }, + { + "epoch": 0.04883, + "grad_norm": 0.637601167920116, + "learning_rate": 0.003, + "loss": 4.1203, + "step": 4883 + }, + { + "epoch": 0.04884, + "grad_norm": 0.6575805356479, + "learning_rate": 0.003, + "loss": 4.1259, + "step": 4884 + }, + { + "epoch": 0.04885, + "grad_norm": 0.8560411189615229, + "learning_rate": 0.003, + "loss": 4.1406, + "step": 4885 + }, + { + "epoch": 0.04886, + "grad_norm": 1.131590855439064, + "learning_rate": 0.003, + "loss": 4.1439, + "step": 4886 + }, + { + "epoch": 0.04887, + "grad_norm": 0.8261960657958597, + "learning_rate": 0.003, + "loss": 4.1505, + "step": 4887 + }, + { + "epoch": 0.04888, + "grad_norm": 0.5889636338082994, + "learning_rate": 0.003, + "loss": 4.147, + "step": 4888 + }, + { + "epoch": 0.04889, + "grad_norm": 0.6875568675189105, + "learning_rate": 0.003, + "loss": 4.1272, + "step": 4889 + }, + { + "epoch": 0.0489, + "grad_norm": 0.8178506331950943, + "learning_rate": 0.003, + "loss": 4.1384, + "step": 4890 + }, + { + "epoch": 0.04891, + "grad_norm": 0.8524321612267298, + "learning_rate": 0.003, + "loss": 4.1499, + "step": 4891 + }, + { + "epoch": 0.04892, + "grad_norm": 0.6402952785912811, + "learning_rate": 0.003, + "loss": 4.1436, + "step": 4892 + }, + { + "epoch": 0.04893, + "grad_norm": 0.49571623322484065, + "learning_rate": 0.003, + "loss": 4.1372, + "step": 4893 + }, + { + "epoch": 0.04894, + "grad_norm": 0.6005303000219224, + "learning_rate": 0.003, + "loss": 4.1485, + "step": 4894 + }, + { + "epoch": 0.04895, + "grad_norm": 0.7874188186491505, + "learning_rate": 0.003, + "loss": 4.1285, + "step": 4895 + }, + { + "epoch": 0.04896, + "grad_norm": 0.8796160513302255, + "learning_rate": 0.003, + "loss": 4.1496, + "step": 4896 + }, + { + "epoch": 0.04897, + "grad_norm": 0.7789626736982245, + "learning_rate": 0.003, + "loss": 4.1546, + "step": 4897 + }, + { + "epoch": 0.04898, + "grad_norm": 0.7938347458338417, + "learning_rate": 0.003, + "loss": 4.1434, + "step": 4898 + }, + { + "epoch": 0.04899, + "grad_norm": 0.6968501707866385, + "learning_rate": 0.003, + "loss": 4.1295, + "step": 4899 + }, + { + "epoch": 0.049, + "grad_norm": 0.6737103261211691, + "learning_rate": 0.003, + "loss": 4.1563, + "step": 4900 + }, + { + "epoch": 0.04901, + "grad_norm": 0.6928180762743227, + "learning_rate": 0.003, + "loss": 4.1181, + "step": 4901 + }, + { + "epoch": 0.04902, + "grad_norm": 0.7387421279656656, + "learning_rate": 0.003, + "loss": 4.1554, + "step": 4902 + }, + { + "epoch": 0.04903, + "grad_norm": 0.7365881444221591, + "learning_rate": 0.003, + "loss": 4.1552, + "step": 4903 + }, + { + "epoch": 0.04904, + "grad_norm": 0.6809368406351045, + "learning_rate": 0.003, + "loss": 4.1193, + "step": 4904 + }, + { + "epoch": 0.04905, + "grad_norm": 0.8596440024174532, + "learning_rate": 0.003, + "loss": 4.1227, + "step": 4905 + }, + { + "epoch": 0.04906, + "grad_norm": 1.1779416676139556, + "learning_rate": 0.003, + "loss": 4.1426, + "step": 4906 + }, + { + "epoch": 0.04907, + "grad_norm": 0.9153726514069693, + "learning_rate": 0.003, + "loss": 4.1244, + "step": 4907 + }, + { + "epoch": 0.04908, + "grad_norm": 0.6705432105024506, + "learning_rate": 0.003, + "loss": 4.1326, + "step": 4908 + }, + { + "epoch": 0.04909, + "grad_norm": 0.7177688849716874, + "learning_rate": 0.003, + "loss": 4.1377, + "step": 4909 + }, + { + "epoch": 0.0491, + "grad_norm": 0.8659923553333108, + "learning_rate": 0.003, + "loss": 4.1347, + "step": 4910 + }, + { + "epoch": 0.04911, + "grad_norm": 1.0871210771334645, + "learning_rate": 0.003, + "loss": 4.1392, + "step": 4911 + }, + { + "epoch": 0.04912, + "grad_norm": 0.825170311776902, + "learning_rate": 0.003, + "loss": 4.1074, + "step": 4912 + }, + { + "epoch": 0.04913, + "grad_norm": 0.6669180449834547, + "learning_rate": 0.003, + "loss": 4.1304, + "step": 4913 + }, + { + "epoch": 0.04914, + "grad_norm": 0.6589662640213191, + "learning_rate": 0.003, + "loss": 4.1349, + "step": 4914 + }, + { + "epoch": 0.04915, + "grad_norm": 0.6453034668867473, + "learning_rate": 0.003, + "loss": 4.1394, + "step": 4915 + }, + { + "epoch": 0.04916, + "grad_norm": 0.6379061303258432, + "learning_rate": 0.003, + "loss": 4.1375, + "step": 4916 + }, + { + "epoch": 0.04917, + "grad_norm": 0.5924781741927655, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 4917 + }, + { + "epoch": 0.04918, + "grad_norm": 0.5392017954625788, + "learning_rate": 0.003, + "loss": 4.1294, + "step": 4918 + }, + { + "epoch": 0.04919, + "grad_norm": 0.4926263926107212, + "learning_rate": 0.003, + "loss": 4.1291, + "step": 4919 + }, + { + "epoch": 0.0492, + "grad_norm": 0.5657039568533931, + "learning_rate": 0.003, + "loss": 4.1282, + "step": 4920 + }, + { + "epoch": 0.04921, + "grad_norm": 0.6408113470012788, + "learning_rate": 0.003, + "loss": 4.1481, + "step": 4921 + }, + { + "epoch": 0.04922, + "grad_norm": 0.7276187339344597, + "learning_rate": 0.003, + "loss": 4.1242, + "step": 4922 + }, + { + "epoch": 0.04923, + "grad_norm": 0.7782906389614147, + "learning_rate": 0.003, + "loss": 4.1051, + "step": 4923 + }, + { + "epoch": 0.04924, + "grad_norm": 0.8311620891758528, + "learning_rate": 0.003, + "loss": 4.1585, + "step": 4924 + }, + { + "epoch": 0.04925, + "grad_norm": 0.9301177021878505, + "learning_rate": 0.003, + "loss": 4.1125, + "step": 4925 + }, + { + "epoch": 0.04926, + "grad_norm": 0.8645596717575721, + "learning_rate": 0.003, + "loss": 4.115, + "step": 4926 + }, + { + "epoch": 0.04927, + "grad_norm": 0.7213043835497918, + "learning_rate": 0.003, + "loss": 4.1488, + "step": 4927 + }, + { + "epoch": 0.04928, + "grad_norm": 0.6797037072765041, + "learning_rate": 0.003, + "loss": 4.1102, + "step": 4928 + }, + { + "epoch": 0.04929, + "grad_norm": 0.5833040699084097, + "learning_rate": 0.003, + "loss": 4.1237, + "step": 4929 + }, + { + "epoch": 0.0493, + "grad_norm": 0.7502027197799604, + "learning_rate": 0.003, + "loss": 4.1211, + "step": 4930 + }, + { + "epoch": 0.04931, + "grad_norm": 0.7729294183641608, + "learning_rate": 0.003, + "loss": 4.1088, + "step": 4931 + }, + { + "epoch": 0.04932, + "grad_norm": 0.7115365383624536, + "learning_rate": 0.003, + "loss": 4.1526, + "step": 4932 + }, + { + "epoch": 0.04933, + "grad_norm": 0.619935413571119, + "learning_rate": 0.003, + "loss": 4.127, + "step": 4933 + }, + { + "epoch": 0.04934, + "grad_norm": 0.6328194752766774, + "learning_rate": 0.003, + "loss": 4.156, + "step": 4934 + }, + { + "epoch": 0.04935, + "grad_norm": 0.666067060827277, + "learning_rate": 0.003, + "loss": 4.153, + "step": 4935 + }, + { + "epoch": 0.04936, + "grad_norm": 0.7455790570370421, + "learning_rate": 0.003, + "loss": 4.1274, + "step": 4936 + }, + { + "epoch": 0.04937, + "grad_norm": 0.8690467307626608, + "learning_rate": 0.003, + "loss": 4.1265, + "step": 4937 + }, + { + "epoch": 0.04938, + "grad_norm": 0.9382478740758116, + "learning_rate": 0.003, + "loss": 4.1384, + "step": 4938 + }, + { + "epoch": 0.04939, + "grad_norm": 0.9269163238429858, + "learning_rate": 0.003, + "loss": 4.15, + "step": 4939 + }, + { + "epoch": 0.0494, + "grad_norm": 0.7535488235404929, + "learning_rate": 0.003, + "loss": 4.1512, + "step": 4940 + }, + { + "epoch": 0.04941, + "grad_norm": 0.6231540125860376, + "learning_rate": 0.003, + "loss": 4.1145, + "step": 4941 + }, + { + "epoch": 0.04942, + "grad_norm": 0.5994421619470709, + "learning_rate": 0.003, + "loss": 4.1284, + "step": 4942 + }, + { + "epoch": 0.04943, + "grad_norm": 0.6316381391994007, + "learning_rate": 0.003, + "loss": 4.1507, + "step": 4943 + }, + { + "epoch": 0.04944, + "grad_norm": 0.6942848124244505, + "learning_rate": 0.003, + "loss": 4.1526, + "step": 4944 + }, + { + "epoch": 0.04945, + "grad_norm": 0.6745431924540877, + "learning_rate": 0.003, + "loss": 4.1442, + "step": 4945 + }, + { + "epoch": 0.04946, + "grad_norm": 0.7148234111999289, + "learning_rate": 0.003, + "loss": 4.1734, + "step": 4946 + }, + { + "epoch": 0.04947, + "grad_norm": 0.8151644855471658, + "learning_rate": 0.003, + "loss": 4.1207, + "step": 4947 + }, + { + "epoch": 0.04948, + "grad_norm": 0.901565456553654, + "learning_rate": 0.003, + "loss": 4.1404, + "step": 4948 + }, + { + "epoch": 0.04949, + "grad_norm": 0.8388445391988596, + "learning_rate": 0.003, + "loss": 4.1455, + "step": 4949 + }, + { + "epoch": 0.0495, + "grad_norm": 0.769498516722048, + "learning_rate": 0.003, + "loss": 4.1069, + "step": 4950 + }, + { + "epoch": 0.04951, + "grad_norm": 0.9271659871328424, + "learning_rate": 0.003, + "loss": 4.1193, + "step": 4951 + }, + { + "epoch": 0.04952, + "grad_norm": 0.9430760383533365, + "learning_rate": 0.003, + "loss": 4.1601, + "step": 4952 + }, + { + "epoch": 0.04953, + "grad_norm": 0.8839519050512392, + "learning_rate": 0.003, + "loss": 4.1458, + "step": 4953 + }, + { + "epoch": 0.04954, + "grad_norm": 0.8691782058986859, + "learning_rate": 0.003, + "loss": 4.1668, + "step": 4954 + }, + { + "epoch": 0.04955, + "grad_norm": 0.7804007935901659, + "learning_rate": 0.003, + "loss": 4.1378, + "step": 4955 + }, + { + "epoch": 0.04956, + "grad_norm": 0.7340937957959927, + "learning_rate": 0.003, + "loss": 4.1545, + "step": 4956 + }, + { + "epoch": 0.04957, + "grad_norm": 0.7502559806635023, + "learning_rate": 0.003, + "loss": 4.123, + "step": 4957 + }, + { + "epoch": 0.04958, + "grad_norm": 0.7469357331628064, + "learning_rate": 0.003, + "loss": 4.1412, + "step": 4958 + }, + { + "epoch": 0.04959, + "grad_norm": 0.7227020566857775, + "learning_rate": 0.003, + "loss": 4.1451, + "step": 4959 + }, + { + "epoch": 0.0496, + "grad_norm": 0.6072369078325202, + "learning_rate": 0.003, + "loss": 4.1095, + "step": 4960 + }, + { + "epoch": 0.04961, + "grad_norm": 0.676652458930749, + "learning_rate": 0.003, + "loss": 4.1185, + "step": 4961 + }, + { + "epoch": 0.04962, + "grad_norm": 0.7166853924943506, + "learning_rate": 0.003, + "loss": 4.1329, + "step": 4962 + }, + { + "epoch": 0.04963, + "grad_norm": 0.8282425384023169, + "learning_rate": 0.003, + "loss": 4.1738, + "step": 4963 + }, + { + "epoch": 0.04964, + "grad_norm": 0.9085445932185267, + "learning_rate": 0.003, + "loss": 4.1454, + "step": 4964 + }, + { + "epoch": 0.04965, + "grad_norm": 0.9373085689145864, + "learning_rate": 0.003, + "loss": 4.149, + "step": 4965 + }, + { + "epoch": 0.04966, + "grad_norm": 0.6879723306931469, + "learning_rate": 0.003, + "loss": 4.114, + "step": 4966 + }, + { + "epoch": 0.04967, + "grad_norm": 0.5346987676889187, + "learning_rate": 0.003, + "loss": 4.1357, + "step": 4967 + }, + { + "epoch": 0.04968, + "grad_norm": 0.5973814985358372, + "learning_rate": 0.003, + "loss": 4.1594, + "step": 4968 + }, + { + "epoch": 0.04969, + "grad_norm": 0.8137665544420138, + "learning_rate": 0.003, + "loss": 4.1371, + "step": 4969 + }, + { + "epoch": 0.0497, + "grad_norm": 0.9213595657452577, + "learning_rate": 0.003, + "loss": 4.1364, + "step": 4970 + }, + { + "epoch": 0.04971, + "grad_norm": 0.8081884217188025, + "learning_rate": 0.003, + "loss": 4.1548, + "step": 4971 + }, + { + "epoch": 0.04972, + "grad_norm": 0.696216159144708, + "learning_rate": 0.003, + "loss": 4.127, + "step": 4972 + }, + { + "epoch": 0.04973, + "grad_norm": 0.8733113357334519, + "learning_rate": 0.003, + "loss": 4.1797, + "step": 4973 + }, + { + "epoch": 0.04974, + "grad_norm": 1.0046232999118463, + "learning_rate": 0.003, + "loss": 4.1469, + "step": 4974 + }, + { + "epoch": 0.04975, + "grad_norm": 0.7903522031468557, + "learning_rate": 0.003, + "loss": 4.1385, + "step": 4975 + }, + { + "epoch": 0.04976, + "grad_norm": 0.6802567984502358, + "learning_rate": 0.003, + "loss": 4.1337, + "step": 4976 + }, + { + "epoch": 0.04977, + "grad_norm": 0.6880693634478566, + "learning_rate": 0.003, + "loss": 4.145, + "step": 4977 + }, + { + "epoch": 0.04978, + "grad_norm": 0.8224558060329078, + "learning_rate": 0.003, + "loss": 4.1443, + "step": 4978 + }, + { + "epoch": 0.04979, + "grad_norm": 0.693860233816605, + "learning_rate": 0.003, + "loss": 4.1505, + "step": 4979 + }, + { + "epoch": 0.0498, + "grad_norm": 0.5722513075301342, + "learning_rate": 0.003, + "loss": 4.1367, + "step": 4980 + }, + { + "epoch": 0.04981, + "grad_norm": 0.5155541843719716, + "learning_rate": 0.003, + "loss": 4.1092, + "step": 4981 + }, + { + "epoch": 0.04982, + "grad_norm": 0.5293279825762658, + "learning_rate": 0.003, + "loss": 4.1319, + "step": 4982 + }, + { + "epoch": 0.04983, + "grad_norm": 0.5031810438056484, + "learning_rate": 0.003, + "loss": 4.0871, + "step": 4983 + }, + { + "epoch": 0.04984, + "grad_norm": 0.50373719850947, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 4984 + }, + { + "epoch": 0.04985, + "grad_norm": 0.4503251156136766, + "learning_rate": 0.003, + "loss": 4.1217, + "step": 4985 + }, + { + "epoch": 0.04986, + "grad_norm": 0.45313125147575284, + "learning_rate": 0.003, + "loss": 4.1387, + "step": 4986 + }, + { + "epoch": 0.04987, + "grad_norm": 0.3893398619956913, + "learning_rate": 0.003, + "loss": 4.1002, + "step": 4987 + }, + { + "epoch": 0.04988, + "grad_norm": 0.38877981714111726, + "learning_rate": 0.003, + "loss": 4.133, + "step": 4988 + }, + { + "epoch": 0.04989, + "grad_norm": 0.4389205190435005, + "learning_rate": 0.003, + "loss": 4.1368, + "step": 4989 + }, + { + "epoch": 0.0499, + "grad_norm": 0.629712543859391, + "learning_rate": 0.003, + "loss": 4.1162, + "step": 4990 + }, + { + "epoch": 0.04991, + "grad_norm": 0.8411643038205673, + "learning_rate": 0.003, + "loss": 4.1109, + "step": 4991 + }, + { + "epoch": 0.04992, + "grad_norm": 1.0789241755274537, + "learning_rate": 0.003, + "loss": 4.1066, + "step": 4992 + }, + { + "epoch": 0.04993, + "grad_norm": 0.894691990453197, + "learning_rate": 0.003, + "loss": 4.1324, + "step": 4993 + }, + { + "epoch": 0.04994, + "grad_norm": 0.768409182457612, + "learning_rate": 0.003, + "loss": 4.1156, + "step": 4994 + }, + { + "epoch": 0.04995, + "grad_norm": 0.7502833338777625, + "learning_rate": 0.003, + "loss": 4.1388, + "step": 4995 + }, + { + "epoch": 0.04996, + "grad_norm": 0.7438036098652065, + "learning_rate": 0.003, + "loss": 4.1195, + "step": 4996 + }, + { + "epoch": 0.04997, + "grad_norm": 0.7542393015064909, + "learning_rate": 0.003, + "loss": 4.137, + "step": 4997 + }, + { + "epoch": 0.04998, + "grad_norm": 0.7987106834768938, + "learning_rate": 0.003, + "loss": 4.1614, + "step": 4998 + }, + { + "epoch": 0.04999, + "grad_norm": 0.7978624560845015, + "learning_rate": 0.003, + "loss": 4.1203, + "step": 4999 + }, + { + "epoch": 0.05, + "grad_norm": 0.9161273312112043, + "learning_rate": 0.003, + "loss": 4.1693, + "step": 5000 + }, + { + "epoch": 0.05001, + "grad_norm": 0.996412080112922, + "learning_rate": 0.003, + "loss": 4.1666, + "step": 5001 + }, + { + "epoch": 0.05002, + "grad_norm": 0.9597914662378922, + "learning_rate": 0.003, + "loss": 4.1413, + "step": 5002 + }, + { + "epoch": 0.05003, + "grad_norm": 0.8551476884140587, + "learning_rate": 0.003, + "loss": 4.1549, + "step": 5003 + }, + { + "epoch": 0.05004, + "grad_norm": 0.8613602446247762, + "learning_rate": 0.003, + "loss": 4.1388, + "step": 5004 + }, + { + "epoch": 0.05005, + "grad_norm": 0.9135522439159383, + "learning_rate": 0.003, + "loss": 4.1599, + "step": 5005 + }, + { + "epoch": 0.05006, + "grad_norm": 1.0256214526238998, + "learning_rate": 0.003, + "loss": 4.1448, + "step": 5006 + }, + { + "epoch": 0.05007, + "grad_norm": 0.9189841175452944, + "learning_rate": 0.003, + "loss": 4.1565, + "step": 5007 + }, + { + "epoch": 0.05008, + "grad_norm": 0.8069809344089516, + "learning_rate": 0.003, + "loss": 4.1222, + "step": 5008 + }, + { + "epoch": 0.05009, + "grad_norm": 0.7363173254372745, + "learning_rate": 0.003, + "loss": 4.1321, + "step": 5009 + }, + { + "epoch": 0.0501, + "grad_norm": 0.780930048247047, + "learning_rate": 0.003, + "loss": 4.1444, + "step": 5010 + }, + { + "epoch": 0.05011, + "grad_norm": 0.8533571714235313, + "learning_rate": 0.003, + "loss": 4.151, + "step": 5011 + }, + { + "epoch": 0.05012, + "grad_norm": 0.7333946370690233, + "learning_rate": 0.003, + "loss": 4.1338, + "step": 5012 + }, + { + "epoch": 0.05013, + "grad_norm": 0.7554255781670896, + "learning_rate": 0.003, + "loss": 4.1335, + "step": 5013 + }, + { + "epoch": 0.05014, + "grad_norm": 0.8053934909287488, + "learning_rate": 0.003, + "loss": 4.1128, + "step": 5014 + }, + { + "epoch": 0.05015, + "grad_norm": 0.6989364159311318, + "learning_rate": 0.003, + "loss": 4.1504, + "step": 5015 + }, + { + "epoch": 0.05016, + "grad_norm": 0.6523590967522017, + "learning_rate": 0.003, + "loss": 4.1287, + "step": 5016 + }, + { + "epoch": 0.05017, + "grad_norm": 0.6801442129276165, + "learning_rate": 0.003, + "loss": 4.1461, + "step": 5017 + }, + { + "epoch": 0.05018, + "grad_norm": 0.7921166720657037, + "learning_rate": 0.003, + "loss": 4.1423, + "step": 5018 + }, + { + "epoch": 0.05019, + "grad_norm": 0.9072386747671375, + "learning_rate": 0.003, + "loss": 4.1649, + "step": 5019 + }, + { + "epoch": 0.0502, + "grad_norm": 1.01701838314669, + "learning_rate": 0.003, + "loss": 4.1346, + "step": 5020 + }, + { + "epoch": 0.05021, + "grad_norm": 0.9399609124786235, + "learning_rate": 0.003, + "loss": 4.1404, + "step": 5021 + }, + { + "epoch": 0.05022, + "grad_norm": 0.9060890151344648, + "learning_rate": 0.003, + "loss": 4.1401, + "step": 5022 + }, + { + "epoch": 0.05023, + "grad_norm": 0.833401199404608, + "learning_rate": 0.003, + "loss": 4.1634, + "step": 5023 + }, + { + "epoch": 0.05024, + "grad_norm": 0.6936980469006718, + "learning_rate": 0.003, + "loss": 4.1185, + "step": 5024 + }, + { + "epoch": 0.05025, + "grad_norm": 0.6485227724397681, + "learning_rate": 0.003, + "loss": 4.1543, + "step": 5025 + }, + { + "epoch": 0.05026, + "grad_norm": 0.6504679913105271, + "learning_rate": 0.003, + "loss": 4.1308, + "step": 5026 + }, + { + "epoch": 0.05027, + "grad_norm": 0.6627173144511802, + "learning_rate": 0.003, + "loss": 4.1474, + "step": 5027 + }, + { + "epoch": 0.05028, + "grad_norm": 0.6951447785935732, + "learning_rate": 0.003, + "loss": 4.1337, + "step": 5028 + }, + { + "epoch": 0.05029, + "grad_norm": 0.6952175839048826, + "learning_rate": 0.003, + "loss": 4.1472, + "step": 5029 + }, + { + "epoch": 0.0503, + "grad_norm": 0.5688496607028097, + "learning_rate": 0.003, + "loss": 4.112, + "step": 5030 + }, + { + "epoch": 0.05031, + "grad_norm": 0.6035161846572389, + "learning_rate": 0.003, + "loss": 4.1318, + "step": 5031 + }, + { + "epoch": 0.05032, + "grad_norm": 0.670286035574246, + "learning_rate": 0.003, + "loss": 4.135, + "step": 5032 + }, + { + "epoch": 0.05033, + "grad_norm": 0.6467306224656293, + "learning_rate": 0.003, + "loss": 4.1099, + "step": 5033 + }, + { + "epoch": 0.05034, + "grad_norm": 0.6063260055449672, + "learning_rate": 0.003, + "loss": 4.1125, + "step": 5034 + }, + { + "epoch": 0.05035, + "grad_norm": 0.629592938048323, + "learning_rate": 0.003, + "loss": 4.1323, + "step": 5035 + }, + { + "epoch": 0.05036, + "grad_norm": 0.6478567980439086, + "learning_rate": 0.003, + "loss": 4.1177, + "step": 5036 + }, + { + "epoch": 0.05037, + "grad_norm": 0.7505964391613944, + "learning_rate": 0.003, + "loss": 4.1165, + "step": 5037 + }, + { + "epoch": 0.05038, + "grad_norm": 0.8528209687755143, + "learning_rate": 0.003, + "loss": 4.1296, + "step": 5038 + }, + { + "epoch": 0.05039, + "grad_norm": 1.0629743487440606, + "learning_rate": 0.003, + "loss": 4.1472, + "step": 5039 + }, + { + "epoch": 0.0504, + "grad_norm": 1.1351742762146553, + "learning_rate": 0.003, + "loss": 4.1529, + "step": 5040 + }, + { + "epoch": 0.05041, + "grad_norm": 0.7653122098370709, + "learning_rate": 0.003, + "loss": 4.1447, + "step": 5041 + }, + { + "epoch": 0.05042, + "grad_norm": 0.7775733262103008, + "learning_rate": 0.003, + "loss": 4.1377, + "step": 5042 + }, + { + "epoch": 0.05043, + "grad_norm": 0.7251140829300723, + "learning_rate": 0.003, + "loss": 4.1332, + "step": 5043 + }, + { + "epoch": 0.05044, + "grad_norm": 0.7675590430218364, + "learning_rate": 0.003, + "loss": 4.1501, + "step": 5044 + }, + { + "epoch": 0.05045, + "grad_norm": 0.796129916502261, + "learning_rate": 0.003, + "loss": 4.1675, + "step": 5045 + }, + { + "epoch": 0.05046, + "grad_norm": 0.7730541689349488, + "learning_rate": 0.003, + "loss": 4.1316, + "step": 5046 + }, + { + "epoch": 0.05047, + "grad_norm": 0.721813094114717, + "learning_rate": 0.003, + "loss": 4.1401, + "step": 5047 + }, + { + "epoch": 0.05048, + "grad_norm": 0.6421472913380964, + "learning_rate": 0.003, + "loss": 4.1331, + "step": 5048 + }, + { + "epoch": 0.05049, + "grad_norm": 0.6622380588723976, + "learning_rate": 0.003, + "loss": 4.0972, + "step": 5049 + }, + { + "epoch": 0.0505, + "grad_norm": 0.5608629241012306, + "learning_rate": 0.003, + "loss": 4.183, + "step": 5050 + }, + { + "epoch": 0.05051, + "grad_norm": 0.5745585056656611, + "learning_rate": 0.003, + "loss": 4.1443, + "step": 5051 + }, + { + "epoch": 0.05052, + "grad_norm": 0.6379797741525732, + "learning_rate": 0.003, + "loss": 4.1186, + "step": 5052 + }, + { + "epoch": 0.05053, + "grad_norm": 0.6833197049521939, + "learning_rate": 0.003, + "loss": 4.0926, + "step": 5053 + }, + { + "epoch": 0.05054, + "grad_norm": 0.7634490766042622, + "learning_rate": 0.003, + "loss": 4.1497, + "step": 5054 + }, + { + "epoch": 0.05055, + "grad_norm": 0.8837284078408866, + "learning_rate": 0.003, + "loss": 4.1298, + "step": 5055 + }, + { + "epoch": 0.05056, + "grad_norm": 0.986372668426662, + "learning_rate": 0.003, + "loss": 4.1493, + "step": 5056 + }, + { + "epoch": 0.05057, + "grad_norm": 1.1002958520848571, + "learning_rate": 0.003, + "loss": 4.1615, + "step": 5057 + }, + { + "epoch": 0.05058, + "grad_norm": 0.8620407065983835, + "learning_rate": 0.003, + "loss": 4.1539, + "step": 5058 + }, + { + "epoch": 0.05059, + "grad_norm": 0.7776023795769754, + "learning_rate": 0.003, + "loss": 4.12, + "step": 5059 + }, + { + "epoch": 0.0506, + "grad_norm": 0.8000191647790404, + "learning_rate": 0.003, + "loss": 4.155, + "step": 5060 + }, + { + "epoch": 0.05061, + "grad_norm": 0.8859446865423543, + "learning_rate": 0.003, + "loss": 4.135, + "step": 5061 + }, + { + "epoch": 0.05062, + "grad_norm": 0.8482279418144665, + "learning_rate": 0.003, + "loss": 4.1209, + "step": 5062 + }, + { + "epoch": 0.05063, + "grad_norm": 0.8922032578947977, + "learning_rate": 0.003, + "loss": 4.1079, + "step": 5063 + }, + { + "epoch": 0.05064, + "grad_norm": 0.8411761082071082, + "learning_rate": 0.003, + "loss": 4.1479, + "step": 5064 + }, + { + "epoch": 0.05065, + "grad_norm": 0.8548177713084265, + "learning_rate": 0.003, + "loss": 4.136, + "step": 5065 + }, + { + "epoch": 0.05066, + "grad_norm": 0.8283313824637577, + "learning_rate": 0.003, + "loss": 4.1234, + "step": 5066 + }, + { + "epoch": 0.05067, + "grad_norm": 0.8012249471762791, + "learning_rate": 0.003, + "loss": 4.1327, + "step": 5067 + }, + { + "epoch": 0.05068, + "grad_norm": 0.7695096288578107, + "learning_rate": 0.003, + "loss": 4.1533, + "step": 5068 + }, + { + "epoch": 0.05069, + "grad_norm": 0.9045793357271396, + "learning_rate": 0.003, + "loss": 4.1473, + "step": 5069 + }, + { + "epoch": 0.0507, + "grad_norm": 0.864648963362809, + "learning_rate": 0.003, + "loss": 4.1598, + "step": 5070 + }, + { + "epoch": 0.05071, + "grad_norm": 0.8168223538297239, + "learning_rate": 0.003, + "loss": 4.1468, + "step": 5071 + }, + { + "epoch": 0.05072, + "grad_norm": 0.7496813210802163, + "learning_rate": 0.003, + "loss": 4.1213, + "step": 5072 + }, + { + "epoch": 0.05073, + "grad_norm": 0.758480840663007, + "learning_rate": 0.003, + "loss": 4.123, + "step": 5073 + }, + { + "epoch": 0.05074, + "grad_norm": 0.696634586083982, + "learning_rate": 0.003, + "loss": 4.1158, + "step": 5074 + }, + { + "epoch": 0.05075, + "grad_norm": 0.7650541290537565, + "learning_rate": 0.003, + "loss": 4.1406, + "step": 5075 + }, + { + "epoch": 0.05076, + "grad_norm": 0.8193433029203343, + "learning_rate": 0.003, + "loss": 4.1544, + "step": 5076 + }, + { + "epoch": 0.05077, + "grad_norm": 0.8494467870191629, + "learning_rate": 0.003, + "loss": 4.1587, + "step": 5077 + }, + { + "epoch": 0.05078, + "grad_norm": 0.746324730026238, + "learning_rate": 0.003, + "loss": 4.1508, + "step": 5078 + }, + { + "epoch": 0.05079, + "grad_norm": 0.7788241307468048, + "learning_rate": 0.003, + "loss": 4.1233, + "step": 5079 + }, + { + "epoch": 0.0508, + "grad_norm": 0.6225546418109031, + "learning_rate": 0.003, + "loss": 4.1123, + "step": 5080 + }, + { + "epoch": 0.05081, + "grad_norm": 0.5890146987434735, + "learning_rate": 0.003, + "loss": 4.1354, + "step": 5081 + }, + { + "epoch": 0.05082, + "grad_norm": 0.7095924514939523, + "learning_rate": 0.003, + "loss": 4.145, + "step": 5082 + }, + { + "epoch": 0.05083, + "grad_norm": 0.8654893728046584, + "learning_rate": 0.003, + "loss": 4.1128, + "step": 5083 + }, + { + "epoch": 0.05084, + "grad_norm": 0.9274741044534023, + "learning_rate": 0.003, + "loss": 4.1349, + "step": 5084 + }, + { + "epoch": 0.05085, + "grad_norm": 0.863892065363985, + "learning_rate": 0.003, + "loss": 4.1004, + "step": 5085 + }, + { + "epoch": 0.05086, + "grad_norm": 0.7754486240701118, + "learning_rate": 0.003, + "loss": 4.1074, + "step": 5086 + }, + { + "epoch": 0.05087, + "grad_norm": 0.6467115070840203, + "learning_rate": 0.003, + "loss": 4.1273, + "step": 5087 + }, + { + "epoch": 0.05088, + "grad_norm": 0.6210040639615062, + "learning_rate": 0.003, + "loss": 4.1254, + "step": 5088 + }, + { + "epoch": 0.05089, + "grad_norm": 0.6024673934619533, + "learning_rate": 0.003, + "loss": 4.1198, + "step": 5089 + }, + { + "epoch": 0.0509, + "grad_norm": 0.6180444267761035, + "learning_rate": 0.003, + "loss": 4.156, + "step": 5090 + }, + { + "epoch": 0.05091, + "grad_norm": 0.6761606632023955, + "learning_rate": 0.003, + "loss": 4.1193, + "step": 5091 + }, + { + "epoch": 0.05092, + "grad_norm": 0.7124865882076785, + "learning_rate": 0.003, + "loss": 4.1217, + "step": 5092 + }, + { + "epoch": 0.05093, + "grad_norm": 0.6974866206345641, + "learning_rate": 0.003, + "loss": 4.1352, + "step": 5093 + }, + { + "epoch": 0.05094, + "grad_norm": 0.6044579073369937, + "learning_rate": 0.003, + "loss": 4.0943, + "step": 5094 + }, + { + "epoch": 0.05095, + "grad_norm": 0.6081394607788756, + "learning_rate": 0.003, + "loss": 4.1312, + "step": 5095 + }, + { + "epoch": 0.05096, + "grad_norm": 0.529589673235904, + "learning_rate": 0.003, + "loss": 4.1165, + "step": 5096 + }, + { + "epoch": 0.05097, + "grad_norm": 0.5564475574602485, + "learning_rate": 0.003, + "loss": 4.1129, + "step": 5097 + }, + { + "epoch": 0.05098, + "grad_norm": 0.6408190676261977, + "learning_rate": 0.003, + "loss": 4.1069, + "step": 5098 + }, + { + "epoch": 0.05099, + "grad_norm": 0.6723429647324713, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 5099 + }, + { + "epoch": 0.051, + "grad_norm": 0.8300243600705887, + "learning_rate": 0.003, + "loss": 4.1109, + "step": 5100 + }, + { + "epoch": 0.05101, + "grad_norm": 1.1968215120059336, + "learning_rate": 0.003, + "loss": 4.1589, + "step": 5101 + }, + { + "epoch": 0.05102, + "grad_norm": 0.9163396896342757, + "learning_rate": 0.003, + "loss": 4.1289, + "step": 5102 + }, + { + "epoch": 0.05103, + "grad_norm": 0.6393734271585462, + "learning_rate": 0.003, + "loss": 4.1356, + "step": 5103 + }, + { + "epoch": 0.05104, + "grad_norm": 0.6360132483972007, + "learning_rate": 0.003, + "loss": 4.1164, + "step": 5104 + }, + { + "epoch": 0.05105, + "grad_norm": 0.7716876792159864, + "learning_rate": 0.003, + "loss": 4.1071, + "step": 5105 + }, + { + "epoch": 0.05106, + "grad_norm": 0.8361810564307343, + "learning_rate": 0.003, + "loss": 4.088, + "step": 5106 + }, + { + "epoch": 0.05107, + "grad_norm": 0.8838910237762364, + "learning_rate": 0.003, + "loss": 4.1191, + "step": 5107 + }, + { + "epoch": 0.05108, + "grad_norm": 0.7935879173083066, + "learning_rate": 0.003, + "loss": 4.1669, + "step": 5108 + }, + { + "epoch": 0.05109, + "grad_norm": 0.8192023142285442, + "learning_rate": 0.003, + "loss": 4.1093, + "step": 5109 + }, + { + "epoch": 0.0511, + "grad_norm": 0.843755803433391, + "learning_rate": 0.003, + "loss": 4.1448, + "step": 5110 + }, + { + "epoch": 0.05111, + "grad_norm": 0.7470385931001929, + "learning_rate": 0.003, + "loss": 4.155, + "step": 5111 + }, + { + "epoch": 0.05112, + "grad_norm": 0.6896006523029916, + "learning_rate": 0.003, + "loss": 4.1399, + "step": 5112 + }, + { + "epoch": 0.05113, + "grad_norm": 0.7291351113373098, + "learning_rate": 0.003, + "loss": 4.1297, + "step": 5113 + }, + { + "epoch": 0.05114, + "grad_norm": 0.6627016065501992, + "learning_rate": 0.003, + "loss": 4.1057, + "step": 5114 + }, + { + "epoch": 0.05115, + "grad_norm": 0.6079619066030912, + "learning_rate": 0.003, + "loss": 4.1316, + "step": 5115 + }, + { + "epoch": 0.05116, + "grad_norm": 0.6260970910935404, + "learning_rate": 0.003, + "loss": 4.1393, + "step": 5116 + }, + { + "epoch": 0.05117, + "grad_norm": 0.6494564538670361, + "learning_rate": 0.003, + "loss": 4.1193, + "step": 5117 + }, + { + "epoch": 0.05118, + "grad_norm": 0.6314756962030549, + "learning_rate": 0.003, + "loss": 4.119, + "step": 5118 + }, + { + "epoch": 0.05119, + "grad_norm": 0.621412047113099, + "learning_rate": 0.003, + "loss": 4.101, + "step": 5119 + }, + { + "epoch": 0.0512, + "grad_norm": 0.6805129862001599, + "learning_rate": 0.003, + "loss": 4.0929, + "step": 5120 + }, + { + "epoch": 0.05121, + "grad_norm": 0.6678436168391524, + "learning_rate": 0.003, + "loss": 4.1117, + "step": 5121 + }, + { + "epoch": 0.05122, + "grad_norm": 0.6883310547366032, + "learning_rate": 0.003, + "loss": 4.1009, + "step": 5122 + }, + { + "epoch": 0.05123, + "grad_norm": 0.6801681468514705, + "learning_rate": 0.003, + "loss": 4.1157, + "step": 5123 + }, + { + "epoch": 0.05124, + "grad_norm": 0.6407786390586026, + "learning_rate": 0.003, + "loss": 4.154, + "step": 5124 + }, + { + "epoch": 0.05125, + "grad_norm": 0.616883516629288, + "learning_rate": 0.003, + "loss": 4.122, + "step": 5125 + }, + { + "epoch": 0.05126, + "grad_norm": 0.7827204877464654, + "learning_rate": 0.003, + "loss": 4.1224, + "step": 5126 + }, + { + "epoch": 0.05127, + "grad_norm": 0.8837164038749966, + "learning_rate": 0.003, + "loss": 4.1311, + "step": 5127 + }, + { + "epoch": 0.05128, + "grad_norm": 0.8294720258543833, + "learning_rate": 0.003, + "loss": 4.1466, + "step": 5128 + }, + { + "epoch": 0.05129, + "grad_norm": 0.8659985985101629, + "learning_rate": 0.003, + "loss": 4.1179, + "step": 5129 + }, + { + "epoch": 0.0513, + "grad_norm": 0.9120789657636011, + "learning_rate": 0.003, + "loss": 4.111, + "step": 5130 + }, + { + "epoch": 0.05131, + "grad_norm": 1.071986933837498, + "learning_rate": 0.003, + "loss": 4.1629, + "step": 5131 + }, + { + "epoch": 0.05132, + "grad_norm": 1.0458562389037795, + "learning_rate": 0.003, + "loss": 4.1506, + "step": 5132 + }, + { + "epoch": 0.05133, + "grad_norm": 0.9786133915152717, + "learning_rate": 0.003, + "loss": 4.1558, + "step": 5133 + }, + { + "epoch": 0.05134, + "grad_norm": 0.9697734611114633, + "learning_rate": 0.003, + "loss": 4.1714, + "step": 5134 + }, + { + "epoch": 0.05135, + "grad_norm": 0.9363895927818596, + "learning_rate": 0.003, + "loss": 4.1397, + "step": 5135 + }, + { + "epoch": 0.05136, + "grad_norm": 1.0170683776594798, + "learning_rate": 0.003, + "loss": 4.125, + "step": 5136 + }, + { + "epoch": 0.05137, + "grad_norm": 1.1145412465734708, + "learning_rate": 0.003, + "loss": 4.1715, + "step": 5137 + }, + { + "epoch": 0.05138, + "grad_norm": 0.9941607682684069, + "learning_rate": 0.003, + "loss": 4.1663, + "step": 5138 + }, + { + "epoch": 0.05139, + "grad_norm": 0.978207963075468, + "learning_rate": 0.003, + "loss": 4.1558, + "step": 5139 + }, + { + "epoch": 0.0514, + "grad_norm": 1.0158981617955298, + "learning_rate": 0.003, + "loss": 4.1585, + "step": 5140 + }, + { + "epoch": 0.05141, + "grad_norm": 0.8431378615881333, + "learning_rate": 0.003, + "loss": 4.2033, + "step": 5141 + }, + { + "epoch": 0.05142, + "grad_norm": 0.9258386606688979, + "learning_rate": 0.003, + "loss": 4.1465, + "step": 5142 + }, + { + "epoch": 0.05143, + "grad_norm": 0.8809664967212562, + "learning_rate": 0.003, + "loss": 4.1529, + "step": 5143 + }, + { + "epoch": 0.05144, + "grad_norm": 0.808685397208462, + "learning_rate": 0.003, + "loss": 4.136, + "step": 5144 + }, + { + "epoch": 0.05145, + "grad_norm": 0.7632108507977365, + "learning_rate": 0.003, + "loss": 4.1352, + "step": 5145 + }, + { + "epoch": 0.05146, + "grad_norm": 0.6541780509288703, + "learning_rate": 0.003, + "loss": 4.1198, + "step": 5146 + }, + { + "epoch": 0.05147, + "grad_norm": 0.5225402009574097, + "learning_rate": 0.003, + "loss": 4.1424, + "step": 5147 + }, + { + "epoch": 0.05148, + "grad_norm": 0.4857663422658965, + "learning_rate": 0.003, + "loss": 4.1408, + "step": 5148 + }, + { + "epoch": 0.05149, + "grad_norm": 0.5849142561064581, + "learning_rate": 0.003, + "loss": 4.1136, + "step": 5149 + }, + { + "epoch": 0.0515, + "grad_norm": 0.6123859001036767, + "learning_rate": 0.003, + "loss": 4.151, + "step": 5150 + }, + { + "epoch": 0.05151, + "grad_norm": 0.7026699600678006, + "learning_rate": 0.003, + "loss": 4.1142, + "step": 5151 + }, + { + "epoch": 0.05152, + "grad_norm": 0.6504209042889696, + "learning_rate": 0.003, + "loss": 4.1508, + "step": 5152 + }, + { + "epoch": 0.05153, + "grad_norm": 0.6252489466134844, + "learning_rate": 0.003, + "loss": 4.124, + "step": 5153 + }, + { + "epoch": 0.05154, + "grad_norm": 0.6371096455455558, + "learning_rate": 0.003, + "loss": 4.1347, + "step": 5154 + }, + { + "epoch": 0.05155, + "grad_norm": 0.5613653424439898, + "learning_rate": 0.003, + "loss": 4.1327, + "step": 5155 + }, + { + "epoch": 0.05156, + "grad_norm": 0.5387099035222069, + "learning_rate": 0.003, + "loss": 4.1562, + "step": 5156 + }, + { + "epoch": 0.05157, + "grad_norm": 0.5796148701609515, + "learning_rate": 0.003, + "loss": 4.1277, + "step": 5157 + }, + { + "epoch": 0.05158, + "grad_norm": 0.6280103262144628, + "learning_rate": 0.003, + "loss": 4.1224, + "step": 5158 + }, + { + "epoch": 0.05159, + "grad_norm": 0.713561476333827, + "learning_rate": 0.003, + "loss": 4.1053, + "step": 5159 + }, + { + "epoch": 0.0516, + "grad_norm": 0.7440717469043273, + "learning_rate": 0.003, + "loss": 4.1151, + "step": 5160 + }, + { + "epoch": 0.05161, + "grad_norm": 0.693854318455779, + "learning_rate": 0.003, + "loss": 4.0878, + "step": 5161 + }, + { + "epoch": 0.05162, + "grad_norm": 0.5713907895082873, + "learning_rate": 0.003, + "loss": 4.1132, + "step": 5162 + }, + { + "epoch": 0.05163, + "grad_norm": 0.6389112592034603, + "learning_rate": 0.003, + "loss": 4.1355, + "step": 5163 + }, + { + "epoch": 0.05164, + "grad_norm": 0.644621023044623, + "learning_rate": 0.003, + "loss": 4.1119, + "step": 5164 + }, + { + "epoch": 0.05165, + "grad_norm": 0.6418172439331881, + "learning_rate": 0.003, + "loss": 4.1297, + "step": 5165 + }, + { + "epoch": 0.05166, + "grad_norm": 0.6129087318474767, + "learning_rate": 0.003, + "loss": 4.1266, + "step": 5166 + }, + { + "epoch": 0.05167, + "grad_norm": 0.6032353120193177, + "learning_rate": 0.003, + "loss": 4.1341, + "step": 5167 + }, + { + "epoch": 0.05168, + "grad_norm": 0.5247219196682845, + "learning_rate": 0.003, + "loss": 4.1612, + "step": 5168 + }, + { + "epoch": 0.05169, + "grad_norm": 0.5175916837961249, + "learning_rate": 0.003, + "loss": 4.1113, + "step": 5169 + }, + { + "epoch": 0.0517, + "grad_norm": 0.6351381415951768, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 5170 + }, + { + "epoch": 0.05171, + "grad_norm": 0.9707906373263301, + "learning_rate": 0.003, + "loss": 4.121, + "step": 5171 + }, + { + "epoch": 0.05172, + "grad_norm": 1.3871012373265488, + "learning_rate": 0.003, + "loss": 4.1567, + "step": 5172 + }, + { + "epoch": 0.05173, + "grad_norm": 0.6859878951673775, + "learning_rate": 0.003, + "loss": 4.1484, + "step": 5173 + }, + { + "epoch": 0.05174, + "grad_norm": 0.688770612830069, + "learning_rate": 0.003, + "loss": 4.1416, + "step": 5174 + }, + { + "epoch": 0.05175, + "grad_norm": 0.7348674861956664, + "learning_rate": 0.003, + "loss": 4.151, + "step": 5175 + }, + { + "epoch": 0.05176, + "grad_norm": 0.7293852449611181, + "learning_rate": 0.003, + "loss": 4.1142, + "step": 5176 + }, + { + "epoch": 0.05177, + "grad_norm": 0.6564692961913474, + "learning_rate": 0.003, + "loss": 4.1059, + "step": 5177 + }, + { + "epoch": 0.05178, + "grad_norm": 0.6743163086709401, + "learning_rate": 0.003, + "loss": 4.104, + "step": 5178 + }, + { + "epoch": 0.05179, + "grad_norm": 0.7124422218567592, + "learning_rate": 0.003, + "loss": 4.1241, + "step": 5179 + }, + { + "epoch": 0.0518, + "grad_norm": 0.7927407152562181, + "learning_rate": 0.003, + "loss": 4.1251, + "step": 5180 + }, + { + "epoch": 0.05181, + "grad_norm": 0.7257533748695947, + "learning_rate": 0.003, + "loss": 4.1324, + "step": 5181 + }, + { + "epoch": 0.05182, + "grad_norm": 0.6857538651077083, + "learning_rate": 0.003, + "loss": 4.1461, + "step": 5182 + }, + { + "epoch": 0.05183, + "grad_norm": 0.6165169903841881, + "learning_rate": 0.003, + "loss": 4.1098, + "step": 5183 + }, + { + "epoch": 0.05184, + "grad_norm": 0.5646217873608306, + "learning_rate": 0.003, + "loss": 4.1087, + "step": 5184 + }, + { + "epoch": 0.05185, + "grad_norm": 0.6765417920434729, + "learning_rate": 0.003, + "loss": 4.1155, + "step": 5185 + }, + { + "epoch": 0.05186, + "grad_norm": 0.763739343016591, + "learning_rate": 0.003, + "loss": 4.1271, + "step": 5186 + }, + { + "epoch": 0.05187, + "grad_norm": 0.8133701961716747, + "learning_rate": 0.003, + "loss": 4.1673, + "step": 5187 + }, + { + "epoch": 0.05188, + "grad_norm": 0.9136528986209898, + "learning_rate": 0.003, + "loss": 4.1414, + "step": 5188 + }, + { + "epoch": 0.05189, + "grad_norm": 1.1309518830748815, + "learning_rate": 0.003, + "loss": 4.1235, + "step": 5189 + }, + { + "epoch": 0.0519, + "grad_norm": 1.1084320754378056, + "learning_rate": 0.003, + "loss": 4.1316, + "step": 5190 + }, + { + "epoch": 0.05191, + "grad_norm": 0.7822117289248245, + "learning_rate": 0.003, + "loss": 4.1232, + "step": 5191 + }, + { + "epoch": 0.05192, + "grad_norm": 0.6054235823299604, + "learning_rate": 0.003, + "loss": 4.1413, + "step": 5192 + }, + { + "epoch": 0.05193, + "grad_norm": 0.7272923940874335, + "learning_rate": 0.003, + "loss": 4.1392, + "step": 5193 + }, + { + "epoch": 0.05194, + "grad_norm": 1.053286129123416, + "learning_rate": 0.003, + "loss": 4.1104, + "step": 5194 + }, + { + "epoch": 0.05195, + "grad_norm": 1.0774389375766364, + "learning_rate": 0.003, + "loss": 4.1542, + "step": 5195 + }, + { + "epoch": 0.05196, + "grad_norm": 0.7935135429431, + "learning_rate": 0.003, + "loss": 4.1262, + "step": 5196 + }, + { + "epoch": 0.05197, + "grad_norm": 0.6765899030219741, + "learning_rate": 0.003, + "loss": 4.146, + "step": 5197 + }, + { + "epoch": 0.05198, + "grad_norm": 0.7288470154808702, + "learning_rate": 0.003, + "loss": 4.1197, + "step": 5198 + }, + { + "epoch": 0.05199, + "grad_norm": 0.7762676094179997, + "learning_rate": 0.003, + "loss": 4.1066, + "step": 5199 + }, + { + "epoch": 0.052, + "grad_norm": 0.8517147173377261, + "learning_rate": 0.003, + "loss": 4.1432, + "step": 5200 + }, + { + "epoch": 0.05201, + "grad_norm": 0.7812345967785519, + "learning_rate": 0.003, + "loss": 4.1391, + "step": 5201 + }, + { + "epoch": 0.05202, + "grad_norm": 0.7915734925208033, + "learning_rate": 0.003, + "loss": 4.1227, + "step": 5202 + }, + { + "epoch": 0.05203, + "grad_norm": 0.779004263274718, + "learning_rate": 0.003, + "loss": 4.1542, + "step": 5203 + }, + { + "epoch": 0.05204, + "grad_norm": 0.8440004396702558, + "learning_rate": 0.003, + "loss": 4.143, + "step": 5204 + }, + { + "epoch": 0.05205, + "grad_norm": 0.8044224640384809, + "learning_rate": 0.003, + "loss": 4.1498, + "step": 5205 + }, + { + "epoch": 0.05206, + "grad_norm": 0.830818456670086, + "learning_rate": 0.003, + "loss": 4.1336, + "step": 5206 + }, + { + "epoch": 0.05207, + "grad_norm": 0.840221891416657, + "learning_rate": 0.003, + "loss": 4.1185, + "step": 5207 + }, + { + "epoch": 0.05208, + "grad_norm": 0.8228927798697641, + "learning_rate": 0.003, + "loss": 4.1353, + "step": 5208 + }, + { + "epoch": 0.05209, + "grad_norm": 0.7538723463798628, + "learning_rate": 0.003, + "loss": 4.1278, + "step": 5209 + }, + { + "epoch": 0.0521, + "grad_norm": 0.7269914660347427, + "learning_rate": 0.003, + "loss": 4.0982, + "step": 5210 + }, + { + "epoch": 0.05211, + "grad_norm": 0.6396845788398585, + "learning_rate": 0.003, + "loss": 4.121, + "step": 5211 + }, + { + "epoch": 0.05212, + "grad_norm": 0.6578986704720922, + "learning_rate": 0.003, + "loss": 4.1682, + "step": 5212 + }, + { + "epoch": 0.05213, + "grad_norm": 0.6665917425071297, + "learning_rate": 0.003, + "loss": 4.124, + "step": 5213 + }, + { + "epoch": 0.05214, + "grad_norm": 0.6696745228858743, + "learning_rate": 0.003, + "loss": 4.1155, + "step": 5214 + }, + { + "epoch": 0.05215, + "grad_norm": 0.8360803674253388, + "learning_rate": 0.003, + "loss": 4.1145, + "step": 5215 + }, + { + "epoch": 0.05216, + "grad_norm": 1.0234029588407536, + "learning_rate": 0.003, + "loss": 4.1326, + "step": 5216 + }, + { + "epoch": 0.05217, + "grad_norm": 0.8990465132650213, + "learning_rate": 0.003, + "loss": 4.1318, + "step": 5217 + }, + { + "epoch": 0.05218, + "grad_norm": 0.7789964318595968, + "learning_rate": 0.003, + "loss": 4.1198, + "step": 5218 + }, + { + "epoch": 0.05219, + "grad_norm": 0.7298439479090568, + "learning_rate": 0.003, + "loss": 4.1481, + "step": 5219 + }, + { + "epoch": 0.0522, + "grad_norm": 0.8002872146937169, + "learning_rate": 0.003, + "loss": 4.115, + "step": 5220 + }, + { + "epoch": 0.05221, + "grad_norm": 0.8077845914697496, + "learning_rate": 0.003, + "loss": 4.1263, + "step": 5221 + }, + { + "epoch": 0.05222, + "grad_norm": 0.8937695724547757, + "learning_rate": 0.003, + "loss": 4.1376, + "step": 5222 + }, + { + "epoch": 0.05223, + "grad_norm": 0.8660989097279149, + "learning_rate": 0.003, + "loss": 4.1496, + "step": 5223 + }, + { + "epoch": 0.05224, + "grad_norm": 0.8551329970671594, + "learning_rate": 0.003, + "loss": 4.1288, + "step": 5224 + }, + { + "epoch": 0.05225, + "grad_norm": 0.7915686333783567, + "learning_rate": 0.003, + "loss": 4.1505, + "step": 5225 + }, + { + "epoch": 0.05226, + "grad_norm": 0.8295912864972166, + "learning_rate": 0.003, + "loss": 4.1414, + "step": 5226 + }, + { + "epoch": 0.05227, + "grad_norm": 0.7362810272638042, + "learning_rate": 0.003, + "loss": 4.1637, + "step": 5227 + }, + { + "epoch": 0.05228, + "grad_norm": 0.7001627229979629, + "learning_rate": 0.003, + "loss": 4.1432, + "step": 5228 + }, + { + "epoch": 0.05229, + "grad_norm": 0.7602202740566725, + "learning_rate": 0.003, + "loss": 4.1086, + "step": 5229 + }, + { + "epoch": 0.0523, + "grad_norm": 0.6912471723136686, + "learning_rate": 0.003, + "loss": 4.0986, + "step": 5230 + }, + { + "epoch": 0.05231, + "grad_norm": 0.7018975883307987, + "learning_rate": 0.003, + "loss": 4.1343, + "step": 5231 + }, + { + "epoch": 0.05232, + "grad_norm": 0.717854420616106, + "learning_rate": 0.003, + "loss": 4.1242, + "step": 5232 + }, + { + "epoch": 0.05233, + "grad_norm": 0.7712695717441904, + "learning_rate": 0.003, + "loss": 4.1086, + "step": 5233 + }, + { + "epoch": 0.05234, + "grad_norm": 0.8385680879325836, + "learning_rate": 0.003, + "loss": 4.1626, + "step": 5234 + }, + { + "epoch": 0.05235, + "grad_norm": 0.9750232594843035, + "learning_rate": 0.003, + "loss": 4.1435, + "step": 5235 + }, + { + "epoch": 0.05236, + "grad_norm": 0.8446359663530114, + "learning_rate": 0.003, + "loss": 4.1204, + "step": 5236 + }, + { + "epoch": 0.05237, + "grad_norm": 0.9096810630918026, + "learning_rate": 0.003, + "loss": 4.1267, + "step": 5237 + }, + { + "epoch": 0.05238, + "grad_norm": 0.9079420042262413, + "learning_rate": 0.003, + "loss": 4.1189, + "step": 5238 + }, + { + "epoch": 0.05239, + "grad_norm": 0.8141480866664513, + "learning_rate": 0.003, + "loss": 4.1388, + "step": 5239 + }, + { + "epoch": 0.0524, + "grad_norm": 0.7056408832787606, + "learning_rate": 0.003, + "loss": 4.1448, + "step": 5240 + }, + { + "epoch": 0.05241, + "grad_norm": 0.6654549467738141, + "learning_rate": 0.003, + "loss": 4.1475, + "step": 5241 + }, + { + "epoch": 0.05242, + "grad_norm": 0.7294744123105587, + "learning_rate": 0.003, + "loss": 4.1449, + "step": 5242 + }, + { + "epoch": 0.05243, + "grad_norm": 0.7203093785089144, + "learning_rate": 0.003, + "loss": 4.1366, + "step": 5243 + }, + { + "epoch": 0.05244, + "grad_norm": 0.6894002827090067, + "learning_rate": 0.003, + "loss": 4.1125, + "step": 5244 + }, + { + "epoch": 0.05245, + "grad_norm": 0.6674737328058699, + "learning_rate": 0.003, + "loss": 4.1193, + "step": 5245 + }, + { + "epoch": 0.05246, + "grad_norm": 0.6789552091924481, + "learning_rate": 0.003, + "loss": 4.1241, + "step": 5246 + }, + { + "epoch": 0.05247, + "grad_norm": 0.7027379397907318, + "learning_rate": 0.003, + "loss": 4.1241, + "step": 5247 + }, + { + "epoch": 0.05248, + "grad_norm": 0.7672915233226524, + "learning_rate": 0.003, + "loss": 4.1316, + "step": 5248 + }, + { + "epoch": 0.05249, + "grad_norm": 0.8600082125105806, + "learning_rate": 0.003, + "loss": 4.1487, + "step": 5249 + }, + { + "epoch": 0.0525, + "grad_norm": 0.8053204409517745, + "learning_rate": 0.003, + "loss": 4.1378, + "step": 5250 + }, + { + "epoch": 0.05251, + "grad_norm": 0.7017613490500446, + "learning_rate": 0.003, + "loss": 4.1246, + "step": 5251 + }, + { + "epoch": 0.05252, + "grad_norm": 0.6533051938670863, + "learning_rate": 0.003, + "loss": 4.1484, + "step": 5252 + }, + { + "epoch": 0.05253, + "grad_norm": 0.6002345144376094, + "learning_rate": 0.003, + "loss": 4.1137, + "step": 5253 + }, + { + "epoch": 0.05254, + "grad_norm": 0.5673372560372355, + "learning_rate": 0.003, + "loss": 4.0996, + "step": 5254 + }, + { + "epoch": 0.05255, + "grad_norm": 0.5743257458749801, + "learning_rate": 0.003, + "loss": 4.1015, + "step": 5255 + }, + { + "epoch": 0.05256, + "grad_norm": 0.5639251305320502, + "learning_rate": 0.003, + "loss": 4.1283, + "step": 5256 + }, + { + "epoch": 0.05257, + "grad_norm": 0.5046284771633743, + "learning_rate": 0.003, + "loss": 4.1166, + "step": 5257 + }, + { + "epoch": 0.05258, + "grad_norm": 0.474168541210226, + "learning_rate": 0.003, + "loss": 4.1167, + "step": 5258 + }, + { + "epoch": 0.05259, + "grad_norm": 0.47547400835554277, + "learning_rate": 0.003, + "loss": 4.1012, + "step": 5259 + }, + { + "epoch": 0.0526, + "grad_norm": 0.5164451968626312, + "learning_rate": 0.003, + "loss": 4.1303, + "step": 5260 + }, + { + "epoch": 0.05261, + "grad_norm": 0.5250991374606924, + "learning_rate": 0.003, + "loss": 4.0976, + "step": 5261 + }, + { + "epoch": 0.05262, + "grad_norm": 0.6851820393823259, + "learning_rate": 0.003, + "loss": 4.0927, + "step": 5262 + }, + { + "epoch": 0.05263, + "grad_norm": 0.9977081433022779, + "learning_rate": 0.003, + "loss": 4.1511, + "step": 5263 + }, + { + "epoch": 0.05264, + "grad_norm": 1.268054358654276, + "learning_rate": 0.003, + "loss": 4.159, + "step": 5264 + }, + { + "epoch": 0.05265, + "grad_norm": 0.5767306016131388, + "learning_rate": 0.003, + "loss": 4.1057, + "step": 5265 + }, + { + "epoch": 0.05266, + "grad_norm": 0.7952501578678153, + "learning_rate": 0.003, + "loss": 4.0961, + "step": 5266 + }, + { + "epoch": 0.05267, + "grad_norm": 1.0070088823226309, + "learning_rate": 0.003, + "loss": 4.152, + "step": 5267 + }, + { + "epoch": 0.05268, + "grad_norm": 0.9932744688310504, + "learning_rate": 0.003, + "loss": 4.1218, + "step": 5268 + }, + { + "epoch": 0.05269, + "grad_norm": 1.0388601409327192, + "learning_rate": 0.003, + "loss": 4.1029, + "step": 5269 + }, + { + "epoch": 0.0527, + "grad_norm": 0.956313275292693, + "learning_rate": 0.003, + "loss": 4.1371, + "step": 5270 + }, + { + "epoch": 0.05271, + "grad_norm": 1.0292710271964594, + "learning_rate": 0.003, + "loss": 4.1075, + "step": 5271 + }, + { + "epoch": 0.05272, + "grad_norm": 0.8930084447233932, + "learning_rate": 0.003, + "loss": 4.1636, + "step": 5272 + }, + { + "epoch": 0.05273, + "grad_norm": 0.9405712293541173, + "learning_rate": 0.003, + "loss": 4.1478, + "step": 5273 + }, + { + "epoch": 0.05274, + "grad_norm": 1.0695246435294683, + "learning_rate": 0.003, + "loss": 4.1665, + "step": 5274 + }, + { + "epoch": 0.05275, + "grad_norm": 0.9058378894957746, + "learning_rate": 0.003, + "loss": 4.135, + "step": 5275 + }, + { + "epoch": 0.05276, + "grad_norm": 0.8493287470524594, + "learning_rate": 0.003, + "loss": 4.1264, + "step": 5276 + }, + { + "epoch": 0.05277, + "grad_norm": 0.9420450553310487, + "learning_rate": 0.003, + "loss": 4.1198, + "step": 5277 + }, + { + "epoch": 0.05278, + "grad_norm": 1.1979114145894763, + "learning_rate": 0.003, + "loss": 4.1498, + "step": 5278 + }, + { + "epoch": 0.05279, + "grad_norm": 0.9851879183907458, + "learning_rate": 0.003, + "loss": 4.1443, + "step": 5279 + }, + { + "epoch": 0.0528, + "grad_norm": 0.9906333358429689, + "learning_rate": 0.003, + "loss": 4.1431, + "step": 5280 + }, + { + "epoch": 0.05281, + "grad_norm": 1.0018246671795275, + "learning_rate": 0.003, + "loss": 4.1561, + "step": 5281 + }, + { + "epoch": 0.05282, + "grad_norm": 0.9088563498086558, + "learning_rate": 0.003, + "loss": 4.1483, + "step": 5282 + }, + { + "epoch": 0.05283, + "grad_norm": 0.8255004872484286, + "learning_rate": 0.003, + "loss": 4.1314, + "step": 5283 + }, + { + "epoch": 0.05284, + "grad_norm": 0.7395222306618988, + "learning_rate": 0.003, + "loss": 4.1586, + "step": 5284 + }, + { + "epoch": 0.05285, + "grad_norm": 0.7519757069394422, + "learning_rate": 0.003, + "loss": 4.1084, + "step": 5285 + }, + { + "epoch": 0.05286, + "grad_norm": 0.8309929954214635, + "learning_rate": 0.003, + "loss": 4.1763, + "step": 5286 + }, + { + "epoch": 0.05287, + "grad_norm": 0.8541420130148721, + "learning_rate": 0.003, + "loss": 4.1453, + "step": 5287 + }, + { + "epoch": 0.05288, + "grad_norm": 0.9074648300179621, + "learning_rate": 0.003, + "loss": 4.1262, + "step": 5288 + }, + { + "epoch": 0.05289, + "grad_norm": 0.934169321459613, + "learning_rate": 0.003, + "loss": 4.1567, + "step": 5289 + }, + { + "epoch": 0.0529, + "grad_norm": 0.8614434047866291, + "learning_rate": 0.003, + "loss": 4.1499, + "step": 5290 + }, + { + "epoch": 0.05291, + "grad_norm": 0.7851323632915874, + "learning_rate": 0.003, + "loss": 4.1592, + "step": 5291 + }, + { + "epoch": 0.05292, + "grad_norm": 0.8846666231031659, + "learning_rate": 0.003, + "loss": 4.1535, + "step": 5292 + }, + { + "epoch": 0.05293, + "grad_norm": 0.8272989171480324, + "learning_rate": 0.003, + "loss": 4.151, + "step": 5293 + }, + { + "epoch": 0.05294, + "grad_norm": 0.7925051182542472, + "learning_rate": 0.003, + "loss": 4.1348, + "step": 5294 + }, + { + "epoch": 0.05295, + "grad_norm": 0.8458687864133015, + "learning_rate": 0.003, + "loss": 4.1485, + "step": 5295 + }, + { + "epoch": 0.05296, + "grad_norm": 0.7329229801657224, + "learning_rate": 0.003, + "loss": 4.1264, + "step": 5296 + }, + { + "epoch": 0.05297, + "grad_norm": 0.590148048628958, + "learning_rate": 0.003, + "loss": 4.1241, + "step": 5297 + }, + { + "epoch": 0.05298, + "grad_norm": 0.5633959457343373, + "learning_rate": 0.003, + "loss": 4.1403, + "step": 5298 + }, + { + "epoch": 0.05299, + "grad_norm": 0.6047041287435936, + "learning_rate": 0.003, + "loss": 4.1167, + "step": 5299 + }, + { + "epoch": 0.053, + "grad_norm": 0.6026083589407333, + "learning_rate": 0.003, + "loss": 4.1334, + "step": 5300 + }, + { + "epoch": 0.05301, + "grad_norm": 0.6614621931672765, + "learning_rate": 0.003, + "loss": 4.1486, + "step": 5301 + }, + { + "epoch": 0.05302, + "grad_norm": 0.667460840133991, + "learning_rate": 0.003, + "loss": 4.1218, + "step": 5302 + }, + { + "epoch": 0.05303, + "grad_norm": 0.7178043539507469, + "learning_rate": 0.003, + "loss": 4.1337, + "step": 5303 + }, + { + "epoch": 0.05304, + "grad_norm": 0.7057622216619723, + "learning_rate": 0.003, + "loss": 4.1228, + "step": 5304 + }, + { + "epoch": 0.05305, + "grad_norm": 0.6395996129884405, + "learning_rate": 0.003, + "loss": 4.1282, + "step": 5305 + }, + { + "epoch": 0.05306, + "grad_norm": 0.6595661147403964, + "learning_rate": 0.003, + "loss": 4.1362, + "step": 5306 + }, + { + "epoch": 0.05307, + "grad_norm": 0.5697470297170749, + "learning_rate": 0.003, + "loss": 4.1297, + "step": 5307 + }, + { + "epoch": 0.05308, + "grad_norm": 0.6269529857264409, + "learning_rate": 0.003, + "loss": 4.1412, + "step": 5308 + }, + { + "epoch": 0.05309, + "grad_norm": 0.7109602470701628, + "learning_rate": 0.003, + "loss": 4.1076, + "step": 5309 + }, + { + "epoch": 0.0531, + "grad_norm": 0.8779710810213476, + "learning_rate": 0.003, + "loss": 4.1214, + "step": 5310 + }, + { + "epoch": 0.05311, + "grad_norm": 0.9865994447177229, + "learning_rate": 0.003, + "loss": 4.161, + "step": 5311 + }, + { + "epoch": 0.05312, + "grad_norm": 0.928167243969855, + "learning_rate": 0.003, + "loss": 4.1182, + "step": 5312 + }, + { + "epoch": 0.05313, + "grad_norm": 0.65760273068672, + "learning_rate": 0.003, + "loss": 4.0955, + "step": 5313 + }, + { + "epoch": 0.05314, + "grad_norm": 0.6095510092870687, + "learning_rate": 0.003, + "loss": 4.1245, + "step": 5314 + }, + { + "epoch": 0.05315, + "grad_norm": 0.7424516970237663, + "learning_rate": 0.003, + "loss": 4.1331, + "step": 5315 + }, + { + "epoch": 0.05316, + "grad_norm": 0.8440091626661397, + "learning_rate": 0.003, + "loss": 4.1369, + "step": 5316 + }, + { + "epoch": 0.05317, + "grad_norm": 0.9378368997983185, + "learning_rate": 0.003, + "loss": 4.1418, + "step": 5317 + }, + { + "epoch": 0.05318, + "grad_norm": 0.8048248745024449, + "learning_rate": 0.003, + "loss": 4.108, + "step": 5318 + }, + { + "epoch": 0.05319, + "grad_norm": 0.7236000931693785, + "learning_rate": 0.003, + "loss": 4.1133, + "step": 5319 + }, + { + "epoch": 0.0532, + "grad_norm": 0.6461031092410674, + "learning_rate": 0.003, + "loss": 4.1188, + "step": 5320 + }, + { + "epoch": 0.05321, + "grad_norm": 0.7679854561007805, + "learning_rate": 0.003, + "loss": 4.1233, + "step": 5321 + }, + { + "epoch": 0.05322, + "grad_norm": 0.7986946965085385, + "learning_rate": 0.003, + "loss": 4.1448, + "step": 5322 + }, + { + "epoch": 0.05323, + "grad_norm": 0.7153375585884025, + "learning_rate": 0.003, + "loss": 4.1679, + "step": 5323 + }, + { + "epoch": 0.05324, + "grad_norm": 0.749330813072559, + "learning_rate": 0.003, + "loss": 4.1102, + "step": 5324 + }, + { + "epoch": 0.05325, + "grad_norm": 0.832758253039198, + "learning_rate": 0.003, + "loss": 4.1332, + "step": 5325 + }, + { + "epoch": 0.05326, + "grad_norm": 0.7759697554212088, + "learning_rate": 0.003, + "loss": 4.1525, + "step": 5326 + }, + { + "epoch": 0.05327, + "grad_norm": 0.7232266024198855, + "learning_rate": 0.003, + "loss": 4.1297, + "step": 5327 + }, + { + "epoch": 0.05328, + "grad_norm": 0.7836949776849659, + "learning_rate": 0.003, + "loss": 4.1406, + "step": 5328 + }, + { + "epoch": 0.05329, + "grad_norm": 0.7124265190034871, + "learning_rate": 0.003, + "loss": 4.1226, + "step": 5329 + }, + { + "epoch": 0.0533, + "grad_norm": 0.7073201953779398, + "learning_rate": 0.003, + "loss": 4.1422, + "step": 5330 + }, + { + "epoch": 0.05331, + "grad_norm": 0.6517329606524963, + "learning_rate": 0.003, + "loss": 4.1286, + "step": 5331 + }, + { + "epoch": 0.05332, + "grad_norm": 0.7675020790673767, + "learning_rate": 0.003, + "loss": 4.1486, + "step": 5332 + }, + { + "epoch": 0.05333, + "grad_norm": 0.8204477024627048, + "learning_rate": 0.003, + "loss": 4.1252, + "step": 5333 + }, + { + "epoch": 0.05334, + "grad_norm": 0.7022074746197576, + "learning_rate": 0.003, + "loss": 4.108, + "step": 5334 + }, + { + "epoch": 0.05335, + "grad_norm": 0.6275459710277902, + "learning_rate": 0.003, + "loss": 4.1391, + "step": 5335 + }, + { + "epoch": 0.05336, + "grad_norm": 0.6160252429147814, + "learning_rate": 0.003, + "loss": 4.147, + "step": 5336 + }, + { + "epoch": 0.05337, + "grad_norm": 0.6645784185963264, + "learning_rate": 0.003, + "loss": 4.1336, + "step": 5337 + }, + { + "epoch": 0.05338, + "grad_norm": 0.755147688591901, + "learning_rate": 0.003, + "loss": 4.1201, + "step": 5338 + }, + { + "epoch": 0.05339, + "grad_norm": 0.8974685697024821, + "learning_rate": 0.003, + "loss": 4.1117, + "step": 5339 + }, + { + "epoch": 0.0534, + "grad_norm": 1.0247940652403054, + "learning_rate": 0.003, + "loss": 4.1358, + "step": 5340 + }, + { + "epoch": 0.05341, + "grad_norm": 0.8398376000828047, + "learning_rate": 0.003, + "loss": 4.136, + "step": 5341 + }, + { + "epoch": 0.05342, + "grad_norm": 0.6251041860327448, + "learning_rate": 0.003, + "loss": 4.1176, + "step": 5342 + }, + { + "epoch": 0.05343, + "grad_norm": 0.6653717662444403, + "learning_rate": 0.003, + "loss": 4.1344, + "step": 5343 + }, + { + "epoch": 0.05344, + "grad_norm": 0.7626824156057184, + "learning_rate": 0.003, + "loss": 4.0965, + "step": 5344 + }, + { + "epoch": 0.05345, + "grad_norm": 0.7412027124944542, + "learning_rate": 0.003, + "loss": 4.1092, + "step": 5345 + }, + { + "epoch": 0.05346, + "grad_norm": 0.8204551363999621, + "learning_rate": 0.003, + "loss": 4.1275, + "step": 5346 + }, + { + "epoch": 0.05347, + "grad_norm": 0.8691311780811011, + "learning_rate": 0.003, + "loss": 4.1195, + "step": 5347 + }, + { + "epoch": 0.05348, + "grad_norm": 0.8297575781225915, + "learning_rate": 0.003, + "loss": 4.1173, + "step": 5348 + }, + { + "epoch": 0.05349, + "grad_norm": 0.7800308176994237, + "learning_rate": 0.003, + "loss": 4.1248, + "step": 5349 + }, + { + "epoch": 0.0535, + "grad_norm": 0.7428195062452232, + "learning_rate": 0.003, + "loss": 4.1134, + "step": 5350 + }, + { + "epoch": 0.05351, + "grad_norm": 0.698577766705363, + "learning_rate": 0.003, + "loss": 4.1171, + "step": 5351 + }, + { + "epoch": 0.05352, + "grad_norm": 0.7463247671330128, + "learning_rate": 0.003, + "loss": 4.114, + "step": 5352 + }, + { + "epoch": 0.05353, + "grad_norm": 0.8427380940568103, + "learning_rate": 0.003, + "loss": 4.1292, + "step": 5353 + }, + { + "epoch": 0.05354, + "grad_norm": 0.8580345957059121, + "learning_rate": 0.003, + "loss": 4.1189, + "step": 5354 + }, + { + "epoch": 0.05355, + "grad_norm": 0.7981948995131939, + "learning_rate": 0.003, + "loss": 4.1077, + "step": 5355 + }, + { + "epoch": 0.05356, + "grad_norm": 0.9987056987093543, + "learning_rate": 0.003, + "loss": 4.1536, + "step": 5356 + }, + { + "epoch": 0.05357, + "grad_norm": 1.0863630771998207, + "learning_rate": 0.003, + "loss": 4.1449, + "step": 5357 + }, + { + "epoch": 0.05358, + "grad_norm": 0.8873987298102782, + "learning_rate": 0.003, + "loss": 4.1316, + "step": 5358 + }, + { + "epoch": 0.05359, + "grad_norm": 0.8982951690299781, + "learning_rate": 0.003, + "loss": 4.1411, + "step": 5359 + }, + { + "epoch": 0.0536, + "grad_norm": 0.7420851849216799, + "learning_rate": 0.003, + "loss": 4.1329, + "step": 5360 + }, + { + "epoch": 0.05361, + "grad_norm": 0.6478690518338635, + "learning_rate": 0.003, + "loss": 4.1369, + "step": 5361 + }, + { + "epoch": 0.05362, + "grad_norm": 0.7363423400186474, + "learning_rate": 0.003, + "loss": 4.1501, + "step": 5362 + }, + { + "epoch": 0.05363, + "grad_norm": 0.8098849889259154, + "learning_rate": 0.003, + "loss": 4.1558, + "step": 5363 + }, + { + "epoch": 0.05364, + "grad_norm": 0.8901440262592046, + "learning_rate": 0.003, + "loss": 4.152, + "step": 5364 + }, + { + "epoch": 0.05365, + "grad_norm": 1.0380282918537116, + "learning_rate": 0.003, + "loss": 4.1511, + "step": 5365 + }, + { + "epoch": 0.05366, + "grad_norm": 1.0788058766201782, + "learning_rate": 0.003, + "loss": 4.1375, + "step": 5366 + }, + { + "epoch": 0.05367, + "grad_norm": 0.7608583479176579, + "learning_rate": 0.003, + "loss": 4.1254, + "step": 5367 + }, + { + "epoch": 0.05368, + "grad_norm": 0.5600149243746851, + "learning_rate": 0.003, + "loss": 4.1146, + "step": 5368 + }, + { + "epoch": 0.05369, + "grad_norm": 0.5618528481117717, + "learning_rate": 0.003, + "loss": 4.1273, + "step": 5369 + }, + { + "epoch": 0.0537, + "grad_norm": 0.6773018479702052, + "learning_rate": 0.003, + "loss": 4.1287, + "step": 5370 + }, + { + "epoch": 0.05371, + "grad_norm": 0.8077481499415716, + "learning_rate": 0.003, + "loss": 4.1506, + "step": 5371 + }, + { + "epoch": 0.05372, + "grad_norm": 0.9098981004391098, + "learning_rate": 0.003, + "loss": 4.1303, + "step": 5372 + }, + { + "epoch": 0.05373, + "grad_norm": 0.8688252009595805, + "learning_rate": 0.003, + "loss": 4.1478, + "step": 5373 + }, + { + "epoch": 0.05374, + "grad_norm": 0.7822281391029792, + "learning_rate": 0.003, + "loss": 4.1242, + "step": 5374 + }, + { + "epoch": 0.05375, + "grad_norm": 0.6393823720543013, + "learning_rate": 0.003, + "loss": 4.1167, + "step": 5375 + }, + { + "epoch": 0.05376, + "grad_norm": 0.6297681205873783, + "learning_rate": 0.003, + "loss": 4.1424, + "step": 5376 + }, + { + "epoch": 0.05377, + "grad_norm": 0.5591158983975534, + "learning_rate": 0.003, + "loss": 4.1287, + "step": 5377 + }, + { + "epoch": 0.05378, + "grad_norm": 0.5827833277734367, + "learning_rate": 0.003, + "loss": 4.1316, + "step": 5378 + }, + { + "epoch": 0.05379, + "grad_norm": 0.5267920451244182, + "learning_rate": 0.003, + "loss": 4.1356, + "step": 5379 + }, + { + "epoch": 0.0538, + "grad_norm": 0.5153469929794081, + "learning_rate": 0.003, + "loss": 4.1031, + "step": 5380 + }, + { + "epoch": 0.05381, + "grad_norm": 0.541236643808538, + "learning_rate": 0.003, + "loss": 4.115, + "step": 5381 + }, + { + "epoch": 0.05382, + "grad_norm": 0.5524397786242985, + "learning_rate": 0.003, + "loss": 4.1335, + "step": 5382 + }, + { + "epoch": 0.05383, + "grad_norm": 0.5171016444028097, + "learning_rate": 0.003, + "loss": 4.1294, + "step": 5383 + }, + { + "epoch": 0.05384, + "grad_norm": 0.4322886075458978, + "learning_rate": 0.003, + "loss": 4.1065, + "step": 5384 + }, + { + "epoch": 0.05385, + "grad_norm": 0.45234091719568803, + "learning_rate": 0.003, + "loss": 4.1332, + "step": 5385 + }, + { + "epoch": 0.05386, + "grad_norm": 0.4669069348140255, + "learning_rate": 0.003, + "loss": 4.1133, + "step": 5386 + }, + { + "epoch": 0.05387, + "grad_norm": 0.505314292670269, + "learning_rate": 0.003, + "loss": 4.0983, + "step": 5387 + }, + { + "epoch": 0.05388, + "grad_norm": 0.6661928342962524, + "learning_rate": 0.003, + "loss": 4.1388, + "step": 5388 + }, + { + "epoch": 0.05389, + "grad_norm": 1.0017958320911928, + "learning_rate": 0.003, + "loss": 4.1206, + "step": 5389 + }, + { + "epoch": 0.0539, + "grad_norm": 1.2992679013311437, + "learning_rate": 0.003, + "loss": 4.1067, + "step": 5390 + }, + { + "epoch": 0.05391, + "grad_norm": 0.5736546179541971, + "learning_rate": 0.003, + "loss": 4.1074, + "step": 5391 + }, + { + "epoch": 0.05392, + "grad_norm": 0.7440996442059347, + "learning_rate": 0.003, + "loss": 4.1139, + "step": 5392 + }, + { + "epoch": 0.05393, + "grad_norm": 0.9950521432118898, + "learning_rate": 0.003, + "loss": 4.1497, + "step": 5393 + }, + { + "epoch": 0.05394, + "grad_norm": 0.7875014530755975, + "learning_rate": 0.003, + "loss": 4.1182, + "step": 5394 + }, + { + "epoch": 0.05395, + "grad_norm": 0.7254081418838102, + "learning_rate": 0.003, + "loss": 4.1479, + "step": 5395 + }, + { + "epoch": 0.05396, + "grad_norm": 0.8612136110888624, + "learning_rate": 0.003, + "loss": 4.143, + "step": 5396 + }, + { + "epoch": 0.05397, + "grad_norm": 0.8986673711381578, + "learning_rate": 0.003, + "loss": 4.122, + "step": 5397 + }, + { + "epoch": 0.05398, + "grad_norm": 0.778809786786069, + "learning_rate": 0.003, + "loss": 4.1161, + "step": 5398 + }, + { + "epoch": 0.05399, + "grad_norm": 0.7995244546793957, + "learning_rate": 0.003, + "loss": 4.1442, + "step": 5399 + }, + { + "epoch": 0.054, + "grad_norm": 0.7896689616881526, + "learning_rate": 0.003, + "loss": 4.1557, + "step": 5400 + }, + { + "epoch": 0.05401, + "grad_norm": 0.7804724282282718, + "learning_rate": 0.003, + "loss": 4.1348, + "step": 5401 + }, + { + "epoch": 0.05402, + "grad_norm": 0.7709619532218206, + "learning_rate": 0.003, + "loss": 4.1489, + "step": 5402 + }, + { + "epoch": 0.05403, + "grad_norm": 0.7494119205086632, + "learning_rate": 0.003, + "loss": 4.133, + "step": 5403 + }, + { + "epoch": 0.05404, + "grad_norm": 0.7695850697537835, + "learning_rate": 0.003, + "loss": 4.128, + "step": 5404 + }, + { + "epoch": 0.05405, + "grad_norm": 0.6494919985184457, + "learning_rate": 0.003, + "loss": 4.1251, + "step": 5405 + }, + { + "epoch": 0.05406, + "grad_norm": 0.7107256413996718, + "learning_rate": 0.003, + "loss": 4.1092, + "step": 5406 + }, + { + "epoch": 0.05407, + "grad_norm": 0.8097891472467551, + "learning_rate": 0.003, + "loss": 4.1255, + "step": 5407 + }, + { + "epoch": 0.05408, + "grad_norm": 0.9502406141177873, + "learning_rate": 0.003, + "loss": 4.1168, + "step": 5408 + }, + { + "epoch": 0.05409, + "grad_norm": 0.9823219241510378, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 5409 + }, + { + "epoch": 0.0541, + "grad_norm": 0.878184005124274, + "learning_rate": 0.003, + "loss": 4.1409, + "step": 5410 + }, + { + "epoch": 0.05411, + "grad_norm": 0.783228910710241, + "learning_rate": 0.003, + "loss": 4.1364, + "step": 5411 + }, + { + "epoch": 0.05412, + "grad_norm": 0.7419337388906813, + "learning_rate": 0.003, + "loss": 4.1309, + "step": 5412 + }, + { + "epoch": 0.05413, + "grad_norm": 0.8677947411304687, + "learning_rate": 0.003, + "loss": 4.1314, + "step": 5413 + }, + { + "epoch": 0.05414, + "grad_norm": 0.754625111832618, + "learning_rate": 0.003, + "loss": 4.1077, + "step": 5414 + }, + { + "epoch": 0.05415, + "grad_norm": 0.6166571551752229, + "learning_rate": 0.003, + "loss": 4.1407, + "step": 5415 + }, + { + "epoch": 0.05416, + "grad_norm": 0.6149867683651459, + "learning_rate": 0.003, + "loss": 4.0953, + "step": 5416 + }, + { + "epoch": 0.05417, + "grad_norm": 0.7707506569643702, + "learning_rate": 0.003, + "loss": 4.1237, + "step": 5417 + }, + { + "epoch": 0.05418, + "grad_norm": 0.8827482178277591, + "learning_rate": 0.003, + "loss": 4.12, + "step": 5418 + }, + { + "epoch": 0.05419, + "grad_norm": 1.0696377805286104, + "learning_rate": 0.003, + "loss": 4.1453, + "step": 5419 + }, + { + "epoch": 0.0542, + "grad_norm": 1.0762623846918855, + "learning_rate": 0.003, + "loss": 4.1305, + "step": 5420 + }, + { + "epoch": 0.05421, + "grad_norm": 1.0370239550620723, + "learning_rate": 0.003, + "loss": 4.1296, + "step": 5421 + }, + { + "epoch": 0.05422, + "grad_norm": 0.8836609588737956, + "learning_rate": 0.003, + "loss": 4.1453, + "step": 5422 + }, + { + "epoch": 0.05423, + "grad_norm": 0.9202200994596317, + "learning_rate": 0.003, + "loss": 4.1371, + "step": 5423 + }, + { + "epoch": 0.05424, + "grad_norm": 0.8512778459324802, + "learning_rate": 0.003, + "loss": 4.1658, + "step": 5424 + }, + { + "epoch": 0.05425, + "grad_norm": 0.9221497425172506, + "learning_rate": 0.003, + "loss": 4.1457, + "step": 5425 + }, + { + "epoch": 0.05426, + "grad_norm": 0.8858038824729005, + "learning_rate": 0.003, + "loss": 4.1296, + "step": 5426 + }, + { + "epoch": 0.05427, + "grad_norm": 0.8461598509812891, + "learning_rate": 0.003, + "loss": 4.1252, + "step": 5427 + }, + { + "epoch": 0.05428, + "grad_norm": 0.7838983529721615, + "learning_rate": 0.003, + "loss": 4.1428, + "step": 5428 + }, + { + "epoch": 0.05429, + "grad_norm": 0.7381870936331452, + "learning_rate": 0.003, + "loss": 4.1432, + "step": 5429 + }, + { + "epoch": 0.0543, + "grad_norm": 0.6675438001202542, + "learning_rate": 0.003, + "loss": 4.1583, + "step": 5430 + }, + { + "epoch": 0.05431, + "grad_norm": 0.7468875444789291, + "learning_rate": 0.003, + "loss": 4.1301, + "step": 5431 + }, + { + "epoch": 0.05432, + "grad_norm": 0.7685334970223217, + "learning_rate": 0.003, + "loss": 4.1248, + "step": 5432 + }, + { + "epoch": 0.05433, + "grad_norm": 0.7841989897169739, + "learning_rate": 0.003, + "loss": 4.0999, + "step": 5433 + }, + { + "epoch": 0.05434, + "grad_norm": 0.7643517687218986, + "learning_rate": 0.003, + "loss": 4.1241, + "step": 5434 + }, + { + "epoch": 0.05435, + "grad_norm": 0.7188445789131351, + "learning_rate": 0.003, + "loss": 4.1133, + "step": 5435 + }, + { + "epoch": 0.05436, + "grad_norm": 0.6666336203457994, + "learning_rate": 0.003, + "loss": 4.1339, + "step": 5436 + }, + { + "epoch": 0.05437, + "grad_norm": 0.5578867371709249, + "learning_rate": 0.003, + "loss": 4.1353, + "step": 5437 + }, + { + "epoch": 0.05438, + "grad_norm": 0.6023052773273868, + "learning_rate": 0.003, + "loss": 4.1276, + "step": 5438 + }, + { + "epoch": 0.05439, + "grad_norm": 0.686819083498365, + "learning_rate": 0.003, + "loss": 4.1193, + "step": 5439 + }, + { + "epoch": 0.0544, + "grad_norm": 0.7877497063451424, + "learning_rate": 0.003, + "loss": 4.1164, + "step": 5440 + }, + { + "epoch": 0.05441, + "grad_norm": 0.8224525297404207, + "learning_rate": 0.003, + "loss": 4.148, + "step": 5441 + }, + { + "epoch": 0.05442, + "grad_norm": 0.8963412844552312, + "learning_rate": 0.003, + "loss": 4.1128, + "step": 5442 + }, + { + "epoch": 0.05443, + "grad_norm": 0.7843413799075878, + "learning_rate": 0.003, + "loss": 4.1267, + "step": 5443 + }, + { + "epoch": 0.05444, + "grad_norm": 0.6677654509581011, + "learning_rate": 0.003, + "loss": 4.1127, + "step": 5444 + }, + { + "epoch": 0.05445, + "grad_norm": 0.5372869992196785, + "learning_rate": 0.003, + "loss": 4.1083, + "step": 5445 + }, + { + "epoch": 0.05446, + "grad_norm": 0.4853458301953657, + "learning_rate": 0.003, + "loss": 4.1094, + "step": 5446 + }, + { + "epoch": 0.05447, + "grad_norm": 0.534235282926048, + "learning_rate": 0.003, + "loss": 4.1032, + "step": 5447 + }, + { + "epoch": 0.05448, + "grad_norm": 0.6051573720194104, + "learning_rate": 0.003, + "loss": 4.1417, + "step": 5448 + }, + { + "epoch": 0.05449, + "grad_norm": 0.7586494998052061, + "learning_rate": 0.003, + "loss": 4.1358, + "step": 5449 + }, + { + "epoch": 0.0545, + "grad_norm": 0.8213010225949955, + "learning_rate": 0.003, + "loss": 4.1342, + "step": 5450 + }, + { + "epoch": 0.05451, + "grad_norm": 0.7188380237062496, + "learning_rate": 0.003, + "loss": 4.1539, + "step": 5451 + }, + { + "epoch": 0.05452, + "grad_norm": 0.5272820320310276, + "learning_rate": 0.003, + "loss": 4.1121, + "step": 5452 + }, + { + "epoch": 0.05453, + "grad_norm": 0.4825898765859038, + "learning_rate": 0.003, + "loss": 4.1071, + "step": 5453 + }, + { + "epoch": 0.05454, + "grad_norm": 0.5231707853046749, + "learning_rate": 0.003, + "loss": 4.1133, + "step": 5454 + }, + { + "epoch": 0.05455, + "grad_norm": 0.6109918111895452, + "learning_rate": 0.003, + "loss": 4.1419, + "step": 5455 + }, + { + "epoch": 0.05456, + "grad_norm": 0.717222959857415, + "learning_rate": 0.003, + "loss": 4.1169, + "step": 5456 + }, + { + "epoch": 0.05457, + "grad_norm": 0.8551566509250617, + "learning_rate": 0.003, + "loss": 4.1442, + "step": 5457 + }, + { + "epoch": 0.05458, + "grad_norm": 0.9372765867290779, + "learning_rate": 0.003, + "loss": 4.1202, + "step": 5458 + }, + { + "epoch": 0.05459, + "grad_norm": 0.9320602167981404, + "learning_rate": 0.003, + "loss": 4.1324, + "step": 5459 + }, + { + "epoch": 0.0546, + "grad_norm": 0.7627661268344466, + "learning_rate": 0.003, + "loss": 4.1183, + "step": 5460 + }, + { + "epoch": 0.05461, + "grad_norm": 0.6315449710507808, + "learning_rate": 0.003, + "loss": 4.1204, + "step": 5461 + }, + { + "epoch": 0.05462, + "grad_norm": 0.7123926231181666, + "learning_rate": 0.003, + "loss": 4.1466, + "step": 5462 + }, + { + "epoch": 0.05463, + "grad_norm": 0.7549211607821605, + "learning_rate": 0.003, + "loss": 4.1081, + "step": 5463 + }, + { + "epoch": 0.05464, + "grad_norm": 0.8370379354643269, + "learning_rate": 0.003, + "loss": 4.1503, + "step": 5464 + }, + { + "epoch": 0.05465, + "grad_norm": 0.9542926928325179, + "learning_rate": 0.003, + "loss": 4.1515, + "step": 5465 + }, + { + "epoch": 0.05466, + "grad_norm": 1.054115432669863, + "learning_rate": 0.003, + "loss": 4.1695, + "step": 5466 + }, + { + "epoch": 0.05467, + "grad_norm": 0.8917210500728532, + "learning_rate": 0.003, + "loss": 4.1439, + "step": 5467 + }, + { + "epoch": 0.05468, + "grad_norm": 0.8412209468493033, + "learning_rate": 0.003, + "loss": 4.1181, + "step": 5468 + }, + { + "epoch": 0.05469, + "grad_norm": 0.7971647310022061, + "learning_rate": 0.003, + "loss": 4.1111, + "step": 5469 + }, + { + "epoch": 0.0547, + "grad_norm": 0.8775383662555644, + "learning_rate": 0.003, + "loss": 4.1182, + "step": 5470 + }, + { + "epoch": 0.05471, + "grad_norm": 0.9125028901592531, + "learning_rate": 0.003, + "loss": 4.1309, + "step": 5471 + }, + { + "epoch": 0.05472, + "grad_norm": 0.7767839416762876, + "learning_rate": 0.003, + "loss": 4.1215, + "step": 5472 + }, + { + "epoch": 0.05473, + "grad_norm": 0.7509158035092801, + "learning_rate": 0.003, + "loss": 4.1403, + "step": 5473 + }, + { + "epoch": 0.05474, + "grad_norm": 0.8240995260680938, + "learning_rate": 0.003, + "loss": 4.1154, + "step": 5474 + }, + { + "epoch": 0.05475, + "grad_norm": 0.9739217618215972, + "learning_rate": 0.003, + "loss": 4.1769, + "step": 5475 + }, + { + "epoch": 0.05476, + "grad_norm": 1.0112035995282798, + "learning_rate": 0.003, + "loss": 4.1044, + "step": 5476 + }, + { + "epoch": 0.05477, + "grad_norm": 0.8701723078990121, + "learning_rate": 0.003, + "loss": 4.1252, + "step": 5477 + }, + { + "epoch": 0.05478, + "grad_norm": 0.745391995716312, + "learning_rate": 0.003, + "loss": 4.12, + "step": 5478 + }, + { + "epoch": 0.05479, + "grad_norm": 0.7047469678929648, + "learning_rate": 0.003, + "loss": 4.1376, + "step": 5479 + }, + { + "epoch": 0.0548, + "grad_norm": 0.7568179897684124, + "learning_rate": 0.003, + "loss": 4.1371, + "step": 5480 + }, + { + "epoch": 0.05481, + "grad_norm": 0.842291746689268, + "learning_rate": 0.003, + "loss": 4.1302, + "step": 5481 + }, + { + "epoch": 0.05482, + "grad_norm": 0.8335768637729064, + "learning_rate": 0.003, + "loss": 4.1229, + "step": 5482 + }, + { + "epoch": 0.05483, + "grad_norm": 0.8356448898440997, + "learning_rate": 0.003, + "loss": 4.1231, + "step": 5483 + }, + { + "epoch": 0.05484, + "grad_norm": 0.8347694767449836, + "learning_rate": 0.003, + "loss": 4.1485, + "step": 5484 + }, + { + "epoch": 0.05485, + "grad_norm": 0.7596611463998084, + "learning_rate": 0.003, + "loss": 4.15, + "step": 5485 + }, + { + "epoch": 0.05486, + "grad_norm": 0.7203868901609732, + "learning_rate": 0.003, + "loss": 4.1318, + "step": 5486 + }, + { + "epoch": 0.05487, + "grad_norm": 0.7045034193470036, + "learning_rate": 0.003, + "loss": 4.1376, + "step": 5487 + }, + { + "epoch": 0.05488, + "grad_norm": 0.7134164831679168, + "learning_rate": 0.003, + "loss": 4.1332, + "step": 5488 + }, + { + "epoch": 0.05489, + "grad_norm": 0.7566204654677959, + "learning_rate": 0.003, + "loss": 4.1554, + "step": 5489 + }, + { + "epoch": 0.0549, + "grad_norm": 0.753089787028579, + "learning_rate": 0.003, + "loss": 4.1274, + "step": 5490 + }, + { + "epoch": 0.05491, + "grad_norm": 0.7900961174130867, + "learning_rate": 0.003, + "loss": 4.0994, + "step": 5491 + }, + { + "epoch": 0.05492, + "grad_norm": 0.7735481675330634, + "learning_rate": 0.003, + "loss": 4.1455, + "step": 5492 + }, + { + "epoch": 0.05493, + "grad_norm": 0.7921324153077896, + "learning_rate": 0.003, + "loss": 4.1113, + "step": 5493 + }, + { + "epoch": 0.05494, + "grad_norm": 0.7783756160879682, + "learning_rate": 0.003, + "loss": 4.1404, + "step": 5494 + }, + { + "epoch": 0.05495, + "grad_norm": 0.8354402801680301, + "learning_rate": 0.003, + "loss": 4.1288, + "step": 5495 + }, + { + "epoch": 0.05496, + "grad_norm": 0.6922018070920165, + "learning_rate": 0.003, + "loss": 4.1145, + "step": 5496 + }, + { + "epoch": 0.05497, + "grad_norm": 0.6446395374277146, + "learning_rate": 0.003, + "loss": 4.1365, + "step": 5497 + }, + { + "epoch": 0.05498, + "grad_norm": 0.6080103999530705, + "learning_rate": 0.003, + "loss": 4.1147, + "step": 5498 + }, + { + "epoch": 0.05499, + "grad_norm": 0.6519375925877654, + "learning_rate": 0.003, + "loss": 4.1023, + "step": 5499 + }, + { + "epoch": 0.055, + "grad_norm": 0.5807988789895889, + "learning_rate": 0.003, + "loss": 4.1058, + "step": 5500 + }, + { + "epoch": 0.05501, + "grad_norm": 0.5252348957768659, + "learning_rate": 0.003, + "loss": 4.1129, + "step": 5501 + }, + { + "epoch": 0.05502, + "grad_norm": 0.6442067499919607, + "learning_rate": 0.003, + "loss": 4.1035, + "step": 5502 + }, + { + "epoch": 0.05503, + "grad_norm": 0.6540019754179668, + "learning_rate": 0.003, + "loss": 4.1065, + "step": 5503 + }, + { + "epoch": 0.05504, + "grad_norm": 0.6146856306540659, + "learning_rate": 0.003, + "loss": 4.1579, + "step": 5504 + }, + { + "epoch": 0.05505, + "grad_norm": 0.6120251915961363, + "learning_rate": 0.003, + "loss": 4.1105, + "step": 5505 + }, + { + "epoch": 0.05506, + "grad_norm": 0.7454876837271709, + "learning_rate": 0.003, + "loss": 4.0868, + "step": 5506 + }, + { + "epoch": 0.05507, + "grad_norm": 0.7761307027748946, + "learning_rate": 0.003, + "loss": 4.1446, + "step": 5507 + }, + { + "epoch": 0.05508, + "grad_norm": 0.7800855127913302, + "learning_rate": 0.003, + "loss": 4.1244, + "step": 5508 + }, + { + "epoch": 0.05509, + "grad_norm": 0.9996587539628538, + "learning_rate": 0.003, + "loss": 4.126, + "step": 5509 + }, + { + "epoch": 0.0551, + "grad_norm": 1.2203215437155306, + "learning_rate": 0.003, + "loss": 4.1171, + "step": 5510 + }, + { + "epoch": 0.05511, + "grad_norm": 0.8793583971610104, + "learning_rate": 0.003, + "loss": 4.1355, + "step": 5511 + }, + { + "epoch": 0.05512, + "grad_norm": 0.7370740348757334, + "learning_rate": 0.003, + "loss": 4.1383, + "step": 5512 + }, + { + "epoch": 0.05513, + "grad_norm": 0.6510573891135065, + "learning_rate": 0.003, + "loss": 4.1211, + "step": 5513 + }, + { + "epoch": 0.05514, + "grad_norm": 0.7249273397693745, + "learning_rate": 0.003, + "loss": 4.1304, + "step": 5514 + }, + { + "epoch": 0.05515, + "grad_norm": 0.7885147514411773, + "learning_rate": 0.003, + "loss": 4.1145, + "step": 5515 + }, + { + "epoch": 0.05516, + "grad_norm": 0.9148149024980577, + "learning_rate": 0.003, + "loss": 4.1517, + "step": 5516 + }, + { + "epoch": 0.05517, + "grad_norm": 0.9500608738728219, + "learning_rate": 0.003, + "loss": 4.1135, + "step": 5517 + }, + { + "epoch": 0.05518, + "grad_norm": 0.8711962258266909, + "learning_rate": 0.003, + "loss": 4.1339, + "step": 5518 + }, + { + "epoch": 0.05519, + "grad_norm": 0.8912312013555587, + "learning_rate": 0.003, + "loss": 4.1629, + "step": 5519 + }, + { + "epoch": 0.0552, + "grad_norm": 0.7356449479297673, + "learning_rate": 0.003, + "loss": 4.1323, + "step": 5520 + }, + { + "epoch": 0.05521, + "grad_norm": 0.652079639982586, + "learning_rate": 0.003, + "loss": 4.1438, + "step": 5521 + }, + { + "epoch": 0.05522, + "grad_norm": 0.7227034463026758, + "learning_rate": 0.003, + "loss": 4.1587, + "step": 5522 + }, + { + "epoch": 0.05523, + "grad_norm": 0.6646962930116027, + "learning_rate": 0.003, + "loss": 4.1304, + "step": 5523 + }, + { + "epoch": 0.05524, + "grad_norm": 0.6345664894540247, + "learning_rate": 0.003, + "loss": 4.1448, + "step": 5524 + }, + { + "epoch": 0.05525, + "grad_norm": 0.7628485020641987, + "learning_rate": 0.003, + "loss": 4.1119, + "step": 5525 + }, + { + "epoch": 0.05526, + "grad_norm": 0.9521465562602499, + "learning_rate": 0.003, + "loss": 4.1243, + "step": 5526 + }, + { + "epoch": 0.05527, + "grad_norm": 0.9654484523561925, + "learning_rate": 0.003, + "loss": 4.0874, + "step": 5527 + }, + { + "epoch": 0.05528, + "grad_norm": 0.9467279300817267, + "learning_rate": 0.003, + "loss": 4.137, + "step": 5528 + }, + { + "epoch": 0.05529, + "grad_norm": 0.8441582406490136, + "learning_rate": 0.003, + "loss": 4.1446, + "step": 5529 + }, + { + "epoch": 0.0553, + "grad_norm": 0.8409264258057348, + "learning_rate": 0.003, + "loss": 4.1574, + "step": 5530 + }, + { + "epoch": 0.05531, + "grad_norm": 0.898368141131664, + "learning_rate": 0.003, + "loss": 4.1485, + "step": 5531 + }, + { + "epoch": 0.05532, + "grad_norm": 0.9193518146488663, + "learning_rate": 0.003, + "loss": 4.1331, + "step": 5532 + }, + { + "epoch": 0.05533, + "grad_norm": 0.9804221843926488, + "learning_rate": 0.003, + "loss": 4.1541, + "step": 5533 + }, + { + "epoch": 0.05534, + "grad_norm": 1.0187659844050676, + "learning_rate": 0.003, + "loss": 4.1449, + "step": 5534 + }, + { + "epoch": 0.05535, + "grad_norm": 0.9901354181855488, + "learning_rate": 0.003, + "loss": 4.1562, + "step": 5535 + }, + { + "epoch": 0.05536, + "grad_norm": 0.8825476878279837, + "learning_rate": 0.003, + "loss": 4.1468, + "step": 5536 + }, + { + "epoch": 0.05537, + "grad_norm": 0.7556391963162472, + "learning_rate": 0.003, + "loss": 4.1375, + "step": 5537 + }, + { + "epoch": 0.05538, + "grad_norm": 0.6817528820545807, + "learning_rate": 0.003, + "loss": 4.1332, + "step": 5538 + }, + { + "epoch": 0.05539, + "grad_norm": 0.7027437428476654, + "learning_rate": 0.003, + "loss": 4.1186, + "step": 5539 + }, + { + "epoch": 0.0554, + "grad_norm": 0.7674376622806265, + "learning_rate": 0.003, + "loss": 4.1747, + "step": 5540 + }, + { + "epoch": 0.05541, + "grad_norm": 0.9105163029193267, + "learning_rate": 0.003, + "loss": 4.1584, + "step": 5541 + }, + { + "epoch": 0.05542, + "grad_norm": 0.8086875293524934, + "learning_rate": 0.003, + "loss": 4.1253, + "step": 5542 + }, + { + "epoch": 0.05543, + "grad_norm": 0.7420293385687333, + "learning_rate": 0.003, + "loss": 4.1174, + "step": 5543 + }, + { + "epoch": 0.05544, + "grad_norm": 0.8559082347379338, + "learning_rate": 0.003, + "loss": 4.1594, + "step": 5544 + }, + { + "epoch": 0.05545, + "grad_norm": 0.8152608790796877, + "learning_rate": 0.003, + "loss": 4.1494, + "step": 5545 + }, + { + "epoch": 0.05546, + "grad_norm": 0.6798845886566196, + "learning_rate": 0.003, + "loss": 4.1289, + "step": 5546 + }, + { + "epoch": 0.05547, + "grad_norm": 0.69937059223718, + "learning_rate": 0.003, + "loss": 4.1266, + "step": 5547 + }, + { + "epoch": 0.05548, + "grad_norm": 0.6313128061744391, + "learning_rate": 0.003, + "loss": 4.0991, + "step": 5548 + }, + { + "epoch": 0.05549, + "grad_norm": 0.6803316037182424, + "learning_rate": 0.003, + "loss": 4.1454, + "step": 5549 + }, + { + "epoch": 0.0555, + "grad_norm": 0.8849647916705993, + "learning_rate": 0.003, + "loss": 4.1291, + "step": 5550 + }, + { + "epoch": 0.05551, + "grad_norm": 0.9946377104603599, + "learning_rate": 0.003, + "loss": 4.1016, + "step": 5551 + }, + { + "epoch": 0.05552, + "grad_norm": 0.9054479922928002, + "learning_rate": 0.003, + "loss": 4.1164, + "step": 5552 + }, + { + "epoch": 0.05553, + "grad_norm": 0.8058698994488415, + "learning_rate": 0.003, + "loss": 4.14, + "step": 5553 + }, + { + "epoch": 0.05554, + "grad_norm": 0.8113670227482532, + "learning_rate": 0.003, + "loss": 4.1253, + "step": 5554 + }, + { + "epoch": 0.05555, + "grad_norm": 0.8623032272003974, + "learning_rate": 0.003, + "loss": 4.1697, + "step": 5555 + }, + { + "epoch": 0.05556, + "grad_norm": 0.7815305295497466, + "learning_rate": 0.003, + "loss": 4.1187, + "step": 5556 + }, + { + "epoch": 0.05557, + "grad_norm": 0.7613158957065711, + "learning_rate": 0.003, + "loss": 4.1482, + "step": 5557 + }, + { + "epoch": 0.05558, + "grad_norm": 0.6848954615857884, + "learning_rate": 0.003, + "loss": 4.1024, + "step": 5558 + }, + { + "epoch": 0.05559, + "grad_norm": 0.6558580775338294, + "learning_rate": 0.003, + "loss": 4.1263, + "step": 5559 + }, + { + "epoch": 0.0556, + "grad_norm": 0.6949517459401227, + "learning_rate": 0.003, + "loss": 4.1522, + "step": 5560 + }, + { + "epoch": 0.05561, + "grad_norm": 0.6020633152546068, + "learning_rate": 0.003, + "loss": 4.1023, + "step": 5561 + }, + { + "epoch": 0.05562, + "grad_norm": 0.6130293848011331, + "learning_rate": 0.003, + "loss": 4.1158, + "step": 5562 + }, + { + "epoch": 0.05563, + "grad_norm": 0.5305748367682133, + "learning_rate": 0.003, + "loss": 4.1211, + "step": 5563 + }, + { + "epoch": 0.05564, + "grad_norm": 0.5360201658815928, + "learning_rate": 0.003, + "loss": 4.1054, + "step": 5564 + }, + { + "epoch": 0.05565, + "grad_norm": 0.5782685165038048, + "learning_rate": 0.003, + "loss": 4.1461, + "step": 5565 + }, + { + "epoch": 0.05566, + "grad_norm": 0.7133895920716543, + "learning_rate": 0.003, + "loss": 4.0986, + "step": 5566 + }, + { + "epoch": 0.05567, + "grad_norm": 0.826072535494595, + "learning_rate": 0.003, + "loss": 4.1109, + "step": 5567 + }, + { + "epoch": 0.05568, + "grad_norm": 0.9865934604781055, + "learning_rate": 0.003, + "loss": 4.1465, + "step": 5568 + }, + { + "epoch": 0.05569, + "grad_norm": 1.1577665175633736, + "learning_rate": 0.003, + "loss": 4.1238, + "step": 5569 + }, + { + "epoch": 0.0557, + "grad_norm": 0.836822155212909, + "learning_rate": 0.003, + "loss": 4.1424, + "step": 5570 + }, + { + "epoch": 0.05571, + "grad_norm": 0.8107309429680243, + "learning_rate": 0.003, + "loss": 4.1347, + "step": 5571 + }, + { + "epoch": 0.05572, + "grad_norm": 0.7471424734568298, + "learning_rate": 0.003, + "loss": 4.1077, + "step": 5572 + }, + { + "epoch": 0.05573, + "grad_norm": 0.834234761336203, + "learning_rate": 0.003, + "loss": 4.1533, + "step": 5573 + }, + { + "epoch": 0.05574, + "grad_norm": 0.9390828719578128, + "learning_rate": 0.003, + "loss": 4.1247, + "step": 5574 + }, + { + "epoch": 0.05575, + "grad_norm": 1.0260723102103353, + "learning_rate": 0.003, + "loss": 4.1331, + "step": 5575 + }, + { + "epoch": 0.05576, + "grad_norm": 0.8611281671804792, + "learning_rate": 0.003, + "loss": 4.1032, + "step": 5576 + }, + { + "epoch": 0.05577, + "grad_norm": 0.7889811692592865, + "learning_rate": 0.003, + "loss": 4.1267, + "step": 5577 + }, + { + "epoch": 0.05578, + "grad_norm": 0.8753881201720198, + "learning_rate": 0.003, + "loss": 4.1268, + "step": 5578 + }, + { + "epoch": 0.05579, + "grad_norm": 0.8075261839099657, + "learning_rate": 0.003, + "loss": 4.1474, + "step": 5579 + }, + { + "epoch": 0.0558, + "grad_norm": 0.6451347078049557, + "learning_rate": 0.003, + "loss": 4.1276, + "step": 5580 + }, + { + "epoch": 0.05581, + "grad_norm": 0.5565812999963974, + "learning_rate": 0.003, + "loss": 4.1399, + "step": 5581 + }, + { + "epoch": 0.05582, + "grad_norm": 0.5875801341192183, + "learning_rate": 0.003, + "loss": 4.1519, + "step": 5582 + }, + { + "epoch": 0.05583, + "grad_norm": 0.6172149180481271, + "learning_rate": 0.003, + "loss": 4.1257, + "step": 5583 + }, + { + "epoch": 0.05584, + "grad_norm": 0.6243564462276252, + "learning_rate": 0.003, + "loss": 4.1224, + "step": 5584 + }, + { + "epoch": 0.05585, + "grad_norm": 0.7526662295082331, + "learning_rate": 0.003, + "loss": 4.0942, + "step": 5585 + }, + { + "epoch": 0.05586, + "grad_norm": 0.926843716234651, + "learning_rate": 0.003, + "loss": 4.1419, + "step": 5586 + }, + { + "epoch": 0.05587, + "grad_norm": 1.0068785999700622, + "learning_rate": 0.003, + "loss": 4.1489, + "step": 5587 + }, + { + "epoch": 0.05588, + "grad_norm": 0.9183554908331668, + "learning_rate": 0.003, + "loss": 4.1226, + "step": 5588 + }, + { + "epoch": 0.05589, + "grad_norm": 0.8399340851508159, + "learning_rate": 0.003, + "loss": 4.1356, + "step": 5589 + }, + { + "epoch": 0.0559, + "grad_norm": 0.702522081897405, + "learning_rate": 0.003, + "loss": 4.126, + "step": 5590 + }, + { + "epoch": 0.05591, + "grad_norm": 0.8554841678862133, + "learning_rate": 0.003, + "loss": 4.1384, + "step": 5591 + }, + { + "epoch": 0.05592, + "grad_norm": 1.0796189206376294, + "learning_rate": 0.003, + "loss": 4.1463, + "step": 5592 + }, + { + "epoch": 0.05593, + "grad_norm": 0.8728376363454068, + "learning_rate": 0.003, + "loss": 4.1311, + "step": 5593 + }, + { + "epoch": 0.05594, + "grad_norm": 0.6937202641640116, + "learning_rate": 0.003, + "loss": 4.1183, + "step": 5594 + }, + { + "epoch": 0.05595, + "grad_norm": 0.66356619963666, + "learning_rate": 0.003, + "loss": 4.098, + "step": 5595 + }, + { + "epoch": 0.05596, + "grad_norm": 0.7927953508675363, + "learning_rate": 0.003, + "loss": 4.112, + "step": 5596 + }, + { + "epoch": 0.05597, + "grad_norm": 1.0914599310465052, + "learning_rate": 0.003, + "loss": 4.1531, + "step": 5597 + }, + { + "epoch": 0.05598, + "grad_norm": 0.979078781931696, + "learning_rate": 0.003, + "loss": 4.1089, + "step": 5598 + }, + { + "epoch": 0.05599, + "grad_norm": 0.8737639680425687, + "learning_rate": 0.003, + "loss": 4.1574, + "step": 5599 + }, + { + "epoch": 0.056, + "grad_norm": 0.9488588659190235, + "learning_rate": 0.003, + "loss": 4.1035, + "step": 5600 + }, + { + "epoch": 0.05601, + "grad_norm": 0.9787141618869671, + "learning_rate": 0.003, + "loss": 4.1211, + "step": 5601 + }, + { + "epoch": 0.05602, + "grad_norm": 1.1091138120552808, + "learning_rate": 0.003, + "loss": 4.1233, + "step": 5602 + }, + { + "epoch": 0.05603, + "grad_norm": 0.8587130498163756, + "learning_rate": 0.003, + "loss": 4.1264, + "step": 5603 + }, + { + "epoch": 0.05604, + "grad_norm": 0.7477443911515989, + "learning_rate": 0.003, + "loss": 4.1576, + "step": 5604 + }, + { + "epoch": 0.05605, + "grad_norm": 0.6578770652147883, + "learning_rate": 0.003, + "loss": 4.0874, + "step": 5605 + }, + { + "epoch": 0.05606, + "grad_norm": 0.6529008726545039, + "learning_rate": 0.003, + "loss": 4.1393, + "step": 5606 + }, + { + "epoch": 0.05607, + "grad_norm": 0.6971218018514623, + "learning_rate": 0.003, + "loss": 4.1562, + "step": 5607 + }, + { + "epoch": 0.05608, + "grad_norm": 0.8592836275322725, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 5608 + }, + { + "epoch": 0.05609, + "grad_norm": 1.0589082640672076, + "learning_rate": 0.003, + "loss": 4.135, + "step": 5609 + }, + { + "epoch": 0.0561, + "grad_norm": 0.8647971161448784, + "learning_rate": 0.003, + "loss": 4.1651, + "step": 5610 + }, + { + "epoch": 0.05611, + "grad_norm": 0.7323461767341091, + "learning_rate": 0.003, + "loss": 4.1315, + "step": 5611 + }, + { + "epoch": 0.05612, + "grad_norm": 0.7346877413349945, + "learning_rate": 0.003, + "loss": 4.1228, + "step": 5612 + }, + { + "epoch": 0.05613, + "grad_norm": 0.7046408743706015, + "learning_rate": 0.003, + "loss": 4.1484, + "step": 5613 + }, + { + "epoch": 0.05614, + "grad_norm": 0.7374788617059296, + "learning_rate": 0.003, + "loss": 4.1303, + "step": 5614 + }, + { + "epoch": 0.05615, + "grad_norm": 0.8226983375167506, + "learning_rate": 0.003, + "loss": 4.0993, + "step": 5615 + }, + { + "epoch": 0.05616, + "grad_norm": 0.8679780386223853, + "learning_rate": 0.003, + "loss": 4.117, + "step": 5616 + }, + { + "epoch": 0.05617, + "grad_norm": 0.9315265116837809, + "learning_rate": 0.003, + "loss": 4.1448, + "step": 5617 + }, + { + "epoch": 0.05618, + "grad_norm": 0.7604395735189772, + "learning_rate": 0.003, + "loss": 4.1175, + "step": 5618 + }, + { + "epoch": 0.05619, + "grad_norm": 0.7485651636552902, + "learning_rate": 0.003, + "loss": 4.177, + "step": 5619 + }, + { + "epoch": 0.0562, + "grad_norm": 0.7181098485002947, + "learning_rate": 0.003, + "loss": 4.1281, + "step": 5620 + }, + { + "epoch": 0.05621, + "grad_norm": 0.6644004914528993, + "learning_rate": 0.003, + "loss": 4.1353, + "step": 5621 + }, + { + "epoch": 0.05622, + "grad_norm": 0.6212809614577611, + "learning_rate": 0.003, + "loss": 4.1164, + "step": 5622 + }, + { + "epoch": 0.05623, + "grad_norm": 0.6362873457411731, + "learning_rate": 0.003, + "loss": 4.1385, + "step": 5623 + }, + { + "epoch": 0.05624, + "grad_norm": 0.6095376423594582, + "learning_rate": 0.003, + "loss": 4.1528, + "step": 5624 + }, + { + "epoch": 0.05625, + "grad_norm": 0.6012390884980424, + "learning_rate": 0.003, + "loss": 4.1461, + "step": 5625 + }, + { + "epoch": 0.05626, + "grad_norm": 0.6373515773309039, + "learning_rate": 0.003, + "loss": 4.1126, + "step": 5626 + }, + { + "epoch": 0.05627, + "grad_norm": 0.6997251309707232, + "learning_rate": 0.003, + "loss": 4.1249, + "step": 5627 + }, + { + "epoch": 0.05628, + "grad_norm": 0.7740087467448155, + "learning_rate": 0.003, + "loss": 4.1205, + "step": 5628 + }, + { + "epoch": 0.05629, + "grad_norm": 0.7635869528942756, + "learning_rate": 0.003, + "loss": 4.1309, + "step": 5629 + }, + { + "epoch": 0.0563, + "grad_norm": 0.8230920756703966, + "learning_rate": 0.003, + "loss": 4.1105, + "step": 5630 + }, + { + "epoch": 0.05631, + "grad_norm": 1.0656938602588153, + "learning_rate": 0.003, + "loss": 4.105, + "step": 5631 + }, + { + "epoch": 0.05632, + "grad_norm": 1.0708267403591085, + "learning_rate": 0.003, + "loss": 4.1391, + "step": 5632 + }, + { + "epoch": 0.05633, + "grad_norm": 0.789050996667616, + "learning_rate": 0.003, + "loss": 4.1317, + "step": 5633 + }, + { + "epoch": 0.05634, + "grad_norm": 0.703640516716815, + "learning_rate": 0.003, + "loss": 4.1114, + "step": 5634 + }, + { + "epoch": 0.05635, + "grad_norm": 0.8170971379255785, + "learning_rate": 0.003, + "loss": 4.1341, + "step": 5635 + }, + { + "epoch": 0.05636, + "grad_norm": 0.7643132742033529, + "learning_rate": 0.003, + "loss": 4.1264, + "step": 5636 + }, + { + "epoch": 0.05637, + "grad_norm": 0.7269418593201121, + "learning_rate": 0.003, + "loss": 4.1621, + "step": 5637 + }, + { + "epoch": 0.05638, + "grad_norm": 0.6425160594147294, + "learning_rate": 0.003, + "loss": 4.1113, + "step": 5638 + }, + { + "epoch": 0.05639, + "grad_norm": 0.59424152487318, + "learning_rate": 0.003, + "loss": 4.1173, + "step": 5639 + }, + { + "epoch": 0.0564, + "grad_norm": 0.6307390916938302, + "learning_rate": 0.003, + "loss": 4.1285, + "step": 5640 + }, + { + "epoch": 0.05641, + "grad_norm": 0.6398464382688712, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 5641 + }, + { + "epoch": 0.05642, + "grad_norm": 0.5607522531731973, + "learning_rate": 0.003, + "loss": 4.1426, + "step": 5642 + }, + { + "epoch": 0.05643, + "grad_norm": 0.5848917477606379, + "learning_rate": 0.003, + "loss": 4.1019, + "step": 5643 + }, + { + "epoch": 0.05644, + "grad_norm": 0.6555662778545387, + "learning_rate": 0.003, + "loss": 4.1172, + "step": 5644 + }, + { + "epoch": 0.05645, + "grad_norm": 0.6055581079132784, + "learning_rate": 0.003, + "loss": 4.1071, + "step": 5645 + }, + { + "epoch": 0.05646, + "grad_norm": 0.5727498420416256, + "learning_rate": 0.003, + "loss": 4.1154, + "step": 5646 + }, + { + "epoch": 0.05647, + "grad_norm": 0.5143080586691178, + "learning_rate": 0.003, + "loss": 4.1149, + "step": 5647 + }, + { + "epoch": 0.05648, + "grad_norm": 0.6091773925484103, + "learning_rate": 0.003, + "loss": 4.13, + "step": 5648 + }, + { + "epoch": 0.05649, + "grad_norm": 0.7510364666789875, + "learning_rate": 0.003, + "loss": 4.1253, + "step": 5649 + }, + { + "epoch": 0.0565, + "grad_norm": 0.8427546995305852, + "learning_rate": 0.003, + "loss": 4.1444, + "step": 5650 + }, + { + "epoch": 0.05651, + "grad_norm": 0.9454344139315918, + "learning_rate": 0.003, + "loss": 4.1095, + "step": 5651 + }, + { + "epoch": 0.05652, + "grad_norm": 1.0512411398620864, + "learning_rate": 0.003, + "loss": 4.1114, + "step": 5652 + }, + { + "epoch": 0.05653, + "grad_norm": 1.0651426178923196, + "learning_rate": 0.003, + "loss": 4.1443, + "step": 5653 + }, + { + "epoch": 0.05654, + "grad_norm": 0.8316486999901311, + "learning_rate": 0.003, + "loss": 4.1458, + "step": 5654 + }, + { + "epoch": 0.05655, + "grad_norm": 0.7989690345947827, + "learning_rate": 0.003, + "loss": 4.1368, + "step": 5655 + }, + { + "epoch": 0.05656, + "grad_norm": 0.7220583330129677, + "learning_rate": 0.003, + "loss": 4.1322, + "step": 5656 + }, + { + "epoch": 0.05657, + "grad_norm": 0.6534481422931128, + "learning_rate": 0.003, + "loss": 4.13, + "step": 5657 + }, + { + "epoch": 0.05658, + "grad_norm": 0.5840017265991315, + "learning_rate": 0.003, + "loss": 4.1165, + "step": 5658 + }, + { + "epoch": 0.05659, + "grad_norm": 0.6445129008912936, + "learning_rate": 0.003, + "loss": 4.1191, + "step": 5659 + }, + { + "epoch": 0.0566, + "grad_norm": 0.6350047659342725, + "learning_rate": 0.003, + "loss": 4.1249, + "step": 5660 + }, + { + "epoch": 0.05661, + "grad_norm": 0.75443344313924, + "learning_rate": 0.003, + "loss": 4.1615, + "step": 5661 + }, + { + "epoch": 0.05662, + "grad_norm": 0.8347625040490713, + "learning_rate": 0.003, + "loss": 4.1044, + "step": 5662 + }, + { + "epoch": 0.05663, + "grad_norm": 0.7844268329201943, + "learning_rate": 0.003, + "loss": 4.1254, + "step": 5663 + }, + { + "epoch": 0.05664, + "grad_norm": 0.6882757704839232, + "learning_rate": 0.003, + "loss": 4.1018, + "step": 5664 + }, + { + "epoch": 0.05665, + "grad_norm": 0.7143954784573894, + "learning_rate": 0.003, + "loss": 4.1122, + "step": 5665 + }, + { + "epoch": 0.05666, + "grad_norm": 0.7697196574672253, + "learning_rate": 0.003, + "loss": 4.1278, + "step": 5666 + }, + { + "epoch": 0.05667, + "grad_norm": 0.6930833704435834, + "learning_rate": 0.003, + "loss": 4.0974, + "step": 5667 + }, + { + "epoch": 0.05668, + "grad_norm": 0.7588668210696813, + "learning_rate": 0.003, + "loss": 4.1428, + "step": 5668 + }, + { + "epoch": 0.05669, + "grad_norm": 0.7785704484779421, + "learning_rate": 0.003, + "loss": 4.0928, + "step": 5669 + }, + { + "epoch": 0.0567, + "grad_norm": 0.7557718596536394, + "learning_rate": 0.003, + "loss": 4.1333, + "step": 5670 + }, + { + "epoch": 0.05671, + "grad_norm": 0.7694361102868486, + "learning_rate": 0.003, + "loss": 4.1168, + "step": 5671 + }, + { + "epoch": 0.05672, + "grad_norm": 0.8007083587054245, + "learning_rate": 0.003, + "loss": 4.1211, + "step": 5672 + }, + { + "epoch": 0.05673, + "grad_norm": 0.8580257925305742, + "learning_rate": 0.003, + "loss": 4.0999, + "step": 5673 + }, + { + "epoch": 0.05674, + "grad_norm": 0.9196710283055469, + "learning_rate": 0.003, + "loss": 4.1366, + "step": 5674 + }, + { + "epoch": 0.05675, + "grad_norm": 0.9492581195457488, + "learning_rate": 0.003, + "loss": 4.129, + "step": 5675 + }, + { + "epoch": 0.05676, + "grad_norm": 1.0914070056598988, + "learning_rate": 0.003, + "loss": 4.146, + "step": 5676 + }, + { + "epoch": 0.05677, + "grad_norm": 0.8885367967761989, + "learning_rate": 0.003, + "loss": 4.1347, + "step": 5677 + }, + { + "epoch": 0.05678, + "grad_norm": 0.7905086473346482, + "learning_rate": 0.003, + "loss": 4.1278, + "step": 5678 + }, + { + "epoch": 0.05679, + "grad_norm": 0.6956868554948951, + "learning_rate": 0.003, + "loss": 4.14, + "step": 5679 + }, + { + "epoch": 0.0568, + "grad_norm": 0.7085015911445326, + "learning_rate": 0.003, + "loss": 4.117, + "step": 5680 + }, + { + "epoch": 0.05681, + "grad_norm": 0.7982290524952671, + "learning_rate": 0.003, + "loss": 4.1228, + "step": 5681 + }, + { + "epoch": 0.05682, + "grad_norm": 0.8702967583238443, + "learning_rate": 0.003, + "loss": 4.138, + "step": 5682 + }, + { + "epoch": 0.05683, + "grad_norm": 0.8957753899417163, + "learning_rate": 0.003, + "loss": 4.0995, + "step": 5683 + }, + { + "epoch": 0.05684, + "grad_norm": 0.7828031111834381, + "learning_rate": 0.003, + "loss": 4.1079, + "step": 5684 + }, + { + "epoch": 0.05685, + "grad_norm": 0.7982392827676595, + "learning_rate": 0.003, + "loss": 4.1506, + "step": 5685 + }, + { + "epoch": 0.05686, + "grad_norm": 0.7549245911981988, + "learning_rate": 0.003, + "loss": 4.1508, + "step": 5686 + }, + { + "epoch": 0.05687, + "grad_norm": 0.7845211375597457, + "learning_rate": 0.003, + "loss": 4.103, + "step": 5687 + }, + { + "epoch": 0.05688, + "grad_norm": 0.8652123527501867, + "learning_rate": 0.003, + "loss": 4.1191, + "step": 5688 + }, + { + "epoch": 0.05689, + "grad_norm": 1.1480876620771052, + "learning_rate": 0.003, + "loss": 4.1102, + "step": 5689 + }, + { + "epoch": 0.0569, + "grad_norm": 1.0790768924826935, + "learning_rate": 0.003, + "loss": 4.1365, + "step": 5690 + }, + { + "epoch": 0.05691, + "grad_norm": 1.015989841166836, + "learning_rate": 0.003, + "loss": 4.1641, + "step": 5691 + }, + { + "epoch": 0.05692, + "grad_norm": 0.8594661410624701, + "learning_rate": 0.003, + "loss": 4.1349, + "step": 5692 + }, + { + "epoch": 0.05693, + "grad_norm": 0.6762044853377914, + "learning_rate": 0.003, + "loss": 4.1319, + "step": 5693 + }, + { + "epoch": 0.05694, + "grad_norm": 0.7821423521203583, + "learning_rate": 0.003, + "loss": 4.1115, + "step": 5694 + }, + { + "epoch": 0.05695, + "grad_norm": 0.8134766301370081, + "learning_rate": 0.003, + "loss": 4.1135, + "step": 5695 + }, + { + "epoch": 0.05696, + "grad_norm": 0.8476984030864564, + "learning_rate": 0.003, + "loss": 4.1189, + "step": 5696 + }, + { + "epoch": 0.05697, + "grad_norm": 0.8589446601517351, + "learning_rate": 0.003, + "loss": 4.1562, + "step": 5697 + }, + { + "epoch": 0.05698, + "grad_norm": 0.7927970537237816, + "learning_rate": 0.003, + "loss": 4.1535, + "step": 5698 + }, + { + "epoch": 0.05699, + "grad_norm": 0.6663534787165875, + "learning_rate": 0.003, + "loss": 4.1331, + "step": 5699 + }, + { + "epoch": 0.057, + "grad_norm": 0.6962826442411336, + "learning_rate": 0.003, + "loss": 4.1161, + "step": 5700 + }, + { + "epoch": 0.05701, + "grad_norm": 0.748056579704057, + "learning_rate": 0.003, + "loss": 4.1391, + "step": 5701 + }, + { + "epoch": 0.05702, + "grad_norm": 0.7276987471205317, + "learning_rate": 0.003, + "loss": 4.119, + "step": 5702 + }, + { + "epoch": 0.05703, + "grad_norm": 0.7077636085898174, + "learning_rate": 0.003, + "loss": 4.159, + "step": 5703 + }, + { + "epoch": 0.05704, + "grad_norm": 0.9037559825169571, + "learning_rate": 0.003, + "loss": 4.1266, + "step": 5704 + }, + { + "epoch": 0.05705, + "grad_norm": 0.8075509768599368, + "learning_rate": 0.003, + "loss": 4.124, + "step": 5705 + }, + { + "epoch": 0.05706, + "grad_norm": 0.7674921460937386, + "learning_rate": 0.003, + "loss": 4.1165, + "step": 5706 + }, + { + "epoch": 0.05707, + "grad_norm": 0.7539599082513933, + "learning_rate": 0.003, + "loss": 4.1188, + "step": 5707 + }, + { + "epoch": 0.05708, + "grad_norm": 0.6799305657607112, + "learning_rate": 0.003, + "loss": 4.1537, + "step": 5708 + }, + { + "epoch": 0.05709, + "grad_norm": 0.5479624493880318, + "learning_rate": 0.003, + "loss": 4.1207, + "step": 5709 + }, + { + "epoch": 0.0571, + "grad_norm": 0.5121373459902853, + "learning_rate": 0.003, + "loss": 4.1015, + "step": 5710 + }, + { + "epoch": 0.05711, + "grad_norm": 0.7280903597835207, + "learning_rate": 0.003, + "loss": 4.1343, + "step": 5711 + }, + { + "epoch": 0.05712, + "grad_norm": 1.0090249543371728, + "learning_rate": 0.003, + "loss": 4.1159, + "step": 5712 + }, + { + "epoch": 0.05713, + "grad_norm": 1.1666935879947706, + "learning_rate": 0.003, + "loss": 4.1256, + "step": 5713 + }, + { + "epoch": 0.05714, + "grad_norm": 0.6574654065182656, + "learning_rate": 0.003, + "loss": 4.1657, + "step": 5714 + }, + { + "epoch": 0.05715, + "grad_norm": 0.6534581937051788, + "learning_rate": 0.003, + "loss": 4.1688, + "step": 5715 + }, + { + "epoch": 0.05716, + "grad_norm": 0.7637170840325201, + "learning_rate": 0.003, + "loss": 4.1401, + "step": 5716 + }, + { + "epoch": 0.05717, + "grad_norm": 0.80055490534206, + "learning_rate": 0.003, + "loss": 4.1548, + "step": 5717 + }, + { + "epoch": 0.05718, + "grad_norm": 0.8922767275182982, + "learning_rate": 0.003, + "loss": 4.1398, + "step": 5718 + }, + { + "epoch": 0.05719, + "grad_norm": 0.8933371654984332, + "learning_rate": 0.003, + "loss": 4.142, + "step": 5719 + }, + { + "epoch": 0.0572, + "grad_norm": 0.8966300715229761, + "learning_rate": 0.003, + "loss": 4.1342, + "step": 5720 + }, + { + "epoch": 0.05721, + "grad_norm": 0.8505332145208457, + "learning_rate": 0.003, + "loss": 4.1181, + "step": 5721 + }, + { + "epoch": 0.05722, + "grad_norm": 0.7349175001912129, + "learning_rate": 0.003, + "loss": 4.1237, + "step": 5722 + }, + { + "epoch": 0.05723, + "grad_norm": 0.723220376175918, + "learning_rate": 0.003, + "loss": 4.1257, + "step": 5723 + }, + { + "epoch": 0.05724, + "grad_norm": 0.680563858030686, + "learning_rate": 0.003, + "loss": 4.1224, + "step": 5724 + }, + { + "epoch": 0.05725, + "grad_norm": 0.8088590173981038, + "learning_rate": 0.003, + "loss": 4.1389, + "step": 5725 + }, + { + "epoch": 0.05726, + "grad_norm": 0.9418455002457423, + "learning_rate": 0.003, + "loss": 4.1317, + "step": 5726 + }, + { + "epoch": 0.05727, + "grad_norm": 1.0373122053831039, + "learning_rate": 0.003, + "loss": 4.1217, + "step": 5727 + }, + { + "epoch": 0.05728, + "grad_norm": 0.9012220097719827, + "learning_rate": 0.003, + "loss": 4.1502, + "step": 5728 + }, + { + "epoch": 0.05729, + "grad_norm": 0.9082669562608017, + "learning_rate": 0.003, + "loss": 4.1295, + "step": 5729 + }, + { + "epoch": 0.0573, + "grad_norm": 0.943356157606432, + "learning_rate": 0.003, + "loss": 4.104, + "step": 5730 + }, + { + "epoch": 0.05731, + "grad_norm": 0.8723864537449718, + "learning_rate": 0.003, + "loss": 4.1203, + "step": 5731 + }, + { + "epoch": 0.05732, + "grad_norm": 0.8570219760726053, + "learning_rate": 0.003, + "loss": 4.1478, + "step": 5732 + }, + { + "epoch": 0.05733, + "grad_norm": 0.8298818226350395, + "learning_rate": 0.003, + "loss": 4.1567, + "step": 5733 + }, + { + "epoch": 0.05734, + "grad_norm": 0.8547854470967996, + "learning_rate": 0.003, + "loss": 4.1537, + "step": 5734 + }, + { + "epoch": 0.05735, + "grad_norm": 0.8145784886674767, + "learning_rate": 0.003, + "loss": 4.1161, + "step": 5735 + }, + { + "epoch": 0.05736, + "grad_norm": 0.6881823470117722, + "learning_rate": 0.003, + "loss": 4.139, + "step": 5736 + }, + { + "epoch": 0.05737, + "grad_norm": 0.5844418060230477, + "learning_rate": 0.003, + "loss": 4.1273, + "step": 5737 + }, + { + "epoch": 0.05738, + "grad_norm": 0.5430304440202582, + "learning_rate": 0.003, + "loss": 4.0985, + "step": 5738 + }, + { + "epoch": 0.05739, + "grad_norm": 0.5591474039370316, + "learning_rate": 0.003, + "loss": 4.1055, + "step": 5739 + }, + { + "epoch": 0.0574, + "grad_norm": 0.5762447209115041, + "learning_rate": 0.003, + "loss": 4.1108, + "step": 5740 + }, + { + "epoch": 0.05741, + "grad_norm": 0.5659495615021296, + "learning_rate": 0.003, + "loss": 4.0992, + "step": 5741 + }, + { + "epoch": 0.05742, + "grad_norm": 0.5868859562225888, + "learning_rate": 0.003, + "loss": 4.1159, + "step": 5742 + }, + { + "epoch": 0.05743, + "grad_norm": 0.6325935486380904, + "learning_rate": 0.003, + "loss": 4.1088, + "step": 5743 + }, + { + "epoch": 0.05744, + "grad_norm": 0.7198789457341803, + "learning_rate": 0.003, + "loss": 4.121, + "step": 5744 + }, + { + "epoch": 0.05745, + "grad_norm": 0.8372703754668815, + "learning_rate": 0.003, + "loss": 4.1213, + "step": 5745 + }, + { + "epoch": 0.05746, + "grad_norm": 1.013456290369712, + "learning_rate": 0.003, + "loss": 4.1139, + "step": 5746 + }, + { + "epoch": 0.05747, + "grad_norm": 0.9926529931921089, + "learning_rate": 0.003, + "loss": 4.1411, + "step": 5747 + }, + { + "epoch": 0.05748, + "grad_norm": 0.8673968423379839, + "learning_rate": 0.003, + "loss": 4.1592, + "step": 5748 + }, + { + "epoch": 0.05749, + "grad_norm": 0.8265215218017055, + "learning_rate": 0.003, + "loss": 4.1121, + "step": 5749 + }, + { + "epoch": 0.0575, + "grad_norm": 0.9141071050046009, + "learning_rate": 0.003, + "loss": 4.1328, + "step": 5750 + }, + { + "epoch": 0.05751, + "grad_norm": 0.8867858497305796, + "learning_rate": 0.003, + "loss": 4.1531, + "step": 5751 + }, + { + "epoch": 0.05752, + "grad_norm": 0.8160084253804927, + "learning_rate": 0.003, + "loss": 4.1369, + "step": 5752 + }, + { + "epoch": 0.05753, + "grad_norm": 0.7570642462201534, + "learning_rate": 0.003, + "loss": 4.1294, + "step": 5753 + }, + { + "epoch": 0.05754, + "grad_norm": 0.6931105461575188, + "learning_rate": 0.003, + "loss": 4.1256, + "step": 5754 + }, + { + "epoch": 0.05755, + "grad_norm": 0.7892100308149763, + "learning_rate": 0.003, + "loss": 4.1134, + "step": 5755 + }, + { + "epoch": 0.05756, + "grad_norm": 0.7797288548736725, + "learning_rate": 0.003, + "loss": 4.1459, + "step": 5756 + }, + { + "epoch": 0.05757, + "grad_norm": 0.9143238759273115, + "learning_rate": 0.003, + "loss": 4.134, + "step": 5757 + }, + { + "epoch": 0.05758, + "grad_norm": 0.9568176868550978, + "learning_rate": 0.003, + "loss": 4.1324, + "step": 5758 + }, + { + "epoch": 0.05759, + "grad_norm": 0.9850995699177724, + "learning_rate": 0.003, + "loss": 4.1335, + "step": 5759 + }, + { + "epoch": 0.0576, + "grad_norm": 0.9114825113358783, + "learning_rate": 0.003, + "loss": 4.1321, + "step": 5760 + }, + { + "epoch": 0.05761, + "grad_norm": 0.8393162691101879, + "learning_rate": 0.003, + "loss": 4.1244, + "step": 5761 + }, + { + "epoch": 0.05762, + "grad_norm": 0.8185722596967431, + "learning_rate": 0.003, + "loss": 4.1103, + "step": 5762 + }, + { + "epoch": 0.05763, + "grad_norm": 0.7941154884714461, + "learning_rate": 0.003, + "loss": 4.1287, + "step": 5763 + }, + { + "epoch": 0.05764, + "grad_norm": 0.8167052006322443, + "learning_rate": 0.003, + "loss": 4.128, + "step": 5764 + }, + { + "epoch": 0.05765, + "grad_norm": 0.9075689951343894, + "learning_rate": 0.003, + "loss": 4.1415, + "step": 5765 + }, + { + "epoch": 0.05766, + "grad_norm": 0.9379058663927884, + "learning_rate": 0.003, + "loss": 4.1443, + "step": 5766 + }, + { + "epoch": 0.05767, + "grad_norm": 0.980509823532266, + "learning_rate": 0.003, + "loss": 4.1339, + "step": 5767 + }, + { + "epoch": 0.05768, + "grad_norm": 1.0177076074132854, + "learning_rate": 0.003, + "loss": 4.1549, + "step": 5768 + }, + { + "epoch": 0.05769, + "grad_norm": 1.0429502276363345, + "learning_rate": 0.003, + "loss": 4.139, + "step": 5769 + }, + { + "epoch": 0.0577, + "grad_norm": 1.1100087019496394, + "learning_rate": 0.003, + "loss": 4.1491, + "step": 5770 + }, + { + "epoch": 0.05771, + "grad_norm": 0.7223949464608548, + "learning_rate": 0.003, + "loss": 4.1454, + "step": 5771 + }, + { + "epoch": 0.05772, + "grad_norm": 0.5382829044206766, + "learning_rate": 0.003, + "loss": 4.1464, + "step": 5772 + }, + { + "epoch": 0.05773, + "grad_norm": 0.6406979044556045, + "learning_rate": 0.003, + "loss": 4.1415, + "step": 5773 + }, + { + "epoch": 0.05774, + "grad_norm": 0.8097734695290948, + "learning_rate": 0.003, + "loss": 4.1152, + "step": 5774 + }, + { + "epoch": 0.05775, + "grad_norm": 1.0604370012894522, + "learning_rate": 0.003, + "loss": 4.1297, + "step": 5775 + }, + { + "epoch": 0.05776, + "grad_norm": 1.0011226290122666, + "learning_rate": 0.003, + "loss": 4.1261, + "step": 5776 + }, + { + "epoch": 0.05777, + "grad_norm": 0.799664737208272, + "learning_rate": 0.003, + "loss": 4.1371, + "step": 5777 + }, + { + "epoch": 0.05778, + "grad_norm": 0.7400439760795068, + "learning_rate": 0.003, + "loss": 4.1222, + "step": 5778 + }, + { + "epoch": 0.05779, + "grad_norm": 0.8139200513919773, + "learning_rate": 0.003, + "loss": 4.1449, + "step": 5779 + }, + { + "epoch": 0.0578, + "grad_norm": 0.7887623553765343, + "learning_rate": 0.003, + "loss": 4.1033, + "step": 5780 + }, + { + "epoch": 0.05781, + "grad_norm": 0.6516840422961846, + "learning_rate": 0.003, + "loss": 4.1126, + "step": 5781 + }, + { + "epoch": 0.05782, + "grad_norm": 0.6059160139010061, + "learning_rate": 0.003, + "loss": 4.1428, + "step": 5782 + }, + { + "epoch": 0.05783, + "grad_norm": 0.6322968665722302, + "learning_rate": 0.003, + "loss": 4.0845, + "step": 5783 + }, + { + "epoch": 0.05784, + "grad_norm": 0.6350140210950438, + "learning_rate": 0.003, + "loss": 4.1248, + "step": 5784 + }, + { + "epoch": 0.05785, + "grad_norm": 0.6503751753400261, + "learning_rate": 0.003, + "loss": 4.1374, + "step": 5785 + }, + { + "epoch": 0.05786, + "grad_norm": 0.6778764586815985, + "learning_rate": 0.003, + "loss": 4.1245, + "step": 5786 + }, + { + "epoch": 0.05787, + "grad_norm": 0.5988251572879624, + "learning_rate": 0.003, + "loss": 4.093, + "step": 5787 + }, + { + "epoch": 0.05788, + "grad_norm": 0.4761288844619093, + "learning_rate": 0.003, + "loss": 4.1109, + "step": 5788 + }, + { + "epoch": 0.05789, + "grad_norm": 0.4810545422328699, + "learning_rate": 0.003, + "loss": 4.1101, + "step": 5789 + }, + { + "epoch": 0.0579, + "grad_norm": 0.48719290090643896, + "learning_rate": 0.003, + "loss": 4.1274, + "step": 5790 + }, + { + "epoch": 0.05791, + "grad_norm": 0.6649633480208823, + "learning_rate": 0.003, + "loss": 4.1083, + "step": 5791 + }, + { + "epoch": 0.05792, + "grad_norm": 0.836880563895898, + "learning_rate": 0.003, + "loss": 4.1126, + "step": 5792 + }, + { + "epoch": 0.05793, + "grad_norm": 0.8850525856438148, + "learning_rate": 0.003, + "loss": 4.1047, + "step": 5793 + }, + { + "epoch": 0.05794, + "grad_norm": 0.7567725522310005, + "learning_rate": 0.003, + "loss": 4.118, + "step": 5794 + }, + { + "epoch": 0.05795, + "grad_norm": 0.6015273097107093, + "learning_rate": 0.003, + "loss": 4.1475, + "step": 5795 + }, + { + "epoch": 0.05796, + "grad_norm": 0.6630617490292271, + "learning_rate": 0.003, + "loss": 4.1181, + "step": 5796 + }, + { + "epoch": 0.05797, + "grad_norm": 0.6842122117299039, + "learning_rate": 0.003, + "loss": 4.1188, + "step": 5797 + }, + { + "epoch": 0.05798, + "grad_norm": 0.6729171219216635, + "learning_rate": 0.003, + "loss": 4.0983, + "step": 5798 + }, + { + "epoch": 0.05799, + "grad_norm": 0.6050205840034845, + "learning_rate": 0.003, + "loss": 4.1294, + "step": 5799 + }, + { + "epoch": 0.058, + "grad_norm": 0.5873051359813098, + "learning_rate": 0.003, + "loss": 4.1282, + "step": 5800 + }, + { + "epoch": 0.05801, + "grad_norm": 0.6839087744451585, + "learning_rate": 0.003, + "loss": 4.1102, + "step": 5801 + }, + { + "epoch": 0.05802, + "grad_norm": 0.7630433000288757, + "learning_rate": 0.003, + "loss": 4.1185, + "step": 5802 + }, + { + "epoch": 0.05803, + "grad_norm": 0.8857270500243197, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 5803 + }, + { + "epoch": 0.05804, + "grad_norm": 0.8401507038678849, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 5804 + }, + { + "epoch": 0.05805, + "grad_norm": 0.7475439434893316, + "learning_rate": 0.003, + "loss": 4.0891, + "step": 5805 + }, + { + "epoch": 0.05806, + "grad_norm": 0.7978687405953363, + "learning_rate": 0.003, + "loss": 4.128, + "step": 5806 + }, + { + "epoch": 0.05807, + "grad_norm": 0.8733737077265966, + "learning_rate": 0.003, + "loss": 4.1388, + "step": 5807 + }, + { + "epoch": 0.05808, + "grad_norm": 0.7983979140429381, + "learning_rate": 0.003, + "loss": 4.1475, + "step": 5808 + }, + { + "epoch": 0.05809, + "grad_norm": 0.8908751163199425, + "learning_rate": 0.003, + "loss": 4.1541, + "step": 5809 + }, + { + "epoch": 0.0581, + "grad_norm": 0.9552572694017188, + "learning_rate": 0.003, + "loss": 4.1022, + "step": 5810 + }, + { + "epoch": 0.05811, + "grad_norm": 0.9248914071476577, + "learning_rate": 0.003, + "loss": 4.1204, + "step": 5811 + }, + { + "epoch": 0.05812, + "grad_norm": 0.9817321870382631, + "learning_rate": 0.003, + "loss": 4.1247, + "step": 5812 + }, + { + "epoch": 0.05813, + "grad_norm": 1.0001250001988298, + "learning_rate": 0.003, + "loss": 4.1355, + "step": 5813 + }, + { + "epoch": 0.05814, + "grad_norm": 0.8827566805540904, + "learning_rate": 0.003, + "loss": 4.1153, + "step": 5814 + }, + { + "epoch": 0.05815, + "grad_norm": 0.8029521049860747, + "learning_rate": 0.003, + "loss": 4.094, + "step": 5815 + }, + { + "epoch": 0.05816, + "grad_norm": 0.70939056436584, + "learning_rate": 0.003, + "loss": 4.1343, + "step": 5816 + }, + { + "epoch": 0.05817, + "grad_norm": 0.726146469102038, + "learning_rate": 0.003, + "loss": 4.1461, + "step": 5817 + }, + { + "epoch": 0.05818, + "grad_norm": 0.6869846356437459, + "learning_rate": 0.003, + "loss": 4.1257, + "step": 5818 + }, + { + "epoch": 0.05819, + "grad_norm": 0.7498344281131131, + "learning_rate": 0.003, + "loss": 4.1189, + "step": 5819 + }, + { + "epoch": 0.0582, + "grad_norm": 0.7991793787271796, + "learning_rate": 0.003, + "loss": 4.1253, + "step": 5820 + }, + { + "epoch": 0.05821, + "grad_norm": 0.8692851524335367, + "learning_rate": 0.003, + "loss": 4.1427, + "step": 5821 + }, + { + "epoch": 0.05822, + "grad_norm": 0.9485327101586195, + "learning_rate": 0.003, + "loss": 4.1088, + "step": 5822 + }, + { + "epoch": 0.05823, + "grad_norm": 0.8570830392648408, + "learning_rate": 0.003, + "loss": 4.1397, + "step": 5823 + }, + { + "epoch": 0.05824, + "grad_norm": 0.8874388461019113, + "learning_rate": 0.003, + "loss": 4.1426, + "step": 5824 + }, + { + "epoch": 0.05825, + "grad_norm": 0.7062702547735038, + "learning_rate": 0.003, + "loss": 4.1608, + "step": 5825 + }, + { + "epoch": 0.05826, + "grad_norm": 0.6393477795637886, + "learning_rate": 0.003, + "loss": 4.1078, + "step": 5826 + }, + { + "epoch": 0.05827, + "grad_norm": 0.8019064704761822, + "learning_rate": 0.003, + "loss": 4.1371, + "step": 5827 + }, + { + "epoch": 0.05828, + "grad_norm": 0.935978602016839, + "learning_rate": 0.003, + "loss": 4.0949, + "step": 5828 + }, + { + "epoch": 0.05829, + "grad_norm": 1.0118867950089439, + "learning_rate": 0.003, + "loss": 4.1439, + "step": 5829 + }, + { + "epoch": 0.0583, + "grad_norm": 0.9012310576760173, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 5830 + }, + { + "epoch": 0.05831, + "grad_norm": 0.7921513486784892, + "learning_rate": 0.003, + "loss": 4.1011, + "step": 5831 + }, + { + "epoch": 0.05832, + "grad_norm": 0.8500687381471804, + "learning_rate": 0.003, + "loss": 4.148, + "step": 5832 + }, + { + "epoch": 0.05833, + "grad_norm": 0.8901324912360562, + "learning_rate": 0.003, + "loss": 4.1358, + "step": 5833 + }, + { + "epoch": 0.05834, + "grad_norm": 0.9308154213152057, + "learning_rate": 0.003, + "loss": 4.1442, + "step": 5834 + }, + { + "epoch": 0.05835, + "grad_norm": 0.7729299081011445, + "learning_rate": 0.003, + "loss": 4.1607, + "step": 5835 + }, + { + "epoch": 0.05836, + "grad_norm": 0.6500047045753582, + "learning_rate": 0.003, + "loss": 4.1399, + "step": 5836 + }, + { + "epoch": 0.05837, + "grad_norm": 0.6365900961231694, + "learning_rate": 0.003, + "loss": 4.135, + "step": 5837 + }, + { + "epoch": 0.05838, + "grad_norm": 0.6835199033150741, + "learning_rate": 0.003, + "loss": 4.1369, + "step": 5838 + }, + { + "epoch": 0.05839, + "grad_norm": 0.6793630131258913, + "learning_rate": 0.003, + "loss": 4.1364, + "step": 5839 + }, + { + "epoch": 0.0584, + "grad_norm": 0.6942313892219114, + "learning_rate": 0.003, + "loss": 4.1543, + "step": 5840 + }, + { + "epoch": 0.05841, + "grad_norm": 0.8257321604395392, + "learning_rate": 0.003, + "loss": 4.1101, + "step": 5841 + }, + { + "epoch": 0.05842, + "grad_norm": 0.9932148908544386, + "learning_rate": 0.003, + "loss": 4.1315, + "step": 5842 + }, + { + "epoch": 0.05843, + "grad_norm": 1.060226393564342, + "learning_rate": 0.003, + "loss": 4.1376, + "step": 5843 + }, + { + "epoch": 0.05844, + "grad_norm": 0.7244723778858251, + "learning_rate": 0.003, + "loss": 4.1253, + "step": 5844 + }, + { + "epoch": 0.05845, + "grad_norm": 0.5661816774133244, + "learning_rate": 0.003, + "loss": 4.1206, + "step": 5845 + }, + { + "epoch": 0.05846, + "grad_norm": 0.6964991790304752, + "learning_rate": 0.003, + "loss": 4.1155, + "step": 5846 + }, + { + "epoch": 0.05847, + "grad_norm": 0.8137963768049039, + "learning_rate": 0.003, + "loss": 4.1448, + "step": 5847 + }, + { + "epoch": 0.05848, + "grad_norm": 0.9321439032747159, + "learning_rate": 0.003, + "loss": 4.1239, + "step": 5848 + }, + { + "epoch": 0.05849, + "grad_norm": 0.9690612770794925, + "learning_rate": 0.003, + "loss": 4.092, + "step": 5849 + }, + { + "epoch": 0.0585, + "grad_norm": 0.8070900144801142, + "learning_rate": 0.003, + "loss": 4.1203, + "step": 5850 + }, + { + "epoch": 0.05851, + "grad_norm": 0.7978154280770444, + "learning_rate": 0.003, + "loss": 4.1098, + "step": 5851 + }, + { + "epoch": 0.05852, + "grad_norm": 0.9108907816600593, + "learning_rate": 0.003, + "loss": 4.1228, + "step": 5852 + }, + { + "epoch": 0.05853, + "grad_norm": 0.8478337317220017, + "learning_rate": 0.003, + "loss": 4.1098, + "step": 5853 + }, + { + "epoch": 0.05854, + "grad_norm": 0.8702591663009297, + "learning_rate": 0.003, + "loss": 4.1814, + "step": 5854 + }, + { + "epoch": 0.05855, + "grad_norm": 0.7888882543910517, + "learning_rate": 0.003, + "loss": 4.1356, + "step": 5855 + }, + { + "epoch": 0.05856, + "grad_norm": 0.7469242966369579, + "learning_rate": 0.003, + "loss": 4.1279, + "step": 5856 + }, + { + "epoch": 0.05857, + "grad_norm": 0.706624864851467, + "learning_rate": 0.003, + "loss": 4.1372, + "step": 5857 + }, + { + "epoch": 0.05858, + "grad_norm": 0.7270168320884999, + "learning_rate": 0.003, + "loss": 4.1077, + "step": 5858 + }, + { + "epoch": 0.05859, + "grad_norm": 0.7529935867004306, + "learning_rate": 0.003, + "loss": 4.1324, + "step": 5859 + }, + { + "epoch": 0.0586, + "grad_norm": 0.6444427467245577, + "learning_rate": 0.003, + "loss": 4.1347, + "step": 5860 + }, + { + "epoch": 0.05861, + "grad_norm": 0.5606393662770783, + "learning_rate": 0.003, + "loss": 4.147, + "step": 5861 + }, + { + "epoch": 0.05862, + "grad_norm": 0.5657377156785917, + "learning_rate": 0.003, + "loss": 4.1369, + "step": 5862 + }, + { + "epoch": 0.05863, + "grad_norm": 0.5760395754129868, + "learning_rate": 0.003, + "loss": 4.1341, + "step": 5863 + }, + { + "epoch": 0.05864, + "grad_norm": 0.6345646643457323, + "learning_rate": 0.003, + "loss": 4.1103, + "step": 5864 + }, + { + "epoch": 0.05865, + "grad_norm": 0.6894044067934254, + "learning_rate": 0.003, + "loss": 4.1062, + "step": 5865 + }, + { + "epoch": 0.05866, + "grad_norm": 0.783353091605477, + "learning_rate": 0.003, + "loss": 4.1302, + "step": 5866 + }, + { + "epoch": 0.05867, + "grad_norm": 0.8917647666324704, + "learning_rate": 0.003, + "loss": 4.1147, + "step": 5867 + }, + { + "epoch": 0.05868, + "grad_norm": 1.0201499997700771, + "learning_rate": 0.003, + "loss": 4.1059, + "step": 5868 + }, + { + "epoch": 0.05869, + "grad_norm": 0.9506659421163607, + "learning_rate": 0.003, + "loss": 4.1166, + "step": 5869 + }, + { + "epoch": 0.0587, + "grad_norm": 0.7664369973041759, + "learning_rate": 0.003, + "loss": 4.1105, + "step": 5870 + }, + { + "epoch": 0.05871, + "grad_norm": 0.7087081834503488, + "learning_rate": 0.003, + "loss": 4.1128, + "step": 5871 + }, + { + "epoch": 0.05872, + "grad_norm": 0.6953730352638628, + "learning_rate": 0.003, + "loss": 4.1093, + "step": 5872 + }, + { + "epoch": 0.05873, + "grad_norm": 0.671656467983226, + "learning_rate": 0.003, + "loss": 4.1224, + "step": 5873 + }, + { + "epoch": 0.05874, + "grad_norm": 0.6955207948522178, + "learning_rate": 0.003, + "loss": 4.1132, + "step": 5874 + }, + { + "epoch": 0.05875, + "grad_norm": 0.6268815574500972, + "learning_rate": 0.003, + "loss": 4.1228, + "step": 5875 + }, + { + "epoch": 0.05876, + "grad_norm": 0.6389310541639931, + "learning_rate": 0.003, + "loss": 4.1187, + "step": 5876 + }, + { + "epoch": 0.05877, + "grad_norm": 0.6217227109072572, + "learning_rate": 0.003, + "loss": 4.1241, + "step": 5877 + }, + { + "epoch": 0.05878, + "grad_norm": 0.7692902630082031, + "learning_rate": 0.003, + "loss": 4.1385, + "step": 5878 + }, + { + "epoch": 0.05879, + "grad_norm": 0.7440526157383079, + "learning_rate": 0.003, + "loss": 4.1519, + "step": 5879 + }, + { + "epoch": 0.0588, + "grad_norm": 0.8420901226461514, + "learning_rate": 0.003, + "loss": 4.1187, + "step": 5880 + }, + { + "epoch": 0.05881, + "grad_norm": 0.781449431655522, + "learning_rate": 0.003, + "loss": 4.1208, + "step": 5881 + }, + { + "epoch": 0.05882, + "grad_norm": 0.7627663058529272, + "learning_rate": 0.003, + "loss": 4.136, + "step": 5882 + }, + { + "epoch": 0.05883, + "grad_norm": 0.8235508005883959, + "learning_rate": 0.003, + "loss": 4.1376, + "step": 5883 + }, + { + "epoch": 0.05884, + "grad_norm": 0.9282232426526248, + "learning_rate": 0.003, + "loss": 4.1275, + "step": 5884 + }, + { + "epoch": 0.05885, + "grad_norm": 0.8846373141454603, + "learning_rate": 0.003, + "loss": 4.1403, + "step": 5885 + }, + { + "epoch": 0.05886, + "grad_norm": 1.1084709963523778, + "learning_rate": 0.003, + "loss": 4.1454, + "step": 5886 + }, + { + "epoch": 0.05887, + "grad_norm": 0.9597419334563931, + "learning_rate": 0.003, + "loss": 4.1302, + "step": 5887 + }, + { + "epoch": 0.05888, + "grad_norm": 1.1484270558519247, + "learning_rate": 0.003, + "loss": 4.0923, + "step": 5888 + }, + { + "epoch": 0.05889, + "grad_norm": 0.7934646401717872, + "learning_rate": 0.003, + "loss": 4.1205, + "step": 5889 + }, + { + "epoch": 0.0589, + "grad_norm": 0.7136730779704626, + "learning_rate": 0.003, + "loss": 4.1228, + "step": 5890 + }, + { + "epoch": 0.05891, + "grad_norm": 0.725739807105299, + "learning_rate": 0.003, + "loss": 4.1214, + "step": 5891 + }, + { + "epoch": 0.05892, + "grad_norm": 0.7256800823290586, + "learning_rate": 0.003, + "loss": 4.1515, + "step": 5892 + }, + { + "epoch": 0.05893, + "grad_norm": 0.7692017239614891, + "learning_rate": 0.003, + "loss": 4.1265, + "step": 5893 + }, + { + "epoch": 0.05894, + "grad_norm": 0.7387008350548316, + "learning_rate": 0.003, + "loss": 4.1008, + "step": 5894 + }, + { + "epoch": 0.05895, + "grad_norm": 0.6999771063166865, + "learning_rate": 0.003, + "loss": 4.1099, + "step": 5895 + }, + { + "epoch": 0.05896, + "grad_norm": 0.6586617304026913, + "learning_rate": 0.003, + "loss": 4.1334, + "step": 5896 + }, + { + "epoch": 0.05897, + "grad_norm": 0.7731203514227012, + "learning_rate": 0.003, + "loss": 4.1445, + "step": 5897 + }, + { + "epoch": 0.05898, + "grad_norm": 0.8305320510139119, + "learning_rate": 0.003, + "loss": 4.1428, + "step": 5898 + }, + { + "epoch": 0.05899, + "grad_norm": 0.7742225479585695, + "learning_rate": 0.003, + "loss": 4.1025, + "step": 5899 + }, + { + "epoch": 0.059, + "grad_norm": 0.9048055534854499, + "learning_rate": 0.003, + "loss": 4.1318, + "step": 5900 + }, + { + "epoch": 0.05901, + "grad_norm": 0.9868685055861205, + "learning_rate": 0.003, + "loss": 4.1192, + "step": 5901 + }, + { + "epoch": 0.05902, + "grad_norm": 1.0603964019326821, + "learning_rate": 0.003, + "loss": 4.1368, + "step": 5902 + }, + { + "epoch": 0.05903, + "grad_norm": 1.0745850428190593, + "learning_rate": 0.003, + "loss": 4.1428, + "step": 5903 + }, + { + "epoch": 0.05904, + "grad_norm": 0.8629637800138508, + "learning_rate": 0.003, + "loss": 4.1233, + "step": 5904 + }, + { + "epoch": 0.05905, + "grad_norm": 0.746934964755465, + "learning_rate": 0.003, + "loss": 4.1213, + "step": 5905 + }, + { + "epoch": 0.05906, + "grad_norm": 0.7381867732393038, + "learning_rate": 0.003, + "loss": 4.1215, + "step": 5906 + }, + { + "epoch": 0.05907, + "grad_norm": 0.8572362096896168, + "learning_rate": 0.003, + "loss": 4.1132, + "step": 5907 + }, + { + "epoch": 0.05908, + "grad_norm": 1.0322237476040848, + "learning_rate": 0.003, + "loss": 4.1436, + "step": 5908 + }, + { + "epoch": 0.05909, + "grad_norm": 0.8902648134987068, + "learning_rate": 0.003, + "loss": 4.1457, + "step": 5909 + }, + { + "epoch": 0.0591, + "grad_norm": 0.773959915980087, + "learning_rate": 0.003, + "loss": 4.1168, + "step": 5910 + }, + { + "epoch": 0.05911, + "grad_norm": 0.8192300954731131, + "learning_rate": 0.003, + "loss": 4.1067, + "step": 5911 + }, + { + "epoch": 0.05912, + "grad_norm": 0.8546496512225619, + "learning_rate": 0.003, + "loss": 4.1508, + "step": 5912 + }, + { + "epoch": 0.05913, + "grad_norm": 0.7554616555250914, + "learning_rate": 0.003, + "loss": 4.1702, + "step": 5913 + }, + { + "epoch": 0.05914, + "grad_norm": 0.7048163682769332, + "learning_rate": 0.003, + "loss": 4.1399, + "step": 5914 + }, + { + "epoch": 0.05915, + "grad_norm": 0.7670702397330968, + "learning_rate": 0.003, + "loss": 4.1385, + "step": 5915 + }, + { + "epoch": 0.05916, + "grad_norm": 0.7678316112541398, + "learning_rate": 0.003, + "loss": 4.1246, + "step": 5916 + }, + { + "epoch": 0.05917, + "grad_norm": 0.6726151262347494, + "learning_rate": 0.003, + "loss": 4.1064, + "step": 5917 + }, + { + "epoch": 0.05918, + "grad_norm": 0.6440357536146859, + "learning_rate": 0.003, + "loss": 4.1188, + "step": 5918 + }, + { + "epoch": 0.05919, + "grad_norm": 0.6990263177794974, + "learning_rate": 0.003, + "loss": 4.1158, + "step": 5919 + }, + { + "epoch": 0.0592, + "grad_norm": 0.7179007290375548, + "learning_rate": 0.003, + "loss": 4.0966, + "step": 5920 + }, + { + "epoch": 0.05921, + "grad_norm": 0.7012296689409627, + "learning_rate": 0.003, + "loss": 4.1342, + "step": 5921 + }, + { + "epoch": 0.05922, + "grad_norm": 0.7174027660375633, + "learning_rate": 0.003, + "loss": 4.126, + "step": 5922 + }, + { + "epoch": 0.05923, + "grad_norm": 0.7806176175934582, + "learning_rate": 0.003, + "loss": 4.0927, + "step": 5923 + }, + { + "epoch": 0.05924, + "grad_norm": 0.8832244353079346, + "learning_rate": 0.003, + "loss": 4.1249, + "step": 5924 + }, + { + "epoch": 0.05925, + "grad_norm": 0.8531781306643916, + "learning_rate": 0.003, + "loss": 4.1126, + "step": 5925 + }, + { + "epoch": 0.05926, + "grad_norm": 0.6880234314584872, + "learning_rate": 0.003, + "loss": 4.1153, + "step": 5926 + }, + { + "epoch": 0.05927, + "grad_norm": 0.5242997783745522, + "learning_rate": 0.003, + "loss": 4.0932, + "step": 5927 + }, + { + "epoch": 0.05928, + "grad_norm": 0.7118263988410835, + "learning_rate": 0.003, + "loss": 4.1155, + "step": 5928 + }, + { + "epoch": 0.05929, + "grad_norm": 0.9051922436752257, + "learning_rate": 0.003, + "loss": 4.1295, + "step": 5929 + }, + { + "epoch": 0.0593, + "grad_norm": 0.9625875611083993, + "learning_rate": 0.003, + "loss": 4.1346, + "step": 5930 + }, + { + "epoch": 0.05931, + "grad_norm": 0.8328206700541043, + "learning_rate": 0.003, + "loss": 4.1154, + "step": 5931 + }, + { + "epoch": 0.05932, + "grad_norm": 0.6522428602922113, + "learning_rate": 0.003, + "loss": 4.1378, + "step": 5932 + }, + { + "epoch": 0.05933, + "grad_norm": 0.8202258652975196, + "learning_rate": 0.003, + "loss": 4.1083, + "step": 5933 + }, + { + "epoch": 0.05934, + "grad_norm": 0.945676463256145, + "learning_rate": 0.003, + "loss": 4.1538, + "step": 5934 + }, + { + "epoch": 0.05935, + "grad_norm": 0.9491618675679084, + "learning_rate": 0.003, + "loss": 4.1398, + "step": 5935 + }, + { + "epoch": 0.05936, + "grad_norm": 0.7346756788314402, + "learning_rate": 0.003, + "loss": 4.1319, + "step": 5936 + }, + { + "epoch": 0.05937, + "grad_norm": 0.6849843625907307, + "learning_rate": 0.003, + "loss": 4.1291, + "step": 5937 + }, + { + "epoch": 0.05938, + "grad_norm": 0.842493843061451, + "learning_rate": 0.003, + "loss": 4.1086, + "step": 5938 + }, + { + "epoch": 0.05939, + "grad_norm": 0.9802131731998531, + "learning_rate": 0.003, + "loss": 4.1259, + "step": 5939 + }, + { + "epoch": 0.0594, + "grad_norm": 1.0740955279537077, + "learning_rate": 0.003, + "loss": 4.1484, + "step": 5940 + }, + { + "epoch": 0.05941, + "grad_norm": 0.80664432417099, + "learning_rate": 0.003, + "loss": 4.1417, + "step": 5941 + }, + { + "epoch": 0.05942, + "grad_norm": 0.6673463401001458, + "learning_rate": 0.003, + "loss": 4.127, + "step": 5942 + }, + { + "epoch": 0.05943, + "grad_norm": 0.7136729446036693, + "learning_rate": 0.003, + "loss": 4.1155, + "step": 5943 + }, + { + "epoch": 0.05944, + "grad_norm": 0.7748772994265909, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 5944 + }, + { + "epoch": 0.05945, + "grad_norm": 0.8319748600017105, + "learning_rate": 0.003, + "loss": 4.1153, + "step": 5945 + }, + { + "epoch": 0.05946, + "grad_norm": 0.7912457127582416, + "learning_rate": 0.003, + "loss": 4.111, + "step": 5946 + }, + { + "epoch": 0.05947, + "grad_norm": 0.7794353450592082, + "learning_rate": 0.003, + "loss": 4.105, + "step": 5947 + }, + { + "epoch": 0.05948, + "grad_norm": 0.8118650335500223, + "learning_rate": 0.003, + "loss": 4.1199, + "step": 5948 + }, + { + "epoch": 0.05949, + "grad_norm": 0.8550107923144452, + "learning_rate": 0.003, + "loss": 4.1159, + "step": 5949 + }, + { + "epoch": 0.0595, + "grad_norm": 0.8915593726429895, + "learning_rate": 0.003, + "loss": 4.1317, + "step": 5950 + }, + { + "epoch": 0.05951, + "grad_norm": 0.8945628866915003, + "learning_rate": 0.003, + "loss": 4.127, + "step": 5951 + }, + { + "epoch": 0.05952, + "grad_norm": 1.0089347086265024, + "learning_rate": 0.003, + "loss": 4.1428, + "step": 5952 + }, + { + "epoch": 0.05953, + "grad_norm": 1.0649366317902482, + "learning_rate": 0.003, + "loss": 4.1271, + "step": 5953 + }, + { + "epoch": 0.05954, + "grad_norm": 1.0265306010895543, + "learning_rate": 0.003, + "loss": 4.1486, + "step": 5954 + }, + { + "epoch": 0.05955, + "grad_norm": 1.0053891091737848, + "learning_rate": 0.003, + "loss": 4.1313, + "step": 5955 + }, + { + "epoch": 0.05956, + "grad_norm": 0.9418944733574852, + "learning_rate": 0.003, + "loss": 4.1144, + "step": 5956 + }, + { + "epoch": 0.05957, + "grad_norm": 0.8074037334150802, + "learning_rate": 0.003, + "loss": 4.1176, + "step": 5957 + }, + { + "epoch": 0.05958, + "grad_norm": 0.835853516513737, + "learning_rate": 0.003, + "loss": 4.1292, + "step": 5958 + }, + { + "epoch": 0.05959, + "grad_norm": 0.9524075368140398, + "learning_rate": 0.003, + "loss": 4.1328, + "step": 5959 + }, + { + "epoch": 0.0596, + "grad_norm": 0.9443691821790957, + "learning_rate": 0.003, + "loss": 4.1497, + "step": 5960 + }, + { + "epoch": 0.05961, + "grad_norm": 0.8502001962303277, + "learning_rate": 0.003, + "loss": 4.1367, + "step": 5961 + }, + { + "epoch": 0.05962, + "grad_norm": 0.7110403654732431, + "learning_rate": 0.003, + "loss": 4.1236, + "step": 5962 + }, + { + "epoch": 0.05963, + "grad_norm": 0.6983877872125744, + "learning_rate": 0.003, + "loss": 4.1316, + "step": 5963 + }, + { + "epoch": 0.05964, + "grad_norm": 0.6706508091828537, + "learning_rate": 0.003, + "loss": 4.124, + "step": 5964 + }, + { + "epoch": 0.05965, + "grad_norm": 0.5516979364941672, + "learning_rate": 0.003, + "loss": 4.1158, + "step": 5965 + }, + { + "epoch": 0.05966, + "grad_norm": 0.5599182409877457, + "learning_rate": 0.003, + "loss": 4.1312, + "step": 5966 + }, + { + "epoch": 0.05967, + "grad_norm": 0.5140228003501242, + "learning_rate": 0.003, + "loss": 4.1081, + "step": 5967 + }, + { + "epoch": 0.05968, + "grad_norm": 0.5043516900053563, + "learning_rate": 0.003, + "loss": 4.1064, + "step": 5968 + }, + { + "epoch": 0.05969, + "grad_norm": 0.4781424578723309, + "learning_rate": 0.003, + "loss": 4.0957, + "step": 5969 + }, + { + "epoch": 0.0597, + "grad_norm": 0.49499059228808473, + "learning_rate": 0.003, + "loss": 4.102, + "step": 5970 + }, + { + "epoch": 0.05971, + "grad_norm": 0.5302688602898687, + "learning_rate": 0.003, + "loss": 4.1204, + "step": 5971 + }, + { + "epoch": 0.05972, + "grad_norm": 0.5628921179090366, + "learning_rate": 0.003, + "loss": 4.1306, + "step": 5972 + }, + { + "epoch": 0.05973, + "grad_norm": 0.6729538698391679, + "learning_rate": 0.003, + "loss": 4.0945, + "step": 5973 + }, + { + "epoch": 0.05974, + "grad_norm": 0.8325991069119943, + "learning_rate": 0.003, + "loss": 4.0936, + "step": 5974 + }, + { + "epoch": 0.05975, + "grad_norm": 1.0067237957827262, + "learning_rate": 0.003, + "loss": 4.1348, + "step": 5975 + }, + { + "epoch": 0.05976, + "grad_norm": 1.2218569863697355, + "learning_rate": 0.003, + "loss": 4.1282, + "step": 5976 + }, + { + "epoch": 0.05977, + "grad_norm": 0.6120282975916178, + "learning_rate": 0.003, + "loss": 4.1255, + "step": 5977 + }, + { + "epoch": 0.05978, + "grad_norm": 0.634934658722155, + "learning_rate": 0.003, + "loss": 4.1114, + "step": 5978 + }, + { + "epoch": 0.05979, + "grad_norm": 0.795528878435463, + "learning_rate": 0.003, + "loss": 4.1217, + "step": 5979 + }, + { + "epoch": 0.0598, + "grad_norm": 0.7850331087825336, + "learning_rate": 0.003, + "loss": 4.0976, + "step": 5980 + }, + { + "epoch": 0.05981, + "grad_norm": 0.7268016197563338, + "learning_rate": 0.003, + "loss": 4.1002, + "step": 5981 + }, + { + "epoch": 0.05982, + "grad_norm": 0.6606189209405257, + "learning_rate": 0.003, + "loss": 4.1138, + "step": 5982 + }, + { + "epoch": 0.05983, + "grad_norm": 0.7268589111503444, + "learning_rate": 0.003, + "loss": 4.1294, + "step": 5983 + }, + { + "epoch": 0.05984, + "grad_norm": 0.7708148438016759, + "learning_rate": 0.003, + "loss": 4.0984, + "step": 5984 + }, + { + "epoch": 0.05985, + "grad_norm": 0.7846506991954134, + "learning_rate": 0.003, + "loss": 4.134, + "step": 5985 + }, + { + "epoch": 0.05986, + "grad_norm": 0.7546183129518631, + "learning_rate": 0.003, + "loss": 4.1308, + "step": 5986 + }, + { + "epoch": 0.05987, + "grad_norm": 0.8496680958314422, + "learning_rate": 0.003, + "loss": 4.105, + "step": 5987 + }, + { + "epoch": 0.05988, + "grad_norm": 0.9751701746085983, + "learning_rate": 0.003, + "loss": 4.121, + "step": 5988 + }, + { + "epoch": 0.05989, + "grad_norm": 1.0265361397551773, + "learning_rate": 0.003, + "loss": 4.1218, + "step": 5989 + }, + { + "epoch": 0.0599, + "grad_norm": 1.032703400806406, + "learning_rate": 0.003, + "loss": 4.1273, + "step": 5990 + }, + { + "epoch": 0.05991, + "grad_norm": 0.9301589844396891, + "learning_rate": 0.003, + "loss": 4.096, + "step": 5991 + }, + { + "epoch": 0.05992, + "grad_norm": 0.8529001190170848, + "learning_rate": 0.003, + "loss": 4.1232, + "step": 5992 + }, + { + "epoch": 0.05993, + "grad_norm": 0.7232314574469758, + "learning_rate": 0.003, + "loss": 4.1272, + "step": 5993 + }, + { + "epoch": 0.05994, + "grad_norm": 0.6055257270988292, + "learning_rate": 0.003, + "loss": 4.1071, + "step": 5994 + }, + { + "epoch": 0.05995, + "grad_norm": 0.6461271491413968, + "learning_rate": 0.003, + "loss": 4.1133, + "step": 5995 + }, + { + "epoch": 0.05996, + "grad_norm": 0.6721002555843104, + "learning_rate": 0.003, + "loss": 4.1485, + "step": 5996 + }, + { + "epoch": 0.05997, + "grad_norm": 0.7240739565812643, + "learning_rate": 0.003, + "loss": 4.0945, + "step": 5997 + }, + { + "epoch": 0.05998, + "grad_norm": 0.6950086390271106, + "learning_rate": 0.003, + "loss": 4.1417, + "step": 5998 + }, + { + "epoch": 0.05999, + "grad_norm": 0.750185481402966, + "learning_rate": 0.003, + "loss": 4.1105, + "step": 5999 + }, + { + "epoch": 0.06, + "grad_norm": 0.8158906083982681, + "learning_rate": 0.003, + "loss": 4.1242, + "step": 6000 + }, + { + "epoch": 0.06001, + "grad_norm": 0.7786341857744299, + "learning_rate": 0.003, + "loss": 4.1079, + "step": 6001 + }, + { + "epoch": 0.06002, + "grad_norm": 0.6362829691059452, + "learning_rate": 0.003, + "loss": 4.1, + "step": 6002 + }, + { + "epoch": 0.06003, + "grad_norm": 0.6439798777608323, + "learning_rate": 0.003, + "loss": 4.1191, + "step": 6003 + }, + { + "epoch": 0.06004, + "grad_norm": 0.7447283430547924, + "learning_rate": 0.003, + "loss": 4.1004, + "step": 6004 + }, + { + "epoch": 0.06005, + "grad_norm": 0.8074281820106757, + "learning_rate": 0.003, + "loss": 4.1189, + "step": 6005 + }, + { + "epoch": 0.06006, + "grad_norm": 0.8310801813091088, + "learning_rate": 0.003, + "loss": 4.103, + "step": 6006 + }, + { + "epoch": 0.06007, + "grad_norm": 0.739959036143615, + "learning_rate": 0.003, + "loss": 4.1522, + "step": 6007 + }, + { + "epoch": 0.06008, + "grad_norm": 0.7512598056369182, + "learning_rate": 0.003, + "loss": 4.1108, + "step": 6008 + }, + { + "epoch": 0.06009, + "grad_norm": 0.8453751856138448, + "learning_rate": 0.003, + "loss": 4.1186, + "step": 6009 + }, + { + "epoch": 0.0601, + "grad_norm": 0.9834836783342701, + "learning_rate": 0.003, + "loss": 4.1254, + "step": 6010 + }, + { + "epoch": 0.06011, + "grad_norm": 1.2294686050706798, + "learning_rate": 0.003, + "loss": 4.1418, + "step": 6011 + }, + { + "epoch": 0.06012, + "grad_norm": 0.6713324153136399, + "learning_rate": 0.003, + "loss": 4.1324, + "step": 6012 + }, + { + "epoch": 0.06013, + "grad_norm": 0.7264483540409177, + "learning_rate": 0.003, + "loss": 4.1253, + "step": 6013 + }, + { + "epoch": 0.06014, + "grad_norm": 0.8963286802156893, + "learning_rate": 0.003, + "loss": 4.1183, + "step": 6014 + }, + { + "epoch": 0.06015, + "grad_norm": 1.0516872025977433, + "learning_rate": 0.003, + "loss": 4.137, + "step": 6015 + }, + { + "epoch": 0.06016, + "grad_norm": 0.9867378972073549, + "learning_rate": 0.003, + "loss": 4.1417, + "step": 6016 + }, + { + "epoch": 0.06017, + "grad_norm": 0.8189207241122054, + "learning_rate": 0.003, + "loss": 4.1409, + "step": 6017 + }, + { + "epoch": 0.06018, + "grad_norm": 0.6248389222598227, + "learning_rate": 0.003, + "loss": 4.1515, + "step": 6018 + }, + { + "epoch": 0.06019, + "grad_norm": 0.6382727015073211, + "learning_rate": 0.003, + "loss": 4.1185, + "step": 6019 + }, + { + "epoch": 0.0602, + "grad_norm": 0.5942681644648503, + "learning_rate": 0.003, + "loss": 4.1121, + "step": 6020 + }, + { + "epoch": 0.06021, + "grad_norm": 0.6307301681376707, + "learning_rate": 0.003, + "loss": 4.1322, + "step": 6021 + }, + { + "epoch": 0.06022, + "grad_norm": 0.6750537233371731, + "learning_rate": 0.003, + "loss": 4.1374, + "step": 6022 + }, + { + "epoch": 0.06023, + "grad_norm": 0.7128799339128652, + "learning_rate": 0.003, + "loss": 4.1231, + "step": 6023 + }, + { + "epoch": 0.06024, + "grad_norm": 0.7520500789879113, + "learning_rate": 0.003, + "loss": 4.0988, + "step": 6024 + }, + { + "epoch": 0.06025, + "grad_norm": 0.7805119674993733, + "learning_rate": 0.003, + "loss": 4.156, + "step": 6025 + }, + { + "epoch": 0.06026, + "grad_norm": 0.7499035678190326, + "learning_rate": 0.003, + "loss": 4.1246, + "step": 6026 + }, + { + "epoch": 0.06027, + "grad_norm": 0.7664750877261561, + "learning_rate": 0.003, + "loss": 4.1167, + "step": 6027 + }, + { + "epoch": 0.06028, + "grad_norm": 0.7776094623285253, + "learning_rate": 0.003, + "loss": 4.1418, + "step": 6028 + }, + { + "epoch": 0.06029, + "grad_norm": 0.8761047706417141, + "learning_rate": 0.003, + "loss": 4.1355, + "step": 6029 + }, + { + "epoch": 0.0603, + "grad_norm": 0.9104716332204089, + "learning_rate": 0.003, + "loss": 4.1337, + "step": 6030 + }, + { + "epoch": 0.06031, + "grad_norm": 0.8922902846451947, + "learning_rate": 0.003, + "loss": 4.127, + "step": 6031 + }, + { + "epoch": 0.06032, + "grad_norm": 0.9770316982513426, + "learning_rate": 0.003, + "loss": 4.1335, + "step": 6032 + }, + { + "epoch": 0.06033, + "grad_norm": 0.9457538473651317, + "learning_rate": 0.003, + "loss": 4.1363, + "step": 6033 + }, + { + "epoch": 0.06034, + "grad_norm": 0.9217776085904755, + "learning_rate": 0.003, + "loss": 4.151, + "step": 6034 + }, + { + "epoch": 0.06035, + "grad_norm": 0.9280484794744475, + "learning_rate": 0.003, + "loss": 4.1415, + "step": 6035 + }, + { + "epoch": 0.06036, + "grad_norm": 1.0009116543043906, + "learning_rate": 0.003, + "loss": 4.0963, + "step": 6036 + }, + { + "epoch": 0.06037, + "grad_norm": 1.100896351699259, + "learning_rate": 0.003, + "loss": 4.1413, + "step": 6037 + }, + { + "epoch": 0.06038, + "grad_norm": 0.751886944643684, + "learning_rate": 0.003, + "loss": 4.1173, + "step": 6038 + }, + { + "epoch": 0.06039, + "grad_norm": 0.6560741836996629, + "learning_rate": 0.003, + "loss": 4.0955, + "step": 6039 + }, + { + "epoch": 0.0604, + "grad_norm": 0.6307042035546973, + "learning_rate": 0.003, + "loss": 4.1208, + "step": 6040 + }, + { + "epoch": 0.06041, + "grad_norm": 0.6583077177213943, + "learning_rate": 0.003, + "loss": 4.1357, + "step": 6041 + }, + { + "epoch": 0.06042, + "grad_norm": 0.5980033750895085, + "learning_rate": 0.003, + "loss": 4.1128, + "step": 6042 + }, + { + "epoch": 0.06043, + "grad_norm": 0.6032908893943834, + "learning_rate": 0.003, + "loss": 4.1176, + "step": 6043 + }, + { + "epoch": 0.06044, + "grad_norm": 0.6209590746215262, + "learning_rate": 0.003, + "loss": 4.1185, + "step": 6044 + }, + { + "epoch": 0.06045, + "grad_norm": 0.6136074895659309, + "learning_rate": 0.003, + "loss": 4.1182, + "step": 6045 + }, + { + "epoch": 0.06046, + "grad_norm": 0.6915063121187296, + "learning_rate": 0.003, + "loss": 4.1221, + "step": 6046 + }, + { + "epoch": 0.06047, + "grad_norm": 0.7812065657084274, + "learning_rate": 0.003, + "loss": 4.103, + "step": 6047 + }, + { + "epoch": 0.06048, + "grad_norm": 0.784591842396142, + "learning_rate": 0.003, + "loss": 4.1017, + "step": 6048 + }, + { + "epoch": 0.06049, + "grad_norm": 0.856638395736723, + "learning_rate": 0.003, + "loss": 4.1, + "step": 6049 + }, + { + "epoch": 0.0605, + "grad_norm": 1.1075444631976008, + "learning_rate": 0.003, + "loss": 4.1364, + "step": 6050 + }, + { + "epoch": 0.06051, + "grad_norm": 1.0022518924058987, + "learning_rate": 0.003, + "loss": 4.1445, + "step": 6051 + }, + { + "epoch": 0.06052, + "grad_norm": 0.9359706571597523, + "learning_rate": 0.003, + "loss": 4.1408, + "step": 6052 + }, + { + "epoch": 0.06053, + "grad_norm": 0.8709162398571225, + "learning_rate": 0.003, + "loss": 4.1005, + "step": 6053 + }, + { + "epoch": 0.06054, + "grad_norm": 0.9149227979278415, + "learning_rate": 0.003, + "loss": 4.138, + "step": 6054 + }, + { + "epoch": 0.06055, + "grad_norm": 1.0346925153764615, + "learning_rate": 0.003, + "loss": 4.1351, + "step": 6055 + }, + { + "epoch": 0.06056, + "grad_norm": 0.9282475689671914, + "learning_rate": 0.003, + "loss": 4.1269, + "step": 6056 + }, + { + "epoch": 0.06057, + "grad_norm": 0.9497057705026418, + "learning_rate": 0.003, + "loss": 4.1123, + "step": 6057 + }, + { + "epoch": 0.06058, + "grad_norm": 1.003018583474644, + "learning_rate": 0.003, + "loss": 4.1316, + "step": 6058 + }, + { + "epoch": 0.06059, + "grad_norm": 0.9574516546287778, + "learning_rate": 0.003, + "loss": 4.1447, + "step": 6059 + }, + { + "epoch": 0.0606, + "grad_norm": 0.9267456613771629, + "learning_rate": 0.003, + "loss": 4.1427, + "step": 6060 + }, + { + "epoch": 0.06061, + "grad_norm": 1.0318438116932864, + "learning_rate": 0.003, + "loss": 4.1518, + "step": 6061 + }, + { + "epoch": 0.06062, + "grad_norm": 1.0400522917116977, + "learning_rate": 0.003, + "loss": 4.1579, + "step": 6062 + }, + { + "epoch": 0.06063, + "grad_norm": 1.0497641878303774, + "learning_rate": 0.003, + "loss": 4.1133, + "step": 6063 + }, + { + "epoch": 0.06064, + "grad_norm": 1.0662498014953086, + "learning_rate": 0.003, + "loss": 4.1748, + "step": 6064 + }, + { + "epoch": 0.06065, + "grad_norm": 0.8198498372995232, + "learning_rate": 0.003, + "loss": 4.1373, + "step": 6065 + }, + { + "epoch": 0.06066, + "grad_norm": 0.7109049478395263, + "learning_rate": 0.003, + "loss": 4.166, + "step": 6066 + }, + { + "epoch": 0.06067, + "grad_norm": 0.6804945226918797, + "learning_rate": 0.003, + "loss": 4.1385, + "step": 6067 + }, + { + "epoch": 0.06068, + "grad_norm": 0.6372476248357833, + "learning_rate": 0.003, + "loss": 4.1206, + "step": 6068 + }, + { + "epoch": 0.06069, + "grad_norm": 0.5605214633243844, + "learning_rate": 0.003, + "loss": 4.1259, + "step": 6069 + }, + { + "epoch": 0.0607, + "grad_norm": 0.5452568213138966, + "learning_rate": 0.003, + "loss": 4.1273, + "step": 6070 + }, + { + "epoch": 0.06071, + "grad_norm": 0.6180224930197478, + "learning_rate": 0.003, + "loss": 4.1005, + "step": 6071 + }, + { + "epoch": 0.06072, + "grad_norm": 0.7813523705166946, + "learning_rate": 0.003, + "loss": 4.1091, + "step": 6072 + }, + { + "epoch": 0.06073, + "grad_norm": 0.9736890155608084, + "learning_rate": 0.003, + "loss": 4.1425, + "step": 6073 + }, + { + "epoch": 0.06074, + "grad_norm": 1.0327347386785892, + "learning_rate": 0.003, + "loss": 4.1275, + "step": 6074 + }, + { + "epoch": 0.06075, + "grad_norm": 0.7472307019465172, + "learning_rate": 0.003, + "loss": 4.1415, + "step": 6075 + }, + { + "epoch": 0.06076, + "grad_norm": 0.6670271053470518, + "learning_rate": 0.003, + "loss": 4.1251, + "step": 6076 + }, + { + "epoch": 0.06077, + "grad_norm": 0.7254836202543244, + "learning_rate": 0.003, + "loss": 4.1225, + "step": 6077 + }, + { + "epoch": 0.06078, + "grad_norm": 0.6208465711650167, + "learning_rate": 0.003, + "loss": 4.1262, + "step": 6078 + }, + { + "epoch": 0.06079, + "grad_norm": 0.7113936555649242, + "learning_rate": 0.003, + "loss": 4.1217, + "step": 6079 + }, + { + "epoch": 0.0608, + "grad_norm": 0.689258626457068, + "learning_rate": 0.003, + "loss": 4.119, + "step": 6080 + }, + { + "epoch": 0.06081, + "grad_norm": 0.6551455177926783, + "learning_rate": 0.003, + "loss": 4.0983, + "step": 6081 + }, + { + "epoch": 0.06082, + "grad_norm": 0.6542633333810457, + "learning_rate": 0.003, + "loss": 4.1454, + "step": 6082 + }, + { + "epoch": 0.06083, + "grad_norm": 0.6010838532362384, + "learning_rate": 0.003, + "loss": 4.1273, + "step": 6083 + }, + { + "epoch": 0.06084, + "grad_norm": 0.5910563656180051, + "learning_rate": 0.003, + "loss": 4.0919, + "step": 6084 + }, + { + "epoch": 0.06085, + "grad_norm": 0.6028680880478657, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 6085 + }, + { + "epoch": 0.06086, + "grad_norm": 0.5204888285801396, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 6086 + }, + { + "epoch": 0.06087, + "grad_norm": 0.5900462834410196, + "learning_rate": 0.003, + "loss": 4.1224, + "step": 6087 + }, + { + "epoch": 0.06088, + "grad_norm": 0.6959450877948858, + "learning_rate": 0.003, + "loss": 4.1019, + "step": 6088 + }, + { + "epoch": 0.06089, + "grad_norm": 0.6784516411313329, + "learning_rate": 0.003, + "loss": 4.1012, + "step": 6089 + }, + { + "epoch": 0.0609, + "grad_norm": 0.7035758936099877, + "learning_rate": 0.003, + "loss": 4.14, + "step": 6090 + }, + { + "epoch": 0.06091, + "grad_norm": 0.8330650414758795, + "learning_rate": 0.003, + "loss": 4.1164, + "step": 6091 + }, + { + "epoch": 0.06092, + "grad_norm": 0.8930098103439321, + "learning_rate": 0.003, + "loss": 4.0994, + "step": 6092 + }, + { + "epoch": 0.06093, + "grad_norm": 1.0224308602932068, + "learning_rate": 0.003, + "loss": 4.115, + "step": 6093 + }, + { + "epoch": 0.06094, + "grad_norm": 0.9894448173843531, + "learning_rate": 0.003, + "loss": 4.1193, + "step": 6094 + }, + { + "epoch": 0.06095, + "grad_norm": 0.8810812235713495, + "learning_rate": 0.003, + "loss": 4.1225, + "step": 6095 + }, + { + "epoch": 0.06096, + "grad_norm": 0.6354430317367544, + "learning_rate": 0.003, + "loss": 4.1062, + "step": 6096 + }, + { + "epoch": 0.06097, + "grad_norm": 0.6376701446414055, + "learning_rate": 0.003, + "loss": 4.1213, + "step": 6097 + }, + { + "epoch": 0.06098, + "grad_norm": 0.8448719370273449, + "learning_rate": 0.003, + "loss": 4.0962, + "step": 6098 + }, + { + "epoch": 0.06099, + "grad_norm": 0.969987418580439, + "learning_rate": 0.003, + "loss": 4.1229, + "step": 6099 + }, + { + "epoch": 0.061, + "grad_norm": 1.0255182095149356, + "learning_rate": 0.003, + "loss": 4.1198, + "step": 6100 + }, + { + "epoch": 0.06101, + "grad_norm": 0.8139563380818391, + "learning_rate": 0.003, + "loss": 4.1465, + "step": 6101 + }, + { + "epoch": 0.06102, + "grad_norm": 0.6862499005656783, + "learning_rate": 0.003, + "loss": 4.1248, + "step": 6102 + }, + { + "epoch": 0.06103, + "grad_norm": 0.7284494857382159, + "learning_rate": 0.003, + "loss": 4.1373, + "step": 6103 + }, + { + "epoch": 0.06104, + "grad_norm": 0.7385823683842133, + "learning_rate": 0.003, + "loss": 4.1067, + "step": 6104 + }, + { + "epoch": 0.06105, + "grad_norm": 0.7289519050865993, + "learning_rate": 0.003, + "loss": 4.115, + "step": 6105 + }, + { + "epoch": 0.06106, + "grad_norm": 0.7418004159768921, + "learning_rate": 0.003, + "loss": 4.1343, + "step": 6106 + }, + { + "epoch": 0.06107, + "grad_norm": 0.683000547263818, + "learning_rate": 0.003, + "loss": 4.1453, + "step": 6107 + }, + { + "epoch": 0.06108, + "grad_norm": 0.5960849690912845, + "learning_rate": 0.003, + "loss": 4.1079, + "step": 6108 + }, + { + "epoch": 0.06109, + "grad_norm": 0.6202810402417694, + "learning_rate": 0.003, + "loss": 4.0857, + "step": 6109 + }, + { + "epoch": 0.0611, + "grad_norm": 0.7084749148334659, + "learning_rate": 0.003, + "loss": 4.1041, + "step": 6110 + }, + { + "epoch": 0.06111, + "grad_norm": 0.8416354572669955, + "learning_rate": 0.003, + "loss": 4.1221, + "step": 6111 + }, + { + "epoch": 0.06112, + "grad_norm": 1.1477396706080005, + "learning_rate": 0.003, + "loss": 4.1278, + "step": 6112 + }, + { + "epoch": 0.06113, + "grad_norm": 0.9298557660649742, + "learning_rate": 0.003, + "loss": 4.1337, + "step": 6113 + }, + { + "epoch": 0.06114, + "grad_norm": 0.8426576582561913, + "learning_rate": 0.003, + "loss": 4.095, + "step": 6114 + }, + { + "epoch": 0.06115, + "grad_norm": 0.8345003178868361, + "learning_rate": 0.003, + "loss": 4.0997, + "step": 6115 + }, + { + "epoch": 0.06116, + "grad_norm": 0.8891402025914743, + "learning_rate": 0.003, + "loss": 4.1225, + "step": 6116 + }, + { + "epoch": 0.06117, + "grad_norm": 0.8733942465346064, + "learning_rate": 0.003, + "loss": 4.1167, + "step": 6117 + }, + { + "epoch": 0.06118, + "grad_norm": 0.8300392838566886, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 6118 + }, + { + "epoch": 0.06119, + "grad_norm": 1.0718155087870083, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 6119 + }, + { + "epoch": 0.0612, + "grad_norm": 1.082038424508775, + "learning_rate": 0.003, + "loss": 4.1363, + "step": 6120 + }, + { + "epoch": 0.06121, + "grad_norm": 1.0039524800797741, + "learning_rate": 0.003, + "loss": 4.12, + "step": 6121 + }, + { + "epoch": 0.06122, + "grad_norm": 0.858861812903392, + "learning_rate": 0.003, + "loss": 4.1343, + "step": 6122 + }, + { + "epoch": 0.06123, + "grad_norm": 0.7517337557672539, + "learning_rate": 0.003, + "loss": 4.131, + "step": 6123 + }, + { + "epoch": 0.06124, + "grad_norm": 0.8054070621318578, + "learning_rate": 0.003, + "loss": 4.1251, + "step": 6124 + }, + { + "epoch": 0.06125, + "grad_norm": 0.8798723198485504, + "learning_rate": 0.003, + "loss": 4.0915, + "step": 6125 + }, + { + "epoch": 0.06126, + "grad_norm": 0.9746332514781364, + "learning_rate": 0.003, + "loss": 4.1116, + "step": 6126 + }, + { + "epoch": 0.06127, + "grad_norm": 1.0519385433383126, + "learning_rate": 0.003, + "loss": 4.1099, + "step": 6127 + }, + { + "epoch": 0.06128, + "grad_norm": 0.7862904704912201, + "learning_rate": 0.003, + "loss": 4.1178, + "step": 6128 + }, + { + "epoch": 0.06129, + "grad_norm": 0.8108064027519418, + "learning_rate": 0.003, + "loss": 4.1301, + "step": 6129 + }, + { + "epoch": 0.0613, + "grad_norm": 0.7803383051940936, + "learning_rate": 0.003, + "loss": 4.1383, + "step": 6130 + }, + { + "epoch": 0.06131, + "grad_norm": 0.807939857772501, + "learning_rate": 0.003, + "loss": 4.136, + "step": 6131 + }, + { + "epoch": 0.06132, + "grad_norm": 0.9270834066209759, + "learning_rate": 0.003, + "loss": 4.1373, + "step": 6132 + }, + { + "epoch": 0.06133, + "grad_norm": 1.057590287106868, + "learning_rate": 0.003, + "loss": 4.1308, + "step": 6133 + }, + { + "epoch": 0.06134, + "grad_norm": 0.9315718021788794, + "learning_rate": 0.003, + "loss": 4.1219, + "step": 6134 + }, + { + "epoch": 0.06135, + "grad_norm": 0.7682639425500988, + "learning_rate": 0.003, + "loss": 4.1459, + "step": 6135 + }, + { + "epoch": 0.06136, + "grad_norm": 0.6453509034899395, + "learning_rate": 0.003, + "loss": 4.1199, + "step": 6136 + }, + { + "epoch": 0.06137, + "grad_norm": 0.5089285133051348, + "learning_rate": 0.003, + "loss": 4.1142, + "step": 6137 + }, + { + "epoch": 0.06138, + "grad_norm": 0.5921446681322661, + "learning_rate": 0.003, + "loss": 4.1093, + "step": 6138 + }, + { + "epoch": 0.06139, + "grad_norm": 0.6572428733185919, + "learning_rate": 0.003, + "loss": 4.1294, + "step": 6139 + }, + { + "epoch": 0.0614, + "grad_norm": 0.7594320601513859, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 6140 + }, + { + "epoch": 0.06141, + "grad_norm": 0.7701960965013912, + "learning_rate": 0.003, + "loss": 4.0918, + "step": 6141 + }, + { + "epoch": 0.06142, + "grad_norm": 0.708324006436222, + "learning_rate": 0.003, + "loss": 4.1139, + "step": 6142 + }, + { + "epoch": 0.06143, + "grad_norm": 0.8072574094920275, + "learning_rate": 0.003, + "loss": 4.1253, + "step": 6143 + }, + { + "epoch": 0.06144, + "grad_norm": 0.9407988342872765, + "learning_rate": 0.003, + "loss": 4.108, + "step": 6144 + }, + { + "epoch": 0.06145, + "grad_norm": 0.9922781126952929, + "learning_rate": 0.003, + "loss": 4.1304, + "step": 6145 + }, + { + "epoch": 0.06146, + "grad_norm": 0.9137503629646905, + "learning_rate": 0.003, + "loss": 4.0974, + "step": 6146 + }, + { + "epoch": 0.06147, + "grad_norm": 0.8208946959727162, + "learning_rate": 0.003, + "loss": 4.1489, + "step": 6147 + }, + { + "epoch": 0.06148, + "grad_norm": 0.7535313704885915, + "learning_rate": 0.003, + "loss": 4.1189, + "step": 6148 + }, + { + "epoch": 0.06149, + "grad_norm": 0.7754246480784573, + "learning_rate": 0.003, + "loss": 4.1117, + "step": 6149 + }, + { + "epoch": 0.0615, + "grad_norm": 0.8086681823805969, + "learning_rate": 0.003, + "loss": 4.1301, + "step": 6150 + }, + { + "epoch": 0.06151, + "grad_norm": 0.7520491914732798, + "learning_rate": 0.003, + "loss": 4.1233, + "step": 6151 + }, + { + "epoch": 0.06152, + "grad_norm": 0.7385892490900827, + "learning_rate": 0.003, + "loss": 4.1083, + "step": 6152 + }, + { + "epoch": 0.06153, + "grad_norm": 0.585306853249163, + "learning_rate": 0.003, + "loss": 4.1309, + "step": 6153 + }, + { + "epoch": 0.06154, + "grad_norm": 0.6075200706981868, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 6154 + }, + { + "epoch": 0.06155, + "grad_norm": 0.7008926353906986, + "learning_rate": 0.003, + "loss": 4.123, + "step": 6155 + }, + { + "epoch": 0.06156, + "grad_norm": 0.7377701370473413, + "learning_rate": 0.003, + "loss": 4.1097, + "step": 6156 + }, + { + "epoch": 0.06157, + "grad_norm": 0.6372062483340378, + "learning_rate": 0.003, + "loss": 4.1258, + "step": 6157 + }, + { + "epoch": 0.06158, + "grad_norm": 0.5711395334899737, + "learning_rate": 0.003, + "loss": 4.147, + "step": 6158 + }, + { + "epoch": 0.06159, + "grad_norm": 0.6974553713335565, + "learning_rate": 0.003, + "loss": 4.106, + "step": 6159 + }, + { + "epoch": 0.0616, + "grad_norm": 0.7914523454155186, + "learning_rate": 0.003, + "loss": 4.1007, + "step": 6160 + }, + { + "epoch": 0.06161, + "grad_norm": 0.9147836100156929, + "learning_rate": 0.003, + "loss": 4.1323, + "step": 6161 + }, + { + "epoch": 0.06162, + "grad_norm": 1.0202822458279153, + "learning_rate": 0.003, + "loss": 4.112, + "step": 6162 + }, + { + "epoch": 0.06163, + "grad_norm": 0.8492609261522512, + "learning_rate": 0.003, + "loss": 4.149, + "step": 6163 + }, + { + "epoch": 0.06164, + "grad_norm": 0.6305588102799871, + "learning_rate": 0.003, + "loss": 4.1197, + "step": 6164 + }, + { + "epoch": 0.06165, + "grad_norm": 0.70222905462584, + "learning_rate": 0.003, + "loss": 4.1037, + "step": 6165 + }, + { + "epoch": 0.06166, + "grad_norm": 0.7330174697865238, + "learning_rate": 0.003, + "loss": 4.1323, + "step": 6166 + }, + { + "epoch": 0.06167, + "grad_norm": 0.8120770794656365, + "learning_rate": 0.003, + "loss": 4.1048, + "step": 6167 + }, + { + "epoch": 0.06168, + "grad_norm": 0.9461948541503907, + "learning_rate": 0.003, + "loss": 4.1221, + "step": 6168 + }, + { + "epoch": 0.06169, + "grad_norm": 0.8911742935865481, + "learning_rate": 0.003, + "loss": 4.1144, + "step": 6169 + }, + { + "epoch": 0.0617, + "grad_norm": 0.7323433640480617, + "learning_rate": 0.003, + "loss": 4.1158, + "step": 6170 + }, + { + "epoch": 0.06171, + "grad_norm": 0.6208020453162558, + "learning_rate": 0.003, + "loss": 4.1142, + "step": 6171 + }, + { + "epoch": 0.06172, + "grad_norm": 0.7291939143005758, + "learning_rate": 0.003, + "loss": 4.0931, + "step": 6172 + }, + { + "epoch": 0.06173, + "grad_norm": 0.7913577473386937, + "learning_rate": 0.003, + "loss": 4.1242, + "step": 6173 + }, + { + "epoch": 0.06174, + "grad_norm": 0.9778156595493965, + "learning_rate": 0.003, + "loss": 4.1249, + "step": 6174 + }, + { + "epoch": 0.06175, + "grad_norm": 1.1932141855839438, + "learning_rate": 0.003, + "loss": 4.1365, + "step": 6175 + }, + { + "epoch": 0.06176, + "grad_norm": 0.7326715787881036, + "learning_rate": 0.003, + "loss": 4.1201, + "step": 6176 + }, + { + "epoch": 0.06177, + "grad_norm": 0.7923716099483298, + "learning_rate": 0.003, + "loss": 4.116, + "step": 6177 + }, + { + "epoch": 0.06178, + "grad_norm": 0.9764360789021225, + "learning_rate": 0.003, + "loss": 4.1127, + "step": 6178 + }, + { + "epoch": 0.06179, + "grad_norm": 1.0887682373744165, + "learning_rate": 0.003, + "loss": 4.135, + "step": 6179 + }, + { + "epoch": 0.0618, + "grad_norm": 0.8700014538602673, + "learning_rate": 0.003, + "loss": 4.1276, + "step": 6180 + }, + { + "epoch": 0.06181, + "grad_norm": 0.8463904835451614, + "learning_rate": 0.003, + "loss": 4.1436, + "step": 6181 + }, + { + "epoch": 0.06182, + "grad_norm": 0.8112163844886693, + "learning_rate": 0.003, + "loss": 4.1031, + "step": 6182 + }, + { + "epoch": 0.06183, + "grad_norm": 0.8784890732030473, + "learning_rate": 0.003, + "loss": 4.1191, + "step": 6183 + }, + { + "epoch": 0.06184, + "grad_norm": 0.8158970894757084, + "learning_rate": 0.003, + "loss": 4.1197, + "step": 6184 + }, + { + "epoch": 0.06185, + "grad_norm": 0.8186525569563441, + "learning_rate": 0.003, + "loss": 4.1077, + "step": 6185 + }, + { + "epoch": 0.06186, + "grad_norm": 0.8640122378248711, + "learning_rate": 0.003, + "loss": 4.1353, + "step": 6186 + }, + { + "epoch": 0.06187, + "grad_norm": 0.9198602507085508, + "learning_rate": 0.003, + "loss": 4.1429, + "step": 6187 + }, + { + "epoch": 0.06188, + "grad_norm": 0.9582358948075637, + "learning_rate": 0.003, + "loss": 4.123, + "step": 6188 + }, + { + "epoch": 0.06189, + "grad_norm": 0.9436060131627309, + "learning_rate": 0.003, + "loss": 4.1141, + "step": 6189 + }, + { + "epoch": 0.0619, + "grad_norm": 0.8106004278533174, + "learning_rate": 0.003, + "loss": 4.1567, + "step": 6190 + }, + { + "epoch": 0.06191, + "grad_norm": 0.7377926481554797, + "learning_rate": 0.003, + "loss": 4.1079, + "step": 6191 + }, + { + "epoch": 0.06192, + "grad_norm": 0.7292550893414801, + "learning_rate": 0.003, + "loss": 4.1096, + "step": 6192 + }, + { + "epoch": 0.06193, + "grad_norm": 0.796675035271722, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 6193 + }, + { + "epoch": 0.06194, + "grad_norm": 0.9834328051273128, + "learning_rate": 0.003, + "loss": 4.1093, + "step": 6194 + }, + { + "epoch": 0.06195, + "grad_norm": 0.9674422632284655, + "learning_rate": 0.003, + "loss": 4.127, + "step": 6195 + }, + { + "epoch": 0.06196, + "grad_norm": 0.9127448537843487, + "learning_rate": 0.003, + "loss": 4.1387, + "step": 6196 + }, + { + "epoch": 0.06197, + "grad_norm": 0.8018597016816854, + "learning_rate": 0.003, + "loss": 4.1196, + "step": 6197 + }, + { + "epoch": 0.06198, + "grad_norm": 0.6492553167687579, + "learning_rate": 0.003, + "loss": 4.1469, + "step": 6198 + }, + { + "epoch": 0.06199, + "grad_norm": 0.500075335938287, + "learning_rate": 0.003, + "loss": 4.1312, + "step": 6199 + }, + { + "epoch": 0.062, + "grad_norm": 0.6408031097259872, + "learning_rate": 0.003, + "loss": 4.1388, + "step": 6200 + }, + { + "epoch": 0.06201, + "grad_norm": 0.6882374713932283, + "learning_rate": 0.003, + "loss": 4.1048, + "step": 6201 + }, + { + "epoch": 0.06202, + "grad_norm": 0.6650459048800811, + "learning_rate": 0.003, + "loss": 4.1492, + "step": 6202 + }, + { + "epoch": 0.06203, + "grad_norm": 0.7279713021245527, + "learning_rate": 0.003, + "loss": 4.1338, + "step": 6203 + }, + { + "epoch": 0.06204, + "grad_norm": 0.8176381462944218, + "learning_rate": 0.003, + "loss": 4.1308, + "step": 6204 + }, + { + "epoch": 0.06205, + "grad_norm": 0.8635078911095838, + "learning_rate": 0.003, + "loss": 4.1327, + "step": 6205 + }, + { + "epoch": 0.06206, + "grad_norm": 0.9051460085260756, + "learning_rate": 0.003, + "loss": 4.1253, + "step": 6206 + }, + { + "epoch": 0.06207, + "grad_norm": 0.9683554461781614, + "learning_rate": 0.003, + "loss": 4.0917, + "step": 6207 + }, + { + "epoch": 0.06208, + "grad_norm": 0.9491242929275403, + "learning_rate": 0.003, + "loss": 4.1203, + "step": 6208 + }, + { + "epoch": 0.06209, + "grad_norm": 0.946544007384069, + "learning_rate": 0.003, + "loss": 4.1078, + "step": 6209 + }, + { + "epoch": 0.0621, + "grad_norm": 0.7496490508161171, + "learning_rate": 0.003, + "loss": 4.1606, + "step": 6210 + }, + { + "epoch": 0.06211, + "grad_norm": 0.7213205165847495, + "learning_rate": 0.003, + "loss": 4.1221, + "step": 6211 + }, + { + "epoch": 0.06212, + "grad_norm": 0.7495412802850021, + "learning_rate": 0.003, + "loss": 4.1295, + "step": 6212 + }, + { + "epoch": 0.06213, + "grad_norm": 0.7517545861084816, + "learning_rate": 0.003, + "loss": 4.1064, + "step": 6213 + }, + { + "epoch": 0.06214, + "grad_norm": 0.8206749672685049, + "learning_rate": 0.003, + "loss": 4.1092, + "step": 6214 + }, + { + "epoch": 0.06215, + "grad_norm": 0.8335784823968934, + "learning_rate": 0.003, + "loss": 4.1283, + "step": 6215 + }, + { + "epoch": 0.06216, + "grad_norm": 0.8600050966450496, + "learning_rate": 0.003, + "loss": 4.1572, + "step": 6216 + }, + { + "epoch": 0.06217, + "grad_norm": 0.7434739467397622, + "learning_rate": 0.003, + "loss": 4.1152, + "step": 6217 + }, + { + "epoch": 0.06218, + "grad_norm": 0.7048668827578362, + "learning_rate": 0.003, + "loss": 4.1013, + "step": 6218 + }, + { + "epoch": 0.06219, + "grad_norm": 0.772659800309845, + "learning_rate": 0.003, + "loss": 4.1341, + "step": 6219 + }, + { + "epoch": 0.0622, + "grad_norm": 0.8655188591453645, + "learning_rate": 0.003, + "loss": 4.1305, + "step": 6220 + }, + { + "epoch": 0.06221, + "grad_norm": 0.855208686912549, + "learning_rate": 0.003, + "loss": 4.1321, + "step": 6221 + }, + { + "epoch": 0.06222, + "grad_norm": 0.7753969968080353, + "learning_rate": 0.003, + "loss": 4.1176, + "step": 6222 + }, + { + "epoch": 0.06223, + "grad_norm": 0.7347873178842909, + "learning_rate": 0.003, + "loss": 4.1243, + "step": 6223 + }, + { + "epoch": 0.06224, + "grad_norm": 0.7453826365575011, + "learning_rate": 0.003, + "loss": 4.095, + "step": 6224 + }, + { + "epoch": 0.06225, + "grad_norm": 0.6725261598977015, + "learning_rate": 0.003, + "loss": 4.1197, + "step": 6225 + }, + { + "epoch": 0.06226, + "grad_norm": 0.7149958935506678, + "learning_rate": 0.003, + "loss": 4.1017, + "step": 6226 + }, + { + "epoch": 0.06227, + "grad_norm": 0.6062843965023491, + "learning_rate": 0.003, + "loss": 4.1101, + "step": 6227 + }, + { + "epoch": 0.06228, + "grad_norm": 0.5956924427962597, + "learning_rate": 0.003, + "loss": 4.1163, + "step": 6228 + }, + { + "epoch": 0.06229, + "grad_norm": 0.49453429887940603, + "learning_rate": 0.003, + "loss": 4.0981, + "step": 6229 + }, + { + "epoch": 0.0623, + "grad_norm": 0.4966784851981883, + "learning_rate": 0.003, + "loss": 4.1103, + "step": 6230 + }, + { + "epoch": 0.06231, + "grad_norm": 0.5841536691365076, + "learning_rate": 0.003, + "loss": 4.1017, + "step": 6231 + }, + { + "epoch": 0.06232, + "grad_norm": 0.6363555647753291, + "learning_rate": 0.003, + "loss": 4.098, + "step": 6232 + }, + { + "epoch": 0.06233, + "grad_norm": 0.739340815218559, + "learning_rate": 0.003, + "loss": 4.0966, + "step": 6233 + }, + { + "epoch": 0.06234, + "grad_norm": 0.983288242920816, + "learning_rate": 0.003, + "loss": 4.1054, + "step": 6234 + }, + { + "epoch": 0.06235, + "grad_norm": 1.256808031451369, + "learning_rate": 0.003, + "loss": 4.121, + "step": 6235 + }, + { + "epoch": 0.06236, + "grad_norm": 0.879699004471445, + "learning_rate": 0.003, + "loss": 4.1267, + "step": 6236 + }, + { + "epoch": 0.06237, + "grad_norm": 0.7521683358827187, + "learning_rate": 0.003, + "loss": 4.1184, + "step": 6237 + }, + { + "epoch": 0.06238, + "grad_norm": 0.9148201380927862, + "learning_rate": 0.003, + "loss": 4.1211, + "step": 6238 + }, + { + "epoch": 0.06239, + "grad_norm": 0.9607286756265107, + "learning_rate": 0.003, + "loss": 4.1067, + "step": 6239 + }, + { + "epoch": 0.0624, + "grad_norm": 0.9539386861231751, + "learning_rate": 0.003, + "loss": 4.1235, + "step": 6240 + }, + { + "epoch": 0.06241, + "grad_norm": 0.9355231782983046, + "learning_rate": 0.003, + "loss": 4.1058, + "step": 6241 + }, + { + "epoch": 0.06242, + "grad_norm": 0.9288421332433013, + "learning_rate": 0.003, + "loss": 4.1331, + "step": 6242 + }, + { + "epoch": 0.06243, + "grad_norm": 0.8502254135089574, + "learning_rate": 0.003, + "loss": 4.123, + "step": 6243 + }, + { + "epoch": 0.06244, + "grad_norm": 0.9732395834043053, + "learning_rate": 0.003, + "loss": 4.1585, + "step": 6244 + }, + { + "epoch": 0.06245, + "grad_norm": 1.1071153201796413, + "learning_rate": 0.003, + "loss": 4.1402, + "step": 6245 + }, + { + "epoch": 0.06246, + "grad_norm": 1.0940195095069707, + "learning_rate": 0.003, + "loss": 4.1576, + "step": 6246 + }, + { + "epoch": 0.06247, + "grad_norm": 0.9412652481549822, + "learning_rate": 0.003, + "loss": 4.1377, + "step": 6247 + }, + { + "epoch": 0.06248, + "grad_norm": 0.895195578619345, + "learning_rate": 0.003, + "loss": 4.136, + "step": 6248 + }, + { + "epoch": 0.06249, + "grad_norm": 0.9518706201636767, + "learning_rate": 0.003, + "loss": 4.1519, + "step": 6249 + }, + { + "epoch": 0.0625, + "grad_norm": 0.9872959824439677, + "learning_rate": 0.003, + "loss": 4.137, + "step": 6250 + }, + { + "epoch": 0.06251, + "grad_norm": 0.8928606640381823, + "learning_rate": 0.003, + "loss": 4.1229, + "step": 6251 + }, + { + "epoch": 0.06252, + "grad_norm": 0.8613095945713102, + "learning_rate": 0.003, + "loss": 4.1447, + "step": 6252 + }, + { + "epoch": 0.06253, + "grad_norm": 0.8745234996707003, + "learning_rate": 0.003, + "loss": 4.1324, + "step": 6253 + }, + { + "epoch": 0.06254, + "grad_norm": 0.9945273839471781, + "learning_rate": 0.003, + "loss": 4.1301, + "step": 6254 + }, + { + "epoch": 0.06255, + "grad_norm": 0.9317981123576337, + "learning_rate": 0.003, + "loss": 4.1423, + "step": 6255 + }, + { + "epoch": 0.06256, + "grad_norm": 0.935398061718941, + "learning_rate": 0.003, + "loss": 4.1146, + "step": 6256 + }, + { + "epoch": 0.06257, + "grad_norm": 0.9753404112298208, + "learning_rate": 0.003, + "loss": 4.095, + "step": 6257 + }, + { + "epoch": 0.06258, + "grad_norm": 0.7914387634721403, + "learning_rate": 0.003, + "loss": 4.1434, + "step": 6258 + }, + { + "epoch": 0.06259, + "grad_norm": 0.8459303686065615, + "learning_rate": 0.003, + "loss": 4.1596, + "step": 6259 + }, + { + "epoch": 0.0626, + "grad_norm": 0.8608798731117852, + "learning_rate": 0.003, + "loss": 4.1331, + "step": 6260 + }, + { + "epoch": 0.06261, + "grad_norm": 0.9015697577574606, + "learning_rate": 0.003, + "loss": 4.1183, + "step": 6261 + }, + { + "epoch": 0.06262, + "grad_norm": 0.8626927231908963, + "learning_rate": 0.003, + "loss": 4.1227, + "step": 6262 + }, + { + "epoch": 0.06263, + "grad_norm": 0.8363061035136914, + "learning_rate": 0.003, + "loss": 4.1492, + "step": 6263 + }, + { + "epoch": 0.06264, + "grad_norm": 0.870156868679017, + "learning_rate": 0.003, + "loss": 4.1245, + "step": 6264 + }, + { + "epoch": 0.06265, + "grad_norm": 0.9555307920316731, + "learning_rate": 0.003, + "loss": 4.145, + "step": 6265 + }, + { + "epoch": 0.06266, + "grad_norm": 0.8998498765721897, + "learning_rate": 0.003, + "loss": 4.1287, + "step": 6266 + }, + { + "epoch": 0.06267, + "grad_norm": 0.8579446764983243, + "learning_rate": 0.003, + "loss": 4.1345, + "step": 6267 + }, + { + "epoch": 0.06268, + "grad_norm": 0.7491708731971836, + "learning_rate": 0.003, + "loss": 4.1014, + "step": 6268 + }, + { + "epoch": 0.06269, + "grad_norm": 0.5769598306382601, + "learning_rate": 0.003, + "loss": 4.1013, + "step": 6269 + }, + { + "epoch": 0.0627, + "grad_norm": 0.612130695073576, + "learning_rate": 0.003, + "loss": 4.1176, + "step": 6270 + }, + { + "epoch": 0.06271, + "grad_norm": 0.5804514862016513, + "learning_rate": 0.003, + "loss": 4.129, + "step": 6271 + }, + { + "epoch": 0.06272, + "grad_norm": 0.6768271666465923, + "learning_rate": 0.003, + "loss": 4.1408, + "step": 6272 + }, + { + "epoch": 0.06273, + "grad_norm": 0.725620628836043, + "learning_rate": 0.003, + "loss": 4.1407, + "step": 6273 + }, + { + "epoch": 0.06274, + "grad_norm": 0.7124587885763234, + "learning_rate": 0.003, + "loss": 4.1181, + "step": 6274 + }, + { + "epoch": 0.06275, + "grad_norm": 0.6549941998076714, + "learning_rate": 0.003, + "loss": 4.1093, + "step": 6275 + }, + { + "epoch": 0.06276, + "grad_norm": 0.651422535989974, + "learning_rate": 0.003, + "loss": 4.1216, + "step": 6276 + }, + { + "epoch": 0.06277, + "grad_norm": 0.7632957652022478, + "learning_rate": 0.003, + "loss": 4.1264, + "step": 6277 + }, + { + "epoch": 0.06278, + "grad_norm": 0.8921311551644955, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 6278 + }, + { + "epoch": 0.06279, + "grad_norm": 0.9316054360840504, + "learning_rate": 0.003, + "loss": 4.1418, + "step": 6279 + }, + { + "epoch": 0.0628, + "grad_norm": 0.9480087249404298, + "learning_rate": 0.003, + "loss": 4.1147, + "step": 6280 + }, + { + "epoch": 0.06281, + "grad_norm": 0.9437555976288077, + "learning_rate": 0.003, + "loss": 4.094, + "step": 6281 + }, + { + "epoch": 0.06282, + "grad_norm": 0.8181620617485008, + "learning_rate": 0.003, + "loss": 4.1155, + "step": 6282 + }, + { + "epoch": 0.06283, + "grad_norm": 0.6530022259763903, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 6283 + }, + { + "epoch": 0.06284, + "grad_norm": 0.5914752081352821, + "learning_rate": 0.003, + "loss": 4.1161, + "step": 6284 + }, + { + "epoch": 0.06285, + "grad_norm": 0.5833107165282243, + "learning_rate": 0.003, + "loss": 4.1141, + "step": 6285 + }, + { + "epoch": 0.06286, + "grad_norm": 0.6373596862276327, + "learning_rate": 0.003, + "loss": 4.1034, + "step": 6286 + }, + { + "epoch": 0.06287, + "grad_norm": 0.6834837784355613, + "learning_rate": 0.003, + "loss": 4.0962, + "step": 6287 + }, + { + "epoch": 0.06288, + "grad_norm": 0.6535796667468495, + "learning_rate": 0.003, + "loss": 4.1224, + "step": 6288 + }, + { + "epoch": 0.06289, + "grad_norm": 0.6174828688423645, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 6289 + }, + { + "epoch": 0.0629, + "grad_norm": 0.4959130253790628, + "learning_rate": 0.003, + "loss": 4.1004, + "step": 6290 + }, + { + "epoch": 0.06291, + "grad_norm": 0.5595818315925254, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 6291 + }, + { + "epoch": 0.06292, + "grad_norm": 0.5248847935016591, + "learning_rate": 0.003, + "loss": 4.0995, + "step": 6292 + }, + { + "epoch": 0.06293, + "grad_norm": 0.5063929729505692, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 6293 + }, + { + "epoch": 0.06294, + "grad_norm": 0.46785323488692016, + "learning_rate": 0.003, + "loss": 4.1102, + "step": 6294 + }, + { + "epoch": 0.06295, + "grad_norm": 0.4202500873613126, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 6295 + }, + { + "epoch": 0.06296, + "grad_norm": 0.42143259177649955, + "learning_rate": 0.003, + "loss": 4.1268, + "step": 6296 + }, + { + "epoch": 0.06297, + "grad_norm": 0.49833908617620376, + "learning_rate": 0.003, + "loss": 4.1005, + "step": 6297 + }, + { + "epoch": 0.06298, + "grad_norm": 0.7251722641880546, + "learning_rate": 0.003, + "loss": 4.1009, + "step": 6298 + }, + { + "epoch": 0.06299, + "grad_norm": 1.1796989504352162, + "learning_rate": 0.003, + "loss": 4.112, + "step": 6299 + }, + { + "epoch": 0.063, + "grad_norm": 1.0778085487061984, + "learning_rate": 0.003, + "loss": 4.1599, + "step": 6300 + }, + { + "epoch": 0.06301, + "grad_norm": 0.890093583413934, + "learning_rate": 0.003, + "loss": 4.1362, + "step": 6301 + }, + { + "epoch": 0.06302, + "grad_norm": 0.8849818441147115, + "learning_rate": 0.003, + "loss": 4.1248, + "step": 6302 + }, + { + "epoch": 0.06303, + "grad_norm": 0.9016935377264436, + "learning_rate": 0.003, + "loss": 4.1265, + "step": 6303 + }, + { + "epoch": 0.06304, + "grad_norm": 0.9975700771133877, + "learning_rate": 0.003, + "loss": 4.1474, + "step": 6304 + }, + { + "epoch": 0.06305, + "grad_norm": 1.078399082415521, + "learning_rate": 0.003, + "loss": 4.1567, + "step": 6305 + }, + { + "epoch": 0.06306, + "grad_norm": 0.8656165382853225, + "learning_rate": 0.003, + "loss": 4.1111, + "step": 6306 + }, + { + "epoch": 0.06307, + "grad_norm": 0.9837508136152356, + "learning_rate": 0.003, + "loss": 4.113, + "step": 6307 + }, + { + "epoch": 0.06308, + "grad_norm": 1.0753663335239507, + "learning_rate": 0.003, + "loss": 4.1403, + "step": 6308 + }, + { + "epoch": 0.06309, + "grad_norm": 0.9920769784954451, + "learning_rate": 0.003, + "loss": 4.1273, + "step": 6309 + }, + { + "epoch": 0.0631, + "grad_norm": 0.8930854451381682, + "learning_rate": 0.003, + "loss": 4.1275, + "step": 6310 + }, + { + "epoch": 0.06311, + "grad_norm": 0.9784115790794385, + "learning_rate": 0.003, + "loss": 4.125, + "step": 6311 + }, + { + "epoch": 0.06312, + "grad_norm": 0.9052068566792085, + "learning_rate": 0.003, + "loss": 4.1182, + "step": 6312 + }, + { + "epoch": 0.06313, + "grad_norm": 0.8062060403102312, + "learning_rate": 0.003, + "loss": 4.1391, + "step": 6313 + }, + { + "epoch": 0.06314, + "grad_norm": 0.8612542206730062, + "learning_rate": 0.003, + "loss": 4.1427, + "step": 6314 + }, + { + "epoch": 0.06315, + "grad_norm": 1.2349281974671547, + "learning_rate": 0.003, + "loss": 4.1291, + "step": 6315 + }, + { + "epoch": 0.06316, + "grad_norm": 0.9748496951834336, + "learning_rate": 0.003, + "loss": 4.0956, + "step": 6316 + }, + { + "epoch": 0.06317, + "grad_norm": 0.8725933515560151, + "learning_rate": 0.003, + "loss": 4.1222, + "step": 6317 + }, + { + "epoch": 0.06318, + "grad_norm": 0.7900742702838245, + "learning_rate": 0.003, + "loss": 4.1613, + "step": 6318 + }, + { + "epoch": 0.06319, + "grad_norm": 0.8185448176929361, + "learning_rate": 0.003, + "loss": 4.1638, + "step": 6319 + }, + { + "epoch": 0.0632, + "grad_norm": 0.7867763226056398, + "learning_rate": 0.003, + "loss": 4.1237, + "step": 6320 + }, + { + "epoch": 0.06321, + "grad_norm": 0.7397415591963721, + "learning_rate": 0.003, + "loss": 4.136, + "step": 6321 + }, + { + "epoch": 0.06322, + "grad_norm": 0.7295018218458792, + "learning_rate": 0.003, + "loss": 4.1205, + "step": 6322 + }, + { + "epoch": 0.06323, + "grad_norm": 0.8038126764870634, + "learning_rate": 0.003, + "loss": 4.1163, + "step": 6323 + }, + { + "epoch": 0.06324, + "grad_norm": 0.8381105892893556, + "learning_rate": 0.003, + "loss": 4.1371, + "step": 6324 + }, + { + "epoch": 0.06325, + "grad_norm": 0.7753841480841397, + "learning_rate": 0.003, + "loss": 4.1158, + "step": 6325 + }, + { + "epoch": 0.06326, + "grad_norm": 0.7397370113159383, + "learning_rate": 0.003, + "loss": 4.1282, + "step": 6326 + }, + { + "epoch": 0.06327, + "grad_norm": 0.8110748971387649, + "learning_rate": 0.003, + "loss": 4.0988, + "step": 6327 + }, + { + "epoch": 0.06328, + "grad_norm": 0.8165737037187061, + "learning_rate": 0.003, + "loss": 4.1491, + "step": 6328 + }, + { + "epoch": 0.06329, + "grad_norm": 0.8459755592595166, + "learning_rate": 0.003, + "loss": 4.1322, + "step": 6329 + }, + { + "epoch": 0.0633, + "grad_norm": 0.7837172627440523, + "learning_rate": 0.003, + "loss": 4.1074, + "step": 6330 + }, + { + "epoch": 0.06331, + "grad_norm": 0.7382526500041345, + "learning_rate": 0.003, + "loss": 4.1036, + "step": 6331 + }, + { + "epoch": 0.06332, + "grad_norm": 0.8446347024742388, + "learning_rate": 0.003, + "loss": 4.1058, + "step": 6332 + }, + { + "epoch": 0.06333, + "grad_norm": 0.8755392261266487, + "learning_rate": 0.003, + "loss": 4.1119, + "step": 6333 + }, + { + "epoch": 0.06334, + "grad_norm": 0.9464338840507761, + "learning_rate": 0.003, + "loss": 4.1143, + "step": 6334 + }, + { + "epoch": 0.06335, + "grad_norm": 0.9871255748238474, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 6335 + }, + { + "epoch": 0.06336, + "grad_norm": 1.011604006822508, + "learning_rate": 0.003, + "loss": 4.1484, + "step": 6336 + }, + { + "epoch": 0.06337, + "grad_norm": 1.054372126242909, + "learning_rate": 0.003, + "loss": 4.1529, + "step": 6337 + }, + { + "epoch": 0.06338, + "grad_norm": 0.8149090879744815, + "learning_rate": 0.003, + "loss": 4.1337, + "step": 6338 + }, + { + "epoch": 0.06339, + "grad_norm": 0.8045420062877509, + "learning_rate": 0.003, + "loss": 4.107, + "step": 6339 + }, + { + "epoch": 0.0634, + "grad_norm": 0.773692028277079, + "learning_rate": 0.003, + "loss": 4.102, + "step": 6340 + }, + { + "epoch": 0.06341, + "grad_norm": 0.834869748376275, + "learning_rate": 0.003, + "loss": 4.1011, + "step": 6341 + }, + { + "epoch": 0.06342, + "grad_norm": 0.8239421334821241, + "learning_rate": 0.003, + "loss": 4.1283, + "step": 6342 + }, + { + "epoch": 0.06343, + "grad_norm": 0.7317579621776237, + "learning_rate": 0.003, + "loss": 4.1343, + "step": 6343 + }, + { + "epoch": 0.06344, + "grad_norm": 0.7009311944721036, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 6344 + }, + { + "epoch": 0.06345, + "grad_norm": 0.6736306168677829, + "learning_rate": 0.003, + "loss": 4.1068, + "step": 6345 + }, + { + "epoch": 0.06346, + "grad_norm": 0.6489992606346455, + "learning_rate": 0.003, + "loss": 4.1217, + "step": 6346 + }, + { + "epoch": 0.06347, + "grad_norm": 0.7318992775914919, + "learning_rate": 0.003, + "loss": 4.1099, + "step": 6347 + }, + { + "epoch": 0.06348, + "grad_norm": 0.8506294348624643, + "learning_rate": 0.003, + "loss": 4.1073, + "step": 6348 + }, + { + "epoch": 0.06349, + "grad_norm": 0.8414498040161846, + "learning_rate": 0.003, + "loss": 4.1078, + "step": 6349 + }, + { + "epoch": 0.0635, + "grad_norm": 0.7853415582358513, + "learning_rate": 0.003, + "loss": 4.0995, + "step": 6350 + }, + { + "epoch": 0.06351, + "grad_norm": 0.8089386421880802, + "learning_rate": 0.003, + "loss": 4.1269, + "step": 6351 + }, + { + "epoch": 0.06352, + "grad_norm": 0.7578096372347558, + "learning_rate": 0.003, + "loss": 4.1038, + "step": 6352 + }, + { + "epoch": 0.06353, + "grad_norm": 0.7726928843900696, + "learning_rate": 0.003, + "loss": 4.1191, + "step": 6353 + }, + { + "epoch": 0.06354, + "grad_norm": 0.8235268790134066, + "learning_rate": 0.003, + "loss": 4.1143, + "step": 6354 + }, + { + "epoch": 0.06355, + "grad_norm": 0.8685123509492527, + "learning_rate": 0.003, + "loss": 4.124, + "step": 6355 + }, + { + "epoch": 0.06356, + "grad_norm": 0.9033483279727317, + "learning_rate": 0.003, + "loss": 4.1175, + "step": 6356 + }, + { + "epoch": 0.06357, + "grad_norm": 0.8189510174670123, + "learning_rate": 0.003, + "loss": 4.1338, + "step": 6357 + }, + { + "epoch": 0.06358, + "grad_norm": 0.7233022129389708, + "learning_rate": 0.003, + "loss": 4.1001, + "step": 6358 + }, + { + "epoch": 0.06359, + "grad_norm": 0.591618308364435, + "learning_rate": 0.003, + "loss": 4.0988, + "step": 6359 + }, + { + "epoch": 0.0636, + "grad_norm": 0.6724366774358936, + "learning_rate": 0.003, + "loss": 4.1258, + "step": 6360 + }, + { + "epoch": 0.06361, + "grad_norm": 0.7710988154121916, + "learning_rate": 0.003, + "loss": 4.1203, + "step": 6361 + }, + { + "epoch": 0.06362, + "grad_norm": 0.8624550458659666, + "learning_rate": 0.003, + "loss": 4.1283, + "step": 6362 + }, + { + "epoch": 0.06363, + "grad_norm": 0.9165629494523708, + "learning_rate": 0.003, + "loss": 4.1128, + "step": 6363 + }, + { + "epoch": 0.06364, + "grad_norm": 0.7471123965901227, + "learning_rate": 0.003, + "loss": 4.1082, + "step": 6364 + }, + { + "epoch": 0.06365, + "grad_norm": 0.674364493786534, + "learning_rate": 0.003, + "loss": 4.0935, + "step": 6365 + }, + { + "epoch": 0.06366, + "grad_norm": 0.7335045602903685, + "learning_rate": 0.003, + "loss": 4.1234, + "step": 6366 + }, + { + "epoch": 0.06367, + "grad_norm": 0.829927307349712, + "learning_rate": 0.003, + "loss": 4.1067, + "step": 6367 + }, + { + "epoch": 0.06368, + "grad_norm": 0.9487139825002684, + "learning_rate": 0.003, + "loss": 4.1467, + "step": 6368 + }, + { + "epoch": 0.06369, + "grad_norm": 0.911983799114357, + "learning_rate": 0.003, + "loss": 4.1529, + "step": 6369 + }, + { + "epoch": 0.0637, + "grad_norm": 1.002706686712587, + "learning_rate": 0.003, + "loss": 4.1168, + "step": 6370 + }, + { + "epoch": 0.06371, + "grad_norm": 1.0290249495706973, + "learning_rate": 0.003, + "loss": 4.1453, + "step": 6371 + }, + { + "epoch": 0.06372, + "grad_norm": 0.9956241176141073, + "learning_rate": 0.003, + "loss": 4.1401, + "step": 6372 + }, + { + "epoch": 0.06373, + "grad_norm": 0.9694674226068865, + "learning_rate": 0.003, + "loss": 4.1382, + "step": 6373 + }, + { + "epoch": 0.06374, + "grad_norm": 0.8725376570060053, + "learning_rate": 0.003, + "loss": 4.0933, + "step": 6374 + }, + { + "epoch": 0.06375, + "grad_norm": 0.8676467464492347, + "learning_rate": 0.003, + "loss": 4.1163, + "step": 6375 + }, + { + "epoch": 0.06376, + "grad_norm": 0.8676286162915308, + "learning_rate": 0.003, + "loss": 4.1383, + "step": 6376 + }, + { + "epoch": 0.06377, + "grad_norm": 0.9874746445128965, + "learning_rate": 0.003, + "loss": 4.1141, + "step": 6377 + }, + { + "epoch": 0.06378, + "grad_norm": 1.0158658512452112, + "learning_rate": 0.003, + "loss": 4.1542, + "step": 6378 + }, + { + "epoch": 0.06379, + "grad_norm": 1.0804519520556228, + "learning_rate": 0.003, + "loss": 4.1544, + "step": 6379 + }, + { + "epoch": 0.0638, + "grad_norm": 1.0517174792581394, + "learning_rate": 0.003, + "loss": 4.1343, + "step": 6380 + }, + { + "epoch": 0.06381, + "grad_norm": 0.9191882933740307, + "learning_rate": 0.003, + "loss": 4.1127, + "step": 6381 + }, + { + "epoch": 0.06382, + "grad_norm": 1.112305209925744, + "learning_rate": 0.003, + "loss": 4.1259, + "step": 6382 + }, + { + "epoch": 0.06383, + "grad_norm": 0.887564897445918, + "learning_rate": 0.003, + "loss": 4.1516, + "step": 6383 + }, + { + "epoch": 0.06384, + "grad_norm": 0.7444199605679525, + "learning_rate": 0.003, + "loss": 4.1456, + "step": 6384 + }, + { + "epoch": 0.06385, + "grad_norm": 0.6861155001402504, + "learning_rate": 0.003, + "loss": 4.1053, + "step": 6385 + }, + { + "epoch": 0.06386, + "grad_norm": 0.6803077477027196, + "learning_rate": 0.003, + "loss": 4.1145, + "step": 6386 + }, + { + "epoch": 0.06387, + "grad_norm": 0.6585413586302139, + "learning_rate": 0.003, + "loss": 4.1229, + "step": 6387 + }, + { + "epoch": 0.06388, + "grad_norm": 0.6526729101850073, + "learning_rate": 0.003, + "loss": 4.1148, + "step": 6388 + }, + { + "epoch": 0.06389, + "grad_norm": 0.6961400897039125, + "learning_rate": 0.003, + "loss": 4.0977, + "step": 6389 + }, + { + "epoch": 0.0639, + "grad_norm": 0.7971727357460624, + "learning_rate": 0.003, + "loss": 4.1202, + "step": 6390 + }, + { + "epoch": 0.06391, + "grad_norm": 0.8100704061167459, + "learning_rate": 0.003, + "loss": 4.1393, + "step": 6391 + }, + { + "epoch": 0.06392, + "grad_norm": 0.7878863222547036, + "learning_rate": 0.003, + "loss": 4.1102, + "step": 6392 + }, + { + "epoch": 0.06393, + "grad_norm": 0.7832197554643082, + "learning_rate": 0.003, + "loss": 4.1226, + "step": 6393 + }, + { + "epoch": 0.06394, + "grad_norm": 0.7676826793962613, + "learning_rate": 0.003, + "loss": 4.1154, + "step": 6394 + }, + { + "epoch": 0.06395, + "grad_norm": 0.7865004179204782, + "learning_rate": 0.003, + "loss": 4.1403, + "step": 6395 + }, + { + "epoch": 0.06396, + "grad_norm": 0.7481149498310921, + "learning_rate": 0.003, + "loss": 4.1519, + "step": 6396 + }, + { + "epoch": 0.06397, + "grad_norm": 0.6578310984328573, + "learning_rate": 0.003, + "loss": 4.1029, + "step": 6397 + }, + { + "epoch": 0.06398, + "grad_norm": 0.7814732896391083, + "learning_rate": 0.003, + "loss": 4.0974, + "step": 6398 + }, + { + "epoch": 0.06399, + "grad_norm": 0.7910300530341595, + "learning_rate": 0.003, + "loss": 4.1275, + "step": 6399 + }, + { + "epoch": 0.064, + "grad_norm": 0.7384277513313977, + "learning_rate": 0.003, + "loss": 4.1199, + "step": 6400 + }, + { + "epoch": 0.06401, + "grad_norm": 0.6044786328825004, + "learning_rate": 0.003, + "loss": 4.0992, + "step": 6401 + }, + { + "epoch": 0.06402, + "grad_norm": 0.5635787730251838, + "learning_rate": 0.003, + "loss": 4.1011, + "step": 6402 + }, + { + "epoch": 0.06403, + "grad_norm": 0.4804036029092371, + "learning_rate": 0.003, + "loss": 4.0947, + "step": 6403 + }, + { + "epoch": 0.06404, + "grad_norm": 0.46851547499916696, + "learning_rate": 0.003, + "loss": 4.1495, + "step": 6404 + }, + { + "epoch": 0.06405, + "grad_norm": 0.4343792187108702, + "learning_rate": 0.003, + "loss": 4.1047, + "step": 6405 + }, + { + "epoch": 0.06406, + "grad_norm": 0.4387967998008624, + "learning_rate": 0.003, + "loss": 4.0936, + "step": 6406 + }, + { + "epoch": 0.06407, + "grad_norm": 0.3763862974231301, + "learning_rate": 0.003, + "loss": 4.0925, + "step": 6407 + }, + { + "epoch": 0.06408, + "grad_norm": 0.3618832104359289, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 6408 + }, + { + "epoch": 0.06409, + "grad_norm": 0.3640424177742778, + "learning_rate": 0.003, + "loss": 4.1159, + "step": 6409 + }, + { + "epoch": 0.0641, + "grad_norm": 0.5071096861354427, + "learning_rate": 0.003, + "loss": 4.1073, + "step": 6410 + }, + { + "epoch": 0.06411, + "grad_norm": 0.7056594230315663, + "learning_rate": 0.003, + "loss": 4.1014, + "step": 6411 + }, + { + "epoch": 0.06412, + "grad_norm": 1.0514700079261672, + "learning_rate": 0.003, + "loss": 4.1031, + "step": 6412 + }, + { + "epoch": 0.06413, + "grad_norm": 1.4637822933171232, + "learning_rate": 0.003, + "loss": 4.1524, + "step": 6413 + }, + { + "epoch": 0.06414, + "grad_norm": 0.6368174491993011, + "learning_rate": 0.003, + "loss": 4.0935, + "step": 6414 + }, + { + "epoch": 0.06415, + "grad_norm": 0.6864691858033185, + "learning_rate": 0.003, + "loss": 4.0922, + "step": 6415 + }, + { + "epoch": 0.06416, + "grad_norm": 0.6967129148461769, + "learning_rate": 0.003, + "loss": 4.13, + "step": 6416 + }, + { + "epoch": 0.06417, + "grad_norm": 0.7877830132689768, + "learning_rate": 0.003, + "loss": 4.1027, + "step": 6417 + }, + { + "epoch": 0.06418, + "grad_norm": 1.1025848418056756, + "learning_rate": 0.003, + "loss": 4.0996, + "step": 6418 + }, + { + "epoch": 0.06419, + "grad_norm": 1.1303087225349653, + "learning_rate": 0.003, + "loss": 4.1096, + "step": 6419 + }, + { + "epoch": 0.0642, + "grad_norm": 0.8179128267985711, + "learning_rate": 0.003, + "loss": 4.1156, + "step": 6420 + }, + { + "epoch": 0.06421, + "grad_norm": 0.7061052538423624, + "learning_rate": 0.003, + "loss": 4.1237, + "step": 6421 + }, + { + "epoch": 0.06422, + "grad_norm": 0.6712943095898026, + "learning_rate": 0.003, + "loss": 4.1112, + "step": 6422 + }, + { + "epoch": 0.06423, + "grad_norm": 0.7842157557770016, + "learning_rate": 0.003, + "loss": 4.1304, + "step": 6423 + }, + { + "epoch": 0.06424, + "grad_norm": 0.8391508667334445, + "learning_rate": 0.003, + "loss": 4.0946, + "step": 6424 + }, + { + "epoch": 0.06425, + "grad_norm": 0.8032559980871449, + "learning_rate": 0.003, + "loss": 4.1154, + "step": 6425 + }, + { + "epoch": 0.06426, + "grad_norm": 0.842568662503113, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 6426 + }, + { + "epoch": 0.06427, + "grad_norm": 0.9659133641186062, + "learning_rate": 0.003, + "loss": 4.1058, + "step": 6427 + }, + { + "epoch": 0.06428, + "grad_norm": 0.9129797056099707, + "learning_rate": 0.003, + "loss": 4.1309, + "step": 6428 + }, + { + "epoch": 0.06429, + "grad_norm": 0.8807646015484044, + "learning_rate": 0.003, + "loss": 4.1292, + "step": 6429 + }, + { + "epoch": 0.0643, + "grad_norm": 0.9116849659316084, + "learning_rate": 0.003, + "loss": 4.102, + "step": 6430 + }, + { + "epoch": 0.06431, + "grad_norm": 0.9401397808995475, + "learning_rate": 0.003, + "loss": 4.1161, + "step": 6431 + }, + { + "epoch": 0.06432, + "grad_norm": 1.094843463270304, + "learning_rate": 0.003, + "loss": 4.1145, + "step": 6432 + }, + { + "epoch": 0.06433, + "grad_norm": 0.97330153788876, + "learning_rate": 0.003, + "loss": 4.1417, + "step": 6433 + }, + { + "epoch": 0.06434, + "grad_norm": 0.8801080101022402, + "learning_rate": 0.003, + "loss": 4.0901, + "step": 6434 + }, + { + "epoch": 0.06435, + "grad_norm": 0.7207019820857952, + "learning_rate": 0.003, + "loss": 4.1034, + "step": 6435 + }, + { + "epoch": 0.06436, + "grad_norm": 0.7301567424764769, + "learning_rate": 0.003, + "loss": 4.1209, + "step": 6436 + }, + { + "epoch": 0.06437, + "grad_norm": 0.7242287759939515, + "learning_rate": 0.003, + "loss": 4.1168, + "step": 6437 + }, + { + "epoch": 0.06438, + "grad_norm": 0.6183084411636348, + "learning_rate": 0.003, + "loss": 4.1088, + "step": 6438 + }, + { + "epoch": 0.06439, + "grad_norm": 0.6942308219572709, + "learning_rate": 0.003, + "loss": 4.1093, + "step": 6439 + }, + { + "epoch": 0.0644, + "grad_norm": 0.6502050855985386, + "learning_rate": 0.003, + "loss": 4.1023, + "step": 6440 + }, + { + "epoch": 0.06441, + "grad_norm": 0.6169845208363806, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 6441 + }, + { + "epoch": 0.06442, + "grad_norm": 0.6761282291621807, + "learning_rate": 0.003, + "loss": 4.0981, + "step": 6442 + }, + { + "epoch": 0.06443, + "grad_norm": 0.7049850390444782, + "learning_rate": 0.003, + "loss": 4.0903, + "step": 6443 + }, + { + "epoch": 0.06444, + "grad_norm": 0.7982502738327323, + "learning_rate": 0.003, + "loss": 4.1005, + "step": 6444 + }, + { + "epoch": 0.06445, + "grad_norm": 0.8849927446065019, + "learning_rate": 0.003, + "loss": 4.1174, + "step": 6445 + }, + { + "epoch": 0.06446, + "grad_norm": 0.9290562203347035, + "learning_rate": 0.003, + "loss": 4.1066, + "step": 6446 + }, + { + "epoch": 0.06447, + "grad_norm": 0.9155039355826416, + "learning_rate": 0.003, + "loss": 4.1222, + "step": 6447 + }, + { + "epoch": 0.06448, + "grad_norm": 0.915311592818687, + "learning_rate": 0.003, + "loss": 4.1287, + "step": 6448 + }, + { + "epoch": 0.06449, + "grad_norm": 0.8915234617599778, + "learning_rate": 0.003, + "loss": 4.1577, + "step": 6449 + }, + { + "epoch": 0.0645, + "grad_norm": 0.8516263875404035, + "learning_rate": 0.003, + "loss": 4.1008, + "step": 6450 + }, + { + "epoch": 0.06451, + "grad_norm": 0.9911826973450533, + "learning_rate": 0.003, + "loss": 4.1539, + "step": 6451 + }, + { + "epoch": 0.06452, + "grad_norm": 1.0818621297011708, + "learning_rate": 0.003, + "loss": 4.1198, + "step": 6452 + }, + { + "epoch": 0.06453, + "grad_norm": 0.9321332755148247, + "learning_rate": 0.003, + "loss": 4.1578, + "step": 6453 + }, + { + "epoch": 0.06454, + "grad_norm": 0.850408071613158, + "learning_rate": 0.003, + "loss": 4.1097, + "step": 6454 + }, + { + "epoch": 0.06455, + "grad_norm": 0.8713116964180613, + "learning_rate": 0.003, + "loss": 4.1244, + "step": 6455 + }, + { + "epoch": 0.06456, + "grad_norm": 0.7964782772027043, + "learning_rate": 0.003, + "loss": 4.1358, + "step": 6456 + }, + { + "epoch": 0.06457, + "grad_norm": 0.8045164932744877, + "learning_rate": 0.003, + "loss": 4.1174, + "step": 6457 + }, + { + "epoch": 0.06458, + "grad_norm": 0.8513566837706239, + "learning_rate": 0.003, + "loss": 4.1279, + "step": 6458 + }, + { + "epoch": 0.06459, + "grad_norm": 0.8890380816261108, + "learning_rate": 0.003, + "loss": 4.1399, + "step": 6459 + }, + { + "epoch": 0.0646, + "grad_norm": 0.9142267341528604, + "learning_rate": 0.003, + "loss": 4.1163, + "step": 6460 + }, + { + "epoch": 0.06461, + "grad_norm": 1.1972918407769095, + "learning_rate": 0.003, + "loss": 4.1696, + "step": 6461 + }, + { + "epoch": 0.06462, + "grad_norm": 1.0336978547847568, + "learning_rate": 0.003, + "loss": 4.1123, + "step": 6462 + }, + { + "epoch": 0.06463, + "grad_norm": 1.0122741524242997, + "learning_rate": 0.003, + "loss": 4.1167, + "step": 6463 + }, + { + "epoch": 0.06464, + "grad_norm": 1.0493974717346164, + "learning_rate": 0.003, + "loss": 4.113, + "step": 6464 + }, + { + "epoch": 0.06465, + "grad_norm": 0.9438301613214977, + "learning_rate": 0.003, + "loss": 4.1057, + "step": 6465 + }, + { + "epoch": 0.06466, + "grad_norm": 0.8498735851960341, + "learning_rate": 0.003, + "loss": 4.1219, + "step": 6466 + }, + { + "epoch": 0.06467, + "grad_norm": 0.8190041939424388, + "learning_rate": 0.003, + "loss": 4.1088, + "step": 6467 + }, + { + "epoch": 0.06468, + "grad_norm": 0.8403763457939923, + "learning_rate": 0.003, + "loss": 4.1341, + "step": 6468 + }, + { + "epoch": 0.06469, + "grad_norm": 0.7964142979953805, + "learning_rate": 0.003, + "loss": 4.1313, + "step": 6469 + }, + { + "epoch": 0.0647, + "grad_norm": 0.7347258244702751, + "learning_rate": 0.003, + "loss": 4.1372, + "step": 6470 + }, + { + "epoch": 0.06471, + "grad_norm": 0.6129309594714507, + "learning_rate": 0.003, + "loss": 4.1107, + "step": 6471 + }, + { + "epoch": 0.06472, + "grad_norm": 0.6076460566383186, + "learning_rate": 0.003, + "loss": 4.1204, + "step": 6472 + }, + { + "epoch": 0.06473, + "grad_norm": 0.5891815040105876, + "learning_rate": 0.003, + "loss": 4.125, + "step": 6473 + }, + { + "epoch": 0.06474, + "grad_norm": 0.5162733573289747, + "learning_rate": 0.003, + "loss": 4.1398, + "step": 6474 + }, + { + "epoch": 0.06475, + "grad_norm": 0.5962513852003754, + "learning_rate": 0.003, + "loss": 4.1372, + "step": 6475 + }, + { + "epoch": 0.06476, + "grad_norm": 0.7146539756300717, + "learning_rate": 0.003, + "loss": 4.1091, + "step": 6476 + }, + { + "epoch": 0.06477, + "grad_norm": 0.8105712416755565, + "learning_rate": 0.003, + "loss": 4.0953, + "step": 6477 + }, + { + "epoch": 0.06478, + "grad_norm": 0.9652008011579472, + "learning_rate": 0.003, + "loss": 4.088, + "step": 6478 + }, + { + "epoch": 0.06479, + "grad_norm": 1.1900520557936207, + "learning_rate": 0.003, + "loss": 4.1137, + "step": 6479 + }, + { + "epoch": 0.0648, + "grad_norm": 0.7249331917094521, + "learning_rate": 0.003, + "loss": 4.0984, + "step": 6480 + }, + { + "epoch": 0.06481, + "grad_norm": 0.6534504012207452, + "learning_rate": 0.003, + "loss": 4.1079, + "step": 6481 + }, + { + "epoch": 0.06482, + "grad_norm": 0.7458945563014072, + "learning_rate": 0.003, + "loss": 4.1061, + "step": 6482 + }, + { + "epoch": 0.06483, + "grad_norm": 0.909230494347854, + "learning_rate": 0.003, + "loss": 4.1186, + "step": 6483 + }, + { + "epoch": 0.06484, + "grad_norm": 0.9786701503465991, + "learning_rate": 0.003, + "loss": 4.1186, + "step": 6484 + }, + { + "epoch": 0.06485, + "grad_norm": 0.8056031175729698, + "learning_rate": 0.003, + "loss": 4.1381, + "step": 6485 + }, + { + "epoch": 0.06486, + "grad_norm": 0.7315640607458516, + "learning_rate": 0.003, + "loss": 4.1369, + "step": 6486 + }, + { + "epoch": 0.06487, + "grad_norm": 0.738123736009974, + "learning_rate": 0.003, + "loss": 4.124, + "step": 6487 + }, + { + "epoch": 0.06488, + "grad_norm": 0.7016088617855374, + "learning_rate": 0.003, + "loss": 4.1303, + "step": 6488 + }, + { + "epoch": 0.06489, + "grad_norm": 0.7706221605393652, + "learning_rate": 0.003, + "loss": 4.1173, + "step": 6489 + }, + { + "epoch": 0.0649, + "grad_norm": 0.8075185716569853, + "learning_rate": 0.003, + "loss": 4.1072, + "step": 6490 + }, + { + "epoch": 0.06491, + "grad_norm": 0.7917671230247577, + "learning_rate": 0.003, + "loss": 4.1121, + "step": 6491 + }, + { + "epoch": 0.06492, + "grad_norm": 0.851038360366963, + "learning_rate": 0.003, + "loss": 4.128, + "step": 6492 + }, + { + "epoch": 0.06493, + "grad_norm": 0.8479749542932367, + "learning_rate": 0.003, + "loss": 4.1155, + "step": 6493 + }, + { + "epoch": 0.06494, + "grad_norm": 0.7857228913822375, + "learning_rate": 0.003, + "loss": 4.1237, + "step": 6494 + }, + { + "epoch": 0.06495, + "grad_norm": 0.7000090472611841, + "learning_rate": 0.003, + "loss": 4.1133, + "step": 6495 + }, + { + "epoch": 0.06496, + "grad_norm": 0.6189238416133499, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 6496 + }, + { + "epoch": 0.06497, + "grad_norm": 0.5745909145349337, + "learning_rate": 0.003, + "loss": 4.1012, + "step": 6497 + }, + { + "epoch": 0.06498, + "grad_norm": 0.5294157379174294, + "learning_rate": 0.003, + "loss": 4.0934, + "step": 6498 + }, + { + "epoch": 0.06499, + "grad_norm": 0.5466886858120876, + "learning_rate": 0.003, + "loss": 4.1253, + "step": 6499 + }, + { + "epoch": 0.065, + "grad_norm": 0.6539813137699607, + "learning_rate": 0.003, + "loss": 4.1022, + "step": 6500 + }, + { + "epoch": 0.06501, + "grad_norm": 0.6647161266915989, + "learning_rate": 0.003, + "loss": 4.0973, + "step": 6501 + }, + { + "epoch": 0.06502, + "grad_norm": 0.6443788622041227, + "learning_rate": 0.003, + "loss": 4.1054, + "step": 6502 + }, + { + "epoch": 0.06503, + "grad_norm": 0.7571231315947718, + "learning_rate": 0.003, + "loss": 4.1221, + "step": 6503 + }, + { + "epoch": 0.06504, + "grad_norm": 0.9637682445146619, + "learning_rate": 0.003, + "loss": 4.1063, + "step": 6504 + }, + { + "epoch": 0.06505, + "grad_norm": 1.3505877682488592, + "learning_rate": 0.003, + "loss": 4.1153, + "step": 6505 + }, + { + "epoch": 0.06506, + "grad_norm": 0.7210710726778544, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 6506 + }, + { + "epoch": 0.06507, + "grad_norm": 0.8243540456063395, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 6507 + }, + { + "epoch": 0.06508, + "grad_norm": 0.9264270578051855, + "learning_rate": 0.003, + "loss": 4.1089, + "step": 6508 + }, + { + "epoch": 0.06509, + "grad_norm": 0.8402439484381861, + "learning_rate": 0.003, + "loss": 4.0974, + "step": 6509 + }, + { + "epoch": 0.0651, + "grad_norm": 0.8452797474799149, + "learning_rate": 0.003, + "loss": 4.1355, + "step": 6510 + }, + { + "epoch": 0.06511, + "grad_norm": 0.8155771099505443, + "learning_rate": 0.003, + "loss": 4.099, + "step": 6511 + }, + { + "epoch": 0.06512, + "grad_norm": 0.8504514854470253, + "learning_rate": 0.003, + "loss": 4.0886, + "step": 6512 + }, + { + "epoch": 0.06513, + "grad_norm": 0.7996056231914204, + "learning_rate": 0.003, + "loss": 4.1182, + "step": 6513 + }, + { + "epoch": 0.06514, + "grad_norm": 0.8483202440944669, + "learning_rate": 0.003, + "loss": 4.1088, + "step": 6514 + }, + { + "epoch": 0.06515, + "grad_norm": 0.729647231890123, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 6515 + }, + { + "epoch": 0.06516, + "grad_norm": 0.7668262347734042, + "learning_rate": 0.003, + "loss": 4.1125, + "step": 6516 + }, + { + "epoch": 0.06517, + "grad_norm": 0.9668803624817275, + "learning_rate": 0.003, + "loss": 4.1157, + "step": 6517 + }, + { + "epoch": 0.06518, + "grad_norm": 0.9221337117494466, + "learning_rate": 0.003, + "loss": 4.1395, + "step": 6518 + }, + { + "epoch": 0.06519, + "grad_norm": 0.8991712754647655, + "learning_rate": 0.003, + "loss": 4.1151, + "step": 6519 + }, + { + "epoch": 0.0652, + "grad_norm": 0.9996762890374128, + "learning_rate": 0.003, + "loss": 4.0873, + "step": 6520 + }, + { + "epoch": 0.06521, + "grad_norm": 0.9694147573902756, + "learning_rate": 0.003, + "loss": 4.1385, + "step": 6521 + }, + { + "epoch": 0.06522, + "grad_norm": 1.1436487696458615, + "learning_rate": 0.003, + "loss": 4.1504, + "step": 6522 + }, + { + "epoch": 0.06523, + "grad_norm": 1.1761311790032607, + "learning_rate": 0.003, + "loss": 4.1357, + "step": 6523 + }, + { + "epoch": 0.06524, + "grad_norm": 0.9935715060905626, + "learning_rate": 0.003, + "loss": 4.1267, + "step": 6524 + }, + { + "epoch": 0.06525, + "grad_norm": 1.0984134338267004, + "learning_rate": 0.003, + "loss": 4.1515, + "step": 6525 + }, + { + "epoch": 0.06526, + "grad_norm": 0.8597157647493189, + "learning_rate": 0.003, + "loss": 4.1287, + "step": 6526 + }, + { + "epoch": 0.06527, + "grad_norm": 0.9122953325479415, + "learning_rate": 0.003, + "loss": 4.1405, + "step": 6527 + }, + { + "epoch": 0.06528, + "grad_norm": 0.9550047890046722, + "learning_rate": 0.003, + "loss": 4.1386, + "step": 6528 + }, + { + "epoch": 0.06529, + "grad_norm": 0.9456502240272384, + "learning_rate": 0.003, + "loss": 4.1265, + "step": 6529 + }, + { + "epoch": 0.0653, + "grad_norm": 0.8144448232250512, + "learning_rate": 0.003, + "loss": 4.1182, + "step": 6530 + }, + { + "epoch": 0.06531, + "grad_norm": 0.7884345470232, + "learning_rate": 0.003, + "loss": 4.1073, + "step": 6531 + }, + { + "epoch": 0.06532, + "grad_norm": 0.7546346553317457, + "learning_rate": 0.003, + "loss": 4.1361, + "step": 6532 + }, + { + "epoch": 0.06533, + "grad_norm": 0.6337607386448925, + "learning_rate": 0.003, + "loss": 4.1334, + "step": 6533 + }, + { + "epoch": 0.06534, + "grad_norm": 0.6571305549922228, + "learning_rate": 0.003, + "loss": 4.1256, + "step": 6534 + }, + { + "epoch": 0.06535, + "grad_norm": 0.6727892441041884, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 6535 + }, + { + "epoch": 0.06536, + "grad_norm": 0.6796596200367593, + "learning_rate": 0.003, + "loss": 4.1024, + "step": 6536 + }, + { + "epoch": 0.06537, + "grad_norm": 0.6850578598200898, + "learning_rate": 0.003, + "loss": 4.1186, + "step": 6537 + }, + { + "epoch": 0.06538, + "grad_norm": 0.6903169559185911, + "learning_rate": 0.003, + "loss": 4.1232, + "step": 6538 + }, + { + "epoch": 0.06539, + "grad_norm": 0.8869816313052468, + "learning_rate": 0.003, + "loss": 4.0958, + "step": 6539 + }, + { + "epoch": 0.0654, + "grad_norm": 1.168073515327997, + "learning_rate": 0.003, + "loss": 4.1617, + "step": 6540 + }, + { + "epoch": 0.06541, + "grad_norm": 0.7182356963906228, + "learning_rate": 0.003, + "loss": 4.1144, + "step": 6541 + }, + { + "epoch": 0.06542, + "grad_norm": 0.6188811083162069, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 6542 + }, + { + "epoch": 0.06543, + "grad_norm": 0.8051980596249964, + "learning_rate": 0.003, + "loss": 4.1256, + "step": 6543 + }, + { + "epoch": 0.06544, + "grad_norm": 0.9132061876425465, + "learning_rate": 0.003, + "loss": 4.1202, + "step": 6544 + }, + { + "epoch": 0.06545, + "grad_norm": 0.9233000070732396, + "learning_rate": 0.003, + "loss": 4.1273, + "step": 6545 + }, + { + "epoch": 0.06546, + "grad_norm": 0.7831417568832402, + "learning_rate": 0.003, + "loss": 4.1092, + "step": 6546 + }, + { + "epoch": 0.06547, + "grad_norm": 0.6532883144447844, + "learning_rate": 0.003, + "loss": 4.1072, + "step": 6547 + }, + { + "epoch": 0.06548, + "grad_norm": 0.6727801173157394, + "learning_rate": 0.003, + "loss": 4.107, + "step": 6548 + }, + { + "epoch": 0.06549, + "grad_norm": 0.6217127018777375, + "learning_rate": 0.003, + "loss": 4.1333, + "step": 6549 + }, + { + "epoch": 0.0655, + "grad_norm": 0.7089420407652894, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 6550 + }, + { + "epoch": 0.06551, + "grad_norm": 0.8434472089359629, + "learning_rate": 0.003, + "loss": 4.093, + "step": 6551 + }, + { + "epoch": 0.06552, + "grad_norm": 0.957288546390429, + "learning_rate": 0.003, + "loss": 4.1361, + "step": 6552 + }, + { + "epoch": 0.06553, + "grad_norm": 0.8538606588662652, + "learning_rate": 0.003, + "loss": 4.1126, + "step": 6553 + }, + { + "epoch": 0.06554, + "grad_norm": 0.6821331682783609, + "learning_rate": 0.003, + "loss": 4.1141, + "step": 6554 + }, + { + "epoch": 0.06555, + "grad_norm": 0.5970317370800081, + "learning_rate": 0.003, + "loss": 4.1065, + "step": 6555 + }, + { + "epoch": 0.06556, + "grad_norm": 0.6339812838968247, + "learning_rate": 0.003, + "loss": 4.1253, + "step": 6556 + }, + { + "epoch": 0.06557, + "grad_norm": 0.7300737435789836, + "learning_rate": 0.003, + "loss": 4.1005, + "step": 6557 + }, + { + "epoch": 0.06558, + "grad_norm": 0.9340751362463283, + "learning_rate": 0.003, + "loss": 4.1393, + "step": 6558 + }, + { + "epoch": 0.06559, + "grad_norm": 1.078857392761123, + "learning_rate": 0.003, + "loss": 4.1249, + "step": 6559 + }, + { + "epoch": 0.0656, + "grad_norm": 0.7783300852395647, + "learning_rate": 0.003, + "loss": 4.1161, + "step": 6560 + }, + { + "epoch": 0.06561, + "grad_norm": 0.6838855361080844, + "learning_rate": 0.003, + "loss": 4.1294, + "step": 6561 + }, + { + "epoch": 0.06562, + "grad_norm": 0.8236313339666234, + "learning_rate": 0.003, + "loss": 4.0906, + "step": 6562 + }, + { + "epoch": 0.06563, + "grad_norm": 0.9128215836077087, + "learning_rate": 0.003, + "loss": 4.113, + "step": 6563 + }, + { + "epoch": 0.06564, + "grad_norm": 1.0183713945436557, + "learning_rate": 0.003, + "loss": 4.123, + "step": 6564 + }, + { + "epoch": 0.06565, + "grad_norm": 0.9875484143170253, + "learning_rate": 0.003, + "loss": 4.0936, + "step": 6565 + }, + { + "epoch": 0.06566, + "grad_norm": 0.8482203740641269, + "learning_rate": 0.003, + "loss": 4.1195, + "step": 6566 + }, + { + "epoch": 0.06567, + "grad_norm": 0.8887069547124475, + "learning_rate": 0.003, + "loss": 4.1341, + "step": 6567 + }, + { + "epoch": 0.06568, + "grad_norm": 1.0056485677045872, + "learning_rate": 0.003, + "loss": 4.1463, + "step": 6568 + }, + { + "epoch": 0.06569, + "grad_norm": 0.9064862101322059, + "learning_rate": 0.003, + "loss": 4.0967, + "step": 6569 + }, + { + "epoch": 0.0657, + "grad_norm": 0.7555124162684835, + "learning_rate": 0.003, + "loss": 4.1195, + "step": 6570 + }, + { + "epoch": 0.06571, + "grad_norm": 0.7348370653563371, + "learning_rate": 0.003, + "loss": 4.1252, + "step": 6571 + }, + { + "epoch": 0.06572, + "grad_norm": 0.8152888496093048, + "learning_rate": 0.003, + "loss": 4.0918, + "step": 6572 + }, + { + "epoch": 0.06573, + "grad_norm": 0.7572678952633992, + "learning_rate": 0.003, + "loss": 4.1201, + "step": 6573 + }, + { + "epoch": 0.06574, + "grad_norm": 0.7173617049609856, + "learning_rate": 0.003, + "loss": 4.1087, + "step": 6574 + }, + { + "epoch": 0.06575, + "grad_norm": 0.7216202084765286, + "learning_rate": 0.003, + "loss": 4.1017, + "step": 6575 + }, + { + "epoch": 0.06576, + "grad_norm": 0.7321185248183836, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 6576 + }, + { + "epoch": 0.06577, + "grad_norm": 0.812337799951959, + "learning_rate": 0.003, + "loss": 4.0993, + "step": 6577 + }, + { + "epoch": 0.06578, + "grad_norm": 1.0097566236813522, + "learning_rate": 0.003, + "loss": 4.1116, + "step": 6578 + }, + { + "epoch": 0.06579, + "grad_norm": 1.1289060393565276, + "learning_rate": 0.003, + "loss": 4.1044, + "step": 6579 + }, + { + "epoch": 0.0658, + "grad_norm": 1.1236108631521293, + "learning_rate": 0.003, + "loss": 4.0992, + "step": 6580 + }, + { + "epoch": 0.06581, + "grad_norm": 0.9095208566603907, + "learning_rate": 0.003, + "loss": 4.1357, + "step": 6581 + }, + { + "epoch": 0.06582, + "grad_norm": 0.7078312720167307, + "learning_rate": 0.003, + "loss": 4.1261, + "step": 6582 + }, + { + "epoch": 0.06583, + "grad_norm": 0.6250796377300359, + "learning_rate": 0.003, + "loss": 4.0935, + "step": 6583 + }, + { + "epoch": 0.06584, + "grad_norm": 0.6182361788610296, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 6584 + }, + { + "epoch": 0.06585, + "grad_norm": 0.6199101947792578, + "learning_rate": 0.003, + "loss": 4.1048, + "step": 6585 + }, + { + "epoch": 0.06586, + "grad_norm": 0.644608198538537, + "learning_rate": 0.003, + "loss": 4.1162, + "step": 6586 + }, + { + "epoch": 0.06587, + "grad_norm": 0.7531896129898413, + "learning_rate": 0.003, + "loss": 4.1225, + "step": 6587 + }, + { + "epoch": 0.06588, + "grad_norm": 0.9463136908207883, + "learning_rate": 0.003, + "loss": 4.1466, + "step": 6588 + }, + { + "epoch": 0.06589, + "grad_norm": 1.0611698029062901, + "learning_rate": 0.003, + "loss": 4.1384, + "step": 6589 + }, + { + "epoch": 0.0659, + "grad_norm": 0.9951512526105761, + "learning_rate": 0.003, + "loss": 4.1488, + "step": 6590 + }, + { + "epoch": 0.06591, + "grad_norm": 0.9622203108833793, + "learning_rate": 0.003, + "loss": 4.1735, + "step": 6591 + }, + { + "epoch": 0.06592, + "grad_norm": 0.7792553434807309, + "learning_rate": 0.003, + "loss": 4.1311, + "step": 6592 + }, + { + "epoch": 0.06593, + "grad_norm": 0.7097135195972493, + "learning_rate": 0.003, + "loss": 4.1091, + "step": 6593 + }, + { + "epoch": 0.06594, + "grad_norm": 0.7407425345106201, + "learning_rate": 0.003, + "loss": 4.1303, + "step": 6594 + }, + { + "epoch": 0.06595, + "grad_norm": 0.7612553261817883, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 6595 + }, + { + "epoch": 0.06596, + "grad_norm": 0.7713906419670148, + "learning_rate": 0.003, + "loss": 4.1343, + "step": 6596 + }, + { + "epoch": 0.06597, + "grad_norm": 0.95305210338701, + "learning_rate": 0.003, + "loss": 4.1171, + "step": 6597 + }, + { + "epoch": 0.06598, + "grad_norm": 0.9912085601243025, + "learning_rate": 0.003, + "loss": 4.1523, + "step": 6598 + }, + { + "epoch": 0.06599, + "grad_norm": 0.9563843538129239, + "learning_rate": 0.003, + "loss": 4.1346, + "step": 6599 + }, + { + "epoch": 0.066, + "grad_norm": 0.9173651825854763, + "learning_rate": 0.003, + "loss": 4.1429, + "step": 6600 + }, + { + "epoch": 0.06601, + "grad_norm": 0.8706761413967128, + "learning_rate": 0.003, + "loss": 4.1266, + "step": 6601 + }, + { + "epoch": 0.06602, + "grad_norm": 0.8530638651302136, + "learning_rate": 0.003, + "loss": 4.1244, + "step": 6602 + }, + { + "epoch": 0.06603, + "grad_norm": 0.7620877613851017, + "learning_rate": 0.003, + "loss": 4.1155, + "step": 6603 + }, + { + "epoch": 0.06604, + "grad_norm": 0.7157412127972154, + "learning_rate": 0.003, + "loss": 4.1114, + "step": 6604 + }, + { + "epoch": 0.06605, + "grad_norm": 0.7664275284896481, + "learning_rate": 0.003, + "loss": 4.1325, + "step": 6605 + }, + { + "epoch": 0.06606, + "grad_norm": 0.8596918832087841, + "learning_rate": 0.003, + "loss": 4.119, + "step": 6606 + }, + { + "epoch": 0.06607, + "grad_norm": 0.9733248442463999, + "learning_rate": 0.003, + "loss": 4.1112, + "step": 6607 + }, + { + "epoch": 0.06608, + "grad_norm": 1.261129457668947, + "learning_rate": 0.003, + "loss": 4.1147, + "step": 6608 + }, + { + "epoch": 0.06609, + "grad_norm": 0.7875463153702758, + "learning_rate": 0.003, + "loss": 4.123, + "step": 6609 + }, + { + "epoch": 0.0661, + "grad_norm": 0.6418585115175501, + "learning_rate": 0.003, + "loss": 4.1225, + "step": 6610 + }, + { + "epoch": 0.06611, + "grad_norm": 0.6118831582806695, + "learning_rate": 0.003, + "loss": 4.1074, + "step": 6611 + }, + { + "epoch": 0.06612, + "grad_norm": 0.6131198639857838, + "learning_rate": 0.003, + "loss": 4.1031, + "step": 6612 + }, + { + "epoch": 0.06613, + "grad_norm": 0.605891802900793, + "learning_rate": 0.003, + "loss": 4.1184, + "step": 6613 + }, + { + "epoch": 0.06614, + "grad_norm": 0.7122334128472662, + "learning_rate": 0.003, + "loss": 4.1152, + "step": 6614 + }, + { + "epoch": 0.06615, + "grad_norm": 0.986281825824359, + "learning_rate": 0.003, + "loss": 4.1009, + "step": 6615 + }, + { + "epoch": 0.06616, + "grad_norm": 1.209509744232019, + "learning_rate": 0.003, + "loss": 4.1629, + "step": 6616 + }, + { + "epoch": 0.06617, + "grad_norm": 0.589186192111392, + "learning_rate": 0.003, + "loss": 4.126, + "step": 6617 + }, + { + "epoch": 0.06618, + "grad_norm": 0.817010663602591, + "learning_rate": 0.003, + "loss": 4.1429, + "step": 6618 + }, + { + "epoch": 0.06619, + "grad_norm": 1.1163167401139134, + "learning_rate": 0.003, + "loss": 4.1143, + "step": 6619 + }, + { + "epoch": 0.0662, + "grad_norm": 0.861557189675986, + "learning_rate": 0.003, + "loss": 4.101, + "step": 6620 + }, + { + "epoch": 0.06621, + "grad_norm": 0.7734302388169766, + "learning_rate": 0.003, + "loss": 4.1324, + "step": 6621 + }, + { + "epoch": 0.06622, + "grad_norm": 0.6681446519688905, + "learning_rate": 0.003, + "loss": 4.1237, + "step": 6622 + }, + { + "epoch": 0.06623, + "grad_norm": 0.6086441119859902, + "learning_rate": 0.003, + "loss": 4.1052, + "step": 6623 + }, + { + "epoch": 0.06624, + "grad_norm": 0.6161665012708084, + "learning_rate": 0.003, + "loss": 4.126, + "step": 6624 + }, + { + "epoch": 0.06625, + "grad_norm": 0.6016793043105603, + "learning_rate": 0.003, + "loss": 4.1109, + "step": 6625 + }, + { + "epoch": 0.06626, + "grad_norm": 0.5731558676794132, + "learning_rate": 0.003, + "loss": 4.129, + "step": 6626 + }, + { + "epoch": 0.06627, + "grad_norm": 0.5748801713145845, + "learning_rate": 0.003, + "loss": 4.1173, + "step": 6627 + }, + { + "epoch": 0.06628, + "grad_norm": 0.5724674668619956, + "learning_rate": 0.003, + "loss": 4.1197, + "step": 6628 + }, + { + "epoch": 0.06629, + "grad_norm": 0.5329925465369826, + "learning_rate": 0.003, + "loss": 4.1118, + "step": 6629 + }, + { + "epoch": 0.0663, + "grad_norm": 0.4945919230759341, + "learning_rate": 0.003, + "loss": 4.1287, + "step": 6630 + }, + { + "epoch": 0.06631, + "grad_norm": 0.45658695378096065, + "learning_rate": 0.003, + "loss": 4.1026, + "step": 6631 + }, + { + "epoch": 0.06632, + "grad_norm": 0.526478004374339, + "learning_rate": 0.003, + "loss": 4.1042, + "step": 6632 + }, + { + "epoch": 0.06633, + "grad_norm": 0.6229089968891325, + "learning_rate": 0.003, + "loss": 4.1235, + "step": 6633 + }, + { + "epoch": 0.06634, + "grad_norm": 0.7440020938095453, + "learning_rate": 0.003, + "loss": 4.1129, + "step": 6634 + }, + { + "epoch": 0.06635, + "grad_norm": 1.0497670831766566, + "learning_rate": 0.003, + "loss": 4.1378, + "step": 6635 + }, + { + "epoch": 0.06636, + "grad_norm": 1.14053551502134, + "learning_rate": 0.003, + "loss": 4.0949, + "step": 6636 + }, + { + "epoch": 0.06637, + "grad_norm": 0.7502849737062908, + "learning_rate": 0.003, + "loss": 4.112, + "step": 6637 + }, + { + "epoch": 0.06638, + "grad_norm": 0.8150413391233561, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 6638 + }, + { + "epoch": 0.06639, + "grad_norm": 0.9898719461642121, + "learning_rate": 0.003, + "loss": 4.1258, + "step": 6639 + }, + { + "epoch": 0.0664, + "grad_norm": 1.1511766317227263, + "learning_rate": 0.003, + "loss": 4.1092, + "step": 6640 + }, + { + "epoch": 0.06641, + "grad_norm": 0.8614640565436265, + "learning_rate": 0.003, + "loss": 4.1251, + "step": 6641 + }, + { + "epoch": 0.06642, + "grad_norm": 0.7977417905197698, + "learning_rate": 0.003, + "loss": 4.1256, + "step": 6642 + }, + { + "epoch": 0.06643, + "grad_norm": 0.7780801354815946, + "learning_rate": 0.003, + "loss": 4.114, + "step": 6643 + }, + { + "epoch": 0.06644, + "grad_norm": 0.7755477432783495, + "learning_rate": 0.003, + "loss": 4.1157, + "step": 6644 + }, + { + "epoch": 0.06645, + "grad_norm": 0.9119617892094356, + "learning_rate": 0.003, + "loss": 4.1069, + "step": 6645 + }, + { + "epoch": 0.06646, + "grad_norm": 0.878567897301363, + "learning_rate": 0.003, + "loss": 4.1099, + "step": 6646 + }, + { + "epoch": 0.06647, + "grad_norm": 0.8867140954494283, + "learning_rate": 0.003, + "loss": 4.1405, + "step": 6647 + }, + { + "epoch": 0.06648, + "grad_norm": 0.8336313880890538, + "learning_rate": 0.003, + "loss": 4.1093, + "step": 6648 + }, + { + "epoch": 0.06649, + "grad_norm": 0.8170814645370212, + "learning_rate": 0.003, + "loss": 4.1201, + "step": 6649 + }, + { + "epoch": 0.0665, + "grad_norm": 0.7834859433585468, + "learning_rate": 0.003, + "loss": 4.1168, + "step": 6650 + }, + { + "epoch": 0.06651, + "grad_norm": 0.8357692571639677, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 6651 + }, + { + "epoch": 0.06652, + "grad_norm": 0.9187801035749852, + "learning_rate": 0.003, + "loss": 4.1187, + "step": 6652 + }, + { + "epoch": 0.06653, + "grad_norm": 0.969813469239403, + "learning_rate": 0.003, + "loss": 4.1504, + "step": 6653 + }, + { + "epoch": 0.06654, + "grad_norm": 1.0521703710878825, + "learning_rate": 0.003, + "loss": 4.1142, + "step": 6654 + }, + { + "epoch": 0.06655, + "grad_norm": 1.115077498474384, + "learning_rate": 0.003, + "loss": 4.133, + "step": 6655 + }, + { + "epoch": 0.06656, + "grad_norm": 0.9993667641705132, + "learning_rate": 0.003, + "loss": 4.1122, + "step": 6656 + }, + { + "epoch": 0.06657, + "grad_norm": 1.1623793190867178, + "learning_rate": 0.003, + "loss": 4.138, + "step": 6657 + }, + { + "epoch": 0.06658, + "grad_norm": 0.8317171031791138, + "learning_rate": 0.003, + "loss": 4.1207, + "step": 6658 + }, + { + "epoch": 0.06659, + "grad_norm": 0.7417282289820908, + "learning_rate": 0.003, + "loss": 4.1346, + "step": 6659 + }, + { + "epoch": 0.0666, + "grad_norm": 0.7996438420969371, + "learning_rate": 0.003, + "loss": 4.1154, + "step": 6660 + }, + { + "epoch": 0.06661, + "grad_norm": 0.8239468144667403, + "learning_rate": 0.003, + "loss": 4.115, + "step": 6661 + }, + { + "epoch": 0.06662, + "grad_norm": 0.8636078960543694, + "learning_rate": 0.003, + "loss": 4.1075, + "step": 6662 + }, + { + "epoch": 0.06663, + "grad_norm": 0.8286265545981787, + "learning_rate": 0.003, + "loss": 4.1754, + "step": 6663 + }, + { + "epoch": 0.06664, + "grad_norm": 0.7760590887675644, + "learning_rate": 0.003, + "loss": 4.1319, + "step": 6664 + }, + { + "epoch": 0.06665, + "grad_norm": 0.6944243205061547, + "learning_rate": 0.003, + "loss": 4.1548, + "step": 6665 + }, + { + "epoch": 0.06666, + "grad_norm": 0.7180657446089408, + "learning_rate": 0.003, + "loss": 4.0845, + "step": 6666 + }, + { + "epoch": 0.06667, + "grad_norm": 0.8974376505212266, + "learning_rate": 0.003, + "loss": 4.1154, + "step": 6667 + }, + { + "epoch": 0.06668, + "grad_norm": 1.0172055502989037, + "learning_rate": 0.003, + "loss": 4.1119, + "step": 6668 + }, + { + "epoch": 0.06669, + "grad_norm": 1.1607157063808713, + "learning_rate": 0.003, + "loss": 4.1234, + "step": 6669 + }, + { + "epoch": 0.0667, + "grad_norm": 0.6558890868526093, + "learning_rate": 0.003, + "loss": 4.1248, + "step": 6670 + }, + { + "epoch": 0.06671, + "grad_norm": 0.5339830424514167, + "learning_rate": 0.003, + "loss": 4.1163, + "step": 6671 + }, + { + "epoch": 0.06672, + "grad_norm": 0.6960933681548969, + "learning_rate": 0.003, + "loss": 4.1, + "step": 6672 + }, + { + "epoch": 0.06673, + "grad_norm": 0.8613148280885923, + "learning_rate": 0.003, + "loss": 4.1033, + "step": 6673 + }, + { + "epoch": 0.06674, + "grad_norm": 1.0944320576632818, + "learning_rate": 0.003, + "loss": 4.136, + "step": 6674 + }, + { + "epoch": 0.06675, + "grad_norm": 0.8624252033051901, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 6675 + }, + { + "epoch": 0.06676, + "grad_norm": 0.6153729593690301, + "learning_rate": 0.003, + "loss": 4.1197, + "step": 6676 + }, + { + "epoch": 0.06677, + "grad_norm": 0.6138367390296136, + "learning_rate": 0.003, + "loss": 4.1283, + "step": 6677 + }, + { + "epoch": 0.06678, + "grad_norm": 0.6766082571603963, + "learning_rate": 0.003, + "loss": 4.1348, + "step": 6678 + }, + { + "epoch": 0.06679, + "grad_norm": 0.7509125610237447, + "learning_rate": 0.003, + "loss": 4.1065, + "step": 6679 + }, + { + "epoch": 0.0668, + "grad_norm": 0.7800006560160819, + "learning_rate": 0.003, + "loss": 4.1423, + "step": 6680 + }, + { + "epoch": 0.06681, + "grad_norm": 0.7276956103034433, + "learning_rate": 0.003, + "loss": 4.0959, + "step": 6681 + }, + { + "epoch": 0.06682, + "grad_norm": 0.7677356222531484, + "learning_rate": 0.003, + "loss": 4.1246, + "step": 6682 + }, + { + "epoch": 0.06683, + "grad_norm": 0.9103583128182188, + "learning_rate": 0.003, + "loss": 4.1183, + "step": 6683 + }, + { + "epoch": 0.06684, + "grad_norm": 0.9184243344640678, + "learning_rate": 0.003, + "loss": 4.1264, + "step": 6684 + }, + { + "epoch": 0.06685, + "grad_norm": 0.8953119859552124, + "learning_rate": 0.003, + "loss": 4.1215, + "step": 6685 + }, + { + "epoch": 0.06686, + "grad_norm": 0.8649576016834076, + "learning_rate": 0.003, + "loss": 4.1023, + "step": 6686 + }, + { + "epoch": 0.06687, + "grad_norm": 0.8737336563045488, + "learning_rate": 0.003, + "loss": 4.0922, + "step": 6687 + }, + { + "epoch": 0.06688, + "grad_norm": 0.8001454108494693, + "learning_rate": 0.003, + "loss": 4.1044, + "step": 6688 + }, + { + "epoch": 0.06689, + "grad_norm": 0.7007608354845569, + "learning_rate": 0.003, + "loss": 4.1377, + "step": 6689 + }, + { + "epoch": 0.0669, + "grad_norm": 0.7346334248821653, + "learning_rate": 0.003, + "loss": 4.1024, + "step": 6690 + }, + { + "epoch": 0.06691, + "grad_norm": 0.8242428244307782, + "learning_rate": 0.003, + "loss": 4.1227, + "step": 6691 + }, + { + "epoch": 0.06692, + "grad_norm": 0.8785059768462102, + "learning_rate": 0.003, + "loss": 4.1006, + "step": 6692 + }, + { + "epoch": 0.06693, + "grad_norm": 0.8419120527636385, + "learning_rate": 0.003, + "loss": 4.116, + "step": 6693 + }, + { + "epoch": 0.06694, + "grad_norm": 0.8364345690497847, + "learning_rate": 0.003, + "loss": 4.112, + "step": 6694 + }, + { + "epoch": 0.06695, + "grad_norm": 0.7282086608438694, + "learning_rate": 0.003, + "loss": 4.1101, + "step": 6695 + }, + { + "epoch": 0.06696, + "grad_norm": 0.7156637313733002, + "learning_rate": 0.003, + "loss": 4.1117, + "step": 6696 + }, + { + "epoch": 0.06697, + "grad_norm": 0.759716727374639, + "learning_rate": 0.003, + "loss": 4.1229, + "step": 6697 + }, + { + "epoch": 0.06698, + "grad_norm": 0.6814818122358287, + "learning_rate": 0.003, + "loss": 4.1075, + "step": 6698 + }, + { + "epoch": 0.06699, + "grad_norm": 0.8881481031361592, + "learning_rate": 0.003, + "loss": 4.1118, + "step": 6699 + }, + { + "epoch": 0.067, + "grad_norm": 0.916609985935385, + "learning_rate": 0.003, + "loss": 4.1344, + "step": 6700 + }, + { + "epoch": 0.06701, + "grad_norm": 0.7647456152832255, + "learning_rate": 0.003, + "loss": 4.1261, + "step": 6701 + }, + { + "epoch": 0.06702, + "grad_norm": 0.9878626082963966, + "learning_rate": 0.003, + "loss": 4.1189, + "step": 6702 + }, + { + "epoch": 0.06703, + "grad_norm": 1.450715682692916, + "learning_rate": 0.003, + "loss": 4.1498, + "step": 6703 + }, + { + "epoch": 0.06704, + "grad_norm": 0.8099658530003969, + "learning_rate": 0.003, + "loss": 4.1068, + "step": 6704 + }, + { + "epoch": 0.06705, + "grad_norm": 0.8003246468539195, + "learning_rate": 0.003, + "loss": 4.114, + "step": 6705 + }, + { + "epoch": 0.06706, + "grad_norm": 0.9277530689587093, + "learning_rate": 0.003, + "loss": 4.165, + "step": 6706 + }, + { + "epoch": 0.06707, + "grad_norm": 1.081334232913671, + "learning_rate": 0.003, + "loss": 4.1046, + "step": 6707 + }, + { + "epoch": 0.06708, + "grad_norm": 0.7824655188582949, + "learning_rate": 0.003, + "loss": 4.1134, + "step": 6708 + }, + { + "epoch": 0.06709, + "grad_norm": 0.6810909588960559, + "learning_rate": 0.003, + "loss": 4.1104, + "step": 6709 + }, + { + "epoch": 0.0671, + "grad_norm": 0.6175677027859562, + "learning_rate": 0.003, + "loss": 4.131, + "step": 6710 + }, + { + "epoch": 0.06711, + "grad_norm": 0.7248156251811534, + "learning_rate": 0.003, + "loss": 4.1407, + "step": 6711 + }, + { + "epoch": 0.06712, + "grad_norm": 0.9236044790837477, + "learning_rate": 0.003, + "loss": 4.0971, + "step": 6712 + }, + { + "epoch": 0.06713, + "grad_norm": 1.1656738764307082, + "learning_rate": 0.003, + "loss": 4.1193, + "step": 6713 + }, + { + "epoch": 0.06714, + "grad_norm": 0.687130908864893, + "learning_rate": 0.003, + "loss": 4.1121, + "step": 6714 + }, + { + "epoch": 0.06715, + "grad_norm": 0.6511308300062618, + "learning_rate": 0.003, + "loss": 4.1105, + "step": 6715 + }, + { + "epoch": 0.06716, + "grad_norm": 0.9028221528538252, + "learning_rate": 0.003, + "loss": 4.1228, + "step": 6716 + }, + { + "epoch": 0.06717, + "grad_norm": 0.9804042876410267, + "learning_rate": 0.003, + "loss": 4.1445, + "step": 6717 + }, + { + "epoch": 0.06718, + "grad_norm": 1.0173300549817423, + "learning_rate": 0.003, + "loss": 4.1064, + "step": 6718 + }, + { + "epoch": 0.06719, + "grad_norm": 0.9220403787562276, + "learning_rate": 0.003, + "loss": 4.1076, + "step": 6719 + }, + { + "epoch": 0.0672, + "grad_norm": 0.7514745228001316, + "learning_rate": 0.003, + "loss": 4.1207, + "step": 6720 + }, + { + "epoch": 0.06721, + "grad_norm": 0.8665408867149632, + "learning_rate": 0.003, + "loss": 4.134, + "step": 6721 + }, + { + "epoch": 0.06722, + "grad_norm": 0.9302053283399239, + "learning_rate": 0.003, + "loss": 4.1493, + "step": 6722 + }, + { + "epoch": 0.06723, + "grad_norm": 0.9301656966667754, + "learning_rate": 0.003, + "loss": 4.1395, + "step": 6723 + }, + { + "epoch": 0.06724, + "grad_norm": 0.8389657280038182, + "learning_rate": 0.003, + "loss": 4.099, + "step": 6724 + }, + { + "epoch": 0.06725, + "grad_norm": 0.7512173727545822, + "learning_rate": 0.003, + "loss": 4.1299, + "step": 6725 + }, + { + "epoch": 0.06726, + "grad_norm": 0.7891627577920859, + "learning_rate": 0.003, + "loss": 4.1563, + "step": 6726 + }, + { + "epoch": 0.06727, + "grad_norm": 0.9042190562513309, + "learning_rate": 0.003, + "loss": 4.1221, + "step": 6727 + }, + { + "epoch": 0.06728, + "grad_norm": 1.0541308359487462, + "learning_rate": 0.003, + "loss": 4.0975, + "step": 6728 + }, + { + "epoch": 0.06729, + "grad_norm": 1.0843675552428607, + "learning_rate": 0.003, + "loss": 4.1202, + "step": 6729 + }, + { + "epoch": 0.0673, + "grad_norm": 1.0045820679439086, + "learning_rate": 0.003, + "loss": 4.128, + "step": 6730 + }, + { + "epoch": 0.06731, + "grad_norm": 0.9002591648778789, + "learning_rate": 0.003, + "loss": 4.1006, + "step": 6731 + }, + { + "epoch": 0.06732, + "grad_norm": 0.646912864765862, + "learning_rate": 0.003, + "loss": 4.1142, + "step": 6732 + }, + { + "epoch": 0.06733, + "grad_norm": 0.7141084225358552, + "learning_rate": 0.003, + "loss": 4.1284, + "step": 6733 + }, + { + "epoch": 0.06734, + "grad_norm": 0.7401896537351272, + "learning_rate": 0.003, + "loss": 4.1231, + "step": 6734 + }, + { + "epoch": 0.06735, + "grad_norm": 0.716630242459992, + "learning_rate": 0.003, + "loss": 4.1002, + "step": 6735 + }, + { + "epoch": 0.06736, + "grad_norm": 0.8346796605661073, + "learning_rate": 0.003, + "loss": 4.1056, + "step": 6736 + }, + { + "epoch": 0.06737, + "grad_norm": 0.8558768913961102, + "learning_rate": 0.003, + "loss": 4.1144, + "step": 6737 + }, + { + "epoch": 0.06738, + "grad_norm": 0.8608671123963944, + "learning_rate": 0.003, + "loss": 4.1098, + "step": 6738 + }, + { + "epoch": 0.06739, + "grad_norm": 0.9596153104235727, + "learning_rate": 0.003, + "loss": 4.1182, + "step": 6739 + }, + { + "epoch": 0.0674, + "grad_norm": 0.937939190033793, + "learning_rate": 0.003, + "loss": 4.1145, + "step": 6740 + }, + { + "epoch": 0.06741, + "grad_norm": 0.7552811761560235, + "learning_rate": 0.003, + "loss": 4.0981, + "step": 6741 + }, + { + "epoch": 0.06742, + "grad_norm": 0.650352878706116, + "learning_rate": 0.003, + "loss": 4.12, + "step": 6742 + }, + { + "epoch": 0.06743, + "grad_norm": 0.6887739825416035, + "learning_rate": 0.003, + "loss": 4.0913, + "step": 6743 + }, + { + "epoch": 0.06744, + "grad_norm": 0.6459118608591984, + "learning_rate": 0.003, + "loss": 4.1247, + "step": 6744 + }, + { + "epoch": 0.06745, + "grad_norm": 0.7004216000552352, + "learning_rate": 0.003, + "loss": 4.0962, + "step": 6745 + }, + { + "epoch": 0.06746, + "grad_norm": 0.6099448791242965, + "learning_rate": 0.003, + "loss": 4.1021, + "step": 6746 + }, + { + "epoch": 0.06747, + "grad_norm": 0.6988500937390398, + "learning_rate": 0.003, + "loss": 4.1065, + "step": 6747 + }, + { + "epoch": 0.06748, + "grad_norm": 0.7156433755323733, + "learning_rate": 0.003, + "loss": 4.1094, + "step": 6748 + }, + { + "epoch": 0.06749, + "grad_norm": 0.801421918564714, + "learning_rate": 0.003, + "loss": 4.1034, + "step": 6749 + }, + { + "epoch": 0.0675, + "grad_norm": 0.9888166246308364, + "learning_rate": 0.003, + "loss": 4.1015, + "step": 6750 + }, + { + "epoch": 0.06751, + "grad_norm": 0.9741131169646507, + "learning_rate": 0.003, + "loss": 4.1407, + "step": 6751 + }, + { + "epoch": 0.06752, + "grad_norm": 0.8319653111077174, + "learning_rate": 0.003, + "loss": 4.1293, + "step": 6752 + }, + { + "epoch": 0.06753, + "grad_norm": 0.7778459734702268, + "learning_rate": 0.003, + "loss": 4.0881, + "step": 6753 + }, + { + "epoch": 0.06754, + "grad_norm": 0.6877593644712623, + "learning_rate": 0.003, + "loss": 4.1132, + "step": 6754 + }, + { + "epoch": 0.06755, + "grad_norm": 0.5706976203238855, + "learning_rate": 0.003, + "loss": 4.0969, + "step": 6755 + }, + { + "epoch": 0.06756, + "grad_norm": 0.6454631853072159, + "learning_rate": 0.003, + "loss": 4.1147, + "step": 6756 + }, + { + "epoch": 0.06757, + "grad_norm": 0.7487310221163939, + "learning_rate": 0.003, + "loss": 4.119, + "step": 6757 + }, + { + "epoch": 0.06758, + "grad_norm": 0.9383683479121699, + "learning_rate": 0.003, + "loss": 4.1131, + "step": 6758 + }, + { + "epoch": 0.06759, + "grad_norm": 1.2488321237213988, + "learning_rate": 0.003, + "loss": 4.152, + "step": 6759 + }, + { + "epoch": 0.0676, + "grad_norm": 0.8044219051814184, + "learning_rate": 0.003, + "loss": 4.1233, + "step": 6760 + }, + { + "epoch": 0.06761, + "grad_norm": 0.6504031656659648, + "learning_rate": 0.003, + "loss": 4.1028, + "step": 6761 + }, + { + "epoch": 0.06762, + "grad_norm": 0.5974848351395243, + "learning_rate": 0.003, + "loss": 4.1069, + "step": 6762 + }, + { + "epoch": 0.06763, + "grad_norm": 0.7658766240767714, + "learning_rate": 0.003, + "loss": 4.1079, + "step": 6763 + }, + { + "epoch": 0.06764, + "grad_norm": 1.0816158497707196, + "learning_rate": 0.003, + "loss": 4.1001, + "step": 6764 + }, + { + "epoch": 0.06765, + "grad_norm": 1.0931948991927563, + "learning_rate": 0.003, + "loss": 4.1191, + "step": 6765 + }, + { + "epoch": 0.06766, + "grad_norm": 0.7550131909735355, + "learning_rate": 0.003, + "loss": 4.1244, + "step": 6766 + }, + { + "epoch": 0.06767, + "grad_norm": 0.6220201794198699, + "learning_rate": 0.003, + "loss": 4.1212, + "step": 6767 + }, + { + "epoch": 0.06768, + "grad_norm": 0.7460575761496198, + "learning_rate": 0.003, + "loss": 4.0848, + "step": 6768 + }, + { + "epoch": 0.06769, + "grad_norm": 0.7893758885707454, + "learning_rate": 0.003, + "loss": 4.0993, + "step": 6769 + }, + { + "epoch": 0.0677, + "grad_norm": 0.7505537103644507, + "learning_rate": 0.003, + "loss": 4.1165, + "step": 6770 + }, + { + "epoch": 0.06771, + "grad_norm": 0.7025624171913118, + "learning_rate": 0.003, + "loss": 4.0995, + "step": 6771 + }, + { + "epoch": 0.06772, + "grad_norm": 0.7560527883448278, + "learning_rate": 0.003, + "loss": 4.1098, + "step": 6772 + }, + { + "epoch": 0.06773, + "grad_norm": 0.7196524539900757, + "learning_rate": 0.003, + "loss": 4.1252, + "step": 6773 + }, + { + "epoch": 0.06774, + "grad_norm": 0.7389219268859994, + "learning_rate": 0.003, + "loss": 4.062, + "step": 6774 + }, + { + "epoch": 0.06775, + "grad_norm": 0.6685763528477642, + "learning_rate": 0.003, + "loss": 4.1096, + "step": 6775 + }, + { + "epoch": 0.06776, + "grad_norm": 0.5686448643650356, + "learning_rate": 0.003, + "loss": 4.1352, + "step": 6776 + }, + { + "epoch": 0.06777, + "grad_norm": 0.5856469192051172, + "learning_rate": 0.003, + "loss": 4.1145, + "step": 6777 + }, + { + "epoch": 0.06778, + "grad_norm": 0.5829647709549176, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 6778 + }, + { + "epoch": 0.06779, + "grad_norm": 0.58878250453052, + "learning_rate": 0.003, + "loss": 4.1078, + "step": 6779 + }, + { + "epoch": 0.0678, + "grad_norm": 0.6705441528405734, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 6780 + }, + { + "epoch": 0.06781, + "grad_norm": 0.694256508354405, + "learning_rate": 0.003, + "loss": 4.1373, + "step": 6781 + }, + { + "epoch": 0.06782, + "grad_norm": 0.7836721563560676, + "learning_rate": 0.003, + "loss": 4.1166, + "step": 6782 + }, + { + "epoch": 0.06783, + "grad_norm": 0.9605929520599132, + "learning_rate": 0.003, + "loss": 4.1272, + "step": 6783 + }, + { + "epoch": 0.06784, + "grad_norm": 0.9687369923954169, + "learning_rate": 0.003, + "loss": 4.1023, + "step": 6784 + }, + { + "epoch": 0.06785, + "grad_norm": 0.8611250521831618, + "learning_rate": 0.003, + "loss": 4.1229, + "step": 6785 + }, + { + "epoch": 0.06786, + "grad_norm": 0.8150784557873674, + "learning_rate": 0.003, + "loss": 4.0913, + "step": 6786 + }, + { + "epoch": 0.06787, + "grad_norm": 0.7294549203147801, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 6787 + }, + { + "epoch": 0.06788, + "grad_norm": 0.7384718901245526, + "learning_rate": 0.003, + "loss": 4.1238, + "step": 6788 + }, + { + "epoch": 0.06789, + "grad_norm": 0.7373796160060775, + "learning_rate": 0.003, + "loss": 4.0898, + "step": 6789 + }, + { + "epoch": 0.0679, + "grad_norm": 0.7304803645302286, + "learning_rate": 0.003, + "loss": 4.1154, + "step": 6790 + }, + { + "epoch": 0.06791, + "grad_norm": 0.7119811531380004, + "learning_rate": 0.003, + "loss": 4.0989, + "step": 6791 + }, + { + "epoch": 0.06792, + "grad_norm": 0.8288761752825445, + "learning_rate": 0.003, + "loss": 4.122, + "step": 6792 + }, + { + "epoch": 0.06793, + "grad_norm": 0.7074297261709009, + "learning_rate": 0.003, + "loss": 4.1159, + "step": 6793 + }, + { + "epoch": 0.06794, + "grad_norm": 0.6917103888708278, + "learning_rate": 0.003, + "loss": 4.1297, + "step": 6794 + }, + { + "epoch": 0.06795, + "grad_norm": 0.9056603235821777, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 6795 + }, + { + "epoch": 0.06796, + "grad_norm": 1.2619528323809053, + "learning_rate": 0.003, + "loss": 4.1357, + "step": 6796 + }, + { + "epoch": 0.06797, + "grad_norm": 0.9779295543791401, + "learning_rate": 0.003, + "loss": 4.1424, + "step": 6797 + }, + { + "epoch": 0.06798, + "grad_norm": 1.0466081925100106, + "learning_rate": 0.003, + "loss": 4.1134, + "step": 6798 + }, + { + "epoch": 0.06799, + "grad_norm": 1.0487678592502534, + "learning_rate": 0.003, + "loss": 4.153, + "step": 6799 + }, + { + "epoch": 0.068, + "grad_norm": 0.9892002508567624, + "learning_rate": 0.003, + "loss": 4.1263, + "step": 6800 + }, + { + "epoch": 0.06801, + "grad_norm": 0.9771195211817344, + "learning_rate": 0.003, + "loss": 4.1038, + "step": 6801 + }, + { + "epoch": 0.06802, + "grad_norm": 0.9110534102852377, + "learning_rate": 0.003, + "loss": 4.1182, + "step": 6802 + }, + { + "epoch": 0.06803, + "grad_norm": 0.9513266489274651, + "learning_rate": 0.003, + "loss": 4.1217, + "step": 6803 + }, + { + "epoch": 0.06804, + "grad_norm": 1.1138225723663524, + "learning_rate": 0.003, + "loss": 4.1412, + "step": 6804 + }, + { + "epoch": 0.06805, + "grad_norm": 1.0317572052095008, + "learning_rate": 0.003, + "loss": 4.1185, + "step": 6805 + }, + { + "epoch": 0.06806, + "grad_norm": 0.797666875905461, + "learning_rate": 0.003, + "loss": 4.1117, + "step": 6806 + }, + { + "epoch": 0.06807, + "grad_norm": 0.7541059465600338, + "learning_rate": 0.003, + "loss": 4.1153, + "step": 6807 + }, + { + "epoch": 0.06808, + "grad_norm": 0.6793248525851846, + "learning_rate": 0.003, + "loss": 4.1179, + "step": 6808 + }, + { + "epoch": 0.06809, + "grad_norm": 0.6873438348421407, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 6809 + }, + { + "epoch": 0.0681, + "grad_norm": 0.6049328221467462, + "learning_rate": 0.003, + "loss": 4.1103, + "step": 6810 + }, + { + "epoch": 0.06811, + "grad_norm": 0.6187722825054933, + "learning_rate": 0.003, + "loss": 4.1381, + "step": 6811 + }, + { + "epoch": 0.06812, + "grad_norm": 0.7495708505892162, + "learning_rate": 0.003, + "loss": 4.1062, + "step": 6812 + }, + { + "epoch": 0.06813, + "grad_norm": 0.8952833843553192, + "learning_rate": 0.003, + "loss": 4.0878, + "step": 6813 + }, + { + "epoch": 0.06814, + "grad_norm": 1.052298391151623, + "learning_rate": 0.003, + "loss": 4.1321, + "step": 6814 + }, + { + "epoch": 0.06815, + "grad_norm": 0.8914337446957912, + "learning_rate": 0.003, + "loss": 4.1363, + "step": 6815 + }, + { + "epoch": 0.06816, + "grad_norm": 0.9633140015019734, + "learning_rate": 0.003, + "loss": 4.1361, + "step": 6816 + }, + { + "epoch": 0.06817, + "grad_norm": 0.8885850734774637, + "learning_rate": 0.003, + "loss": 4.1044, + "step": 6817 + }, + { + "epoch": 0.06818, + "grad_norm": 0.7384023274306349, + "learning_rate": 0.003, + "loss": 4.1212, + "step": 6818 + }, + { + "epoch": 0.06819, + "grad_norm": 0.7256819329745252, + "learning_rate": 0.003, + "loss": 4.1145, + "step": 6819 + }, + { + "epoch": 0.0682, + "grad_norm": 0.8464960000732561, + "learning_rate": 0.003, + "loss": 4.1449, + "step": 6820 + }, + { + "epoch": 0.06821, + "grad_norm": 0.959885711235584, + "learning_rate": 0.003, + "loss": 4.1381, + "step": 6821 + }, + { + "epoch": 0.06822, + "grad_norm": 1.0407940708061996, + "learning_rate": 0.003, + "loss": 4.1013, + "step": 6822 + }, + { + "epoch": 0.06823, + "grad_norm": 0.9542538781486196, + "learning_rate": 0.003, + "loss": 4.1156, + "step": 6823 + }, + { + "epoch": 0.06824, + "grad_norm": 0.8548426620050126, + "learning_rate": 0.003, + "loss": 4.0933, + "step": 6824 + }, + { + "epoch": 0.06825, + "grad_norm": 0.7562330738587587, + "learning_rate": 0.003, + "loss": 4.1082, + "step": 6825 + }, + { + "epoch": 0.06826, + "grad_norm": 0.6682601342553707, + "learning_rate": 0.003, + "loss": 4.1002, + "step": 6826 + }, + { + "epoch": 0.06827, + "grad_norm": 0.5882550844092198, + "learning_rate": 0.003, + "loss": 4.087, + "step": 6827 + }, + { + "epoch": 0.06828, + "grad_norm": 0.7012417787965474, + "learning_rate": 0.003, + "loss": 4.1137, + "step": 6828 + }, + { + "epoch": 0.06829, + "grad_norm": 0.8327606673237082, + "learning_rate": 0.003, + "loss": 4.1202, + "step": 6829 + }, + { + "epoch": 0.0683, + "grad_norm": 0.8528489072611283, + "learning_rate": 0.003, + "loss": 4.1112, + "step": 6830 + }, + { + "epoch": 0.06831, + "grad_norm": 0.7377079020858234, + "learning_rate": 0.003, + "loss": 4.1327, + "step": 6831 + }, + { + "epoch": 0.06832, + "grad_norm": 0.6999485550253548, + "learning_rate": 0.003, + "loss": 4.0981, + "step": 6832 + }, + { + "epoch": 0.06833, + "grad_norm": 0.7712453382887129, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 6833 + }, + { + "epoch": 0.06834, + "grad_norm": 0.9741841802598422, + "learning_rate": 0.003, + "loss": 4.1248, + "step": 6834 + }, + { + "epoch": 0.06835, + "grad_norm": 1.0151191229361403, + "learning_rate": 0.003, + "loss": 4.1314, + "step": 6835 + }, + { + "epoch": 0.06836, + "grad_norm": 0.9928135661127082, + "learning_rate": 0.003, + "loss": 4.1557, + "step": 6836 + }, + { + "epoch": 0.06837, + "grad_norm": 1.0176229330263218, + "learning_rate": 0.003, + "loss": 4.1103, + "step": 6837 + }, + { + "epoch": 0.06838, + "grad_norm": 0.8062582072461082, + "learning_rate": 0.003, + "loss": 4.11, + "step": 6838 + }, + { + "epoch": 0.06839, + "grad_norm": 1.0001169867154327, + "learning_rate": 0.003, + "loss": 4.1073, + "step": 6839 + }, + { + "epoch": 0.0684, + "grad_norm": 1.2793247468125468, + "learning_rate": 0.003, + "loss": 4.1399, + "step": 6840 + }, + { + "epoch": 0.06841, + "grad_norm": 0.9352650749931306, + "learning_rate": 0.003, + "loss": 4.1298, + "step": 6841 + }, + { + "epoch": 0.06842, + "grad_norm": 0.873691761023133, + "learning_rate": 0.003, + "loss": 4.1281, + "step": 6842 + }, + { + "epoch": 0.06843, + "grad_norm": 0.943397035481235, + "learning_rate": 0.003, + "loss": 4.1205, + "step": 6843 + }, + { + "epoch": 0.06844, + "grad_norm": 1.2419554124329863, + "learning_rate": 0.003, + "loss": 4.1288, + "step": 6844 + }, + { + "epoch": 0.06845, + "grad_norm": 0.7627006179431822, + "learning_rate": 0.003, + "loss": 4.1313, + "step": 6845 + }, + { + "epoch": 0.06846, + "grad_norm": 0.7761262500841255, + "learning_rate": 0.003, + "loss": 4.1212, + "step": 6846 + }, + { + "epoch": 0.06847, + "grad_norm": 0.9754798916992613, + "learning_rate": 0.003, + "loss": 4.1099, + "step": 6847 + }, + { + "epoch": 0.06848, + "grad_norm": 1.2952128685404893, + "learning_rate": 0.003, + "loss": 4.1332, + "step": 6848 + }, + { + "epoch": 0.06849, + "grad_norm": 0.8183956076455051, + "learning_rate": 0.003, + "loss": 4.1125, + "step": 6849 + }, + { + "epoch": 0.0685, + "grad_norm": 0.7298588998388881, + "learning_rate": 0.003, + "loss": 4.1113, + "step": 6850 + }, + { + "epoch": 0.06851, + "grad_norm": 0.7720044301801032, + "learning_rate": 0.003, + "loss": 4.0917, + "step": 6851 + }, + { + "epoch": 0.06852, + "grad_norm": 0.9117508720265394, + "learning_rate": 0.003, + "loss": 4.104, + "step": 6852 + }, + { + "epoch": 0.06853, + "grad_norm": 1.0533798726154795, + "learning_rate": 0.003, + "loss": 4.1073, + "step": 6853 + }, + { + "epoch": 0.06854, + "grad_norm": 1.027625478858702, + "learning_rate": 0.003, + "loss": 4.1418, + "step": 6854 + }, + { + "epoch": 0.06855, + "grad_norm": 0.9135813747015725, + "learning_rate": 0.003, + "loss": 4.1489, + "step": 6855 + }, + { + "epoch": 0.06856, + "grad_norm": 0.9132419414958513, + "learning_rate": 0.003, + "loss": 4.1486, + "step": 6856 + }, + { + "epoch": 0.06857, + "grad_norm": 0.9212345733341637, + "learning_rate": 0.003, + "loss": 4.1283, + "step": 6857 + }, + { + "epoch": 0.06858, + "grad_norm": 0.8783800930563433, + "learning_rate": 0.003, + "loss": 4.0902, + "step": 6858 + }, + { + "epoch": 0.06859, + "grad_norm": 0.6985668013167255, + "learning_rate": 0.003, + "loss": 4.0979, + "step": 6859 + }, + { + "epoch": 0.0686, + "grad_norm": 0.6082758464053263, + "learning_rate": 0.003, + "loss": 4.1047, + "step": 6860 + }, + { + "epoch": 0.06861, + "grad_norm": 0.5184883931489526, + "learning_rate": 0.003, + "loss": 4.0956, + "step": 6861 + }, + { + "epoch": 0.06862, + "grad_norm": 0.5465866791407239, + "learning_rate": 0.003, + "loss": 4.0935, + "step": 6862 + }, + { + "epoch": 0.06863, + "grad_norm": 0.6294940342821269, + "learning_rate": 0.003, + "loss": 4.1001, + "step": 6863 + }, + { + "epoch": 0.06864, + "grad_norm": 0.819027721570075, + "learning_rate": 0.003, + "loss": 4.1144, + "step": 6864 + }, + { + "epoch": 0.06865, + "grad_norm": 0.9537722441632093, + "learning_rate": 0.003, + "loss": 4.1017, + "step": 6865 + }, + { + "epoch": 0.06866, + "grad_norm": 0.9579871191173921, + "learning_rate": 0.003, + "loss": 4.1245, + "step": 6866 + }, + { + "epoch": 0.06867, + "grad_norm": 0.7874535648082873, + "learning_rate": 0.003, + "loss": 4.132, + "step": 6867 + }, + { + "epoch": 0.06868, + "grad_norm": 0.7207326410237167, + "learning_rate": 0.003, + "loss": 4.0927, + "step": 6868 + }, + { + "epoch": 0.06869, + "grad_norm": 0.8921760331303482, + "learning_rate": 0.003, + "loss": 4.1235, + "step": 6869 + }, + { + "epoch": 0.0687, + "grad_norm": 1.0524247874039665, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 6870 + }, + { + "epoch": 0.06871, + "grad_norm": 0.9959937136708353, + "learning_rate": 0.003, + "loss": 4.0958, + "step": 6871 + }, + { + "epoch": 0.06872, + "grad_norm": 0.7683343036479461, + "learning_rate": 0.003, + "loss": 4.093, + "step": 6872 + }, + { + "epoch": 0.06873, + "grad_norm": 0.6815142284894331, + "learning_rate": 0.003, + "loss": 4.1016, + "step": 6873 + }, + { + "epoch": 0.06874, + "grad_norm": 0.681503918141428, + "learning_rate": 0.003, + "loss": 4.1157, + "step": 6874 + }, + { + "epoch": 0.06875, + "grad_norm": 0.671319808930849, + "learning_rate": 0.003, + "loss": 4.1171, + "step": 6875 + }, + { + "epoch": 0.06876, + "grad_norm": 0.6432032608458409, + "learning_rate": 0.003, + "loss": 4.1207, + "step": 6876 + }, + { + "epoch": 0.06877, + "grad_norm": 0.681246314917716, + "learning_rate": 0.003, + "loss": 4.0902, + "step": 6877 + }, + { + "epoch": 0.06878, + "grad_norm": 0.6862800830361652, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 6878 + }, + { + "epoch": 0.06879, + "grad_norm": 0.6785172447705301, + "learning_rate": 0.003, + "loss": 4.0935, + "step": 6879 + }, + { + "epoch": 0.0688, + "grad_norm": 0.6523592214155444, + "learning_rate": 0.003, + "loss": 4.1044, + "step": 6880 + }, + { + "epoch": 0.06881, + "grad_norm": 0.7257213543926093, + "learning_rate": 0.003, + "loss": 4.0942, + "step": 6881 + }, + { + "epoch": 0.06882, + "grad_norm": 0.8060377847691599, + "learning_rate": 0.003, + "loss": 4.1028, + "step": 6882 + }, + { + "epoch": 0.06883, + "grad_norm": 0.792778596117736, + "learning_rate": 0.003, + "loss": 4.1251, + "step": 6883 + }, + { + "epoch": 0.06884, + "grad_norm": 0.6894553250899708, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 6884 + }, + { + "epoch": 0.06885, + "grad_norm": 0.7168844072539149, + "learning_rate": 0.003, + "loss": 4.1098, + "step": 6885 + }, + { + "epoch": 0.06886, + "grad_norm": 0.7480318972493727, + "learning_rate": 0.003, + "loss": 4.0949, + "step": 6886 + }, + { + "epoch": 0.06887, + "grad_norm": 0.7617857801277111, + "learning_rate": 0.003, + "loss": 4.1175, + "step": 6887 + }, + { + "epoch": 0.06888, + "grad_norm": 0.861679021778327, + "learning_rate": 0.003, + "loss": 4.0962, + "step": 6888 + }, + { + "epoch": 0.06889, + "grad_norm": 0.8718407312295603, + "learning_rate": 0.003, + "loss": 4.1072, + "step": 6889 + }, + { + "epoch": 0.0689, + "grad_norm": 1.0130605630972407, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 6890 + }, + { + "epoch": 0.06891, + "grad_norm": 1.3688291187731094, + "learning_rate": 0.003, + "loss": 4.1356, + "step": 6891 + }, + { + "epoch": 0.06892, + "grad_norm": 0.7799161275909664, + "learning_rate": 0.003, + "loss": 4.1038, + "step": 6892 + }, + { + "epoch": 0.06893, + "grad_norm": 0.7045235058505965, + "learning_rate": 0.003, + "loss": 4.1136, + "step": 6893 + }, + { + "epoch": 0.06894, + "grad_norm": 0.6684728435369993, + "learning_rate": 0.003, + "loss": 4.0971, + "step": 6894 + }, + { + "epoch": 0.06895, + "grad_norm": 0.6945126017991131, + "learning_rate": 0.003, + "loss": 4.0937, + "step": 6895 + }, + { + "epoch": 0.06896, + "grad_norm": 0.7482855504159163, + "learning_rate": 0.003, + "loss": 4.0913, + "step": 6896 + }, + { + "epoch": 0.06897, + "grad_norm": 0.8746143480093368, + "learning_rate": 0.003, + "loss": 4.1417, + "step": 6897 + }, + { + "epoch": 0.06898, + "grad_norm": 0.8993489919013548, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 6898 + }, + { + "epoch": 0.06899, + "grad_norm": 1.0857843804381477, + "learning_rate": 0.003, + "loss": 4.1538, + "step": 6899 + }, + { + "epoch": 0.069, + "grad_norm": 1.0648749747194601, + "learning_rate": 0.003, + "loss": 4.1172, + "step": 6900 + }, + { + "epoch": 0.06901, + "grad_norm": 0.9006029334239385, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 6901 + }, + { + "epoch": 0.06902, + "grad_norm": 0.8416068012359819, + "learning_rate": 0.003, + "loss": 4.1022, + "step": 6902 + }, + { + "epoch": 0.06903, + "grad_norm": 1.006337020622227, + "learning_rate": 0.003, + "loss": 4.1141, + "step": 6903 + }, + { + "epoch": 0.06904, + "grad_norm": 1.045403521519028, + "learning_rate": 0.003, + "loss": 4.1181, + "step": 6904 + }, + { + "epoch": 0.06905, + "grad_norm": 1.0035690205519932, + "learning_rate": 0.003, + "loss": 4.1433, + "step": 6905 + }, + { + "epoch": 0.06906, + "grad_norm": 1.0866394477802588, + "learning_rate": 0.003, + "loss": 4.0927, + "step": 6906 + }, + { + "epoch": 0.06907, + "grad_norm": 0.9795306254573748, + "learning_rate": 0.003, + "loss": 4.1113, + "step": 6907 + }, + { + "epoch": 0.06908, + "grad_norm": 0.921664480593037, + "learning_rate": 0.003, + "loss": 4.0912, + "step": 6908 + }, + { + "epoch": 0.06909, + "grad_norm": 0.7625577740017703, + "learning_rate": 0.003, + "loss": 4.1224, + "step": 6909 + }, + { + "epoch": 0.0691, + "grad_norm": 0.7611464664895455, + "learning_rate": 0.003, + "loss": 4.09, + "step": 6910 + }, + { + "epoch": 0.06911, + "grad_norm": 0.7481854469654662, + "learning_rate": 0.003, + "loss": 4.1344, + "step": 6911 + }, + { + "epoch": 0.06912, + "grad_norm": 0.9007946005791158, + "learning_rate": 0.003, + "loss": 4.1152, + "step": 6912 + }, + { + "epoch": 0.06913, + "grad_norm": 1.0181585830069129, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 6913 + }, + { + "epoch": 0.06914, + "grad_norm": 1.1870130455108574, + "learning_rate": 0.003, + "loss": 4.1449, + "step": 6914 + }, + { + "epoch": 0.06915, + "grad_norm": 0.9236319005490047, + "learning_rate": 0.003, + "loss": 4.1477, + "step": 6915 + }, + { + "epoch": 0.06916, + "grad_norm": 0.8302058218315981, + "learning_rate": 0.003, + "loss": 4.1137, + "step": 6916 + }, + { + "epoch": 0.06917, + "grad_norm": 0.8449547905827482, + "learning_rate": 0.003, + "loss": 4.0982, + "step": 6917 + }, + { + "epoch": 0.06918, + "grad_norm": 1.0042057313137678, + "learning_rate": 0.003, + "loss": 4.1405, + "step": 6918 + }, + { + "epoch": 0.06919, + "grad_norm": 1.0948578415550971, + "learning_rate": 0.003, + "loss": 4.1374, + "step": 6919 + }, + { + "epoch": 0.0692, + "grad_norm": 0.8651976841261361, + "learning_rate": 0.003, + "loss": 4.1192, + "step": 6920 + }, + { + "epoch": 0.06921, + "grad_norm": 0.843906688583314, + "learning_rate": 0.003, + "loss": 4.1275, + "step": 6921 + }, + { + "epoch": 0.06922, + "grad_norm": 0.7494458953280497, + "learning_rate": 0.003, + "loss": 4.1176, + "step": 6922 + }, + { + "epoch": 0.06923, + "grad_norm": 0.6493697075436607, + "learning_rate": 0.003, + "loss": 4.0968, + "step": 6923 + }, + { + "epoch": 0.06924, + "grad_norm": 0.5960115740450397, + "learning_rate": 0.003, + "loss": 4.1368, + "step": 6924 + }, + { + "epoch": 0.06925, + "grad_norm": 0.5439383509954068, + "learning_rate": 0.003, + "loss": 4.1275, + "step": 6925 + }, + { + "epoch": 0.06926, + "grad_norm": 0.5566009950142305, + "learning_rate": 0.003, + "loss": 4.1205, + "step": 6926 + }, + { + "epoch": 0.06927, + "grad_norm": 0.5929207701665649, + "learning_rate": 0.003, + "loss": 4.1134, + "step": 6927 + }, + { + "epoch": 0.06928, + "grad_norm": 0.6959312150674015, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 6928 + }, + { + "epoch": 0.06929, + "grad_norm": 0.9859706292355523, + "learning_rate": 0.003, + "loss": 4.1045, + "step": 6929 + }, + { + "epoch": 0.0693, + "grad_norm": 1.2296463232455344, + "learning_rate": 0.003, + "loss": 4.1608, + "step": 6930 + }, + { + "epoch": 0.06931, + "grad_norm": 0.610382372425946, + "learning_rate": 0.003, + "loss": 4.1152, + "step": 6931 + }, + { + "epoch": 0.06932, + "grad_norm": 0.6524546502395882, + "learning_rate": 0.003, + "loss": 4.0908, + "step": 6932 + }, + { + "epoch": 0.06933, + "grad_norm": 0.7923321668003368, + "learning_rate": 0.003, + "loss": 4.1315, + "step": 6933 + }, + { + "epoch": 0.06934, + "grad_norm": 0.824582431481253, + "learning_rate": 0.003, + "loss": 4.0968, + "step": 6934 + }, + { + "epoch": 0.06935, + "grad_norm": 0.8727764667568022, + "learning_rate": 0.003, + "loss": 4.1197, + "step": 6935 + }, + { + "epoch": 0.06936, + "grad_norm": 0.8326712177028426, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 6936 + }, + { + "epoch": 0.06937, + "grad_norm": 0.6666993557326405, + "learning_rate": 0.003, + "loss": 4.1008, + "step": 6937 + }, + { + "epoch": 0.06938, + "grad_norm": 0.7059003515766912, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 6938 + }, + { + "epoch": 0.06939, + "grad_norm": 0.7910834610590093, + "learning_rate": 0.003, + "loss": 4.1098, + "step": 6939 + }, + { + "epoch": 0.0694, + "grad_norm": 0.755416937286377, + "learning_rate": 0.003, + "loss": 4.0981, + "step": 6940 + }, + { + "epoch": 0.06941, + "grad_norm": 0.6585307567475693, + "learning_rate": 0.003, + "loss": 4.1137, + "step": 6941 + }, + { + "epoch": 0.06942, + "grad_norm": 0.686786611772453, + "learning_rate": 0.003, + "loss": 4.0949, + "step": 6942 + }, + { + "epoch": 0.06943, + "grad_norm": 0.6748173855765259, + "learning_rate": 0.003, + "loss": 4.1205, + "step": 6943 + }, + { + "epoch": 0.06944, + "grad_norm": 0.8309412479867045, + "learning_rate": 0.003, + "loss": 4.1066, + "step": 6944 + }, + { + "epoch": 0.06945, + "grad_norm": 0.8694006981278489, + "learning_rate": 0.003, + "loss": 4.1363, + "step": 6945 + }, + { + "epoch": 0.06946, + "grad_norm": 0.9735883766446164, + "learning_rate": 0.003, + "loss": 4.1176, + "step": 6946 + }, + { + "epoch": 0.06947, + "grad_norm": 0.9569405363329243, + "learning_rate": 0.003, + "loss": 4.1025, + "step": 6947 + }, + { + "epoch": 0.06948, + "grad_norm": 0.9911481171151727, + "learning_rate": 0.003, + "loss": 4.1157, + "step": 6948 + }, + { + "epoch": 0.06949, + "grad_norm": 1.1434234191553132, + "learning_rate": 0.003, + "loss": 4.1296, + "step": 6949 + }, + { + "epoch": 0.0695, + "grad_norm": 0.8225443853899382, + "learning_rate": 0.003, + "loss": 4.1008, + "step": 6950 + }, + { + "epoch": 0.06951, + "grad_norm": 0.7771139496183902, + "learning_rate": 0.003, + "loss": 4.1334, + "step": 6951 + }, + { + "epoch": 0.06952, + "grad_norm": 0.8827808082345756, + "learning_rate": 0.003, + "loss": 4.1382, + "step": 6952 + }, + { + "epoch": 0.06953, + "grad_norm": 1.0344317147347284, + "learning_rate": 0.003, + "loss": 4.1402, + "step": 6953 + }, + { + "epoch": 0.06954, + "grad_norm": 0.9087885968248024, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 6954 + }, + { + "epoch": 0.06955, + "grad_norm": 0.8948872903159871, + "learning_rate": 0.003, + "loss": 4.129, + "step": 6955 + }, + { + "epoch": 0.06956, + "grad_norm": 0.8801466457771367, + "learning_rate": 0.003, + "loss": 4.133, + "step": 6956 + }, + { + "epoch": 0.06957, + "grad_norm": 0.7407954654118085, + "learning_rate": 0.003, + "loss": 4.1092, + "step": 6957 + }, + { + "epoch": 0.06958, + "grad_norm": 0.6110627923040244, + "learning_rate": 0.003, + "loss": 4.122, + "step": 6958 + }, + { + "epoch": 0.06959, + "grad_norm": 0.640461355001315, + "learning_rate": 0.003, + "loss": 4.1159, + "step": 6959 + }, + { + "epoch": 0.0696, + "grad_norm": 0.7327706567363643, + "learning_rate": 0.003, + "loss": 4.1274, + "step": 6960 + }, + { + "epoch": 0.06961, + "grad_norm": 0.8343628226365843, + "learning_rate": 0.003, + "loss": 4.1274, + "step": 6961 + }, + { + "epoch": 0.06962, + "grad_norm": 0.8816790067550263, + "learning_rate": 0.003, + "loss": 4.1198, + "step": 6962 + }, + { + "epoch": 0.06963, + "grad_norm": 1.029137591999145, + "learning_rate": 0.003, + "loss": 4.1277, + "step": 6963 + }, + { + "epoch": 0.06964, + "grad_norm": 1.083546288789342, + "learning_rate": 0.003, + "loss": 4.1255, + "step": 6964 + }, + { + "epoch": 0.06965, + "grad_norm": 1.0972587017397695, + "learning_rate": 0.003, + "loss": 4.1103, + "step": 6965 + }, + { + "epoch": 0.06966, + "grad_norm": 0.8505331999829698, + "learning_rate": 0.003, + "loss": 4.1279, + "step": 6966 + }, + { + "epoch": 0.06967, + "grad_norm": 0.6926913786836009, + "learning_rate": 0.003, + "loss": 4.1058, + "step": 6967 + }, + { + "epoch": 0.06968, + "grad_norm": 0.7470348630991248, + "learning_rate": 0.003, + "loss": 4.1134, + "step": 6968 + }, + { + "epoch": 0.06969, + "grad_norm": 0.7260002360079123, + "learning_rate": 0.003, + "loss": 4.1279, + "step": 6969 + }, + { + "epoch": 0.0697, + "grad_norm": 0.6766317452834478, + "learning_rate": 0.003, + "loss": 4.095, + "step": 6970 + }, + { + "epoch": 0.06971, + "grad_norm": 0.6911223172142554, + "learning_rate": 0.003, + "loss": 4.1041, + "step": 6971 + }, + { + "epoch": 0.06972, + "grad_norm": 0.6699177971563128, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 6972 + }, + { + "epoch": 0.06973, + "grad_norm": 0.694687314086879, + "learning_rate": 0.003, + "loss": 4.1132, + "step": 6973 + }, + { + "epoch": 0.06974, + "grad_norm": 0.9488840337040796, + "learning_rate": 0.003, + "loss": 4.1058, + "step": 6974 + }, + { + "epoch": 0.06975, + "grad_norm": 1.2258549533377359, + "learning_rate": 0.003, + "loss": 4.1239, + "step": 6975 + }, + { + "epoch": 0.06976, + "grad_norm": 0.7740323563047606, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 6976 + }, + { + "epoch": 0.06977, + "grad_norm": 0.575041977095538, + "learning_rate": 0.003, + "loss": 4.1458, + "step": 6977 + }, + { + "epoch": 0.06978, + "grad_norm": 0.728998390696451, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 6978 + }, + { + "epoch": 0.06979, + "grad_norm": 1.013780311861753, + "learning_rate": 0.003, + "loss": 4.1357, + "step": 6979 + }, + { + "epoch": 0.0698, + "grad_norm": 1.1763620482776194, + "learning_rate": 0.003, + "loss": 4.1272, + "step": 6980 + }, + { + "epoch": 0.06981, + "grad_norm": 0.815900592117822, + "learning_rate": 0.003, + "loss": 4.0939, + "step": 6981 + }, + { + "epoch": 0.06982, + "grad_norm": 0.7122151392371082, + "learning_rate": 0.003, + "loss": 4.1027, + "step": 6982 + }, + { + "epoch": 0.06983, + "grad_norm": 0.6205206572999914, + "learning_rate": 0.003, + "loss": 4.0848, + "step": 6983 + }, + { + "epoch": 0.06984, + "grad_norm": 0.6501948602208925, + "learning_rate": 0.003, + "loss": 4.0876, + "step": 6984 + }, + { + "epoch": 0.06985, + "grad_norm": 0.6936542630188876, + "learning_rate": 0.003, + "loss": 4.1203, + "step": 6985 + }, + { + "epoch": 0.06986, + "grad_norm": 0.7630402930183048, + "learning_rate": 0.003, + "loss": 4.1136, + "step": 6986 + }, + { + "epoch": 0.06987, + "grad_norm": 0.8307718661807078, + "learning_rate": 0.003, + "loss": 4.1145, + "step": 6987 + }, + { + "epoch": 0.06988, + "grad_norm": 0.8259567088608353, + "learning_rate": 0.003, + "loss": 4.0991, + "step": 6988 + }, + { + "epoch": 0.06989, + "grad_norm": 0.8383919214928979, + "learning_rate": 0.003, + "loss": 4.0891, + "step": 6989 + }, + { + "epoch": 0.0699, + "grad_norm": 0.8394764313790253, + "learning_rate": 0.003, + "loss": 4.0876, + "step": 6990 + }, + { + "epoch": 0.06991, + "grad_norm": 0.7966606342946648, + "learning_rate": 0.003, + "loss": 4.1182, + "step": 6991 + }, + { + "epoch": 0.06992, + "grad_norm": 0.6596406930588895, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 6992 + }, + { + "epoch": 0.06993, + "grad_norm": 0.6295320798498751, + "learning_rate": 0.003, + "loss": 4.0855, + "step": 6993 + }, + { + "epoch": 0.06994, + "grad_norm": 0.6843612644051112, + "learning_rate": 0.003, + "loss": 4.0919, + "step": 6994 + }, + { + "epoch": 0.06995, + "grad_norm": 0.7889219514350806, + "learning_rate": 0.003, + "loss": 4.1197, + "step": 6995 + }, + { + "epoch": 0.06996, + "grad_norm": 0.8402572785911753, + "learning_rate": 0.003, + "loss": 4.1067, + "step": 6996 + }, + { + "epoch": 0.06997, + "grad_norm": 0.7656897240238402, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 6997 + }, + { + "epoch": 0.06998, + "grad_norm": 0.8838339815381696, + "learning_rate": 0.003, + "loss": 4.1032, + "step": 6998 + }, + { + "epoch": 0.06999, + "grad_norm": 1.0521884576052078, + "learning_rate": 0.003, + "loss": 4.0955, + "step": 6999 + }, + { + "epoch": 0.07, + "grad_norm": 1.3428947555527946, + "learning_rate": 0.003, + "loss": 4.1077, + "step": 7000 + }, + { + "epoch": 0.07001, + "grad_norm": 0.7773020396247668, + "learning_rate": 0.003, + "loss": 4.152, + "step": 7001 + }, + { + "epoch": 0.07002, + "grad_norm": 0.6360859962376562, + "learning_rate": 0.003, + "loss": 4.0991, + "step": 7002 + }, + { + "epoch": 0.07003, + "grad_norm": 0.71049436819903, + "learning_rate": 0.003, + "loss": 4.0899, + "step": 7003 + }, + { + "epoch": 0.07004, + "grad_norm": 0.7207589244507122, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 7004 + }, + { + "epoch": 0.07005, + "grad_norm": 0.8029274080334843, + "learning_rate": 0.003, + "loss": 4.1037, + "step": 7005 + }, + { + "epoch": 0.07006, + "grad_norm": 0.9621397059134545, + "learning_rate": 0.003, + "loss": 4.1107, + "step": 7006 + }, + { + "epoch": 0.07007, + "grad_norm": 1.0548395954148817, + "learning_rate": 0.003, + "loss": 4.1166, + "step": 7007 + }, + { + "epoch": 0.07008, + "grad_norm": 0.9292949868098707, + "learning_rate": 0.003, + "loss": 4.1088, + "step": 7008 + }, + { + "epoch": 0.07009, + "grad_norm": 1.1487800774682821, + "learning_rate": 0.003, + "loss": 4.1326, + "step": 7009 + }, + { + "epoch": 0.0701, + "grad_norm": 1.0137599916747433, + "learning_rate": 0.003, + "loss": 4.1066, + "step": 7010 + }, + { + "epoch": 0.07011, + "grad_norm": 0.9682428887494761, + "learning_rate": 0.003, + "loss": 4.1013, + "step": 7011 + }, + { + "epoch": 0.07012, + "grad_norm": 0.8914447047092996, + "learning_rate": 0.003, + "loss": 4.1336, + "step": 7012 + }, + { + "epoch": 0.07013, + "grad_norm": 0.9039463647228102, + "learning_rate": 0.003, + "loss": 4.1038, + "step": 7013 + }, + { + "epoch": 0.07014, + "grad_norm": 1.134635997819684, + "learning_rate": 0.003, + "loss": 4.1003, + "step": 7014 + }, + { + "epoch": 0.07015, + "grad_norm": 1.031076929207217, + "learning_rate": 0.003, + "loss": 4.1057, + "step": 7015 + }, + { + "epoch": 0.07016, + "grad_norm": 1.0116310847444454, + "learning_rate": 0.003, + "loss": 4.1117, + "step": 7016 + }, + { + "epoch": 0.07017, + "grad_norm": 1.005782143604759, + "learning_rate": 0.003, + "loss": 4.0986, + "step": 7017 + }, + { + "epoch": 0.07018, + "grad_norm": 0.9578767335403211, + "learning_rate": 0.003, + "loss": 4.1274, + "step": 7018 + }, + { + "epoch": 0.07019, + "grad_norm": 0.8935381942583269, + "learning_rate": 0.003, + "loss": 4.1152, + "step": 7019 + }, + { + "epoch": 0.0702, + "grad_norm": 0.744623485765837, + "learning_rate": 0.003, + "loss": 4.1031, + "step": 7020 + }, + { + "epoch": 0.07021, + "grad_norm": 0.801305747893465, + "learning_rate": 0.003, + "loss": 4.121, + "step": 7021 + }, + { + "epoch": 0.07022, + "grad_norm": 0.7083760005593353, + "learning_rate": 0.003, + "loss": 4.1213, + "step": 7022 + }, + { + "epoch": 0.07023, + "grad_norm": 0.621039433188525, + "learning_rate": 0.003, + "loss": 4.1125, + "step": 7023 + }, + { + "epoch": 0.07024, + "grad_norm": 0.5806951914320432, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 7024 + }, + { + "epoch": 0.07025, + "grad_norm": 0.6152547374263034, + "learning_rate": 0.003, + "loss": 4.1165, + "step": 7025 + }, + { + "epoch": 0.07026, + "grad_norm": 0.6694178964744939, + "learning_rate": 0.003, + "loss": 4.0887, + "step": 7026 + }, + { + "epoch": 0.07027, + "grad_norm": 0.9841630681827739, + "learning_rate": 0.003, + "loss": 4.1042, + "step": 7027 + }, + { + "epoch": 0.07028, + "grad_norm": 1.200254843619916, + "learning_rate": 0.003, + "loss": 4.115, + "step": 7028 + }, + { + "epoch": 0.07029, + "grad_norm": 0.7174951451462476, + "learning_rate": 0.003, + "loss": 4.0925, + "step": 7029 + }, + { + "epoch": 0.0703, + "grad_norm": 0.6728970196461604, + "learning_rate": 0.003, + "loss": 4.1239, + "step": 7030 + }, + { + "epoch": 0.07031, + "grad_norm": 0.7070157774496603, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 7031 + }, + { + "epoch": 0.07032, + "grad_norm": 0.7387898161770746, + "learning_rate": 0.003, + "loss": 4.1128, + "step": 7032 + }, + { + "epoch": 0.07033, + "grad_norm": 0.8901816616568644, + "learning_rate": 0.003, + "loss": 4.1179, + "step": 7033 + }, + { + "epoch": 0.07034, + "grad_norm": 1.0284878793130177, + "learning_rate": 0.003, + "loss": 4.1098, + "step": 7034 + }, + { + "epoch": 0.07035, + "grad_norm": 1.058686540675132, + "learning_rate": 0.003, + "loss": 4.1186, + "step": 7035 + }, + { + "epoch": 0.07036, + "grad_norm": 0.9359391304214265, + "learning_rate": 0.003, + "loss": 4.1052, + "step": 7036 + }, + { + "epoch": 0.07037, + "grad_norm": 0.9481158673322754, + "learning_rate": 0.003, + "loss": 4.0974, + "step": 7037 + }, + { + "epoch": 0.07038, + "grad_norm": 0.8334487757118696, + "learning_rate": 0.003, + "loss": 4.1106, + "step": 7038 + }, + { + "epoch": 0.07039, + "grad_norm": 0.6326431662983385, + "learning_rate": 0.003, + "loss": 4.1031, + "step": 7039 + }, + { + "epoch": 0.0704, + "grad_norm": 0.702514824071445, + "learning_rate": 0.003, + "loss": 4.1267, + "step": 7040 + }, + { + "epoch": 0.07041, + "grad_norm": 0.7400349502050564, + "learning_rate": 0.003, + "loss": 4.1032, + "step": 7041 + }, + { + "epoch": 0.07042, + "grad_norm": 0.7099648121514438, + "learning_rate": 0.003, + "loss": 4.0884, + "step": 7042 + }, + { + "epoch": 0.07043, + "grad_norm": 0.6483290540107833, + "learning_rate": 0.003, + "loss": 4.121, + "step": 7043 + }, + { + "epoch": 0.07044, + "grad_norm": 0.6620145629117125, + "learning_rate": 0.003, + "loss": 4.1202, + "step": 7044 + }, + { + "epoch": 0.07045, + "grad_norm": 0.5978749302357278, + "learning_rate": 0.003, + "loss": 4.1042, + "step": 7045 + }, + { + "epoch": 0.07046, + "grad_norm": 0.5888111174272556, + "learning_rate": 0.003, + "loss": 4.1089, + "step": 7046 + }, + { + "epoch": 0.07047, + "grad_norm": 0.5784250123506132, + "learning_rate": 0.003, + "loss": 4.1062, + "step": 7047 + }, + { + "epoch": 0.07048, + "grad_norm": 0.6072433980268588, + "learning_rate": 0.003, + "loss": 4.0959, + "step": 7048 + }, + { + "epoch": 0.07049, + "grad_norm": 0.7094471050898912, + "learning_rate": 0.003, + "loss": 4.097, + "step": 7049 + }, + { + "epoch": 0.0705, + "grad_norm": 0.953711232001365, + "learning_rate": 0.003, + "loss": 4.1206, + "step": 7050 + }, + { + "epoch": 0.07051, + "grad_norm": 1.4359270563806132, + "learning_rate": 0.003, + "loss": 4.1404, + "step": 7051 + }, + { + "epoch": 0.07052, + "grad_norm": 0.6731996114775755, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 7052 + }, + { + "epoch": 0.07053, + "grad_norm": 0.8452128279853137, + "learning_rate": 0.003, + "loss": 4.0998, + "step": 7053 + }, + { + "epoch": 0.07054, + "grad_norm": 0.9361316507438332, + "learning_rate": 0.003, + "loss": 4.1027, + "step": 7054 + }, + { + "epoch": 0.07055, + "grad_norm": 1.0791674499784152, + "learning_rate": 0.003, + "loss": 4.1101, + "step": 7055 + }, + { + "epoch": 0.07056, + "grad_norm": 0.9840050657142604, + "learning_rate": 0.003, + "loss": 4.0995, + "step": 7056 + }, + { + "epoch": 0.07057, + "grad_norm": 0.7477218913081056, + "learning_rate": 0.003, + "loss": 4.118, + "step": 7057 + }, + { + "epoch": 0.07058, + "grad_norm": 0.6446497473373758, + "learning_rate": 0.003, + "loss": 4.1011, + "step": 7058 + }, + { + "epoch": 0.07059, + "grad_norm": 0.7865830553034692, + "learning_rate": 0.003, + "loss": 4.1074, + "step": 7059 + }, + { + "epoch": 0.0706, + "grad_norm": 1.0013324953332357, + "learning_rate": 0.003, + "loss": 4.1092, + "step": 7060 + }, + { + "epoch": 0.07061, + "grad_norm": 0.9809126840231154, + "learning_rate": 0.003, + "loss": 4.1046, + "step": 7061 + }, + { + "epoch": 0.07062, + "grad_norm": 1.0431446716134858, + "learning_rate": 0.003, + "loss": 4.1436, + "step": 7062 + }, + { + "epoch": 0.07063, + "grad_norm": 0.9216298810073756, + "learning_rate": 0.003, + "loss": 4.1163, + "step": 7063 + }, + { + "epoch": 0.07064, + "grad_norm": 0.9309447457188175, + "learning_rate": 0.003, + "loss": 4.1322, + "step": 7064 + }, + { + "epoch": 0.07065, + "grad_norm": 0.9478643723606038, + "learning_rate": 0.003, + "loss": 4.1255, + "step": 7065 + }, + { + "epoch": 0.07066, + "grad_norm": 0.8659841976692164, + "learning_rate": 0.003, + "loss": 4.1274, + "step": 7066 + }, + { + "epoch": 0.07067, + "grad_norm": 0.820962997351535, + "learning_rate": 0.003, + "loss": 4.1571, + "step": 7067 + }, + { + "epoch": 0.07068, + "grad_norm": 0.8086119331768552, + "learning_rate": 0.003, + "loss": 4.0912, + "step": 7068 + }, + { + "epoch": 0.07069, + "grad_norm": 0.7230672452463993, + "learning_rate": 0.003, + "loss": 4.0936, + "step": 7069 + }, + { + "epoch": 0.0707, + "grad_norm": 0.6999038291572481, + "learning_rate": 0.003, + "loss": 4.1175, + "step": 7070 + }, + { + "epoch": 0.07071, + "grad_norm": 0.646981573011097, + "learning_rate": 0.003, + "loss": 4.0935, + "step": 7071 + }, + { + "epoch": 0.07072, + "grad_norm": 0.6484005206388396, + "learning_rate": 0.003, + "loss": 4.0994, + "step": 7072 + }, + { + "epoch": 0.07073, + "grad_norm": 0.650657762847781, + "learning_rate": 0.003, + "loss": 4.1142, + "step": 7073 + }, + { + "epoch": 0.07074, + "grad_norm": 0.6775125603608778, + "learning_rate": 0.003, + "loss": 4.0996, + "step": 7074 + }, + { + "epoch": 0.07075, + "grad_norm": 0.6083872877550612, + "learning_rate": 0.003, + "loss": 4.1356, + "step": 7075 + }, + { + "epoch": 0.07076, + "grad_norm": 0.6452208544105434, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 7076 + }, + { + "epoch": 0.07077, + "grad_norm": 0.6649673970366402, + "learning_rate": 0.003, + "loss": 4.0958, + "step": 7077 + }, + { + "epoch": 0.07078, + "grad_norm": 0.6783088752414806, + "learning_rate": 0.003, + "loss": 4.1197, + "step": 7078 + }, + { + "epoch": 0.07079, + "grad_norm": 0.7710592718558568, + "learning_rate": 0.003, + "loss": 4.1241, + "step": 7079 + }, + { + "epoch": 0.0708, + "grad_norm": 0.958668145233532, + "learning_rate": 0.003, + "loss": 4.071, + "step": 7080 + }, + { + "epoch": 0.07081, + "grad_norm": 1.2533157696452137, + "learning_rate": 0.003, + "loss": 4.1191, + "step": 7081 + }, + { + "epoch": 0.07082, + "grad_norm": 0.7341633652839094, + "learning_rate": 0.003, + "loss": 4.1009, + "step": 7082 + }, + { + "epoch": 0.07083, + "grad_norm": 0.7125445429868206, + "learning_rate": 0.003, + "loss": 4.0874, + "step": 7083 + }, + { + "epoch": 0.07084, + "grad_norm": 0.8323202978555945, + "learning_rate": 0.003, + "loss": 4.1028, + "step": 7084 + }, + { + "epoch": 0.07085, + "grad_norm": 0.8775075656235146, + "learning_rate": 0.003, + "loss": 4.1231, + "step": 7085 + }, + { + "epoch": 0.07086, + "grad_norm": 0.8415506665597984, + "learning_rate": 0.003, + "loss": 4.1061, + "step": 7086 + }, + { + "epoch": 0.07087, + "grad_norm": 0.8886268254263243, + "learning_rate": 0.003, + "loss": 4.1043, + "step": 7087 + }, + { + "epoch": 0.07088, + "grad_norm": 1.1007254498444334, + "learning_rate": 0.003, + "loss": 4.083, + "step": 7088 + }, + { + "epoch": 0.07089, + "grad_norm": 1.077560245637077, + "learning_rate": 0.003, + "loss": 4.1117, + "step": 7089 + }, + { + "epoch": 0.0709, + "grad_norm": 1.0164885049934373, + "learning_rate": 0.003, + "loss": 4.1116, + "step": 7090 + }, + { + "epoch": 0.07091, + "grad_norm": 0.9408820922271797, + "learning_rate": 0.003, + "loss": 4.1018, + "step": 7091 + }, + { + "epoch": 0.07092, + "grad_norm": 0.9903430667983368, + "learning_rate": 0.003, + "loss": 4.1063, + "step": 7092 + }, + { + "epoch": 0.07093, + "grad_norm": 1.0927513972609133, + "learning_rate": 0.003, + "loss": 4.1388, + "step": 7093 + }, + { + "epoch": 0.07094, + "grad_norm": 0.9057260963699147, + "learning_rate": 0.003, + "loss": 4.1507, + "step": 7094 + }, + { + "epoch": 0.07095, + "grad_norm": 0.961594432402731, + "learning_rate": 0.003, + "loss": 4.1407, + "step": 7095 + }, + { + "epoch": 0.07096, + "grad_norm": 0.9337093837213043, + "learning_rate": 0.003, + "loss": 4.1069, + "step": 7096 + }, + { + "epoch": 0.07097, + "grad_norm": 0.8978322204376042, + "learning_rate": 0.003, + "loss": 4.1303, + "step": 7097 + }, + { + "epoch": 0.07098, + "grad_norm": 0.8250215892500493, + "learning_rate": 0.003, + "loss": 4.1318, + "step": 7098 + }, + { + "epoch": 0.07099, + "grad_norm": 0.7161931173541564, + "learning_rate": 0.003, + "loss": 4.1273, + "step": 7099 + }, + { + "epoch": 0.071, + "grad_norm": 0.7049995268154481, + "learning_rate": 0.003, + "loss": 4.1194, + "step": 7100 + }, + { + "epoch": 0.07101, + "grad_norm": 0.7656370792822328, + "learning_rate": 0.003, + "loss": 4.0995, + "step": 7101 + }, + { + "epoch": 0.07102, + "grad_norm": 0.7875547696719094, + "learning_rate": 0.003, + "loss": 4.1158, + "step": 7102 + }, + { + "epoch": 0.07103, + "grad_norm": 0.8916531396092162, + "learning_rate": 0.003, + "loss": 4.1292, + "step": 7103 + }, + { + "epoch": 0.07104, + "grad_norm": 1.0114205653924595, + "learning_rate": 0.003, + "loss": 4.1284, + "step": 7104 + }, + { + "epoch": 0.07105, + "grad_norm": 1.1741497513128247, + "learning_rate": 0.003, + "loss": 4.1241, + "step": 7105 + }, + { + "epoch": 0.07106, + "grad_norm": 0.7625197875421197, + "learning_rate": 0.003, + "loss": 4.1261, + "step": 7106 + }, + { + "epoch": 0.07107, + "grad_norm": 0.6257457801610031, + "learning_rate": 0.003, + "loss": 4.1208, + "step": 7107 + }, + { + "epoch": 0.07108, + "grad_norm": 0.722291112408107, + "learning_rate": 0.003, + "loss": 4.1298, + "step": 7108 + }, + { + "epoch": 0.07109, + "grad_norm": 0.8199337468043636, + "learning_rate": 0.003, + "loss": 4.1157, + "step": 7109 + }, + { + "epoch": 0.0711, + "grad_norm": 0.9809541172196407, + "learning_rate": 0.003, + "loss": 4.093, + "step": 7110 + }, + { + "epoch": 0.07111, + "grad_norm": 1.3461787795853564, + "learning_rate": 0.003, + "loss": 4.1208, + "step": 7111 + }, + { + "epoch": 0.07112, + "grad_norm": 0.8890390770134524, + "learning_rate": 0.003, + "loss": 4.1142, + "step": 7112 + }, + { + "epoch": 0.07113, + "grad_norm": 0.795752170005651, + "learning_rate": 0.003, + "loss": 4.1472, + "step": 7113 + }, + { + "epoch": 0.07114, + "grad_norm": 0.7764623741527109, + "learning_rate": 0.003, + "loss": 4.1218, + "step": 7114 + }, + { + "epoch": 0.07115, + "grad_norm": 0.7626873101755763, + "learning_rate": 0.003, + "loss": 4.1063, + "step": 7115 + }, + { + "epoch": 0.07116, + "grad_norm": 0.709900661004303, + "learning_rate": 0.003, + "loss": 4.1096, + "step": 7116 + }, + { + "epoch": 0.07117, + "grad_norm": 0.7447218585864936, + "learning_rate": 0.003, + "loss": 4.1227, + "step": 7117 + }, + { + "epoch": 0.07118, + "grad_norm": 0.8145999556594223, + "learning_rate": 0.003, + "loss": 4.1298, + "step": 7118 + }, + { + "epoch": 0.07119, + "grad_norm": 0.969296790631849, + "learning_rate": 0.003, + "loss": 4.1101, + "step": 7119 + }, + { + "epoch": 0.0712, + "grad_norm": 1.110298419738427, + "learning_rate": 0.003, + "loss": 4.1131, + "step": 7120 + }, + { + "epoch": 0.07121, + "grad_norm": 0.7765027521208008, + "learning_rate": 0.003, + "loss": 4.1076, + "step": 7121 + }, + { + "epoch": 0.07122, + "grad_norm": 0.6976330840832459, + "learning_rate": 0.003, + "loss": 4.0984, + "step": 7122 + }, + { + "epoch": 0.07123, + "grad_norm": 0.7282209002398681, + "learning_rate": 0.003, + "loss": 4.0927, + "step": 7123 + }, + { + "epoch": 0.07124, + "grad_norm": 0.8034457748599896, + "learning_rate": 0.003, + "loss": 4.0898, + "step": 7124 + }, + { + "epoch": 0.07125, + "grad_norm": 0.9445995479508306, + "learning_rate": 0.003, + "loss": 4.0992, + "step": 7125 + }, + { + "epoch": 0.07126, + "grad_norm": 1.1268646555573658, + "learning_rate": 0.003, + "loss": 4.1371, + "step": 7126 + }, + { + "epoch": 0.07127, + "grad_norm": 0.8675499183490605, + "learning_rate": 0.003, + "loss": 4.0989, + "step": 7127 + }, + { + "epoch": 0.07128, + "grad_norm": 0.7495770001240798, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 7128 + }, + { + "epoch": 0.07129, + "grad_norm": 0.7927069334940464, + "learning_rate": 0.003, + "loss": 4.1188, + "step": 7129 + }, + { + "epoch": 0.0713, + "grad_norm": 0.8376734219621176, + "learning_rate": 0.003, + "loss": 4.1083, + "step": 7130 + }, + { + "epoch": 0.07131, + "grad_norm": 0.9019229212023059, + "learning_rate": 0.003, + "loss": 4.0929, + "step": 7131 + }, + { + "epoch": 0.07132, + "grad_norm": 0.958492302536442, + "learning_rate": 0.003, + "loss": 4.1359, + "step": 7132 + }, + { + "epoch": 0.07133, + "grad_norm": 0.9703595503667748, + "learning_rate": 0.003, + "loss": 4.1434, + "step": 7133 + }, + { + "epoch": 0.07134, + "grad_norm": 1.0020602010967854, + "learning_rate": 0.003, + "loss": 4.0864, + "step": 7134 + }, + { + "epoch": 0.07135, + "grad_norm": 0.8801413035960939, + "learning_rate": 0.003, + "loss": 4.1129, + "step": 7135 + }, + { + "epoch": 0.07136, + "grad_norm": 0.7990610722753618, + "learning_rate": 0.003, + "loss": 4.147, + "step": 7136 + }, + { + "epoch": 0.07137, + "grad_norm": 0.793252959668317, + "learning_rate": 0.003, + "loss": 4.11, + "step": 7137 + }, + { + "epoch": 0.07138, + "grad_norm": 0.9362895477284106, + "learning_rate": 0.003, + "loss": 4.1098, + "step": 7138 + }, + { + "epoch": 0.07139, + "grad_norm": 1.0323509656605805, + "learning_rate": 0.003, + "loss": 4.1163, + "step": 7139 + }, + { + "epoch": 0.0714, + "grad_norm": 0.9495162380496068, + "learning_rate": 0.003, + "loss": 4.1211, + "step": 7140 + }, + { + "epoch": 0.07141, + "grad_norm": 0.9575955550409052, + "learning_rate": 0.003, + "loss": 4.101, + "step": 7141 + }, + { + "epoch": 0.07142, + "grad_norm": 0.8744238864306674, + "learning_rate": 0.003, + "loss": 4.0901, + "step": 7142 + }, + { + "epoch": 0.07143, + "grad_norm": 0.8469591921258507, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 7143 + }, + { + "epoch": 0.07144, + "grad_norm": 0.8348336779926869, + "learning_rate": 0.003, + "loss": 4.0936, + "step": 7144 + }, + { + "epoch": 0.07145, + "grad_norm": 1.20426435156208, + "learning_rate": 0.003, + "loss": 4.1252, + "step": 7145 + }, + { + "epoch": 0.07146, + "grad_norm": 0.9331877874118307, + "learning_rate": 0.003, + "loss": 4.1185, + "step": 7146 + }, + { + "epoch": 0.07147, + "grad_norm": 0.7956424973642607, + "learning_rate": 0.003, + "loss": 4.1237, + "step": 7147 + }, + { + "epoch": 0.07148, + "grad_norm": 0.6333045591280563, + "learning_rate": 0.003, + "loss": 4.1239, + "step": 7148 + }, + { + "epoch": 0.07149, + "grad_norm": 0.6566262136153223, + "learning_rate": 0.003, + "loss": 4.1273, + "step": 7149 + }, + { + "epoch": 0.0715, + "grad_norm": 0.7318802572187021, + "learning_rate": 0.003, + "loss": 4.1087, + "step": 7150 + }, + { + "epoch": 0.07151, + "grad_norm": 0.7829355969313044, + "learning_rate": 0.003, + "loss": 4.1096, + "step": 7151 + }, + { + "epoch": 0.07152, + "grad_norm": 0.8583613260253945, + "learning_rate": 0.003, + "loss": 4.1282, + "step": 7152 + }, + { + "epoch": 0.07153, + "grad_norm": 0.92982352203367, + "learning_rate": 0.003, + "loss": 4.1147, + "step": 7153 + }, + { + "epoch": 0.07154, + "grad_norm": 0.9644437837333677, + "learning_rate": 0.003, + "loss": 4.1498, + "step": 7154 + }, + { + "epoch": 0.07155, + "grad_norm": 0.978358778668515, + "learning_rate": 0.003, + "loss": 4.1149, + "step": 7155 + }, + { + "epoch": 0.07156, + "grad_norm": 0.7507926446560859, + "learning_rate": 0.003, + "loss": 4.1172, + "step": 7156 + }, + { + "epoch": 0.07157, + "grad_norm": 0.6945396009037457, + "learning_rate": 0.003, + "loss": 4.1004, + "step": 7157 + }, + { + "epoch": 0.07158, + "grad_norm": 0.6202573792874552, + "learning_rate": 0.003, + "loss": 4.1053, + "step": 7158 + }, + { + "epoch": 0.07159, + "grad_norm": 0.6406745858919619, + "learning_rate": 0.003, + "loss": 4.1111, + "step": 7159 + }, + { + "epoch": 0.0716, + "grad_norm": 0.6759721738325275, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 7160 + }, + { + "epoch": 0.07161, + "grad_norm": 0.6773350977546571, + "learning_rate": 0.003, + "loss": 4.1124, + "step": 7161 + }, + { + "epoch": 0.07162, + "grad_norm": 0.8134936196821616, + "learning_rate": 0.003, + "loss": 4.1165, + "step": 7162 + }, + { + "epoch": 0.07163, + "grad_norm": 1.0238124509153523, + "learning_rate": 0.003, + "loss": 4.1117, + "step": 7163 + }, + { + "epoch": 0.07164, + "grad_norm": 1.0689438667971007, + "learning_rate": 0.003, + "loss": 4.0973, + "step": 7164 + }, + { + "epoch": 0.07165, + "grad_norm": 0.7928160801351044, + "learning_rate": 0.003, + "loss": 4.0961, + "step": 7165 + }, + { + "epoch": 0.07166, + "grad_norm": 0.7238005610299342, + "learning_rate": 0.003, + "loss": 4.0966, + "step": 7166 + }, + { + "epoch": 0.07167, + "grad_norm": 0.7943353284460141, + "learning_rate": 0.003, + "loss": 4.1138, + "step": 7167 + }, + { + "epoch": 0.07168, + "grad_norm": 0.9366138493792299, + "learning_rate": 0.003, + "loss": 4.1346, + "step": 7168 + }, + { + "epoch": 0.07169, + "grad_norm": 0.9317917412995604, + "learning_rate": 0.003, + "loss": 4.1022, + "step": 7169 + }, + { + "epoch": 0.0717, + "grad_norm": 0.9498234616222482, + "learning_rate": 0.003, + "loss": 4.0961, + "step": 7170 + }, + { + "epoch": 0.07171, + "grad_norm": 1.0293570458877537, + "learning_rate": 0.003, + "loss": 4.1234, + "step": 7171 + }, + { + "epoch": 0.07172, + "grad_norm": 0.9825020297184559, + "learning_rate": 0.003, + "loss": 4.1273, + "step": 7172 + }, + { + "epoch": 0.07173, + "grad_norm": 0.8679097068915133, + "learning_rate": 0.003, + "loss": 4.1242, + "step": 7173 + }, + { + "epoch": 0.07174, + "grad_norm": 0.8695176732805833, + "learning_rate": 0.003, + "loss": 4.0988, + "step": 7174 + }, + { + "epoch": 0.07175, + "grad_norm": 0.9453745541920484, + "learning_rate": 0.003, + "loss": 4.1366, + "step": 7175 + }, + { + "epoch": 0.07176, + "grad_norm": 0.9192625870625667, + "learning_rate": 0.003, + "loss": 4.1125, + "step": 7176 + }, + { + "epoch": 0.07177, + "grad_norm": 0.9892316081377227, + "learning_rate": 0.003, + "loss": 4.1181, + "step": 7177 + }, + { + "epoch": 0.07178, + "grad_norm": 1.1173320244326939, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 7178 + }, + { + "epoch": 0.07179, + "grad_norm": 0.9062788363818038, + "learning_rate": 0.003, + "loss": 4.1165, + "step": 7179 + }, + { + "epoch": 0.0718, + "grad_norm": 0.8000050025573026, + "learning_rate": 0.003, + "loss": 4.1228, + "step": 7180 + }, + { + "epoch": 0.07181, + "grad_norm": 0.873878962358314, + "learning_rate": 0.003, + "loss": 4.117, + "step": 7181 + }, + { + "epoch": 0.07182, + "grad_norm": 0.8940125268309896, + "learning_rate": 0.003, + "loss": 4.128, + "step": 7182 + }, + { + "epoch": 0.07183, + "grad_norm": 0.9553838594697793, + "learning_rate": 0.003, + "loss": 4.1263, + "step": 7183 + }, + { + "epoch": 0.07184, + "grad_norm": 0.8409034686100991, + "learning_rate": 0.003, + "loss": 4.1007, + "step": 7184 + }, + { + "epoch": 0.07185, + "grad_norm": 0.8965763690380733, + "learning_rate": 0.003, + "loss": 4.1097, + "step": 7185 + }, + { + "epoch": 0.07186, + "grad_norm": 1.0738828307599333, + "learning_rate": 0.003, + "loss": 4.1354, + "step": 7186 + }, + { + "epoch": 0.07187, + "grad_norm": 1.121879584063851, + "learning_rate": 0.003, + "loss": 4.1261, + "step": 7187 + }, + { + "epoch": 0.07188, + "grad_norm": 0.7967451613183767, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 7188 + }, + { + "epoch": 0.07189, + "grad_norm": 0.8216455172743733, + "learning_rate": 0.003, + "loss": 4.1332, + "step": 7189 + }, + { + "epoch": 0.0719, + "grad_norm": 0.8712547518094813, + "learning_rate": 0.003, + "loss": 4.1318, + "step": 7190 + }, + { + "epoch": 0.07191, + "grad_norm": 0.7876295918843358, + "learning_rate": 0.003, + "loss": 4.136, + "step": 7191 + }, + { + "epoch": 0.07192, + "grad_norm": 0.8009874929822709, + "learning_rate": 0.003, + "loss": 4.0981, + "step": 7192 + }, + { + "epoch": 0.07193, + "grad_norm": 0.8465208583922181, + "learning_rate": 0.003, + "loss": 4.1224, + "step": 7193 + }, + { + "epoch": 0.07194, + "grad_norm": 0.919178707859649, + "learning_rate": 0.003, + "loss": 4.1214, + "step": 7194 + }, + { + "epoch": 0.07195, + "grad_norm": 1.0483728768665097, + "learning_rate": 0.003, + "loss": 4.1011, + "step": 7195 + }, + { + "epoch": 0.07196, + "grad_norm": 1.0529880709135624, + "learning_rate": 0.003, + "loss": 4.1114, + "step": 7196 + }, + { + "epoch": 0.07197, + "grad_norm": 0.8929728066328368, + "learning_rate": 0.003, + "loss": 4.1305, + "step": 7197 + }, + { + "epoch": 0.07198, + "grad_norm": 0.8510797490053401, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 7198 + }, + { + "epoch": 0.07199, + "grad_norm": 0.7701251238667544, + "learning_rate": 0.003, + "loss": 4.0992, + "step": 7199 + }, + { + "epoch": 0.072, + "grad_norm": 0.5992356375970112, + "learning_rate": 0.003, + "loss": 4.1285, + "step": 7200 + }, + { + "epoch": 0.07201, + "grad_norm": 0.6127454964693743, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 7201 + }, + { + "epoch": 0.07202, + "grad_norm": 0.5750069845188995, + "learning_rate": 0.003, + "loss": 4.1225, + "step": 7202 + }, + { + "epoch": 0.07203, + "grad_norm": 0.5644356454850563, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 7203 + }, + { + "epoch": 0.07204, + "grad_norm": 0.607948596229715, + "learning_rate": 0.003, + "loss": 4.1266, + "step": 7204 + }, + { + "epoch": 0.07205, + "grad_norm": 0.575336251782507, + "learning_rate": 0.003, + "loss": 4.1212, + "step": 7205 + }, + { + "epoch": 0.07206, + "grad_norm": 0.6206756982347824, + "learning_rate": 0.003, + "loss": 4.1002, + "step": 7206 + }, + { + "epoch": 0.07207, + "grad_norm": 0.6744817053594698, + "learning_rate": 0.003, + "loss": 4.0991, + "step": 7207 + }, + { + "epoch": 0.07208, + "grad_norm": 0.7220807677197786, + "learning_rate": 0.003, + "loss": 4.0955, + "step": 7208 + }, + { + "epoch": 0.07209, + "grad_norm": 0.8377268449123019, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 7209 + }, + { + "epoch": 0.0721, + "grad_norm": 1.0721743343425194, + "learning_rate": 0.003, + "loss": 4.0992, + "step": 7210 + }, + { + "epoch": 0.07211, + "grad_norm": 0.9661201701564104, + "learning_rate": 0.003, + "loss": 4.1127, + "step": 7211 + }, + { + "epoch": 0.07212, + "grad_norm": 0.7690576613887181, + "learning_rate": 0.003, + "loss": 4.1186, + "step": 7212 + }, + { + "epoch": 0.07213, + "grad_norm": 0.581038878715819, + "learning_rate": 0.003, + "loss": 4.1265, + "step": 7213 + }, + { + "epoch": 0.07214, + "grad_norm": 0.6435911406628065, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 7214 + }, + { + "epoch": 0.07215, + "grad_norm": 0.8429785330520374, + "learning_rate": 0.003, + "loss": 4.1038, + "step": 7215 + }, + { + "epoch": 0.07216, + "grad_norm": 1.0387688679965494, + "learning_rate": 0.003, + "loss": 4.1198, + "step": 7216 + }, + { + "epoch": 0.07217, + "grad_norm": 1.0443475300961371, + "learning_rate": 0.003, + "loss": 4.1133, + "step": 7217 + }, + { + "epoch": 0.07218, + "grad_norm": 0.8473614781155715, + "learning_rate": 0.003, + "loss": 4.112, + "step": 7218 + }, + { + "epoch": 0.07219, + "grad_norm": 0.8315491811072131, + "learning_rate": 0.003, + "loss": 4.0906, + "step": 7219 + }, + { + "epoch": 0.0722, + "grad_norm": 1.0296189666178754, + "learning_rate": 0.003, + "loss": 4.122, + "step": 7220 + }, + { + "epoch": 0.07221, + "grad_norm": 1.058643563273535, + "learning_rate": 0.003, + "loss": 4.1232, + "step": 7221 + }, + { + "epoch": 0.07222, + "grad_norm": 0.8979290937756685, + "learning_rate": 0.003, + "loss": 4.1093, + "step": 7222 + }, + { + "epoch": 0.07223, + "grad_norm": 0.7842228557729956, + "learning_rate": 0.003, + "loss": 4.1202, + "step": 7223 + }, + { + "epoch": 0.07224, + "grad_norm": 0.774784015172213, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 7224 + }, + { + "epoch": 0.07225, + "grad_norm": 0.77416880062085, + "learning_rate": 0.003, + "loss": 4.1137, + "step": 7225 + }, + { + "epoch": 0.07226, + "grad_norm": 0.8594028278800052, + "learning_rate": 0.003, + "loss": 4.1196, + "step": 7226 + }, + { + "epoch": 0.07227, + "grad_norm": 0.9339883501438095, + "learning_rate": 0.003, + "loss": 4.1026, + "step": 7227 + }, + { + "epoch": 0.07228, + "grad_norm": 1.009238406341426, + "learning_rate": 0.003, + "loss": 4.1009, + "step": 7228 + }, + { + "epoch": 0.07229, + "grad_norm": 0.9248237524022925, + "learning_rate": 0.003, + "loss": 4.1162, + "step": 7229 + }, + { + "epoch": 0.0723, + "grad_norm": 0.7879348638628303, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 7230 + }, + { + "epoch": 0.07231, + "grad_norm": 0.8125590881523109, + "learning_rate": 0.003, + "loss": 4.1249, + "step": 7231 + }, + { + "epoch": 0.07232, + "grad_norm": 0.7397244009053785, + "learning_rate": 0.003, + "loss": 4.1143, + "step": 7232 + }, + { + "epoch": 0.07233, + "grad_norm": 0.8631670605854277, + "learning_rate": 0.003, + "loss": 4.1198, + "step": 7233 + }, + { + "epoch": 0.07234, + "grad_norm": 0.9590374718321888, + "learning_rate": 0.003, + "loss": 4.0979, + "step": 7234 + }, + { + "epoch": 0.07235, + "grad_norm": 1.1821140388987355, + "learning_rate": 0.003, + "loss": 4.152, + "step": 7235 + }, + { + "epoch": 0.07236, + "grad_norm": 1.0912993445758663, + "learning_rate": 0.003, + "loss": 4.1468, + "step": 7236 + }, + { + "epoch": 0.07237, + "grad_norm": 1.0186661817583151, + "learning_rate": 0.003, + "loss": 4.1103, + "step": 7237 + }, + { + "epoch": 0.07238, + "grad_norm": 0.8827875731907374, + "learning_rate": 0.003, + "loss": 4.132, + "step": 7238 + }, + { + "epoch": 0.07239, + "grad_norm": 1.0101708637839817, + "learning_rate": 0.003, + "loss": 4.1397, + "step": 7239 + }, + { + "epoch": 0.0724, + "grad_norm": 0.9436996667559744, + "learning_rate": 0.003, + "loss": 4.0983, + "step": 7240 + }, + { + "epoch": 0.07241, + "grad_norm": 0.924594063482519, + "learning_rate": 0.003, + "loss": 4.1432, + "step": 7241 + }, + { + "epoch": 0.07242, + "grad_norm": 1.0534829249112354, + "learning_rate": 0.003, + "loss": 4.1018, + "step": 7242 + }, + { + "epoch": 0.07243, + "grad_norm": 0.9856885879894006, + "learning_rate": 0.003, + "loss": 4.1349, + "step": 7243 + }, + { + "epoch": 0.07244, + "grad_norm": 0.8968354162645473, + "learning_rate": 0.003, + "loss": 4.1167, + "step": 7244 + }, + { + "epoch": 0.07245, + "grad_norm": 0.7432909812609705, + "learning_rate": 0.003, + "loss": 4.1134, + "step": 7245 + }, + { + "epoch": 0.07246, + "grad_norm": 0.7581401567213042, + "learning_rate": 0.003, + "loss": 4.1226, + "step": 7246 + }, + { + "epoch": 0.07247, + "grad_norm": 0.7827779191397117, + "learning_rate": 0.003, + "loss": 4.141, + "step": 7247 + }, + { + "epoch": 0.07248, + "grad_norm": 0.8301715949958813, + "learning_rate": 0.003, + "loss": 4.1055, + "step": 7248 + }, + { + "epoch": 0.07249, + "grad_norm": 0.8525741406819738, + "learning_rate": 0.003, + "loss": 4.0884, + "step": 7249 + }, + { + "epoch": 0.0725, + "grad_norm": 0.8907335082596, + "learning_rate": 0.003, + "loss": 4.1023, + "step": 7250 + }, + { + "epoch": 0.07251, + "grad_norm": 0.8802924249074248, + "learning_rate": 0.003, + "loss": 4.1195, + "step": 7251 + }, + { + "epoch": 0.07252, + "grad_norm": 0.9494861985507634, + "learning_rate": 0.003, + "loss": 4.059, + "step": 7252 + }, + { + "epoch": 0.07253, + "grad_norm": 0.9816725505864071, + "learning_rate": 0.003, + "loss": 4.0898, + "step": 7253 + }, + { + "epoch": 0.07254, + "grad_norm": 1.0443814177090267, + "learning_rate": 0.003, + "loss": 4.139, + "step": 7254 + }, + { + "epoch": 0.07255, + "grad_norm": 0.9747578800346165, + "learning_rate": 0.003, + "loss": 4.1331, + "step": 7255 + }, + { + "epoch": 0.07256, + "grad_norm": 0.9623975370695055, + "learning_rate": 0.003, + "loss": 4.1353, + "step": 7256 + }, + { + "epoch": 0.07257, + "grad_norm": 0.8844084755688933, + "learning_rate": 0.003, + "loss": 4.1195, + "step": 7257 + }, + { + "epoch": 0.07258, + "grad_norm": 0.7873250461581538, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 7258 + }, + { + "epoch": 0.07259, + "grad_norm": 0.7048185689672526, + "learning_rate": 0.003, + "loss": 4.1117, + "step": 7259 + }, + { + "epoch": 0.0726, + "grad_norm": 0.7212279689840713, + "learning_rate": 0.003, + "loss": 4.088, + "step": 7260 + }, + { + "epoch": 0.07261, + "grad_norm": 0.6900512375900392, + "learning_rate": 0.003, + "loss": 4.1137, + "step": 7261 + }, + { + "epoch": 0.07262, + "grad_norm": 0.7136129301757733, + "learning_rate": 0.003, + "loss": 4.1071, + "step": 7262 + }, + { + "epoch": 0.07263, + "grad_norm": 0.7469633157532889, + "learning_rate": 0.003, + "loss": 4.1134, + "step": 7263 + }, + { + "epoch": 0.07264, + "grad_norm": 0.8012211214960027, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 7264 + }, + { + "epoch": 0.07265, + "grad_norm": 0.7552911178426823, + "learning_rate": 0.003, + "loss": 4.1091, + "step": 7265 + }, + { + "epoch": 0.07266, + "grad_norm": 0.7077347824210926, + "learning_rate": 0.003, + "loss": 4.1258, + "step": 7266 + }, + { + "epoch": 0.07267, + "grad_norm": 0.6760989582415581, + "learning_rate": 0.003, + "loss": 4.1103, + "step": 7267 + }, + { + "epoch": 0.07268, + "grad_norm": 0.7437332060079201, + "learning_rate": 0.003, + "loss": 4.118, + "step": 7268 + }, + { + "epoch": 0.07269, + "grad_norm": 0.9444827297437564, + "learning_rate": 0.003, + "loss": 4.1084, + "step": 7269 + }, + { + "epoch": 0.0727, + "grad_norm": 1.1220594233552328, + "learning_rate": 0.003, + "loss": 4.0989, + "step": 7270 + }, + { + "epoch": 0.07271, + "grad_norm": 0.8200379597299912, + "learning_rate": 0.003, + "loss": 4.1044, + "step": 7271 + }, + { + "epoch": 0.07272, + "grad_norm": 0.7596490807886606, + "learning_rate": 0.003, + "loss": 4.0966, + "step": 7272 + }, + { + "epoch": 0.07273, + "grad_norm": 0.7884580146314756, + "learning_rate": 0.003, + "loss": 4.1131, + "step": 7273 + }, + { + "epoch": 0.07274, + "grad_norm": 0.7179271579858106, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 7274 + }, + { + "epoch": 0.07275, + "grad_norm": 0.6409693363218573, + "learning_rate": 0.003, + "loss": 4.0857, + "step": 7275 + }, + { + "epoch": 0.07276, + "grad_norm": 0.561214096824011, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 7276 + }, + { + "epoch": 0.07277, + "grad_norm": 0.5256694600157773, + "learning_rate": 0.003, + "loss": 4.0947, + "step": 7277 + }, + { + "epoch": 0.07278, + "grad_norm": 0.513949681102309, + "learning_rate": 0.003, + "loss": 4.1015, + "step": 7278 + }, + { + "epoch": 0.07279, + "grad_norm": 0.49570529414604636, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 7279 + }, + { + "epoch": 0.0728, + "grad_norm": 0.6204492806369919, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 7280 + }, + { + "epoch": 0.07281, + "grad_norm": 0.9083869384535228, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 7281 + }, + { + "epoch": 0.07282, + "grad_norm": 1.2773527610872655, + "learning_rate": 0.003, + "loss": 4.1165, + "step": 7282 + }, + { + "epoch": 0.07283, + "grad_norm": 0.7627102174193735, + "learning_rate": 0.003, + "loss": 4.1234, + "step": 7283 + }, + { + "epoch": 0.07284, + "grad_norm": 0.6303760922439519, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 7284 + }, + { + "epoch": 0.07285, + "grad_norm": 0.703121523726995, + "learning_rate": 0.003, + "loss": 4.1106, + "step": 7285 + }, + { + "epoch": 0.07286, + "grad_norm": 0.7516732999345842, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 7286 + }, + { + "epoch": 0.07287, + "grad_norm": 0.8402680521780728, + "learning_rate": 0.003, + "loss": 4.1095, + "step": 7287 + }, + { + "epoch": 0.07288, + "grad_norm": 0.8639792163628891, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 7288 + }, + { + "epoch": 0.07289, + "grad_norm": 0.9106770231204078, + "learning_rate": 0.003, + "loss": 4.111, + "step": 7289 + }, + { + "epoch": 0.0729, + "grad_norm": 0.8569039805553732, + "learning_rate": 0.003, + "loss": 4.133, + "step": 7290 + }, + { + "epoch": 0.07291, + "grad_norm": 0.8741502241373251, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 7291 + }, + { + "epoch": 0.07292, + "grad_norm": 0.9093181247050962, + "learning_rate": 0.003, + "loss": 4.1104, + "step": 7292 + }, + { + "epoch": 0.07293, + "grad_norm": 1.0546113868958757, + "learning_rate": 0.003, + "loss": 4.1031, + "step": 7293 + }, + { + "epoch": 0.07294, + "grad_norm": 1.0193664307418668, + "learning_rate": 0.003, + "loss": 4.092, + "step": 7294 + }, + { + "epoch": 0.07295, + "grad_norm": 1.0910195655187844, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 7295 + }, + { + "epoch": 0.07296, + "grad_norm": 0.8335499181973589, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 7296 + }, + { + "epoch": 0.07297, + "grad_norm": 0.6580676089467326, + "learning_rate": 0.003, + "loss": 4.126, + "step": 7297 + }, + { + "epoch": 0.07298, + "grad_norm": 0.5832643079122204, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 7298 + }, + { + "epoch": 0.07299, + "grad_norm": 0.6303194883819615, + "learning_rate": 0.003, + "loss": 4.0988, + "step": 7299 + }, + { + "epoch": 0.073, + "grad_norm": 0.6469636223827426, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 7300 + }, + { + "epoch": 0.07301, + "grad_norm": 0.6597994798980442, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 7301 + }, + { + "epoch": 0.07302, + "grad_norm": 0.6982812827344924, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 7302 + }, + { + "epoch": 0.07303, + "grad_norm": 0.7027447626804699, + "learning_rate": 0.003, + "loss": 4.1082, + "step": 7303 + }, + { + "epoch": 0.07304, + "grad_norm": 0.7898566798558829, + "learning_rate": 0.003, + "loss": 4.0905, + "step": 7304 + }, + { + "epoch": 0.07305, + "grad_norm": 0.8619601714981096, + "learning_rate": 0.003, + "loss": 4.0858, + "step": 7305 + }, + { + "epoch": 0.07306, + "grad_norm": 1.0874296421242524, + "learning_rate": 0.003, + "loss": 4.1274, + "step": 7306 + }, + { + "epoch": 0.07307, + "grad_norm": 1.1513689835246426, + "learning_rate": 0.003, + "loss": 4.1405, + "step": 7307 + }, + { + "epoch": 0.07308, + "grad_norm": 0.8638122639112095, + "learning_rate": 0.003, + "loss": 4.1022, + "step": 7308 + }, + { + "epoch": 0.07309, + "grad_norm": 0.8255250333131889, + "learning_rate": 0.003, + "loss": 4.1088, + "step": 7309 + }, + { + "epoch": 0.0731, + "grad_norm": 0.9337338907492057, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 7310 + }, + { + "epoch": 0.07311, + "grad_norm": 0.9258041196216139, + "learning_rate": 0.003, + "loss": 4.1147, + "step": 7311 + }, + { + "epoch": 0.07312, + "grad_norm": 0.951937894172149, + "learning_rate": 0.003, + "loss": 4.1268, + "step": 7312 + }, + { + "epoch": 0.07313, + "grad_norm": 1.0078253665907535, + "learning_rate": 0.003, + "loss": 4.1176, + "step": 7313 + }, + { + "epoch": 0.07314, + "grad_norm": 0.901224372401486, + "learning_rate": 0.003, + "loss": 4.1279, + "step": 7314 + }, + { + "epoch": 0.07315, + "grad_norm": 0.7944852254095126, + "learning_rate": 0.003, + "loss": 4.1111, + "step": 7315 + }, + { + "epoch": 0.07316, + "grad_norm": 0.783421376743893, + "learning_rate": 0.003, + "loss": 4.1285, + "step": 7316 + }, + { + "epoch": 0.07317, + "grad_norm": 0.8368936906337524, + "learning_rate": 0.003, + "loss": 4.1126, + "step": 7317 + }, + { + "epoch": 0.07318, + "grad_norm": 1.033174150909009, + "learning_rate": 0.003, + "loss": 4.1026, + "step": 7318 + }, + { + "epoch": 0.07319, + "grad_norm": 1.27108174493455, + "learning_rate": 0.003, + "loss": 4.1282, + "step": 7319 + }, + { + "epoch": 0.0732, + "grad_norm": 0.7406462978297551, + "learning_rate": 0.003, + "loss": 4.0973, + "step": 7320 + }, + { + "epoch": 0.07321, + "grad_norm": 0.7025965454631923, + "learning_rate": 0.003, + "loss": 4.1042, + "step": 7321 + }, + { + "epoch": 0.07322, + "grad_norm": 0.8144605025587333, + "learning_rate": 0.003, + "loss": 4.1241, + "step": 7322 + }, + { + "epoch": 0.07323, + "grad_norm": 0.8053334048255597, + "learning_rate": 0.003, + "loss": 4.1074, + "step": 7323 + }, + { + "epoch": 0.07324, + "grad_norm": 0.775372831647096, + "learning_rate": 0.003, + "loss": 4.104, + "step": 7324 + }, + { + "epoch": 0.07325, + "grad_norm": 0.8357997029570158, + "learning_rate": 0.003, + "loss": 4.0848, + "step": 7325 + }, + { + "epoch": 0.07326, + "grad_norm": 0.8489307022162469, + "learning_rate": 0.003, + "loss": 4.1221, + "step": 7326 + }, + { + "epoch": 0.07327, + "grad_norm": 0.8228446709367002, + "learning_rate": 0.003, + "loss": 4.1236, + "step": 7327 + }, + { + "epoch": 0.07328, + "grad_norm": 0.7838796840554372, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 7328 + }, + { + "epoch": 0.07329, + "grad_norm": 0.7251399802311324, + "learning_rate": 0.003, + "loss": 4.1018, + "step": 7329 + }, + { + "epoch": 0.0733, + "grad_norm": 0.766770759186389, + "learning_rate": 0.003, + "loss": 4.0919, + "step": 7330 + }, + { + "epoch": 0.07331, + "grad_norm": 0.943369746260282, + "learning_rate": 0.003, + "loss": 4.0916, + "step": 7331 + }, + { + "epoch": 0.07332, + "grad_norm": 1.106302603979319, + "learning_rate": 0.003, + "loss": 4.1229, + "step": 7332 + }, + { + "epoch": 0.07333, + "grad_norm": 0.9429802103066166, + "learning_rate": 0.003, + "loss": 4.1125, + "step": 7333 + }, + { + "epoch": 0.07334, + "grad_norm": 1.0320307959064938, + "learning_rate": 0.003, + "loss": 4.1108, + "step": 7334 + }, + { + "epoch": 0.07335, + "grad_norm": 0.9938214659378108, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 7335 + }, + { + "epoch": 0.07336, + "grad_norm": 1.0914354610999535, + "learning_rate": 0.003, + "loss": 4.1311, + "step": 7336 + }, + { + "epoch": 0.07337, + "grad_norm": 1.1269724240813392, + "learning_rate": 0.003, + "loss": 4.1287, + "step": 7337 + }, + { + "epoch": 0.07338, + "grad_norm": 0.8570775179663218, + "learning_rate": 0.003, + "loss": 4.1056, + "step": 7338 + }, + { + "epoch": 0.07339, + "grad_norm": 0.8031931079538867, + "learning_rate": 0.003, + "loss": 4.138, + "step": 7339 + }, + { + "epoch": 0.0734, + "grad_norm": 0.8935576371259307, + "learning_rate": 0.003, + "loss": 4.0947, + "step": 7340 + }, + { + "epoch": 0.07341, + "grad_norm": 0.9083095502174247, + "learning_rate": 0.003, + "loss": 4.1152, + "step": 7341 + }, + { + "epoch": 0.07342, + "grad_norm": 0.9576486900412259, + "learning_rate": 0.003, + "loss": 4.1373, + "step": 7342 + }, + { + "epoch": 0.07343, + "grad_norm": 0.9182147356345189, + "learning_rate": 0.003, + "loss": 4.1243, + "step": 7343 + }, + { + "epoch": 0.07344, + "grad_norm": 0.8421322703250478, + "learning_rate": 0.003, + "loss": 4.1232, + "step": 7344 + }, + { + "epoch": 0.07345, + "grad_norm": 0.7537817256778269, + "learning_rate": 0.003, + "loss": 4.0943, + "step": 7345 + }, + { + "epoch": 0.07346, + "grad_norm": 0.7033688234808623, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 7346 + }, + { + "epoch": 0.07347, + "grad_norm": 0.7413990535788959, + "learning_rate": 0.003, + "loss": 4.105, + "step": 7347 + }, + { + "epoch": 0.07348, + "grad_norm": 0.7773942391883343, + "learning_rate": 0.003, + "loss": 4.1073, + "step": 7348 + }, + { + "epoch": 0.07349, + "grad_norm": 0.9329173997319489, + "learning_rate": 0.003, + "loss": 4.1157, + "step": 7349 + }, + { + "epoch": 0.0735, + "grad_norm": 1.093903209357276, + "learning_rate": 0.003, + "loss": 4.1047, + "step": 7350 + }, + { + "epoch": 0.07351, + "grad_norm": 0.8109483676035741, + "learning_rate": 0.003, + "loss": 4.0939, + "step": 7351 + }, + { + "epoch": 0.07352, + "grad_norm": 0.7939494918683481, + "learning_rate": 0.003, + "loss": 4.114, + "step": 7352 + }, + { + "epoch": 0.07353, + "grad_norm": 0.7869344084686208, + "learning_rate": 0.003, + "loss": 4.1333, + "step": 7353 + }, + { + "epoch": 0.07354, + "grad_norm": 0.7788006156502201, + "learning_rate": 0.003, + "loss": 4.0946, + "step": 7354 + }, + { + "epoch": 0.07355, + "grad_norm": 0.8552053349081179, + "learning_rate": 0.003, + "loss": 4.1168, + "step": 7355 + }, + { + "epoch": 0.07356, + "grad_norm": 1.0518730489917372, + "learning_rate": 0.003, + "loss": 4.1226, + "step": 7356 + }, + { + "epoch": 0.07357, + "grad_norm": 1.3412824434070465, + "learning_rate": 0.003, + "loss": 4.1117, + "step": 7357 + }, + { + "epoch": 0.07358, + "grad_norm": 0.9208430362293964, + "learning_rate": 0.003, + "loss": 4.1214, + "step": 7358 + }, + { + "epoch": 0.07359, + "grad_norm": 0.8344769448824529, + "learning_rate": 0.003, + "loss": 4.1195, + "step": 7359 + }, + { + "epoch": 0.0736, + "grad_norm": 0.6793410562113039, + "learning_rate": 0.003, + "loss": 4.1272, + "step": 7360 + }, + { + "epoch": 0.07361, + "grad_norm": 0.6267991527813198, + "learning_rate": 0.003, + "loss": 4.0986, + "step": 7361 + }, + { + "epoch": 0.07362, + "grad_norm": 0.690161715844307, + "learning_rate": 0.003, + "loss": 4.1237, + "step": 7362 + }, + { + "epoch": 0.07363, + "grad_norm": 0.7869927998336663, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 7363 + }, + { + "epoch": 0.07364, + "grad_norm": 0.8027534998077103, + "learning_rate": 0.003, + "loss": 4.1247, + "step": 7364 + }, + { + "epoch": 0.07365, + "grad_norm": 0.6955286691239817, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 7365 + }, + { + "epoch": 0.07366, + "grad_norm": 0.7372804832228109, + "learning_rate": 0.003, + "loss": 4.1043, + "step": 7366 + }, + { + "epoch": 0.07367, + "grad_norm": 0.6394799490289566, + "learning_rate": 0.003, + "loss": 4.1128, + "step": 7367 + }, + { + "epoch": 0.07368, + "grad_norm": 0.509023284762404, + "learning_rate": 0.003, + "loss": 4.1263, + "step": 7368 + }, + { + "epoch": 0.07369, + "grad_norm": 0.5616436196591941, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 7369 + }, + { + "epoch": 0.0737, + "grad_norm": 0.6425475395508851, + "learning_rate": 0.003, + "loss": 4.1058, + "step": 7370 + }, + { + "epoch": 0.07371, + "grad_norm": 0.7368588731675243, + "learning_rate": 0.003, + "loss": 4.1263, + "step": 7371 + }, + { + "epoch": 0.07372, + "grad_norm": 0.9085719201340855, + "learning_rate": 0.003, + "loss": 4.1092, + "step": 7372 + }, + { + "epoch": 0.07373, + "grad_norm": 1.2408545702032314, + "learning_rate": 0.003, + "loss": 4.1187, + "step": 7373 + }, + { + "epoch": 0.07374, + "grad_norm": 0.7374902067189273, + "learning_rate": 0.003, + "loss": 4.0977, + "step": 7374 + }, + { + "epoch": 0.07375, + "grad_norm": 0.6163993980166736, + "learning_rate": 0.003, + "loss": 4.1025, + "step": 7375 + }, + { + "epoch": 0.07376, + "grad_norm": 0.6423268145367569, + "learning_rate": 0.003, + "loss": 4.1068, + "step": 7376 + }, + { + "epoch": 0.07377, + "grad_norm": 0.7533167250559268, + "learning_rate": 0.003, + "loss": 4.1055, + "step": 7377 + }, + { + "epoch": 0.07378, + "grad_norm": 0.8584299528911417, + "learning_rate": 0.003, + "loss": 4.1159, + "step": 7378 + }, + { + "epoch": 0.07379, + "grad_norm": 1.0096294225047744, + "learning_rate": 0.003, + "loss": 4.0922, + "step": 7379 + }, + { + "epoch": 0.0738, + "grad_norm": 1.0223154674326658, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 7380 + }, + { + "epoch": 0.07381, + "grad_norm": 0.8907126146182821, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 7381 + }, + { + "epoch": 0.07382, + "grad_norm": 0.7207053071921802, + "learning_rate": 0.003, + "loss": 4.1337, + "step": 7382 + }, + { + "epoch": 0.07383, + "grad_norm": 0.9145179387624196, + "learning_rate": 0.003, + "loss": 4.1137, + "step": 7383 + }, + { + "epoch": 0.07384, + "grad_norm": 1.0098905122982873, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 7384 + }, + { + "epoch": 0.07385, + "grad_norm": 1.198216983974359, + "learning_rate": 0.003, + "loss": 4.1127, + "step": 7385 + }, + { + "epoch": 0.07386, + "grad_norm": 0.7055995485988885, + "learning_rate": 0.003, + "loss": 4.1143, + "step": 7386 + }, + { + "epoch": 0.07387, + "grad_norm": 0.6341184291168993, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 7387 + }, + { + "epoch": 0.07388, + "grad_norm": 0.6305102045443429, + "learning_rate": 0.003, + "loss": 4.1095, + "step": 7388 + }, + { + "epoch": 0.07389, + "grad_norm": 0.7729018243590356, + "learning_rate": 0.003, + "loss": 4.0967, + "step": 7389 + }, + { + "epoch": 0.0739, + "grad_norm": 1.046784353921616, + "learning_rate": 0.003, + "loss": 4.093, + "step": 7390 + }, + { + "epoch": 0.07391, + "grad_norm": 1.003334204394043, + "learning_rate": 0.003, + "loss": 4.1185, + "step": 7391 + }, + { + "epoch": 0.07392, + "grad_norm": 0.867653951162738, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 7392 + }, + { + "epoch": 0.07393, + "grad_norm": 0.7970634630521443, + "learning_rate": 0.003, + "loss": 4.1089, + "step": 7393 + }, + { + "epoch": 0.07394, + "grad_norm": 0.9057360666683115, + "learning_rate": 0.003, + "loss": 4.1067, + "step": 7394 + }, + { + "epoch": 0.07395, + "grad_norm": 0.9547098856968925, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 7395 + }, + { + "epoch": 0.07396, + "grad_norm": 0.9234933056161622, + "learning_rate": 0.003, + "loss": 4.1095, + "step": 7396 + }, + { + "epoch": 0.07397, + "grad_norm": 1.0368747780491556, + "learning_rate": 0.003, + "loss": 4.1127, + "step": 7397 + }, + { + "epoch": 0.07398, + "grad_norm": 1.0403397267434613, + "learning_rate": 0.003, + "loss": 4.1108, + "step": 7398 + }, + { + "epoch": 0.07399, + "grad_norm": 0.9544666764070975, + "learning_rate": 0.003, + "loss": 4.0884, + "step": 7399 + }, + { + "epoch": 0.074, + "grad_norm": 0.8872295611618851, + "learning_rate": 0.003, + "loss": 4.1285, + "step": 7400 + }, + { + "epoch": 0.07401, + "grad_norm": 0.9340892887170521, + "learning_rate": 0.003, + "loss": 4.0935, + "step": 7401 + }, + { + "epoch": 0.07402, + "grad_norm": 0.8784569523984895, + "learning_rate": 0.003, + "loss": 4.1129, + "step": 7402 + }, + { + "epoch": 0.07403, + "grad_norm": 0.8660676350594653, + "learning_rate": 0.003, + "loss": 4.1211, + "step": 7403 + }, + { + "epoch": 0.07404, + "grad_norm": 0.9009419907324966, + "learning_rate": 0.003, + "loss": 4.1556, + "step": 7404 + }, + { + "epoch": 0.07405, + "grad_norm": 0.8106715007698777, + "learning_rate": 0.003, + "loss": 4.1102, + "step": 7405 + }, + { + "epoch": 0.07406, + "grad_norm": 0.807586011085066, + "learning_rate": 0.003, + "loss": 4.1109, + "step": 7406 + }, + { + "epoch": 0.07407, + "grad_norm": 0.8628667672682003, + "learning_rate": 0.003, + "loss": 4.1302, + "step": 7407 + }, + { + "epoch": 0.07408, + "grad_norm": 0.9415613388154968, + "learning_rate": 0.003, + "loss": 4.1262, + "step": 7408 + }, + { + "epoch": 0.07409, + "grad_norm": 0.9481965047839528, + "learning_rate": 0.003, + "loss": 4.1119, + "step": 7409 + }, + { + "epoch": 0.0741, + "grad_norm": 1.0223140305985579, + "learning_rate": 0.003, + "loss": 4.1325, + "step": 7410 + }, + { + "epoch": 0.07411, + "grad_norm": 1.1261250470212472, + "learning_rate": 0.003, + "loss": 4.1286, + "step": 7411 + }, + { + "epoch": 0.07412, + "grad_norm": 0.8452821990703466, + "learning_rate": 0.003, + "loss": 4.1093, + "step": 7412 + }, + { + "epoch": 0.07413, + "grad_norm": 0.7070647319303838, + "learning_rate": 0.003, + "loss": 4.106, + "step": 7413 + }, + { + "epoch": 0.07414, + "grad_norm": 0.6602409738515731, + "learning_rate": 0.003, + "loss": 4.1058, + "step": 7414 + }, + { + "epoch": 0.07415, + "grad_norm": 0.6813711085208377, + "learning_rate": 0.003, + "loss": 4.1048, + "step": 7415 + }, + { + "epoch": 0.07416, + "grad_norm": 0.7813248550198105, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 7416 + }, + { + "epoch": 0.07417, + "grad_norm": 0.7903284361595316, + "learning_rate": 0.003, + "loss": 4.111, + "step": 7417 + }, + { + "epoch": 0.07418, + "grad_norm": 0.8776997591398797, + "learning_rate": 0.003, + "loss": 4.0908, + "step": 7418 + }, + { + "epoch": 0.07419, + "grad_norm": 0.9346374935990277, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 7419 + }, + { + "epoch": 0.0742, + "grad_norm": 0.8167496983343525, + "learning_rate": 0.003, + "loss": 4.1193, + "step": 7420 + }, + { + "epoch": 0.07421, + "grad_norm": 0.7740749507997752, + "learning_rate": 0.003, + "loss": 4.0893, + "step": 7421 + }, + { + "epoch": 0.07422, + "grad_norm": 0.7797923880019133, + "learning_rate": 0.003, + "loss": 4.111, + "step": 7422 + }, + { + "epoch": 0.07423, + "grad_norm": 0.7246765372540556, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 7423 + }, + { + "epoch": 0.07424, + "grad_norm": 0.7739364594038591, + "learning_rate": 0.003, + "loss": 4.1145, + "step": 7424 + }, + { + "epoch": 0.07425, + "grad_norm": 0.8561033384244009, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 7425 + }, + { + "epoch": 0.07426, + "grad_norm": 1.0873991072382838, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 7426 + }, + { + "epoch": 0.07427, + "grad_norm": 1.0969759409269335, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 7427 + }, + { + "epoch": 0.07428, + "grad_norm": 0.7321830844983641, + "learning_rate": 0.003, + "loss": 4.1327, + "step": 7428 + }, + { + "epoch": 0.07429, + "grad_norm": 0.6644474298927815, + "learning_rate": 0.003, + "loss": 4.1055, + "step": 7429 + }, + { + "epoch": 0.0743, + "grad_norm": 0.6883336057236158, + "learning_rate": 0.003, + "loss": 4.0949, + "step": 7430 + }, + { + "epoch": 0.07431, + "grad_norm": 0.7448399088605852, + "learning_rate": 0.003, + "loss": 4.0903, + "step": 7431 + }, + { + "epoch": 0.07432, + "grad_norm": 0.8377015855878724, + "learning_rate": 0.003, + "loss": 4.1129, + "step": 7432 + }, + { + "epoch": 0.07433, + "grad_norm": 0.8455466117001035, + "learning_rate": 0.003, + "loss": 4.0954, + "step": 7433 + }, + { + "epoch": 0.07434, + "grad_norm": 0.8038998713277352, + "learning_rate": 0.003, + "loss": 4.069, + "step": 7434 + }, + { + "epoch": 0.07435, + "grad_norm": 0.7628824415481451, + "learning_rate": 0.003, + "loss": 4.1063, + "step": 7435 + }, + { + "epoch": 0.07436, + "grad_norm": 0.8471666681896792, + "learning_rate": 0.003, + "loss": 4.1293, + "step": 7436 + }, + { + "epoch": 0.07437, + "grad_norm": 1.022193249171131, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 7437 + }, + { + "epoch": 0.07438, + "grad_norm": 0.954279749528698, + "learning_rate": 0.003, + "loss": 4.1191, + "step": 7438 + }, + { + "epoch": 0.07439, + "grad_norm": 0.9359744466674262, + "learning_rate": 0.003, + "loss": 4.1056, + "step": 7439 + }, + { + "epoch": 0.0744, + "grad_norm": 0.985750832333834, + "learning_rate": 0.003, + "loss": 4.1168, + "step": 7440 + }, + { + "epoch": 0.07441, + "grad_norm": 1.0387568082669436, + "learning_rate": 0.003, + "loss": 4.1552, + "step": 7441 + }, + { + "epoch": 0.07442, + "grad_norm": 0.9354486578980633, + "learning_rate": 0.003, + "loss": 4.1282, + "step": 7442 + }, + { + "epoch": 0.07443, + "grad_norm": 0.8937562700947757, + "learning_rate": 0.003, + "loss": 4.1114, + "step": 7443 + }, + { + "epoch": 0.07444, + "grad_norm": 0.7862746074501281, + "learning_rate": 0.003, + "loss": 4.1268, + "step": 7444 + }, + { + "epoch": 0.07445, + "grad_norm": 0.7580214165836215, + "learning_rate": 0.003, + "loss": 4.1077, + "step": 7445 + }, + { + "epoch": 0.07446, + "grad_norm": 0.826490781957247, + "learning_rate": 0.003, + "loss": 4.0912, + "step": 7446 + }, + { + "epoch": 0.07447, + "grad_norm": 0.78561092743045, + "learning_rate": 0.003, + "loss": 4.0969, + "step": 7447 + }, + { + "epoch": 0.07448, + "grad_norm": 0.8720738755090752, + "learning_rate": 0.003, + "loss": 4.0934, + "step": 7448 + }, + { + "epoch": 0.07449, + "grad_norm": 1.1356579541218612, + "learning_rate": 0.003, + "loss": 4.1163, + "step": 7449 + }, + { + "epoch": 0.0745, + "grad_norm": 1.194362940520485, + "learning_rate": 0.003, + "loss": 4.1263, + "step": 7450 + }, + { + "epoch": 0.07451, + "grad_norm": 0.8425811609172174, + "learning_rate": 0.003, + "loss": 4.115, + "step": 7451 + }, + { + "epoch": 0.07452, + "grad_norm": 0.9146392058972959, + "learning_rate": 0.003, + "loss": 4.0891, + "step": 7452 + }, + { + "epoch": 0.07453, + "grad_norm": 1.0026955605148151, + "learning_rate": 0.003, + "loss": 4.1274, + "step": 7453 + }, + { + "epoch": 0.07454, + "grad_norm": 1.0112038508522885, + "learning_rate": 0.003, + "loss": 4.1031, + "step": 7454 + }, + { + "epoch": 0.07455, + "grad_norm": 1.168191155504265, + "learning_rate": 0.003, + "loss": 4.1204, + "step": 7455 + }, + { + "epoch": 0.07456, + "grad_norm": 0.7449918001899162, + "learning_rate": 0.003, + "loss": 4.1173, + "step": 7456 + }, + { + "epoch": 0.07457, + "grad_norm": 0.7986467621349783, + "learning_rate": 0.003, + "loss": 4.1105, + "step": 7457 + }, + { + "epoch": 0.07458, + "grad_norm": 0.6606779110991354, + "learning_rate": 0.003, + "loss": 4.1158, + "step": 7458 + }, + { + "epoch": 0.07459, + "grad_norm": 0.74600487207374, + "learning_rate": 0.003, + "loss": 4.0996, + "step": 7459 + }, + { + "epoch": 0.0746, + "grad_norm": 0.7705606291540048, + "learning_rate": 0.003, + "loss": 4.1096, + "step": 7460 + }, + { + "epoch": 0.07461, + "grad_norm": 0.8369060675928409, + "learning_rate": 0.003, + "loss": 4.1182, + "step": 7461 + }, + { + "epoch": 0.07462, + "grad_norm": 0.9604504411667463, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 7462 + }, + { + "epoch": 0.07463, + "grad_norm": 1.018164117686845, + "learning_rate": 0.003, + "loss": 4.1448, + "step": 7463 + }, + { + "epoch": 0.07464, + "grad_norm": 0.8462415172519894, + "learning_rate": 0.003, + "loss": 4.0874, + "step": 7464 + }, + { + "epoch": 0.07465, + "grad_norm": 0.7280333713094582, + "learning_rate": 0.003, + "loss": 4.1067, + "step": 7465 + }, + { + "epoch": 0.07466, + "grad_norm": 0.6527117875066514, + "learning_rate": 0.003, + "loss": 4.1199, + "step": 7466 + }, + { + "epoch": 0.07467, + "grad_norm": 0.7283692800044493, + "learning_rate": 0.003, + "loss": 4.0912, + "step": 7467 + }, + { + "epoch": 0.07468, + "grad_norm": 0.7286232701262282, + "learning_rate": 0.003, + "loss": 4.1121, + "step": 7468 + }, + { + "epoch": 0.07469, + "grad_norm": 0.7554001099044895, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 7469 + }, + { + "epoch": 0.0747, + "grad_norm": 0.6494769448278457, + "learning_rate": 0.003, + "loss": 4.1238, + "step": 7470 + }, + { + "epoch": 0.07471, + "grad_norm": 0.582239848281011, + "learning_rate": 0.003, + "loss": 4.1034, + "step": 7471 + }, + { + "epoch": 0.07472, + "grad_norm": 0.6054171618519452, + "learning_rate": 0.003, + "loss": 4.1059, + "step": 7472 + }, + { + "epoch": 0.07473, + "grad_norm": 0.5800597140442034, + "learning_rate": 0.003, + "loss": 4.1072, + "step": 7473 + }, + { + "epoch": 0.07474, + "grad_norm": 0.6418775672195469, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 7474 + }, + { + "epoch": 0.07475, + "grad_norm": 0.8209219098567517, + "learning_rate": 0.003, + "loss": 4.1065, + "step": 7475 + }, + { + "epoch": 0.07476, + "grad_norm": 1.0927372326738478, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 7476 + }, + { + "epoch": 0.07477, + "grad_norm": 1.1813159129321458, + "learning_rate": 0.003, + "loss": 4.1158, + "step": 7477 + }, + { + "epoch": 0.07478, + "grad_norm": 0.7297682403726699, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 7478 + }, + { + "epoch": 0.07479, + "grad_norm": 0.6410942452556853, + "learning_rate": 0.003, + "loss": 4.074, + "step": 7479 + }, + { + "epoch": 0.0748, + "grad_norm": 0.8698329147222141, + "learning_rate": 0.003, + "loss": 4.0961, + "step": 7480 + }, + { + "epoch": 0.07481, + "grad_norm": 1.0082944565315157, + "learning_rate": 0.003, + "loss": 4.1108, + "step": 7481 + }, + { + "epoch": 0.07482, + "grad_norm": 0.9628391840892444, + "learning_rate": 0.003, + "loss": 4.1099, + "step": 7482 + }, + { + "epoch": 0.07483, + "grad_norm": 0.8641925302412802, + "learning_rate": 0.003, + "loss": 4.1281, + "step": 7483 + }, + { + "epoch": 0.07484, + "grad_norm": 0.7533282082846949, + "learning_rate": 0.003, + "loss": 4.1103, + "step": 7484 + }, + { + "epoch": 0.07485, + "grad_norm": 0.7703693706278185, + "learning_rate": 0.003, + "loss": 4.1, + "step": 7485 + }, + { + "epoch": 0.07486, + "grad_norm": 0.8291299861969627, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 7486 + }, + { + "epoch": 0.07487, + "grad_norm": 0.9151584554564175, + "learning_rate": 0.003, + "loss": 4.107, + "step": 7487 + }, + { + "epoch": 0.07488, + "grad_norm": 1.0172079023902525, + "learning_rate": 0.003, + "loss": 4.1177, + "step": 7488 + }, + { + "epoch": 0.07489, + "grad_norm": 0.985794974411023, + "learning_rate": 0.003, + "loss": 4.1284, + "step": 7489 + }, + { + "epoch": 0.0749, + "grad_norm": 0.7878742848581212, + "learning_rate": 0.003, + "loss": 4.093, + "step": 7490 + }, + { + "epoch": 0.07491, + "grad_norm": 0.7562126231512927, + "learning_rate": 0.003, + "loss": 4.0998, + "step": 7491 + }, + { + "epoch": 0.07492, + "grad_norm": 0.7316445879371943, + "learning_rate": 0.003, + "loss": 4.1174, + "step": 7492 + }, + { + "epoch": 0.07493, + "grad_norm": 0.7015263206527975, + "learning_rate": 0.003, + "loss": 4.1081, + "step": 7493 + }, + { + "epoch": 0.07494, + "grad_norm": 0.7094674720974151, + "learning_rate": 0.003, + "loss": 4.1219, + "step": 7494 + }, + { + "epoch": 0.07495, + "grad_norm": 0.8005067312090948, + "learning_rate": 0.003, + "loss": 4.0987, + "step": 7495 + }, + { + "epoch": 0.07496, + "grad_norm": 1.0221046723780693, + "learning_rate": 0.003, + "loss": 4.0976, + "step": 7496 + }, + { + "epoch": 0.07497, + "grad_norm": 1.1782769790208016, + "learning_rate": 0.003, + "loss": 4.1157, + "step": 7497 + }, + { + "epoch": 0.07498, + "grad_norm": 0.8031446202765408, + "learning_rate": 0.003, + "loss": 4.0992, + "step": 7498 + }, + { + "epoch": 0.07499, + "grad_norm": 0.7861635761156994, + "learning_rate": 0.003, + "loss": 4.103, + "step": 7499 + }, + { + "epoch": 0.075, + "grad_norm": 0.7398205403713403, + "learning_rate": 0.003, + "loss": 4.1212, + "step": 7500 + }, + { + "epoch": 0.07501, + "grad_norm": 0.8734378274009751, + "learning_rate": 0.003, + "loss": 4.0961, + "step": 7501 + }, + { + "epoch": 0.07502, + "grad_norm": 0.8830879959651403, + "learning_rate": 0.003, + "loss": 4.095, + "step": 7502 + }, + { + "epoch": 0.07503, + "grad_norm": 0.8281346250649004, + "learning_rate": 0.003, + "loss": 4.0943, + "step": 7503 + }, + { + "epoch": 0.07504, + "grad_norm": 0.8072671794855871, + "learning_rate": 0.003, + "loss": 4.0967, + "step": 7504 + }, + { + "epoch": 0.07505, + "grad_norm": 0.8666820809776766, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 7505 + }, + { + "epoch": 0.07506, + "grad_norm": 1.0055552931832379, + "learning_rate": 0.003, + "loss": 4.101, + "step": 7506 + }, + { + "epoch": 0.07507, + "grad_norm": 0.9690975893829946, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 7507 + }, + { + "epoch": 0.07508, + "grad_norm": 0.9561250009010052, + "learning_rate": 0.003, + "loss": 4.1303, + "step": 7508 + }, + { + "epoch": 0.07509, + "grad_norm": 1.0877559050066379, + "learning_rate": 0.003, + "loss": 4.1123, + "step": 7509 + }, + { + "epoch": 0.0751, + "grad_norm": 1.0027154307652972, + "learning_rate": 0.003, + "loss": 4.1077, + "step": 7510 + }, + { + "epoch": 0.07511, + "grad_norm": 1.0165739974701828, + "learning_rate": 0.003, + "loss": 4.1, + "step": 7511 + }, + { + "epoch": 0.07512, + "grad_norm": 1.0057419295722907, + "learning_rate": 0.003, + "loss": 4.1394, + "step": 7512 + }, + { + "epoch": 0.07513, + "grad_norm": 0.9433174355296526, + "learning_rate": 0.003, + "loss": 4.12, + "step": 7513 + }, + { + "epoch": 0.07514, + "grad_norm": 1.0313875668855543, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 7514 + }, + { + "epoch": 0.07515, + "grad_norm": 1.0147222353644576, + "learning_rate": 0.003, + "loss": 4.1449, + "step": 7515 + }, + { + "epoch": 0.07516, + "grad_norm": 0.9245083215799488, + "learning_rate": 0.003, + "loss": 4.0957, + "step": 7516 + }, + { + "epoch": 0.07517, + "grad_norm": 0.8259475443651306, + "learning_rate": 0.003, + "loss": 4.1198, + "step": 7517 + }, + { + "epoch": 0.07518, + "grad_norm": 0.7587280215699812, + "learning_rate": 0.003, + "loss": 4.0948, + "step": 7518 + }, + { + "epoch": 0.07519, + "grad_norm": 0.8033857629693775, + "learning_rate": 0.003, + "loss": 4.1064, + "step": 7519 + }, + { + "epoch": 0.0752, + "grad_norm": 0.6891155253835879, + "learning_rate": 0.003, + "loss": 4.1064, + "step": 7520 + }, + { + "epoch": 0.07521, + "grad_norm": 0.6668629114515775, + "learning_rate": 0.003, + "loss": 4.1085, + "step": 7521 + }, + { + "epoch": 0.07522, + "grad_norm": 0.6646628095819561, + "learning_rate": 0.003, + "loss": 4.0908, + "step": 7522 + }, + { + "epoch": 0.07523, + "grad_norm": 0.6810426992617306, + "learning_rate": 0.003, + "loss": 4.0872, + "step": 7523 + }, + { + "epoch": 0.07524, + "grad_norm": 0.5957519856201996, + "learning_rate": 0.003, + "loss": 4.112, + "step": 7524 + }, + { + "epoch": 0.07525, + "grad_norm": 0.6304965471697975, + "learning_rate": 0.003, + "loss": 4.122, + "step": 7525 + }, + { + "epoch": 0.07526, + "grad_norm": 0.7044075744869454, + "learning_rate": 0.003, + "loss": 4.1186, + "step": 7526 + }, + { + "epoch": 0.07527, + "grad_norm": 0.8184433815359451, + "learning_rate": 0.003, + "loss": 4.1181, + "step": 7527 + }, + { + "epoch": 0.07528, + "grad_norm": 0.8298479343150946, + "learning_rate": 0.003, + "loss": 4.1225, + "step": 7528 + }, + { + "epoch": 0.07529, + "grad_norm": 0.7555105418948039, + "learning_rate": 0.003, + "loss": 4.0916, + "step": 7529 + }, + { + "epoch": 0.0753, + "grad_norm": 0.62755290323034, + "learning_rate": 0.003, + "loss": 4.0864, + "step": 7530 + }, + { + "epoch": 0.07531, + "grad_norm": 0.6259328988343698, + "learning_rate": 0.003, + "loss": 4.1175, + "step": 7531 + }, + { + "epoch": 0.07532, + "grad_norm": 0.7024264501793085, + "learning_rate": 0.003, + "loss": 4.1066, + "step": 7532 + }, + { + "epoch": 0.07533, + "grad_norm": 0.7976246816821072, + "learning_rate": 0.003, + "loss": 4.092, + "step": 7533 + }, + { + "epoch": 0.07534, + "grad_norm": 0.9491149711829591, + "learning_rate": 0.003, + "loss": 4.1022, + "step": 7534 + }, + { + "epoch": 0.07535, + "grad_norm": 0.9702246438073734, + "learning_rate": 0.003, + "loss": 4.1152, + "step": 7535 + }, + { + "epoch": 0.07536, + "grad_norm": 0.8833175400088759, + "learning_rate": 0.003, + "loss": 4.0986, + "step": 7536 + }, + { + "epoch": 0.07537, + "grad_norm": 1.0566640021173932, + "learning_rate": 0.003, + "loss": 4.0928, + "step": 7537 + }, + { + "epoch": 0.07538, + "grad_norm": 1.0524050892778398, + "learning_rate": 0.003, + "loss": 4.1158, + "step": 7538 + }, + { + "epoch": 0.07539, + "grad_norm": 0.9461058512204453, + "learning_rate": 0.003, + "loss": 4.087, + "step": 7539 + }, + { + "epoch": 0.0754, + "grad_norm": 0.8671896332276945, + "learning_rate": 0.003, + "loss": 4.1062, + "step": 7540 + }, + { + "epoch": 0.07541, + "grad_norm": 0.9370741051157787, + "learning_rate": 0.003, + "loss": 4.0907, + "step": 7541 + }, + { + "epoch": 0.07542, + "grad_norm": 0.7655033736787794, + "learning_rate": 0.003, + "loss": 4.1052, + "step": 7542 + }, + { + "epoch": 0.07543, + "grad_norm": 0.9638854244860205, + "learning_rate": 0.003, + "loss": 4.1295, + "step": 7543 + }, + { + "epoch": 0.07544, + "grad_norm": 1.151245583846751, + "learning_rate": 0.003, + "loss": 4.1281, + "step": 7544 + }, + { + "epoch": 0.07545, + "grad_norm": 1.045720500701298, + "learning_rate": 0.003, + "loss": 4.1496, + "step": 7545 + }, + { + "epoch": 0.07546, + "grad_norm": 0.8508437267643452, + "learning_rate": 0.003, + "loss": 4.1, + "step": 7546 + }, + { + "epoch": 0.07547, + "grad_norm": 0.8758265191573648, + "learning_rate": 0.003, + "loss": 4.1213, + "step": 7547 + }, + { + "epoch": 0.07548, + "grad_norm": 1.0308655234346675, + "learning_rate": 0.003, + "loss": 4.1395, + "step": 7548 + }, + { + "epoch": 0.07549, + "grad_norm": 1.1222648672525313, + "learning_rate": 0.003, + "loss": 4.1407, + "step": 7549 + }, + { + "epoch": 0.0755, + "grad_norm": 1.0347446053481122, + "learning_rate": 0.003, + "loss": 4.1338, + "step": 7550 + }, + { + "epoch": 0.07551, + "grad_norm": 0.8951008390390673, + "learning_rate": 0.003, + "loss": 4.1237, + "step": 7551 + }, + { + "epoch": 0.07552, + "grad_norm": 0.867925078672574, + "learning_rate": 0.003, + "loss": 4.1268, + "step": 7552 + }, + { + "epoch": 0.07553, + "grad_norm": 0.8525722259854441, + "learning_rate": 0.003, + "loss": 4.1311, + "step": 7553 + }, + { + "epoch": 0.07554, + "grad_norm": 0.9114938094018381, + "learning_rate": 0.003, + "loss": 4.1279, + "step": 7554 + }, + { + "epoch": 0.07555, + "grad_norm": 0.9344191657970983, + "learning_rate": 0.003, + "loss": 4.1158, + "step": 7555 + }, + { + "epoch": 0.07556, + "grad_norm": 1.0145169044440256, + "learning_rate": 0.003, + "loss": 4.1376, + "step": 7556 + }, + { + "epoch": 0.07557, + "grad_norm": 1.008268955978688, + "learning_rate": 0.003, + "loss": 4.1073, + "step": 7557 + }, + { + "epoch": 0.07558, + "grad_norm": 0.9718709803852565, + "learning_rate": 0.003, + "loss": 4.1114, + "step": 7558 + }, + { + "epoch": 0.07559, + "grad_norm": 1.056325136300485, + "learning_rate": 0.003, + "loss": 4.082, + "step": 7559 + }, + { + "epoch": 0.0756, + "grad_norm": 0.9740547207080791, + "learning_rate": 0.003, + "loss": 4.1236, + "step": 7560 + }, + { + "epoch": 0.07561, + "grad_norm": 0.9083193763007056, + "learning_rate": 0.003, + "loss": 4.1057, + "step": 7561 + }, + { + "epoch": 0.07562, + "grad_norm": 0.8448439316271258, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 7562 + }, + { + "epoch": 0.07563, + "grad_norm": 0.9065754257807024, + "learning_rate": 0.003, + "loss": 4.1157, + "step": 7563 + }, + { + "epoch": 0.07564, + "grad_norm": 0.8560571418992496, + "learning_rate": 0.003, + "loss": 4.1322, + "step": 7564 + }, + { + "epoch": 0.07565, + "grad_norm": 0.8274853334246817, + "learning_rate": 0.003, + "loss": 4.1122, + "step": 7565 + }, + { + "epoch": 0.07566, + "grad_norm": 0.8273427382041143, + "learning_rate": 0.003, + "loss": 4.1077, + "step": 7566 + }, + { + "epoch": 0.07567, + "grad_norm": 0.8118867625373324, + "learning_rate": 0.003, + "loss": 4.1221, + "step": 7567 + }, + { + "epoch": 0.07568, + "grad_norm": 0.6681343989273923, + "learning_rate": 0.003, + "loss": 4.1111, + "step": 7568 + }, + { + "epoch": 0.07569, + "grad_norm": 0.5803814722946369, + "learning_rate": 0.003, + "loss": 4.0888, + "step": 7569 + }, + { + "epoch": 0.0757, + "grad_norm": 0.5510764495582413, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 7570 + }, + { + "epoch": 0.07571, + "grad_norm": 0.6627010068288498, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 7571 + }, + { + "epoch": 0.07572, + "grad_norm": 0.6455742336359287, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 7572 + }, + { + "epoch": 0.07573, + "grad_norm": 0.7410211759586838, + "learning_rate": 0.003, + "loss": 4.0982, + "step": 7573 + }, + { + "epoch": 0.07574, + "grad_norm": 0.7937533836416206, + "learning_rate": 0.003, + "loss": 4.1165, + "step": 7574 + }, + { + "epoch": 0.07575, + "grad_norm": 0.7151578069937585, + "learning_rate": 0.003, + "loss": 4.0948, + "step": 7575 + }, + { + "epoch": 0.07576, + "grad_norm": 0.6757485878022385, + "learning_rate": 0.003, + "loss": 4.0884, + "step": 7576 + }, + { + "epoch": 0.07577, + "grad_norm": 0.7377828856053927, + "learning_rate": 0.003, + "loss": 4.122, + "step": 7577 + }, + { + "epoch": 0.07578, + "grad_norm": 0.9295354995808791, + "learning_rate": 0.003, + "loss": 4.1085, + "step": 7578 + }, + { + "epoch": 0.07579, + "grad_norm": 1.1308441721629587, + "learning_rate": 0.003, + "loss": 4.1034, + "step": 7579 + }, + { + "epoch": 0.0758, + "grad_norm": 0.9537849948969392, + "learning_rate": 0.003, + "loss": 4.1163, + "step": 7580 + }, + { + "epoch": 0.07581, + "grad_norm": 0.9648991721783183, + "learning_rate": 0.003, + "loss": 4.1097, + "step": 7581 + }, + { + "epoch": 0.07582, + "grad_norm": 0.926845739832675, + "learning_rate": 0.003, + "loss": 4.1037, + "step": 7582 + }, + { + "epoch": 0.07583, + "grad_norm": 0.894884269204141, + "learning_rate": 0.003, + "loss": 4.1021, + "step": 7583 + }, + { + "epoch": 0.07584, + "grad_norm": 0.8924573565922338, + "learning_rate": 0.003, + "loss": 4.1477, + "step": 7584 + }, + { + "epoch": 0.07585, + "grad_norm": 0.9888084614023944, + "learning_rate": 0.003, + "loss": 4.1151, + "step": 7585 + }, + { + "epoch": 0.07586, + "grad_norm": 1.1114233056422533, + "learning_rate": 0.003, + "loss": 4.1001, + "step": 7586 + }, + { + "epoch": 0.07587, + "grad_norm": 0.8144693201284348, + "learning_rate": 0.003, + "loss": 4.1327, + "step": 7587 + }, + { + "epoch": 0.07588, + "grad_norm": 0.8305575966505442, + "learning_rate": 0.003, + "loss": 4.1078, + "step": 7588 + }, + { + "epoch": 0.07589, + "grad_norm": 0.8498435104769524, + "learning_rate": 0.003, + "loss": 4.1152, + "step": 7589 + }, + { + "epoch": 0.0759, + "grad_norm": 0.8188810092794875, + "learning_rate": 0.003, + "loss": 4.1095, + "step": 7590 + }, + { + "epoch": 0.07591, + "grad_norm": 0.7698614390712419, + "learning_rate": 0.003, + "loss": 4.1173, + "step": 7591 + }, + { + "epoch": 0.07592, + "grad_norm": 0.5909542261154095, + "learning_rate": 0.003, + "loss": 4.1137, + "step": 7592 + }, + { + "epoch": 0.07593, + "grad_norm": 0.6169915126769421, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 7593 + }, + { + "epoch": 0.07594, + "grad_norm": 0.6183226824646945, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 7594 + }, + { + "epoch": 0.07595, + "grad_norm": 0.6295258139827735, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 7595 + }, + { + "epoch": 0.07596, + "grad_norm": 0.5647538604253549, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 7596 + }, + { + "epoch": 0.07597, + "grad_norm": 0.5488719464323173, + "learning_rate": 0.003, + "loss": 4.1142, + "step": 7597 + }, + { + "epoch": 0.07598, + "grad_norm": 0.5190318492649775, + "learning_rate": 0.003, + "loss": 4.1075, + "step": 7598 + }, + { + "epoch": 0.07599, + "grad_norm": 0.5944308410673468, + "learning_rate": 0.003, + "loss": 4.1082, + "step": 7599 + }, + { + "epoch": 0.076, + "grad_norm": 0.6645839697886495, + "learning_rate": 0.003, + "loss": 4.0989, + "step": 7600 + }, + { + "epoch": 0.07601, + "grad_norm": 0.9191079894373272, + "learning_rate": 0.003, + "loss": 4.1219, + "step": 7601 + }, + { + "epoch": 0.07602, + "grad_norm": 1.2588970299218996, + "learning_rate": 0.003, + "loss": 4.1049, + "step": 7602 + }, + { + "epoch": 0.07603, + "grad_norm": 0.9641328600344435, + "learning_rate": 0.003, + "loss": 4.1006, + "step": 7603 + }, + { + "epoch": 0.07604, + "grad_norm": 1.0002475506914794, + "learning_rate": 0.003, + "loss": 4.1197, + "step": 7604 + }, + { + "epoch": 0.07605, + "grad_norm": 0.9041839404022576, + "learning_rate": 0.003, + "loss": 4.11, + "step": 7605 + }, + { + "epoch": 0.07606, + "grad_norm": 0.8495943134663567, + "learning_rate": 0.003, + "loss": 4.1653, + "step": 7606 + }, + { + "epoch": 0.07607, + "grad_norm": 0.800744853771563, + "learning_rate": 0.003, + "loss": 4.1199, + "step": 7607 + }, + { + "epoch": 0.07608, + "grad_norm": 0.7315846867215643, + "learning_rate": 0.003, + "loss": 4.13, + "step": 7608 + }, + { + "epoch": 0.07609, + "grad_norm": 0.6443317008016733, + "learning_rate": 0.003, + "loss": 4.1027, + "step": 7609 + }, + { + "epoch": 0.0761, + "grad_norm": 0.6907982352452112, + "learning_rate": 0.003, + "loss": 4.098, + "step": 7610 + }, + { + "epoch": 0.07611, + "grad_norm": 0.857723111168813, + "learning_rate": 0.003, + "loss": 4.1294, + "step": 7611 + }, + { + "epoch": 0.07612, + "grad_norm": 1.0992578425673403, + "learning_rate": 0.003, + "loss": 4.1009, + "step": 7612 + }, + { + "epoch": 0.07613, + "grad_norm": 0.9526866939763878, + "learning_rate": 0.003, + "loss": 4.0971, + "step": 7613 + }, + { + "epoch": 0.07614, + "grad_norm": 0.7803177657645249, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 7614 + }, + { + "epoch": 0.07615, + "grad_norm": 0.7081390563241489, + "learning_rate": 0.003, + "loss": 4.1014, + "step": 7615 + }, + { + "epoch": 0.07616, + "grad_norm": 0.7252993228592656, + "learning_rate": 0.003, + "loss": 4.1019, + "step": 7616 + }, + { + "epoch": 0.07617, + "grad_norm": 0.7475989787312242, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 7617 + }, + { + "epoch": 0.07618, + "grad_norm": 0.7573711078639473, + "learning_rate": 0.003, + "loss": 4.1336, + "step": 7618 + }, + { + "epoch": 0.07619, + "grad_norm": 1.0541262061063688, + "learning_rate": 0.003, + "loss": 4.1161, + "step": 7619 + }, + { + "epoch": 0.0762, + "grad_norm": 1.1895056232233716, + "learning_rate": 0.003, + "loss": 4.0943, + "step": 7620 + }, + { + "epoch": 0.07621, + "grad_norm": 0.8742054868789058, + "learning_rate": 0.003, + "loss": 4.1209, + "step": 7621 + }, + { + "epoch": 0.07622, + "grad_norm": 0.8444291423398331, + "learning_rate": 0.003, + "loss": 4.1195, + "step": 7622 + }, + { + "epoch": 0.07623, + "grad_norm": 0.8922019849127748, + "learning_rate": 0.003, + "loss": 4.1048, + "step": 7623 + }, + { + "epoch": 0.07624, + "grad_norm": 0.803083822831012, + "learning_rate": 0.003, + "loss": 4.1052, + "step": 7624 + }, + { + "epoch": 0.07625, + "grad_norm": 0.7385150275748038, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 7625 + }, + { + "epoch": 0.07626, + "grad_norm": 0.6711142915115071, + "learning_rate": 0.003, + "loss": 4.1038, + "step": 7626 + }, + { + "epoch": 0.07627, + "grad_norm": 0.812731441964542, + "learning_rate": 0.003, + "loss": 4.1105, + "step": 7627 + }, + { + "epoch": 0.07628, + "grad_norm": 0.8485441320364617, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 7628 + }, + { + "epoch": 0.07629, + "grad_norm": 0.8442461477013375, + "learning_rate": 0.003, + "loss": 4.1021, + "step": 7629 + }, + { + "epoch": 0.0763, + "grad_norm": 0.8784878480520506, + "learning_rate": 0.003, + "loss": 4.0907, + "step": 7630 + }, + { + "epoch": 0.07631, + "grad_norm": 0.8831785338162262, + "learning_rate": 0.003, + "loss": 4.094, + "step": 7631 + }, + { + "epoch": 0.07632, + "grad_norm": 0.8267222772195404, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 7632 + }, + { + "epoch": 0.07633, + "grad_norm": 0.8474903779702966, + "learning_rate": 0.003, + "loss": 4.0926, + "step": 7633 + }, + { + "epoch": 0.07634, + "grad_norm": 0.8527662645558675, + "learning_rate": 0.003, + "loss": 4.0845, + "step": 7634 + }, + { + "epoch": 0.07635, + "grad_norm": 0.7539827705409249, + "learning_rate": 0.003, + "loss": 4.1066, + "step": 7635 + }, + { + "epoch": 0.07636, + "grad_norm": 0.7979423225524935, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 7636 + }, + { + "epoch": 0.07637, + "grad_norm": 0.9896367410513814, + "learning_rate": 0.003, + "loss": 4.1276, + "step": 7637 + }, + { + "epoch": 0.07638, + "grad_norm": 1.23693930151839, + "learning_rate": 0.003, + "loss": 4.1123, + "step": 7638 + }, + { + "epoch": 0.07639, + "grad_norm": 0.8951052515565378, + "learning_rate": 0.003, + "loss": 4.1123, + "step": 7639 + }, + { + "epoch": 0.0764, + "grad_norm": 0.9661295603094655, + "learning_rate": 0.003, + "loss": 4.1432, + "step": 7640 + }, + { + "epoch": 0.07641, + "grad_norm": 1.0381140300663276, + "learning_rate": 0.003, + "loss": 4.1436, + "step": 7641 + }, + { + "epoch": 0.07642, + "grad_norm": 0.8657178028651203, + "learning_rate": 0.003, + "loss": 4.132, + "step": 7642 + }, + { + "epoch": 0.07643, + "grad_norm": 0.7585312382956262, + "learning_rate": 0.003, + "loss": 4.1041, + "step": 7643 + }, + { + "epoch": 0.07644, + "grad_norm": 0.8424021918604633, + "learning_rate": 0.003, + "loss": 4.0905, + "step": 7644 + }, + { + "epoch": 0.07645, + "grad_norm": 1.0547127355883397, + "learning_rate": 0.003, + "loss": 4.114, + "step": 7645 + }, + { + "epoch": 0.07646, + "grad_norm": 1.0501940720210954, + "learning_rate": 0.003, + "loss": 4.1182, + "step": 7646 + }, + { + "epoch": 0.07647, + "grad_norm": 1.0089732479233036, + "learning_rate": 0.003, + "loss": 4.0995, + "step": 7647 + }, + { + "epoch": 0.07648, + "grad_norm": 0.9700747028421894, + "learning_rate": 0.003, + "loss": 4.1172, + "step": 7648 + }, + { + "epoch": 0.07649, + "grad_norm": 0.8831977035938575, + "learning_rate": 0.003, + "loss": 4.1133, + "step": 7649 + }, + { + "epoch": 0.0765, + "grad_norm": 0.7985816216412632, + "learning_rate": 0.003, + "loss": 4.0975, + "step": 7650 + }, + { + "epoch": 0.07651, + "grad_norm": 0.7812576551419401, + "learning_rate": 0.003, + "loss": 4.1046, + "step": 7651 + }, + { + "epoch": 0.07652, + "grad_norm": 0.863619677787746, + "learning_rate": 0.003, + "loss": 4.1032, + "step": 7652 + }, + { + "epoch": 0.07653, + "grad_norm": 1.033937451094598, + "learning_rate": 0.003, + "loss": 4.088, + "step": 7653 + }, + { + "epoch": 0.07654, + "grad_norm": 1.0055330888443008, + "learning_rate": 0.003, + "loss": 4.1136, + "step": 7654 + }, + { + "epoch": 0.07655, + "grad_norm": 0.961819876357498, + "learning_rate": 0.003, + "loss": 4.1176, + "step": 7655 + }, + { + "epoch": 0.07656, + "grad_norm": 0.8931684820162469, + "learning_rate": 0.003, + "loss": 4.1241, + "step": 7656 + }, + { + "epoch": 0.07657, + "grad_norm": 0.8072740880367048, + "learning_rate": 0.003, + "loss": 4.1257, + "step": 7657 + }, + { + "epoch": 0.07658, + "grad_norm": 0.8027134754140552, + "learning_rate": 0.003, + "loss": 4.1072, + "step": 7658 + }, + { + "epoch": 0.07659, + "grad_norm": 0.8125620251732755, + "learning_rate": 0.003, + "loss": 4.1028, + "step": 7659 + }, + { + "epoch": 0.0766, + "grad_norm": 0.7220469429506398, + "learning_rate": 0.003, + "loss": 4.14, + "step": 7660 + }, + { + "epoch": 0.07661, + "grad_norm": 0.715054421826916, + "learning_rate": 0.003, + "loss": 4.0987, + "step": 7661 + }, + { + "epoch": 0.07662, + "grad_norm": 0.7493362311346902, + "learning_rate": 0.003, + "loss": 4.1261, + "step": 7662 + }, + { + "epoch": 0.07663, + "grad_norm": 0.8714009180373697, + "learning_rate": 0.003, + "loss": 4.1258, + "step": 7663 + }, + { + "epoch": 0.07664, + "grad_norm": 0.9901602291385457, + "learning_rate": 0.003, + "loss": 4.1237, + "step": 7664 + }, + { + "epoch": 0.07665, + "grad_norm": 1.0074073013289961, + "learning_rate": 0.003, + "loss": 4.0996, + "step": 7665 + }, + { + "epoch": 0.07666, + "grad_norm": 0.953985834092227, + "learning_rate": 0.003, + "loss": 4.1152, + "step": 7666 + }, + { + "epoch": 0.07667, + "grad_norm": 0.9324987692217996, + "learning_rate": 0.003, + "loss": 4.1243, + "step": 7667 + }, + { + "epoch": 0.07668, + "grad_norm": 0.9181312329680767, + "learning_rate": 0.003, + "loss": 4.1146, + "step": 7668 + }, + { + "epoch": 0.07669, + "grad_norm": 1.0636826877445138, + "learning_rate": 0.003, + "loss": 4.1336, + "step": 7669 + }, + { + "epoch": 0.0767, + "grad_norm": 1.1446064632763406, + "learning_rate": 0.003, + "loss": 4.101, + "step": 7670 + }, + { + "epoch": 0.07671, + "grad_norm": 1.1178580340464093, + "learning_rate": 0.003, + "loss": 4.138, + "step": 7671 + }, + { + "epoch": 0.07672, + "grad_norm": 1.0178300133444778, + "learning_rate": 0.003, + "loss": 4.116, + "step": 7672 + }, + { + "epoch": 0.07673, + "grad_norm": 1.2016623636923163, + "learning_rate": 0.003, + "loss": 4.151, + "step": 7673 + }, + { + "epoch": 0.07674, + "grad_norm": 0.8240382566249935, + "learning_rate": 0.003, + "loss": 4.1123, + "step": 7674 + }, + { + "epoch": 0.07675, + "grad_norm": 0.8028023964885046, + "learning_rate": 0.003, + "loss": 4.1011, + "step": 7675 + }, + { + "epoch": 0.07676, + "grad_norm": 0.8231955050487717, + "learning_rate": 0.003, + "loss": 4.1009, + "step": 7676 + }, + { + "epoch": 0.07677, + "grad_norm": 0.7115485042666164, + "learning_rate": 0.003, + "loss": 4.0987, + "step": 7677 + }, + { + "epoch": 0.07678, + "grad_norm": 0.6936604353261299, + "learning_rate": 0.003, + "loss": 4.1315, + "step": 7678 + }, + { + "epoch": 0.07679, + "grad_norm": 0.8220436974564044, + "learning_rate": 0.003, + "loss": 4.1265, + "step": 7679 + }, + { + "epoch": 0.0768, + "grad_norm": 0.8824999345056386, + "learning_rate": 0.003, + "loss": 4.0991, + "step": 7680 + }, + { + "epoch": 0.07681, + "grad_norm": 1.074897714216047, + "learning_rate": 0.003, + "loss": 4.1173, + "step": 7681 + }, + { + "epoch": 0.07682, + "grad_norm": 1.1253938246986106, + "learning_rate": 0.003, + "loss": 4.106, + "step": 7682 + }, + { + "epoch": 0.07683, + "grad_norm": 0.8445187360108007, + "learning_rate": 0.003, + "loss": 4.1142, + "step": 7683 + }, + { + "epoch": 0.07684, + "grad_norm": 0.6711564131616432, + "learning_rate": 0.003, + "loss": 4.1229, + "step": 7684 + }, + { + "epoch": 0.07685, + "grad_norm": 0.8607141168125007, + "learning_rate": 0.003, + "loss": 4.1041, + "step": 7685 + }, + { + "epoch": 0.07686, + "grad_norm": 0.990613718786834, + "learning_rate": 0.003, + "loss": 4.1055, + "step": 7686 + }, + { + "epoch": 0.07687, + "grad_norm": 1.2572290057789604, + "learning_rate": 0.003, + "loss": 4.1289, + "step": 7687 + }, + { + "epoch": 0.07688, + "grad_norm": 0.8215779647578441, + "learning_rate": 0.003, + "loss": 4.1125, + "step": 7688 + }, + { + "epoch": 0.07689, + "grad_norm": 0.6891001411750275, + "learning_rate": 0.003, + "loss": 4.0926, + "step": 7689 + }, + { + "epoch": 0.0769, + "grad_norm": 0.6166117212029528, + "learning_rate": 0.003, + "loss": 4.1167, + "step": 7690 + }, + { + "epoch": 0.07691, + "grad_norm": 0.5677700647993942, + "learning_rate": 0.003, + "loss": 4.1085, + "step": 7691 + }, + { + "epoch": 0.07692, + "grad_norm": 0.7724897932547341, + "learning_rate": 0.003, + "loss": 4.1195, + "step": 7692 + }, + { + "epoch": 0.07693, + "grad_norm": 1.0012551819046691, + "learning_rate": 0.003, + "loss": 4.0968, + "step": 7693 + }, + { + "epoch": 0.07694, + "grad_norm": 1.1907585124420144, + "learning_rate": 0.003, + "loss": 4.1061, + "step": 7694 + }, + { + "epoch": 0.07695, + "grad_norm": 0.7010072537738787, + "learning_rate": 0.003, + "loss": 4.1189, + "step": 7695 + }, + { + "epoch": 0.07696, + "grad_norm": 0.6208430723305157, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 7696 + }, + { + "epoch": 0.07697, + "grad_norm": 0.6196777653103698, + "learning_rate": 0.003, + "loss": 4.134, + "step": 7697 + }, + { + "epoch": 0.07698, + "grad_norm": 0.700128726476087, + "learning_rate": 0.003, + "loss": 4.1068, + "step": 7698 + }, + { + "epoch": 0.07699, + "grad_norm": 0.6686751131176947, + "learning_rate": 0.003, + "loss": 4.1315, + "step": 7699 + }, + { + "epoch": 0.077, + "grad_norm": 0.7342255589833349, + "learning_rate": 0.003, + "loss": 4.1056, + "step": 7700 + }, + { + "epoch": 0.07701, + "grad_norm": 0.756854445847, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 7701 + }, + { + "epoch": 0.07702, + "grad_norm": 0.88266150116856, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 7702 + }, + { + "epoch": 0.07703, + "grad_norm": 0.9388488634222688, + "learning_rate": 0.003, + "loss": 4.1075, + "step": 7703 + }, + { + "epoch": 0.07704, + "grad_norm": 0.8712150471498338, + "learning_rate": 0.003, + "loss": 4.1084, + "step": 7704 + }, + { + "epoch": 0.07705, + "grad_norm": 0.7905979661669048, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 7705 + }, + { + "epoch": 0.07706, + "grad_norm": 0.8554435529838232, + "learning_rate": 0.003, + "loss": 4.1148, + "step": 7706 + }, + { + "epoch": 0.07707, + "grad_norm": 0.950211545207407, + "learning_rate": 0.003, + "loss": 4.1114, + "step": 7707 + }, + { + "epoch": 0.07708, + "grad_norm": 0.9941562838099525, + "learning_rate": 0.003, + "loss": 4.1271, + "step": 7708 + }, + { + "epoch": 0.07709, + "grad_norm": 0.9405303791524773, + "learning_rate": 0.003, + "loss": 4.1115, + "step": 7709 + }, + { + "epoch": 0.0771, + "grad_norm": 0.8363759541957757, + "learning_rate": 0.003, + "loss": 4.0856, + "step": 7710 + }, + { + "epoch": 0.07711, + "grad_norm": 0.733875954007421, + "learning_rate": 0.003, + "loss": 4.0902, + "step": 7711 + }, + { + "epoch": 0.07712, + "grad_norm": 0.6560253796119527, + "learning_rate": 0.003, + "loss": 4.0968, + "step": 7712 + }, + { + "epoch": 0.07713, + "grad_norm": 0.6759517680276719, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 7713 + }, + { + "epoch": 0.07714, + "grad_norm": 1.0143647874314274, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 7714 + }, + { + "epoch": 0.07715, + "grad_norm": 1.232675077798191, + "learning_rate": 0.003, + "loss": 4.1085, + "step": 7715 + }, + { + "epoch": 0.07716, + "grad_norm": 0.8270400357958588, + "learning_rate": 0.003, + "loss": 4.0902, + "step": 7716 + }, + { + "epoch": 0.07717, + "grad_norm": 0.7208914160909232, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 7717 + }, + { + "epoch": 0.07718, + "grad_norm": 0.7014829688376748, + "learning_rate": 0.003, + "loss": 4.1251, + "step": 7718 + }, + { + "epoch": 0.07719, + "grad_norm": 0.7014046527951289, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 7719 + }, + { + "epoch": 0.0772, + "grad_norm": 0.746499130996974, + "learning_rate": 0.003, + "loss": 4.1248, + "step": 7720 + }, + { + "epoch": 0.07721, + "grad_norm": 0.7004391345875081, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 7721 + }, + { + "epoch": 0.07722, + "grad_norm": 0.8408283424603382, + "learning_rate": 0.003, + "loss": 4.1201, + "step": 7722 + }, + { + "epoch": 0.07723, + "grad_norm": 1.110388676041958, + "learning_rate": 0.003, + "loss": 4.1266, + "step": 7723 + }, + { + "epoch": 0.07724, + "grad_norm": 0.9338754060953655, + "learning_rate": 0.003, + "loss": 4.1047, + "step": 7724 + }, + { + "epoch": 0.07725, + "grad_norm": 0.8624819322150851, + "learning_rate": 0.003, + "loss": 4.0946, + "step": 7725 + }, + { + "epoch": 0.07726, + "grad_norm": 0.8365430711688485, + "learning_rate": 0.003, + "loss": 4.1189, + "step": 7726 + }, + { + "epoch": 0.07727, + "grad_norm": 0.8249395305754468, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 7727 + }, + { + "epoch": 0.07728, + "grad_norm": 0.7844147351205248, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 7728 + }, + { + "epoch": 0.07729, + "grad_norm": 0.7973591326922866, + "learning_rate": 0.003, + "loss": 4.1152, + "step": 7729 + }, + { + "epoch": 0.0773, + "grad_norm": 0.7225338885856214, + "learning_rate": 0.003, + "loss": 4.1046, + "step": 7730 + }, + { + "epoch": 0.07731, + "grad_norm": 0.73397785517986, + "learning_rate": 0.003, + "loss": 4.0953, + "step": 7731 + }, + { + "epoch": 0.07732, + "grad_norm": 0.7538960206734072, + "learning_rate": 0.003, + "loss": 4.1068, + "step": 7732 + }, + { + "epoch": 0.07733, + "grad_norm": 0.9299182384721361, + "learning_rate": 0.003, + "loss": 4.1229, + "step": 7733 + }, + { + "epoch": 0.07734, + "grad_norm": 1.1092844376628306, + "learning_rate": 0.003, + "loss": 4.1306, + "step": 7734 + }, + { + "epoch": 0.07735, + "grad_norm": 0.9361685548610317, + "learning_rate": 0.003, + "loss": 4.0998, + "step": 7735 + }, + { + "epoch": 0.07736, + "grad_norm": 0.8101474819136633, + "learning_rate": 0.003, + "loss": 4.119, + "step": 7736 + }, + { + "epoch": 0.07737, + "grad_norm": 0.8203573852291894, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 7737 + }, + { + "epoch": 0.07738, + "grad_norm": 1.169891214994567, + "learning_rate": 0.003, + "loss": 4.1043, + "step": 7738 + }, + { + "epoch": 0.07739, + "grad_norm": 1.2019762033842287, + "learning_rate": 0.003, + "loss": 4.1176, + "step": 7739 + }, + { + "epoch": 0.0774, + "grad_norm": 0.9191096276408175, + "learning_rate": 0.003, + "loss": 4.0938, + "step": 7740 + }, + { + "epoch": 0.07741, + "grad_norm": 0.8019281817921025, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 7741 + }, + { + "epoch": 0.07742, + "grad_norm": 0.9593283919942299, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 7742 + }, + { + "epoch": 0.07743, + "grad_norm": 1.0190652900336852, + "learning_rate": 0.003, + "loss": 4.112, + "step": 7743 + }, + { + "epoch": 0.07744, + "grad_norm": 1.1716932492490297, + "learning_rate": 0.003, + "loss": 4.1179, + "step": 7744 + }, + { + "epoch": 0.07745, + "grad_norm": 1.0937600539399255, + "learning_rate": 0.003, + "loss": 4.1167, + "step": 7745 + }, + { + "epoch": 0.07746, + "grad_norm": 0.9602047963501511, + "learning_rate": 0.003, + "loss": 4.1159, + "step": 7746 + }, + { + "epoch": 0.07747, + "grad_norm": 0.7943597561692926, + "learning_rate": 0.003, + "loss": 4.1141, + "step": 7747 + }, + { + "epoch": 0.07748, + "grad_norm": 0.7757213618554147, + "learning_rate": 0.003, + "loss": 4.1393, + "step": 7748 + }, + { + "epoch": 0.07749, + "grad_norm": 0.791302635220645, + "learning_rate": 0.003, + "loss": 4.132, + "step": 7749 + }, + { + "epoch": 0.0775, + "grad_norm": 0.794756302153131, + "learning_rate": 0.003, + "loss": 4.0899, + "step": 7750 + }, + { + "epoch": 0.07751, + "grad_norm": 0.8515667360808754, + "learning_rate": 0.003, + "loss": 4.105, + "step": 7751 + }, + { + "epoch": 0.07752, + "grad_norm": 0.8271299496217495, + "learning_rate": 0.003, + "loss": 4.1247, + "step": 7752 + }, + { + "epoch": 0.07753, + "grad_norm": 0.8707061104805746, + "learning_rate": 0.003, + "loss": 4.0965, + "step": 7753 + }, + { + "epoch": 0.07754, + "grad_norm": 0.8698105492883605, + "learning_rate": 0.003, + "loss": 4.1031, + "step": 7754 + }, + { + "epoch": 0.07755, + "grad_norm": 0.9757058824840338, + "learning_rate": 0.003, + "loss": 4.1107, + "step": 7755 + }, + { + "epoch": 0.07756, + "grad_norm": 1.0996218275953191, + "learning_rate": 0.003, + "loss": 4.1267, + "step": 7756 + }, + { + "epoch": 0.07757, + "grad_norm": 1.0046296348538857, + "learning_rate": 0.003, + "loss": 4.0942, + "step": 7757 + }, + { + "epoch": 0.07758, + "grad_norm": 1.088474067480404, + "learning_rate": 0.003, + "loss": 4.0957, + "step": 7758 + }, + { + "epoch": 0.07759, + "grad_norm": 0.6838363547835872, + "learning_rate": 0.003, + "loss": 4.0978, + "step": 7759 + }, + { + "epoch": 0.0776, + "grad_norm": 0.6387940327331244, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 7760 + }, + { + "epoch": 0.07761, + "grad_norm": 0.6801202292166391, + "learning_rate": 0.003, + "loss": 4.1163, + "step": 7761 + }, + { + "epoch": 0.07762, + "grad_norm": 0.7524204138228072, + "learning_rate": 0.003, + "loss": 4.0995, + "step": 7762 + }, + { + "epoch": 0.07763, + "grad_norm": 0.8422682117877314, + "learning_rate": 0.003, + "loss": 4.1163, + "step": 7763 + }, + { + "epoch": 0.07764, + "grad_norm": 0.9464082829191608, + "learning_rate": 0.003, + "loss": 4.1254, + "step": 7764 + }, + { + "epoch": 0.07765, + "grad_norm": 1.0400578267063363, + "learning_rate": 0.003, + "loss": 4.1103, + "step": 7765 + }, + { + "epoch": 0.07766, + "grad_norm": 0.8570240484946164, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 7766 + }, + { + "epoch": 0.07767, + "grad_norm": 0.7886939048717178, + "learning_rate": 0.003, + "loss": 4.0876, + "step": 7767 + }, + { + "epoch": 0.07768, + "grad_norm": 0.7327426017118384, + "learning_rate": 0.003, + "loss": 4.0971, + "step": 7768 + }, + { + "epoch": 0.07769, + "grad_norm": 0.7224926065792775, + "learning_rate": 0.003, + "loss": 4.102, + "step": 7769 + }, + { + "epoch": 0.0777, + "grad_norm": 0.6812310474755818, + "learning_rate": 0.003, + "loss": 4.1155, + "step": 7770 + }, + { + "epoch": 0.07771, + "grad_norm": 0.6297454383181756, + "learning_rate": 0.003, + "loss": 4.1123, + "step": 7771 + }, + { + "epoch": 0.07772, + "grad_norm": 0.727035568405193, + "learning_rate": 0.003, + "loss": 4.1049, + "step": 7772 + }, + { + "epoch": 0.07773, + "grad_norm": 0.7440750703020748, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 7773 + }, + { + "epoch": 0.07774, + "grad_norm": 0.6795627443083461, + "learning_rate": 0.003, + "loss": 4.1006, + "step": 7774 + }, + { + "epoch": 0.07775, + "grad_norm": 0.6784508298159274, + "learning_rate": 0.003, + "loss": 4.1091, + "step": 7775 + }, + { + "epoch": 0.07776, + "grad_norm": 0.6385790918136247, + "learning_rate": 0.003, + "loss": 4.0967, + "step": 7776 + }, + { + "epoch": 0.07777, + "grad_norm": 0.6872009152821507, + "learning_rate": 0.003, + "loss": 4.1109, + "step": 7777 + }, + { + "epoch": 0.07778, + "grad_norm": 0.6675405617900448, + "learning_rate": 0.003, + "loss": 4.1189, + "step": 7778 + }, + { + "epoch": 0.07779, + "grad_norm": 0.7357317176753875, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 7779 + }, + { + "epoch": 0.0778, + "grad_norm": 0.8022677370632643, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 7780 + }, + { + "epoch": 0.07781, + "grad_norm": 0.9434523581478621, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 7781 + }, + { + "epoch": 0.07782, + "grad_norm": 1.5623982009752648, + "learning_rate": 0.003, + "loss": 4.0998, + "step": 7782 + }, + { + "epoch": 0.07783, + "grad_norm": 0.7044738718569362, + "learning_rate": 0.003, + "loss": 4.0939, + "step": 7783 + }, + { + "epoch": 0.07784, + "grad_norm": 0.7245143997153488, + "learning_rate": 0.003, + "loss": 4.0936, + "step": 7784 + }, + { + "epoch": 0.07785, + "grad_norm": 0.716698916381954, + "learning_rate": 0.003, + "loss": 4.1197, + "step": 7785 + }, + { + "epoch": 0.07786, + "grad_norm": 0.6921187171985951, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 7786 + }, + { + "epoch": 0.07787, + "grad_norm": 0.6045099384961314, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 7787 + }, + { + "epoch": 0.07788, + "grad_norm": 0.6083002796628949, + "learning_rate": 0.003, + "loss": 4.0855, + "step": 7788 + }, + { + "epoch": 0.07789, + "grad_norm": 0.6450261280850079, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 7789 + }, + { + "epoch": 0.0779, + "grad_norm": 0.7723456314060048, + "learning_rate": 0.003, + "loss": 4.1255, + "step": 7790 + }, + { + "epoch": 0.07791, + "grad_norm": 0.9626934092811175, + "learning_rate": 0.003, + "loss": 4.0936, + "step": 7791 + }, + { + "epoch": 0.07792, + "grad_norm": 1.0213202281359437, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 7792 + }, + { + "epoch": 0.07793, + "grad_norm": 0.9823991856862937, + "learning_rate": 0.003, + "loss": 4.0858, + "step": 7793 + }, + { + "epoch": 0.07794, + "grad_norm": 1.0324749912970457, + "learning_rate": 0.003, + "loss": 4.1365, + "step": 7794 + }, + { + "epoch": 0.07795, + "grad_norm": 1.2976167109237968, + "learning_rate": 0.003, + "loss": 4.0864, + "step": 7795 + }, + { + "epoch": 0.07796, + "grad_norm": 0.8902788450024939, + "learning_rate": 0.003, + "loss": 4.0992, + "step": 7796 + }, + { + "epoch": 0.07797, + "grad_norm": 0.9291964989058917, + "learning_rate": 0.003, + "loss": 4.0915, + "step": 7797 + }, + { + "epoch": 0.07798, + "grad_norm": 0.9627682686730562, + "learning_rate": 0.003, + "loss": 4.132, + "step": 7798 + }, + { + "epoch": 0.07799, + "grad_norm": 1.1456962518858615, + "learning_rate": 0.003, + "loss": 4.1401, + "step": 7799 + }, + { + "epoch": 0.078, + "grad_norm": 1.0251765436245048, + "learning_rate": 0.003, + "loss": 4.0986, + "step": 7800 + }, + { + "epoch": 0.07801, + "grad_norm": 0.9973687128151619, + "learning_rate": 0.003, + "loss": 4.1117, + "step": 7801 + }, + { + "epoch": 0.07802, + "grad_norm": 0.948086135583704, + "learning_rate": 0.003, + "loss": 4.0913, + "step": 7802 + }, + { + "epoch": 0.07803, + "grad_norm": 1.1277990578479882, + "learning_rate": 0.003, + "loss": 4.1049, + "step": 7803 + }, + { + "epoch": 0.07804, + "grad_norm": 0.9774623897022733, + "learning_rate": 0.003, + "loss": 4.1292, + "step": 7804 + }, + { + "epoch": 0.07805, + "grad_norm": 0.8320985331874762, + "learning_rate": 0.003, + "loss": 4.1153, + "step": 7805 + }, + { + "epoch": 0.07806, + "grad_norm": 0.8657196694590319, + "learning_rate": 0.003, + "loss": 4.1292, + "step": 7806 + }, + { + "epoch": 0.07807, + "grad_norm": 0.9492463229383535, + "learning_rate": 0.003, + "loss": 4.1325, + "step": 7807 + }, + { + "epoch": 0.07808, + "grad_norm": 1.0394516274093006, + "learning_rate": 0.003, + "loss": 4.1039, + "step": 7808 + }, + { + "epoch": 0.07809, + "grad_norm": 0.9335393900367771, + "learning_rate": 0.003, + "loss": 4.1157, + "step": 7809 + }, + { + "epoch": 0.0781, + "grad_norm": 0.9443482076739536, + "learning_rate": 0.003, + "loss": 4.0987, + "step": 7810 + }, + { + "epoch": 0.07811, + "grad_norm": 0.9123874992413225, + "learning_rate": 0.003, + "loss": 4.1105, + "step": 7811 + }, + { + "epoch": 0.07812, + "grad_norm": 1.1264587870913032, + "learning_rate": 0.003, + "loss": 4.1125, + "step": 7812 + }, + { + "epoch": 0.07813, + "grad_norm": 1.1305952516582176, + "learning_rate": 0.003, + "loss": 4.1298, + "step": 7813 + }, + { + "epoch": 0.07814, + "grad_norm": 0.8414805101214488, + "learning_rate": 0.003, + "loss": 4.1511, + "step": 7814 + }, + { + "epoch": 0.07815, + "grad_norm": 0.9502914015578153, + "learning_rate": 0.003, + "loss": 4.1164, + "step": 7815 + }, + { + "epoch": 0.07816, + "grad_norm": 1.1760246778581924, + "learning_rate": 0.003, + "loss": 4.1092, + "step": 7816 + }, + { + "epoch": 0.07817, + "grad_norm": 0.9408621866890481, + "learning_rate": 0.003, + "loss": 4.1087, + "step": 7817 + }, + { + "epoch": 0.07818, + "grad_norm": 0.8296428274925994, + "learning_rate": 0.003, + "loss": 4.0989, + "step": 7818 + }, + { + "epoch": 0.07819, + "grad_norm": 0.7920801052079187, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 7819 + }, + { + "epoch": 0.0782, + "grad_norm": 0.8823722442311102, + "learning_rate": 0.003, + "loss": 4.1285, + "step": 7820 + }, + { + "epoch": 0.07821, + "grad_norm": 0.9516339518218203, + "learning_rate": 0.003, + "loss": 4.0937, + "step": 7821 + }, + { + "epoch": 0.07822, + "grad_norm": 1.0547110747707245, + "learning_rate": 0.003, + "loss": 4.0876, + "step": 7822 + }, + { + "epoch": 0.07823, + "grad_norm": 0.989883022467342, + "learning_rate": 0.003, + "loss": 4.1007, + "step": 7823 + }, + { + "epoch": 0.07824, + "grad_norm": 0.8914027725354553, + "learning_rate": 0.003, + "loss": 4.131, + "step": 7824 + }, + { + "epoch": 0.07825, + "grad_norm": 0.8076840807799915, + "learning_rate": 0.003, + "loss": 4.0934, + "step": 7825 + }, + { + "epoch": 0.07826, + "grad_norm": 0.7415708695700829, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 7826 + }, + { + "epoch": 0.07827, + "grad_norm": 0.7214257388849613, + "learning_rate": 0.003, + "loss": 4.1111, + "step": 7827 + }, + { + "epoch": 0.07828, + "grad_norm": 0.8136232450263843, + "learning_rate": 0.003, + "loss": 4.1138, + "step": 7828 + }, + { + "epoch": 0.07829, + "grad_norm": 0.9199584413946614, + "learning_rate": 0.003, + "loss": 4.1221, + "step": 7829 + }, + { + "epoch": 0.0783, + "grad_norm": 0.8115857466773752, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 7830 + }, + { + "epoch": 0.07831, + "grad_norm": 0.7591405213641907, + "learning_rate": 0.003, + "loss": 4.0935, + "step": 7831 + }, + { + "epoch": 0.07832, + "grad_norm": 0.6490760526643984, + "learning_rate": 0.003, + "loss": 4.0997, + "step": 7832 + }, + { + "epoch": 0.07833, + "grad_norm": 0.703850585915276, + "learning_rate": 0.003, + "loss": 4.099, + "step": 7833 + }, + { + "epoch": 0.07834, + "grad_norm": 0.8284668702922542, + "learning_rate": 0.003, + "loss": 4.1513, + "step": 7834 + }, + { + "epoch": 0.07835, + "grad_norm": 0.8475382307293997, + "learning_rate": 0.003, + "loss": 4.1145, + "step": 7835 + }, + { + "epoch": 0.07836, + "grad_norm": 0.800674520894362, + "learning_rate": 0.003, + "loss": 4.0968, + "step": 7836 + }, + { + "epoch": 0.07837, + "grad_norm": 0.7804816932384945, + "learning_rate": 0.003, + "loss": 4.107, + "step": 7837 + }, + { + "epoch": 0.07838, + "grad_norm": 0.7719832162406756, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 7838 + }, + { + "epoch": 0.07839, + "grad_norm": 0.8103102203464976, + "learning_rate": 0.003, + "loss": 4.0871, + "step": 7839 + }, + { + "epoch": 0.0784, + "grad_norm": 0.847847320739481, + "learning_rate": 0.003, + "loss": 4.1155, + "step": 7840 + }, + { + "epoch": 0.07841, + "grad_norm": 0.9217204569862923, + "learning_rate": 0.003, + "loss": 4.0931, + "step": 7841 + }, + { + "epoch": 0.07842, + "grad_norm": 1.0755759756691639, + "learning_rate": 0.003, + "loss": 4.0903, + "step": 7842 + }, + { + "epoch": 0.07843, + "grad_norm": 1.0372058631482444, + "learning_rate": 0.003, + "loss": 4.1073, + "step": 7843 + }, + { + "epoch": 0.07844, + "grad_norm": 0.8852971276699896, + "learning_rate": 0.003, + "loss": 4.1061, + "step": 7844 + }, + { + "epoch": 0.07845, + "grad_norm": 0.7984782327828842, + "learning_rate": 0.003, + "loss": 4.1157, + "step": 7845 + }, + { + "epoch": 0.07846, + "grad_norm": 0.8806483710377667, + "learning_rate": 0.003, + "loss": 4.1337, + "step": 7846 + }, + { + "epoch": 0.07847, + "grad_norm": 1.0378492921152178, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 7847 + }, + { + "epoch": 0.07848, + "grad_norm": 0.895360780588699, + "learning_rate": 0.003, + "loss": 4.1303, + "step": 7848 + }, + { + "epoch": 0.07849, + "grad_norm": 0.8289639586531552, + "learning_rate": 0.003, + "loss": 4.1355, + "step": 7849 + }, + { + "epoch": 0.0785, + "grad_norm": 0.8305063229582818, + "learning_rate": 0.003, + "loss": 4.1039, + "step": 7850 + }, + { + "epoch": 0.07851, + "grad_norm": 0.7058129266076502, + "learning_rate": 0.003, + "loss": 4.0919, + "step": 7851 + }, + { + "epoch": 0.07852, + "grad_norm": 0.6418192413394789, + "learning_rate": 0.003, + "loss": 4.1283, + "step": 7852 + }, + { + "epoch": 0.07853, + "grad_norm": 0.7531346386515196, + "learning_rate": 0.003, + "loss": 4.0957, + "step": 7853 + }, + { + "epoch": 0.07854, + "grad_norm": 0.8398583249558569, + "learning_rate": 0.003, + "loss": 4.1222, + "step": 7854 + }, + { + "epoch": 0.07855, + "grad_norm": 0.7774066949599959, + "learning_rate": 0.003, + "loss": 4.1053, + "step": 7855 + }, + { + "epoch": 0.07856, + "grad_norm": 1.0200945417324516, + "learning_rate": 0.003, + "loss": 4.1135, + "step": 7856 + }, + { + "epoch": 0.07857, + "grad_norm": 1.075009695811694, + "learning_rate": 0.003, + "loss": 4.1381, + "step": 7857 + }, + { + "epoch": 0.07858, + "grad_norm": 0.7910404278344594, + "learning_rate": 0.003, + "loss": 4.1217, + "step": 7858 + }, + { + "epoch": 0.07859, + "grad_norm": 0.7300829664682413, + "learning_rate": 0.003, + "loss": 4.0898, + "step": 7859 + }, + { + "epoch": 0.0786, + "grad_norm": 0.7197539823867891, + "learning_rate": 0.003, + "loss": 4.102, + "step": 7860 + }, + { + "epoch": 0.07861, + "grad_norm": 0.9620546769547311, + "learning_rate": 0.003, + "loss": 4.1242, + "step": 7861 + }, + { + "epoch": 0.07862, + "grad_norm": 1.490251553426132, + "learning_rate": 0.003, + "loss": 4.1592, + "step": 7862 + }, + { + "epoch": 0.07863, + "grad_norm": 0.7857679852194924, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 7863 + }, + { + "epoch": 0.07864, + "grad_norm": 0.8634479975354046, + "learning_rate": 0.003, + "loss": 4.118, + "step": 7864 + }, + { + "epoch": 0.07865, + "grad_norm": 0.7964685970693094, + "learning_rate": 0.003, + "loss": 4.0956, + "step": 7865 + }, + { + "epoch": 0.07866, + "grad_norm": 0.6828020542596382, + "learning_rate": 0.003, + "loss": 4.0973, + "step": 7866 + }, + { + "epoch": 0.07867, + "grad_norm": 0.6092921432164138, + "learning_rate": 0.003, + "loss": 4.0848, + "step": 7867 + }, + { + "epoch": 0.07868, + "grad_norm": 0.5633101550595739, + "learning_rate": 0.003, + "loss": 4.1168, + "step": 7868 + }, + { + "epoch": 0.07869, + "grad_norm": 0.6431111329973238, + "learning_rate": 0.003, + "loss": 4.0962, + "step": 7869 + }, + { + "epoch": 0.0787, + "grad_norm": 0.901303356812522, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 7870 + }, + { + "epoch": 0.07871, + "grad_norm": 1.2191547805657925, + "learning_rate": 0.003, + "loss": 4.1382, + "step": 7871 + }, + { + "epoch": 0.07872, + "grad_norm": 0.8262977027782229, + "learning_rate": 0.003, + "loss": 4.1284, + "step": 7872 + }, + { + "epoch": 0.07873, + "grad_norm": 0.6724946366349702, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 7873 + }, + { + "epoch": 0.07874, + "grad_norm": 0.7244287405483971, + "learning_rate": 0.003, + "loss": 4.0977, + "step": 7874 + }, + { + "epoch": 0.07875, + "grad_norm": 0.8227798028339546, + "learning_rate": 0.003, + "loss": 4.1129, + "step": 7875 + }, + { + "epoch": 0.07876, + "grad_norm": 0.9349875722236806, + "learning_rate": 0.003, + "loss": 4.1097, + "step": 7876 + }, + { + "epoch": 0.07877, + "grad_norm": 1.0794801674884056, + "learning_rate": 0.003, + "loss": 4.099, + "step": 7877 + }, + { + "epoch": 0.07878, + "grad_norm": 0.9276580973578815, + "learning_rate": 0.003, + "loss": 4.1074, + "step": 7878 + }, + { + "epoch": 0.07879, + "grad_norm": 0.8251333456721903, + "learning_rate": 0.003, + "loss": 4.1068, + "step": 7879 + }, + { + "epoch": 0.0788, + "grad_norm": 0.7060382543658212, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 7880 + }, + { + "epoch": 0.07881, + "grad_norm": 0.67303988544837, + "learning_rate": 0.003, + "loss": 4.0959, + "step": 7881 + }, + { + "epoch": 0.07882, + "grad_norm": 0.6233971310328372, + "learning_rate": 0.003, + "loss": 4.0906, + "step": 7882 + }, + { + "epoch": 0.07883, + "grad_norm": 0.6298409951352602, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 7883 + }, + { + "epoch": 0.07884, + "grad_norm": 0.7947655306250354, + "learning_rate": 0.003, + "loss": 4.0973, + "step": 7884 + }, + { + "epoch": 0.07885, + "grad_norm": 0.8497398170009447, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 7885 + }, + { + "epoch": 0.07886, + "grad_norm": 0.8619830961416387, + "learning_rate": 0.003, + "loss": 4.1118, + "step": 7886 + }, + { + "epoch": 0.07887, + "grad_norm": 1.0484039664176932, + "learning_rate": 0.003, + "loss": 4.0927, + "step": 7887 + }, + { + "epoch": 0.07888, + "grad_norm": 1.1847395590743321, + "learning_rate": 0.003, + "loss": 4.1116, + "step": 7888 + }, + { + "epoch": 0.07889, + "grad_norm": 0.8493272011871066, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 7889 + }, + { + "epoch": 0.0789, + "grad_norm": 0.8008301667726648, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 7890 + }, + { + "epoch": 0.07891, + "grad_norm": 0.930077598807836, + "learning_rate": 0.003, + "loss": 4.0961, + "step": 7891 + }, + { + "epoch": 0.07892, + "grad_norm": 1.0989482964593247, + "learning_rate": 0.003, + "loss": 4.1012, + "step": 7892 + }, + { + "epoch": 0.07893, + "grad_norm": 1.1363252018029342, + "learning_rate": 0.003, + "loss": 4.1224, + "step": 7893 + }, + { + "epoch": 0.07894, + "grad_norm": 1.0228512594068824, + "learning_rate": 0.003, + "loss": 4.0919, + "step": 7894 + }, + { + "epoch": 0.07895, + "grad_norm": 0.9218498361406187, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 7895 + }, + { + "epoch": 0.07896, + "grad_norm": 0.9251940784692912, + "learning_rate": 0.003, + "loss": 4.1029, + "step": 7896 + }, + { + "epoch": 0.07897, + "grad_norm": 0.931960102854229, + "learning_rate": 0.003, + "loss": 4.1241, + "step": 7897 + }, + { + "epoch": 0.07898, + "grad_norm": 0.8464573363229492, + "learning_rate": 0.003, + "loss": 4.1031, + "step": 7898 + }, + { + "epoch": 0.07899, + "grad_norm": 0.7028868962428811, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 7899 + }, + { + "epoch": 0.079, + "grad_norm": 0.7441885103587236, + "learning_rate": 0.003, + "loss": 4.0898, + "step": 7900 + }, + { + "epoch": 0.07901, + "grad_norm": 0.8664872116804634, + "learning_rate": 0.003, + "loss": 4.1245, + "step": 7901 + }, + { + "epoch": 0.07902, + "grad_norm": 0.9184607477010853, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 7902 + }, + { + "epoch": 0.07903, + "grad_norm": 0.7590321607729549, + "learning_rate": 0.003, + "loss": 4.0939, + "step": 7903 + }, + { + "epoch": 0.07904, + "grad_norm": 0.7531494902293769, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 7904 + }, + { + "epoch": 0.07905, + "grad_norm": 0.70438167321181, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 7905 + }, + { + "epoch": 0.07906, + "grad_norm": 0.6511721703071448, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 7906 + }, + { + "epoch": 0.07907, + "grad_norm": 0.716900467492343, + "learning_rate": 0.003, + "loss": 4.1048, + "step": 7907 + }, + { + "epoch": 0.07908, + "grad_norm": 0.7085281032769419, + "learning_rate": 0.003, + "loss": 4.1069, + "step": 7908 + }, + { + "epoch": 0.07909, + "grad_norm": 0.8416217578074326, + "learning_rate": 0.003, + "loss": 4.1123, + "step": 7909 + }, + { + "epoch": 0.0791, + "grad_norm": 1.1433040371010987, + "learning_rate": 0.003, + "loss": 4.0978, + "step": 7910 + }, + { + "epoch": 0.07911, + "grad_norm": 1.1835699988724921, + "learning_rate": 0.003, + "loss": 4.1226, + "step": 7911 + }, + { + "epoch": 0.07912, + "grad_norm": 0.9223643758933799, + "learning_rate": 0.003, + "loss": 4.0931, + "step": 7912 + }, + { + "epoch": 0.07913, + "grad_norm": 1.089125989023986, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 7913 + }, + { + "epoch": 0.07914, + "grad_norm": 0.8390123983066416, + "learning_rate": 0.003, + "loss": 4.0968, + "step": 7914 + }, + { + "epoch": 0.07915, + "grad_norm": 0.672914923150026, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 7915 + }, + { + "epoch": 0.07916, + "grad_norm": 0.7017447792175915, + "learning_rate": 0.003, + "loss": 4.108, + "step": 7916 + }, + { + "epoch": 0.07917, + "grad_norm": 0.7051019103527106, + "learning_rate": 0.003, + "loss": 4.1085, + "step": 7917 + }, + { + "epoch": 0.07918, + "grad_norm": 0.9855647114410215, + "learning_rate": 0.003, + "loss": 4.1062, + "step": 7918 + }, + { + "epoch": 0.07919, + "grad_norm": 1.3116973662720455, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 7919 + }, + { + "epoch": 0.0792, + "grad_norm": 0.7402493790216023, + "learning_rate": 0.003, + "loss": 4.0949, + "step": 7920 + }, + { + "epoch": 0.07921, + "grad_norm": 0.6988160486580384, + "learning_rate": 0.003, + "loss": 4.1027, + "step": 7921 + }, + { + "epoch": 0.07922, + "grad_norm": 0.7456591280144469, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 7922 + }, + { + "epoch": 0.07923, + "grad_norm": 0.8033173795113849, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 7923 + }, + { + "epoch": 0.07924, + "grad_norm": 0.7216426094382513, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 7924 + }, + { + "epoch": 0.07925, + "grad_norm": 0.7373410578068488, + "learning_rate": 0.003, + "loss": 4.1117, + "step": 7925 + }, + { + "epoch": 0.07926, + "grad_norm": 0.9580792267793108, + "learning_rate": 0.003, + "loss": 4.1202, + "step": 7926 + }, + { + "epoch": 0.07927, + "grad_norm": 1.1508484267353751, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 7927 + }, + { + "epoch": 0.07928, + "grad_norm": 0.8413926446028052, + "learning_rate": 0.003, + "loss": 4.1065, + "step": 7928 + }, + { + "epoch": 0.07929, + "grad_norm": 0.8045322762280693, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 7929 + }, + { + "epoch": 0.0793, + "grad_norm": 0.7334642856257453, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 7930 + }, + { + "epoch": 0.07931, + "grad_norm": 0.7666083653707652, + "learning_rate": 0.003, + "loss": 4.1106, + "step": 7931 + }, + { + "epoch": 0.07932, + "grad_norm": 0.845631543538524, + "learning_rate": 0.003, + "loss": 4.1118, + "step": 7932 + }, + { + "epoch": 0.07933, + "grad_norm": 1.1053771346780619, + "learning_rate": 0.003, + "loss": 4.1051, + "step": 7933 + }, + { + "epoch": 0.07934, + "grad_norm": 1.0697166718848607, + "learning_rate": 0.003, + "loss": 4.1234, + "step": 7934 + }, + { + "epoch": 0.07935, + "grad_norm": 0.780788335506962, + "learning_rate": 0.003, + "loss": 4.0996, + "step": 7935 + }, + { + "epoch": 0.07936, + "grad_norm": 0.6719369309095249, + "learning_rate": 0.003, + "loss": 4.0947, + "step": 7936 + }, + { + "epoch": 0.07937, + "grad_norm": 0.7129234476370259, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 7937 + }, + { + "epoch": 0.07938, + "grad_norm": 0.7579710871562575, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 7938 + }, + { + "epoch": 0.07939, + "grad_norm": 0.8414949927890002, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 7939 + }, + { + "epoch": 0.0794, + "grad_norm": 0.873475867512058, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 7940 + }, + { + "epoch": 0.07941, + "grad_norm": 0.7753297409475207, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 7941 + }, + { + "epoch": 0.07942, + "grad_norm": 0.7946685821619549, + "learning_rate": 0.003, + "loss": 4.073, + "step": 7942 + }, + { + "epoch": 0.07943, + "grad_norm": 0.9007048298260407, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 7943 + }, + { + "epoch": 0.07944, + "grad_norm": 1.071812362091642, + "learning_rate": 0.003, + "loss": 4.1278, + "step": 7944 + }, + { + "epoch": 0.07945, + "grad_norm": 1.0693712200612246, + "learning_rate": 0.003, + "loss": 4.1057, + "step": 7945 + }, + { + "epoch": 0.07946, + "grad_norm": 0.8795848429684096, + "learning_rate": 0.003, + "loss": 4.1052, + "step": 7946 + }, + { + "epoch": 0.07947, + "grad_norm": 0.800378114738204, + "learning_rate": 0.003, + "loss": 4.1186, + "step": 7947 + }, + { + "epoch": 0.07948, + "grad_norm": 0.9321588061699683, + "learning_rate": 0.003, + "loss": 4.1055, + "step": 7948 + }, + { + "epoch": 0.07949, + "grad_norm": 0.991434583531255, + "learning_rate": 0.003, + "loss": 4.1241, + "step": 7949 + }, + { + "epoch": 0.0795, + "grad_norm": 0.9816512046959835, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 7950 + }, + { + "epoch": 0.07951, + "grad_norm": 0.7677971687728063, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 7951 + }, + { + "epoch": 0.07952, + "grad_norm": 0.8051868602715272, + "learning_rate": 0.003, + "loss": 4.1082, + "step": 7952 + }, + { + "epoch": 0.07953, + "grad_norm": 0.9001609812725201, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 7953 + }, + { + "epoch": 0.07954, + "grad_norm": 1.1106187952067725, + "learning_rate": 0.003, + "loss": 4.1309, + "step": 7954 + }, + { + "epoch": 0.07955, + "grad_norm": 0.9649928236706383, + "learning_rate": 0.003, + "loss": 4.1288, + "step": 7955 + }, + { + "epoch": 0.07956, + "grad_norm": 1.1226837346899992, + "learning_rate": 0.003, + "loss": 4.1112, + "step": 7956 + }, + { + "epoch": 0.07957, + "grad_norm": 1.1705169222423675, + "learning_rate": 0.003, + "loss": 4.1027, + "step": 7957 + }, + { + "epoch": 0.07958, + "grad_norm": 0.8899304842940169, + "learning_rate": 0.003, + "loss": 4.1091, + "step": 7958 + }, + { + "epoch": 0.07959, + "grad_norm": 0.7210543206553331, + "learning_rate": 0.003, + "loss": 4.1089, + "step": 7959 + }, + { + "epoch": 0.0796, + "grad_norm": 0.8474860169251741, + "learning_rate": 0.003, + "loss": 4.1099, + "step": 7960 + }, + { + "epoch": 0.07961, + "grad_norm": 0.8510145192053272, + "learning_rate": 0.003, + "loss": 4.0956, + "step": 7961 + }, + { + "epoch": 0.07962, + "grad_norm": 0.9477244770512336, + "learning_rate": 0.003, + "loss": 4.1214, + "step": 7962 + }, + { + "epoch": 0.07963, + "grad_norm": 0.9694960562391948, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 7963 + }, + { + "epoch": 0.07964, + "grad_norm": 1.0096222672390198, + "learning_rate": 0.003, + "loss": 4.1071, + "step": 7964 + }, + { + "epoch": 0.07965, + "grad_norm": 0.9843752313745823, + "learning_rate": 0.003, + "loss": 4.1184, + "step": 7965 + }, + { + "epoch": 0.07966, + "grad_norm": 0.9595021912517606, + "learning_rate": 0.003, + "loss": 4.095, + "step": 7966 + }, + { + "epoch": 0.07967, + "grad_norm": 1.0016620302362607, + "learning_rate": 0.003, + "loss": 4.1206, + "step": 7967 + }, + { + "epoch": 0.07968, + "grad_norm": 0.9784507337152613, + "learning_rate": 0.003, + "loss": 4.0908, + "step": 7968 + }, + { + "epoch": 0.07969, + "grad_norm": 0.9283133075078502, + "learning_rate": 0.003, + "loss": 4.1292, + "step": 7969 + }, + { + "epoch": 0.0797, + "grad_norm": 0.9854134391658027, + "learning_rate": 0.003, + "loss": 4.1122, + "step": 7970 + }, + { + "epoch": 0.07971, + "grad_norm": 1.0491313465828465, + "learning_rate": 0.003, + "loss": 4.1012, + "step": 7971 + }, + { + "epoch": 0.07972, + "grad_norm": 1.0232565756589487, + "learning_rate": 0.003, + "loss": 4.1043, + "step": 7972 + }, + { + "epoch": 0.07973, + "grad_norm": 0.9389427379261678, + "learning_rate": 0.003, + "loss": 4.1196, + "step": 7973 + }, + { + "epoch": 0.07974, + "grad_norm": 0.9196943064318163, + "learning_rate": 0.003, + "loss": 4.1003, + "step": 7974 + }, + { + "epoch": 0.07975, + "grad_norm": 0.8757962873318504, + "learning_rate": 0.003, + "loss": 4.1336, + "step": 7975 + }, + { + "epoch": 0.07976, + "grad_norm": 0.8223659364709709, + "learning_rate": 0.003, + "loss": 4.1146, + "step": 7976 + }, + { + "epoch": 0.07977, + "grad_norm": 0.8719111818094638, + "learning_rate": 0.003, + "loss": 4.108, + "step": 7977 + }, + { + "epoch": 0.07978, + "grad_norm": 0.8824234915401362, + "learning_rate": 0.003, + "loss": 4.1158, + "step": 7978 + }, + { + "epoch": 0.07979, + "grad_norm": 0.9247964543100683, + "learning_rate": 0.003, + "loss": 4.099, + "step": 7979 + }, + { + "epoch": 0.0798, + "grad_norm": 0.9397634918432941, + "learning_rate": 0.003, + "loss": 4.1299, + "step": 7980 + }, + { + "epoch": 0.07981, + "grad_norm": 1.0543674923134152, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 7981 + }, + { + "epoch": 0.07982, + "grad_norm": 0.9626911947278476, + "learning_rate": 0.003, + "loss": 4.1173, + "step": 7982 + }, + { + "epoch": 0.07983, + "grad_norm": 1.0080240905133067, + "learning_rate": 0.003, + "loss": 4.1443, + "step": 7983 + }, + { + "epoch": 0.07984, + "grad_norm": 1.2190721528512718, + "learning_rate": 0.003, + "loss": 4.1319, + "step": 7984 + }, + { + "epoch": 0.07985, + "grad_norm": 0.8286713865087831, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 7985 + }, + { + "epoch": 0.07986, + "grad_norm": 0.8677596520468425, + "learning_rate": 0.003, + "loss": 4.1251, + "step": 7986 + }, + { + "epoch": 0.07987, + "grad_norm": 0.839953559356892, + "learning_rate": 0.003, + "loss": 4.0953, + "step": 7987 + }, + { + "epoch": 0.07988, + "grad_norm": 0.8616747266554645, + "learning_rate": 0.003, + "loss": 4.1127, + "step": 7988 + }, + { + "epoch": 0.07989, + "grad_norm": 0.8768670980849567, + "learning_rate": 0.003, + "loss": 4.0857, + "step": 7989 + }, + { + "epoch": 0.0799, + "grad_norm": 1.060320785667829, + "learning_rate": 0.003, + "loss": 4.1073, + "step": 7990 + }, + { + "epoch": 0.07991, + "grad_norm": 1.2397246116873737, + "learning_rate": 0.003, + "loss": 4.1086, + "step": 7991 + }, + { + "epoch": 0.07992, + "grad_norm": 0.8564943829168591, + "learning_rate": 0.003, + "loss": 4.1067, + "step": 7992 + }, + { + "epoch": 0.07993, + "grad_norm": 0.6329653057644131, + "learning_rate": 0.003, + "loss": 4.0928, + "step": 7993 + }, + { + "epoch": 0.07994, + "grad_norm": 0.5495879883814838, + "learning_rate": 0.003, + "loss": 4.1109, + "step": 7994 + }, + { + "epoch": 0.07995, + "grad_norm": 0.6418923191491803, + "learning_rate": 0.003, + "loss": 4.1091, + "step": 7995 + }, + { + "epoch": 0.07996, + "grad_norm": 0.6360153656004878, + "learning_rate": 0.003, + "loss": 4.1003, + "step": 7996 + }, + { + "epoch": 0.07997, + "grad_norm": 0.6967070433022678, + "learning_rate": 0.003, + "loss": 4.0979, + "step": 7997 + }, + { + "epoch": 0.07998, + "grad_norm": 0.8401425534763325, + "learning_rate": 0.003, + "loss": 4.1299, + "step": 7998 + }, + { + "epoch": 0.07999, + "grad_norm": 0.9140910976472046, + "learning_rate": 0.003, + "loss": 4.0999, + "step": 7999 + }, + { + "epoch": 0.08, + "grad_norm": 0.8341075710754534, + "learning_rate": 0.003, + "loss": 4.1081, + "step": 8000 + }, + { + "epoch": 0.08001, + "grad_norm": 0.6762636347053551, + "learning_rate": 0.003, + "loss": 4.0902, + "step": 8001 + }, + { + "epoch": 0.08002, + "grad_norm": 0.7219182957365917, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 8002 + }, + { + "epoch": 0.08003, + "grad_norm": 0.7307059911401975, + "learning_rate": 0.003, + "loss": 4.1094, + "step": 8003 + }, + { + "epoch": 0.08004, + "grad_norm": 0.7690134766684371, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 8004 + }, + { + "epoch": 0.08005, + "grad_norm": 0.7032812824393433, + "learning_rate": 0.003, + "loss": 4.1121, + "step": 8005 + }, + { + "epoch": 0.08006, + "grad_norm": 0.8083021905718699, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 8006 + }, + { + "epoch": 0.08007, + "grad_norm": 1.0155775443805966, + "learning_rate": 0.003, + "loss": 4.0881, + "step": 8007 + }, + { + "epoch": 0.08008, + "grad_norm": 1.2028469175430232, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 8008 + }, + { + "epoch": 0.08009, + "grad_norm": 0.7132237916366507, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 8009 + }, + { + "epoch": 0.0801, + "grad_norm": 0.5404697854674432, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 8010 + }, + { + "epoch": 0.08011, + "grad_norm": 0.5947072885434764, + "learning_rate": 0.003, + "loss": 4.0933, + "step": 8011 + }, + { + "epoch": 0.08012, + "grad_norm": 0.7363521666233862, + "learning_rate": 0.003, + "loss": 4.094, + "step": 8012 + }, + { + "epoch": 0.08013, + "grad_norm": 1.0063376322547504, + "learning_rate": 0.003, + "loss": 4.1003, + "step": 8013 + }, + { + "epoch": 0.08014, + "grad_norm": 1.1160436669300626, + "learning_rate": 0.003, + "loss": 4.09, + "step": 8014 + }, + { + "epoch": 0.08015, + "grad_norm": 0.7989320079754313, + "learning_rate": 0.003, + "loss": 4.1159, + "step": 8015 + }, + { + "epoch": 0.08016, + "grad_norm": 0.7682760666118215, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 8016 + }, + { + "epoch": 0.08017, + "grad_norm": 0.7989671209318877, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 8017 + }, + { + "epoch": 0.08018, + "grad_norm": 0.8900732101990633, + "learning_rate": 0.003, + "loss": 4.0946, + "step": 8018 + }, + { + "epoch": 0.08019, + "grad_norm": 0.7809140627865704, + "learning_rate": 0.003, + "loss": 4.1184, + "step": 8019 + }, + { + "epoch": 0.0802, + "grad_norm": 0.8370791311704688, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 8020 + }, + { + "epoch": 0.08021, + "grad_norm": 0.7244688057921231, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 8021 + }, + { + "epoch": 0.08022, + "grad_norm": 0.7032261987394863, + "learning_rate": 0.003, + "loss": 4.0991, + "step": 8022 + }, + { + "epoch": 0.08023, + "grad_norm": 0.7746372857782606, + "learning_rate": 0.003, + "loss": 4.0878, + "step": 8023 + }, + { + "epoch": 0.08024, + "grad_norm": 0.980871731492932, + "learning_rate": 0.003, + "loss": 4.1094, + "step": 8024 + }, + { + "epoch": 0.08025, + "grad_norm": 1.1967229295248476, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 8025 + }, + { + "epoch": 0.08026, + "grad_norm": 0.8874537036909782, + "learning_rate": 0.003, + "loss": 4.0908, + "step": 8026 + }, + { + "epoch": 0.08027, + "grad_norm": 0.7886044707135144, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 8027 + }, + { + "epoch": 0.08028, + "grad_norm": 0.7195647660407796, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 8028 + }, + { + "epoch": 0.08029, + "grad_norm": 0.8677533765236772, + "learning_rate": 0.003, + "loss": 4.1126, + "step": 8029 + }, + { + "epoch": 0.0803, + "grad_norm": 1.006614001349766, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 8030 + }, + { + "epoch": 0.08031, + "grad_norm": 1.1001052865444716, + "learning_rate": 0.003, + "loss": 4.1064, + "step": 8031 + }, + { + "epoch": 0.08032, + "grad_norm": 0.9468984767405869, + "learning_rate": 0.003, + "loss": 4.0902, + "step": 8032 + }, + { + "epoch": 0.08033, + "grad_norm": 0.7392931914642568, + "learning_rate": 0.003, + "loss": 4.1144, + "step": 8033 + }, + { + "epoch": 0.08034, + "grad_norm": 0.8147657234401885, + "learning_rate": 0.003, + "loss": 4.1125, + "step": 8034 + }, + { + "epoch": 0.08035, + "grad_norm": 0.9473798784474897, + "learning_rate": 0.003, + "loss": 4.1259, + "step": 8035 + }, + { + "epoch": 0.08036, + "grad_norm": 0.9663624382781947, + "learning_rate": 0.003, + "loss": 4.1423, + "step": 8036 + }, + { + "epoch": 0.08037, + "grad_norm": 0.9851396854632779, + "learning_rate": 0.003, + "loss": 4.1002, + "step": 8037 + }, + { + "epoch": 0.08038, + "grad_norm": 0.9226417242283115, + "learning_rate": 0.003, + "loss": 4.0977, + "step": 8038 + }, + { + "epoch": 0.08039, + "grad_norm": 0.9417189662091079, + "learning_rate": 0.003, + "loss": 4.1065, + "step": 8039 + }, + { + "epoch": 0.0804, + "grad_norm": 1.1112977827509558, + "learning_rate": 0.003, + "loss": 4.1227, + "step": 8040 + }, + { + "epoch": 0.08041, + "grad_norm": 0.7739549217316514, + "learning_rate": 0.003, + "loss": 4.1207, + "step": 8041 + }, + { + "epoch": 0.08042, + "grad_norm": 0.6673108385892449, + "learning_rate": 0.003, + "loss": 4.1019, + "step": 8042 + }, + { + "epoch": 0.08043, + "grad_norm": 0.7412544947569171, + "learning_rate": 0.003, + "loss": 4.0956, + "step": 8043 + }, + { + "epoch": 0.08044, + "grad_norm": 0.7934062901474336, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 8044 + }, + { + "epoch": 0.08045, + "grad_norm": 0.939659696654444, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 8045 + }, + { + "epoch": 0.08046, + "grad_norm": 1.0779216512774807, + "learning_rate": 0.003, + "loss": 4.1042, + "step": 8046 + }, + { + "epoch": 0.08047, + "grad_norm": 1.0583855488249307, + "learning_rate": 0.003, + "loss": 4.101, + "step": 8047 + }, + { + "epoch": 0.08048, + "grad_norm": 1.010247891168986, + "learning_rate": 0.003, + "loss": 4.111, + "step": 8048 + }, + { + "epoch": 0.08049, + "grad_norm": 0.8559934548803988, + "learning_rate": 0.003, + "loss": 4.091, + "step": 8049 + }, + { + "epoch": 0.0805, + "grad_norm": 0.785962682018867, + "learning_rate": 0.003, + "loss": 4.1154, + "step": 8050 + }, + { + "epoch": 0.08051, + "grad_norm": 0.8243401452192997, + "learning_rate": 0.003, + "loss": 4.0936, + "step": 8051 + }, + { + "epoch": 0.08052, + "grad_norm": 0.8118236172127069, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 8052 + }, + { + "epoch": 0.08053, + "grad_norm": 0.7919496704600435, + "learning_rate": 0.003, + "loss": 4.1142, + "step": 8053 + }, + { + "epoch": 0.08054, + "grad_norm": 0.7931762362063768, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 8054 + }, + { + "epoch": 0.08055, + "grad_norm": 0.8667155845252095, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 8055 + }, + { + "epoch": 0.08056, + "grad_norm": 1.0124039809076317, + "learning_rate": 0.003, + "loss": 4.1054, + "step": 8056 + }, + { + "epoch": 0.08057, + "grad_norm": 1.1336788875699602, + "learning_rate": 0.003, + "loss": 4.1141, + "step": 8057 + }, + { + "epoch": 0.08058, + "grad_norm": 0.8138250876786765, + "learning_rate": 0.003, + "loss": 4.1001, + "step": 8058 + }, + { + "epoch": 0.08059, + "grad_norm": 0.7527191665858058, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 8059 + }, + { + "epoch": 0.0806, + "grad_norm": 0.8398808621055186, + "learning_rate": 0.003, + "loss": 4.0922, + "step": 8060 + }, + { + "epoch": 0.08061, + "grad_norm": 0.825612767978505, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 8061 + }, + { + "epoch": 0.08062, + "grad_norm": 0.9747793071523012, + "learning_rate": 0.003, + "loss": 4.11, + "step": 8062 + }, + { + "epoch": 0.08063, + "grad_norm": 1.1323711262853373, + "learning_rate": 0.003, + "loss": 4.1409, + "step": 8063 + }, + { + "epoch": 0.08064, + "grad_norm": 0.9414405685374689, + "learning_rate": 0.003, + "loss": 4.0977, + "step": 8064 + }, + { + "epoch": 0.08065, + "grad_norm": 0.877182385950829, + "learning_rate": 0.003, + "loss": 4.1089, + "step": 8065 + }, + { + "epoch": 0.08066, + "grad_norm": 0.9823013557480436, + "learning_rate": 0.003, + "loss": 4.1285, + "step": 8066 + }, + { + "epoch": 0.08067, + "grad_norm": 1.0076994124987724, + "learning_rate": 0.003, + "loss": 4.1172, + "step": 8067 + }, + { + "epoch": 0.08068, + "grad_norm": 0.9543931097825955, + "learning_rate": 0.003, + "loss": 4.1151, + "step": 8068 + }, + { + "epoch": 0.08069, + "grad_norm": 0.860639718383474, + "learning_rate": 0.003, + "loss": 4.1147, + "step": 8069 + }, + { + "epoch": 0.0807, + "grad_norm": 0.8988959057627193, + "learning_rate": 0.003, + "loss": 4.1014, + "step": 8070 + }, + { + "epoch": 0.08071, + "grad_norm": 0.9534116382487039, + "learning_rate": 0.003, + "loss": 4.1059, + "step": 8071 + }, + { + "epoch": 0.08072, + "grad_norm": 0.8660070014644025, + "learning_rate": 0.003, + "loss": 4.0918, + "step": 8072 + }, + { + "epoch": 0.08073, + "grad_norm": 0.8206611525482264, + "learning_rate": 0.003, + "loss": 4.0941, + "step": 8073 + }, + { + "epoch": 0.08074, + "grad_norm": 0.965813130304323, + "learning_rate": 0.003, + "loss": 4.084, + "step": 8074 + }, + { + "epoch": 0.08075, + "grad_norm": 1.1442526973314016, + "learning_rate": 0.003, + "loss": 4.1309, + "step": 8075 + }, + { + "epoch": 0.08076, + "grad_norm": 0.7971529185891382, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 8076 + }, + { + "epoch": 0.08077, + "grad_norm": 0.6441126050923031, + "learning_rate": 0.003, + "loss": 4.1244, + "step": 8077 + }, + { + "epoch": 0.08078, + "grad_norm": 0.711624279954515, + "learning_rate": 0.003, + "loss": 4.111, + "step": 8078 + }, + { + "epoch": 0.08079, + "grad_norm": 0.8468946223392866, + "learning_rate": 0.003, + "loss": 4.1025, + "step": 8079 + }, + { + "epoch": 0.0808, + "grad_norm": 0.955200727057764, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 8080 + }, + { + "epoch": 0.08081, + "grad_norm": 1.153055759299516, + "learning_rate": 0.003, + "loss": 4.1146, + "step": 8081 + }, + { + "epoch": 0.08082, + "grad_norm": 0.7784581084262547, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 8082 + }, + { + "epoch": 0.08083, + "grad_norm": 0.7994938331488759, + "learning_rate": 0.003, + "loss": 4.0935, + "step": 8083 + }, + { + "epoch": 0.08084, + "grad_norm": 0.7985218153568882, + "learning_rate": 0.003, + "loss": 4.0949, + "step": 8084 + }, + { + "epoch": 0.08085, + "grad_norm": 0.7650441422106858, + "learning_rate": 0.003, + "loss": 4.1222, + "step": 8085 + }, + { + "epoch": 0.08086, + "grad_norm": 0.9072526574743619, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 8086 + }, + { + "epoch": 0.08087, + "grad_norm": 1.0019552972566608, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 8087 + }, + { + "epoch": 0.08088, + "grad_norm": 1.0698186825083669, + "learning_rate": 0.003, + "loss": 4.1086, + "step": 8088 + }, + { + "epoch": 0.08089, + "grad_norm": 0.8487971751746695, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 8089 + }, + { + "epoch": 0.0809, + "grad_norm": 0.7375163925623175, + "learning_rate": 0.003, + "loss": 4.096, + "step": 8090 + }, + { + "epoch": 0.08091, + "grad_norm": 0.8077965976484879, + "learning_rate": 0.003, + "loss": 4.1169, + "step": 8091 + }, + { + "epoch": 0.08092, + "grad_norm": 0.8806655777433894, + "learning_rate": 0.003, + "loss": 4.118, + "step": 8092 + }, + { + "epoch": 0.08093, + "grad_norm": 0.8846095215059361, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 8093 + }, + { + "epoch": 0.08094, + "grad_norm": 1.0263143758049644, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 8094 + }, + { + "epoch": 0.08095, + "grad_norm": 0.9975182671338138, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 8095 + }, + { + "epoch": 0.08096, + "grad_norm": 0.9243866628279602, + "learning_rate": 0.003, + "loss": 4.1112, + "step": 8096 + }, + { + "epoch": 0.08097, + "grad_norm": 0.764690705974969, + "learning_rate": 0.003, + "loss": 4.0992, + "step": 8097 + }, + { + "epoch": 0.08098, + "grad_norm": 0.8123826720903898, + "learning_rate": 0.003, + "loss": 4.1283, + "step": 8098 + }, + { + "epoch": 0.08099, + "grad_norm": 0.8307382789621239, + "learning_rate": 0.003, + "loss": 4.1016, + "step": 8099 + }, + { + "epoch": 0.081, + "grad_norm": 0.683295591217391, + "learning_rate": 0.003, + "loss": 4.1022, + "step": 8100 + }, + { + "epoch": 0.08101, + "grad_norm": 0.762072793349242, + "learning_rate": 0.003, + "loss": 4.1253, + "step": 8101 + }, + { + "epoch": 0.08102, + "grad_norm": 0.8417363332108121, + "learning_rate": 0.003, + "loss": 4.1029, + "step": 8102 + }, + { + "epoch": 0.08103, + "grad_norm": 1.0026749696403685, + "learning_rate": 0.003, + "loss": 4.1107, + "step": 8103 + }, + { + "epoch": 0.08104, + "grad_norm": 1.3055013665970854, + "learning_rate": 0.003, + "loss": 4.1224, + "step": 8104 + }, + { + "epoch": 0.08105, + "grad_norm": 0.7284350470055041, + "learning_rate": 0.003, + "loss": 4.1211, + "step": 8105 + }, + { + "epoch": 0.08106, + "grad_norm": 0.7042488841506047, + "learning_rate": 0.003, + "loss": 4.1078, + "step": 8106 + }, + { + "epoch": 0.08107, + "grad_norm": 0.6662191545969933, + "learning_rate": 0.003, + "loss": 4.1052, + "step": 8107 + }, + { + "epoch": 0.08108, + "grad_norm": 0.7428305715535388, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 8108 + }, + { + "epoch": 0.08109, + "grad_norm": 0.7934277909736256, + "learning_rate": 0.003, + "loss": 4.0994, + "step": 8109 + }, + { + "epoch": 0.0811, + "grad_norm": 0.8818863948526912, + "learning_rate": 0.003, + "loss": 4.1196, + "step": 8110 + }, + { + "epoch": 0.08111, + "grad_norm": 1.0271099724103216, + "learning_rate": 0.003, + "loss": 4.115, + "step": 8111 + }, + { + "epoch": 0.08112, + "grad_norm": 1.2432013981414785, + "learning_rate": 0.003, + "loss": 4.0916, + "step": 8112 + }, + { + "epoch": 0.08113, + "grad_norm": 0.8995583170658074, + "learning_rate": 0.003, + "loss": 4.101, + "step": 8113 + }, + { + "epoch": 0.08114, + "grad_norm": 1.0954559019247962, + "learning_rate": 0.003, + "loss": 4.1321, + "step": 8114 + }, + { + "epoch": 0.08115, + "grad_norm": 1.261265478089719, + "learning_rate": 0.003, + "loss": 4.1249, + "step": 8115 + }, + { + "epoch": 0.08116, + "grad_norm": 0.8815832900165901, + "learning_rate": 0.003, + "loss": 4.1289, + "step": 8116 + }, + { + "epoch": 0.08117, + "grad_norm": 0.9087111097091439, + "learning_rate": 0.003, + "loss": 4.1278, + "step": 8117 + }, + { + "epoch": 0.08118, + "grad_norm": 1.010760403401548, + "learning_rate": 0.003, + "loss": 4.1141, + "step": 8118 + }, + { + "epoch": 0.08119, + "grad_norm": 1.056578294217066, + "learning_rate": 0.003, + "loss": 4.1093, + "step": 8119 + }, + { + "epoch": 0.0812, + "grad_norm": 0.9182887201534551, + "learning_rate": 0.003, + "loss": 4.0887, + "step": 8120 + }, + { + "epoch": 0.08121, + "grad_norm": 0.9041701432881559, + "learning_rate": 0.003, + "loss": 4.1182, + "step": 8121 + }, + { + "epoch": 0.08122, + "grad_norm": 0.8561224947474484, + "learning_rate": 0.003, + "loss": 4.1108, + "step": 8122 + }, + { + "epoch": 0.08123, + "grad_norm": 0.7614632039050264, + "learning_rate": 0.003, + "loss": 4.1016, + "step": 8123 + }, + { + "epoch": 0.08124, + "grad_norm": 0.7907531416628623, + "learning_rate": 0.003, + "loss": 4.1148, + "step": 8124 + }, + { + "epoch": 0.08125, + "grad_norm": 0.7300964775523906, + "learning_rate": 0.003, + "loss": 4.1313, + "step": 8125 + }, + { + "epoch": 0.08126, + "grad_norm": 0.827569261113533, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 8126 + }, + { + "epoch": 0.08127, + "grad_norm": 1.0339006016723764, + "learning_rate": 0.003, + "loss": 4.0945, + "step": 8127 + }, + { + "epoch": 0.08128, + "grad_norm": 1.2837248118220423, + "learning_rate": 0.003, + "loss": 4.1243, + "step": 8128 + }, + { + "epoch": 0.08129, + "grad_norm": 0.831463644855417, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 8129 + }, + { + "epoch": 0.0813, + "grad_norm": 0.7313691024451621, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 8130 + }, + { + "epoch": 0.08131, + "grad_norm": 0.7449363031040399, + "learning_rate": 0.003, + "loss": 4.1078, + "step": 8131 + }, + { + "epoch": 0.08132, + "grad_norm": 0.8923893884805761, + "learning_rate": 0.003, + "loss": 4.1007, + "step": 8132 + }, + { + "epoch": 0.08133, + "grad_norm": 1.00858302579857, + "learning_rate": 0.003, + "loss": 4.0976, + "step": 8133 + }, + { + "epoch": 0.08134, + "grad_norm": 1.0839591261263513, + "learning_rate": 0.003, + "loss": 4.0935, + "step": 8134 + }, + { + "epoch": 0.08135, + "grad_norm": 0.9551176992279036, + "learning_rate": 0.003, + "loss": 4.0953, + "step": 8135 + }, + { + "epoch": 0.08136, + "grad_norm": 1.1427717477515151, + "learning_rate": 0.003, + "loss": 4.1348, + "step": 8136 + }, + { + "epoch": 0.08137, + "grad_norm": 1.1608247537070675, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 8137 + }, + { + "epoch": 0.08138, + "grad_norm": 0.8036743748727992, + "learning_rate": 0.003, + "loss": 4.0979, + "step": 8138 + }, + { + "epoch": 0.08139, + "grad_norm": 0.7123613175658509, + "learning_rate": 0.003, + "loss": 4.1044, + "step": 8139 + }, + { + "epoch": 0.0814, + "grad_norm": 0.7693281819705117, + "learning_rate": 0.003, + "loss": 4.1192, + "step": 8140 + }, + { + "epoch": 0.08141, + "grad_norm": 0.6727495096734566, + "learning_rate": 0.003, + "loss": 4.1112, + "step": 8141 + }, + { + "epoch": 0.08142, + "grad_norm": 0.6845081706658788, + "learning_rate": 0.003, + "loss": 4.1214, + "step": 8142 + }, + { + "epoch": 0.08143, + "grad_norm": 0.7460442553490216, + "learning_rate": 0.003, + "loss": 4.1158, + "step": 8143 + }, + { + "epoch": 0.08144, + "grad_norm": 0.8245361382413854, + "learning_rate": 0.003, + "loss": 4.0854, + "step": 8144 + }, + { + "epoch": 0.08145, + "grad_norm": 0.8230934857490788, + "learning_rate": 0.003, + "loss": 4.0918, + "step": 8145 + }, + { + "epoch": 0.08146, + "grad_norm": 0.7712994219712946, + "learning_rate": 0.003, + "loss": 4.1229, + "step": 8146 + }, + { + "epoch": 0.08147, + "grad_norm": 0.7460865071832761, + "learning_rate": 0.003, + "loss": 4.1233, + "step": 8147 + }, + { + "epoch": 0.08148, + "grad_norm": 0.7735141713283167, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 8148 + }, + { + "epoch": 0.08149, + "grad_norm": 0.7896814609713348, + "learning_rate": 0.003, + "loss": 4.088, + "step": 8149 + }, + { + "epoch": 0.0815, + "grad_norm": 0.7747886996009599, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 8150 + }, + { + "epoch": 0.08151, + "grad_norm": 0.8866325001517937, + "learning_rate": 0.003, + "loss": 4.109, + "step": 8151 + }, + { + "epoch": 0.08152, + "grad_norm": 0.9586498492742512, + "learning_rate": 0.003, + "loss": 4.1149, + "step": 8152 + }, + { + "epoch": 0.08153, + "grad_norm": 0.8759445141936913, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 8153 + }, + { + "epoch": 0.08154, + "grad_norm": 0.8213012411475468, + "learning_rate": 0.003, + "loss": 4.104, + "step": 8154 + }, + { + "epoch": 0.08155, + "grad_norm": 0.7343561260089444, + "learning_rate": 0.003, + "loss": 4.1183, + "step": 8155 + }, + { + "epoch": 0.08156, + "grad_norm": 0.7055042559161329, + "learning_rate": 0.003, + "loss": 4.1006, + "step": 8156 + }, + { + "epoch": 0.08157, + "grad_norm": 0.6906999012739224, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 8157 + }, + { + "epoch": 0.08158, + "grad_norm": 0.6797037737186805, + "learning_rate": 0.003, + "loss": 4.1267, + "step": 8158 + }, + { + "epoch": 0.08159, + "grad_norm": 0.6440815279956414, + "learning_rate": 0.003, + "loss": 4.1222, + "step": 8159 + }, + { + "epoch": 0.0816, + "grad_norm": 0.6557581564589431, + "learning_rate": 0.003, + "loss": 4.0944, + "step": 8160 + }, + { + "epoch": 0.08161, + "grad_norm": 0.6961649850170736, + "learning_rate": 0.003, + "loss": 4.1012, + "step": 8161 + }, + { + "epoch": 0.08162, + "grad_norm": 0.7152753880484393, + "learning_rate": 0.003, + "loss": 4.1055, + "step": 8162 + }, + { + "epoch": 0.08163, + "grad_norm": 0.7444333284957504, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 8163 + }, + { + "epoch": 0.08164, + "grad_norm": 0.8297049679085736, + "learning_rate": 0.003, + "loss": 4.0913, + "step": 8164 + }, + { + "epoch": 0.08165, + "grad_norm": 0.9254543973261257, + "learning_rate": 0.003, + "loss": 4.0958, + "step": 8165 + }, + { + "epoch": 0.08166, + "grad_norm": 1.135495751545436, + "learning_rate": 0.003, + "loss": 4.1029, + "step": 8166 + }, + { + "epoch": 0.08167, + "grad_norm": 1.003264644227776, + "learning_rate": 0.003, + "loss": 4.1002, + "step": 8167 + }, + { + "epoch": 0.08168, + "grad_norm": 1.3151569568484487, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 8168 + }, + { + "epoch": 0.08169, + "grad_norm": 0.7417217462858907, + "learning_rate": 0.003, + "loss": 4.1065, + "step": 8169 + }, + { + "epoch": 0.0817, + "grad_norm": 0.6967246388224767, + "learning_rate": 0.003, + "loss": 4.1001, + "step": 8170 + }, + { + "epoch": 0.08171, + "grad_norm": 0.8159574604205326, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 8171 + }, + { + "epoch": 0.08172, + "grad_norm": 0.8338929792539445, + "learning_rate": 0.003, + "loss": 4.1141, + "step": 8172 + }, + { + "epoch": 0.08173, + "grad_norm": 0.98094890794689, + "learning_rate": 0.003, + "loss": 4.141, + "step": 8173 + }, + { + "epoch": 0.08174, + "grad_norm": 1.0004268057358918, + "learning_rate": 0.003, + "loss": 4.1362, + "step": 8174 + }, + { + "epoch": 0.08175, + "grad_norm": 1.1607043587315327, + "learning_rate": 0.003, + "loss": 4.105, + "step": 8175 + }, + { + "epoch": 0.08176, + "grad_norm": 0.9605737836969909, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 8176 + }, + { + "epoch": 0.08177, + "grad_norm": 0.8834558793913779, + "learning_rate": 0.003, + "loss": 4.0903, + "step": 8177 + }, + { + "epoch": 0.08178, + "grad_norm": 0.9463401693916579, + "learning_rate": 0.003, + "loss": 4.1009, + "step": 8178 + }, + { + "epoch": 0.08179, + "grad_norm": 0.9414403348160015, + "learning_rate": 0.003, + "loss": 4.1263, + "step": 8179 + }, + { + "epoch": 0.0818, + "grad_norm": 0.9984246787560999, + "learning_rate": 0.003, + "loss": 4.1026, + "step": 8180 + }, + { + "epoch": 0.08181, + "grad_norm": 1.1775258869822263, + "learning_rate": 0.003, + "loss": 4.0968, + "step": 8181 + }, + { + "epoch": 0.08182, + "grad_norm": 0.8435168559607914, + "learning_rate": 0.003, + "loss": 4.1339, + "step": 8182 + }, + { + "epoch": 0.08183, + "grad_norm": 0.8045360298546768, + "learning_rate": 0.003, + "loss": 4.0901, + "step": 8183 + }, + { + "epoch": 0.08184, + "grad_norm": 0.7940871155944753, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 8184 + }, + { + "epoch": 0.08185, + "grad_norm": 0.7578411281597589, + "learning_rate": 0.003, + "loss": 4.1063, + "step": 8185 + }, + { + "epoch": 0.08186, + "grad_norm": 0.7267884989595176, + "learning_rate": 0.003, + "loss": 4.1003, + "step": 8186 + }, + { + "epoch": 0.08187, + "grad_norm": 0.9028879506424374, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 8187 + }, + { + "epoch": 0.08188, + "grad_norm": 1.0836114502579879, + "learning_rate": 0.003, + "loss": 4.1278, + "step": 8188 + }, + { + "epoch": 0.08189, + "grad_norm": 1.2100347103121765, + "learning_rate": 0.003, + "loss": 4.1286, + "step": 8189 + }, + { + "epoch": 0.0819, + "grad_norm": 0.7309510071375395, + "learning_rate": 0.003, + "loss": 4.0928, + "step": 8190 + }, + { + "epoch": 0.08191, + "grad_norm": 0.6311549475723786, + "learning_rate": 0.003, + "loss": 4.0938, + "step": 8191 + }, + { + "epoch": 0.08192, + "grad_norm": 0.7435668383560015, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 8192 + }, + { + "epoch": 0.08193, + "grad_norm": 0.8580955152495726, + "learning_rate": 0.003, + "loss": 4.1014, + "step": 8193 + }, + { + "epoch": 0.08194, + "grad_norm": 0.8737720108493068, + "learning_rate": 0.003, + "loss": 4.1122, + "step": 8194 + }, + { + "epoch": 0.08195, + "grad_norm": 0.7368558728662361, + "learning_rate": 0.003, + "loss": 4.0903, + "step": 8195 + }, + { + "epoch": 0.08196, + "grad_norm": 0.648364175407553, + "learning_rate": 0.003, + "loss": 4.1175, + "step": 8196 + }, + { + "epoch": 0.08197, + "grad_norm": 0.7511314273266356, + "learning_rate": 0.003, + "loss": 4.1014, + "step": 8197 + }, + { + "epoch": 0.08198, + "grad_norm": 0.9036375613694632, + "learning_rate": 0.003, + "loss": 4.0962, + "step": 8198 + }, + { + "epoch": 0.08199, + "grad_norm": 1.0500737128901654, + "learning_rate": 0.003, + "loss": 4.112, + "step": 8199 + }, + { + "epoch": 0.082, + "grad_norm": 1.1226888117313103, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 8200 + }, + { + "epoch": 0.08201, + "grad_norm": 0.9559312403994046, + "learning_rate": 0.003, + "loss": 4.0925, + "step": 8201 + }, + { + "epoch": 0.08202, + "grad_norm": 0.9864224976849971, + "learning_rate": 0.003, + "loss": 4.0857, + "step": 8202 + }, + { + "epoch": 0.08203, + "grad_norm": 1.0665254798248887, + "learning_rate": 0.003, + "loss": 4.1128, + "step": 8203 + }, + { + "epoch": 0.08204, + "grad_norm": 0.9332708500174354, + "learning_rate": 0.003, + "loss": 4.0993, + "step": 8204 + }, + { + "epoch": 0.08205, + "grad_norm": 0.9114402954224472, + "learning_rate": 0.003, + "loss": 4.095, + "step": 8205 + }, + { + "epoch": 0.08206, + "grad_norm": 0.9578612535805147, + "learning_rate": 0.003, + "loss": 4.1306, + "step": 8206 + }, + { + "epoch": 0.08207, + "grad_norm": 0.8491904917863199, + "learning_rate": 0.003, + "loss": 4.1012, + "step": 8207 + }, + { + "epoch": 0.08208, + "grad_norm": 0.6965883444740882, + "learning_rate": 0.003, + "loss": 4.0969, + "step": 8208 + }, + { + "epoch": 0.08209, + "grad_norm": 0.7953053984407373, + "learning_rate": 0.003, + "loss": 4.1146, + "step": 8209 + }, + { + "epoch": 0.0821, + "grad_norm": 0.7485874491455442, + "learning_rate": 0.003, + "loss": 4.0998, + "step": 8210 + }, + { + "epoch": 0.08211, + "grad_norm": 0.6959203077939378, + "learning_rate": 0.003, + "loss": 4.0982, + "step": 8211 + }, + { + "epoch": 0.08212, + "grad_norm": 0.8449995767744634, + "learning_rate": 0.003, + "loss": 4.0973, + "step": 8212 + }, + { + "epoch": 0.08213, + "grad_norm": 1.1015900099708935, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 8213 + }, + { + "epoch": 0.08214, + "grad_norm": 1.0315791177365499, + "learning_rate": 0.003, + "loss": 4.1219, + "step": 8214 + }, + { + "epoch": 0.08215, + "grad_norm": 0.9380557040904997, + "learning_rate": 0.003, + "loss": 4.1229, + "step": 8215 + }, + { + "epoch": 0.08216, + "grad_norm": 0.8923877134844612, + "learning_rate": 0.003, + "loss": 4.1007, + "step": 8216 + }, + { + "epoch": 0.08217, + "grad_norm": 0.8575588533424483, + "learning_rate": 0.003, + "loss": 4.1298, + "step": 8217 + }, + { + "epoch": 0.08218, + "grad_norm": 0.8232596706369313, + "learning_rate": 0.003, + "loss": 4.0913, + "step": 8218 + }, + { + "epoch": 0.08219, + "grad_norm": 0.9081979548545349, + "learning_rate": 0.003, + "loss": 4.091, + "step": 8219 + }, + { + "epoch": 0.0822, + "grad_norm": 0.8155798907038242, + "learning_rate": 0.003, + "loss": 4.0994, + "step": 8220 + }, + { + "epoch": 0.08221, + "grad_norm": 0.8942358339851308, + "learning_rate": 0.003, + "loss": 4.1209, + "step": 8221 + }, + { + "epoch": 0.08222, + "grad_norm": 1.1787055400411643, + "learning_rate": 0.003, + "loss": 4.1132, + "step": 8222 + }, + { + "epoch": 0.08223, + "grad_norm": 0.9509702723098798, + "learning_rate": 0.003, + "loss": 4.1173, + "step": 8223 + }, + { + "epoch": 0.08224, + "grad_norm": 0.888798577292018, + "learning_rate": 0.003, + "loss": 4.1138, + "step": 8224 + }, + { + "epoch": 0.08225, + "grad_norm": 0.8859713724373705, + "learning_rate": 0.003, + "loss": 4.1253, + "step": 8225 + }, + { + "epoch": 0.08226, + "grad_norm": 0.9453256114459335, + "learning_rate": 0.003, + "loss": 4.0908, + "step": 8226 + }, + { + "epoch": 0.08227, + "grad_norm": 1.0285023285363262, + "learning_rate": 0.003, + "loss": 4.1358, + "step": 8227 + }, + { + "epoch": 0.08228, + "grad_norm": 1.0441982441165125, + "learning_rate": 0.003, + "loss": 4.1232, + "step": 8228 + }, + { + "epoch": 0.08229, + "grad_norm": 0.9285745096914537, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 8229 + }, + { + "epoch": 0.0823, + "grad_norm": 0.8612854507864894, + "learning_rate": 0.003, + "loss": 4.1279, + "step": 8230 + }, + { + "epoch": 0.08231, + "grad_norm": 0.9539626109153166, + "learning_rate": 0.003, + "loss": 4.083, + "step": 8231 + }, + { + "epoch": 0.08232, + "grad_norm": 1.1127495391486235, + "learning_rate": 0.003, + "loss": 4.0966, + "step": 8232 + }, + { + "epoch": 0.08233, + "grad_norm": 0.9309241874844985, + "learning_rate": 0.003, + "loss": 4.1035, + "step": 8233 + }, + { + "epoch": 0.08234, + "grad_norm": 0.8586480291175957, + "learning_rate": 0.003, + "loss": 4.1162, + "step": 8234 + }, + { + "epoch": 0.08235, + "grad_norm": 0.8943116440399151, + "learning_rate": 0.003, + "loss": 4.136, + "step": 8235 + }, + { + "epoch": 0.08236, + "grad_norm": 0.8765191739384368, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 8236 + }, + { + "epoch": 0.08237, + "grad_norm": 0.8771079726635557, + "learning_rate": 0.003, + "loss": 4.1005, + "step": 8237 + }, + { + "epoch": 0.08238, + "grad_norm": 0.983672922309625, + "learning_rate": 0.003, + "loss": 4.1149, + "step": 8238 + }, + { + "epoch": 0.08239, + "grad_norm": 1.0592369601456073, + "learning_rate": 0.003, + "loss": 4.0891, + "step": 8239 + }, + { + "epoch": 0.0824, + "grad_norm": 1.0147386504894589, + "learning_rate": 0.003, + "loss": 4.1256, + "step": 8240 + }, + { + "epoch": 0.08241, + "grad_norm": 0.8937176926338467, + "learning_rate": 0.003, + "loss": 4.1036, + "step": 8241 + }, + { + "epoch": 0.08242, + "grad_norm": 0.8211844129412678, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 8242 + }, + { + "epoch": 0.08243, + "grad_norm": 0.941400186703893, + "learning_rate": 0.003, + "loss": 4.1256, + "step": 8243 + }, + { + "epoch": 0.08244, + "grad_norm": 1.147358753730803, + "learning_rate": 0.003, + "loss": 4.1014, + "step": 8244 + }, + { + "epoch": 0.08245, + "grad_norm": 1.0164893306751752, + "learning_rate": 0.003, + "loss": 4.0998, + "step": 8245 + }, + { + "epoch": 0.08246, + "grad_norm": 0.8548549249810995, + "learning_rate": 0.003, + "loss": 4.0985, + "step": 8246 + }, + { + "epoch": 0.08247, + "grad_norm": 0.7706214458931752, + "learning_rate": 0.003, + "loss": 4.1073, + "step": 8247 + }, + { + "epoch": 0.08248, + "grad_norm": 0.8475424364775245, + "learning_rate": 0.003, + "loss": 4.0948, + "step": 8248 + }, + { + "epoch": 0.08249, + "grad_norm": 0.8469414618345413, + "learning_rate": 0.003, + "loss": 4.1146, + "step": 8249 + }, + { + "epoch": 0.0825, + "grad_norm": 0.8229434139477997, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 8250 + }, + { + "epoch": 0.08251, + "grad_norm": 0.8011704320030776, + "learning_rate": 0.003, + "loss": 4.1089, + "step": 8251 + }, + { + "epoch": 0.08252, + "grad_norm": 0.6866093619498361, + "learning_rate": 0.003, + "loss": 4.1101, + "step": 8252 + }, + { + "epoch": 0.08253, + "grad_norm": 0.7062240053088432, + "learning_rate": 0.003, + "loss": 4.1009, + "step": 8253 + }, + { + "epoch": 0.08254, + "grad_norm": 0.6263882996334876, + "learning_rate": 0.003, + "loss": 4.079, + "step": 8254 + }, + { + "epoch": 0.08255, + "grad_norm": 0.6406212013874797, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 8255 + }, + { + "epoch": 0.08256, + "grad_norm": 0.6332253598165484, + "learning_rate": 0.003, + "loss": 4.1104, + "step": 8256 + }, + { + "epoch": 0.08257, + "grad_norm": 0.7298881894270829, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 8257 + }, + { + "epoch": 0.08258, + "grad_norm": 0.8061858783513308, + "learning_rate": 0.003, + "loss": 4.0887, + "step": 8258 + }, + { + "epoch": 0.08259, + "grad_norm": 0.9631605130652663, + "learning_rate": 0.003, + "loss": 4.1326, + "step": 8259 + }, + { + "epoch": 0.0826, + "grad_norm": 1.312679616746302, + "learning_rate": 0.003, + "loss": 4.0868, + "step": 8260 + }, + { + "epoch": 0.08261, + "grad_norm": 0.7878732930444392, + "learning_rate": 0.003, + "loss": 4.0916, + "step": 8261 + }, + { + "epoch": 0.08262, + "grad_norm": 0.7074636050563899, + "learning_rate": 0.003, + "loss": 4.1289, + "step": 8262 + }, + { + "epoch": 0.08263, + "grad_norm": 0.6997390004949057, + "learning_rate": 0.003, + "loss": 4.0925, + "step": 8263 + }, + { + "epoch": 0.08264, + "grad_norm": 0.772169379776631, + "learning_rate": 0.003, + "loss": 4.1065, + "step": 8264 + }, + { + "epoch": 0.08265, + "grad_norm": 0.7993877169518088, + "learning_rate": 0.003, + "loss": 4.0952, + "step": 8265 + }, + { + "epoch": 0.08266, + "grad_norm": 1.0950952503503433, + "learning_rate": 0.003, + "loss": 4.105, + "step": 8266 + }, + { + "epoch": 0.08267, + "grad_norm": 1.1929897335482988, + "learning_rate": 0.003, + "loss": 4.1079, + "step": 8267 + }, + { + "epoch": 0.08268, + "grad_norm": 0.8778501205471867, + "learning_rate": 0.003, + "loss": 4.0854, + "step": 8268 + }, + { + "epoch": 0.08269, + "grad_norm": 0.892710429777809, + "learning_rate": 0.003, + "loss": 4.1214, + "step": 8269 + }, + { + "epoch": 0.0827, + "grad_norm": 1.0149545588000088, + "learning_rate": 0.003, + "loss": 4.1018, + "step": 8270 + }, + { + "epoch": 0.08271, + "grad_norm": 1.124408719720435, + "learning_rate": 0.003, + "loss": 4.1285, + "step": 8271 + }, + { + "epoch": 0.08272, + "grad_norm": 0.8266826354793325, + "learning_rate": 0.003, + "loss": 4.1071, + "step": 8272 + }, + { + "epoch": 0.08273, + "grad_norm": 1.0123352772430743, + "learning_rate": 0.003, + "loss": 4.1082, + "step": 8273 + }, + { + "epoch": 0.08274, + "grad_norm": 1.0192036235033302, + "learning_rate": 0.003, + "loss": 4.1246, + "step": 8274 + }, + { + "epoch": 0.08275, + "grad_norm": 0.9102355182979404, + "learning_rate": 0.003, + "loss": 4.0994, + "step": 8275 + }, + { + "epoch": 0.08276, + "grad_norm": 0.7090653355389134, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 8276 + }, + { + "epoch": 0.08277, + "grad_norm": 0.6881786586872917, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 8277 + }, + { + "epoch": 0.08278, + "grad_norm": 0.7699897377296708, + "learning_rate": 0.003, + "loss": 4.0967, + "step": 8278 + }, + { + "epoch": 0.08279, + "grad_norm": 0.8358151974396222, + "learning_rate": 0.003, + "loss": 4.0957, + "step": 8279 + }, + { + "epoch": 0.0828, + "grad_norm": 0.9018628671728572, + "learning_rate": 0.003, + "loss": 4.1249, + "step": 8280 + }, + { + "epoch": 0.08281, + "grad_norm": 1.1343593814872228, + "learning_rate": 0.003, + "loss": 4.089, + "step": 8281 + }, + { + "epoch": 0.08282, + "grad_norm": 0.9781923188910686, + "learning_rate": 0.003, + "loss": 4.0923, + "step": 8282 + }, + { + "epoch": 0.08283, + "grad_norm": 0.9927329287124252, + "learning_rate": 0.003, + "loss": 4.1112, + "step": 8283 + }, + { + "epoch": 0.08284, + "grad_norm": 0.98118308584172, + "learning_rate": 0.003, + "loss": 4.0933, + "step": 8284 + }, + { + "epoch": 0.08285, + "grad_norm": 0.9263465527068231, + "learning_rate": 0.003, + "loss": 4.1095, + "step": 8285 + }, + { + "epoch": 0.08286, + "grad_norm": 0.8673110587039227, + "learning_rate": 0.003, + "loss": 4.1219, + "step": 8286 + }, + { + "epoch": 0.08287, + "grad_norm": 0.8419508037976696, + "learning_rate": 0.003, + "loss": 4.0988, + "step": 8287 + }, + { + "epoch": 0.08288, + "grad_norm": 0.8220681636415317, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 8288 + }, + { + "epoch": 0.08289, + "grad_norm": 0.7505932152286665, + "learning_rate": 0.003, + "loss": 4.1028, + "step": 8289 + }, + { + "epoch": 0.0829, + "grad_norm": 0.8010371678478732, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 8290 + }, + { + "epoch": 0.08291, + "grad_norm": 0.7956640617851198, + "learning_rate": 0.003, + "loss": 4.091, + "step": 8291 + }, + { + "epoch": 0.08292, + "grad_norm": 0.9655748890689609, + "learning_rate": 0.003, + "loss": 4.0982, + "step": 8292 + }, + { + "epoch": 0.08293, + "grad_norm": 1.2599943864429868, + "learning_rate": 0.003, + "loss": 4.1187, + "step": 8293 + }, + { + "epoch": 0.08294, + "grad_norm": 0.9706360443217623, + "learning_rate": 0.003, + "loss": 4.0958, + "step": 8294 + }, + { + "epoch": 0.08295, + "grad_norm": 1.048320155699331, + "learning_rate": 0.003, + "loss": 4.0989, + "step": 8295 + }, + { + "epoch": 0.08296, + "grad_norm": 0.9595325746578597, + "learning_rate": 0.003, + "loss": 4.1017, + "step": 8296 + }, + { + "epoch": 0.08297, + "grad_norm": 0.9381047024976223, + "learning_rate": 0.003, + "loss": 4.112, + "step": 8297 + }, + { + "epoch": 0.08298, + "grad_norm": 0.863050580789861, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 8298 + }, + { + "epoch": 0.08299, + "grad_norm": 0.7240303327199505, + "learning_rate": 0.003, + "loss": 4.0939, + "step": 8299 + }, + { + "epoch": 0.083, + "grad_norm": 0.7823461180480966, + "learning_rate": 0.003, + "loss": 4.1312, + "step": 8300 + }, + { + "epoch": 0.08301, + "grad_norm": 0.8990430145286981, + "learning_rate": 0.003, + "loss": 4.1109, + "step": 8301 + }, + { + "epoch": 0.08302, + "grad_norm": 0.938888405100819, + "learning_rate": 0.003, + "loss": 4.1236, + "step": 8302 + }, + { + "epoch": 0.08303, + "grad_norm": 0.8734564423642468, + "learning_rate": 0.003, + "loss": 4.0935, + "step": 8303 + }, + { + "epoch": 0.08304, + "grad_norm": 0.8158703165239903, + "learning_rate": 0.003, + "loss": 4.1239, + "step": 8304 + }, + { + "epoch": 0.08305, + "grad_norm": 0.75620484767062, + "learning_rate": 0.003, + "loss": 4.0936, + "step": 8305 + }, + { + "epoch": 0.08306, + "grad_norm": 0.6540375187690641, + "learning_rate": 0.003, + "loss": 4.1012, + "step": 8306 + }, + { + "epoch": 0.08307, + "grad_norm": 0.7220457038710596, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 8307 + }, + { + "epoch": 0.08308, + "grad_norm": 0.7585842959907338, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 8308 + }, + { + "epoch": 0.08309, + "grad_norm": 0.7562956967912909, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 8309 + }, + { + "epoch": 0.0831, + "grad_norm": 0.8620656108784794, + "learning_rate": 0.003, + "loss": 4.0958, + "step": 8310 + }, + { + "epoch": 0.08311, + "grad_norm": 1.0392515900505281, + "learning_rate": 0.003, + "loss": 4.0976, + "step": 8311 + }, + { + "epoch": 0.08312, + "grad_norm": 1.1358410812232829, + "learning_rate": 0.003, + "loss": 4.1248, + "step": 8312 + }, + { + "epoch": 0.08313, + "grad_norm": 0.8856911622270947, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 8313 + }, + { + "epoch": 0.08314, + "grad_norm": 0.8971402484163674, + "learning_rate": 0.003, + "loss": 4.1155, + "step": 8314 + }, + { + "epoch": 0.08315, + "grad_norm": 0.9089314400577564, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 8315 + }, + { + "epoch": 0.08316, + "grad_norm": 0.9935520424178989, + "learning_rate": 0.003, + "loss": 4.095, + "step": 8316 + }, + { + "epoch": 0.08317, + "grad_norm": 0.9643025611664917, + "learning_rate": 0.003, + "loss": 4.0928, + "step": 8317 + }, + { + "epoch": 0.08318, + "grad_norm": 0.9787470048990922, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 8318 + }, + { + "epoch": 0.08319, + "grad_norm": 1.196041471896753, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 8319 + }, + { + "epoch": 0.0832, + "grad_norm": 0.7528163771935631, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 8320 + }, + { + "epoch": 0.08321, + "grad_norm": 0.6113115162451314, + "learning_rate": 0.003, + "loss": 4.1063, + "step": 8321 + }, + { + "epoch": 0.08322, + "grad_norm": 0.6302210737347822, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 8322 + }, + { + "epoch": 0.08323, + "grad_norm": 0.6846448753101897, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 8323 + }, + { + "epoch": 0.08324, + "grad_norm": 0.81282422542553, + "learning_rate": 0.003, + "loss": 4.0976, + "step": 8324 + }, + { + "epoch": 0.08325, + "grad_norm": 0.9233311436381315, + "learning_rate": 0.003, + "loss": 4.1047, + "step": 8325 + }, + { + "epoch": 0.08326, + "grad_norm": 0.954642418929674, + "learning_rate": 0.003, + "loss": 4.0934, + "step": 8326 + }, + { + "epoch": 0.08327, + "grad_norm": 0.8277691448246721, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 8327 + }, + { + "epoch": 0.08328, + "grad_norm": 0.7651088353198385, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 8328 + }, + { + "epoch": 0.08329, + "grad_norm": 0.8129905787749772, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 8329 + }, + { + "epoch": 0.0833, + "grad_norm": 0.7705105903193307, + "learning_rate": 0.003, + "loss": 4.0982, + "step": 8330 + }, + { + "epoch": 0.08331, + "grad_norm": 0.8783516141923371, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 8331 + }, + { + "epoch": 0.08332, + "grad_norm": 0.9500635724943229, + "learning_rate": 0.003, + "loss": 4.1073, + "step": 8332 + }, + { + "epoch": 0.08333, + "grad_norm": 1.0022282494956223, + "learning_rate": 0.003, + "loss": 4.1082, + "step": 8333 + }, + { + "epoch": 0.08334, + "grad_norm": 0.9868081102540992, + "learning_rate": 0.003, + "loss": 4.1067, + "step": 8334 + }, + { + "epoch": 0.08335, + "grad_norm": 0.9702541505884114, + "learning_rate": 0.003, + "loss": 4.0996, + "step": 8335 + }, + { + "epoch": 0.08336, + "grad_norm": 0.9616697241003944, + "learning_rate": 0.003, + "loss": 4.1271, + "step": 8336 + }, + { + "epoch": 0.08337, + "grad_norm": 0.9197392338204791, + "learning_rate": 0.003, + "loss": 4.1158, + "step": 8337 + }, + { + "epoch": 0.08338, + "grad_norm": 0.9953727673941519, + "learning_rate": 0.003, + "loss": 4.1025, + "step": 8338 + }, + { + "epoch": 0.08339, + "grad_norm": 1.0143098515036326, + "learning_rate": 0.003, + "loss": 4.1025, + "step": 8339 + }, + { + "epoch": 0.0834, + "grad_norm": 0.8294567785378771, + "learning_rate": 0.003, + "loss": 4.0963, + "step": 8340 + }, + { + "epoch": 0.08341, + "grad_norm": 0.6099348529213143, + "learning_rate": 0.003, + "loss": 4.1201, + "step": 8341 + }, + { + "epoch": 0.08342, + "grad_norm": 0.7224131427612864, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 8342 + }, + { + "epoch": 0.08343, + "grad_norm": 0.8439433910256732, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 8343 + }, + { + "epoch": 0.08344, + "grad_norm": 0.9868340222493676, + "learning_rate": 0.003, + "loss": 4.0964, + "step": 8344 + }, + { + "epoch": 0.08345, + "grad_norm": 0.9401831390198352, + "learning_rate": 0.003, + "loss": 4.0864, + "step": 8345 + }, + { + "epoch": 0.08346, + "grad_norm": 1.0302321487078914, + "learning_rate": 0.003, + "loss": 4.1087, + "step": 8346 + }, + { + "epoch": 0.08347, + "grad_norm": 0.8650226664698788, + "learning_rate": 0.003, + "loss": 4.1029, + "step": 8347 + }, + { + "epoch": 0.08348, + "grad_norm": 0.7651090535509265, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 8348 + }, + { + "epoch": 0.08349, + "grad_norm": 0.804167869373704, + "learning_rate": 0.003, + "loss": 4.1046, + "step": 8349 + }, + { + "epoch": 0.0835, + "grad_norm": 0.907923736734775, + "learning_rate": 0.003, + "loss": 4.1326, + "step": 8350 + }, + { + "epoch": 0.08351, + "grad_norm": 0.9166006948131097, + "learning_rate": 0.003, + "loss": 4.121, + "step": 8351 + }, + { + "epoch": 0.08352, + "grad_norm": 0.9742286942493177, + "learning_rate": 0.003, + "loss": 4.1174, + "step": 8352 + }, + { + "epoch": 0.08353, + "grad_norm": 1.0443086898250653, + "learning_rate": 0.003, + "loss": 4.1032, + "step": 8353 + }, + { + "epoch": 0.08354, + "grad_norm": 1.0482003041090164, + "learning_rate": 0.003, + "loss": 4.1117, + "step": 8354 + }, + { + "epoch": 0.08355, + "grad_norm": 0.9948528176450955, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 8355 + }, + { + "epoch": 0.08356, + "grad_norm": 0.9930403012130316, + "learning_rate": 0.003, + "loss": 4.1109, + "step": 8356 + }, + { + "epoch": 0.08357, + "grad_norm": 1.058290503425188, + "learning_rate": 0.003, + "loss": 4.12, + "step": 8357 + }, + { + "epoch": 0.08358, + "grad_norm": 0.7888495179404684, + "learning_rate": 0.003, + "loss": 4.1032, + "step": 8358 + }, + { + "epoch": 0.08359, + "grad_norm": 0.7764006655298166, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 8359 + }, + { + "epoch": 0.0836, + "grad_norm": 0.9269785078441903, + "learning_rate": 0.003, + "loss": 4.0881, + "step": 8360 + }, + { + "epoch": 0.08361, + "grad_norm": 1.0448000920569085, + "learning_rate": 0.003, + "loss": 4.1102, + "step": 8361 + }, + { + "epoch": 0.08362, + "grad_norm": 1.0677595443672623, + "learning_rate": 0.003, + "loss": 4.1246, + "step": 8362 + }, + { + "epoch": 0.08363, + "grad_norm": 0.9762108759601814, + "learning_rate": 0.003, + "loss": 4.1332, + "step": 8363 + }, + { + "epoch": 0.08364, + "grad_norm": 0.9250190436823729, + "learning_rate": 0.003, + "loss": 4.0939, + "step": 8364 + }, + { + "epoch": 0.08365, + "grad_norm": 0.6667015272636464, + "learning_rate": 0.003, + "loss": 4.1102, + "step": 8365 + }, + { + "epoch": 0.08366, + "grad_norm": 0.6845211638503894, + "learning_rate": 0.003, + "loss": 4.1198, + "step": 8366 + }, + { + "epoch": 0.08367, + "grad_norm": 0.5939518428147165, + "learning_rate": 0.003, + "loss": 4.092, + "step": 8367 + }, + { + "epoch": 0.08368, + "grad_norm": 0.5996368389783523, + "learning_rate": 0.003, + "loss": 4.1098, + "step": 8368 + }, + { + "epoch": 0.08369, + "grad_norm": 0.6523282835388826, + "learning_rate": 0.003, + "loss": 4.0961, + "step": 8369 + }, + { + "epoch": 0.0837, + "grad_norm": 0.8159208873071091, + "learning_rate": 0.003, + "loss": 4.0989, + "step": 8370 + }, + { + "epoch": 0.08371, + "grad_norm": 0.9860834398447139, + "learning_rate": 0.003, + "loss": 4.11, + "step": 8371 + }, + { + "epoch": 0.08372, + "grad_norm": 1.0940596921770058, + "learning_rate": 0.003, + "loss": 4.1339, + "step": 8372 + }, + { + "epoch": 0.08373, + "grad_norm": 0.9442033751036637, + "learning_rate": 0.003, + "loss": 4.1091, + "step": 8373 + }, + { + "epoch": 0.08374, + "grad_norm": 0.8663695859388787, + "learning_rate": 0.003, + "loss": 4.0856, + "step": 8374 + }, + { + "epoch": 0.08375, + "grad_norm": 0.7415517737245889, + "learning_rate": 0.003, + "loss": 4.087, + "step": 8375 + }, + { + "epoch": 0.08376, + "grad_norm": 0.8577754980749478, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 8376 + }, + { + "epoch": 0.08377, + "grad_norm": 1.0199862392503956, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 8377 + }, + { + "epoch": 0.08378, + "grad_norm": 1.2449008964290007, + "learning_rate": 0.003, + "loss": 4.1124, + "step": 8378 + }, + { + "epoch": 0.08379, + "grad_norm": 0.9419961904078284, + "learning_rate": 0.003, + "loss": 4.0992, + "step": 8379 + }, + { + "epoch": 0.0838, + "grad_norm": 0.9566083866982931, + "learning_rate": 0.003, + "loss": 4.1002, + "step": 8380 + }, + { + "epoch": 0.08381, + "grad_norm": 0.921682176354393, + "learning_rate": 0.003, + "loss": 4.1022, + "step": 8381 + }, + { + "epoch": 0.08382, + "grad_norm": 0.8446653628794919, + "learning_rate": 0.003, + "loss": 4.1172, + "step": 8382 + }, + { + "epoch": 0.08383, + "grad_norm": 0.817980279883183, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 8383 + }, + { + "epoch": 0.08384, + "grad_norm": 0.7858769081287543, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 8384 + }, + { + "epoch": 0.08385, + "grad_norm": 0.7428564824532166, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 8385 + }, + { + "epoch": 0.08386, + "grad_norm": 0.7507352447976503, + "learning_rate": 0.003, + "loss": 4.0891, + "step": 8386 + }, + { + "epoch": 0.08387, + "grad_norm": 0.8467430532110711, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 8387 + }, + { + "epoch": 0.08388, + "grad_norm": 0.9680594439287274, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 8388 + }, + { + "epoch": 0.08389, + "grad_norm": 1.2252522777923969, + "learning_rate": 0.003, + "loss": 4.1204, + "step": 8389 + }, + { + "epoch": 0.0839, + "grad_norm": 0.7533692382713882, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 8390 + }, + { + "epoch": 0.08391, + "grad_norm": 0.578854889997608, + "learning_rate": 0.003, + "loss": 4.108, + "step": 8391 + }, + { + "epoch": 0.08392, + "grad_norm": 0.7310406618044913, + "learning_rate": 0.003, + "loss": 4.078, + "step": 8392 + }, + { + "epoch": 0.08393, + "grad_norm": 0.8080163613542798, + "learning_rate": 0.003, + "loss": 4.1083, + "step": 8393 + }, + { + "epoch": 0.08394, + "grad_norm": 0.9313902145452904, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 8394 + }, + { + "epoch": 0.08395, + "grad_norm": 0.8774904667389832, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 8395 + }, + { + "epoch": 0.08396, + "grad_norm": 0.7336808643576519, + "learning_rate": 0.003, + "loss": 4.1099, + "step": 8396 + }, + { + "epoch": 0.08397, + "grad_norm": 0.6526993931050207, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 8397 + }, + { + "epoch": 0.08398, + "grad_norm": 0.6704748477002462, + "learning_rate": 0.003, + "loss": 4.0987, + "step": 8398 + }, + { + "epoch": 0.08399, + "grad_norm": 0.7009862189779927, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 8399 + }, + { + "epoch": 0.084, + "grad_norm": 0.7959197464027524, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 8400 + }, + { + "epoch": 0.08401, + "grad_norm": 0.7942806861956023, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 8401 + }, + { + "epoch": 0.08402, + "grad_norm": 0.7901087516905125, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 8402 + }, + { + "epoch": 0.08403, + "grad_norm": 0.7878285742447195, + "learning_rate": 0.003, + "loss": 4.0848, + "step": 8403 + }, + { + "epoch": 0.08404, + "grad_norm": 0.760912884794821, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 8404 + }, + { + "epoch": 0.08405, + "grad_norm": 0.9006130071518619, + "learning_rate": 0.003, + "loss": 4.077, + "step": 8405 + }, + { + "epoch": 0.08406, + "grad_norm": 1.0016144008572414, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 8406 + }, + { + "epoch": 0.08407, + "grad_norm": 1.3443814703857433, + "learning_rate": 0.003, + "loss": 4.1531, + "step": 8407 + }, + { + "epoch": 0.08408, + "grad_norm": 0.9056353298369414, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 8408 + }, + { + "epoch": 0.08409, + "grad_norm": 0.8063158095511546, + "learning_rate": 0.003, + "loss": 4.111, + "step": 8409 + }, + { + "epoch": 0.0841, + "grad_norm": 0.9219305008449818, + "learning_rate": 0.003, + "loss": 4.1093, + "step": 8410 + }, + { + "epoch": 0.08411, + "grad_norm": 1.1026094058352998, + "learning_rate": 0.003, + "loss": 4.1366, + "step": 8411 + }, + { + "epoch": 0.08412, + "grad_norm": 0.9534553488792078, + "learning_rate": 0.003, + "loss": 4.114, + "step": 8412 + }, + { + "epoch": 0.08413, + "grad_norm": 0.942913373390524, + "learning_rate": 0.003, + "loss": 4.074, + "step": 8413 + }, + { + "epoch": 0.08414, + "grad_norm": 1.0244828961575672, + "learning_rate": 0.003, + "loss": 4.1186, + "step": 8414 + }, + { + "epoch": 0.08415, + "grad_norm": 0.9400339571658545, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 8415 + }, + { + "epoch": 0.08416, + "grad_norm": 0.8710315727738005, + "learning_rate": 0.003, + "loss": 4.1112, + "step": 8416 + }, + { + "epoch": 0.08417, + "grad_norm": 0.7950437016706896, + "learning_rate": 0.003, + "loss": 4.1064, + "step": 8417 + }, + { + "epoch": 0.08418, + "grad_norm": 0.8098581252447725, + "learning_rate": 0.003, + "loss": 4.0923, + "step": 8418 + }, + { + "epoch": 0.08419, + "grad_norm": 0.9057985377788469, + "learning_rate": 0.003, + "loss": 4.1036, + "step": 8419 + }, + { + "epoch": 0.0842, + "grad_norm": 0.9778312033623336, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 8420 + }, + { + "epoch": 0.08421, + "grad_norm": 0.8982644728463484, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 8421 + }, + { + "epoch": 0.08422, + "grad_norm": 0.9376171311714971, + "learning_rate": 0.003, + "loss": 4.1191, + "step": 8422 + }, + { + "epoch": 0.08423, + "grad_norm": 0.9399936667089988, + "learning_rate": 0.003, + "loss": 4.1017, + "step": 8423 + }, + { + "epoch": 0.08424, + "grad_norm": 0.8765907774353715, + "learning_rate": 0.003, + "loss": 4.0948, + "step": 8424 + }, + { + "epoch": 0.08425, + "grad_norm": 0.8093336331696083, + "learning_rate": 0.003, + "loss": 4.1123, + "step": 8425 + }, + { + "epoch": 0.08426, + "grad_norm": 0.7177444716043226, + "learning_rate": 0.003, + "loss": 4.092, + "step": 8426 + }, + { + "epoch": 0.08427, + "grad_norm": 0.8420947027796308, + "learning_rate": 0.003, + "loss": 4.1152, + "step": 8427 + }, + { + "epoch": 0.08428, + "grad_norm": 1.0570206327016962, + "learning_rate": 0.003, + "loss": 4.103, + "step": 8428 + }, + { + "epoch": 0.08429, + "grad_norm": 1.0693132771916776, + "learning_rate": 0.003, + "loss": 4.0961, + "step": 8429 + }, + { + "epoch": 0.0843, + "grad_norm": 1.037364188618762, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 8430 + }, + { + "epoch": 0.08431, + "grad_norm": 0.9849908084466207, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 8431 + }, + { + "epoch": 0.08432, + "grad_norm": 0.9291586130000788, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 8432 + }, + { + "epoch": 0.08433, + "grad_norm": 0.865132194731464, + "learning_rate": 0.003, + "loss": 4.1126, + "step": 8433 + }, + { + "epoch": 0.08434, + "grad_norm": 0.8122049399169294, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 8434 + }, + { + "epoch": 0.08435, + "grad_norm": 0.747566312505787, + "learning_rate": 0.003, + "loss": 4.0972, + "step": 8435 + }, + { + "epoch": 0.08436, + "grad_norm": 0.7617380609411367, + "learning_rate": 0.003, + "loss": 4.0959, + "step": 8436 + }, + { + "epoch": 0.08437, + "grad_norm": 0.7640674198661149, + "learning_rate": 0.003, + "loss": 4.108, + "step": 8437 + }, + { + "epoch": 0.08438, + "grad_norm": 0.7897561671508848, + "learning_rate": 0.003, + "loss": 4.0912, + "step": 8438 + }, + { + "epoch": 0.08439, + "grad_norm": 0.9805961893177073, + "learning_rate": 0.003, + "loss": 4.1115, + "step": 8439 + }, + { + "epoch": 0.0844, + "grad_norm": 1.2397471986974102, + "learning_rate": 0.003, + "loss": 4.1255, + "step": 8440 + }, + { + "epoch": 0.08441, + "grad_norm": 0.7657531909009802, + "learning_rate": 0.003, + "loss": 4.0912, + "step": 8441 + }, + { + "epoch": 0.08442, + "grad_norm": 0.6761237170315371, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 8442 + }, + { + "epoch": 0.08443, + "grad_norm": 0.6698413991893777, + "learning_rate": 0.003, + "loss": 4.0963, + "step": 8443 + }, + { + "epoch": 0.08444, + "grad_norm": 0.6538638151947389, + "learning_rate": 0.003, + "loss": 4.1241, + "step": 8444 + }, + { + "epoch": 0.08445, + "grad_norm": 0.7819289887319408, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 8445 + }, + { + "epoch": 0.08446, + "grad_norm": 0.7360808672824343, + "learning_rate": 0.003, + "loss": 4.095, + "step": 8446 + }, + { + "epoch": 0.08447, + "grad_norm": 0.8192930714654896, + "learning_rate": 0.003, + "loss": 4.1053, + "step": 8447 + }, + { + "epoch": 0.08448, + "grad_norm": 0.9656464757675268, + "learning_rate": 0.003, + "loss": 4.109, + "step": 8448 + }, + { + "epoch": 0.08449, + "grad_norm": 1.1936151486401498, + "learning_rate": 0.003, + "loss": 4.1264, + "step": 8449 + }, + { + "epoch": 0.0845, + "grad_norm": 0.685549098243575, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 8450 + }, + { + "epoch": 0.08451, + "grad_norm": 0.9139246196778028, + "learning_rate": 0.003, + "loss": 4.1179, + "step": 8451 + }, + { + "epoch": 0.08452, + "grad_norm": 1.2202645871735718, + "learning_rate": 0.003, + "loss": 4.1024, + "step": 8452 + }, + { + "epoch": 0.08453, + "grad_norm": 1.011747247457234, + "learning_rate": 0.003, + "loss": 4.1189, + "step": 8453 + }, + { + "epoch": 0.08454, + "grad_norm": 1.0262226366034506, + "learning_rate": 0.003, + "loss": 4.1133, + "step": 8454 + }, + { + "epoch": 0.08455, + "grad_norm": 1.0099082871033074, + "learning_rate": 0.003, + "loss": 4.0874, + "step": 8455 + }, + { + "epoch": 0.08456, + "grad_norm": 0.90849087232309, + "learning_rate": 0.003, + "loss": 4.0955, + "step": 8456 + }, + { + "epoch": 0.08457, + "grad_norm": 0.8383502937124733, + "learning_rate": 0.003, + "loss": 4.1304, + "step": 8457 + }, + { + "epoch": 0.08458, + "grad_norm": 0.8440455792853685, + "learning_rate": 0.003, + "loss": 4.1004, + "step": 8458 + }, + { + "epoch": 0.08459, + "grad_norm": 0.7112059922093875, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 8459 + }, + { + "epoch": 0.0846, + "grad_norm": 0.649515091750979, + "learning_rate": 0.003, + "loss": 4.1027, + "step": 8460 + }, + { + "epoch": 0.08461, + "grad_norm": 0.6333241585768099, + "learning_rate": 0.003, + "loss": 4.107, + "step": 8461 + }, + { + "epoch": 0.08462, + "grad_norm": 0.6384025767344915, + "learning_rate": 0.003, + "loss": 4.105, + "step": 8462 + }, + { + "epoch": 0.08463, + "grad_norm": 0.6400437078939833, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 8463 + }, + { + "epoch": 0.08464, + "grad_norm": 0.6759399211580148, + "learning_rate": 0.003, + "loss": 4.1134, + "step": 8464 + }, + { + "epoch": 0.08465, + "grad_norm": 0.7413271665798129, + "learning_rate": 0.003, + "loss": 4.0935, + "step": 8465 + }, + { + "epoch": 0.08466, + "grad_norm": 0.8771869007648114, + "learning_rate": 0.003, + "loss": 4.1166, + "step": 8466 + }, + { + "epoch": 0.08467, + "grad_norm": 0.9227181671457924, + "learning_rate": 0.003, + "loss": 4.1193, + "step": 8467 + }, + { + "epoch": 0.08468, + "grad_norm": 0.9695793195002803, + "learning_rate": 0.003, + "loss": 4.1123, + "step": 8468 + }, + { + "epoch": 0.08469, + "grad_norm": 1.2889696763476837, + "learning_rate": 0.003, + "loss": 4.1225, + "step": 8469 + }, + { + "epoch": 0.0847, + "grad_norm": 0.9448980455530938, + "learning_rate": 0.003, + "loss": 4.0985, + "step": 8470 + }, + { + "epoch": 0.08471, + "grad_norm": 1.1057721964654461, + "learning_rate": 0.003, + "loss": 4.087, + "step": 8471 + }, + { + "epoch": 0.08472, + "grad_norm": 1.1484149282990468, + "learning_rate": 0.003, + "loss": 4.0923, + "step": 8472 + }, + { + "epoch": 0.08473, + "grad_norm": 1.0503023171846668, + "learning_rate": 0.003, + "loss": 4.0876, + "step": 8473 + }, + { + "epoch": 0.08474, + "grad_norm": 0.8628642552887439, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 8474 + }, + { + "epoch": 0.08475, + "grad_norm": 0.7509423112660792, + "learning_rate": 0.003, + "loss": 4.0941, + "step": 8475 + }, + { + "epoch": 0.08476, + "grad_norm": 0.70961640106593, + "learning_rate": 0.003, + "loss": 4.1127, + "step": 8476 + }, + { + "epoch": 0.08477, + "grad_norm": 0.7231357262197663, + "learning_rate": 0.003, + "loss": 4.0945, + "step": 8477 + }, + { + "epoch": 0.08478, + "grad_norm": 0.804332121451016, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 8478 + }, + { + "epoch": 0.08479, + "grad_norm": 0.9887354629105892, + "learning_rate": 0.003, + "loss": 4.1141, + "step": 8479 + }, + { + "epoch": 0.0848, + "grad_norm": 0.9881963393542088, + "learning_rate": 0.003, + "loss": 4.1153, + "step": 8480 + }, + { + "epoch": 0.08481, + "grad_norm": 0.9379461518994493, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 8481 + }, + { + "epoch": 0.08482, + "grad_norm": 0.7805227298805099, + "learning_rate": 0.003, + "loss": 4.1036, + "step": 8482 + }, + { + "epoch": 0.08483, + "grad_norm": 0.7856593793741489, + "learning_rate": 0.003, + "loss": 4.0977, + "step": 8483 + }, + { + "epoch": 0.08484, + "grad_norm": 0.8932525879434696, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 8484 + }, + { + "epoch": 0.08485, + "grad_norm": 0.8851276814593144, + "learning_rate": 0.003, + "loss": 4.1077, + "step": 8485 + }, + { + "epoch": 0.08486, + "grad_norm": 0.8944513894921937, + "learning_rate": 0.003, + "loss": 4.1289, + "step": 8486 + }, + { + "epoch": 0.08487, + "grad_norm": 1.0310888734066477, + "learning_rate": 0.003, + "loss": 4.1311, + "step": 8487 + }, + { + "epoch": 0.08488, + "grad_norm": 1.0686836420532142, + "learning_rate": 0.003, + "loss": 4.074, + "step": 8488 + }, + { + "epoch": 0.08489, + "grad_norm": 1.0317207317630044, + "learning_rate": 0.003, + "loss": 4.1129, + "step": 8489 + }, + { + "epoch": 0.0849, + "grad_norm": 1.1495485663194194, + "learning_rate": 0.003, + "loss": 4.1236, + "step": 8490 + }, + { + "epoch": 0.08491, + "grad_norm": 1.0263416248655877, + "learning_rate": 0.003, + "loss": 4.1202, + "step": 8491 + }, + { + "epoch": 0.08492, + "grad_norm": 0.9700428254885138, + "learning_rate": 0.003, + "loss": 4.1051, + "step": 8492 + }, + { + "epoch": 0.08493, + "grad_norm": 0.9948556669000941, + "learning_rate": 0.003, + "loss": 4.1167, + "step": 8493 + }, + { + "epoch": 0.08494, + "grad_norm": 0.910141922117321, + "learning_rate": 0.003, + "loss": 4.0986, + "step": 8494 + }, + { + "epoch": 0.08495, + "grad_norm": 0.9597026790665053, + "learning_rate": 0.003, + "loss": 4.1105, + "step": 8495 + }, + { + "epoch": 0.08496, + "grad_norm": 0.9275785612350264, + "learning_rate": 0.003, + "loss": 4.0845, + "step": 8496 + }, + { + "epoch": 0.08497, + "grad_norm": 1.01867706095219, + "learning_rate": 0.003, + "loss": 4.1037, + "step": 8497 + }, + { + "epoch": 0.08498, + "grad_norm": 1.1590197748031095, + "learning_rate": 0.003, + "loss": 4.118, + "step": 8498 + }, + { + "epoch": 0.08499, + "grad_norm": 0.9880096011902544, + "learning_rate": 0.003, + "loss": 4.1334, + "step": 8499 + }, + { + "epoch": 0.085, + "grad_norm": 0.8266327675260843, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 8500 + }, + { + "epoch": 0.08501, + "grad_norm": 0.7411428596681484, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 8501 + }, + { + "epoch": 0.08502, + "grad_norm": 0.6949138984131208, + "learning_rate": 0.003, + "loss": 4.0995, + "step": 8502 + }, + { + "epoch": 0.08503, + "grad_norm": 0.7536288112312667, + "learning_rate": 0.003, + "loss": 4.1006, + "step": 8503 + }, + { + "epoch": 0.08504, + "grad_norm": 0.7223239288762903, + "learning_rate": 0.003, + "loss": 4.1265, + "step": 8504 + }, + { + "epoch": 0.08505, + "grad_norm": 0.675765276936943, + "learning_rate": 0.003, + "loss": 4.0868, + "step": 8505 + }, + { + "epoch": 0.08506, + "grad_norm": 0.7576978694249362, + "learning_rate": 0.003, + "loss": 4.1066, + "step": 8506 + }, + { + "epoch": 0.08507, + "grad_norm": 0.8515649801971872, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 8507 + }, + { + "epoch": 0.08508, + "grad_norm": 1.057062705401568, + "learning_rate": 0.003, + "loss": 4.088, + "step": 8508 + }, + { + "epoch": 0.08509, + "grad_norm": 1.121630687562224, + "learning_rate": 0.003, + "loss": 4.1326, + "step": 8509 + }, + { + "epoch": 0.0851, + "grad_norm": 0.9104321815273234, + "learning_rate": 0.003, + "loss": 4.1003, + "step": 8510 + }, + { + "epoch": 0.08511, + "grad_norm": 0.8342488080751641, + "learning_rate": 0.003, + "loss": 4.1064, + "step": 8511 + }, + { + "epoch": 0.08512, + "grad_norm": 0.7988070005118341, + "learning_rate": 0.003, + "loss": 4.1328, + "step": 8512 + }, + { + "epoch": 0.08513, + "grad_norm": 0.799837965249621, + "learning_rate": 0.003, + "loss": 4.0938, + "step": 8513 + }, + { + "epoch": 0.08514, + "grad_norm": 0.835432784668365, + "learning_rate": 0.003, + "loss": 4.1025, + "step": 8514 + }, + { + "epoch": 0.08515, + "grad_norm": 0.8955465769825964, + "learning_rate": 0.003, + "loss": 4.0935, + "step": 8515 + }, + { + "epoch": 0.08516, + "grad_norm": 0.8928688903157641, + "learning_rate": 0.003, + "loss": 4.1252, + "step": 8516 + }, + { + "epoch": 0.08517, + "grad_norm": 0.8226476531644744, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 8517 + }, + { + "epoch": 0.08518, + "grad_norm": 0.9561707034998627, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 8518 + }, + { + "epoch": 0.08519, + "grad_norm": 1.1902153338579717, + "learning_rate": 0.003, + "loss": 4.0934, + "step": 8519 + }, + { + "epoch": 0.0852, + "grad_norm": 0.8018095905157278, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 8520 + }, + { + "epoch": 0.08521, + "grad_norm": 0.5533864795410459, + "learning_rate": 0.003, + "loss": 4.1046, + "step": 8521 + }, + { + "epoch": 0.08522, + "grad_norm": 0.7791152111447774, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 8522 + }, + { + "epoch": 0.08523, + "grad_norm": 0.900500602043729, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 8523 + }, + { + "epoch": 0.08524, + "grad_norm": 0.9074705833253267, + "learning_rate": 0.003, + "loss": 4.0906, + "step": 8524 + }, + { + "epoch": 0.08525, + "grad_norm": 0.9112667601876601, + "learning_rate": 0.003, + "loss": 4.0881, + "step": 8525 + }, + { + "epoch": 0.08526, + "grad_norm": 1.0212625794257977, + "learning_rate": 0.003, + "loss": 4.0955, + "step": 8526 + }, + { + "epoch": 0.08527, + "grad_norm": 1.0584452877128767, + "learning_rate": 0.003, + "loss": 4.1014, + "step": 8527 + }, + { + "epoch": 0.08528, + "grad_norm": 0.8561020391089984, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 8528 + }, + { + "epoch": 0.08529, + "grad_norm": 0.8298501875520379, + "learning_rate": 0.003, + "loss": 4.106, + "step": 8529 + }, + { + "epoch": 0.0853, + "grad_norm": 0.9089572166905378, + "learning_rate": 0.003, + "loss": 4.1068, + "step": 8530 + }, + { + "epoch": 0.08531, + "grad_norm": 0.8653272986614832, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 8531 + }, + { + "epoch": 0.08532, + "grad_norm": 0.788926984023568, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 8532 + }, + { + "epoch": 0.08533, + "grad_norm": 0.8903035626600532, + "learning_rate": 0.003, + "loss": 4.117, + "step": 8533 + }, + { + "epoch": 0.08534, + "grad_norm": 0.963034125025281, + "learning_rate": 0.003, + "loss": 4.115, + "step": 8534 + }, + { + "epoch": 0.08535, + "grad_norm": 1.2879128764669079, + "learning_rate": 0.003, + "loss": 4.1075, + "step": 8535 + }, + { + "epoch": 0.08536, + "grad_norm": 0.8902797022173426, + "learning_rate": 0.003, + "loss": 4.1148, + "step": 8536 + }, + { + "epoch": 0.08537, + "grad_norm": 0.8657020447888398, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 8537 + }, + { + "epoch": 0.08538, + "grad_norm": 0.849426185382095, + "learning_rate": 0.003, + "loss": 4.0868, + "step": 8538 + }, + { + "epoch": 0.08539, + "grad_norm": 0.8702155984928356, + "learning_rate": 0.003, + "loss": 4.1264, + "step": 8539 + }, + { + "epoch": 0.0854, + "grad_norm": 0.7764427649934481, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 8540 + }, + { + "epoch": 0.08541, + "grad_norm": 0.7058501386903097, + "learning_rate": 0.003, + "loss": 4.081, + "step": 8541 + }, + { + "epoch": 0.08542, + "grad_norm": 0.8663633824046528, + "learning_rate": 0.003, + "loss": 4.1205, + "step": 8542 + }, + { + "epoch": 0.08543, + "grad_norm": 0.906265069337281, + "learning_rate": 0.003, + "loss": 4.1076, + "step": 8543 + }, + { + "epoch": 0.08544, + "grad_norm": 0.9275930195253327, + "learning_rate": 0.003, + "loss": 4.089, + "step": 8544 + }, + { + "epoch": 0.08545, + "grad_norm": 0.8078784033785711, + "learning_rate": 0.003, + "loss": 4.0949, + "step": 8545 + }, + { + "epoch": 0.08546, + "grad_norm": 0.9066354249548785, + "learning_rate": 0.003, + "loss": 4.1001, + "step": 8546 + }, + { + "epoch": 0.08547, + "grad_norm": 1.3364123020627683, + "learning_rate": 0.003, + "loss": 4.1259, + "step": 8547 + }, + { + "epoch": 0.08548, + "grad_norm": 0.9435393460362433, + "learning_rate": 0.003, + "loss": 4.1209, + "step": 8548 + }, + { + "epoch": 0.08549, + "grad_norm": 0.9332959081861278, + "learning_rate": 0.003, + "loss": 4.1112, + "step": 8549 + }, + { + "epoch": 0.0855, + "grad_norm": 0.8374993962866415, + "learning_rate": 0.003, + "loss": 4.1259, + "step": 8550 + }, + { + "epoch": 0.08551, + "grad_norm": 0.72433894811856, + "learning_rate": 0.003, + "loss": 4.1202, + "step": 8551 + }, + { + "epoch": 0.08552, + "grad_norm": 0.7603328491561405, + "learning_rate": 0.003, + "loss": 4.1029, + "step": 8552 + }, + { + "epoch": 0.08553, + "grad_norm": 0.8165853461739714, + "learning_rate": 0.003, + "loss": 4.1123, + "step": 8553 + }, + { + "epoch": 0.08554, + "grad_norm": 0.8329371861835474, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 8554 + }, + { + "epoch": 0.08555, + "grad_norm": 0.8836247642143207, + "learning_rate": 0.003, + "loss": 4.1223, + "step": 8555 + }, + { + "epoch": 0.08556, + "grad_norm": 1.0655224233717422, + "learning_rate": 0.003, + "loss": 4.1118, + "step": 8556 + }, + { + "epoch": 0.08557, + "grad_norm": 1.1033400346275346, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 8557 + }, + { + "epoch": 0.08558, + "grad_norm": 0.9185812974874371, + "learning_rate": 0.003, + "loss": 4.1023, + "step": 8558 + }, + { + "epoch": 0.08559, + "grad_norm": 0.8449525867089718, + "learning_rate": 0.003, + "loss": 4.108, + "step": 8559 + }, + { + "epoch": 0.0856, + "grad_norm": 0.8791279440183213, + "learning_rate": 0.003, + "loss": 4.0994, + "step": 8560 + }, + { + "epoch": 0.08561, + "grad_norm": 1.007009657834021, + "learning_rate": 0.003, + "loss": 4.0971, + "step": 8561 + }, + { + "epoch": 0.08562, + "grad_norm": 1.2506411131458524, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 8562 + }, + { + "epoch": 0.08563, + "grad_norm": 0.8595663427695304, + "learning_rate": 0.003, + "loss": 4.1006, + "step": 8563 + }, + { + "epoch": 0.08564, + "grad_norm": 0.8258432628148261, + "learning_rate": 0.003, + "loss": 4.0966, + "step": 8564 + }, + { + "epoch": 0.08565, + "grad_norm": 0.8232028179529909, + "learning_rate": 0.003, + "loss": 4.0845, + "step": 8565 + }, + { + "epoch": 0.08566, + "grad_norm": 0.7588530988097654, + "learning_rate": 0.003, + "loss": 4.0996, + "step": 8566 + }, + { + "epoch": 0.08567, + "grad_norm": 0.6824147570028768, + "learning_rate": 0.003, + "loss": 4.1034, + "step": 8567 + }, + { + "epoch": 0.08568, + "grad_norm": 0.5996215117346594, + "learning_rate": 0.003, + "loss": 4.1149, + "step": 8568 + }, + { + "epoch": 0.08569, + "grad_norm": 0.5807584231599277, + "learning_rate": 0.003, + "loss": 4.118, + "step": 8569 + }, + { + "epoch": 0.0857, + "grad_norm": 0.5266785999312754, + "learning_rate": 0.003, + "loss": 4.1146, + "step": 8570 + }, + { + "epoch": 0.08571, + "grad_norm": 0.4724117155463532, + "learning_rate": 0.003, + "loss": 4.0977, + "step": 8571 + }, + { + "epoch": 0.08572, + "grad_norm": 0.49806214531928805, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 8572 + }, + { + "epoch": 0.08573, + "grad_norm": 0.5210491883768703, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 8573 + }, + { + "epoch": 0.08574, + "grad_norm": 0.580899421582982, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 8574 + }, + { + "epoch": 0.08575, + "grad_norm": 0.759338722165245, + "learning_rate": 0.003, + "loss": 4.0987, + "step": 8575 + }, + { + "epoch": 0.08576, + "grad_norm": 1.0882621390033222, + "learning_rate": 0.003, + "loss": 4.0931, + "step": 8576 + }, + { + "epoch": 0.08577, + "grad_norm": 1.286177814911776, + "learning_rate": 0.003, + "loss": 4.0995, + "step": 8577 + }, + { + "epoch": 0.08578, + "grad_norm": 0.649434769796102, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 8578 + }, + { + "epoch": 0.08579, + "grad_norm": 0.7955247063051825, + "learning_rate": 0.003, + "loss": 4.0919, + "step": 8579 + }, + { + "epoch": 0.0858, + "grad_norm": 0.9332228897387346, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 8580 + }, + { + "epoch": 0.08581, + "grad_norm": 0.9286534200000857, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 8581 + }, + { + "epoch": 0.08582, + "grad_norm": 0.8515796422164925, + "learning_rate": 0.003, + "loss": 4.1029, + "step": 8582 + }, + { + "epoch": 0.08583, + "grad_norm": 0.9496730587144366, + "learning_rate": 0.003, + "loss": 4.1166, + "step": 8583 + }, + { + "epoch": 0.08584, + "grad_norm": 1.0376797351944236, + "learning_rate": 0.003, + "loss": 4.1119, + "step": 8584 + }, + { + "epoch": 0.08585, + "grad_norm": 1.0733355797768753, + "learning_rate": 0.003, + "loss": 4.1171, + "step": 8585 + }, + { + "epoch": 0.08586, + "grad_norm": 1.0898909465211588, + "learning_rate": 0.003, + "loss": 4.1357, + "step": 8586 + }, + { + "epoch": 0.08587, + "grad_norm": 0.8541629848679168, + "learning_rate": 0.003, + "loss": 4.1036, + "step": 8587 + }, + { + "epoch": 0.08588, + "grad_norm": 0.9259074296023726, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 8588 + }, + { + "epoch": 0.08589, + "grad_norm": 1.0133166847780468, + "learning_rate": 0.003, + "loss": 4.1106, + "step": 8589 + }, + { + "epoch": 0.0859, + "grad_norm": 0.9952675282959716, + "learning_rate": 0.003, + "loss": 4.0997, + "step": 8590 + }, + { + "epoch": 0.08591, + "grad_norm": 0.8432804668392331, + "learning_rate": 0.003, + "loss": 4.1007, + "step": 8591 + }, + { + "epoch": 0.08592, + "grad_norm": 0.7927775995190869, + "learning_rate": 0.003, + "loss": 4.1234, + "step": 8592 + }, + { + "epoch": 0.08593, + "grad_norm": 0.9135931190636065, + "learning_rate": 0.003, + "loss": 4.1467, + "step": 8593 + }, + { + "epoch": 0.08594, + "grad_norm": 0.89183466708387, + "learning_rate": 0.003, + "loss": 4.1019, + "step": 8594 + }, + { + "epoch": 0.08595, + "grad_norm": 1.0748167455019701, + "learning_rate": 0.003, + "loss": 4.1136, + "step": 8595 + }, + { + "epoch": 0.08596, + "grad_norm": 0.9757553423551145, + "learning_rate": 0.003, + "loss": 4.1293, + "step": 8596 + }, + { + "epoch": 0.08597, + "grad_norm": 1.1029927439937899, + "learning_rate": 0.003, + "loss": 4.1324, + "step": 8597 + }, + { + "epoch": 0.08598, + "grad_norm": 1.2202072285107675, + "learning_rate": 0.003, + "loss": 4.1056, + "step": 8598 + }, + { + "epoch": 0.08599, + "grad_norm": 1.0472074968226261, + "learning_rate": 0.003, + "loss": 4.1223, + "step": 8599 + }, + { + "epoch": 0.086, + "grad_norm": 0.9436651521834845, + "learning_rate": 0.003, + "loss": 4.1399, + "step": 8600 + }, + { + "epoch": 0.08601, + "grad_norm": 0.8978086885499934, + "learning_rate": 0.003, + "loss": 4.122, + "step": 8601 + }, + { + "epoch": 0.08602, + "grad_norm": 0.9046422898166396, + "learning_rate": 0.003, + "loss": 4.1342, + "step": 8602 + }, + { + "epoch": 0.08603, + "grad_norm": 0.9940609256714368, + "learning_rate": 0.003, + "loss": 4.1248, + "step": 8603 + }, + { + "epoch": 0.08604, + "grad_norm": 1.2914507269620887, + "learning_rate": 0.003, + "loss": 4.1102, + "step": 8604 + }, + { + "epoch": 0.08605, + "grad_norm": 1.0010719188912025, + "learning_rate": 0.003, + "loss": 4.1236, + "step": 8605 + }, + { + "epoch": 0.08606, + "grad_norm": 0.851064370373218, + "learning_rate": 0.003, + "loss": 4.0857, + "step": 8606 + }, + { + "epoch": 0.08607, + "grad_norm": 0.8084048446777777, + "learning_rate": 0.003, + "loss": 4.129, + "step": 8607 + }, + { + "epoch": 0.08608, + "grad_norm": 0.8260223223516866, + "learning_rate": 0.003, + "loss": 4.1001, + "step": 8608 + }, + { + "epoch": 0.08609, + "grad_norm": 0.800858297527019, + "learning_rate": 0.003, + "loss": 4.1214, + "step": 8609 + }, + { + "epoch": 0.0861, + "grad_norm": 0.7751744199950412, + "learning_rate": 0.003, + "loss": 4.1195, + "step": 8610 + }, + { + "epoch": 0.08611, + "grad_norm": 0.7850692774747576, + "learning_rate": 0.003, + "loss": 4.1007, + "step": 8611 + }, + { + "epoch": 0.08612, + "grad_norm": 0.8034287087644159, + "learning_rate": 0.003, + "loss": 4.1349, + "step": 8612 + }, + { + "epoch": 0.08613, + "grad_norm": 0.7907981566162866, + "learning_rate": 0.003, + "loss": 4.1094, + "step": 8613 + }, + { + "epoch": 0.08614, + "grad_norm": 0.7648908788973424, + "learning_rate": 0.003, + "loss": 4.0918, + "step": 8614 + }, + { + "epoch": 0.08615, + "grad_norm": 0.8058658371668642, + "learning_rate": 0.003, + "loss": 4.0927, + "step": 8615 + }, + { + "epoch": 0.08616, + "grad_norm": 1.0179169747185717, + "learning_rate": 0.003, + "loss": 4.075, + "step": 8616 + }, + { + "epoch": 0.08617, + "grad_norm": 1.3401947590337657, + "learning_rate": 0.003, + "loss": 4.0957, + "step": 8617 + }, + { + "epoch": 0.08618, + "grad_norm": 0.6850314006469229, + "learning_rate": 0.003, + "loss": 4.0983, + "step": 8618 + }, + { + "epoch": 0.08619, + "grad_norm": 0.8070821056107416, + "learning_rate": 0.003, + "loss": 4.1173, + "step": 8619 + }, + { + "epoch": 0.0862, + "grad_norm": 1.0123939132301667, + "learning_rate": 0.003, + "loss": 4.107, + "step": 8620 + }, + { + "epoch": 0.08621, + "grad_norm": 1.1347343173557283, + "learning_rate": 0.003, + "loss": 4.0967, + "step": 8621 + }, + { + "epoch": 0.08622, + "grad_norm": 0.9266592246218266, + "learning_rate": 0.003, + "loss": 4.12, + "step": 8622 + }, + { + "epoch": 0.08623, + "grad_norm": 0.9044983849980115, + "learning_rate": 0.003, + "loss": 4.1097, + "step": 8623 + }, + { + "epoch": 0.08624, + "grad_norm": 0.8761897564386483, + "learning_rate": 0.003, + "loss": 4.0926, + "step": 8624 + }, + { + "epoch": 0.08625, + "grad_norm": 0.8620326850964458, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 8625 + }, + { + "epoch": 0.08626, + "grad_norm": 0.7720603843449916, + "learning_rate": 0.003, + "loss": 4.1196, + "step": 8626 + }, + { + "epoch": 0.08627, + "grad_norm": 0.8313063670205044, + "learning_rate": 0.003, + "loss": 4.0995, + "step": 8627 + }, + { + "epoch": 0.08628, + "grad_norm": 0.7788816276172811, + "learning_rate": 0.003, + "loss": 4.1021, + "step": 8628 + }, + { + "epoch": 0.08629, + "grad_norm": 0.8225879339157376, + "learning_rate": 0.003, + "loss": 4.097, + "step": 8629 + }, + { + "epoch": 0.0863, + "grad_norm": 0.9503798474325023, + "learning_rate": 0.003, + "loss": 4.0917, + "step": 8630 + }, + { + "epoch": 0.08631, + "grad_norm": 1.0311613535749917, + "learning_rate": 0.003, + "loss": 4.1118, + "step": 8631 + }, + { + "epoch": 0.08632, + "grad_norm": 1.097072302177679, + "learning_rate": 0.003, + "loss": 4.0961, + "step": 8632 + }, + { + "epoch": 0.08633, + "grad_norm": 0.905010614567883, + "learning_rate": 0.003, + "loss": 4.1291, + "step": 8633 + }, + { + "epoch": 0.08634, + "grad_norm": 0.9280254071681021, + "learning_rate": 0.003, + "loss": 4.1147, + "step": 8634 + }, + { + "epoch": 0.08635, + "grad_norm": 0.9486845314400906, + "learning_rate": 0.003, + "loss": 4.1067, + "step": 8635 + }, + { + "epoch": 0.08636, + "grad_norm": 0.965675528971389, + "learning_rate": 0.003, + "loss": 4.1059, + "step": 8636 + }, + { + "epoch": 0.08637, + "grad_norm": 0.9874360924938551, + "learning_rate": 0.003, + "loss": 4.1314, + "step": 8637 + }, + { + "epoch": 0.08638, + "grad_norm": 0.9015114006705611, + "learning_rate": 0.003, + "loss": 4.12, + "step": 8638 + }, + { + "epoch": 0.08639, + "grad_norm": 0.8776646457013992, + "learning_rate": 0.003, + "loss": 4.0976, + "step": 8639 + }, + { + "epoch": 0.0864, + "grad_norm": 1.1686490796312563, + "learning_rate": 0.003, + "loss": 4.1087, + "step": 8640 + }, + { + "epoch": 0.08641, + "grad_norm": 1.1835266668134277, + "learning_rate": 0.003, + "loss": 4.1163, + "step": 8641 + }, + { + "epoch": 0.08642, + "grad_norm": 0.7789856451482106, + "learning_rate": 0.003, + "loss": 4.1074, + "step": 8642 + }, + { + "epoch": 0.08643, + "grad_norm": 0.7159387374682472, + "learning_rate": 0.003, + "loss": 4.0962, + "step": 8643 + }, + { + "epoch": 0.08644, + "grad_norm": 0.7127204757723076, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 8644 + }, + { + "epoch": 0.08645, + "grad_norm": 0.8068921966237395, + "learning_rate": 0.003, + "loss": 4.0903, + "step": 8645 + }, + { + "epoch": 0.08646, + "grad_norm": 0.9115412811447862, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 8646 + }, + { + "epoch": 0.08647, + "grad_norm": 0.9852125911616223, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 8647 + }, + { + "epoch": 0.08648, + "grad_norm": 0.9203439827975389, + "learning_rate": 0.003, + "loss": 4.1004, + "step": 8648 + }, + { + "epoch": 0.08649, + "grad_norm": 0.8551894521735353, + "learning_rate": 0.003, + "loss": 4.123, + "step": 8649 + }, + { + "epoch": 0.0865, + "grad_norm": 0.752190786342537, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 8650 + }, + { + "epoch": 0.08651, + "grad_norm": 0.736051245410185, + "learning_rate": 0.003, + "loss": 4.0992, + "step": 8651 + }, + { + "epoch": 0.08652, + "grad_norm": 0.6463683221296629, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 8652 + }, + { + "epoch": 0.08653, + "grad_norm": 0.6704399865370819, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 8653 + }, + { + "epoch": 0.08654, + "grad_norm": 0.7075584672548234, + "learning_rate": 0.003, + "loss": 4.0918, + "step": 8654 + }, + { + "epoch": 0.08655, + "grad_norm": 0.7589064409569638, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 8655 + }, + { + "epoch": 0.08656, + "grad_norm": 0.8797294460543287, + "learning_rate": 0.003, + "loss": 4.1078, + "step": 8656 + }, + { + "epoch": 0.08657, + "grad_norm": 1.0746650206858512, + "learning_rate": 0.003, + "loss": 4.1251, + "step": 8657 + }, + { + "epoch": 0.08658, + "grad_norm": 1.0381315037441246, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 8658 + }, + { + "epoch": 0.08659, + "grad_norm": 0.9216590731223901, + "learning_rate": 0.003, + "loss": 4.0936, + "step": 8659 + }, + { + "epoch": 0.0866, + "grad_norm": 0.9060927251078454, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 8660 + }, + { + "epoch": 0.08661, + "grad_norm": 0.8365841556281326, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 8661 + }, + { + "epoch": 0.08662, + "grad_norm": 0.940796854284168, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 8662 + }, + { + "epoch": 0.08663, + "grad_norm": 1.1158443930232451, + "learning_rate": 0.003, + "loss": 4.1147, + "step": 8663 + }, + { + "epoch": 0.08664, + "grad_norm": 1.0186570224186573, + "learning_rate": 0.003, + "loss": 4.0994, + "step": 8664 + }, + { + "epoch": 0.08665, + "grad_norm": 1.0348438398835733, + "learning_rate": 0.003, + "loss": 4.1048, + "step": 8665 + }, + { + "epoch": 0.08666, + "grad_norm": 0.9620834884254315, + "learning_rate": 0.003, + "loss": 4.0989, + "step": 8666 + }, + { + "epoch": 0.08667, + "grad_norm": 1.006500389771506, + "learning_rate": 0.003, + "loss": 4.0898, + "step": 8667 + }, + { + "epoch": 0.08668, + "grad_norm": 0.9850331378360856, + "learning_rate": 0.003, + "loss": 4.0902, + "step": 8668 + }, + { + "epoch": 0.08669, + "grad_norm": 0.6894471054181678, + "learning_rate": 0.003, + "loss": 4.084, + "step": 8669 + }, + { + "epoch": 0.0867, + "grad_norm": 0.6694894610529343, + "learning_rate": 0.003, + "loss": 4.0854, + "step": 8670 + }, + { + "epoch": 0.08671, + "grad_norm": 0.5962672000226233, + "learning_rate": 0.003, + "loss": 4.0969, + "step": 8671 + }, + { + "epoch": 0.08672, + "grad_norm": 0.6763083203036772, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 8672 + }, + { + "epoch": 0.08673, + "grad_norm": 0.7676489015109336, + "learning_rate": 0.003, + "loss": 4.0941, + "step": 8673 + }, + { + "epoch": 0.08674, + "grad_norm": 0.7787151403749619, + "learning_rate": 0.003, + "loss": 4.1001, + "step": 8674 + }, + { + "epoch": 0.08675, + "grad_norm": 0.7891229989108475, + "learning_rate": 0.003, + "loss": 4.1025, + "step": 8675 + }, + { + "epoch": 0.08676, + "grad_norm": 0.8927860274538245, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 8676 + }, + { + "epoch": 0.08677, + "grad_norm": 1.067464395193778, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 8677 + }, + { + "epoch": 0.08678, + "grad_norm": 0.8827877751142491, + "learning_rate": 0.003, + "loss": 4.0855, + "step": 8678 + }, + { + "epoch": 0.08679, + "grad_norm": 0.7244826450369652, + "learning_rate": 0.003, + "loss": 4.0942, + "step": 8679 + }, + { + "epoch": 0.0868, + "grad_norm": 0.633006551065808, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 8680 + }, + { + "epoch": 0.08681, + "grad_norm": 0.7719505462142406, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 8681 + }, + { + "epoch": 0.08682, + "grad_norm": 1.0884331781139083, + "learning_rate": 0.003, + "loss": 4.0983, + "step": 8682 + }, + { + "epoch": 0.08683, + "grad_norm": 1.1714352580907677, + "learning_rate": 0.003, + "loss": 4.1194, + "step": 8683 + }, + { + "epoch": 0.08684, + "grad_norm": 1.0196218025826278, + "learning_rate": 0.003, + "loss": 4.1086, + "step": 8684 + }, + { + "epoch": 0.08685, + "grad_norm": 1.0458752270945595, + "learning_rate": 0.003, + "loss": 4.1085, + "step": 8685 + }, + { + "epoch": 0.08686, + "grad_norm": 0.9641978455934338, + "learning_rate": 0.003, + "loss": 4.109, + "step": 8686 + }, + { + "epoch": 0.08687, + "grad_norm": 0.9135236516973639, + "learning_rate": 0.003, + "loss": 4.1102, + "step": 8687 + }, + { + "epoch": 0.08688, + "grad_norm": 0.8512259905499174, + "learning_rate": 0.003, + "loss": 4.0977, + "step": 8688 + }, + { + "epoch": 0.08689, + "grad_norm": 0.8190255834805528, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 8689 + }, + { + "epoch": 0.0869, + "grad_norm": 0.8398166759029391, + "learning_rate": 0.003, + "loss": 4.1031, + "step": 8690 + }, + { + "epoch": 0.08691, + "grad_norm": 0.8574277247566975, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 8691 + }, + { + "epoch": 0.08692, + "grad_norm": 0.9518906449211479, + "learning_rate": 0.003, + "loss": 4.1055, + "step": 8692 + }, + { + "epoch": 0.08693, + "grad_norm": 0.9069434842808899, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 8693 + }, + { + "epoch": 0.08694, + "grad_norm": 0.7908401280606225, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 8694 + }, + { + "epoch": 0.08695, + "grad_norm": 0.875710856836608, + "learning_rate": 0.003, + "loss": 4.0952, + "step": 8695 + }, + { + "epoch": 0.08696, + "grad_norm": 1.0582108487000337, + "learning_rate": 0.003, + "loss": 4.1254, + "step": 8696 + }, + { + "epoch": 0.08697, + "grad_norm": 1.1048858177228473, + "learning_rate": 0.003, + "loss": 4.118, + "step": 8697 + }, + { + "epoch": 0.08698, + "grad_norm": 1.0885309061884358, + "learning_rate": 0.003, + "loss": 4.0962, + "step": 8698 + }, + { + "epoch": 0.08699, + "grad_norm": 1.0780097630336325, + "learning_rate": 0.003, + "loss": 4.1093, + "step": 8699 + }, + { + "epoch": 0.087, + "grad_norm": 0.8462640972072897, + "learning_rate": 0.003, + "loss": 4.1286, + "step": 8700 + }, + { + "epoch": 0.08701, + "grad_norm": 0.7500583295478122, + "learning_rate": 0.003, + "loss": 4.1077, + "step": 8701 + }, + { + "epoch": 0.08702, + "grad_norm": 0.8192338624119824, + "learning_rate": 0.003, + "loss": 4.0979, + "step": 8702 + }, + { + "epoch": 0.08703, + "grad_norm": 0.9761783341836241, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 8703 + }, + { + "epoch": 0.08704, + "grad_norm": 1.0637726019049591, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 8704 + }, + { + "epoch": 0.08705, + "grad_norm": 1.039691640283121, + "learning_rate": 0.003, + "loss": 4.1296, + "step": 8705 + }, + { + "epoch": 0.08706, + "grad_norm": 1.0345969448915813, + "learning_rate": 0.003, + "loss": 4.0988, + "step": 8706 + }, + { + "epoch": 0.08707, + "grad_norm": 1.073442317904122, + "learning_rate": 0.003, + "loss": 4.1015, + "step": 8707 + }, + { + "epoch": 0.08708, + "grad_norm": 1.0167968758055734, + "learning_rate": 0.003, + "loss": 4.1084, + "step": 8708 + }, + { + "epoch": 0.08709, + "grad_norm": 1.1281785403043407, + "learning_rate": 0.003, + "loss": 4.1207, + "step": 8709 + }, + { + "epoch": 0.0871, + "grad_norm": 0.856237067855571, + "learning_rate": 0.003, + "loss": 4.0919, + "step": 8710 + }, + { + "epoch": 0.08711, + "grad_norm": 0.7245271989076703, + "learning_rate": 0.003, + "loss": 4.104, + "step": 8711 + }, + { + "epoch": 0.08712, + "grad_norm": 0.6348264826820577, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 8712 + }, + { + "epoch": 0.08713, + "grad_norm": 0.7213305248825544, + "learning_rate": 0.003, + "loss": 4.1101, + "step": 8713 + }, + { + "epoch": 0.08714, + "grad_norm": 0.7267634259694989, + "learning_rate": 0.003, + "loss": 4.0963, + "step": 8714 + }, + { + "epoch": 0.08715, + "grad_norm": 0.7226602431965007, + "learning_rate": 0.003, + "loss": 4.1079, + "step": 8715 + }, + { + "epoch": 0.08716, + "grad_norm": 0.749512188800938, + "learning_rate": 0.003, + "loss": 4.1092, + "step": 8716 + }, + { + "epoch": 0.08717, + "grad_norm": 0.7945844637136416, + "learning_rate": 0.003, + "loss": 4.1148, + "step": 8717 + }, + { + "epoch": 0.08718, + "grad_norm": 0.9455334040170957, + "learning_rate": 0.003, + "loss": 4.107, + "step": 8718 + }, + { + "epoch": 0.08719, + "grad_norm": 1.2301600366994065, + "learning_rate": 0.003, + "loss": 4.1003, + "step": 8719 + }, + { + "epoch": 0.0872, + "grad_norm": 0.9688272507046074, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 8720 + }, + { + "epoch": 0.08721, + "grad_norm": 0.882130678243584, + "learning_rate": 0.003, + "loss": 4.1033, + "step": 8721 + }, + { + "epoch": 0.08722, + "grad_norm": 0.7624823510784595, + "learning_rate": 0.003, + "loss": 4.0923, + "step": 8722 + }, + { + "epoch": 0.08723, + "grad_norm": 0.8208175149976452, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 8723 + }, + { + "epoch": 0.08724, + "grad_norm": 1.0208221289653825, + "learning_rate": 0.003, + "loss": 4.0965, + "step": 8724 + }, + { + "epoch": 0.08725, + "grad_norm": 1.1583066056086238, + "learning_rate": 0.003, + "loss": 4.1105, + "step": 8725 + }, + { + "epoch": 0.08726, + "grad_norm": 0.8636528049466422, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 8726 + }, + { + "epoch": 0.08727, + "grad_norm": 0.7315027868053344, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 8727 + }, + { + "epoch": 0.08728, + "grad_norm": 0.7105524649657913, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 8728 + }, + { + "epoch": 0.08729, + "grad_norm": 0.8027256735287427, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 8729 + }, + { + "epoch": 0.0873, + "grad_norm": 0.7642689263014413, + "learning_rate": 0.003, + "loss": 4.1013, + "step": 8730 + }, + { + "epoch": 0.08731, + "grad_norm": 0.8054605331616608, + "learning_rate": 0.003, + "loss": 4.1007, + "step": 8731 + }, + { + "epoch": 0.08732, + "grad_norm": 0.8920977986550327, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 8732 + }, + { + "epoch": 0.08733, + "grad_norm": 0.8107240409545715, + "learning_rate": 0.003, + "loss": 4.082, + "step": 8733 + }, + { + "epoch": 0.08734, + "grad_norm": 0.8548040032308308, + "learning_rate": 0.003, + "loss": 4.1162, + "step": 8734 + }, + { + "epoch": 0.08735, + "grad_norm": 1.2015525237447153, + "learning_rate": 0.003, + "loss": 4.1058, + "step": 8735 + }, + { + "epoch": 0.08736, + "grad_norm": 1.0254473175716212, + "learning_rate": 0.003, + "loss": 4.1049, + "step": 8736 + }, + { + "epoch": 0.08737, + "grad_norm": 0.9567188089886024, + "learning_rate": 0.003, + "loss": 4.1247, + "step": 8737 + }, + { + "epoch": 0.08738, + "grad_norm": 1.0271105193607937, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 8738 + }, + { + "epoch": 0.08739, + "grad_norm": 0.9282295321228008, + "learning_rate": 0.003, + "loss": 4.0979, + "step": 8739 + }, + { + "epoch": 0.0874, + "grad_norm": 0.8558073220103064, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 8740 + }, + { + "epoch": 0.08741, + "grad_norm": 0.8429315708538133, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 8741 + }, + { + "epoch": 0.08742, + "grad_norm": 0.8121185242286065, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 8742 + }, + { + "epoch": 0.08743, + "grad_norm": 0.9197196989946418, + "learning_rate": 0.003, + "loss": 4.081, + "step": 8743 + }, + { + "epoch": 0.08744, + "grad_norm": 1.1348863902424295, + "learning_rate": 0.003, + "loss": 4.1278, + "step": 8744 + }, + { + "epoch": 0.08745, + "grad_norm": 0.9755220575900516, + "learning_rate": 0.003, + "loss": 4.1005, + "step": 8745 + }, + { + "epoch": 0.08746, + "grad_norm": 1.1617288318988086, + "learning_rate": 0.003, + "loss": 4.1137, + "step": 8746 + }, + { + "epoch": 0.08747, + "grad_norm": 0.8515070893905221, + "learning_rate": 0.003, + "loss": 4.1078, + "step": 8747 + }, + { + "epoch": 0.08748, + "grad_norm": 0.8210706822748797, + "learning_rate": 0.003, + "loss": 4.0965, + "step": 8748 + }, + { + "epoch": 0.08749, + "grad_norm": 0.8533026928432937, + "learning_rate": 0.003, + "loss": 4.0963, + "step": 8749 + }, + { + "epoch": 0.0875, + "grad_norm": 0.7247474091640602, + "learning_rate": 0.003, + "loss": 4.0967, + "step": 8750 + }, + { + "epoch": 0.08751, + "grad_norm": 0.7613678855591323, + "learning_rate": 0.003, + "loss": 4.1013, + "step": 8751 + }, + { + "epoch": 0.08752, + "grad_norm": 0.8298586709080547, + "learning_rate": 0.003, + "loss": 4.1247, + "step": 8752 + }, + { + "epoch": 0.08753, + "grad_norm": 0.9510227968156119, + "learning_rate": 0.003, + "loss": 4.1277, + "step": 8753 + }, + { + "epoch": 0.08754, + "grad_norm": 1.1661555141798292, + "learning_rate": 0.003, + "loss": 4.1142, + "step": 8754 + }, + { + "epoch": 0.08755, + "grad_norm": 0.9742950157697924, + "learning_rate": 0.003, + "loss": 4.093, + "step": 8755 + }, + { + "epoch": 0.08756, + "grad_norm": 1.0948999240878925, + "learning_rate": 0.003, + "loss": 4.0949, + "step": 8756 + }, + { + "epoch": 0.08757, + "grad_norm": 1.0747066998581816, + "learning_rate": 0.003, + "loss": 4.1255, + "step": 8757 + }, + { + "epoch": 0.08758, + "grad_norm": 0.9803826009713699, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 8758 + }, + { + "epoch": 0.08759, + "grad_norm": 0.8978411068790542, + "learning_rate": 0.003, + "loss": 4.1019, + "step": 8759 + }, + { + "epoch": 0.0876, + "grad_norm": 0.8777548601783808, + "learning_rate": 0.003, + "loss": 4.1014, + "step": 8760 + }, + { + "epoch": 0.08761, + "grad_norm": 0.8296530204156158, + "learning_rate": 0.003, + "loss": 4.1319, + "step": 8761 + }, + { + "epoch": 0.08762, + "grad_norm": 0.9503211821723436, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 8762 + }, + { + "epoch": 0.08763, + "grad_norm": 0.9498831438790756, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 8763 + }, + { + "epoch": 0.08764, + "grad_norm": 0.9656491920428545, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 8764 + }, + { + "epoch": 0.08765, + "grad_norm": 1.0147165951625998, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 8765 + }, + { + "epoch": 0.08766, + "grad_norm": 0.9826351379893413, + "learning_rate": 0.003, + "loss": 4.1237, + "step": 8766 + }, + { + "epoch": 0.08767, + "grad_norm": 1.0458878090436103, + "learning_rate": 0.003, + "loss": 4.1043, + "step": 8767 + }, + { + "epoch": 0.08768, + "grad_norm": 1.0251752707071955, + "learning_rate": 0.003, + "loss": 4.1011, + "step": 8768 + }, + { + "epoch": 0.08769, + "grad_norm": 0.9012853792156551, + "learning_rate": 0.003, + "loss": 4.0907, + "step": 8769 + }, + { + "epoch": 0.0877, + "grad_norm": 0.8035752435955863, + "learning_rate": 0.003, + "loss": 4.1069, + "step": 8770 + }, + { + "epoch": 0.08771, + "grad_norm": 0.7518314170540558, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 8771 + }, + { + "epoch": 0.08772, + "grad_norm": 0.8180447348392333, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 8772 + }, + { + "epoch": 0.08773, + "grad_norm": 0.7147251860859791, + "learning_rate": 0.003, + "loss": 4.1148, + "step": 8773 + }, + { + "epoch": 0.08774, + "grad_norm": 0.5867039223653453, + "learning_rate": 0.003, + "loss": 4.0916, + "step": 8774 + }, + { + "epoch": 0.08775, + "grad_norm": 0.5527229194986593, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 8775 + }, + { + "epoch": 0.08776, + "grad_norm": 0.5844947741413182, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 8776 + }, + { + "epoch": 0.08777, + "grad_norm": 0.6563370684489898, + "learning_rate": 0.003, + "loss": 4.1063, + "step": 8777 + }, + { + "epoch": 0.08778, + "grad_norm": 0.7712524036141793, + "learning_rate": 0.003, + "loss": 4.0946, + "step": 8778 + }, + { + "epoch": 0.08779, + "grad_norm": 0.7807517822630886, + "learning_rate": 0.003, + "loss": 4.1099, + "step": 8779 + }, + { + "epoch": 0.0878, + "grad_norm": 0.6904453164163412, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 8780 + }, + { + "epoch": 0.08781, + "grad_norm": 0.767764779325477, + "learning_rate": 0.003, + "loss": 4.0872, + "step": 8781 + }, + { + "epoch": 0.08782, + "grad_norm": 0.9600307430515294, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 8782 + }, + { + "epoch": 0.08783, + "grad_norm": 1.286330289921701, + "learning_rate": 0.003, + "loss": 4.1144, + "step": 8783 + }, + { + "epoch": 0.08784, + "grad_norm": 0.9521055086549031, + "learning_rate": 0.003, + "loss": 4.0954, + "step": 8784 + }, + { + "epoch": 0.08785, + "grad_norm": 1.3646540084787944, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 8785 + }, + { + "epoch": 0.08786, + "grad_norm": 0.9165165437571451, + "learning_rate": 0.003, + "loss": 4.1215, + "step": 8786 + }, + { + "epoch": 0.08787, + "grad_norm": 0.8674027071829526, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 8787 + }, + { + "epoch": 0.08788, + "grad_norm": 0.8808990624383837, + "learning_rate": 0.003, + "loss": 4.1186, + "step": 8788 + }, + { + "epoch": 0.08789, + "grad_norm": 0.9789530665994413, + "learning_rate": 0.003, + "loss": 4.0907, + "step": 8789 + }, + { + "epoch": 0.0879, + "grad_norm": 1.05003444606044, + "learning_rate": 0.003, + "loss": 4.115, + "step": 8790 + }, + { + "epoch": 0.08791, + "grad_norm": 0.9193191175375216, + "learning_rate": 0.003, + "loss": 4.0954, + "step": 8791 + }, + { + "epoch": 0.08792, + "grad_norm": 0.8116887310714239, + "learning_rate": 0.003, + "loss": 4.1052, + "step": 8792 + }, + { + "epoch": 0.08793, + "grad_norm": 0.8040951439376696, + "learning_rate": 0.003, + "loss": 4.0946, + "step": 8793 + }, + { + "epoch": 0.08794, + "grad_norm": 0.8728936717925508, + "learning_rate": 0.003, + "loss": 4.0891, + "step": 8794 + }, + { + "epoch": 0.08795, + "grad_norm": 0.9425258213709625, + "learning_rate": 0.003, + "loss": 4.0929, + "step": 8795 + }, + { + "epoch": 0.08796, + "grad_norm": 0.9269620549507115, + "learning_rate": 0.003, + "loss": 4.0978, + "step": 8796 + }, + { + "epoch": 0.08797, + "grad_norm": 0.9274274055304775, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 8797 + }, + { + "epoch": 0.08798, + "grad_norm": 1.0495786479990563, + "learning_rate": 0.003, + "loss": 4.1215, + "step": 8798 + }, + { + "epoch": 0.08799, + "grad_norm": 1.0941402572479257, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 8799 + }, + { + "epoch": 0.088, + "grad_norm": 1.0409088434122218, + "learning_rate": 0.003, + "loss": 4.1182, + "step": 8800 + }, + { + "epoch": 0.08801, + "grad_norm": 1.0008095022296148, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 8801 + }, + { + "epoch": 0.08802, + "grad_norm": 0.9260235726665792, + "learning_rate": 0.003, + "loss": 4.1157, + "step": 8802 + }, + { + "epoch": 0.08803, + "grad_norm": 0.959188194655516, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 8803 + }, + { + "epoch": 0.08804, + "grad_norm": 1.0435507338866736, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 8804 + }, + { + "epoch": 0.08805, + "grad_norm": 0.9112940402613895, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 8805 + }, + { + "epoch": 0.08806, + "grad_norm": 0.9708013111525423, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 8806 + }, + { + "epoch": 0.08807, + "grad_norm": 1.0648214390050283, + "learning_rate": 0.003, + "loss": 4.114, + "step": 8807 + }, + { + "epoch": 0.08808, + "grad_norm": 0.996034423121219, + "learning_rate": 0.003, + "loss": 4.0943, + "step": 8808 + }, + { + "epoch": 0.08809, + "grad_norm": 1.049344011106456, + "learning_rate": 0.003, + "loss": 4.0977, + "step": 8809 + }, + { + "epoch": 0.0881, + "grad_norm": 0.8742010348146958, + "learning_rate": 0.003, + "loss": 4.1013, + "step": 8810 + }, + { + "epoch": 0.08811, + "grad_norm": 0.8862670461095695, + "learning_rate": 0.003, + "loss": 4.0965, + "step": 8811 + }, + { + "epoch": 0.08812, + "grad_norm": 0.8360246704981421, + "learning_rate": 0.003, + "loss": 4.1038, + "step": 8812 + }, + { + "epoch": 0.08813, + "grad_norm": 0.8761385565588603, + "learning_rate": 0.003, + "loss": 4.073, + "step": 8813 + }, + { + "epoch": 0.08814, + "grad_norm": 0.8913146587128812, + "learning_rate": 0.003, + "loss": 4.1124, + "step": 8814 + }, + { + "epoch": 0.08815, + "grad_norm": 0.9207388295031492, + "learning_rate": 0.003, + "loss": 4.1102, + "step": 8815 + }, + { + "epoch": 0.08816, + "grad_norm": 1.0838936761521594, + "learning_rate": 0.003, + "loss": 4.1003, + "step": 8816 + }, + { + "epoch": 0.08817, + "grad_norm": 0.967762286433456, + "learning_rate": 0.003, + "loss": 4.082, + "step": 8817 + }, + { + "epoch": 0.08818, + "grad_norm": 0.8962809858840016, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 8818 + }, + { + "epoch": 0.08819, + "grad_norm": 0.7701740177256127, + "learning_rate": 0.003, + "loss": 4.1215, + "step": 8819 + }, + { + "epoch": 0.0882, + "grad_norm": 0.7258766499158136, + "learning_rate": 0.003, + "loss": 4.094, + "step": 8820 + }, + { + "epoch": 0.08821, + "grad_norm": 0.7058532115848066, + "learning_rate": 0.003, + "loss": 4.1208, + "step": 8821 + }, + { + "epoch": 0.08822, + "grad_norm": 0.7289621370989584, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 8822 + }, + { + "epoch": 0.08823, + "grad_norm": 0.9370302189185644, + "learning_rate": 0.003, + "loss": 4.0915, + "step": 8823 + }, + { + "epoch": 0.08824, + "grad_norm": 1.1037702165909034, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 8824 + }, + { + "epoch": 0.08825, + "grad_norm": 0.8276194064577156, + "learning_rate": 0.003, + "loss": 4.0973, + "step": 8825 + }, + { + "epoch": 0.08826, + "grad_norm": 0.7237687402112217, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 8826 + }, + { + "epoch": 0.08827, + "grad_norm": 0.6068281322897773, + "learning_rate": 0.003, + "loss": 4.0962, + "step": 8827 + }, + { + "epoch": 0.08828, + "grad_norm": 0.6236195415565339, + "learning_rate": 0.003, + "loss": 4.098, + "step": 8828 + }, + { + "epoch": 0.08829, + "grad_norm": 0.6501295996843758, + "learning_rate": 0.003, + "loss": 4.0957, + "step": 8829 + }, + { + "epoch": 0.0883, + "grad_norm": 0.7182692035416562, + "learning_rate": 0.003, + "loss": 4.088, + "step": 8830 + }, + { + "epoch": 0.08831, + "grad_norm": 0.825778877664003, + "learning_rate": 0.003, + "loss": 4.0899, + "step": 8831 + }, + { + "epoch": 0.08832, + "grad_norm": 0.8409707101378571, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 8832 + }, + { + "epoch": 0.08833, + "grad_norm": 0.8688218779665827, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 8833 + }, + { + "epoch": 0.08834, + "grad_norm": 1.053985971333072, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 8834 + }, + { + "epoch": 0.08835, + "grad_norm": 1.2011317648997946, + "learning_rate": 0.003, + "loss": 4.1244, + "step": 8835 + }, + { + "epoch": 0.08836, + "grad_norm": 0.8675347336538084, + "learning_rate": 0.003, + "loss": 4.1047, + "step": 8836 + }, + { + "epoch": 0.08837, + "grad_norm": 0.6847724769015546, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 8837 + }, + { + "epoch": 0.08838, + "grad_norm": 0.7664193052446167, + "learning_rate": 0.003, + "loss": 4.0935, + "step": 8838 + }, + { + "epoch": 0.08839, + "grad_norm": 0.7849476506437728, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 8839 + }, + { + "epoch": 0.0884, + "grad_norm": 0.8581420457247539, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 8840 + }, + { + "epoch": 0.08841, + "grad_norm": 1.1437255891991425, + "learning_rate": 0.003, + "loss": 4.1094, + "step": 8841 + }, + { + "epoch": 0.08842, + "grad_norm": 1.0640839104480644, + "learning_rate": 0.003, + "loss": 4.0886, + "step": 8842 + }, + { + "epoch": 0.08843, + "grad_norm": 0.8177173898310196, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 8843 + }, + { + "epoch": 0.08844, + "grad_norm": 0.7006811082685729, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 8844 + }, + { + "epoch": 0.08845, + "grad_norm": 0.6654586857775031, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 8845 + }, + { + "epoch": 0.08846, + "grad_norm": 0.7828442286838105, + "learning_rate": 0.003, + "loss": 4.1247, + "step": 8846 + }, + { + "epoch": 0.08847, + "grad_norm": 1.002596697776703, + "learning_rate": 0.003, + "loss": 4.1223, + "step": 8847 + }, + { + "epoch": 0.08848, + "grad_norm": 1.0769646848289371, + "learning_rate": 0.003, + "loss": 4.1192, + "step": 8848 + }, + { + "epoch": 0.08849, + "grad_norm": 0.9588089835995077, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 8849 + }, + { + "epoch": 0.0885, + "grad_norm": 0.9736041532230283, + "learning_rate": 0.003, + "loss": 4.1066, + "step": 8850 + }, + { + "epoch": 0.08851, + "grad_norm": 1.0038135324969604, + "learning_rate": 0.003, + "loss": 4.1078, + "step": 8851 + }, + { + "epoch": 0.08852, + "grad_norm": 0.8144136857171164, + "learning_rate": 0.003, + "loss": 4.0928, + "step": 8852 + }, + { + "epoch": 0.08853, + "grad_norm": 0.7492869712538301, + "learning_rate": 0.003, + "loss": 4.1103, + "step": 8853 + }, + { + "epoch": 0.08854, + "grad_norm": 0.7964882824236851, + "learning_rate": 0.003, + "loss": 4.1266, + "step": 8854 + }, + { + "epoch": 0.08855, + "grad_norm": 0.8940261132351949, + "learning_rate": 0.003, + "loss": 4.086, + "step": 8855 + }, + { + "epoch": 0.08856, + "grad_norm": 0.954650133885222, + "learning_rate": 0.003, + "loss": 4.1049, + "step": 8856 + }, + { + "epoch": 0.08857, + "grad_norm": 0.9861048909441799, + "learning_rate": 0.003, + "loss": 4.1074, + "step": 8857 + }, + { + "epoch": 0.08858, + "grad_norm": 1.0148434704216471, + "learning_rate": 0.003, + "loss": 4.0939, + "step": 8858 + }, + { + "epoch": 0.08859, + "grad_norm": 1.2410690598958254, + "learning_rate": 0.003, + "loss": 4.1131, + "step": 8859 + }, + { + "epoch": 0.0886, + "grad_norm": 0.9416605613389235, + "learning_rate": 0.003, + "loss": 4.116, + "step": 8860 + }, + { + "epoch": 0.08861, + "grad_norm": 0.8801276255413449, + "learning_rate": 0.003, + "loss": 4.0913, + "step": 8861 + }, + { + "epoch": 0.08862, + "grad_norm": 0.867667458558974, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 8862 + }, + { + "epoch": 0.08863, + "grad_norm": 0.9625185944766314, + "learning_rate": 0.003, + "loss": 4.1077, + "step": 8863 + }, + { + "epoch": 0.08864, + "grad_norm": 1.025113767580424, + "learning_rate": 0.003, + "loss": 4.1198, + "step": 8864 + }, + { + "epoch": 0.08865, + "grad_norm": 1.0305392893083485, + "learning_rate": 0.003, + "loss": 4.1141, + "step": 8865 + }, + { + "epoch": 0.08866, + "grad_norm": 1.0855251168096982, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 8866 + }, + { + "epoch": 0.08867, + "grad_norm": 0.9690901901735053, + "learning_rate": 0.003, + "loss": 4.1148, + "step": 8867 + }, + { + "epoch": 0.08868, + "grad_norm": 0.9930947313572007, + "learning_rate": 0.003, + "loss": 4.095, + "step": 8868 + }, + { + "epoch": 0.08869, + "grad_norm": 1.0329659460825737, + "learning_rate": 0.003, + "loss": 4.1513, + "step": 8869 + }, + { + "epoch": 0.0887, + "grad_norm": 0.962624300853848, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 8870 + }, + { + "epoch": 0.08871, + "grad_norm": 0.8967371579665262, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 8871 + }, + { + "epoch": 0.08872, + "grad_norm": 0.8242663665328195, + "learning_rate": 0.003, + "loss": 4.1344, + "step": 8872 + }, + { + "epoch": 0.08873, + "grad_norm": 0.8080330802630397, + "learning_rate": 0.003, + "loss": 4.087, + "step": 8873 + }, + { + "epoch": 0.08874, + "grad_norm": 0.8320866327498992, + "learning_rate": 0.003, + "loss": 4.0975, + "step": 8874 + }, + { + "epoch": 0.08875, + "grad_norm": 0.8237590696825445, + "learning_rate": 0.003, + "loss": 4.1174, + "step": 8875 + }, + { + "epoch": 0.08876, + "grad_norm": 0.7670191226856026, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 8876 + }, + { + "epoch": 0.08877, + "grad_norm": 0.8150234000038336, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 8877 + }, + { + "epoch": 0.08878, + "grad_norm": 0.9188804850843868, + "learning_rate": 0.003, + "loss": 4.0976, + "step": 8878 + }, + { + "epoch": 0.08879, + "grad_norm": 1.1505868854127517, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 8879 + }, + { + "epoch": 0.0888, + "grad_norm": 1.0381859644112361, + "learning_rate": 0.003, + "loss": 4.086, + "step": 8880 + }, + { + "epoch": 0.08881, + "grad_norm": 0.8126207042081777, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 8881 + }, + { + "epoch": 0.08882, + "grad_norm": 0.6375023727555839, + "learning_rate": 0.003, + "loss": 4.0995, + "step": 8882 + }, + { + "epoch": 0.08883, + "grad_norm": 0.8121388182245645, + "learning_rate": 0.003, + "loss": 4.1415, + "step": 8883 + }, + { + "epoch": 0.08884, + "grad_norm": 1.0502326054062787, + "learning_rate": 0.003, + "loss": 4.0997, + "step": 8884 + }, + { + "epoch": 0.08885, + "grad_norm": 1.2283294749042548, + "learning_rate": 0.003, + "loss": 4.0973, + "step": 8885 + }, + { + "epoch": 0.08886, + "grad_norm": 0.749971714680993, + "learning_rate": 0.003, + "loss": 4.0972, + "step": 8886 + }, + { + "epoch": 0.08887, + "grad_norm": 0.8691386698939016, + "learning_rate": 0.003, + "loss": 4.0913, + "step": 8887 + }, + { + "epoch": 0.08888, + "grad_norm": 0.9459570536683495, + "learning_rate": 0.003, + "loss": 4.0887, + "step": 8888 + }, + { + "epoch": 0.08889, + "grad_norm": 1.0034071726990192, + "learning_rate": 0.003, + "loss": 4.0901, + "step": 8889 + }, + { + "epoch": 0.0889, + "grad_norm": 1.1036828603076412, + "learning_rate": 0.003, + "loss": 4.0931, + "step": 8890 + }, + { + "epoch": 0.08891, + "grad_norm": 0.7547248732778705, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 8891 + }, + { + "epoch": 0.08892, + "grad_norm": 0.7125455249994503, + "learning_rate": 0.003, + "loss": 4.1096, + "step": 8892 + }, + { + "epoch": 0.08893, + "grad_norm": 0.9110389886249487, + "learning_rate": 0.003, + "loss": 4.0871, + "step": 8893 + }, + { + "epoch": 0.08894, + "grad_norm": 1.0085325989909744, + "learning_rate": 0.003, + "loss": 4.079, + "step": 8894 + }, + { + "epoch": 0.08895, + "grad_norm": 0.9395630146430425, + "learning_rate": 0.003, + "loss": 4.0917, + "step": 8895 + }, + { + "epoch": 0.08896, + "grad_norm": 0.932005587780262, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 8896 + }, + { + "epoch": 0.08897, + "grad_norm": 0.9683101599385363, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 8897 + }, + { + "epoch": 0.08898, + "grad_norm": 0.8778233714006216, + "learning_rate": 0.003, + "loss": 4.0975, + "step": 8898 + }, + { + "epoch": 0.08899, + "grad_norm": 0.8843104049066195, + "learning_rate": 0.003, + "loss": 4.0961, + "step": 8899 + }, + { + "epoch": 0.089, + "grad_norm": 0.8611736022905074, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 8900 + }, + { + "epoch": 0.08901, + "grad_norm": 0.7675741709312146, + "learning_rate": 0.003, + "loss": 4.0967, + "step": 8901 + }, + { + "epoch": 0.08902, + "grad_norm": 0.7542143956414811, + "learning_rate": 0.003, + "loss": 4.0963, + "step": 8902 + }, + { + "epoch": 0.08903, + "grad_norm": 0.824143498268071, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 8903 + }, + { + "epoch": 0.08904, + "grad_norm": 1.0036741311146173, + "learning_rate": 0.003, + "loss": 4.1018, + "step": 8904 + }, + { + "epoch": 0.08905, + "grad_norm": 1.1665427882326276, + "learning_rate": 0.003, + "loss": 4.1198, + "step": 8905 + }, + { + "epoch": 0.08906, + "grad_norm": 0.9167446591164601, + "learning_rate": 0.003, + "loss": 4.1021, + "step": 8906 + }, + { + "epoch": 0.08907, + "grad_norm": 0.8151193889750405, + "learning_rate": 0.003, + "loss": 4.0966, + "step": 8907 + }, + { + "epoch": 0.08908, + "grad_norm": 0.8148807348539009, + "learning_rate": 0.003, + "loss": 4.116, + "step": 8908 + }, + { + "epoch": 0.08909, + "grad_norm": 0.8006264904937266, + "learning_rate": 0.003, + "loss": 4.1071, + "step": 8909 + }, + { + "epoch": 0.0891, + "grad_norm": 0.7847183342716237, + "learning_rate": 0.003, + "loss": 4.0888, + "step": 8910 + }, + { + "epoch": 0.08911, + "grad_norm": 0.871137922082431, + "learning_rate": 0.003, + "loss": 4.0945, + "step": 8911 + }, + { + "epoch": 0.08912, + "grad_norm": 0.9186358754285994, + "learning_rate": 0.003, + "loss": 4.0941, + "step": 8912 + }, + { + "epoch": 0.08913, + "grad_norm": 0.8792963429484404, + "learning_rate": 0.003, + "loss": 4.1138, + "step": 8913 + }, + { + "epoch": 0.08914, + "grad_norm": 0.8389522069822098, + "learning_rate": 0.003, + "loss": 4.1074, + "step": 8914 + }, + { + "epoch": 0.08915, + "grad_norm": 0.7438660035575635, + "learning_rate": 0.003, + "loss": 4.111, + "step": 8915 + }, + { + "epoch": 0.08916, + "grad_norm": 0.81031482315623, + "learning_rate": 0.003, + "loss": 4.0994, + "step": 8916 + }, + { + "epoch": 0.08917, + "grad_norm": 0.9879431715632052, + "learning_rate": 0.003, + "loss": 4.0864, + "step": 8917 + }, + { + "epoch": 0.08918, + "grad_norm": 1.1263164984050484, + "learning_rate": 0.003, + "loss": 4.1033, + "step": 8918 + }, + { + "epoch": 0.08919, + "grad_norm": 0.8217791465004585, + "learning_rate": 0.003, + "loss": 4.1052, + "step": 8919 + }, + { + "epoch": 0.0892, + "grad_norm": 0.838741117076949, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 8920 + }, + { + "epoch": 0.08921, + "grad_norm": 0.8013123325103196, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 8921 + }, + { + "epoch": 0.08922, + "grad_norm": 0.8829008386490941, + "learning_rate": 0.003, + "loss": 4.1102, + "step": 8922 + }, + { + "epoch": 0.08923, + "grad_norm": 0.831010513545994, + "learning_rate": 0.003, + "loss": 4.1134, + "step": 8923 + }, + { + "epoch": 0.08924, + "grad_norm": 0.8026665557631941, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 8924 + }, + { + "epoch": 0.08925, + "grad_norm": 0.7516895419563575, + "learning_rate": 0.003, + "loss": 4.0955, + "step": 8925 + }, + { + "epoch": 0.08926, + "grad_norm": 0.6859708086212067, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 8926 + }, + { + "epoch": 0.08927, + "grad_norm": 0.705876228465091, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 8927 + }, + { + "epoch": 0.08928, + "grad_norm": 0.8197442137549819, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 8928 + }, + { + "epoch": 0.08929, + "grad_norm": 1.158197035763089, + "learning_rate": 0.003, + "loss": 4.0983, + "step": 8929 + }, + { + "epoch": 0.0893, + "grad_norm": 1.2835489288131519, + "learning_rate": 0.003, + "loss": 4.1078, + "step": 8930 + }, + { + "epoch": 0.08931, + "grad_norm": 0.9255470549413722, + "learning_rate": 0.003, + "loss": 4.0873, + "step": 8931 + }, + { + "epoch": 0.08932, + "grad_norm": 0.9494109894231941, + "learning_rate": 0.003, + "loss": 4.1048, + "step": 8932 + }, + { + "epoch": 0.08933, + "grad_norm": 0.9325764456947372, + "learning_rate": 0.003, + "loss": 4.0954, + "step": 8933 + }, + { + "epoch": 0.08934, + "grad_norm": 0.9120790589940708, + "learning_rate": 0.003, + "loss": 4.0916, + "step": 8934 + }, + { + "epoch": 0.08935, + "grad_norm": 0.9622123789895787, + "learning_rate": 0.003, + "loss": 4.106, + "step": 8935 + }, + { + "epoch": 0.08936, + "grad_norm": 0.9982089757464768, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 8936 + }, + { + "epoch": 0.08937, + "grad_norm": 0.8488183553743273, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 8937 + }, + { + "epoch": 0.08938, + "grad_norm": 0.711168048268646, + "learning_rate": 0.003, + "loss": 4.106, + "step": 8938 + }, + { + "epoch": 0.08939, + "grad_norm": 0.7145069662449071, + "learning_rate": 0.003, + "loss": 4.1228, + "step": 8939 + }, + { + "epoch": 0.0894, + "grad_norm": 0.9351641300422199, + "learning_rate": 0.003, + "loss": 4.0905, + "step": 8940 + }, + { + "epoch": 0.08941, + "grad_norm": 0.9908374405712305, + "learning_rate": 0.003, + "loss": 4.1068, + "step": 8941 + }, + { + "epoch": 0.08942, + "grad_norm": 1.0369857459287184, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 8942 + }, + { + "epoch": 0.08943, + "grad_norm": 0.9815132194398201, + "learning_rate": 0.003, + "loss": 4.1022, + "step": 8943 + }, + { + "epoch": 0.08944, + "grad_norm": 0.8275387834639889, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 8944 + }, + { + "epoch": 0.08945, + "grad_norm": 0.7363945481987325, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 8945 + }, + { + "epoch": 0.08946, + "grad_norm": 0.7493687013662418, + "learning_rate": 0.003, + "loss": 4.1161, + "step": 8946 + }, + { + "epoch": 0.08947, + "grad_norm": 0.8635191868013845, + "learning_rate": 0.003, + "loss": 4.0948, + "step": 8947 + }, + { + "epoch": 0.08948, + "grad_norm": 0.8462786471943283, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 8948 + }, + { + "epoch": 0.08949, + "grad_norm": 0.8671422196356308, + "learning_rate": 0.003, + "loss": 4.0903, + "step": 8949 + }, + { + "epoch": 0.0895, + "grad_norm": 0.9409364514795033, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 8950 + }, + { + "epoch": 0.08951, + "grad_norm": 1.090641876171059, + "learning_rate": 0.003, + "loss": 4.0984, + "step": 8951 + }, + { + "epoch": 0.08952, + "grad_norm": 1.0606427376589673, + "learning_rate": 0.003, + "loss": 4.1129, + "step": 8952 + }, + { + "epoch": 0.08953, + "grad_norm": 0.9918378904457558, + "learning_rate": 0.003, + "loss": 4.1043, + "step": 8953 + }, + { + "epoch": 0.08954, + "grad_norm": 0.9774618750373327, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 8954 + }, + { + "epoch": 0.08955, + "grad_norm": 1.2355673972258312, + "learning_rate": 0.003, + "loss": 4.1081, + "step": 8955 + }, + { + "epoch": 0.08956, + "grad_norm": 0.958337187094943, + "learning_rate": 0.003, + "loss": 4.1043, + "step": 8956 + }, + { + "epoch": 0.08957, + "grad_norm": 1.1059278113358553, + "learning_rate": 0.003, + "loss": 4.1176, + "step": 8957 + }, + { + "epoch": 0.08958, + "grad_norm": 1.0600369480191711, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 8958 + }, + { + "epoch": 0.08959, + "grad_norm": 0.9044461430911768, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 8959 + }, + { + "epoch": 0.0896, + "grad_norm": 0.751672034162579, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 8960 + }, + { + "epoch": 0.08961, + "grad_norm": 0.6890812720881869, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 8961 + }, + { + "epoch": 0.08962, + "grad_norm": 0.6420758804789315, + "learning_rate": 0.003, + "loss": 4.0848, + "step": 8962 + }, + { + "epoch": 0.08963, + "grad_norm": 0.7715824227745989, + "learning_rate": 0.003, + "loss": 4.119, + "step": 8963 + }, + { + "epoch": 0.08964, + "grad_norm": 0.9769914979893802, + "learning_rate": 0.003, + "loss": 4.1009, + "step": 8964 + }, + { + "epoch": 0.08965, + "grad_norm": 1.1055411204729655, + "learning_rate": 0.003, + "loss": 4.1159, + "step": 8965 + }, + { + "epoch": 0.08966, + "grad_norm": 0.8858925381363988, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 8966 + }, + { + "epoch": 0.08967, + "grad_norm": 0.973405476101832, + "learning_rate": 0.003, + "loss": 4.125, + "step": 8967 + }, + { + "epoch": 0.08968, + "grad_norm": 1.246488899818171, + "learning_rate": 0.003, + "loss": 4.1068, + "step": 8968 + }, + { + "epoch": 0.08969, + "grad_norm": 0.7919830012222802, + "learning_rate": 0.003, + "loss": 4.092, + "step": 8969 + }, + { + "epoch": 0.0897, + "grad_norm": 0.783905488272791, + "learning_rate": 0.003, + "loss": 4.1039, + "step": 8970 + }, + { + "epoch": 0.08971, + "grad_norm": 0.8587360970358755, + "learning_rate": 0.003, + "loss": 4.0975, + "step": 8971 + }, + { + "epoch": 0.08972, + "grad_norm": 1.1812204010197107, + "learning_rate": 0.003, + "loss": 4.0953, + "step": 8972 + }, + { + "epoch": 0.08973, + "grad_norm": 1.072340183548216, + "learning_rate": 0.003, + "loss": 4.1285, + "step": 8973 + }, + { + "epoch": 0.08974, + "grad_norm": 0.9624889318642255, + "learning_rate": 0.003, + "loss": 4.1191, + "step": 8974 + }, + { + "epoch": 0.08975, + "grad_norm": 1.103329859592797, + "learning_rate": 0.003, + "loss": 4.1042, + "step": 8975 + }, + { + "epoch": 0.08976, + "grad_norm": 0.9226100631400868, + "learning_rate": 0.003, + "loss": 4.1147, + "step": 8976 + }, + { + "epoch": 0.08977, + "grad_norm": 0.9050852208024028, + "learning_rate": 0.003, + "loss": 4.0989, + "step": 8977 + }, + { + "epoch": 0.08978, + "grad_norm": 0.9593748971336933, + "learning_rate": 0.003, + "loss": 4.1056, + "step": 8978 + }, + { + "epoch": 0.08979, + "grad_norm": 0.917813099791213, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 8979 + }, + { + "epoch": 0.0898, + "grad_norm": 0.850359408142017, + "learning_rate": 0.003, + "loss": 4.1055, + "step": 8980 + }, + { + "epoch": 0.08981, + "grad_norm": 0.7873512624929737, + "learning_rate": 0.003, + "loss": 4.08, + "step": 8981 + }, + { + "epoch": 0.08982, + "grad_norm": 0.7460904455327514, + "learning_rate": 0.003, + "loss": 4.0876, + "step": 8982 + }, + { + "epoch": 0.08983, + "grad_norm": 0.813398130626127, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 8983 + }, + { + "epoch": 0.08984, + "grad_norm": 1.0379519181428254, + "learning_rate": 0.003, + "loss": 4.1201, + "step": 8984 + }, + { + "epoch": 0.08985, + "grad_norm": 1.2369546672505778, + "learning_rate": 0.003, + "loss": 4.107, + "step": 8985 + }, + { + "epoch": 0.08986, + "grad_norm": 0.7482892312093569, + "learning_rate": 0.003, + "loss": 4.1, + "step": 8986 + }, + { + "epoch": 0.08987, + "grad_norm": 0.6543486223493022, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 8987 + }, + { + "epoch": 0.08988, + "grad_norm": 0.6451367251570368, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 8988 + }, + { + "epoch": 0.08989, + "grad_norm": 0.601214981373186, + "learning_rate": 0.003, + "loss": 4.1077, + "step": 8989 + }, + { + "epoch": 0.0899, + "grad_norm": 0.670157839844488, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 8990 + }, + { + "epoch": 0.08991, + "grad_norm": 0.6344666888841743, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 8991 + }, + { + "epoch": 0.08992, + "grad_norm": 0.6087663409550896, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 8992 + }, + { + "epoch": 0.08993, + "grad_norm": 0.6007106864127649, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 8993 + }, + { + "epoch": 0.08994, + "grad_norm": 0.6569404032054016, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 8994 + }, + { + "epoch": 0.08995, + "grad_norm": 0.8011788589422568, + "learning_rate": 0.003, + "loss": 4.0907, + "step": 8995 + }, + { + "epoch": 0.08996, + "grad_norm": 1.090537283208205, + "learning_rate": 0.003, + "loss": 4.1194, + "step": 8996 + }, + { + "epoch": 0.08997, + "grad_norm": 1.259139252903192, + "learning_rate": 0.003, + "loss": 4.096, + "step": 8997 + }, + { + "epoch": 0.08998, + "grad_norm": 0.7711113115206871, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 8998 + }, + { + "epoch": 0.08999, + "grad_norm": 0.7350565093378323, + "learning_rate": 0.003, + "loss": 4.0998, + "step": 8999 + }, + { + "epoch": 0.09, + "grad_norm": 0.8212463674203402, + "learning_rate": 0.003, + "loss": 4.093, + "step": 9000 + }, + { + "epoch": 0.09001, + "grad_norm": 0.9909623667562864, + "learning_rate": 0.003, + "loss": 4.0903, + "step": 9001 + }, + { + "epoch": 0.09002, + "grad_norm": 1.1337359570952938, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 9002 + }, + { + "epoch": 0.09003, + "grad_norm": 1.066025616693887, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 9003 + }, + { + "epoch": 0.09004, + "grad_norm": 1.022429577767667, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 9004 + }, + { + "epoch": 0.09005, + "grad_norm": 0.8748962510761829, + "learning_rate": 0.003, + "loss": 4.1092, + "step": 9005 + }, + { + "epoch": 0.09006, + "grad_norm": 0.7098491602250373, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 9006 + }, + { + "epoch": 0.09007, + "grad_norm": 0.5692107949898508, + "learning_rate": 0.003, + "loss": 4.1054, + "step": 9007 + }, + { + "epoch": 0.09008, + "grad_norm": 0.5772606567632318, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 9008 + }, + { + "epoch": 0.09009, + "grad_norm": 0.6669070549022329, + "learning_rate": 0.003, + "loss": 4.0955, + "step": 9009 + }, + { + "epoch": 0.0901, + "grad_norm": 0.8245290219344589, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 9010 + }, + { + "epoch": 0.09011, + "grad_norm": 0.9106422213480005, + "learning_rate": 0.003, + "loss": 4.0857, + "step": 9011 + }, + { + "epoch": 0.09012, + "grad_norm": 0.9526910382185314, + "learning_rate": 0.003, + "loss": 4.0873, + "step": 9012 + }, + { + "epoch": 0.09013, + "grad_norm": 0.9182834414849824, + "learning_rate": 0.003, + "loss": 4.081, + "step": 9013 + }, + { + "epoch": 0.09014, + "grad_norm": 0.8705840436567862, + "learning_rate": 0.003, + "loss": 4.1167, + "step": 9014 + }, + { + "epoch": 0.09015, + "grad_norm": 0.9199037787555114, + "learning_rate": 0.003, + "loss": 4.1304, + "step": 9015 + }, + { + "epoch": 0.09016, + "grad_norm": 0.8979198281963656, + "learning_rate": 0.003, + "loss": 4.1133, + "step": 9016 + }, + { + "epoch": 0.09017, + "grad_norm": 0.9760173547686396, + "learning_rate": 0.003, + "loss": 4.122, + "step": 9017 + }, + { + "epoch": 0.09018, + "grad_norm": 1.0721385512227886, + "learning_rate": 0.003, + "loss": 4.1147, + "step": 9018 + }, + { + "epoch": 0.09019, + "grad_norm": 0.9082460234466702, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 9019 + }, + { + "epoch": 0.0902, + "grad_norm": 0.9401369695786034, + "learning_rate": 0.003, + "loss": 4.0984, + "step": 9020 + }, + { + "epoch": 0.09021, + "grad_norm": 0.9948251941982775, + "learning_rate": 0.003, + "loss": 4.1138, + "step": 9021 + }, + { + "epoch": 0.09022, + "grad_norm": 1.1842814118486442, + "learning_rate": 0.003, + "loss": 4.1091, + "step": 9022 + }, + { + "epoch": 0.09023, + "grad_norm": 0.7972897768988289, + "learning_rate": 0.003, + "loss": 4.0942, + "step": 9023 + }, + { + "epoch": 0.09024, + "grad_norm": 0.7421033982019426, + "learning_rate": 0.003, + "loss": 4.079, + "step": 9024 + }, + { + "epoch": 0.09025, + "grad_norm": 0.7738259101346501, + "learning_rate": 0.003, + "loss": 4.0991, + "step": 9025 + }, + { + "epoch": 0.09026, + "grad_norm": 0.8970150206090364, + "learning_rate": 0.003, + "loss": 4.0937, + "step": 9026 + }, + { + "epoch": 0.09027, + "grad_norm": 1.0492341252458535, + "learning_rate": 0.003, + "loss": 4.0907, + "step": 9027 + }, + { + "epoch": 0.09028, + "grad_norm": 1.2454585937869171, + "learning_rate": 0.003, + "loss": 4.1123, + "step": 9028 + }, + { + "epoch": 0.09029, + "grad_norm": 0.9252985895422703, + "learning_rate": 0.003, + "loss": 4.0959, + "step": 9029 + }, + { + "epoch": 0.0903, + "grad_norm": 0.9241514183975191, + "learning_rate": 0.003, + "loss": 4.1248, + "step": 9030 + }, + { + "epoch": 0.09031, + "grad_norm": 0.8763169048046952, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 9031 + }, + { + "epoch": 0.09032, + "grad_norm": 0.9386352079429895, + "learning_rate": 0.003, + "loss": 4.1293, + "step": 9032 + }, + { + "epoch": 0.09033, + "grad_norm": 0.9783200260031812, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 9033 + }, + { + "epoch": 0.09034, + "grad_norm": 0.956003427831922, + "learning_rate": 0.003, + "loss": 4.0942, + "step": 9034 + }, + { + "epoch": 0.09035, + "grad_norm": 0.8041920674510835, + "learning_rate": 0.003, + "loss": 4.0918, + "step": 9035 + }, + { + "epoch": 0.09036, + "grad_norm": 0.8190103140470621, + "learning_rate": 0.003, + "loss": 4.0878, + "step": 9036 + }, + { + "epoch": 0.09037, + "grad_norm": 0.8735918968075451, + "learning_rate": 0.003, + "loss": 4.0966, + "step": 9037 + }, + { + "epoch": 0.09038, + "grad_norm": 0.8950344717934043, + "learning_rate": 0.003, + "loss": 4.0977, + "step": 9038 + }, + { + "epoch": 0.09039, + "grad_norm": 0.8598737751053982, + "learning_rate": 0.003, + "loss": 4.1071, + "step": 9039 + }, + { + "epoch": 0.0904, + "grad_norm": 0.909321180715369, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 9040 + }, + { + "epoch": 0.09041, + "grad_norm": 0.8046317272372915, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 9041 + }, + { + "epoch": 0.09042, + "grad_norm": 0.7635821438024983, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 9042 + }, + { + "epoch": 0.09043, + "grad_norm": 0.9240410411420402, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 9043 + }, + { + "epoch": 0.09044, + "grad_norm": 1.15129774565637, + "learning_rate": 0.003, + "loss": 4.0891, + "step": 9044 + }, + { + "epoch": 0.09045, + "grad_norm": 1.0176507372571526, + "learning_rate": 0.003, + "loss": 4.0963, + "step": 9045 + }, + { + "epoch": 0.09046, + "grad_norm": 1.0168928227228877, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 9046 + }, + { + "epoch": 0.09047, + "grad_norm": 1.0855871224553029, + "learning_rate": 0.003, + "loss": 4.0938, + "step": 9047 + }, + { + "epoch": 0.09048, + "grad_norm": 0.9615447887028904, + "learning_rate": 0.003, + "loss": 4.0888, + "step": 9048 + }, + { + "epoch": 0.09049, + "grad_norm": 0.9780305253510888, + "learning_rate": 0.003, + "loss": 4.1164, + "step": 9049 + }, + { + "epoch": 0.0905, + "grad_norm": 0.919712660817147, + "learning_rate": 0.003, + "loss": 4.0926, + "step": 9050 + }, + { + "epoch": 0.09051, + "grad_norm": 0.9279054815642093, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 9051 + }, + { + "epoch": 0.09052, + "grad_norm": 0.9418205046994602, + "learning_rate": 0.003, + "loss": 4.0995, + "step": 9052 + }, + { + "epoch": 0.09053, + "grad_norm": 0.9469408829030873, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 9053 + }, + { + "epoch": 0.09054, + "grad_norm": 0.9051974077534375, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 9054 + }, + { + "epoch": 0.09055, + "grad_norm": 1.024878799010692, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 9055 + }, + { + "epoch": 0.09056, + "grad_norm": 0.9922488913468747, + "learning_rate": 0.003, + "loss": 4.0868, + "step": 9056 + }, + { + "epoch": 0.09057, + "grad_norm": 0.881282710691066, + "learning_rate": 0.003, + "loss": 4.1119, + "step": 9057 + }, + { + "epoch": 0.09058, + "grad_norm": 0.9505568921531226, + "learning_rate": 0.003, + "loss": 4.1212, + "step": 9058 + }, + { + "epoch": 0.09059, + "grad_norm": 0.9853808181646652, + "learning_rate": 0.003, + "loss": 4.1086, + "step": 9059 + }, + { + "epoch": 0.0906, + "grad_norm": 1.085605727215449, + "learning_rate": 0.003, + "loss": 4.1289, + "step": 9060 + }, + { + "epoch": 0.09061, + "grad_norm": 0.9708980429037549, + "learning_rate": 0.003, + "loss": 4.1086, + "step": 9061 + }, + { + "epoch": 0.09062, + "grad_norm": 0.9437630749155476, + "learning_rate": 0.003, + "loss": 4.077, + "step": 9062 + }, + { + "epoch": 0.09063, + "grad_norm": 1.15349103657829, + "learning_rate": 0.003, + "loss": 4.0964, + "step": 9063 + }, + { + "epoch": 0.09064, + "grad_norm": 0.9708247162873537, + "learning_rate": 0.003, + "loss": 4.1191, + "step": 9064 + }, + { + "epoch": 0.09065, + "grad_norm": 0.8902413777847998, + "learning_rate": 0.003, + "loss": 4.1096, + "step": 9065 + }, + { + "epoch": 0.09066, + "grad_norm": 0.737181277087745, + "learning_rate": 0.003, + "loss": 4.0956, + "step": 9066 + }, + { + "epoch": 0.09067, + "grad_norm": 0.6279056212510534, + "learning_rate": 0.003, + "loss": 4.0873, + "step": 9067 + }, + { + "epoch": 0.09068, + "grad_norm": 0.6385574242216696, + "learning_rate": 0.003, + "loss": 4.0991, + "step": 9068 + }, + { + "epoch": 0.09069, + "grad_norm": 0.6483082983150563, + "learning_rate": 0.003, + "loss": 4.1367, + "step": 9069 + }, + { + "epoch": 0.0907, + "grad_norm": 0.8253555420224653, + "learning_rate": 0.003, + "loss": 4.1165, + "step": 9070 + }, + { + "epoch": 0.09071, + "grad_norm": 1.0346138256073594, + "learning_rate": 0.003, + "loss": 4.1003, + "step": 9071 + }, + { + "epoch": 0.09072, + "grad_norm": 0.9313433051670195, + "learning_rate": 0.003, + "loss": 4.074, + "step": 9072 + }, + { + "epoch": 0.09073, + "grad_norm": 1.0276295171442573, + "learning_rate": 0.003, + "loss": 4.1075, + "step": 9073 + }, + { + "epoch": 0.09074, + "grad_norm": 1.1532562018862866, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 9074 + }, + { + "epoch": 0.09075, + "grad_norm": 1.0895198666143424, + "learning_rate": 0.003, + "loss": 4.1231, + "step": 9075 + }, + { + "epoch": 0.09076, + "grad_norm": 0.9624509207513962, + "learning_rate": 0.003, + "loss": 4.1227, + "step": 9076 + }, + { + "epoch": 0.09077, + "grad_norm": 1.0339320521373805, + "learning_rate": 0.003, + "loss": 4.1254, + "step": 9077 + }, + { + "epoch": 0.09078, + "grad_norm": 1.0567530098664932, + "learning_rate": 0.003, + "loss": 4.1014, + "step": 9078 + }, + { + "epoch": 0.09079, + "grad_norm": 0.9774925721478938, + "learning_rate": 0.003, + "loss": 4.1089, + "step": 9079 + }, + { + "epoch": 0.0908, + "grad_norm": 1.0195847641854132, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 9080 + }, + { + "epoch": 0.09081, + "grad_norm": 1.0092501737148587, + "learning_rate": 0.003, + "loss": 4.1127, + "step": 9081 + }, + { + "epoch": 0.09082, + "grad_norm": 1.2418266763070203, + "learning_rate": 0.003, + "loss": 4.0955, + "step": 9082 + }, + { + "epoch": 0.09083, + "grad_norm": 0.7841216118125688, + "learning_rate": 0.003, + "loss": 4.1031, + "step": 9083 + }, + { + "epoch": 0.09084, + "grad_norm": 0.6720714850940325, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 9084 + }, + { + "epoch": 0.09085, + "grad_norm": 0.8612009953168612, + "learning_rate": 0.003, + "loss": 4.1132, + "step": 9085 + }, + { + "epoch": 0.09086, + "grad_norm": 1.0531662786767666, + "learning_rate": 0.003, + "loss": 4.087, + "step": 9086 + }, + { + "epoch": 0.09087, + "grad_norm": 1.0678011915356584, + "learning_rate": 0.003, + "loss": 4.0907, + "step": 9087 + }, + { + "epoch": 0.09088, + "grad_norm": 0.8076834182848771, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 9088 + }, + { + "epoch": 0.09089, + "grad_norm": 0.7750777971721545, + "learning_rate": 0.003, + "loss": 4.0956, + "step": 9089 + }, + { + "epoch": 0.0909, + "grad_norm": 0.8512525308462379, + "learning_rate": 0.003, + "loss": 4.0864, + "step": 9090 + }, + { + "epoch": 0.09091, + "grad_norm": 1.0582316561541651, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 9091 + }, + { + "epoch": 0.09092, + "grad_norm": 0.898113597368355, + "learning_rate": 0.003, + "loss": 4.1276, + "step": 9092 + }, + { + "epoch": 0.09093, + "grad_norm": 0.8629700654684181, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 9093 + }, + { + "epoch": 0.09094, + "grad_norm": 0.8939556429591207, + "learning_rate": 0.003, + "loss": 4.0975, + "step": 9094 + }, + { + "epoch": 0.09095, + "grad_norm": 0.8038507515784818, + "learning_rate": 0.003, + "loss": 4.1036, + "step": 9095 + }, + { + "epoch": 0.09096, + "grad_norm": 0.7779918254246555, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 9096 + }, + { + "epoch": 0.09097, + "grad_norm": 0.7898567013017581, + "learning_rate": 0.003, + "loss": 4.0962, + "step": 9097 + }, + { + "epoch": 0.09098, + "grad_norm": 0.7469344511651337, + "learning_rate": 0.003, + "loss": 4.068, + "step": 9098 + }, + { + "epoch": 0.09099, + "grad_norm": 0.81782758465319, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 9099 + }, + { + "epoch": 0.091, + "grad_norm": 0.8663932475321643, + "learning_rate": 0.003, + "loss": 4.071, + "step": 9100 + }, + { + "epoch": 0.09101, + "grad_norm": 0.9040908007482058, + "learning_rate": 0.003, + "loss": 4.0922, + "step": 9101 + }, + { + "epoch": 0.09102, + "grad_norm": 0.8974025166455672, + "learning_rate": 0.003, + "loss": 4.1023, + "step": 9102 + }, + { + "epoch": 0.09103, + "grad_norm": 0.8432481209256596, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 9103 + }, + { + "epoch": 0.09104, + "grad_norm": 0.7839104837670768, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 9104 + }, + { + "epoch": 0.09105, + "grad_norm": 0.74968201955386, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 9105 + }, + { + "epoch": 0.09106, + "grad_norm": 0.862192217242314, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 9106 + }, + { + "epoch": 0.09107, + "grad_norm": 0.8987410447628759, + "learning_rate": 0.003, + "loss": 4.0857, + "step": 9107 + }, + { + "epoch": 0.09108, + "grad_norm": 0.834199691676135, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 9108 + }, + { + "epoch": 0.09109, + "grad_norm": 0.8002673934636545, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 9109 + }, + { + "epoch": 0.0911, + "grad_norm": 0.8991637907546197, + "learning_rate": 0.003, + "loss": 4.0872, + "step": 9110 + }, + { + "epoch": 0.09111, + "grad_norm": 1.0150826030428626, + "learning_rate": 0.003, + "loss": 4.1061, + "step": 9111 + }, + { + "epoch": 0.09112, + "grad_norm": 1.0652838006542034, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 9112 + }, + { + "epoch": 0.09113, + "grad_norm": 0.8745809726912822, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 9113 + }, + { + "epoch": 0.09114, + "grad_norm": 0.891905966917683, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 9114 + }, + { + "epoch": 0.09115, + "grad_norm": 1.0521275368839003, + "learning_rate": 0.003, + "loss": 4.0925, + "step": 9115 + }, + { + "epoch": 0.09116, + "grad_norm": 0.8815584303045687, + "learning_rate": 0.003, + "loss": 4.1368, + "step": 9116 + }, + { + "epoch": 0.09117, + "grad_norm": 1.0372762229109376, + "learning_rate": 0.003, + "loss": 4.1088, + "step": 9117 + }, + { + "epoch": 0.09118, + "grad_norm": 0.9807823101415436, + "learning_rate": 0.003, + "loss": 4.1009, + "step": 9118 + }, + { + "epoch": 0.09119, + "grad_norm": 1.0463681239596228, + "learning_rate": 0.003, + "loss": 4.0978, + "step": 9119 + }, + { + "epoch": 0.0912, + "grad_norm": 1.1558735058794685, + "learning_rate": 0.003, + "loss": 4.117, + "step": 9120 + }, + { + "epoch": 0.09121, + "grad_norm": 0.9471193935945554, + "learning_rate": 0.003, + "loss": 4.1303, + "step": 9121 + }, + { + "epoch": 0.09122, + "grad_norm": 0.903406256690084, + "learning_rate": 0.003, + "loss": 4.1164, + "step": 9122 + }, + { + "epoch": 0.09123, + "grad_norm": 0.9082666157548727, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 9123 + }, + { + "epoch": 0.09124, + "grad_norm": 1.0288774290050668, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 9124 + }, + { + "epoch": 0.09125, + "grad_norm": 1.1563589417293623, + "learning_rate": 0.003, + "loss": 4.1232, + "step": 9125 + }, + { + "epoch": 0.09126, + "grad_norm": 0.9664619902252881, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 9126 + }, + { + "epoch": 0.09127, + "grad_norm": 0.9629651919979498, + "learning_rate": 0.003, + "loss": 4.1087, + "step": 9127 + }, + { + "epoch": 0.09128, + "grad_norm": 1.0135807050519996, + "learning_rate": 0.003, + "loss": 4.103, + "step": 9128 + }, + { + "epoch": 0.09129, + "grad_norm": 1.1527226810083393, + "learning_rate": 0.003, + "loss": 4.0973, + "step": 9129 + }, + { + "epoch": 0.0913, + "grad_norm": 0.8454458533144666, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 9130 + }, + { + "epoch": 0.09131, + "grad_norm": 0.6217315912216276, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 9131 + }, + { + "epoch": 0.09132, + "grad_norm": 0.6894799817711807, + "learning_rate": 0.003, + "loss": 4.1128, + "step": 9132 + }, + { + "epoch": 0.09133, + "grad_norm": 0.9164420443264122, + "learning_rate": 0.003, + "loss": 4.0936, + "step": 9133 + }, + { + "epoch": 0.09134, + "grad_norm": 1.118137516546724, + "learning_rate": 0.003, + "loss": 4.0987, + "step": 9134 + }, + { + "epoch": 0.09135, + "grad_norm": 0.9451104479671788, + "learning_rate": 0.003, + "loss": 4.1064, + "step": 9135 + }, + { + "epoch": 0.09136, + "grad_norm": 0.8741863870514021, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 9136 + }, + { + "epoch": 0.09137, + "grad_norm": 0.6559657083546983, + "learning_rate": 0.003, + "loss": 4.1186, + "step": 9137 + }, + { + "epoch": 0.09138, + "grad_norm": 0.6591621056680503, + "learning_rate": 0.003, + "loss": 4.085, + "step": 9138 + }, + { + "epoch": 0.09139, + "grad_norm": 0.7172388233296422, + "learning_rate": 0.003, + "loss": 4.0864, + "step": 9139 + }, + { + "epoch": 0.0914, + "grad_norm": 0.74698785265184, + "learning_rate": 0.003, + "loss": 4.0891, + "step": 9140 + }, + { + "epoch": 0.09141, + "grad_norm": 0.7383928212536637, + "learning_rate": 0.003, + "loss": 4.1203, + "step": 9141 + }, + { + "epoch": 0.09142, + "grad_norm": 0.8365703977523314, + "learning_rate": 0.003, + "loss": 4.1106, + "step": 9142 + }, + { + "epoch": 0.09143, + "grad_norm": 0.8481652164716268, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 9143 + }, + { + "epoch": 0.09144, + "grad_norm": 0.7795528073061184, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 9144 + }, + { + "epoch": 0.09145, + "grad_norm": 0.8962209591026435, + "learning_rate": 0.003, + "loss": 4.1122, + "step": 9145 + }, + { + "epoch": 0.09146, + "grad_norm": 0.977738810966034, + "learning_rate": 0.003, + "loss": 4.1051, + "step": 9146 + }, + { + "epoch": 0.09147, + "grad_norm": 1.1158176717774442, + "learning_rate": 0.003, + "loss": 4.0971, + "step": 9147 + }, + { + "epoch": 0.09148, + "grad_norm": 0.851745175787143, + "learning_rate": 0.003, + "loss": 4.1069, + "step": 9148 + }, + { + "epoch": 0.09149, + "grad_norm": 0.7540599482452148, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 9149 + }, + { + "epoch": 0.0915, + "grad_norm": 0.8969362552019574, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 9150 + }, + { + "epoch": 0.09151, + "grad_norm": 0.9897498994292178, + "learning_rate": 0.003, + "loss": 4.1018, + "step": 9151 + }, + { + "epoch": 0.09152, + "grad_norm": 1.0440364187474465, + "learning_rate": 0.003, + "loss": 4.1119, + "step": 9152 + }, + { + "epoch": 0.09153, + "grad_norm": 1.1474475308871095, + "learning_rate": 0.003, + "loss": 4.1149, + "step": 9153 + }, + { + "epoch": 0.09154, + "grad_norm": 0.9705935043490699, + "learning_rate": 0.003, + "loss": 4.1128, + "step": 9154 + }, + { + "epoch": 0.09155, + "grad_norm": 0.9298282598002533, + "learning_rate": 0.003, + "loss": 4.1163, + "step": 9155 + }, + { + "epoch": 0.09156, + "grad_norm": 0.9723217314936119, + "learning_rate": 0.003, + "loss": 4.1139, + "step": 9156 + }, + { + "epoch": 0.09157, + "grad_norm": 1.0948870894323015, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 9157 + }, + { + "epoch": 0.09158, + "grad_norm": 1.082303366161037, + "learning_rate": 0.003, + "loss": 4.102, + "step": 9158 + }, + { + "epoch": 0.09159, + "grad_norm": 1.1823492886345546, + "learning_rate": 0.003, + "loss": 4.1154, + "step": 9159 + }, + { + "epoch": 0.0916, + "grad_norm": 1.1448872296285921, + "learning_rate": 0.003, + "loss": 4.0891, + "step": 9160 + }, + { + "epoch": 0.09161, + "grad_norm": 0.7917308786382252, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 9161 + }, + { + "epoch": 0.09162, + "grad_norm": 0.7101340083428053, + "learning_rate": 0.003, + "loss": 4.1247, + "step": 9162 + }, + { + "epoch": 0.09163, + "grad_norm": 0.5734667662994228, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 9163 + }, + { + "epoch": 0.09164, + "grad_norm": 0.6463742170108849, + "learning_rate": 0.003, + "loss": 4.0888, + "step": 9164 + }, + { + "epoch": 0.09165, + "grad_norm": 0.8791425623296205, + "learning_rate": 0.003, + "loss": 4.1211, + "step": 9165 + }, + { + "epoch": 0.09166, + "grad_norm": 1.2896040065899281, + "learning_rate": 0.003, + "loss": 4.1028, + "step": 9166 + }, + { + "epoch": 0.09167, + "grad_norm": 0.777626739229617, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 9167 + }, + { + "epoch": 0.09168, + "grad_norm": 0.6059045588929811, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 9168 + }, + { + "epoch": 0.09169, + "grad_norm": 0.6211082886708152, + "learning_rate": 0.003, + "loss": 4.0999, + "step": 9169 + }, + { + "epoch": 0.0917, + "grad_norm": 0.6201173250342321, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 9170 + }, + { + "epoch": 0.09171, + "grad_norm": 0.65420012036481, + "learning_rate": 0.003, + "loss": 4.1143, + "step": 9171 + }, + { + "epoch": 0.09172, + "grad_norm": 0.6193382027516, + "learning_rate": 0.003, + "loss": 4.1106, + "step": 9172 + }, + { + "epoch": 0.09173, + "grad_norm": 0.726634103434902, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 9173 + }, + { + "epoch": 0.09174, + "grad_norm": 0.75555391022193, + "learning_rate": 0.003, + "loss": 4.096, + "step": 9174 + }, + { + "epoch": 0.09175, + "grad_norm": 0.791110324469862, + "learning_rate": 0.003, + "loss": 4.0913, + "step": 9175 + }, + { + "epoch": 0.09176, + "grad_norm": 0.785799126594617, + "learning_rate": 0.003, + "loss": 4.0962, + "step": 9176 + }, + { + "epoch": 0.09177, + "grad_norm": 0.7612319543811192, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 9177 + }, + { + "epoch": 0.09178, + "grad_norm": 0.7565419822439869, + "learning_rate": 0.003, + "loss": 4.1, + "step": 9178 + }, + { + "epoch": 0.09179, + "grad_norm": 0.8420544599698648, + "learning_rate": 0.003, + "loss": 4.0873, + "step": 9179 + }, + { + "epoch": 0.0918, + "grad_norm": 1.084438958299728, + "learning_rate": 0.003, + "loss": 4.091, + "step": 9180 + }, + { + "epoch": 0.09181, + "grad_norm": 1.1059617305324483, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 9181 + }, + { + "epoch": 0.09182, + "grad_norm": 1.047379795485612, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 9182 + }, + { + "epoch": 0.09183, + "grad_norm": 0.9756788460870816, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 9183 + }, + { + "epoch": 0.09184, + "grad_norm": 0.8475124797770287, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 9184 + }, + { + "epoch": 0.09185, + "grad_norm": 0.7818162898325053, + "learning_rate": 0.003, + "loss": 4.1069, + "step": 9185 + }, + { + "epoch": 0.09186, + "grad_norm": 0.7766932782419979, + "learning_rate": 0.003, + "loss": 4.05, + "step": 9186 + }, + { + "epoch": 0.09187, + "grad_norm": 0.8825420516075367, + "learning_rate": 0.003, + "loss": 4.1197, + "step": 9187 + }, + { + "epoch": 0.09188, + "grad_norm": 0.9631123401662338, + "learning_rate": 0.003, + "loss": 4.0989, + "step": 9188 + }, + { + "epoch": 0.09189, + "grad_norm": 1.0225689776098181, + "learning_rate": 0.003, + "loss": 4.1105, + "step": 9189 + }, + { + "epoch": 0.0919, + "grad_norm": 1.0782850500259313, + "learning_rate": 0.003, + "loss": 4.1264, + "step": 9190 + }, + { + "epoch": 0.09191, + "grad_norm": 0.9063112295904093, + "learning_rate": 0.003, + "loss": 4.0957, + "step": 9191 + }, + { + "epoch": 0.09192, + "grad_norm": 1.055089562560554, + "learning_rate": 0.003, + "loss": 4.0876, + "step": 9192 + }, + { + "epoch": 0.09193, + "grad_norm": 1.014344237878195, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 9193 + }, + { + "epoch": 0.09194, + "grad_norm": 1.3217721747559106, + "learning_rate": 0.003, + "loss": 4.1042, + "step": 9194 + }, + { + "epoch": 0.09195, + "grad_norm": 0.9570521535388152, + "learning_rate": 0.003, + "loss": 4.0985, + "step": 9195 + }, + { + "epoch": 0.09196, + "grad_norm": 0.9200244611664627, + "learning_rate": 0.003, + "loss": 4.1116, + "step": 9196 + }, + { + "epoch": 0.09197, + "grad_norm": 0.962705740113485, + "learning_rate": 0.003, + "loss": 4.106, + "step": 9197 + }, + { + "epoch": 0.09198, + "grad_norm": 1.0994040596641663, + "learning_rate": 0.003, + "loss": 4.1027, + "step": 9198 + }, + { + "epoch": 0.09199, + "grad_norm": 0.9897723274209836, + "learning_rate": 0.003, + "loss": 4.112, + "step": 9199 + }, + { + "epoch": 0.092, + "grad_norm": 0.9542387690632056, + "learning_rate": 0.003, + "loss": 4.1077, + "step": 9200 + }, + { + "epoch": 0.09201, + "grad_norm": 1.073948721463555, + "learning_rate": 0.003, + "loss": 4.0998, + "step": 9201 + }, + { + "epoch": 0.09202, + "grad_norm": 0.9997228282811497, + "learning_rate": 0.003, + "loss": 4.1149, + "step": 9202 + }, + { + "epoch": 0.09203, + "grad_norm": 0.8556418820489938, + "learning_rate": 0.003, + "loss": 4.1072, + "step": 9203 + }, + { + "epoch": 0.09204, + "grad_norm": 0.7789051082427513, + "learning_rate": 0.003, + "loss": 4.1096, + "step": 9204 + }, + { + "epoch": 0.09205, + "grad_norm": 0.7717492134597749, + "learning_rate": 0.003, + "loss": 4.0978, + "step": 9205 + }, + { + "epoch": 0.09206, + "grad_norm": 0.8428849907020309, + "learning_rate": 0.003, + "loss": 4.107, + "step": 9206 + }, + { + "epoch": 0.09207, + "grad_norm": 0.9040648516741944, + "learning_rate": 0.003, + "loss": 4.1254, + "step": 9207 + }, + { + "epoch": 0.09208, + "grad_norm": 1.0913270867403186, + "learning_rate": 0.003, + "loss": 4.0903, + "step": 9208 + }, + { + "epoch": 0.09209, + "grad_norm": 1.0048770200215928, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 9209 + }, + { + "epoch": 0.0921, + "grad_norm": 1.0392328935432598, + "learning_rate": 0.003, + "loss": 4.116, + "step": 9210 + }, + { + "epoch": 0.09211, + "grad_norm": 0.9204701850435161, + "learning_rate": 0.003, + "loss": 4.0902, + "step": 9211 + }, + { + "epoch": 0.09212, + "grad_norm": 0.9402955316963719, + "learning_rate": 0.003, + "loss": 4.1128, + "step": 9212 + }, + { + "epoch": 0.09213, + "grad_norm": 1.048761874355027, + "learning_rate": 0.003, + "loss": 4.0984, + "step": 9213 + }, + { + "epoch": 0.09214, + "grad_norm": 1.0562819020868857, + "learning_rate": 0.003, + "loss": 4.084, + "step": 9214 + }, + { + "epoch": 0.09215, + "grad_norm": 1.085719329634591, + "learning_rate": 0.003, + "loss": 4.0949, + "step": 9215 + }, + { + "epoch": 0.09216, + "grad_norm": 0.9924362432989838, + "learning_rate": 0.003, + "loss": 4.1003, + "step": 9216 + }, + { + "epoch": 0.09217, + "grad_norm": 0.9416068771195782, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 9217 + }, + { + "epoch": 0.09218, + "grad_norm": 0.9216157764389595, + "learning_rate": 0.003, + "loss": 4.1165, + "step": 9218 + }, + { + "epoch": 0.09219, + "grad_norm": 1.0578289251402284, + "learning_rate": 0.003, + "loss": 4.1044, + "step": 9219 + }, + { + "epoch": 0.0922, + "grad_norm": 0.8494550851933941, + "learning_rate": 0.003, + "loss": 4.0996, + "step": 9220 + }, + { + "epoch": 0.09221, + "grad_norm": 0.7004867483817171, + "learning_rate": 0.003, + "loss": 4.1055, + "step": 9221 + }, + { + "epoch": 0.09222, + "grad_norm": 0.7348105532675423, + "learning_rate": 0.003, + "loss": 4.0927, + "step": 9222 + }, + { + "epoch": 0.09223, + "grad_norm": 0.9417547593730853, + "learning_rate": 0.003, + "loss": 4.094, + "step": 9223 + }, + { + "epoch": 0.09224, + "grad_norm": 1.049648042435812, + "learning_rate": 0.003, + "loss": 4.1138, + "step": 9224 + }, + { + "epoch": 0.09225, + "grad_norm": 1.0276279527925067, + "learning_rate": 0.003, + "loss": 4.0919, + "step": 9225 + }, + { + "epoch": 0.09226, + "grad_norm": 0.9585527602924722, + "learning_rate": 0.003, + "loss": 4.076, + "step": 9226 + }, + { + "epoch": 0.09227, + "grad_norm": 0.9337166610853146, + "learning_rate": 0.003, + "loss": 4.0935, + "step": 9227 + }, + { + "epoch": 0.09228, + "grad_norm": 0.9027418181317508, + "learning_rate": 0.003, + "loss": 4.0947, + "step": 9228 + }, + { + "epoch": 0.09229, + "grad_norm": 0.9198727309478711, + "learning_rate": 0.003, + "loss": 4.1011, + "step": 9229 + }, + { + "epoch": 0.0923, + "grad_norm": 0.9045649479252982, + "learning_rate": 0.003, + "loss": 4.0985, + "step": 9230 + }, + { + "epoch": 0.09231, + "grad_norm": 0.9974652662909131, + "learning_rate": 0.003, + "loss": 4.0971, + "step": 9231 + }, + { + "epoch": 0.09232, + "grad_norm": 1.0912379899673035, + "learning_rate": 0.003, + "loss": 4.1094, + "step": 9232 + }, + { + "epoch": 0.09233, + "grad_norm": 1.0915766025748115, + "learning_rate": 0.003, + "loss": 4.1243, + "step": 9233 + }, + { + "epoch": 0.09234, + "grad_norm": 0.9888878322311199, + "learning_rate": 0.003, + "loss": 4.1278, + "step": 9234 + }, + { + "epoch": 0.09235, + "grad_norm": 0.8256741524071066, + "learning_rate": 0.003, + "loss": 4.0888, + "step": 9235 + }, + { + "epoch": 0.09236, + "grad_norm": 0.7623729808775097, + "learning_rate": 0.003, + "loss": 4.1281, + "step": 9236 + }, + { + "epoch": 0.09237, + "grad_norm": 0.784646386208245, + "learning_rate": 0.003, + "loss": 4.087, + "step": 9237 + }, + { + "epoch": 0.09238, + "grad_norm": 0.781447671218314, + "learning_rate": 0.003, + "loss": 4.074, + "step": 9238 + }, + { + "epoch": 0.09239, + "grad_norm": 0.7580100064874866, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 9239 + }, + { + "epoch": 0.0924, + "grad_norm": 0.7887138742063122, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 9240 + }, + { + "epoch": 0.09241, + "grad_norm": 0.9232343150935588, + "learning_rate": 0.003, + "loss": 4.0991, + "step": 9241 + }, + { + "epoch": 0.09242, + "grad_norm": 1.1881392195909175, + "learning_rate": 0.003, + "loss": 4.1256, + "step": 9242 + }, + { + "epoch": 0.09243, + "grad_norm": 0.9840929580945444, + "learning_rate": 0.003, + "loss": 4.12, + "step": 9243 + }, + { + "epoch": 0.09244, + "grad_norm": 1.304430856896462, + "learning_rate": 0.003, + "loss": 4.1126, + "step": 9244 + }, + { + "epoch": 0.09245, + "grad_norm": 0.9351259127699327, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 9245 + }, + { + "epoch": 0.09246, + "grad_norm": 0.9793836268379075, + "learning_rate": 0.003, + "loss": 4.1026, + "step": 9246 + }, + { + "epoch": 0.09247, + "grad_norm": 0.9197456699785576, + "learning_rate": 0.003, + "loss": 4.0905, + "step": 9247 + }, + { + "epoch": 0.09248, + "grad_norm": 0.8196313202466566, + "learning_rate": 0.003, + "loss": 4.0982, + "step": 9248 + }, + { + "epoch": 0.09249, + "grad_norm": 0.7928866730787516, + "learning_rate": 0.003, + "loss": 4.1449, + "step": 9249 + }, + { + "epoch": 0.0925, + "grad_norm": 0.7531883062490411, + "learning_rate": 0.003, + "loss": 4.1114, + "step": 9250 + }, + { + "epoch": 0.09251, + "grad_norm": 0.7666898332256793, + "learning_rate": 0.003, + "loss": 4.1092, + "step": 9251 + }, + { + "epoch": 0.09252, + "grad_norm": 0.7988522492516801, + "learning_rate": 0.003, + "loss": 4.1198, + "step": 9252 + }, + { + "epoch": 0.09253, + "grad_norm": 0.7491102789906501, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 9253 + }, + { + "epoch": 0.09254, + "grad_norm": 0.6784653406918104, + "learning_rate": 0.003, + "loss": 4.1062, + "step": 9254 + }, + { + "epoch": 0.09255, + "grad_norm": 0.5654398621536241, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 9255 + }, + { + "epoch": 0.09256, + "grad_norm": 0.6643608382759081, + "learning_rate": 0.003, + "loss": 4.0957, + "step": 9256 + }, + { + "epoch": 0.09257, + "grad_norm": 0.7193700387398151, + "learning_rate": 0.003, + "loss": 4.0907, + "step": 9257 + }, + { + "epoch": 0.09258, + "grad_norm": 0.8759808055403874, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 9258 + }, + { + "epoch": 0.09259, + "grad_norm": 1.2012930245710616, + "learning_rate": 0.003, + "loss": 4.1075, + "step": 9259 + }, + { + "epoch": 0.0926, + "grad_norm": 0.7418114937003976, + "learning_rate": 0.003, + "loss": 4.0931, + "step": 9260 + }, + { + "epoch": 0.09261, + "grad_norm": 0.5474250675303859, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 9261 + }, + { + "epoch": 0.09262, + "grad_norm": 0.6829621284506816, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 9262 + }, + { + "epoch": 0.09263, + "grad_norm": 0.7768878158809396, + "learning_rate": 0.003, + "loss": 4.0916, + "step": 9263 + }, + { + "epoch": 0.09264, + "grad_norm": 0.9996107855545961, + "learning_rate": 0.003, + "loss": 4.1261, + "step": 9264 + }, + { + "epoch": 0.09265, + "grad_norm": 1.163827375444901, + "learning_rate": 0.003, + "loss": 4.0995, + "step": 9265 + }, + { + "epoch": 0.09266, + "grad_norm": 0.7349870746113485, + "learning_rate": 0.003, + "loss": 4.0908, + "step": 9266 + }, + { + "epoch": 0.09267, + "grad_norm": 0.6853508114179676, + "learning_rate": 0.003, + "loss": 4.07, + "step": 9267 + }, + { + "epoch": 0.09268, + "grad_norm": 0.6333065813900194, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 9268 + }, + { + "epoch": 0.09269, + "grad_norm": 0.7044393984561579, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 9269 + }, + { + "epoch": 0.0927, + "grad_norm": 0.7829797104270343, + "learning_rate": 0.003, + "loss": 4.1113, + "step": 9270 + }, + { + "epoch": 0.09271, + "grad_norm": 0.8257575052405663, + "learning_rate": 0.003, + "loss": 4.0963, + "step": 9271 + }, + { + "epoch": 0.09272, + "grad_norm": 0.9439694552484319, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 9272 + }, + { + "epoch": 0.09273, + "grad_norm": 1.0602176422418417, + "learning_rate": 0.003, + "loss": 4.1077, + "step": 9273 + }, + { + "epoch": 0.09274, + "grad_norm": 0.9858527050122837, + "learning_rate": 0.003, + "loss": 4.0971, + "step": 9274 + }, + { + "epoch": 0.09275, + "grad_norm": 0.9316290417225153, + "learning_rate": 0.003, + "loss": 4.1187, + "step": 9275 + }, + { + "epoch": 0.09276, + "grad_norm": 0.9228744715959912, + "learning_rate": 0.003, + "loss": 4.1045, + "step": 9276 + }, + { + "epoch": 0.09277, + "grad_norm": 0.97661764276748, + "learning_rate": 0.003, + "loss": 4.1017, + "step": 9277 + }, + { + "epoch": 0.09278, + "grad_norm": 0.9667223874049936, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 9278 + }, + { + "epoch": 0.09279, + "grad_norm": 1.0509045802580743, + "learning_rate": 0.003, + "loss": 4.1021, + "step": 9279 + }, + { + "epoch": 0.0928, + "grad_norm": 1.199654598508013, + "learning_rate": 0.003, + "loss": 4.1025, + "step": 9280 + }, + { + "epoch": 0.09281, + "grad_norm": 0.8190040066327821, + "learning_rate": 0.003, + "loss": 4.0984, + "step": 9281 + }, + { + "epoch": 0.09282, + "grad_norm": 0.8005381258667182, + "learning_rate": 0.003, + "loss": 4.1265, + "step": 9282 + }, + { + "epoch": 0.09283, + "grad_norm": 0.9731528357234364, + "learning_rate": 0.003, + "loss": 4.121, + "step": 9283 + }, + { + "epoch": 0.09284, + "grad_norm": 1.0699122649133415, + "learning_rate": 0.003, + "loss": 4.1056, + "step": 9284 + }, + { + "epoch": 0.09285, + "grad_norm": 0.9736646539104704, + "learning_rate": 0.003, + "loss": 4.0955, + "step": 9285 + }, + { + "epoch": 0.09286, + "grad_norm": 0.7534181301780545, + "learning_rate": 0.003, + "loss": 4.1069, + "step": 9286 + }, + { + "epoch": 0.09287, + "grad_norm": 0.7039247323481843, + "learning_rate": 0.003, + "loss": 4.1049, + "step": 9287 + }, + { + "epoch": 0.09288, + "grad_norm": 0.6153847411143628, + "learning_rate": 0.003, + "loss": 4.1008, + "step": 9288 + }, + { + "epoch": 0.09289, + "grad_norm": 0.7439491494868004, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 9289 + }, + { + "epoch": 0.0929, + "grad_norm": 1.0211178389308402, + "learning_rate": 0.003, + "loss": 4.1044, + "step": 9290 + }, + { + "epoch": 0.09291, + "grad_norm": 1.2912396043684187, + "learning_rate": 0.003, + "loss": 4.1027, + "step": 9291 + }, + { + "epoch": 0.09292, + "grad_norm": 1.0329590152181434, + "learning_rate": 0.003, + "loss": 4.115, + "step": 9292 + }, + { + "epoch": 0.09293, + "grad_norm": 0.925671296507913, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 9293 + }, + { + "epoch": 0.09294, + "grad_norm": 0.8753174570663704, + "learning_rate": 0.003, + "loss": 4.0968, + "step": 9294 + }, + { + "epoch": 0.09295, + "grad_norm": 0.7750429431472656, + "learning_rate": 0.003, + "loss": 4.0975, + "step": 9295 + }, + { + "epoch": 0.09296, + "grad_norm": 0.8498716958126399, + "learning_rate": 0.003, + "loss": 4.084, + "step": 9296 + }, + { + "epoch": 0.09297, + "grad_norm": 0.9547075919931944, + "learning_rate": 0.003, + "loss": 4.0993, + "step": 9297 + }, + { + "epoch": 0.09298, + "grad_norm": 1.0547931478455412, + "learning_rate": 0.003, + "loss": 4.1016, + "step": 9298 + }, + { + "epoch": 0.09299, + "grad_norm": 0.9874863882567291, + "learning_rate": 0.003, + "loss": 4.0972, + "step": 9299 + }, + { + "epoch": 0.093, + "grad_norm": 1.0320217787422357, + "learning_rate": 0.003, + "loss": 4.1013, + "step": 9300 + }, + { + "epoch": 0.09301, + "grad_norm": 0.868256523126305, + "learning_rate": 0.003, + "loss": 4.0922, + "step": 9301 + }, + { + "epoch": 0.09302, + "grad_norm": 0.8911189359509369, + "learning_rate": 0.003, + "loss": 4.1232, + "step": 9302 + }, + { + "epoch": 0.09303, + "grad_norm": 1.1184167860096568, + "learning_rate": 0.003, + "loss": 4.1016, + "step": 9303 + }, + { + "epoch": 0.09304, + "grad_norm": 1.0616222963872175, + "learning_rate": 0.003, + "loss": 4.0933, + "step": 9304 + }, + { + "epoch": 0.09305, + "grad_norm": 1.0167967680320442, + "learning_rate": 0.003, + "loss": 4.0971, + "step": 9305 + }, + { + "epoch": 0.09306, + "grad_norm": 0.9100621741165005, + "learning_rate": 0.003, + "loss": 4.1025, + "step": 9306 + }, + { + "epoch": 0.09307, + "grad_norm": 0.7277198010566454, + "learning_rate": 0.003, + "loss": 4.0987, + "step": 9307 + }, + { + "epoch": 0.09308, + "grad_norm": 0.7642184186042998, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 9308 + }, + { + "epoch": 0.09309, + "grad_norm": 0.7898375897779486, + "learning_rate": 0.003, + "loss": 4.1093, + "step": 9309 + }, + { + "epoch": 0.0931, + "grad_norm": 0.8313627098153431, + "learning_rate": 0.003, + "loss": 4.1156, + "step": 9310 + }, + { + "epoch": 0.09311, + "grad_norm": 0.9180361447599314, + "learning_rate": 0.003, + "loss": 4.1021, + "step": 9311 + }, + { + "epoch": 0.09312, + "grad_norm": 1.0722183150254507, + "learning_rate": 0.003, + "loss": 4.1053, + "step": 9312 + }, + { + "epoch": 0.09313, + "grad_norm": 1.0053080965881522, + "learning_rate": 0.003, + "loss": 4.1345, + "step": 9313 + }, + { + "epoch": 0.09314, + "grad_norm": 1.0429150067918742, + "learning_rate": 0.003, + "loss": 4.1065, + "step": 9314 + }, + { + "epoch": 0.09315, + "grad_norm": 1.0275188761589462, + "learning_rate": 0.003, + "loss": 4.1035, + "step": 9315 + }, + { + "epoch": 0.09316, + "grad_norm": 0.9852252098429989, + "learning_rate": 0.003, + "loss": 4.1047, + "step": 9316 + }, + { + "epoch": 0.09317, + "grad_norm": 1.0943356019238426, + "learning_rate": 0.003, + "loss": 4.1259, + "step": 9317 + }, + { + "epoch": 0.09318, + "grad_norm": 0.8892700832028245, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 9318 + }, + { + "epoch": 0.09319, + "grad_norm": 0.8440332983350739, + "learning_rate": 0.003, + "loss": 4.1321, + "step": 9319 + }, + { + "epoch": 0.0932, + "grad_norm": 0.9136370265528279, + "learning_rate": 0.003, + "loss": 4.1065, + "step": 9320 + }, + { + "epoch": 0.09321, + "grad_norm": 1.0652363081449419, + "learning_rate": 0.003, + "loss": 4.1019, + "step": 9321 + }, + { + "epoch": 0.09322, + "grad_norm": 0.9657802025171371, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 9322 + }, + { + "epoch": 0.09323, + "grad_norm": 0.9541349646760839, + "learning_rate": 0.003, + "loss": 4.1003, + "step": 9323 + }, + { + "epoch": 0.09324, + "grad_norm": 0.940690991603989, + "learning_rate": 0.003, + "loss": 4.1185, + "step": 9324 + }, + { + "epoch": 0.09325, + "grad_norm": 0.9773641494477271, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 9325 + }, + { + "epoch": 0.09326, + "grad_norm": 1.1123850382527165, + "learning_rate": 0.003, + "loss": 4.0887, + "step": 9326 + }, + { + "epoch": 0.09327, + "grad_norm": 0.8756313815561116, + "learning_rate": 0.003, + "loss": 4.0956, + "step": 9327 + }, + { + "epoch": 0.09328, + "grad_norm": 0.7389172317899884, + "learning_rate": 0.003, + "loss": 4.0941, + "step": 9328 + }, + { + "epoch": 0.09329, + "grad_norm": 0.6983046686762022, + "learning_rate": 0.003, + "loss": 4.0969, + "step": 9329 + }, + { + "epoch": 0.0933, + "grad_norm": 0.6752597877286427, + "learning_rate": 0.003, + "loss": 4.1111, + "step": 9330 + }, + { + "epoch": 0.09331, + "grad_norm": 0.8385408030611338, + "learning_rate": 0.003, + "loss": 4.081, + "step": 9331 + }, + { + "epoch": 0.09332, + "grad_norm": 0.92425769636424, + "learning_rate": 0.003, + "loss": 4.1071, + "step": 9332 + }, + { + "epoch": 0.09333, + "grad_norm": 1.1149373386342603, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 9333 + }, + { + "epoch": 0.09334, + "grad_norm": 0.9061499154659839, + "learning_rate": 0.003, + "loss": 4.1063, + "step": 9334 + }, + { + "epoch": 0.09335, + "grad_norm": 0.7814181455603038, + "learning_rate": 0.003, + "loss": 4.0905, + "step": 9335 + }, + { + "epoch": 0.09336, + "grad_norm": 0.6185969547295879, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 9336 + }, + { + "epoch": 0.09337, + "grad_norm": 0.570098829353087, + "learning_rate": 0.003, + "loss": 4.091, + "step": 9337 + }, + { + "epoch": 0.09338, + "grad_norm": 0.6307398718763437, + "learning_rate": 0.003, + "loss": 4.0956, + "step": 9338 + }, + { + "epoch": 0.09339, + "grad_norm": 0.5876765823170635, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 9339 + }, + { + "epoch": 0.0934, + "grad_norm": 0.6057455120022778, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 9340 + }, + { + "epoch": 0.09341, + "grad_norm": 0.7819289757235965, + "learning_rate": 0.003, + "loss": 4.0898, + "step": 9341 + }, + { + "epoch": 0.09342, + "grad_norm": 1.0525239903626875, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 9342 + }, + { + "epoch": 0.09343, + "grad_norm": 0.922797767764872, + "learning_rate": 0.003, + "loss": 4.0887, + "step": 9343 + }, + { + "epoch": 0.09344, + "grad_norm": 0.7651325270309342, + "learning_rate": 0.003, + "loss": 4.0864, + "step": 9344 + }, + { + "epoch": 0.09345, + "grad_norm": 0.8307201489790069, + "learning_rate": 0.003, + "loss": 4.1062, + "step": 9345 + }, + { + "epoch": 0.09346, + "grad_norm": 0.9268559341614948, + "learning_rate": 0.003, + "loss": 4.0995, + "step": 9346 + }, + { + "epoch": 0.09347, + "grad_norm": 0.8843815142269438, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 9347 + }, + { + "epoch": 0.09348, + "grad_norm": 0.7511422034725228, + "learning_rate": 0.003, + "loss": 4.1036, + "step": 9348 + }, + { + "epoch": 0.09349, + "grad_norm": 0.7681336246456821, + "learning_rate": 0.003, + "loss": 4.0989, + "step": 9349 + }, + { + "epoch": 0.0935, + "grad_norm": 0.8391732024120159, + "learning_rate": 0.003, + "loss": 4.0878, + "step": 9350 + }, + { + "epoch": 0.09351, + "grad_norm": 0.8572535723613135, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 9351 + }, + { + "epoch": 0.09352, + "grad_norm": 1.1345129108011323, + "learning_rate": 0.003, + "loss": 4.0954, + "step": 9352 + }, + { + "epoch": 0.09353, + "grad_norm": 1.188322335200505, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 9353 + }, + { + "epoch": 0.09354, + "grad_norm": 0.7871694631960734, + "learning_rate": 0.003, + "loss": 4.0855, + "step": 9354 + }, + { + "epoch": 0.09355, + "grad_norm": 0.7942202356512671, + "learning_rate": 0.003, + "loss": 4.109, + "step": 9355 + }, + { + "epoch": 0.09356, + "grad_norm": 0.8720300671130099, + "learning_rate": 0.003, + "loss": 4.095, + "step": 9356 + }, + { + "epoch": 0.09357, + "grad_norm": 0.9400770356953755, + "learning_rate": 0.003, + "loss": 4.0942, + "step": 9357 + }, + { + "epoch": 0.09358, + "grad_norm": 0.872908592359249, + "learning_rate": 0.003, + "loss": 4.0944, + "step": 9358 + }, + { + "epoch": 0.09359, + "grad_norm": 0.9729065406434361, + "learning_rate": 0.003, + "loss": 4.1064, + "step": 9359 + }, + { + "epoch": 0.0936, + "grad_norm": 1.4002450511532056, + "learning_rate": 0.003, + "loss": 4.1089, + "step": 9360 + }, + { + "epoch": 0.09361, + "grad_norm": 0.8007667671323881, + "learning_rate": 0.003, + "loss": 4.1014, + "step": 9361 + }, + { + "epoch": 0.09362, + "grad_norm": 0.8253755648015588, + "learning_rate": 0.003, + "loss": 4.1318, + "step": 9362 + }, + { + "epoch": 0.09363, + "grad_norm": 0.8603986662115061, + "learning_rate": 0.003, + "loss": 4.12, + "step": 9363 + }, + { + "epoch": 0.09364, + "grad_norm": 0.8888437458688937, + "learning_rate": 0.003, + "loss": 4.089, + "step": 9364 + }, + { + "epoch": 0.09365, + "grad_norm": 1.062809120441484, + "learning_rate": 0.003, + "loss": 4.1185, + "step": 9365 + }, + { + "epoch": 0.09366, + "grad_norm": 0.9302023322409313, + "learning_rate": 0.003, + "loss": 4.069, + "step": 9366 + }, + { + "epoch": 0.09367, + "grad_norm": 1.1396009232198598, + "learning_rate": 0.003, + "loss": 4.1054, + "step": 9367 + }, + { + "epoch": 0.09368, + "grad_norm": 0.9961165393209213, + "learning_rate": 0.003, + "loss": 4.1149, + "step": 9368 + }, + { + "epoch": 0.09369, + "grad_norm": 0.9522147427617241, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 9369 + }, + { + "epoch": 0.0937, + "grad_norm": 1.0784020973346404, + "learning_rate": 0.003, + "loss": 4.1296, + "step": 9370 + }, + { + "epoch": 0.09371, + "grad_norm": 1.1071187911940317, + "learning_rate": 0.003, + "loss": 4.1135, + "step": 9371 + }, + { + "epoch": 0.09372, + "grad_norm": 0.94833464425284, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 9372 + }, + { + "epoch": 0.09373, + "grad_norm": 0.9480880293079869, + "learning_rate": 0.003, + "loss": 4.1405, + "step": 9373 + }, + { + "epoch": 0.09374, + "grad_norm": 0.994929884790511, + "learning_rate": 0.003, + "loss": 4.1413, + "step": 9374 + }, + { + "epoch": 0.09375, + "grad_norm": 0.988811266369445, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 9375 + }, + { + "epoch": 0.09376, + "grad_norm": 0.9324886958686643, + "learning_rate": 0.003, + "loss": 4.124, + "step": 9376 + }, + { + "epoch": 0.09377, + "grad_norm": 0.9548553533271001, + "learning_rate": 0.003, + "loss": 4.1048, + "step": 9377 + }, + { + "epoch": 0.09378, + "grad_norm": 0.9562248687261611, + "learning_rate": 0.003, + "loss": 4.1138, + "step": 9378 + }, + { + "epoch": 0.09379, + "grad_norm": 1.0222988140200817, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 9379 + }, + { + "epoch": 0.0938, + "grad_norm": 0.9124887754159823, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 9380 + }, + { + "epoch": 0.09381, + "grad_norm": 1.037020146160175, + "learning_rate": 0.003, + "loss": 4.0908, + "step": 9381 + }, + { + "epoch": 0.09382, + "grad_norm": 0.9546631751003297, + "learning_rate": 0.003, + "loss": 4.1103, + "step": 9382 + }, + { + "epoch": 0.09383, + "grad_norm": 1.040184676131133, + "learning_rate": 0.003, + "loss": 4.0971, + "step": 9383 + }, + { + "epoch": 0.09384, + "grad_norm": 1.029833944161637, + "learning_rate": 0.003, + "loss": 4.1261, + "step": 9384 + }, + { + "epoch": 0.09385, + "grad_norm": 0.899541831452752, + "learning_rate": 0.003, + "loss": 4.103, + "step": 9385 + }, + { + "epoch": 0.09386, + "grad_norm": 0.8629312944127356, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 9386 + }, + { + "epoch": 0.09387, + "grad_norm": 1.0409426461274158, + "learning_rate": 0.003, + "loss": 4.1109, + "step": 9387 + }, + { + "epoch": 0.09388, + "grad_norm": 1.006589469006825, + "learning_rate": 0.003, + "loss": 4.1055, + "step": 9388 + }, + { + "epoch": 0.09389, + "grad_norm": 1.1077114456174915, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 9389 + }, + { + "epoch": 0.0939, + "grad_norm": 1.011899263145205, + "learning_rate": 0.003, + "loss": 4.111, + "step": 9390 + }, + { + "epoch": 0.09391, + "grad_norm": 0.8545868658356651, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 9391 + }, + { + "epoch": 0.09392, + "grad_norm": 0.7530450948177452, + "learning_rate": 0.003, + "loss": 4.1191, + "step": 9392 + }, + { + "epoch": 0.09393, + "grad_norm": 0.8389311586343143, + "learning_rate": 0.003, + "loss": 4.108, + "step": 9393 + }, + { + "epoch": 0.09394, + "grad_norm": 0.8905715644273344, + "learning_rate": 0.003, + "loss": 4.0956, + "step": 9394 + }, + { + "epoch": 0.09395, + "grad_norm": 0.9931804256938302, + "learning_rate": 0.003, + "loss": 4.1021, + "step": 9395 + }, + { + "epoch": 0.09396, + "grad_norm": 1.1886533323261048, + "learning_rate": 0.003, + "loss": 4.1355, + "step": 9396 + }, + { + "epoch": 0.09397, + "grad_norm": 0.8490709939667962, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 9397 + }, + { + "epoch": 0.09398, + "grad_norm": 0.9054704841108464, + "learning_rate": 0.003, + "loss": 4.1157, + "step": 9398 + }, + { + "epoch": 0.09399, + "grad_norm": 0.8063846844294164, + "learning_rate": 0.003, + "loss": 4.1027, + "step": 9399 + }, + { + "epoch": 0.094, + "grad_norm": 0.69451549125735, + "learning_rate": 0.003, + "loss": 4.0933, + "step": 9400 + }, + { + "epoch": 0.09401, + "grad_norm": 0.6661444422351755, + "learning_rate": 0.003, + "loss": 4.1062, + "step": 9401 + }, + { + "epoch": 0.09402, + "grad_norm": 0.7717530463401683, + "learning_rate": 0.003, + "loss": 4.0935, + "step": 9402 + }, + { + "epoch": 0.09403, + "grad_norm": 0.9162649629518274, + "learning_rate": 0.003, + "loss": 4.117, + "step": 9403 + }, + { + "epoch": 0.09404, + "grad_norm": 1.0198930452293649, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 9404 + }, + { + "epoch": 0.09405, + "grad_norm": 0.9665735513307437, + "learning_rate": 0.003, + "loss": 4.1156, + "step": 9405 + }, + { + "epoch": 0.09406, + "grad_norm": 0.9242295093220471, + "learning_rate": 0.003, + "loss": 4.1116, + "step": 9406 + }, + { + "epoch": 0.09407, + "grad_norm": 0.8899999098808963, + "learning_rate": 0.003, + "loss": 4.0989, + "step": 9407 + }, + { + "epoch": 0.09408, + "grad_norm": 0.7998331174379482, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 9408 + }, + { + "epoch": 0.09409, + "grad_norm": 0.7813089937027473, + "learning_rate": 0.003, + "loss": 4.1042, + "step": 9409 + }, + { + "epoch": 0.0941, + "grad_norm": 0.8667120250947528, + "learning_rate": 0.003, + "loss": 4.1145, + "step": 9410 + }, + { + "epoch": 0.09411, + "grad_norm": 0.9269402848773428, + "learning_rate": 0.003, + "loss": 4.1221, + "step": 9411 + }, + { + "epoch": 0.09412, + "grad_norm": 0.9835401553473229, + "learning_rate": 0.003, + "loss": 4.1174, + "step": 9412 + }, + { + "epoch": 0.09413, + "grad_norm": 0.9172083233778977, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 9413 + }, + { + "epoch": 0.09414, + "grad_norm": 1.006134207979243, + "learning_rate": 0.003, + "loss": 4.1099, + "step": 9414 + }, + { + "epoch": 0.09415, + "grad_norm": 1.132729901318295, + "learning_rate": 0.003, + "loss": 4.08, + "step": 9415 + }, + { + "epoch": 0.09416, + "grad_norm": 0.8404655613518899, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 9416 + }, + { + "epoch": 0.09417, + "grad_norm": 0.7194877963786109, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 9417 + }, + { + "epoch": 0.09418, + "grad_norm": 0.7730057932177794, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 9418 + }, + { + "epoch": 0.09419, + "grad_norm": 1.1292343134560585, + "learning_rate": 0.003, + "loss": 4.1207, + "step": 9419 + }, + { + "epoch": 0.0942, + "grad_norm": 1.1774763095295602, + "learning_rate": 0.003, + "loss": 4.0976, + "step": 9420 + }, + { + "epoch": 0.09421, + "grad_norm": 0.7857554926583439, + "learning_rate": 0.003, + "loss": 4.0891, + "step": 9421 + }, + { + "epoch": 0.09422, + "grad_norm": 0.7323939073029049, + "learning_rate": 0.003, + "loss": 4.0925, + "step": 9422 + }, + { + "epoch": 0.09423, + "grad_norm": 0.7380592179333626, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 9423 + }, + { + "epoch": 0.09424, + "grad_norm": 0.811775438330701, + "learning_rate": 0.003, + "loss": 4.0955, + "step": 9424 + }, + { + "epoch": 0.09425, + "grad_norm": 0.7513687697939088, + "learning_rate": 0.003, + "loss": 4.1115, + "step": 9425 + }, + { + "epoch": 0.09426, + "grad_norm": 0.7219130870944211, + "learning_rate": 0.003, + "loss": 4.1169, + "step": 9426 + }, + { + "epoch": 0.09427, + "grad_norm": 0.7513581740192289, + "learning_rate": 0.003, + "loss": 4.1019, + "step": 9427 + }, + { + "epoch": 0.09428, + "grad_norm": 0.8363301820337448, + "learning_rate": 0.003, + "loss": 4.1118, + "step": 9428 + }, + { + "epoch": 0.09429, + "grad_norm": 0.8654532174390273, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 9429 + }, + { + "epoch": 0.0943, + "grad_norm": 1.071277664835941, + "learning_rate": 0.003, + "loss": 4.0947, + "step": 9430 + }, + { + "epoch": 0.09431, + "grad_norm": 1.2678205157799658, + "learning_rate": 0.003, + "loss": 4.1126, + "step": 9431 + }, + { + "epoch": 0.09432, + "grad_norm": 0.754680518969369, + "learning_rate": 0.003, + "loss": 4.0858, + "step": 9432 + }, + { + "epoch": 0.09433, + "grad_norm": 0.8341523731988698, + "learning_rate": 0.003, + "loss": 4.09, + "step": 9433 + }, + { + "epoch": 0.09434, + "grad_norm": 0.9342603548990726, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 9434 + }, + { + "epoch": 0.09435, + "grad_norm": 0.9289091636637407, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 9435 + }, + { + "epoch": 0.09436, + "grad_norm": 1.107766105780556, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 9436 + }, + { + "epoch": 0.09437, + "grad_norm": 0.9085956133024019, + "learning_rate": 0.003, + "loss": 4.0912, + "step": 9437 + }, + { + "epoch": 0.09438, + "grad_norm": 0.7903225243057821, + "learning_rate": 0.003, + "loss": 4.1089, + "step": 9438 + }, + { + "epoch": 0.09439, + "grad_norm": 0.8100969592812339, + "learning_rate": 0.003, + "loss": 4.1025, + "step": 9439 + }, + { + "epoch": 0.0944, + "grad_norm": 0.9111468297674603, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 9440 + }, + { + "epoch": 0.09441, + "grad_norm": 0.8331316815484991, + "learning_rate": 0.003, + "loss": 4.0926, + "step": 9441 + }, + { + "epoch": 0.09442, + "grad_norm": 0.870678637091265, + "learning_rate": 0.003, + "loss": 4.0857, + "step": 9442 + }, + { + "epoch": 0.09443, + "grad_norm": 0.9311560253694179, + "learning_rate": 0.003, + "loss": 4.1204, + "step": 9443 + }, + { + "epoch": 0.09444, + "grad_norm": 1.0214536095553628, + "learning_rate": 0.003, + "loss": 4.111, + "step": 9444 + }, + { + "epoch": 0.09445, + "grad_norm": 1.1766483143623587, + "learning_rate": 0.003, + "loss": 4.1068, + "step": 9445 + }, + { + "epoch": 0.09446, + "grad_norm": 0.9519894255979084, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 9446 + }, + { + "epoch": 0.09447, + "grad_norm": 0.9198353317350884, + "learning_rate": 0.003, + "loss": 4.1122, + "step": 9447 + }, + { + "epoch": 0.09448, + "grad_norm": 1.05453779469519, + "learning_rate": 0.003, + "loss": 4.1114, + "step": 9448 + }, + { + "epoch": 0.09449, + "grad_norm": 1.0266876101640614, + "learning_rate": 0.003, + "loss": 4.1013, + "step": 9449 + }, + { + "epoch": 0.0945, + "grad_norm": 1.029243760537115, + "learning_rate": 0.003, + "loss": 4.0999, + "step": 9450 + }, + { + "epoch": 0.09451, + "grad_norm": 1.0088314651093355, + "learning_rate": 0.003, + "loss": 4.1092, + "step": 9451 + }, + { + "epoch": 0.09452, + "grad_norm": 0.8598300778559472, + "learning_rate": 0.003, + "loss": 4.1151, + "step": 9452 + }, + { + "epoch": 0.09453, + "grad_norm": 0.8624461571472201, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 9453 + }, + { + "epoch": 0.09454, + "grad_norm": 1.0851623978196565, + "learning_rate": 0.003, + "loss": 4.086, + "step": 9454 + }, + { + "epoch": 0.09455, + "grad_norm": 1.1706846745804926, + "learning_rate": 0.003, + "loss": 4.0997, + "step": 9455 + }, + { + "epoch": 0.09456, + "grad_norm": 0.720398550854311, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 9456 + }, + { + "epoch": 0.09457, + "grad_norm": 0.6428844036374872, + "learning_rate": 0.003, + "loss": 4.1011, + "step": 9457 + }, + { + "epoch": 0.09458, + "grad_norm": 0.6854122980561818, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 9458 + }, + { + "epoch": 0.09459, + "grad_norm": 0.7557470276669376, + "learning_rate": 0.003, + "loss": 4.1019, + "step": 9459 + }, + { + "epoch": 0.0946, + "grad_norm": 0.9400745909919421, + "learning_rate": 0.003, + "loss": 4.1177, + "step": 9460 + }, + { + "epoch": 0.09461, + "grad_norm": 0.8272490781668856, + "learning_rate": 0.003, + "loss": 4.1247, + "step": 9461 + }, + { + "epoch": 0.09462, + "grad_norm": 0.7264239631357388, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 9462 + }, + { + "epoch": 0.09463, + "grad_norm": 0.6420461092891769, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 9463 + }, + { + "epoch": 0.09464, + "grad_norm": 0.6005295130137699, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 9464 + }, + { + "epoch": 0.09465, + "grad_norm": 0.7261508186379931, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 9465 + }, + { + "epoch": 0.09466, + "grad_norm": 0.7958128917859779, + "learning_rate": 0.003, + "loss": 4.0981, + "step": 9466 + }, + { + "epoch": 0.09467, + "grad_norm": 0.7585869693775182, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 9467 + }, + { + "epoch": 0.09468, + "grad_norm": 0.7806606873704798, + "learning_rate": 0.003, + "loss": 4.0968, + "step": 9468 + }, + { + "epoch": 0.09469, + "grad_norm": 0.9838410664741073, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 9469 + }, + { + "epoch": 0.0947, + "grad_norm": 1.1761299535165681, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 9470 + }, + { + "epoch": 0.09471, + "grad_norm": 1.0889544845009487, + "learning_rate": 0.003, + "loss": 4.1187, + "step": 9471 + }, + { + "epoch": 0.09472, + "grad_norm": 1.0307683311785911, + "learning_rate": 0.003, + "loss": 4.0934, + "step": 9472 + }, + { + "epoch": 0.09473, + "grad_norm": 0.9564223700183074, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 9473 + }, + { + "epoch": 0.09474, + "grad_norm": 0.9376809113489067, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 9474 + }, + { + "epoch": 0.09475, + "grad_norm": 0.9618524959475816, + "learning_rate": 0.003, + "loss": 4.0907, + "step": 9475 + }, + { + "epoch": 0.09476, + "grad_norm": 0.9554715312019628, + "learning_rate": 0.003, + "loss": 4.0931, + "step": 9476 + }, + { + "epoch": 0.09477, + "grad_norm": 1.000035475540716, + "learning_rate": 0.003, + "loss": 4.093, + "step": 9477 + }, + { + "epoch": 0.09478, + "grad_norm": 1.0839210902354595, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 9478 + }, + { + "epoch": 0.09479, + "grad_norm": 0.869487169656702, + "learning_rate": 0.003, + "loss": 4.0948, + "step": 9479 + }, + { + "epoch": 0.0948, + "grad_norm": 0.7926358942114886, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 9480 + }, + { + "epoch": 0.09481, + "grad_norm": 0.6884497755914634, + "learning_rate": 0.003, + "loss": 4.1179, + "step": 9481 + }, + { + "epoch": 0.09482, + "grad_norm": 0.6845896021916732, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 9482 + }, + { + "epoch": 0.09483, + "grad_norm": 0.6566997735021866, + "learning_rate": 0.003, + "loss": 4.054, + "step": 9483 + }, + { + "epoch": 0.09484, + "grad_norm": 0.616530773620674, + "learning_rate": 0.003, + "loss": 4.1012, + "step": 9484 + }, + { + "epoch": 0.09485, + "grad_norm": 0.7642856486603846, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 9485 + }, + { + "epoch": 0.09486, + "grad_norm": 0.8665981874168233, + "learning_rate": 0.003, + "loss": 4.0973, + "step": 9486 + }, + { + "epoch": 0.09487, + "grad_norm": 1.0909398291701577, + "learning_rate": 0.003, + "loss": 4.0915, + "step": 9487 + }, + { + "epoch": 0.09488, + "grad_norm": 1.1491216476616155, + "learning_rate": 0.003, + "loss": 4.1348, + "step": 9488 + }, + { + "epoch": 0.09489, + "grad_norm": 0.8354603918248583, + "learning_rate": 0.003, + "loss": 4.0975, + "step": 9489 + }, + { + "epoch": 0.0949, + "grad_norm": 0.6895784339169947, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 9490 + }, + { + "epoch": 0.09491, + "grad_norm": 0.711524009729495, + "learning_rate": 0.003, + "loss": 4.072, + "step": 9491 + }, + { + "epoch": 0.09492, + "grad_norm": 0.8394489207958683, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 9492 + }, + { + "epoch": 0.09493, + "grad_norm": 1.2414418185211027, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 9493 + }, + { + "epoch": 0.09494, + "grad_norm": 0.8710054135901388, + "learning_rate": 0.003, + "loss": 4.1035, + "step": 9494 + }, + { + "epoch": 0.09495, + "grad_norm": 0.8645492616541143, + "learning_rate": 0.003, + "loss": 4.0936, + "step": 9495 + }, + { + "epoch": 0.09496, + "grad_norm": 0.83427100248925, + "learning_rate": 0.003, + "loss": 4.1095, + "step": 9496 + }, + { + "epoch": 0.09497, + "grad_norm": 0.7271594302410049, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 9497 + }, + { + "epoch": 0.09498, + "grad_norm": 0.641946567264643, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 9498 + }, + { + "epoch": 0.09499, + "grad_norm": 0.6170199490496553, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 9499 + }, + { + "epoch": 0.095, + "grad_norm": 0.6200424975500217, + "learning_rate": 0.003, + "loss": 4.1149, + "step": 9500 + }, + { + "epoch": 0.09501, + "grad_norm": 0.809057988961176, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 9501 + }, + { + "epoch": 0.09502, + "grad_norm": 1.1020241879146306, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 9502 + }, + { + "epoch": 0.09503, + "grad_norm": 1.030306126051077, + "learning_rate": 0.003, + "loss": 4.049, + "step": 9503 + }, + { + "epoch": 0.09504, + "grad_norm": 0.9735652853838207, + "learning_rate": 0.003, + "loss": 4.1009, + "step": 9504 + }, + { + "epoch": 0.09505, + "grad_norm": 0.9236443544745729, + "learning_rate": 0.003, + "loss": 4.0903, + "step": 9505 + }, + { + "epoch": 0.09506, + "grad_norm": 0.8592569358203569, + "learning_rate": 0.003, + "loss": 4.104, + "step": 9506 + }, + { + "epoch": 0.09507, + "grad_norm": 0.9643353093501955, + "learning_rate": 0.003, + "loss": 4.116, + "step": 9507 + }, + { + "epoch": 0.09508, + "grad_norm": 1.0137276574092238, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 9508 + }, + { + "epoch": 0.09509, + "grad_norm": 1.1660412011916976, + "learning_rate": 0.003, + "loss": 4.0934, + "step": 9509 + }, + { + "epoch": 0.0951, + "grad_norm": 1.082728289928524, + "learning_rate": 0.003, + "loss": 4.0988, + "step": 9510 + }, + { + "epoch": 0.09511, + "grad_norm": 0.9624008478957283, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 9511 + }, + { + "epoch": 0.09512, + "grad_norm": 1.0923795540142767, + "learning_rate": 0.003, + "loss": 4.0925, + "step": 9512 + }, + { + "epoch": 0.09513, + "grad_norm": 1.0328382351919922, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 9513 + }, + { + "epoch": 0.09514, + "grad_norm": 1.0089078992511966, + "learning_rate": 0.003, + "loss": 4.0887, + "step": 9514 + }, + { + "epoch": 0.09515, + "grad_norm": 0.9704873672647005, + "learning_rate": 0.003, + "loss": 4.1029, + "step": 9515 + }, + { + "epoch": 0.09516, + "grad_norm": 0.8985075408487389, + "learning_rate": 0.003, + "loss": 4.0934, + "step": 9516 + }, + { + "epoch": 0.09517, + "grad_norm": 0.9130742158343249, + "learning_rate": 0.003, + "loss": 4.0927, + "step": 9517 + }, + { + "epoch": 0.09518, + "grad_norm": 0.9244616783687439, + "learning_rate": 0.003, + "loss": 4.1091, + "step": 9518 + }, + { + "epoch": 0.09519, + "grad_norm": 0.9668574431326041, + "learning_rate": 0.003, + "loss": 4.1233, + "step": 9519 + }, + { + "epoch": 0.0952, + "grad_norm": 0.9637024791097162, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 9520 + }, + { + "epoch": 0.09521, + "grad_norm": 1.016108766606305, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 9521 + }, + { + "epoch": 0.09522, + "grad_norm": 1.0754104899740695, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 9522 + }, + { + "epoch": 0.09523, + "grad_norm": 0.9460439865375799, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 9523 + }, + { + "epoch": 0.09524, + "grad_norm": 0.8972736244759736, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 9524 + }, + { + "epoch": 0.09525, + "grad_norm": 0.8707423847198635, + "learning_rate": 0.003, + "loss": 4.0978, + "step": 9525 + }, + { + "epoch": 0.09526, + "grad_norm": 0.9298574511865992, + "learning_rate": 0.003, + "loss": 4.0934, + "step": 9526 + }, + { + "epoch": 0.09527, + "grad_norm": 0.873097150552636, + "learning_rate": 0.003, + "loss": 4.0942, + "step": 9527 + }, + { + "epoch": 0.09528, + "grad_norm": 0.861692752888038, + "learning_rate": 0.003, + "loss": 4.1144, + "step": 9528 + }, + { + "epoch": 0.09529, + "grad_norm": 0.8130110024800541, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 9529 + }, + { + "epoch": 0.0953, + "grad_norm": 0.7905443102217827, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 9530 + }, + { + "epoch": 0.09531, + "grad_norm": 0.8291019771418112, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 9531 + }, + { + "epoch": 0.09532, + "grad_norm": 0.8459154353642524, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 9532 + }, + { + "epoch": 0.09533, + "grad_norm": 0.9922895216920463, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 9533 + }, + { + "epoch": 0.09534, + "grad_norm": 1.185301105499309, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 9534 + }, + { + "epoch": 0.09535, + "grad_norm": 1.0236593208473932, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 9535 + }, + { + "epoch": 0.09536, + "grad_norm": 1.1070204390625342, + "learning_rate": 0.003, + "loss": 4.1128, + "step": 9536 + }, + { + "epoch": 0.09537, + "grad_norm": 0.9574983262523417, + "learning_rate": 0.003, + "loss": 4.1175, + "step": 9537 + }, + { + "epoch": 0.09538, + "grad_norm": 0.8493144641874563, + "learning_rate": 0.003, + "loss": 4.096, + "step": 9538 + }, + { + "epoch": 0.09539, + "grad_norm": 0.835531151551209, + "learning_rate": 0.003, + "loss": 4.1131, + "step": 9539 + }, + { + "epoch": 0.0954, + "grad_norm": 0.8283289116606958, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 9540 + }, + { + "epoch": 0.09541, + "grad_norm": 0.9341466910775755, + "learning_rate": 0.003, + "loss": 4.1036, + "step": 9541 + }, + { + "epoch": 0.09542, + "grad_norm": 1.0314622632534094, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 9542 + }, + { + "epoch": 0.09543, + "grad_norm": 1.0142586101903608, + "learning_rate": 0.003, + "loss": 4.0876, + "step": 9543 + }, + { + "epoch": 0.09544, + "grad_norm": 1.4582749640876167, + "learning_rate": 0.003, + "loss": 4.0953, + "step": 9544 + }, + { + "epoch": 0.09545, + "grad_norm": 0.6203000789802795, + "learning_rate": 0.003, + "loss": 4.0932, + "step": 9545 + }, + { + "epoch": 0.09546, + "grad_norm": 0.7933224015182839, + "learning_rate": 0.003, + "loss": 4.1106, + "step": 9546 + }, + { + "epoch": 0.09547, + "grad_norm": 0.7738375177165676, + "learning_rate": 0.003, + "loss": 4.0975, + "step": 9547 + }, + { + "epoch": 0.09548, + "grad_norm": 0.8381207027995874, + "learning_rate": 0.003, + "loss": 4.0915, + "step": 9548 + }, + { + "epoch": 0.09549, + "grad_norm": 0.8454264400374799, + "learning_rate": 0.003, + "loss": 4.1284, + "step": 9549 + }, + { + "epoch": 0.0955, + "grad_norm": 0.8076905957792103, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 9550 + }, + { + "epoch": 0.09551, + "grad_norm": 0.8632802824577473, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 9551 + }, + { + "epoch": 0.09552, + "grad_norm": 1.0768644429113419, + "learning_rate": 0.003, + "loss": 4.1435, + "step": 9552 + }, + { + "epoch": 0.09553, + "grad_norm": 1.1900364276456132, + "learning_rate": 0.003, + "loss": 4.1021, + "step": 9553 + }, + { + "epoch": 0.09554, + "grad_norm": 0.6910437093615704, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 9554 + }, + { + "epoch": 0.09555, + "grad_norm": 0.677559408017465, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 9555 + }, + { + "epoch": 0.09556, + "grad_norm": 0.8863789977775014, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 9556 + }, + { + "epoch": 0.09557, + "grad_norm": 1.2130858830229994, + "learning_rate": 0.003, + "loss": 4.1126, + "step": 9557 + }, + { + "epoch": 0.09558, + "grad_norm": 0.8078845755680898, + "learning_rate": 0.003, + "loss": 4.0944, + "step": 9558 + }, + { + "epoch": 0.09559, + "grad_norm": 0.6888474933216685, + "learning_rate": 0.003, + "loss": 4.1144, + "step": 9559 + }, + { + "epoch": 0.0956, + "grad_norm": 0.7444419985119464, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 9560 + }, + { + "epoch": 0.09561, + "grad_norm": 0.7331440318600505, + "learning_rate": 0.003, + "loss": 4.1056, + "step": 9561 + }, + { + "epoch": 0.09562, + "grad_norm": 0.6745064106478187, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 9562 + }, + { + "epoch": 0.09563, + "grad_norm": 0.7223110614602103, + "learning_rate": 0.003, + "loss": 4.1217, + "step": 9563 + }, + { + "epoch": 0.09564, + "grad_norm": 0.7954354073035838, + "learning_rate": 0.003, + "loss": 4.1098, + "step": 9564 + }, + { + "epoch": 0.09565, + "grad_norm": 0.9518848575929585, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 9565 + }, + { + "epoch": 0.09566, + "grad_norm": 1.1360554053001062, + "learning_rate": 0.003, + "loss": 4.1018, + "step": 9566 + }, + { + "epoch": 0.09567, + "grad_norm": 1.0190304984027616, + "learning_rate": 0.003, + "loss": 4.0982, + "step": 9567 + }, + { + "epoch": 0.09568, + "grad_norm": 1.1111115745846918, + "learning_rate": 0.003, + "loss": 4.1073, + "step": 9568 + }, + { + "epoch": 0.09569, + "grad_norm": 1.060656585628691, + "learning_rate": 0.003, + "loss": 4.1115, + "step": 9569 + }, + { + "epoch": 0.0957, + "grad_norm": 0.9102915228109848, + "learning_rate": 0.003, + "loss": 4.1046, + "step": 9570 + }, + { + "epoch": 0.09571, + "grad_norm": 1.018964046699271, + "learning_rate": 0.003, + "loss": 4.1063, + "step": 9571 + }, + { + "epoch": 0.09572, + "grad_norm": 1.2527068883296468, + "learning_rate": 0.003, + "loss": 4.1042, + "step": 9572 + }, + { + "epoch": 0.09573, + "grad_norm": 0.9851106487089002, + "learning_rate": 0.003, + "loss": 4.1041, + "step": 9573 + }, + { + "epoch": 0.09574, + "grad_norm": 1.0626664552153582, + "learning_rate": 0.003, + "loss": 4.1093, + "step": 9574 + }, + { + "epoch": 0.09575, + "grad_norm": 1.1535018572622826, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 9575 + }, + { + "epoch": 0.09576, + "grad_norm": 0.866406401028726, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 9576 + }, + { + "epoch": 0.09577, + "grad_norm": 0.7598448539401876, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 9577 + }, + { + "epoch": 0.09578, + "grad_norm": 0.76846069859649, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 9578 + }, + { + "epoch": 0.09579, + "grad_norm": 0.8043603561549131, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 9579 + }, + { + "epoch": 0.0958, + "grad_norm": 0.8020399563357902, + "learning_rate": 0.003, + "loss": 4.1078, + "step": 9580 + }, + { + "epoch": 0.09581, + "grad_norm": 0.7682141048439647, + "learning_rate": 0.003, + "loss": 4.1397, + "step": 9581 + }, + { + "epoch": 0.09582, + "grad_norm": 0.8015394924731275, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 9582 + }, + { + "epoch": 0.09583, + "grad_norm": 0.8226749457094206, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 9583 + }, + { + "epoch": 0.09584, + "grad_norm": 0.94933199587187, + "learning_rate": 0.003, + "loss": 4.1012, + "step": 9584 + }, + { + "epoch": 0.09585, + "grad_norm": 0.9918679044193051, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 9585 + }, + { + "epoch": 0.09586, + "grad_norm": 0.9705822187192004, + "learning_rate": 0.003, + "loss": 4.1163, + "step": 9586 + }, + { + "epoch": 0.09587, + "grad_norm": 0.9451328209074761, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 9587 + }, + { + "epoch": 0.09588, + "grad_norm": 1.2495169152311112, + "learning_rate": 0.003, + "loss": 4.1044, + "step": 9588 + }, + { + "epoch": 0.09589, + "grad_norm": 0.8554058248225164, + "learning_rate": 0.003, + "loss": 4.0993, + "step": 9589 + }, + { + "epoch": 0.0959, + "grad_norm": 0.8694247907519619, + "learning_rate": 0.003, + "loss": 4.1135, + "step": 9590 + }, + { + "epoch": 0.09591, + "grad_norm": 0.8764253050367747, + "learning_rate": 0.003, + "loss": 4.1419, + "step": 9591 + }, + { + "epoch": 0.09592, + "grad_norm": 0.928129824027457, + "learning_rate": 0.003, + "loss": 4.1067, + "step": 9592 + }, + { + "epoch": 0.09593, + "grad_norm": 0.9505672638184878, + "learning_rate": 0.003, + "loss": 4.1212, + "step": 9593 + }, + { + "epoch": 0.09594, + "grad_norm": 1.0714724754192935, + "learning_rate": 0.003, + "loss": 4.087, + "step": 9594 + }, + { + "epoch": 0.09595, + "grad_norm": 0.9819378603136486, + "learning_rate": 0.003, + "loss": 4.0884, + "step": 9595 + }, + { + "epoch": 0.09596, + "grad_norm": 1.026865992142624, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 9596 + }, + { + "epoch": 0.09597, + "grad_norm": 0.973648595397517, + "learning_rate": 0.003, + "loss": 4.0922, + "step": 9597 + }, + { + "epoch": 0.09598, + "grad_norm": 0.9850517417503974, + "learning_rate": 0.003, + "loss": 4.1032, + "step": 9598 + }, + { + "epoch": 0.09599, + "grad_norm": 0.9928584309344902, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 9599 + }, + { + "epoch": 0.096, + "grad_norm": 0.9952109010549817, + "learning_rate": 0.003, + "loss": 4.0908, + "step": 9600 + }, + { + "epoch": 0.09601, + "grad_norm": 0.9715688246997389, + "learning_rate": 0.003, + "loss": 4.0956, + "step": 9601 + }, + { + "epoch": 0.09602, + "grad_norm": 0.896038866298406, + "learning_rate": 0.003, + "loss": 4.1225, + "step": 9602 + }, + { + "epoch": 0.09603, + "grad_norm": 0.9394621805775981, + "learning_rate": 0.003, + "loss": 4.1036, + "step": 9603 + }, + { + "epoch": 0.09604, + "grad_norm": 0.8832102278330413, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 9604 + }, + { + "epoch": 0.09605, + "grad_norm": 0.8768838219761711, + "learning_rate": 0.003, + "loss": 4.1059, + "step": 9605 + }, + { + "epoch": 0.09606, + "grad_norm": 0.8777392831472427, + "learning_rate": 0.003, + "loss": 4.103, + "step": 9606 + }, + { + "epoch": 0.09607, + "grad_norm": 1.076257004258557, + "learning_rate": 0.003, + "loss": 4.1166, + "step": 9607 + }, + { + "epoch": 0.09608, + "grad_norm": 0.9490087215698528, + "learning_rate": 0.003, + "loss": 4.1224, + "step": 9608 + }, + { + "epoch": 0.09609, + "grad_norm": 0.8957046130890897, + "learning_rate": 0.003, + "loss": 4.0976, + "step": 9609 + }, + { + "epoch": 0.0961, + "grad_norm": 0.7984430379991309, + "learning_rate": 0.003, + "loss": 4.0956, + "step": 9610 + }, + { + "epoch": 0.09611, + "grad_norm": 0.7499563857935081, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 9611 + }, + { + "epoch": 0.09612, + "grad_norm": 0.7348626128667639, + "learning_rate": 0.003, + "loss": 4.0906, + "step": 9612 + }, + { + "epoch": 0.09613, + "grad_norm": 0.7649035088697457, + "learning_rate": 0.003, + "loss": 4.104, + "step": 9613 + }, + { + "epoch": 0.09614, + "grad_norm": 0.8970754465663441, + "learning_rate": 0.003, + "loss": 4.0959, + "step": 9614 + }, + { + "epoch": 0.09615, + "grad_norm": 1.0353388858633117, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 9615 + }, + { + "epoch": 0.09616, + "grad_norm": 1.0422928251395278, + "learning_rate": 0.003, + "loss": 4.103, + "step": 9616 + }, + { + "epoch": 0.09617, + "grad_norm": 0.9330622686075548, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 9617 + }, + { + "epoch": 0.09618, + "grad_norm": 0.8079910426137322, + "learning_rate": 0.003, + "loss": 4.096, + "step": 9618 + }, + { + "epoch": 0.09619, + "grad_norm": 0.7658297024476686, + "learning_rate": 0.003, + "loss": 4.0891, + "step": 9619 + }, + { + "epoch": 0.0962, + "grad_norm": 0.7549136355891398, + "learning_rate": 0.003, + "loss": 4.078, + "step": 9620 + }, + { + "epoch": 0.09621, + "grad_norm": 0.7392852138332249, + "learning_rate": 0.003, + "loss": 4.0955, + "step": 9621 + }, + { + "epoch": 0.09622, + "grad_norm": 0.8832584494387006, + "learning_rate": 0.003, + "loss": 4.082, + "step": 9622 + }, + { + "epoch": 0.09623, + "grad_norm": 1.0636469145202627, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 9623 + }, + { + "epoch": 0.09624, + "grad_norm": 1.0064354117819914, + "learning_rate": 0.003, + "loss": 4.0983, + "step": 9624 + }, + { + "epoch": 0.09625, + "grad_norm": 0.9371314262078005, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 9625 + }, + { + "epoch": 0.09626, + "grad_norm": 0.8253170716175934, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 9626 + }, + { + "epoch": 0.09627, + "grad_norm": 0.7030524329715849, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 9627 + }, + { + "epoch": 0.09628, + "grad_norm": 0.7250410432973622, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 9628 + }, + { + "epoch": 0.09629, + "grad_norm": 0.7346067220792749, + "learning_rate": 0.003, + "loss": 4.097, + "step": 9629 + }, + { + "epoch": 0.0963, + "grad_norm": 0.9046057127262703, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 9630 + }, + { + "epoch": 0.09631, + "grad_norm": 0.894854230452782, + "learning_rate": 0.003, + "loss": 4.1255, + "step": 9631 + }, + { + "epoch": 0.09632, + "grad_norm": 0.9480273147630979, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 9632 + }, + { + "epoch": 0.09633, + "grad_norm": 0.9786150037379322, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 9633 + }, + { + "epoch": 0.09634, + "grad_norm": 1.0282461128507772, + "learning_rate": 0.003, + "loss": 4.067, + "step": 9634 + }, + { + "epoch": 0.09635, + "grad_norm": 1.2837836327015333, + "learning_rate": 0.003, + "loss": 4.115, + "step": 9635 + }, + { + "epoch": 0.09636, + "grad_norm": 1.0371256794904447, + "learning_rate": 0.003, + "loss": 4.1127, + "step": 9636 + }, + { + "epoch": 0.09637, + "grad_norm": 0.9109444824213261, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 9637 + }, + { + "epoch": 0.09638, + "grad_norm": 0.8349568525937817, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 9638 + }, + { + "epoch": 0.09639, + "grad_norm": 0.8952363870351221, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 9639 + }, + { + "epoch": 0.0964, + "grad_norm": 0.8817057896870121, + "learning_rate": 0.003, + "loss": 4.1278, + "step": 9640 + }, + { + "epoch": 0.09641, + "grad_norm": 1.0165123777415066, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 9641 + }, + { + "epoch": 0.09642, + "grad_norm": 1.2088791393181662, + "learning_rate": 0.003, + "loss": 4.1269, + "step": 9642 + }, + { + "epoch": 0.09643, + "grad_norm": 0.8298242740488613, + "learning_rate": 0.003, + "loss": 4.089, + "step": 9643 + }, + { + "epoch": 0.09644, + "grad_norm": 0.7360761912273628, + "learning_rate": 0.003, + "loss": 4.1078, + "step": 9644 + }, + { + "epoch": 0.09645, + "grad_norm": 0.7427508177687894, + "learning_rate": 0.003, + "loss": 4.0972, + "step": 9645 + }, + { + "epoch": 0.09646, + "grad_norm": 0.8169170909806851, + "learning_rate": 0.003, + "loss": 4.0957, + "step": 9646 + }, + { + "epoch": 0.09647, + "grad_norm": 0.873096623162937, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 9647 + }, + { + "epoch": 0.09648, + "grad_norm": 0.9301514895225974, + "learning_rate": 0.003, + "loss": 4.1311, + "step": 9648 + }, + { + "epoch": 0.09649, + "grad_norm": 1.0261303129344521, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 9649 + }, + { + "epoch": 0.0965, + "grad_norm": 0.9858489832150986, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 9650 + }, + { + "epoch": 0.09651, + "grad_norm": 0.9207520107530078, + "learning_rate": 0.003, + "loss": 4.0982, + "step": 9651 + }, + { + "epoch": 0.09652, + "grad_norm": 0.8934066228016666, + "learning_rate": 0.003, + "loss": 4.0845, + "step": 9652 + }, + { + "epoch": 0.09653, + "grad_norm": 0.8604206377048786, + "learning_rate": 0.003, + "loss": 4.1113, + "step": 9653 + }, + { + "epoch": 0.09654, + "grad_norm": 0.9663619779941607, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 9654 + }, + { + "epoch": 0.09655, + "grad_norm": 0.91146406489948, + "learning_rate": 0.003, + "loss": 4.1002, + "step": 9655 + }, + { + "epoch": 0.09656, + "grad_norm": 0.8975020800835469, + "learning_rate": 0.003, + "loss": 4.1102, + "step": 9656 + }, + { + "epoch": 0.09657, + "grad_norm": 1.0608421766317369, + "learning_rate": 0.003, + "loss": 4.1116, + "step": 9657 + }, + { + "epoch": 0.09658, + "grad_norm": 0.9751782896772202, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 9658 + }, + { + "epoch": 0.09659, + "grad_norm": 0.8507331684637727, + "learning_rate": 0.003, + "loss": 4.1144, + "step": 9659 + }, + { + "epoch": 0.0966, + "grad_norm": 0.7797760426559466, + "learning_rate": 0.003, + "loss": 4.1066, + "step": 9660 + }, + { + "epoch": 0.09661, + "grad_norm": 0.7280346537301909, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 9661 + }, + { + "epoch": 0.09662, + "grad_norm": 0.7687521057717993, + "learning_rate": 0.003, + "loss": 4.078, + "step": 9662 + }, + { + "epoch": 0.09663, + "grad_norm": 0.8765165791530211, + "learning_rate": 0.003, + "loss": 4.1116, + "step": 9663 + }, + { + "epoch": 0.09664, + "grad_norm": 1.213269447787462, + "learning_rate": 0.003, + "loss": 4.111, + "step": 9664 + }, + { + "epoch": 0.09665, + "grad_norm": 0.9511980015499266, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 9665 + }, + { + "epoch": 0.09666, + "grad_norm": 0.8932643356990735, + "learning_rate": 0.003, + "loss": 4.1002, + "step": 9666 + }, + { + "epoch": 0.09667, + "grad_norm": 1.1991501443292183, + "learning_rate": 0.003, + "loss": 4.1147, + "step": 9667 + }, + { + "epoch": 0.09668, + "grad_norm": 0.9122924207412219, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 9668 + }, + { + "epoch": 0.09669, + "grad_norm": 0.8177666074816453, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 9669 + }, + { + "epoch": 0.0967, + "grad_norm": 0.8642756088338996, + "learning_rate": 0.003, + "loss": 4.1086, + "step": 9670 + }, + { + "epoch": 0.09671, + "grad_norm": 0.9387142926253513, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 9671 + }, + { + "epoch": 0.09672, + "grad_norm": 0.9876148943907853, + "learning_rate": 0.003, + "loss": 4.1201, + "step": 9672 + }, + { + "epoch": 0.09673, + "grad_norm": 1.183065157513859, + "learning_rate": 0.003, + "loss": 4.1216, + "step": 9673 + }, + { + "epoch": 0.09674, + "grad_norm": 0.8981167917963628, + "learning_rate": 0.003, + "loss": 4.1436, + "step": 9674 + }, + { + "epoch": 0.09675, + "grad_norm": 0.7908901876516276, + "learning_rate": 0.003, + "loss": 4.077, + "step": 9675 + }, + { + "epoch": 0.09676, + "grad_norm": 0.7836781101706645, + "learning_rate": 0.003, + "loss": 4.0989, + "step": 9676 + }, + { + "epoch": 0.09677, + "grad_norm": 0.9396215534952582, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 9677 + }, + { + "epoch": 0.09678, + "grad_norm": 1.097205685443494, + "learning_rate": 0.003, + "loss": 4.1384, + "step": 9678 + }, + { + "epoch": 0.09679, + "grad_norm": 0.9990488663831774, + "learning_rate": 0.003, + "loss": 4.1153, + "step": 9679 + }, + { + "epoch": 0.0968, + "grad_norm": 0.9171443826425181, + "learning_rate": 0.003, + "loss": 4.1057, + "step": 9680 + }, + { + "epoch": 0.09681, + "grad_norm": 1.009511628313745, + "learning_rate": 0.003, + "loss": 4.1129, + "step": 9681 + }, + { + "epoch": 0.09682, + "grad_norm": 1.2431943193297423, + "learning_rate": 0.003, + "loss": 4.1022, + "step": 9682 + }, + { + "epoch": 0.09683, + "grad_norm": 0.9350989569045836, + "learning_rate": 0.003, + "loss": 4.0919, + "step": 9683 + }, + { + "epoch": 0.09684, + "grad_norm": 0.9827551754050654, + "learning_rate": 0.003, + "loss": 4.1214, + "step": 9684 + }, + { + "epoch": 0.09685, + "grad_norm": 1.0518093494350942, + "learning_rate": 0.003, + "loss": 4.1023, + "step": 9685 + }, + { + "epoch": 0.09686, + "grad_norm": 1.1163299364855512, + "learning_rate": 0.003, + "loss": 4.1101, + "step": 9686 + }, + { + "epoch": 0.09687, + "grad_norm": 0.8195919781179394, + "learning_rate": 0.003, + "loss": 4.105, + "step": 9687 + }, + { + "epoch": 0.09688, + "grad_norm": 0.8190869432151955, + "learning_rate": 0.003, + "loss": 4.1337, + "step": 9688 + }, + { + "epoch": 0.09689, + "grad_norm": 0.916773205951842, + "learning_rate": 0.003, + "loss": 4.0902, + "step": 9689 + }, + { + "epoch": 0.0969, + "grad_norm": 0.8066088249172502, + "learning_rate": 0.003, + "loss": 4.1004, + "step": 9690 + }, + { + "epoch": 0.09691, + "grad_norm": 0.7401326149708389, + "learning_rate": 0.003, + "loss": 4.108, + "step": 9691 + }, + { + "epoch": 0.09692, + "grad_norm": 0.6424151800103749, + "learning_rate": 0.003, + "loss": 4.1052, + "step": 9692 + }, + { + "epoch": 0.09693, + "grad_norm": 0.6299822425823625, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 9693 + }, + { + "epoch": 0.09694, + "grad_norm": 0.6362006844765989, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 9694 + }, + { + "epoch": 0.09695, + "grad_norm": 0.6132233963780619, + "learning_rate": 0.003, + "loss": 4.0906, + "step": 9695 + }, + { + "epoch": 0.09696, + "grad_norm": 0.6660368458929525, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 9696 + }, + { + "epoch": 0.09697, + "grad_norm": 0.7544139461670122, + "learning_rate": 0.003, + "loss": 4.0888, + "step": 9697 + }, + { + "epoch": 0.09698, + "grad_norm": 0.8497357368030164, + "learning_rate": 0.003, + "loss": 4.1028, + "step": 9698 + }, + { + "epoch": 0.09699, + "grad_norm": 0.9145155762891021, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 9699 + }, + { + "epoch": 0.097, + "grad_norm": 0.8975185216089709, + "learning_rate": 0.003, + "loss": 4.1153, + "step": 9700 + }, + { + "epoch": 0.09701, + "grad_norm": 0.9835477897845034, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 9701 + }, + { + "epoch": 0.09702, + "grad_norm": 1.2588913749038257, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 9702 + }, + { + "epoch": 0.09703, + "grad_norm": 1.231889122377567, + "learning_rate": 0.003, + "loss": 4.0918, + "step": 9703 + }, + { + "epoch": 0.09704, + "grad_norm": 0.9026679969753263, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 9704 + }, + { + "epoch": 0.09705, + "grad_norm": 0.917096876663794, + "learning_rate": 0.003, + "loss": 4.08, + "step": 9705 + }, + { + "epoch": 0.09706, + "grad_norm": 1.1210979563105794, + "learning_rate": 0.003, + "loss": 4.1014, + "step": 9706 + }, + { + "epoch": 0.09707, + "grad_norm": 1.046684770518191, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 9707 + }, + { + "epoch": 0.09708, + "grad_norm": 0.8586665214097383, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 9708 + }, + { + "epoch": 0.09709, + "grad_norm": 0.8399659702694784, + "learning_rate": 0.003, + "loss": 4.0948, + "step": 9709 + }, + { + "epoch": 0.0971, + "grad_norm": 0.9061619367314612, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 9710 + }, + { + "epoch": 0.09711, + "grad_norm": 0.869420827619381, + "learning_rate": 0.003, + "loss": 4.0923, + "step": 9711 + }, + { + "epoch": 0.09712, + "grad_norm": 0.9484632391522392, + "learning_rate": 0.003, + "loss": 4.1323, + "step": 9712 + }, + { + "epoch": 0.09713, + "grad_norm": 0.9344386880847007, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 9713 + }, + { + "epoch": 0.09714, + "grad_norm": 0.8191433536044963, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 9714 + }, + { + "epoch": 0.09715, + "grad_norm": 0.9260497381898702, + "learning_rate": 0.003, + "loss": 4.0997, + "step": 9715 + }, + { + "epoch": 0.09716, + "grad_norm": 1.2168835302778682, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 9716 + }, + { + "epoch": 0.09717, + "grad_norm": 1.1755192129415277, + "learning_rate": 0.003, + "loss": 4.125, + "step": 9717 + }, + { + "epoch": 0.09718, + "grad_norm": 0.8578207270174794, + "learning_rate": 0.003, + "loss": 4.0891, + "step": 9718 + }, + { + "epoch": 0.09719, + "grad_norm": 0.7982551490113904, + "learning_rate": 0.003, + "loss": 4.079, + "step": 9719 + }, + { + "epoch": 0.0972, + "grad_norm": 0.8181582924013971, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 9720 + }, + { + "epoch": 0.09721, + "grad_norm": 0.7803993304194405, + "learning_rate": 0.003, + "loss": 4.1149, + "step": 9721 + }, + { + "epoch": 0.09722, + "grad_norm": 0.7217245724741701, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 9722 + }, + { + "epoch": 0.09723, + "grad_norm": 0.8297726003767051, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 9723 + }, + { + "epoch": 0.09724, + "grad_norm": 0.8385216713500718, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 9724 + }, + { + "epoch": 0.09725, + "grad_norm": 0.8535894035344416, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 9725 + }, + { + "epoch": 0.09726, + "grad_norm": 0.9997627703805871, + "learning_rate": 0.003, + "loss": 4.1014, + "step": 9726 + }, + { + "epoch": 0.09727, + "grad_norm": 1.2871027951432406, + "learning_rate": 0.003, + "loss": 4.1, + "step": 9727 + }, + { + "epoch": 0.09728, + "grad_norm": 0.9786561845767965, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 9728 + }, + { + "epoch": 0.09729, + "grad_norm": 1.056518217713997, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 9729 + }, + { + "epoch": 0.0973, + "grad_norm": 1.0746377651628496, + "learning_rate": 0.003, + "loss": 4.1021, + "step": 9730 + }, + { + "epoch": 0.09731, + "grad_norm": 1.1074902816443135, + "learning_rate": 0.003, + "loss": 4.057, + "step": 9731 + }, + { + "epoch": 0.09732, + "grad_norm": 0.9021480346313361, + "learning_rate": 0.003, + "loss": 4.0933, + "step": 9732 + }, + { + "epoch": 0.09733, + "grad_norm": 0.8698953887525058, + "learning_rate": 0.003, + "loss": 4.0854, + "step": 9733 + }, + { + "epoch": 0.09734, + "grad_norm": 0.7657365794014724, + "learning_rate": 0.003, + "loss": 4.1132, + "step": 9734 + }, + { + "epoch": 0.09735, + "grad_norm": 0.7136532065218357, + "learning_rate": 0.003, + "loss": 4.1063, + "step": 9735 + }, + { + "epoch": 0.09736, + "grad_norm": 0.6717357230030461, + "learning_rate": 0.003, + "loss": 4.0819, + "step": 9736 + }, + { + "epoch": 0.09737, + "grad_norm": 0.5638315775801472, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 9737 + }, + { + "epoch": 0.09738, + "grad_norm": 0.6110116723814848, + "learning_rate": 0.003, + "loss": 4.0982, + "step": 9738 + }, + { + "epoch": 0.09739, + "grad_norm": 0.6086506139891856, + "learning_rate": 0.003, + "loss": 4.0946, + "step": 9739 + }, + { + "epoch": 0.0974, + "grad_norm": 0.7935050152052187, + "learning_rate": 0.003, + "loss": 4.0868, + "step": 9740 + }, + { + "epoch": 0.09741, + "grad_norm": 1.1123363599125007, + "learning_rate": 0.003, + "loss": 4.1375, + "step": 9741 + }, + { + "epoch": 0.09742, + "grad_norm": 1.0030554366325666, + "learning_rate": 0.003, + "loss": 4.076, + "step": 9742 + }, + { + "epoch": 0.09743, + "grad_norm": 0.8267397016080698, + "learning_rate": 0.003, + "loss": 4.0996, + "step": 9743 + }, + { + "epoch": 0.09744, + "grad_norm": 0.7717073469243041, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 9744 + }, + { + "epoch": 0.09745, + "grad_norm": 0.798472215693979, + "learning_rate": 0.003, + "loss": 4.1156, + "step": 9745 + }, + { + "epoch": 0.09746, + "grad_norm": 0.9589191423737808, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 9746 + }, + { + "epoch": 0.09747, + "grad_norm": 1.116733109628546, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 9747 + }, + { + "epoch": 0.09748, + "grad_norm": 0.8375628533401334, + "learning_rate": 0.003, + "loss": 4.119, + "step": 9748 + }, + { + "epoch": 0.09749, + "grad_norm": 0.8553001862226853, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 9749 + }, + { + "epoch": 0.0975, + "grad_norm": 0.9088453607552779, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 9750 + }, + { + "epoch": 0.09751, + "grad_norm": 0.8569486483584623, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 9751 + }, + { + "epoch": 0.09752, + "grad_norm": 0.7380045417599184, + "learning_rate": 0.003, + "loss": 4.0938, + "step": 9752 + }, + { + "epoch": 0.09753, + "grad_norm": 0.8851823127379947, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 9753 + }, + { + "epoch": 0.09754, + "grad_norm": 1.0358214518222049, + "learning_rate": 0.003, + "loss": 4.102, + "step": 9754 + }, + { + "epoch": 0.09755, + "grad_norm": 0.9439521181468254, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 9755 + }, + { + "epoch": 0.09756, + "grad_norm": 0.8198513848838082, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 9756 + }, + { + "epoch": 0.09757, + "grad_norm": 0.829860888973659, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 9757 + }, + { + "epoch": 0.09758, + "grad_norm": 0.9561785595312966, + "learning_rate": 0.003, + "loss": 4.0857, + "step": 9758 + }, + { + "epoch": 0.09759, + "grad_norm": 1.2593905860698673, + "learning_rate": 0.003, + "loss": 4.0945, + "step": 9759 + }, + { + "epoch": 0.0976, + "grad_norm": 1.046642575447866, + "learning_rate": 0.003, + "loss": 4.0956, + "step": 9760 + }, + { + "epoch": 0.09761, + "grad_norm": 1.207472022721697, + "learning_rate": 0.003, + "loss": 4.0947, + "step": 9761 + }, + { + "epoch": 0.09762, + "grad_norm": 0.9869421396722704, + "learning_rate": 0.003, + "loss": 4.1097, + "step": 9762 + }, + { + "epoch": 0.09763, + "grad_norm": 1.0380173466462297, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 9763 + }, + { + "epoch": 0.09764, + "grad_norm": 1.1529962238971903, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 9764 + }, + { + "epoch": 0.09765, + "grad_norm": 0.8259926007525762, + "learning_rate": 0.003, + "loss": 4.1206, + "step": 9765 + }, + { + "epoch": 0.09766, + "grad_norm": 0.8887761626705335, + "learning_rate": 0.003, + "loss": 4.0958, + "step": 9766 + }, + { + "epoch": 0.09767, + "grad_norm": 0.8128663843712138, + "learning_rate": 0.003, + "loss": 4.1044, + "step": 9767 + }, + { + "epoch": 0.09768, + "grad_norm": 0.7660324180116674, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 9768 + }, + { + "epoch": 0.09769, + "grad_norm": 0.7219697561948825, + "learning_rate": 0.003, + "loss": 4.1111, + "step": 9769 + }, + { + "epoch": 0.0977, + "grad_norm": 0.7328719015179833, + "learning_rate": 0.003, + "loss": 4.1082, + "step": 9770 + }, + { + "epoch": 0.09771, + "grad_norm": 0.6934477189530808, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 9771 + }, + { + "epoch": 0.09772, + "grad_norm": 0.8226483433677332, + "learning_rate": 0.003, + "loss": 4.073, + "step": 9772 + }, + { + "epoch": 0.09773, + "grad_norm": 1.0407467139689701, + "learning_rate": 0.003, + "loss": 4.0947, + "step": 9773 + }, + { + "epoch": 0.09774, + "grad_norm": 1.379965710605147, + "learning_rate": 0.003, + "loss": 4.1208, + "step": 9774 + }, + { + "epoch": 0.09775, + "grad_norm": 0.743770947000131, + "learning_rate": 0.003, + "loss": 4.1085, + "step": 9775 + }, + { + "epoch": 0.09776, + "grad_norm": 0.7186119882928109, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 9776 + }, + { + "epoch": 0.09777, + "grad_norm": 0.7494079596750043, + "learning_rate": 0.003, + "loss": 4.107, + "step": 9777 + }, + { + "epoch": 0.09778, + "grad_norm": 0.8250803113950317, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 9778 + }, + { + "epoch": 0.09779, + "grad_norm": 0.9604869205387597, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 9779 + }, + { + "epoch": 0.0978, + "grad_norm": 1.2294622769995196, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 9780 + }, + { + "epoch": 0.09781, + "grad_norm": 0.8713194051271618, + "learning_rate": 0.003, + "loss": 4.0985, + "step": 9781 + }, + { + "epoch": 0.09782, + "grad_norm": 0.8347924061909529, + "learning_rate": 0.003, + "loss": 4.0967, + "step": 9782 + }, + { + "epoch": 0.09783, + "grad_norm": 1.1395358663076178, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 9783 + }, + { + "epoch": 0.09784, + "grad_norm": 1.1019088831778014, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 9784 + }, + { + "epoch": 0.09785, + "grad_norm": 1.169417153963683, + "learning_rate": 0.003, + "loss": 4.094, + "step": 9785 + }, + { + "epoch": 0.09786, + "grad_norm": 1.0034794245848353, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 9786 + }, + { + "epoch": 0.09787, + "grad_norm": 1.0284399680460898, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 9787 + }, + { + "epoch": 0.09788, + "grad_norm": 0.9057070287775122, + "learning_rate": 0.003, + "loss": 4.1, + "step": 9788 + }, + { + "epoch": 0.09789, + "grad_norm": 0.8560028986393307, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 9789 + }, + { + "epoch": 0.0979, + "grad_norm": 0.9296582713729994, + "learning_rate": 0.003, + "loss": 4.0996, + "step": 9790 + }, + { + "epoch": 0.09791, + "grad_norm": 1.1827366857116655, + "learning_rate": 0.003, + "loss": 4.1113, + "step": 9791 + }, + { + "epoch": 0.09792, + "grad_norm": 1.0204010831990022, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 9792 + }, + { + "epoch": 0.09793, + "grad_norm": 0.9153519628354887, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 9793 + }, + { + "epoch": 0.09794, + "grad_norm": 0.9068739502438202, + "learning_rate": 0.003, + "loss": 4.1016, + "step": 9794 + }, + { + "epoch": 0.09795, + "grad_norm": 0.9496301652018498, + "learning_rate": 0.003, + "loss": 4.1131, + "step": 9795 + }, + { + "epoch": 0.09796, + "grad_norm": 0.9958118680271525, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 9796 + }, + { + "epoch": 0.09797, + "grad_norm": 0.9557782092056588, + "learning_rate": 0.003, + "loss": 4.094, + "step": 9797 + }, + { + "epoch": 0.09798, + "grad_norm": 0.8533142469787229, + "learning_rate": 0.003, + "loss": 4.0878, + "step": 9798 + }, + { + "epoch": 0.09799, + "grad_norm": 0.8190713601356191, + "learning_rate": 0.003, + "loss": 4.12, + "step": 9799 + }, + { + "epoch": 0.098, + "grad_norm": 0.9429777135714619, + "learning_rate": 0.003, + "loss": 4.0969, + "step": 9800 + }, + { + "epoch": 0.09801, + "grad_norm": 1.0012945094084325, + "learning_rate": 0.003, + "loss": 4.1072, + "step": 9801 + }, + { + "epoch": 0.09802, + "grad_norm": 0.9521250021764239, + "learning_rate": 0.003, + "loss": 4.0886, + "step": 9802 + }, + { + "epoch": 0.09803, + "grad_norm": 1.2251473492074867, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 9803 + }, + { + "epoch": 0.09804, + "grad_norm": 0.9134016069934916, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 9804 + }, + { + "epoch": 0.09805, + "grad_norm": 0.8286943653648622, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 9805 + }, + { + "epoch": 0.09806, + "grad_norm": 0.7519693958000754, + "learning_rate": 0.003, + "loss": 4.069, + "step": 9806 + }, + { + "epoch": 0.09807, + "grad_norm": 0.7722973199567947, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 9807 + }, + { + "epoch": 0.09808, + "grad_norm": 0.74781146199721, + "learning_rate": 0.003, + "loss": 4.1142, + "step": 9808 + }, + { + "epoch": 0.09809, + "grad_norm": 0.7364400647725448, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 9809 + }, + { + "epoch": 0.0981, + "grad_norm": 0.7200212527489623, + "learning_rate": 0.003, + "loss": 4.1092, + "step": 9810 + }, + { + "epoch": 0.09811, + "grad_norm": 0.7544356048465244, + "learning_rate": 0.003, + "loss": 4.1094, + "step": 9811 + }, + { + "epoch": 0.09812, + "grad_norm": 0.7686580462245453, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 9812 + }, + { + "epoch": 0.09813, + "grad_norm": 0.984346883906706, + "learning_rate": 0.003, + "loss": 4.0965, + "step": 9813 + }, + { + "epoch": 0.09814, + "grad_norm": 1.2700083355049891, + "learning_rate": 0.003, + "loss": 4.0855, + "step": 9814 + }, + { + "epoch": 0.09815, + "grad_norm": 0.9442514607634342, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 9815 + }, + { + "epoch": 0.09816, + "grad_norm": 1.0077025937052435, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 9816 + }, + { + "epoch": 0.09817, + "grad_norm": 0.971550909174614, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 9817 + }, + { + "epoch": 0.09818, + "grad_norm": 1.0193923220149106, + "learning_rate": 0.003, + "loss": 4.1056, + "step": 9818 + }, + { + "epoch": 0.09819, + "grad_norm": 1.3161986032712076, + "learning_rate": 0.003, + "loss": 4.1133, + "step": 9819 + }, + { + "epoch": 0.0982, + "grad_norm": 0.8557332761380659, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 9820 + }, + { + "epoch": 0.09821, + "grad_norm": 0.7233644886267359, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 9821 + }, + { + "epoch": 0.09822, + "grad_norm": 0.8326436608697041, + "learning_rate": 0.003, + "loss": 4.1006, + "step": 9822 + }, + { + "epoch": 0.09823, + "grad_norm": 0.8980854204875377, + "learning_rate": 0.003, + "loss": 4.1099, + "step": 9823 + }, + { + "epoch": 0.09824, + "grad_norm": 1.2283773516871788, + "learning_rate": 0.003, + "loss": 4.105, + "step": 9824 + }, + { + "epoch": 0.09825, + "grad_norm": 0.7785325214207814, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 9825 + }, + { + "epoch": 0.09826, + "grad_norm": 0.597043372236362, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 9826 + }, + { + "epoch": 0.09827, + "grad_norm": 0.6464758597311613, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 9827 + }, + { + "epoch": 0.09828, + "grad_norm": 0.8139034532487643, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 9828 + }, + { + "epoch": 0.09829, + "grad_norm": 1.244278860848288, + "learning_rate": 0.003, + "loss": 4.0872, + "step": 9829 + }, + { + "epoch": 0.0983, + "grad_norm": 0.856108216172038, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 9830 + }, + { + "epoch": 0.09831, + "grad_norm": 0.6433844409257128, + "learning_rate": 0.003, + "loss": 4.0957, + "step": 9831 + }, + { + "epoch": 0.09832, + "grad_norm": 0.7326548261726062, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 9832 + }, + { + "epoch": 0.09833, + "grad_norm": 0.7997173970955105, + "learning_rate": 0.003, + "loss": 4.1088, + "step": 9833 + }, + { + "epoch": 0.09834, + "grad_norm": 0.9405671127384347, + "learning_rate": 0.003, + "loss": 4.0984, + "step": 9834 + }, + { + "epoch": 0.09835, + "grad_norm": 1.131659672629919, + "learning_rate": 0.003, + "loss": 4.0995, + "step": 9835 + }, + { + "epoch": 0.09836, + "grad_norm": 0.997538927089926, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 9836 + }, + { + "epoch": 0.09837, + "grad_norm": 1.0175411430188779, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 9837 + }, + { + "epoch": 0.09838, + "grad_norm": 0.9799166460402474, + "learning_rate": 0.003, + "loss": 4.085, + "step": 9838 + }, + { + "epoch": 0.09839, + "grad_norm": 1.0396590201522609, + "learning_rate": 0.003, + "loss": 4.1259, + "step": 9839 + }, + { + "epoch": 0.0984, + "grad_norm": 1.1282086069881085, + "learning_rate": 0.003, + "loss": 4.0965, + "step": 9840 + }, + { + "epoch": 0.09841, + "grad_norm": 0.8198231903899644, + "learning_rate": 0.003, + "loss": 4.1053, + "step": 9841 + }, + { + "epoch": 0.09842, + "grad_norm": 0.9217921401090745, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 9842 + }, + { + "epoch": 0.09843, + "grad_norm": 1.0197186165919299, + "learning_rate": 0.003, + "loss": 4.0887, + "step": 9843 + }, + { + "epoch": 0.09844, + "grad_norm": 1.2101350812829232, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 9844 + }, + { + "epoch": 0.09845, + "grad_norm": 0.8651149977324143, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 9845 + }, + { + "epoch": 0.09846, + "grad_norm": 0.7265952518823648, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 9846 + }, + { + "epoch": 0.09847, + "grad_norm": 0.6887147254332203, + "learning_rate": 0.003, + "loss": 4.0926, + "step": 9847 + }, + { + "epoch": 0.09848, + "grad_norm": 0.7507853589418222, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 9848 + }, + { + "epoch": 0.09849, + "grad_norm": 1.0036628039255613, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 9849 + }, + { + "epoch": 0.0985, + "grad_norm": 1.2748454320535545, + "learning_rate": 0.003, + "loss": 4.1194, + "step": 9850 + }, + { + "epoch": 0.09851, + "grad_norm": 0.8314804970803586, + "learning_rate": 0.003, + "loss": 4.1185, + "step": 9851 + }, + { + "epoch": 0.09852, + "grad_norm": 0.817675377067337, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 9852 + }, + { + "epoch": 0.09853, + "grad_norm": 0.7841091435659291, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 9853 + }, + { + "epoch": 0.09854, + "grad_norm": 0.6372546033409912, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 9854 + }, + { + "epoch": 0.09855, + "grad_norm": 0.679618174317228, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 9855 + }, + { + "epoch": 0.09856, + "grad_norm": 0.7500362277289165, + "learning_rate": 0.003, + "loss": 4.11, + "step": 9856 + }, + { + "epoch": 0.09857, + "grad_norm": 0.8068873392409563, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 9857 + }, + { + "epoch": 0.09858, + "grad_norm": 0.7652166764948078, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 9858 + }, + { + "epoch": 0.09859, + "grad_norm": 0.8218029235936964, + "learning_rate": 0.003, + "loss": 4.1203, + "step": 9859 + }, + { + "epoch": 0.0986, + "grad_norm": 0.843394170131785, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 9860 + }, + { + "epoch": 0.09861, + "grad_norm": 0.9379505462471198, + "learning_rate": 0.003, + "loss": 4.0881, + "step": 9861 + }, + { + "epoch": 0.09862, + "grad_norm": 1.32072231867818, + "learning_rate": 0.003, + "loss": 4.112, + "step": 9862 + }, + { + "epoch": 0.09863, + "grad_norm": 1.0523750210753542, + "learning_rate": 0.003, + "loss": 4.1123, + "step": 9863 + }, + { + "epoch": 0.09864, + "grad_norm": 0.9406173426354154, + "learning_rate": 0.003, + "loss": 4.1149, + "step": 9864 + }, + { + "epoch": 0.09865, + "grad_norm": 0.8931432490745203, + "learning_rate": 0.003, + "loss": 4.093, + "step": 9865 + }, + { + "epoch": 0.09866, + "grad_norm": 0.9699347096958343, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 9866 + }, + { + "epoch": 0.09867, + "grad_norm": 1.046711651416207, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 9867 + }, + { + "epoch": 0.09868, + "grad_norm": 1.0808042490594871, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 9868 + }, + { + "epoch": 0.09869, + "grad_norm": 1.1961822678848502, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 9869 + }, + { + "epoch": 0.0987, + "grad_norm": 0.9228079748238792, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 9870 + }, + { + "epoch": 0.09871, + "grad_norm": 0.8545983336178743, + "learning_rate": 0.003, + "loss": 4.1474, + "step": 9871 + }, + { + "epoch": 0.09872, + "grad_norm": 0.8784832785791439, + "learning_rate": 0.003, + "loss": 4.1026, + "step": 9872 + }, + { + "epoch": 0.09873, + "grad_norm": 1.070280498696221, + "learning_rate": 0.003, + "loss": 4.0946, + "step": 9873 + }, + { + "epoch": 0.09874, + "grad_norm": 1.0439215572833065, + "learning_rate": 0.003, + "loss": 4.1139, + "step": 9874 + }, + { + "epoch": 0.09875, + "grad_norm": 1.0814551372745504, + "learning_rate": 0.003, + "loss": 4.0915, + "step": 9875 + }, + { + "epoch": 0.09876, + "grad_norm": 0.8511905866205883, + "learning_rate": 0.003, + "loss": 4.0915, + "step": 9876 + }, + { + "epoch": 0.09877, + "grad_norm": 0.811275306253899, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 9877 + }, + { + "epoch": 0.09878, + "grad_norm": 1.249209967465114, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 9878 + }, + { + "epoch": 0.09879, + "grad_norm": 0.9304246999674388, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 9879 + }, + { + "epoch": 0.0988, + "grad_norm": 0.9414369214716679, + "learning_rate": 0.003, + "loss": 4.069, + "step": 9880 + }, + { + "epoch": 0.09881, + "grad_norm": 1.0878622028222777, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 9881 + }, + { + "epoch": 0.09882, + "grad_norm": 0.9240651030772346, + "learning_rate": 0.003, + "loss": 4.1051, + "step": 9882 + }, + { + "epoch": 0.09883, + "grad_norm": 0.7991058410005727, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 9883 + }, + { + "epoch": 0.09884, + "grad_norm": 0.9050441156700046, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 9884 + }, + { + "epoch": 0.09885, + "grad_norm": 1.1399203818579307, + "learning_rate": 0.003, + "loss": 4.0881, + "step": 9885 + }, + { + "epoch": 0.09886, + "grad_norm": 0.8766884786296881, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 9886 + }, + { + "epoch": 0.09887, + "grad_norm": 0.801700624949135, + "learning_rate": 0.003, + "loss": 4.1047, + "step": 9887 + }, + { + "epoch": 0.09888, + "grad_norm": 0.8628961038287333, + "learning_rate": 0.003, + "loss": 4.1255, + "step": 9888 + }, + { + "epoch": 0.09889, + "grad_norm": 0.7467079097142902, + "learning_rate": 0.003, + "loss": 4.071, + "step": 9889 + }, + { + "epoch": 0.0989, + "grad_norm": 0.7726770182387934, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 9890 + }, + { + "epoch": 0.09891, + "grad_norm": 0.9462651212194271, + "learning_rate": 0.003, + "loss": 4.1005, + "step": 9891 + }, + { + "epoch": 0.09892, + "grad_norm": 1.139679261250144, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 9892 + }, + { + "epoch": 0.09893, + "grad_norm": 1.0951119397642, + "learning_rate": 0.003, + "loss": 4.0947, + "step": 9893 + }, + { + "epoch": 0.09894, + "grad_norm": 0.905062924614346, + "learning_rate": 0.003, + "loss": 4.1049, + "step": 9894 + }, + { + "epoch": 0.09895, + "grad_norm": 0.7933902578399411, + "learning_rate": 0.003, + "loss": 4.1002, + "step": 9895 + }, + { + "epoch": 0.09896, + "grad_norm": 0.704722792642583, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 9896 + }, + { + "epoch": 0.09897, + "grad_norm": 0.6940718562411224, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 9897 + }, + { + "epoch": 0.09898, + "grad_norm": 0.8845748690268396, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 9898 + }, + { + "epoch": 0.09899, + "grad_norm": 1.0988377109535674, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 9899 + }, + { + "epoch": 0.099, + "grad_norm": 1.1105673173122586, + "learning_rate": 0.003, + "loss": 4.1227, + "step": 9900 + }, + { + "epoch": 0.09901, + "grad_norm": 0.9810554294045679, + "learning_rate": 0.003, + "loss": 4.1031, + "step": 9901 + }, + { + "epoch": 0.09902, + "grad_norm": 1.088761843451906, + "learning_rate": 0.003, + "loss": 4.1059, + "step": 9902 + }, + { + "epoch": 0.09903, + "grad_norm": 0.8865852858603028, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 9903 + }, + { + "epoch": 0.09904, + "grad_norm": 1.043616089079261, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 9904 + }, + { + "epoch": 0.09905, + "grad_norm": 1.068953951671368, + "learning_rate": 0.003, + "loss": 4.0947, + "step": 9905 + }, + { + "epoch": 0.09906, + "grad_norm": 1.044686025986228, + "learning_rate": 0.003, + "loss": 4.1099, + "step": 9906 + }, + { + "epoch": 0.09907, + "grad_norm": 0.8886077091593347, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 9907 + }, + { + "epoch": 0.09908, + "grad_norm": 0.8555950937756621, + "learning_rate": 0.003, + "loss": 4.1026, + "step": 9908 + }, + { + "epoch": 0.09909, + "grad_norm": 0.7817570855188947, + "learning_rate": 0.003, + "loss": 4.0984, + "step": 9909 + }, + { + "epoch": 0.0991, + "grad_norm": 0.8362250403473259, + "learning_rate": 0.003, + "loss": 4.0942, + "step": 9910 + }, + { + "epoch": 0.09911, + "grad_norm": 0.9500328133313455, + "learning_rate": 0.003, + "loss": 4.1008, + "step": 9911 + }, + { + "epoch": 0.09912, + "grad_norm": 1.031805969191571, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 9912 + }, + { + "epoch": 0.09913, + "grad_norm": 1.0138965400557471, + "learning_rate": 0.003, + "loss": 4.1154, + "step": 9913 + }, + { + "epoch": 0.09914, + "grad_norm": 0.9678001875339941, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 9914 + }, + { + "epoch": 0.09915, + "grad_norm": 0.9475987221695131, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 9915 + }, + { + "epoch": 0.09916, + "grad_norm": 0.9451880390051262, + "learning_rate": 0.003, + "loss": 4.1057, + "step": 9916 + }, + { + "epoch": 0.09917, + "grad_norm": 0.9072128377197956, + "learning_rate": 0.003, + "loss": 4.108, + "step": 9917 + }, + { + "epoch": 0.09918, + "grad_norm": 0.9058874365902825, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 9918 + }, + { + "epoch": 0.09919, + "grad_norm": 0.9988923576329186, + "learning_rate": 0.003, + "loss": 4.122, + "step": 9919 + }, + { + "epoch": 0.0992, + "grad_norm": 0.972034979186135, + "learning_rate": 0.003, + "loss": 4.1062, + "step": 9920 + }, + { + "epoch": 0.09921, + "grad_norm": 0.9407505238435098, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 9921 + }, + { + "epoch": 0.09922, + "grad_norm": 1.0709663672999241, + "learning_rate": 0.003, + "loss": 4.1022, + "step": 9922 + }, + { + "epoch": 0.09923, + "grad_norm": 1.011390937859059, + "learning_rate": 0.003, + "loss": 4.1031, + "step": 9923 + }, + { + "epoch": 0.09924, + "grad_norm": 1.0734808599050338, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 9924 + }, + { + "epoch": 0.09925, + "grad_norm": 0.9724775683011624, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 9925 + }, + { + "epoch": 0.09926, + "grad_norm": 0.9858754523172633, + "learning_rate": 0.003, + "loss": 4.097, + "step": 9926 + }, + { + "epoch": 0.09927, + "grad_norm": 1.0026350425415425, + "learning_rate": 0.003, + "loss": 4.1296, + "step": 9927 + }, + { + "epoch": 0.09928, + "grad_norm": 0.9618825289266347, + "learning_rate": 0.003, + "loss": 4.085, + "step": 9928 + }, + { + "epoch": 0.09929, + "grad_norm": 0.8720935243540162, + "learning_rate": 0.003, + "loss": 4.0939, + "step": 9929 + }, + { + "epoch": 0.0993, + "grad_norm": 0.8729170421206854, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 9930 + }, + { + "epoch": 0.09931, + "grad_norm": 0.9387248454992504, + "learning_rate": 0.003, + "loss": 4.0886, + "step": 9931 + }, + { + "epoch": 0.09932, + "grad_norm": 1.0565356457114414, + "learning_rate": 0.003, + "loss": 4.0997, + "step": 9932 + }, + { + "epoch": 0.09933, + "grad_norm": 0.8625813163715004, + "learning_rate": 0.003, + "loss": 4.0926, + "step": 9933 + }, + { + "epoch": 0.09934, + "grad_norm": 0.7002720895141619, + "learning_rate": 0.003, + "loss": 4.0966, + "step": 9934 + }, + { + "epoch": 0.09935, + "grad_norm": 0.7613232133923477, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 9935 + }, + { + "epoch": 0.09936, + "grad_norm": 1.0484981670115654, + "learning_rate": 0.003, + "loss": 4.0926, + "step": 9936 + }, + { + "epoch": 0.09937, + "grad_norm": 1.2541580708606934, + "learning_rate": 0.003, + "loss": 4.064, + "step": 9937 + }, + { + "epoch": 0.09938, + "grad_norm": 0.7679423053752109, + "learning_rate": 0.003, + "loss": 4.0958, + "step": 9938 + }, + { + "epoch": 0.09939, + "grad_norm": 0.7623951555699858, + "learning_rate": 0.003, + "loss": 4.0961, + "step": 9939 + }, + { + "epoch": 0.0994, + "grad_norm": 0.743467827926508, + "learning_rate": 0.003, + "loss": 4.1149, + "step": 9940 + }, + { + "epoch": 0.09941, + "grad_norm": 0.8158817381706474, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 9941 + }, + { + "epoch": 0.09942, + "grad_norm": 0.9477928068113599, + "learning_rate": 0.003, + "loss": 4.087, + "step": 9942 + }, + { + "epoch": 0.09943, + "grad_norm": 0.8989443009796476, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 9943 + }, + { + "epoch": 0.09944, + "grad_norm": 0.8465160184457518, + "learning_rate": 0.003, + "loss": 4.1195, + "step": 9944 + }, + { + "epoch": 0.09945, + "grad_norm": 0.96077143421336, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 9945 + }, + { + "epoch": 0.09946, + "grad_norm": 0.982634646562586, + "learning_rate": 0.003, + "loss": 4.1104, + "step": 9946 + }, + { + "epoch": 0.09947, + "grad_norm": 0.9829649395908524, + "learning_rate": 0.003, + "loss": 4.1031, + "step": 9947 + }, + { + "epoch": 0.09948, + "grad_norm": 1.0904657767502883, + "learning_rate": 0.003, + "loss": 4.1043, + "step": 9948 + }, + { + "epoch": 0.09949, + "grad_norm": 0.958754249548164, + "learning_rate": 0.003, + "loss": 4.1023, + "step": 9949 + }, + { + "epoch": 0.0995, + "grad_norm": 0.941102523885678, + "learning_rate": 0.003, + "loss": 4.0933, + "step": 9950 + }, + { + "epoch": 0.09951, + "grad_norm": 0.9547561697534392, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 9951 + }, + { + "epoch": 0.09952, + "grad_norm": 1.0369210267712394, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 9952 + }, + { + "epoch": 0.09953, + "grad_norm": 1.1666110255591098, + "learning_rate": 0.003, + "loss": 4.118, + "step": 9953 + }, + { + "epoch": 0.09954, + "grad_norm": 0.8404966208187018, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 9954 + }, + { + "epoch": 0.09955, + "grad_norm": 0.6348854161168744, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 9955 + }, + { + "epoch": 0.09956, + "grad_norm": 0.6821965250062815, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 9956 + }, + { + "epoch": 0.09957, + "grad_norm": 0.6331997868974779, + "learning_rate": 0.003, + "loss": 4.086, + "step": 9957 + }, + { + "epoch": 0.09958, + "grad_norm": 0.777484394684235, + "learning_rate": 0.003, + "loss": 4.1028, + "step": 9958 + }, + { + "epoch": 0.09959, + "grad_norm": 1.0101723616966292, + "learning_rate": 0.003, + "loss": 4.1088, + "step": 9959 + }, + { + "epoch": 0.0996, + "grad_norm": 1.2396789474457461, + "learning_rate": 0.003, + "loss": 4.1022, + "step": 9960 + }, + { + "epoch": 0.09961, + "grad_norm": 0.883683117211385, + "learning_rate": 0.003, + "loss": 4.0983, + "step": 9961 + }, + { + "epoch": 0.09962, + "grad_norm": 0.9018658759893716, + "learning_rate": 0.003, + "loss": 4.1046, + "step": 9962 + }, + { + "epoch": 0.09963, + "grad_norm": 1.0876146413308398, + "learning_rate": 0.003, + "loss": 4.1091, + "step": 9963 + }, + { + "epoch": 0.09964, + "grad_norm": 1.0287603306521411, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 9964 + }, + { + "epoch": 0.09965, + "grad_norm": 1.0182673045567394, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 9965 + }, + { + "epoch": 0.09966, + "grad_norm": 1.1527631254244322, + "learning_rate": 0.003, + "loss": 4.1112, + "step": 9966 + }, + { + "epoch": 0.09967, + "grad_norm": 0.8709975266827947, + "learning_rate": 0.003, + "loss": 4.0987, + "step": 9967 + }, + { + "epoch": 0.09968, + "grad_norm": 0.8331778651716775, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 9968 + }, + { + "epoch": 0.09969, + "grad_norm": 0.9011465361487814, + "learning_rate": 0.003, + "loss": 4.0948, + "step": 9969 + }, + { + "epoch": 0.0997, + "grad_norm": 1.0080570307471883, + "learning_rate": 0.003, + "loss": 4.086, + "step": 9970 + }, + { + "epoch": 0.09971, + "grad_norm": 1.1140173135877225, + "learning_rate": 0.003, + "loss": 4.0976, + "step": 9971 + }, + { + "epoch": 0.09972, + "grad_norm": 0.9393638818310338, + "learning_rate": 0.003, + "loss": 4.0819, + "step": 9972 + }, + { + "epoch": 0.09973, + "grad_norm": 1.0649082720147478, + "learning_rate": 0.003, + "loss": 4.0968, + "step": 9973 + }, + { + "epoch": 0.09974, + "grad_norm": 1.130092450772193, + "learning_rate": 0.003, + "loss": 4.0932, + "step": 9974 + }, + { + "epoch": 0.09975, + "grad_norm": 0.88585944847689, + "learning_rate": 0.003, + "loss": 4.099, + "step": 9975 + }, + { + "epoch": 0.09976, + "grad_norm": 0.857801840723658, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 9976 + }, + { + "epoch": 0.09977, + "grad_norm": 0.8296564805689629, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 9977 + }, + { + "epoch": 0.09978, + "grad_norm": 0.8072297066940064, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 9978 + }, + { + "epoch": 0.09979, + "grad_norm": 0.8227232337051928, + "learning_rate": 0.003, + "loss": 4.09, + "step": 9979 + }, + { + "epoch": 0.0998, + "grad_norm": 0.8476832328670773, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 9980 + }, + { + "epoch": 0.09981, + "grad_norm": 0.9357206908798009, + "learning_rate": 0.003, + "loss": 4.0903, + "step": 9981 + }, + { + "epoch": 0.09982, + "grad_norm": 1.143225588531109, + "learning_rate": 0.003, + "loss": 4.099, + "step": 9982 + }, + { + "epoch": 0.09983, + "grad_norm": 0.8237550880806321, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 9983 + }, + { + "epoch": 0.09984, + "grad_norm": 0.7811421447324085, + "learning_rate": 0.003, + "loss": 4.1046, + "step": 9984 + }, + { + "epoch": 0.09985, + "grad_norm": 0.8438712038256704, + "learning_rate": 0.003, + "loss": 4.0998, + "step": 9985 + }, + { + "epoch": 0.09986, + "grad_norm": 0.8254292728857884, + "learning_rate": 0.003, + "loss": 4.1023, + "step": 9986 + }, + { + "epoch": 0.09987, + "grad_norm": 0.8131007944980652, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 9987 + }, + { + "epoch": 0.09988, + "grad_norm": 0.8584836605861299, + "learning_rate": 0.003, + "loss": 4.1098, + "step": 9988 + }, + { + "epoch": 0.09989, + "grad_norm": 1.089542214562673, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 9989 + }, + { + "epoch": 0.0999, + "grad_norm": 1.2008176887832276, + "learning_rate": 0.003, + "loss": 4.0943, + "step": 9990 + }, + { + "epoch": 0.09991, + "grad_norm": 0.9071696842404575, + "learning_rate": 0.003, + "loss": 4.1082, + "step": 9991 + }, + { + "epoch": 0.09992, + "grad_norm": 0.9778468657106637, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 9992 + }, + { + "epoch": 0.09993, + "grad_norm": 1.018266367771181, + "learning_rate": 0.003, + "loss": 4.095, + "step": 9993 + }, + { + "epoch": 0.09994, + "grad_norm": 0.9996060334571353, + "learning_rate": 0.003, + "loss": 4.1086, + "step": 9994 + }, + { + "epoch": 0.09995, + "grad_norm": 0.8639535717682494, + "learning_rate": 0.003, + "loss": 4.1042, + "step": 9995 + }, + { + "epoch": 0.09996, + "grad_norm": 0.752574714119443, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 9996 + }, + { + "epoch": 0.09997, + "grad_norm": 0.7765764977536432, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 9997 + }, + { + "epoch": 0.09998, + "grad_norm": 0.8178576426181308, + "learning_rate": 0.003, + "loss": 4.0958, + "step": 9998 + }, + { + "epoch": 0.09999, + "grad_norm": 1.007703696887012, + "learning_rate": 0.003, + "loss": 4.0974, + "step": 9999 + }, + { + "epoch": 0.1, + "grad_norm": 1.2762926674519628, + "learning_rate": 0.003, + "loss": 4.1068, + "step": 10000 + }, + { + "epoch": 0.10001, + "grad_norm": 0.846407795052644, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 10001 + }, + { + "epoch": 0.10002, + "grad_norm": 0.7905366312062068, + "learning_rate": 0.003, + "loss": 4.1085, + "step": 10002 + }, + { + "epoch": 0.10003, + "grad_norm": 0.7626553182720586, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 10003 + }, + { + "epoch": 0.10004, + "grad_norm": 0.7357040183466467, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 10004 + }, + { + "epoch": 0.10005, + "grad_norm": 0.8552529891304915, + "learning_rate": 0.003, + "loss": 4.087, + "step": 10005 + }, + { + "epoch": 0.10006, + "grad_norm": 0.8393237904177807, + "learning_rate": 0.003, + "loss": 4.0925, + "step": 10006 + }, + { + "epoch": 0.10007, + "grad_norm": 0.7039732095352877, + "learning_rate": 0.003, + "loss": 4.066, + "step": 10007 + }, + { + "epoch": 0.10008, + "grad_norm": 0.7449168028515943, + "learning_rate": 0.003, + "loss": 4.0958, + "step": 10008 + }, + { + "epoch": 0.10009, + "grad_norm": 0.9329505822221974, + "learning_rate": 0.003, + "loss": 4.083, + "step": 10009 + }, + { + "epoch": 0.1001, + "grad_norm": 1.1263928068893259, + "learning_rate": 0.003, + "loss": 4.1063, + "step": 10010 + }, + { + "epoch": 0.10011, + "grad_norm": 1.268614328489737, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 10011 + }, + { + "epoch": 0.10012, + "grad_norm": 0.9122036076799189, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 10012 + }, + { + "epoch": 0.10013, + "grad_norm": 0.8253984253411619, + "learning_rate": 0.003, + "loss": 4.1077, + "step": 10013 + }, + { + "epoch": 0.10014, + "grad_norm": 0.7242857930477015, + "learning_rate": 0.003, + "loss": 4.0965, + "step": 10014 + }, + { + "epoch": 0.10015, + "grad_norm": 0.6973871854077252, + "learning_rate": 0.003, + "loss": 4.1107, + "step": 10015 + }, + { + "epoch": 0.10016, + "grad_norm": 0.8688761600896016, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 10016 + }, + { + "epoch": 0.10017, + "grad_norm": 1.0003375826499101, + "learning_rate": 0.003, + "loss": 4.0957, + "step": 10017 + }, + { + "epoch": 0.10018, + "grad_norm": 1.0687424330551845, + "learning_rate": 0.003, + "loss": 4.1051, + "step": 10018 + }, + { + "epoch": 0.10019, + "grad_norm": 0.9963046067938195, + "learning_rate": 0.003, + "loss": 4.097, + "step": 10019 + }, + { + "epoch": 0.1002, + "grad_norm": 1.1687033860831646, + "learning_rate": 0.003, + "loss": 4.072, + "step": 10020 + }, + { + "epoch": 0.10021, + "grad_norm": 0.8842502711008842, + "learning_rate": 0.003, + "loss": 4.104, + "step": 10021 + }, + { + "epoch": 0.10022, + "grad_norm": 0.9407554761722029, + "learning_rate": 0.003, + "loss": 4.0919, + "step": 10022 + }, + { + "epoch": 0.10023, + "grad_norm": 0.8765452325739109, + "learning_rate": 0.003, + "loss": 4.0959, + "step": 10023 + }, + { + "epoch": 0.10024, + "grad_norm": 0.8525654707148999, + "learning_rate": 0.003, + "loss": 4.1094, + "step": 10024 + }, + { + "epoch": 0.10025, + "grad_norm": 0.9396760999742726, + "learning_rate": 0.003, + "loss": 4.1022, + "step": 10025 + }, + { + "epoch": 0.10026, + "grad_norm": 1.1803601102228167, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 10026 + }, + { + "epoch": 0.10027, + "grad_norm": 0.923241626794587, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 10027 + }, + { + "epoch": 0.10028, + "grad_norm": 0.9963175020977256, + "learning_rate": 0.003, + "loss": 4.1303, + "step": 10028 + }, + { + "epoch": 0.10029, + "grad_norm": 1.0199938399966424, + "learning_rate": 0.003, + "loss": 4.0981, + "step": 10029 + }, + { + "epoch": 0.1003, + "grad_norm": 1.090921827661979, + "learning_rate": 0.003, + "loss": 4.1059, + "step": 10030 + }, + { + "epoch": 0.10031, + "grad_norm": 0.9319954550156979, + "learning_rate": 0.003, + "loss": 4.1302, + "step": 10031 + }, + { + "epoch": 0.10032, + "grad_norm": 1.1047516979284653, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 10032 + }, + { + "epoch": 0.10033, + "grad_norm": 1.1692589400749065, + "learning_rate": 0.003, + "loss": 4.0949, + "step": 10033 + }, + { + "epoch": 0.10034, + "grad_norm": 0.9770333150957505, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 10034 + }, + { + "epoch": 0.10035, + "grad_norm": 0.9954119825979576, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 10035 + }, + { + "epoch": 0.10036, + "grad_norm": 1.044590472870181, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 10036 + }, + { + "epoch": 0.10037, + "grad_norm": 0.8036975893761776, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 10037 + }, + { + "epoch": 0.10038, + "grad_norm": 0.6197785706472615, + "learning_rate": 0.003, + "loss": 4.1055, + "step": 10038 + }, + { + "epoch": 0.10039, + "grad_norm": 0.6559386956265669, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 10039 + }, + { + "epoch": 0.1004, + "grad_norm": 0.8526993389567112, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 10040 + }, + { + "epoch": 0.10041, + "grad_norm": 1.1346763408530436, + "learning_rate": 0.003, + "loss": 4.1013, + "step": 10041 + }, + { + "epoch": 0.10042, + "grad_norm": 0.8675374262995108, + "learning_rate": 0.003, + "loss": 4.1189, + "step": 10042 + }, + { + "epoch": 0.10043, + "grad_norm": 0.8251228533053213, + "learning_rate": 0.003, + "loss": 4.1171, + "step": 10043 + }, + { + "epoch": 0.10044, + "grad_norm": 0.9042772731919037, + "learning_rate": 0.003, + "loss": 4.102, + "step": 10044 + }, + { + "epoch": 0.10045, + "grad_norm": 0.8668844113260905, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 10045 + }, + { + "epoch": 0.10046, + "grad_norm": 0.7490664234115638, + "learning_rate": 0.003, + "loss": 4.0995, + "step": 10046 + }, + { + "epoch": 0.10047, + "grad_norm": 0.7241753646886957, + "learning_rate": 0.003, + "loss": 4.0926, + "step": 10047 + }, + { + "epoch": 0.10048, + "grad_norm": 0.710951375353978, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 10048 + }, + { + "epoch": 0.10049, + "grad_norm": 0.7362753119687154, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 10049 + }, + { + "epoch": 0.1005, + "grad_norm": 0.6549130434053738, + "learning_rate": 0.003, + "loss": 4.0858, + "step": 10050 + }, + { + "epoch": 0.10051, + "grad_norm": 0.6504551821577659, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 10051 + }, + { + "epoch": 0.10052, + "grad_norm": 0.7851754123050869, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 10052 + }, + { + "epoch": 0.10053, + "grad_norm": 0.9100083067673433, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 10053 + }, + { + "epoch": 0.10054, + "grad_norm": 1.0440423818992077, + "learning_rate": 0.003, + "loss": 4.0971, + "step": 10054 + }, + { + "epoch": 0.10055, + "grad_norm": 1.199189154881823, + "learning_rate": 0.003, + "loss": 4.1061, + "step": 10055 + }, + { + "epoch": 0.10056, + "grad_norm": 0.8228820126030579, + "learning_rate": 0.003, + "loss": 4.1089, + "step": 10056 + }, + { + "epoch": 0.10057, + "grad_norm": 0.6751385531616468, + "learning_rate": 0.003, + "loss": 4.0938, + "step": 10057 + }, + { + "epoch": 0.10058, + "grad_norm": 0.6459866651885418, + "learning_rate": 0.003, + "loss": 4.088, + "step": 10058 + }, + { + "epoch": 0.10059, + "grad_norm": 0.7163827085044409, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 10059 + }, + { + "epoch": 0.1006, + "grad_norm": 0.7498357936226233, + "learning_rate": 0.003, + "loss": 4.0871, + "step": 10060 + }, + { + "epoch": 0.10061, + "grad_norm": 0.9199026629500379, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 10061 + }, + { + "epoch": 0.10062, + "grad_norm": 0.9339006252420677, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 10062 + }, + { + "epoch": 0.10063, + "grad_norm": 0.9825515007821343, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 10063 + }, + { + "epoch": 0.10064, + "grad_norm": 0.9083355003285981, + "learning_rate": 0.003, + "loss": 4.0871, + "step": 10064 + }, + { + "epoch": 0.10065, + "grad_norm": 0.9044707268344537, + "learning_rate": 0.003, + "loss": 4.107, + "step": 10065 + }, + { + "epoch": 0.10066, + "grad_norm": 1.1109715579619182, + "learning_rate": 0.003, + "loss": 4.0868, + "step": 10066 + }, + { + "epoch": 0.10067, + "grad_norm": 1.1784060335963866, + "learning_rate": 0.003, + "loss": 4.102, + "step": 10067 + }, + { + "epoch": 0.10068, + "grad_norm": 0.8376423509316573, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 10068 + }, + { + "epoch": 0.10069, + "grad_norm": 0.7242588761496445, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 10069 + }, + { + "epoch": 0.1007, + "grad_norm": 0.8576404093344348, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 10070 + }, + { + "epoch": 0.10071, + "grad_norm": 0.9946862714080159, + "learning_rate": 0.003, + "loss": 4.0857, + "step": 10071 + }, + { + "epoch": 0.10072, + "grad_norm": 1.0828968607567901, + "learning_rate": 0.003, + "loss": 4.0884, + "step": 10072 + }, + { + "epoch": 0.10073, + "grad_norm": 1.1050819072213203, + "learning_rate": 0.003, + "loss": 4.1051, + "step": 10073 + }, + { + "epoch": 0.10074, + "grad_norm": 0.9410259771807634, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 10074 + }, + { + "epoch": 0.10075, + "grad_norm": 1.0981600397772833, + "learning_rate": 0.003, + "loss": 4.1233, + "step": 10075 + }, + { + "epoch": 0.10076, + "grad_norm": 0.9142745269772177, + "learning_rate": 0.003, + "loss": 4.0987, + "step": 10076 + }, + { + "epoch": 0.10077, + "grad_norm": 0.9943875686970955, + "learning_rate": 0.003, + "loss": 4.1187, + "step": 10077 + }, + { + "epoch": 0.10078, + "grad_norm": 1.1167502119378494, + "learning_rate": 0.003, + "loss": 4.076, + "step": 10078 + }, + { + "epoch": 0.10079, + "grad_norm": 1.2695113736982493, + "learning_rate": 0.003, + "loss": 4.1033, + "step": 10079 + }, + { + "epoch": 0.1008, + "grad_norm": 0.9837623919758819, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 10080 + }, + { + "epoch": 0.10081, + "grad_norm": 0.9628725238779852, + "learning_rate": 0.003, + "loss": 4.1185, + "step": 10081 + }, + { + "epoch": 0.10082, + "grad_norm": 0.9550994930520159, + "learning_rate": 0.003, + "loss": 4.0963, + "step": 10082 + }, + { + "epoch": 0.10083, + "grad_norm": 1.1134296419199943, + "learning_rate": 0.003, + "loss": 4.1102, + "step": 10083 + }, + { + "epoch": 0.10084, + "grad_norm": 0.9435992423381032, + "learning_rate": 0.003, + "loss": 4.066, + "step": 10084 + }, + { + "epoch": 0.10085, + "grad_norm": 0.7825874323570984, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 10085 + }, + { + "epoch": 0.10086, + "grad_norm": 0.6921370331084226, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 10086 + }, + { + "epoch": 0.10087, + "grad_norm": 0.7278203894261717, + "learning_rate": 0.003, + "loss": 4.1061, + "step": 10087 + }, + { + "epoch": 0.10088, + "grad_norm": 0.757891744237821, + "learning_rate": 0.003, + "loss": 4.083, + "step": 10088 + }, + { + "epoch": 0.10089, + "grad_norm": 0.8360256616730807, + "learning_rate": 0.003, + "loss": 4.0899, + "step": 10089 + }, + { + "epoch": 0.1009, + "grad_norm": 1.0181286638800873, + "learning_rate": 0.003, + "loss": 4.1204, + "step": 10090 + }, + { + "epoch": 0.10091, + "grad_norm": 1.387952676248201, + "learning_rate": 0.003, + "loss": 4.0926, + "step": 10091 + }, + { + "epoch": 0.10092, + "grad_norm": 0.6362639069092441, + "learning_rate": 0.003, + "loss": 4.1002, + "step": 10092 + }, + { + "epoch": 0.10093, + "grad_norm": 0.743374062070198, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 10093 + }, + { + "epoch": 0.10094, + "grad_norm": 0.9731460449457097, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 10094 + }, + { + "epoch": 0.10095, + "grad_norm": 1.3362207611604642, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 10095 + }, + { + "epoch": 0.10096, + "grad_norm": 0.9238856240977876, + "learning_rate": 0.003, + "loss": 4.0898, + "step": 10096 + }, + { + "epoch": 0.10097, + "grad_norm": 0.9398288617159353, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 10097 + }, + { + "epoch": 0.10098, + "grad_norm": 0.8926736862540934, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 10098 + }, + { + "epoch": 0.10099, + "grad_norm": 0.8547347149934348, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 10099 + }, + { + "epoch": 0.101, + "grad_norm": 0.8322868296554979, + "learning_rate": 0.003, + "loss": 4.0972, + "step": 10100 + }, + { + "epoch": 0.10101, + "grad_norm": 0.8029418806784904, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 10101 + }, + { + "epoch": 0.10102, + "grad_norm": 0.7466775681047372, + "learning_rate": 0.003, + "loss": 4.1016, + "step": 10102 + }, + { + "epoch": 0.10103, + "grad_norm": 0.7762877847621028, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 10103 + }, + { + "epoch": 0.10104, + "grad_norm": 0.8239056171472944, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 10104 + }, + { + "epoch": 0.10105, + "grad_norm": 0.8635790264439557, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 10105 + }, + { + "epoch": 0.10106, + "grad_norm": 0.9705585934825562, + "learning_rate": 0.003, + "loss": 4.0953, + "step": 10106 + }, + { + "epoch": 0.10107, + "grad_norm": 1.1925319618292862, + "learning_rate": 0.003, + "loss": 4.0983, + "step": 10107 + }, + { + "epoch": 0.10108, + "grad_norm": 0.930885951872856, + "learning_rate": 0.003, + "loss": 4.1003, + "step": 10108 + }, + { + "epoch": 0.10109, + "grad_norm": 0.9478659253890813, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 10109 + }, + { + "epoch": 0.1011, + "grad_norm": 0.9458668708439815, + "learning_rate": 0.003, + "loss": 4.1112, + "step": 10110 + }, + { + "epoch": 0.10111, + "grad_norm": 1.0465386282766551, + "learning_rate": 0.003, + "loss": 4.1059, + "step": 10111 + }, + { + "epoch": 0.10112, + "grad_norm": 0.941800413849234, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 10112 + }, + { + "epoch": 0.10113, + "grad_norm": 0.9975021944083188, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 10113 + }, + { + "epoch": 0.10114, + "grad_norm": 1.0568025070552394, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 10114 + }, + { + "epoch": 0.10115, + "grad_norm": 0.9662005716462198, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 10115 + }, + { + "epoch": 0.10116, + "grad_norm": 0.9964430409371794, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 10116 + }, + { + "epoch": 0.10117, + "grad_norm": 1.095577021129824, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 10117 + }, + { + "epoch": 0.10118, + "grad_norm": 0.9058740337483966, + "learning_rate": 0.003, + "loss": 4.1179, + "step": 10118 + }, + { + "epoch": 0.10119, + "grad_norm": 0.7651533816220512, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 10119 + }, + { + "epoch": 0.1012, + "grad_norm": 0.69007560749185, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 10120 + }, + { + "epoch": 0.10121, + "grad_norm": 0.7001010703742647, + "learning_rate": 0.003, + "loss": 4.0918, + "step": 10121 + }, + { + "epoch": 0.10122, + "grad_norm": 0.761442702518915, + "learning_rate": 0.003, + "loss": 4.058, + "step": 10122 + }, + { + "epoch": 0.10123, + "grad_norm": 0.8785099107356117, + "learning_rate": 0.003, + "loss": 4.08, + "step": 10123 + }, + { + "epoch": 0.10124, + "grad_norm": 0.9478766112748606, + "learning_rate": 0.003, + "loss": 4.1102, + "step": 10124 + }, + { + "epoch": 0.10125, + "grad_norm": 1.063224559233055, + "learning_rate": 0.003, + "loss": 4.1041, + "step": 10125 + }, + { + "epoch": 0.10126, + "grad_norm": 0.8729013926215154, + "learning_rate": 0.003, + "loss": 4.078, + "step": 10126 + }, + { + "epoch": 0.10127, + "grad_norm": 0.9790642546904503, + "learning_rate": 0.003, + "loss": 4.1071, + "step": 10127 + }, + { + "epoch": 0.10128, + "grad_norm": 1.194142388457261, + "learning_rate": 0.003, + "loss": 4.1146, + "step": 10128 + }, + { + "epoch": 0.10129, + "grad_norm": 0.8758113085129664, + "learning_rate": 0.003, + "loss": 4.1063, + "step": 10129 + }, + { + "epoch": 0.1013, + "grad_norm": 0.7716520659651475, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 10130 + }, + { + "epoch": 0.10131, + "grad_norm": 0.7456677268457248, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 10131 + }, + { + "epoch": 0.10132, + "grad_norm": 0.7444245827863488, + "learning_rate": 0.003, + "loss": 4.0956, + "step": 10132 + }, + { + "epoch": 0.10133, + "grad_norm": 0.8624467509580721, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 10133 + }, + { + "epoch": 0.10134, + "grad_norm": 1.068676946502593, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 10134 + }, + { + "epoch": 0.10135, + "grad_norm": 0.884273594664609, + "learning_rate": 0.003, + "loss": 4.0906, + "step": 10135 + }, + { + "epoch": 0.10136, + "grad_norm": 0.9370719522559154, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 10136 + }, + { + "epoch": 0.10137, + "grad_norm": 1.1271997457372382, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 10137 + }, + { + "epoch": 0.10138, + "grad_norm": 1.2033883396503056, + "learning_rate": 0.003, + "loss": 4.0958, + "step": 10138 + }, + { + "epoch": 0.10139, + "grad_norm": 1.0032022116088837, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 10139 + }, + { + "epoch": 0.1014, + "grad_norm": 1.0298997230954023, + "learning_rate": 0.003, + "loss": 4.1213, + "step": 10140 + }, + { + "epoch": 0.10141, + "grad_norm": 1.0787132264495785, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 10141 + }, + { + "epoch": 0.10142, + "grad_norm": 0.9439503852040803, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 10142 + }, + { + "epoch": 0.10143, + "grad_norm": 0.793813851397082, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 10143 + }, + { + "epoch": 0.10144, + "grad_norm": 0.9068837449137608, + "learning_rate": 0.003, + "loss": 4.1059, + "step": 10144 + }, + { + "epoch": 0.10145, + "grad_norm": 1.0209227302234767, + "learning_rate": 0.003, + "loss": 4.0979, + "step": 10145 + }, + { + "epoch": 0.10146, + "grad_norm": 0.9666772286362464, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 10146 + }, + { + "epoch": 0.10147, + "grad_norm": 0.9315175908743543, + "learning_rate": 0.003, + "loss": 4.0998, + "step": 10147 + }, + { + "epoch": 0.10148, + "grad_norm": 0.8260305551602328, + "learning_rate": 0.003, + "loss": 4.0981, + "step": 10148 + }, + { + "epoch": 0.10149, + "grad_norm": 0.8482729309499758, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 10149 + }, + { + "epoch": 0.1015, + "grad_norm": 0.9112844755131481, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 10150 + }, + { + "epoch": 0.10151, + "grad_norm": 0.895305709305238, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 10151 + }, + { + "epoch": 0.10152, + "grad_norm": 0.859191763614006, + "learning_rate": 0.003, + "loss": 4.0913, + "step": 10152 + }, + { + "epoch": 0.10153, + "grad_norm": 1.291353413201306, + "learning_rate": 0.003, + "loss": 4.1019, + "step": 10153 + }, + { + "epoch": 0.10154, + "grad_norm": 1.0044755702579264, + "learning_rate": 0.003, + "loss": 4.1053, + "step": 10154 + }, + { + "epoch": 0.10155, + "grad_norm": 1.034686950726131, + "learning_rate": 0.003, + "loss": 4.1003, + "step": 10155 + }, + { + "epoch": 0.10156, + "grad_norm": 1.0424601916900678, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 10156 + }, + { + "epoch": 0.10157, + "grad_norm": 0.9688813053768359, + "learning_rate": 0.003, + "loss": 4.08, + "step": 10157 + }, + { + "epoch": 0.10158, + "grad_norm": 0.9594621123977223, + "learning_rate": 0.003, + "loss": 4.0973, + "step": 10158 + }, + { + "epoch": 0.10159, + "grad_norm": 0.982207587013617, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 10159 + }, + { + "epoch": 0.1016, + "grad_norm": 0.8819333080484042, + "learning_rate": 0.003, + "loss": 4.1011, + "step": 10160 + }, + { + "epoch": 0.10161, + "grad_norm": 0.9293751028457353, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 10161 + }, + { + "epoch": 0.10162, + "grad_norm": 0.8952355474969211, + "learning_rate": 0.003, + "loss": 4.0931, + "step": 10162 + }, + { + "epoch": 0.10163, + "grad_norm": 1.102164324382523, + "learning_rate": 0.003, + "loss": 4.0982, + "step": 10163 + }, + { + "epoch": 0.10164, + "grad_norm": 1.0887008704431567, + "learning_rate": 0.003, + "loss": 4.1141, + "step": 10164 + }, + { + "epoch": 0.10165, + "grad_norm": 0.889951599291344, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 10165 + }, + { + "epoch": 0.10166, + "grad_norm": 0.6952125204021279, + "learning_rate": 0.003, + "loss": 4.0983, + "step": 10166 + }, + { + "epoch": 0.10167, + "grad_norm": 0.6693318934840289, + "learning_rate": 0.003, + "loss": 4.089, + "step": 10167 + }, + { + "epoch": 0.10168, + "grad_norm": 0.7047376173064419, + "learning_rate": 0.003, + "loss": 4.058, + "step": 10168 + }, + { + "epoch": 0.10169, + "grad_norm": 0.7183980469373401, + "learning_rate": 0.003, + "loss": 4.0856, + "step": 10169 + }, + { + "epoch": 0.1017, + "grad_norm": 0.8427166782664123, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 10170 + }, + { + "epoch": 0.10171, + "grad_norm": 1.0410810119415947, + "learning_rate": 0.003, + "loss": 4.071, + "step": 10171 + }, + { + "epoch": 0.10172, + "grad_norm": 1.177683177477319, + "learning_rate": 0.003, + "loss": 4.0978, + "step": 10172 + }, + { + "epoch": 0.10173, + "grad_norm": 0.7571093543560289, + "learning_rate": 0.003, + "loss": 4.0905, + "step": 10173 + }, + { + "epoch": 0.10174, + "grad_norm": 0.6929601496343557, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 10174 + }, + { + "epoch": 0.10175, + "grad_norm": 0.7234447952472367, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 10175 + }, + { + "epoch": 0.10176, + "grad_norm": 0.6916513903179555, + "learning_rate": 0.003, + "loss": 4.081, + "step": 10176 + }, + { + "epoch": 0.10177, + "grad_norm": 0.7736851572569, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 10177 + }, + { + "epoch": 0.10178, + "grad_norm": 0.9485288338897027, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 10178 + }, + { + "epoch": 0.10179, + "grad_norm": 1.2922094110390194, + "learning_rate": 0.003, + "loss": 4.0997, + "step": 10179 + }, + { + "epoch": 0.1018, + "grad_norm": 0.9787810271178282, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 10180 + }, + { + "epoch": 0.10181, + "grad_norm": 1.0716844007885504, + "learning_rate": 0.003, + "loss": 4.0888, + "step": 10181 + }, + { + "epoch": 0.10182, + "grad_norm": 0.9833533126055853, + "learning_rate": 0.003, + "loss": 4.1013, + "step": 10182 + }, + { + "epoch": 0.10183, + "grad_norm": 0.9311939217302757, + "learning_rate": 0.003, + "loss": 4.104, + "step": 10183 + }, + { + "epoch": 0.10184, + "grad_norm": 0.8641055498098966, + "learning_rate": 0.003, + "loss": 4.1074, + "step": 10184 + }, + { + "epoch": 0.10185, + "grad_norm": 0.9267918824165212, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 10185 + }, + { + "epoch": 0.10186, + "grad_norm": 1.0410114798244103, + "learning_rate": 0.003, + "loss": 4.1039, + "step": 10186 + }, + { + "epoch": 0.10187, + "grad_norm": 1.143927178798348, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 10187 + }, + { + "epoch": 0.10188, + "grad_norm": 0.982291560845215, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 10188 + }, + { + "epoch": 0.10189, + "grad_norm": 1.08197977590208, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 10189 + }, + { + "epoch": 0.1019, + "grad_norm": 0.949237752231189, + "learning_rate": 0.003, + "loss": 4.0931, + "step": 10190 + }, + { + "epoch": 0.10191, + "grad_norm": 0.913554083512254, + "learning_rate": 0.003, + "loss": 4.0989, + "step": 10191 + }, + { + "epoch": 0.10192, + "grad_norm": 0.8646445003569243, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 10192 + }, + { + "epoch": 0.10193, + "grad_norm": 0.7784523543751104, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 10193 + }, + { + "epoch": 0.10194, + "grad_norm": 0.6786411486503773, + "learning_rate": 0.003, + "loss": 4.0968, + "step": 10194 + }, + { + "epoch": 0.10195, + "grad_norm": 0.7186344914911345, + "learning_rate": 0.003, + "loss": 4.098, + "step": 10195 + }, + { + "epoch": 0.10196, + "grad_norm": 0.7387234544886367, + "learning_rate": 0.003, + "loss": 4.0931, + "step": 10196 + }, + { + "epoch": 0.10197, + "grad_norm": 0.8102276874615613, + "learning_rate": 0.003, + "loss": 4.0918, + "step": 10197 + }, + { + "epoch": 0.10198, + "grad_norm": 0.8705933283950565, + "learning_rate": 0.003, + "loss": 4.1112, + "step": 10198 + }, + { + "epoch": 0.10199, + "grad_norm": 1.1180920548650946, + "learning_rate": 0.003, + "loss": 4.0918, + "step": 10199 + }, + { + "epoch": 0.102, + "grad_norm": 1.0768424207018248, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 10200 + }, + { + "epoch": 0.10201, + "grad_norm": 0.959117318640648, + "learning_rate": 0.003, + "loss": 4.088, + "step": 10201 + }, + { + "epoch": 0.10202, + "grad_norm": 0.9452210486184167, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 10202 + }, + { + "epoch": 0.10203, + "grad_norm": 0.8744290998180672, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 10203 + }, + { + "epoch": 0.10204, + "grad_norm": 0.8814145368301949, + "learning_rate": 0.003, + "loss": 4.1052, + "step": 10204 + }, + { + "epoch": 0.10205, + "grad_norm": 0.9746557209823046, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 10205 + }, + { + "epoch": 0.10206, + "grad_norm": 1.0418540613710794, + "learning_rate": 0.003, + "loss": 4.1012, + "step": 10206 + }, + { + "epoch": 0.10207, + "grad_norm": 0.9196772923298814, + "learning_rate": 0.003, + "loss": 4.0925, + "step": 10207 + }, + { + "epoch": 0.10208, + "grad_norm": 0.9531610441504913, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 10208 + }, + { + "epoch": 0.10209, + "grad_norm": 0.9572183588800921, + "learning_rate": 0.003, + "loss": 4.1071, + "step": 10209 + }, + { + "epoch": 0.1021, + "grad_norm": 1.018140555117966, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 10210 + }, + { + "epoch": 0.10211, + "grad_norm": 1.1267802119609027, + "learning_rate": 0.003, + "loss": 4.1202, + "step": 10211 + }, + { + "epoch": 0.10212, + "grad_norm": 0.8712823440020605, + "learning_rate": 0.003, + "loss": 4.0925, + "step": 10212 + }, + { + "epoch": 0.10213, + "grad_norm": 0.9072106602979279, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 10213 + }, + { + "epoch": 0.10214, + "grad_norm": 0.9486234408720666, + "learning_rate": 0.003, + "loss": 4.1001, + "step": 10214 + }, + { + "epoch": 0.10215, + "grad_norm": 0.9759771051816137, + "learning_rate": 0.003, + "loss": 4.1146, + "step": 10215 + }, + { + "epoch": 0.10216, + "grad_norm": 1.110106437639336, + "learning_rate": 0.003, + "loss": 4.1099, + "step": 10216 + }, + { + "epoch": 0.10217, + "grad_norm": 1.2216993058746834, + "learning_rate": 0.003, + "loss": 4.0968, + "step": 10217 + }, + { + "epoch": 0.10218, + "grad_norm": 1.2710114431184163, + "learning_rate": 0.003, + "loss": 4.1259, + "step": 10218 + }, + { + "epoch": 0.10219, + "grad_norm": 0.8302554254819106, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 10219 + }, + { + "epoch": 0.1022, + "grad_norm": 0.7863333021144477, + "learning_rate": 0.003, + "loss": 4.0969, + "step": 10220 + }, + { + "epoch": 0.10221, + "grad_norm": 0.8329096323996931, + "learning_rate": 0.003, + "loss": 4.111, + "step": 10221 + }, + { + "epoch": 0.10222, + "grad_norm": 0.8950092316956775, + "learning_rate": 0.003, + "loss": 4.0993, + "step": 10222 + }, + { + "epoch": 0.10223, + "grad_norm": 1.0101933742905977, + "learning_rate": 0.003, + "loss": 4.1071, + "step": 10223 + }, + { + "epoch": 0.10224, + "grad_norm": 1.214233787040993, + "learning_rate": 0.003, + "loss": 4.1075, + "step": 10224 + }, + { + "epoch": 0.10225, + "grad_norm": 0.8237743424691844, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 10225 + }, + { + "epoch": 0.10226, + "grad_norm": 0.8053271054595286, + "learning_rate": 0.003, + "loss": 4.1408, + "step": 10226 + }, + { + "epoch": 0.10227, + "grad_norm": 0.9020570009142984, + "learning_rate": 0.003, + "loss": 4.0994, + "step": 10227 + }, + { + "epoch": 0.10228, + "grad_norm": 1.0004416710517987, + "learning_rate": 0.003, + "loss": 4.1068, + "step": 10228 + }, + { + "epoch": 0.10229, + "grad_norm": 1.0319751729551627, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 10229 + }, + { + "epoch": 0.1023, + "grad_norm": 1.03776059312104, + "learning_rate": 0.003, + "loss": 4.0967, + "step": 10230 + }, + { + "epoch": 0.10231, + "grad_norm": 1.0092313022554864, + "learning_rate": 0.003, + "loss": 4.0995, + "step": 10231 + }, + { + "epoch": 0.10232, + "grad_norm": 0.8963093858201144, + "learning_rate": 0.003, + "loss": 4.0979, + "step": 10232 + }, + { + "epoch": 0.10233, + "grad_norm": 0.7652529049894464, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 10233 + }, + { + "epoch": 0.10234, + "grad_norm": 0.6732069744826151, + "learning_rate": 0.003, + "loss": 4.1028, + "step": 10234 + }, + { + "epoch": 0.10235, + "grad_norm": 0.6850274953707035, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 10235 + }, + { + "epoch": 0.10236, + "grad_norm": 0.7461795196732801, + "learning_rate": 0.003, + "loss": 4.1197, + "step": 10236 + }, + { + "epoch": 0.10237, + "grad_norm": 0.7771855708889348, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 10237 + }, + { + "epoch": 0.10238, + "grad_norm": 0.8645414091849927, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 10238 + }, + { + "epoch": 0.10239, + "grad_norm": 0.9598711609970192, + "learning_rate": 0.003, + "loss": 4.1237, + "step": 10239 + }, + { + "epoch": 0.1024, + "grad_norm": 1.1783930818865855, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 10240 + }, + { + "epoch": 0.10241, + "grad_norm": 0.9512353642002936, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 10241 + }, + { + "epoch": 0.10242, + "grad_norm": 1.1076182241363002, + "learning_rate": 0.003, + "loss": 4.086, + "step": 10242 + }, + { + "epoch": 0.10243, + "grad_norm": 0.8140071943334923, + "learning_rate": 0.003, + "loss": 4.0944, + "step": 10243 + }, + { + "epoch": 0.10244, + "grad_norm": 0.7868081031641362, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 10244 + }, + { + "epoch": 0.10245, + "grad_norm": 0.8277890846490785, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 10245 + }, + { + "epoch": 0.10246, + "grad_norm": 0.8578316539852658, + "learning_rate": 0.003, + "loss": 4.1138, + "step": 10246 + }, + { + "epoch": 0.10247, + "grad_norm": 0.8324285743959013, + "learning_rate": 0.003, + "loss": 4.1172, + "step": 10247 + }, + { + "epoch": 0.10248, + "grad_norm": 0.8750930826936256, + "learning_rate": 0.003, + "loss": 4.1076, + "step": 10248 + }, + { + "epoch": 0.10249, + "grad_norm": 1.060036727312048, + "learning_rate": 0.003, + "loss": 4.097, + "step": 10249 + }, + { + "epoch": 0.1025, + "grad_norm": 1.061281725833956, + "learning_rate": 0.003, + "loss": 4.1081, + "step": 10250 + }, + { + "epoch": 0.10251, + "grad_norm": 0.9071148674826301, + "learning_rate": 0.003, + "loss": 4.0908, + "step": 10251 + }, + { + "epoch": 0.10252, + "grad_norm": 1.1016410599173725, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 10252 + }, + { + "epoch": 0.10253, + "grad_norm": 1.0552690649769536, + "learning_rate": 0.003, + "loss": 4.0948, + "step": 10253 + }, + { + "epoch": 0.10254, + "grad_norm": 0.9150936122477507, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 10254 + }, + { + "epoch": 0.10255, + "grad_norm": 0.8251742844607864, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 10255 + }, + { + "epoch": 0.10256, + "grad_norm": 0.769901480267319, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 10256 + }, + { + "epoch": 0.10257, + "grad_norm": 0.7276745815286125, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 10257 + }, + { + "epoch": 0.10258, + "grad_norm": 0.7332149497476529, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 10258 + }, + { + "epoch": 0.10259, + "grad_norm": 0.7518788065657007, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 10259 + }, + { + "epoch": 0.1026, + "grad_norm": 0.7023632334353492, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 10260 + }, + { + "epoch": 0.10261, + "grad_norm": 0.7804783704951214, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 10261 + }, + { + "epoch": 0.10262, + "grad_norm": 0.8399725680772053, + "learning_rate": 0.003, + "loss": 4.0874, + "step": 10262 + }, + { + "epoch": 0.10263, + "grad_norm": 0.9808468708799891, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 10263 + }, + { + "epoch": 0.10264, + "grad_norm": 1.1134356366421605, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 10264 + }, + { + "epoch": 0.10265, + "grad_norm": 0.9727628676105254, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 10265 + }, + { + "epoch": 0.10266, + "grad_norm": 1.1457311148339535, + "learning_rate": 0.003, + "loss": 4.0919, + "step": 10266 + }, + { + "epoch": 0.10267, + "grad_norm": 1.0369946720443313, + "learning_rate": 0.003, + "loss": 4.1149, + "step": 10267 + }, + { + "epoch": 0.10268, + "grad_norm": 0.9235251394184828, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 10268 + }, + { + "epoch": 0.10269, + "grad_norm": 0.8040182456689232, + "learning_rate": 0.003, + "loss": 4.084, + "step": 10269 + }, + { + "epoch": 0.1027, + "grad_norm": 0.782968238530311, + "learning_rate": 0.003, + "loss": 4.0874, + "step": 10270 + }, + { + "epoch": 0.10271, + "grad_norm": 0.9407811113989156, + "learning_rate": 0.003, + "loss": 4.1023, + "step": 10271 + }, + { + "epoch": 0.10272, + "grad_norm": 1.3167169512817753, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 10272 + }, + { + "epoch": 0.10273, + "grad_norm": 0.901902815771997, + "learning_rate": 0.003, + "loss": 4.1054, + "step": 10273 + }, + { + "epoch": 0.10274, + "grad_norm": 1.031370760218058, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 10274 + }, + { + "epoch": 0.10275, + "grad_norm": 1.1133697077886784, + "learning_rate": 0.003, + "loss": 4.112, + "step": 10275 + }, + { + "epoch": 0.10276, + "grad_norm": 0.9337682201872862, + "learning_rate": 0.003, + "loss": 4.0975, + "step": 10276 + }, + { + "epoch": 0.10277, + "grad_norm": 0.8577556801384054, + "learning_rate": 0.003, + "loss": 4.082, + "step": 10277 + }, + { + "epoch": 0.10278, + "grad_norm": 0.8329130008061313, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 10278 + }, + { + "epoch": 0.10279, + "grad_norm": 0.8008000768493235, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 10279 + }, + { + "epoch": 0.1028, + "grad_norm": 0.9586013002825486, + "learning_rate": 0.003, + "loss": 4.0977, + "step": 10280 + }, + { + "epoch": 0.10281, + "grad_norm": 0.9362380399816307, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 10281 + }, + { + "epoch": 0.10282, + "grad_norm": 0.8034496491265808, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 10282 + }, + { + "epoch": 0.10283, + "grad_norm": 0.8175026392274294, + "learning_rate": 0.003, + "loss": 4.0886, + "step": 10283 + }, + { + "epoch": 0.10284, + "grad_norm": 0.9220091553084664, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 10284 + }, + { + "epoch": 0.10285, + "grad_norm": 1.1745060376041516, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 10285 + }, + { + "epoch": 0.10286, + "grad_norm": 1.0970725530778482, + "learning_rate": 0.003, + "loss": 4.0912, + "step": 10286 + }, + { + "epoch": 0.10287, + "grad_norm": 0.8877853599799197, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 10287 + }, + { + "epoch": 0.10288, + "grad_norm": 0.6729305691229556, + "learning_rate": 0.003, + "loss": 4.1185, + "step": 10288 + }, + { + "epoch": 0.10289, + "grad_norm": 0.6101513929016343, + "learning_rate": 0.003, + "loss": 4.1033, + "step": 10289 + }, + { + "epoch": 0.1029, + "grad_norm": 0.6006184830680509, + "learning_rate": 0.003, + "loss": 4.0989, + "step": 10290 + }, + { + "epoch": 0.10291, + "grad_norm": 0.6580410217249946, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 10291 + }, + { + "epoch": 0.10292, + "grad_norm": 0.7948565223500166, + "learning_rate": 0.003, + "loss": 4.1057, + "step": 10292 + }, + { + "epoch": 0.10293, + "grad_norm": 0.873562469179876, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 10293 + }, + { + "epoch": 0.10294, + "grad_norm": 0.9164282281892033, + "learning_rate": 0.003, + "loss": 4.1132, + "step": 10294 + }, + { + "epoch": 0.10295, + "grad_norm": 0.9618027035487724, + "learning_rate": 0.003, + "loss": 4.0972, + "step": 10295 + }, + { + "epoch": 0.10296, + "grad_norm": 1.270173812912686, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 10296 + }, + { + "epoch": 0.10297, + "grad_norm": 0.9443619634500702, + "learning_rate": 0.003, + "loss": 4.1005, + "step": 10297 + }, + { + "epoch": 0.10298, + "grad_norm": 0.9717393974386974, + "learning_rate": 0.003, + "loss": 4.1239, + "step": 10298 + }, + { + "epoch": 0.10299, + "grad_norm": 0.9770324878575616, + "learning_rate": 0.003, + "loss": 4.0881, + "step": 10299 + }, + { + "epoch": 0.103, + "grad_norm": 1.1101050604150264, + "learning_rate": 0.003, + "loss": 4.0985, + "step": 10300 + }, + { + "epoch": 0.10301, + "grad_norm": 0.9011446491453848, + "learning_rate": 0.003, + "loss": 4.11, + "step": 10301 + }, + { + "epoch": 0.10302, + "grad_norm": 0.9141554564065553, + "learning_rate": 0.003, + "loss": 4.117, + "step": 10302 + }, + { + "epoch": 0.10303, + "grad_norm": 1.004795254729966, + "learning_rate": 0.003, + "loss": 4.0864, + "step": 10303 + }, + { + "epoch": 0.10304, + "grad_norm": 1.1104199011303157, + "learning_rate": 0.003, + "loss": 4.1072, + "step": 10304 + }, + { + "epoch": 0.10305, + "grad_norm": 1.0779701470417002, + "learning_rate": 0.003, + "loss": 4.0927, + "step": 10305 + }, + { + "epoch": 0.10306, + "grad_norm": 0.9929838596058995, + "learning_rate": 0.003, + "loss": 4.1205, + "step": 10306 + }, + { + "epoch": 0.10307, + "grad_norm": 1.0378001284680718, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 10307 + }, + { + "epoch": 0.10308, + "grad_norm": 0.8451606285227234, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 10308 + }, + { + "epoch": 0.10309, + "grad_norm": 0.7280653251404869, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 10309 + }, + { + "epoch": 0.1031, + "grad_norm": 0.696794451745512, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 10310 + }, + { + "epoch": 0.10311, + "grad_norm": 0.7400519052708209, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 10311 + }, + { + "epoch": 0.10312, + "grad_norm": 0.8022018489520792, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 10312 + }, + { + "epoch": 0.10313, + "grad_norm": 0.9665547403935929, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 10313 + }, + { + "epoch": 0.10314, + "grad_norm": 1.5536139006797307, + "learning_rate": 0.003, + "loss": 4.1142, + "step": 10314 + }, + { + "epoch": 0.10315, + "grad_norm": 0.8228479217445895, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 10315 + }, + { + "epoch": 0.10316, + "grad_norm": 0.8344809726813904, + "learning_rate": 0.003, + "loss": 4.1249, + "step": 10316 + }, + { + "epoch": 0.10317, + "grad_norm": 0.9317615186683637, + "learning_rate": 0.003, + "loss": 4.086, + "step": 10317 + }, + { + "epoch": 0.10318, + "grad_norm": 1.166290443357865, + "learning_rate": 0.003, + "loss": 4.087, + "step": 10318 + }, + { + "epoch": 0.10319, + "grad_norm": 1.0667534185590877, + "learning_rate": 0.003, + "loss": 4.1009, + "step": 10319 + }, + { + "epoch": 0.1032, + "grad_norm": 0.9679233979201338, + "learning_rate": 0.003, + "loss": 4.1053, + "step": 10320 + }, + { + "epoch": 0.10321, + "grad_norm": 1.2156018152822397, + "learning_rate": 0.003, + "loss": 4.1098, + "step": 10321 + }, + { + "epoch": 0.10322, + "grad_norm": 0.8694306030100604, + "learning_rate": 0.003, + "loss": 4.1211, + "step": 10322 + }, + { + "epoch": 0.10323, + "grad_norm": 0.77390810530893, + "learning_rate": 0.003, + "loss": 4.105, + "step": 10323 + }, + { + "epoch": 0.10324, + "grad_norm": 0.8459945001619534, + "learning_rate": 0.003, + "loss": 4.1031, + "step": 10324 + }, + { + "epoch": 0.10325, + "grad_norm": 0.9598819363584302, + "learning_rate": 0.003, + "loss": 4.0898, + "step": 10325 + }, + { + "epoch": 0.10326, + "grad_norm": 1.270741833731687, + "learning_rate": 0.003, + "loss": 4.1145, + "step": 10326 + }, + { + "epoch": 0.10327, + "grad_norm": 1.0250394175893829, + "learning_rate": 0.003, + "loss": 4.0943, + "step": 10327 + }, + { + "epoch": 0.10328, + "grad_norm": 1.0646571407533487, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 10328 + }, + { + "epoch": 0.10329, + "grad_norm": 0.8357225132410019, + "learning_rate": 0.003, + "loss": 4.1105, + "step": 10329 + }, + { + "epoch": 0.1033, + "grad_norm": 0.7911799851858787, + "learning_rate": 0.003, + "loss": 4.1004, + "step": 10330 + }, + { + "epoch": 0.10331, + "grad_norm": 0.8219366496745777, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 10331 + }, + { + "epoch": 0.10332, + "grad_norm": 0.8597720644451482, + "learning_rate": 0.003, + "loss": 4.0819, + "step": 10332 + }, + { + "epoch": 0.10333, + "grad_norm": 0.890201673423854, + "learning_rate": 0.003, + "loss": 4.0881, + "step": 10333 + }, + { + "epoch": 0.10334, + "grad_norm": 0.8822162452290838, + "learning_rate": 0.003, + "loss": 4.1203, + "step": 10334 + }, + { + "epoch": 0.10335, + "grad_norm": 1.0434026016363345, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 10335 + }, + { + "epoch": 0.10336, + "grad_norm": 1.0508762574327049, + "learning_rate": 0.003, + "loss": 4.1138, + "step": 10336 + }, + { + "epoch": 0.10337, + "grad_norm": 1.023518325100827, + "learning_rate": 0.003, + "loss": 4.0977, + "step": 10337 + }, + { + "epoch": 0.10338, + "grad_norm": 1.0222939920443586, + "learning_rate": 0.003, + "loss": 4.0856, + "step": 10338 + }, + { + "epoch": 0.10339, + "grad_norm": 1.0200772185596512, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 10339 + }, + { + "epoch": 0.1034, + "grad_norm": 1.1118315208165592, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 10340 + }, + { + "epoch": 0.10341, + "grad_norm": 1.229788890194771, + "learning_rate": 0.003, + "loss": 4.0871, + "step": 10341 + }, + { + "epoch": 0.10342, + "grad_norm": 0.881320621902087, + "learning_rate": 0.003, + "loss": 4.064, + "step": 10342 + }, + { + "epoch": 0.10343, + "grad_norm": 0.8012287495926537, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 10343 + }, + { + "epoch": 0.10344, + "grad_norm": 0.7057244410169525, + "learning_rate": 0.003, + "loss": 4.0938, + "step": 10344 + }, + { + "epoch": 0.10345, + "grad_norm": 0.6831154260246173, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 10345 + }, + { + "epoch": 0.10346, + "grad_norm": 0.7390159629576106, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 10346 + }, + { + "epoch": 0.10347, + "grad_norm": 0.8881991932306049, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 10347 + }, + { + "epoch": 0.10348, + "grad_norm": 1.0870558090042928, + "learning_rate": 0.003, + "loss": 4.1146, + "step": 10348 + }, + { + "epoch": 0.10349, + "grad_norm": 0.9674704344374843, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 10349 + }, + { + "epoch": 0.1035, + "grad_norm": 1.3103851543543703, + "learning_rate": 0.003, + "loss": 4.0953, + "step": 10350 + }, + { + "epoch": 0.10351, + "grad_norm": 0.7752164378221075, + "learning_rate": 0.003, + "loss": 4.0956, + "step": 10351 + }, + { + "epoch": 0.10352, + "grad_norm": 0.6333669676096207, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 10352 + }, + { + "epoch": 0.10353, + "grad_norm": 0.6218713823135109, + "learning_rate": 0.003, + "loss": 4.1173, + "step": 10353 + }, + { + "epoch": 0.10354, + "grad_norm": 0.6760083791815406, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 10354 + }, + { + "epoch": 0.10355, + "grad_norm": 0.9646643200322085, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 10355 + }, + { + "epoch": 0.10356, + "grad_norm": 1.4592855584956232, + "learning_rate": 0.003, + "loss": 4.1022, + "step": 10356 + }, + { + "epoch": 0.10357, + "grad_norm": 0.5555068019093737, + "learning_rate": 0.003, + "loss": 4.062, + "step": 10357 + }, + { + "epoch": 0.10358, + "grad_norm": 0.7512404459799188, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 10358 + }, + { + "epoch": 0.10359, + "grad_norm": 1.0729200127973797, + "learning_rate": 0.003, + "loss": 4.0856, + "step": 10359 + }, + { + "epoch": 0.1036, + "grad_norm": 1.03771510025134, + "learning_rate": 0.003, + "loss": 4.0922, + "step": 10360 + }, + { + "epoch": 0.10361, + "grad_norm": 0.9377908762803766, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 10361 + }, + { + "epoch": 0.10362, + "grad_norm": 0.9079898615078676, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 10362 + }, + { + "epoch": 0.10363, + "grad_norm": 0.8542438216920811, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 10363 + }, + { + "epoch": 0.10364, + "grad_norm": 0.819741082498773, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 10364 + }, + { + "epoch": 0.10365, + "grad_norm": 0.9387758253093952, + "learning_rate": 0.003, + "loss": 4.1098, + "step": 10365 + }, + { + "epoch": 0.10366, + "grad_norm": 0.8595582015203754, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 10366 + }, + { + "epoch": 0.10367, + "grad_norm": 0.9195985652318767, + "learning_rate": 0.003, + "loss": 4.1058, + "step": 10367 + }, + { + "epoch": 0.10368, + "grad_norm": 1.041561976873367, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 10368 + }, + { + "epoch": 0.10369, + "grad_norm": 0.9769289654804704, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 10369 + }, + { + "epoch": 0.1037, + "grad_norm": 0.8689024508593368, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 10370 + }, + { + "epoch": 0.10371, + "grad_norm": 0.9127703576736774, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 10371 + }, + { + "epoch": 0.10372, + "grad_norm": 1.0840217364480413, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 10372 + }, + { + "epoch": 0.10373, + "grad_norm": 0.8589979313681667, + "learning_rate": 0.003, + "loss": 4.0908, + "step": 10373 + }, + { + "epoch": 0.10374, + "grad_norm": 0.9209999042655965, + "learning_rate": 0.003, + "loss": 4.081, + "step": 10374 + }, + { + "epoch": 0.10375, + "grad_norm": 0.8201839455652276, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 10375 + }, + { + "epoch": 0.10376, + "grad_norm": 0.9253860615344267, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 10376 + }, + { + "epoch": 0.10377, + "grad_norm": 0.9786803772735758, + "learning_rate": 0.003, + "loss": 4.0971, + "step": 10377 + }, + { + "epoch": 0.10378, + "grad_norm": 1.1604859081188932, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 10378 + }, + { + "epoch": 0.10379, + "grad_norm": 1.2662343585458882, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 10379 + }, + { + "epoch": 0.1038, + "grad_norm": 0.8440684963312072, + "learning_rate": 0.003, + "loss": 4.1155, + "step": 10380 + }, + { + "epoch": 0.10381, + "grad_norm": 0.7206245307283129, + "learning_rate": 0.003, + "loss": 4.0898, + "step": 10381 + }, + { + "epoch": 0.10382, + "grad_norm": 0.6830006926833568, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 10382 + }, + { + "epoch": 0.10383, + "grad_norm": 0.7129059802346774, + "learning_rate": 0.003, + "loss": 4.0926, + "step": 10383 + }, + { + "epoch": 0.10384, + "grad_norm": 0.7952669183573826, + "learning_rate": 0.003, + "loss": 4.0978, + "step": 10384 + }, + { + "epoch": 0.10385, + "grad_norm": 0.9501299747016804, + "learning_rate": 0.003, + "loss": 4.087, + "step": 10385 + }, + { + "epoch": 0.10386, + "grad_norm": 1.1086117644137294, + "learning_rate": 0.003, + "loss": 4.0994, + "step": 10386 + }, + { + "epoch": 0.10387, + "grad_norm": 0.9932383976281319, + "learning_rate": 0.003, + "loss": 4.0927, + "step": 10387 + }, + { + "epoch": 0.10388, + "grad_norm": 0.9586921010142302, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 10388 + }, + { + "epoch": 0.10389, + "grad_norm": 0.9711866848407964, + "learning_rate": 0.003, + "loss": 4.0999, + "step": 10389 + }, + { + "epoch": 0.1039, + "grad_norm": 1.0147092924912415, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 10390 + }, + { + "epoch": 0.10391, + "grad_norm": 1.0506648630838256, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 10391 + }, + { + "epoch": 0.10392, + "grad_norm": 1.0256891304025917, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 10392 + }, + { + "epoch": 0.10393, + "grad_norm": 0.9068803727043103, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 10393 + }, + { + "epoch": 0.10394, + "grad_norm": 0.7973461045268242, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 10394 + }, + { + "epoch": 0.10395, + "grad_norm": 0.9194292373243754, + "learning_rate": 0.003, + "loss": 4.1117, + "step": 10395 + }, + { + "epoch": 0.10396, + "grad_norm": 1.0102356587327825, + "learning_rate": 0.003, + "loss": 4.0918, + "step": 10396 + }, + { + "epoch": 0.10397, + "grad_norm": 0.9871504508361195, + "learning_rate": 0.003, + "loss": 4.1097, + "step": 10397 + }, + { + "epoch": 0.10398, + "grad_norm": 1.1580641926834954, + "learning_rate": 0.003, + "loss": 4.1054, + "step": 10398 + }, + { + "epoch": 0.10399, + "grad_norm": 1.1865606079170086, + "learning_rate": 0.003, + "loss": 4.1108, + "step": 10399 + }, + { + "epoch": 0.104, + "grad_norm": 0.7309099903593418, + "learning_rate": 0.003, + "loss": 4.0907, + "step": 10400 + }, + { + "epoch": 0.10401, + "grad_norm": 0.797549535904677, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 10401 + }, + { + "epoch": 0.10402, + "grad_norm": 0.8334605859661031, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 10402 + }, + { + "epoch": 0.10403, + "grad_norm": 1.1129497288268397, + "learning_rate": 0.003, + "loss": 4.0916, + "step": 10403 + }, + { + "epoch": 0.10404, + "grad_norm": 0.8866013454742103, + "learning_rate": 0.003, + "loss": 4.0893, + "step": 10404 + }, + { + "epoch": 0.10405, + "grad_norm": 0.8113657361504197, + "learning_rate": 0.003, + "loss": 4.0962, + "step": 10405 + }, + { + "epoch": 0.10406, + "grad_norm": 0.9630667687327077, + "learning_rate": 0.003, + "loss": 4.0881, + "step": 10406 + }, + { + "epoch": 0.10407, + "grad_norm": 1.0018852512268608, + "learning_rate": 0.003, + "loss": 4.1131, + "step": 10407 + }, + { + "epoch": 0.10408, + "grad_norm": 1.210761873171873, + "learning_rate": 0.003, + "loss": 4.0891, + "step": 10408 + }, + { + "epoch": 0.10409, + "grad_norm": 0.911527687025421, + "learning_rate": 0.003, + "loss": 4.0978, + "step": 10409 + }, + { + "epoch": 0.1041, + "grad_norm": 1.1093559437661817, + "learning_rate": 0.003, + "loss": 4.1279, + "step": 10410 + }, + { + "epoch": 0.10411, + "grad_norm": 1.0983001926114488, + "learning_rate": 0.003, + "loss": 4.0925, + "step": 10411 + }, + { + "epoch": 0.10412, + "grad_norm": 1.0418812052286917, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 10412 + }, + { + "epoch": 0.10413, + "grad_norm": 1.034562901350855, + "learning_rate": 0.003, + "loss": 4.1108, + "step": 10413 + }, + { + "epoch": 0.10414, + "grad_norm": 1.070942141831756, + "learning_rate": 0.003, + "loss": 4.0959, + "step": 10414 + }, + { + "epoch": 0.10415, + "grad_norm": 1.0000094759743325, + "learning_rate": 0.003, + "loss": 4.1014, + "step": 10415 + }, + { + "epoch": 0.10416, + "grad_norm": 1.2072050655719908, + "learning_rate": 0.003, + "loss": 4.1171, + "step": 10416 + }, + { + "epoch": 0.10417, + "grad_norm": 0.9528389999280306, + "learning_rate": 0.003, + "loss": 4.1217, + "step": 10417 + }, + { + "epoch": 0.10418, + "grad_norm": 0.9028057108858157, + "learning_rate": 0.003, + "loss": 4.073, + "step": 10418 + }, + { + "epoch": 0.10419, + "grad_norm": 1.0157180076503933, + "learning_rate": 0.003, + "loss": 4.1412, + "step": 10419 + }, + { + "epoch": 0.1042, + "grad_norm": 1.1403458363137668, + "learning_rate": 0.003, + "loss": 4.1199, + "step": 10420 + }, + { + "epoch": 0.10421, + "grad_norm": 0.8704646720456675, + "learning_rate": 0.003, + "loss": 4.0868, + "step": 10421 + }, + { + "epoch": 0.10422, + "grad_norm": 0.9395189062284242, + "learning_rate": 0.003, + "loss": 4.1015, + "step": 10422 + }, + { + "epoch": 0.10423, + "grad_norm": 0.8738072199709813, + "learning_rate": 0.003, + "loss": 4.1102, + "step": 10423 + }, + { + "epoch": 0.10424, + "grad_norm": 0.8289481896392907, + "learning_rate": 0.003, + "loss": 4.088, + "step": 10424 + }, + { + "epoch": 0.10425, + "grad_norm": 0.8642726662440525, + "learning_rate": 0.003, + "loss": 4.0922, + "step": 10425 + }, + { + "epoch": 0.10426, + "grad_norm": 0.7664947327264443, + "learning_rate": 0.003, + "loss": 4.0872, + "step": 10426 + }, + { + "epoch": 0.10427, + "grad_norm": 0.7452710697371259, + "learning_rate": 0.003, + "loss": 4.1004, + "step": 10427 + }, + { + "epoch": 0.10428, + "grad_norm": 0.8401015273709832, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 10428 + }, + { + "epoch": 0.10429, + "grad_norm": 1.274648878168851, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 10429 + }, + { + "epoch": 0.1043, + "grad_norm": 1.0367563780328006, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 10430 + }, + { + "epoch": 0.10431, + "grad_norm": 0.8734572144042883, + "learning_rate": 0.003, + "loss": 4.0891, + "step": 10431 + }, + { + "epoch": 0.10432, + "grad_norm": 0.7657409637891003, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 10432 + }, + { + "epoch": 0.10433, + "grad_norm": 0.7628644277180485, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 10433 + }, + { + "epoch": 0.10434, + "grad_norm": 0.7793533094592869, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 10434 + }, + { + "epoch": 0.10435, + "grad_norm": 0.8094318483713879, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 10435 + }, + { + "epoch": 0.10436, + "grad_norm": 0.7429219594582074, + "learning_rate": 0.003, + "loss": 4.0868, + "step": 10436 + }, + { + "epoch": 0.10437, + "grad_norm": 0.870889314049426, + "learning_rate": 0.003, + "loss": 4.0927, + "step": 10437 + }, + { + "epoch": 0.10438, + "grad_norm": 1.0949239572834308, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 10438 + }, + { + "epoch": 0.10439, + "grad_norm": 0.9283027635978875, + "learning_rate": 0.003, + "loss": 4.0961, + "step": 10439 + }, + { + "epoch": 0.1044, + "grad_norm": 0.7349818655914907, + "learning_rate": 0.003, + "loss": 4.0899, + "step": 10440 + }, + { + "epoch": 0.10441, + "grad_norm": 0.6879154908672748, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 10441 + }, + { + "epoch": 0.10442, + "grad_norm": 0.7006052403529107, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 10442 + }, + { + "epoch": 0.10443, + "grad_norm": 0.6857927068182457, + "learning_rate": 0.003, + "loss": 4.1009, + "step": 10443 + }, + { + "epoch": 0.10444, + "grad_norm": 0.7066994752516409, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 10444 + }, + { + "epoch": 0.10445, + "grad_norm": 0.74121727875003, + "learning_rate": 0.003, + "loss": 4.1009, + "step": 10445 + }, + { + "epoch": 0.10446, + "grad_norm": 0.710351575703391, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 10446 + }, + { + "epoch": 0.10447, + "grad_norm": 0.7727332338705983, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 10447 + }, + { + "epoch": 0.10448, + "grad_norm": 0.882691468705699, + "learning_rate": 0.003, + "loss": 4.0856, + "step": 10448 + }, + { + "epoch": 0.10449, + "grad_norm": 0.8149626492002201, + "learning_rate": 0.003, + "loss": 4.078, + "step": 10449 + }, + { + "epoch": 0.1045, + "grad_norm": 0.9691586098285396, + "learning_rate": 0.003, + "loss": 4.1076, + "step": 10450 + }, + { + "epoch": 0.10451, + "grad_norm": 0.9247313474112963, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 10451 + }, + { + "epoch": 0.10452, + "grad_norm": 0.8506424868739227, + "learning_rate": 0.003, + "loss": 4.084, + "step": 10452 + }, + { + "epoch": 0.10453, + "grad_norm": 0.752106512429951, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 10453 + }, + { + "epoch": 0.10454, + "grad_norm": 0.7107856595454773, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 10454 + }, + { + "epoch": 0.10455, + "grad_norm": 0.7982776348533294, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 10455 + }, + { + "epoch": 0.10456, + "grad_norm": 1.0871946318716093, + "learning_rate": 0.003, + "loss": 4.1016, + "step": 10456 + }, + { + "epoch": 0.10457, + "grad_norm": 1.5826783176326193, + "learning_rate": 0.003, + "loss": 4.1048, + "step": 10457 + }, + { + "epoch": 0.10458, + "grad_norm": 0.8613053967994393, + "learning_rate": 0.003, + "loss": 4.1055, + "step": 10458 + }, + { + "epoch": 0.10459, + "grad_norm": 0.8721872695668961, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 10459 + }, + { + "epoch": 0.1046, + "grad_norm": 1.0061046478692643, + "learning_rate": 0.003, + "loss": 4.1091, + "step": 10460 + }, + { + "epoch": 0.10461, + "grad_norm": 1.1381946172024242, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 10461 + }, + { + "epoch": 0.10462, + "grad_norm": 0.7896362690927083, + "learning_rate": 0.003, + "loss": 4.0988, + "step": 10462 + }, + { + "epoch": 0.10463, + "grad_norm": 0.7815393759341526, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 10463 + }, + { + "epoch": 0.10464, + "grad_norm": 0.8833649367417272, + "learning_rate": 0.003, + "loss": 4.1015, + "step": 10464 + }, + { + "epoch": 0.10465, + "grad_norm": 1.1155972200711728, + "learning_rate": 0.003, + "loss": 4.0974, + "step": 10465 + }, + { + "epoch": 0.10466, + "grad_norm": 1.319418185782679, + "learning_rate": 0.003, + "loss": 4.0957, + "step": 10466 + }, + { + "epoch": 0.10467, + "grad_norm": 0.7881547485849394, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 10467 + }, + { + "epoch": 0.10468, + "grad_norm": 0.7531093485609334, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 10468 + }, + { + "epoch": 0.10469, + "grad_norm": 0.892332422756942, + "learning_rate": 0.003, + "loss": 4.1008, + "step": 10469 + }, + { + "epoch": 0.1047, + "grad_norm": 1.0166321105509875, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 10470 + }, + { + "epoch": 0.10471, + "grad_norm": 1.1889958685212445, + "learning_rate": 0.003, + "loss": 4.1065, + "step": 10471 + }, + { + "epoch": 0.10472, + "grad_norm": 1.0835868860376183, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 10472 + }, + { + "epoch": 0.10473, + "grad_norm": 1.1079367137366658, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 10473 + }, + { + "epoch": 0.10474, + "grad_norm": 1.1561399937082553, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 10474 + }, + { + "epoch": 0.10475, + "grad_norm": 1.0207723549623584, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 10475 + }, + { + "epoch": 0.10476, + "grad_norm": 0.9672839209760528, + "learning_rate": 0.003, + "loss": 4.092, + "step": 10476 + }, + { + "epoch": 0.10477, + "grad_norm": 0.9696363368652143, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 10477 + }, + { + "epoch": 0.10478, + "grad_norm": 0.8984628711179327, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 10478 + }, + { + "epoch": 0.10479, + "grad_norm": 0.9510122995658213, + "learning_rate": 0.003, + "loss": 4.09, + "step": 10479 + }, + { + "epoch": 0.1048, + "grad_norm": 1.1218902993180204, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 10480 + }, + { + "epoch": 0.10481, + "grad_norm": 0.9719132562926983, + "learning_rate": 0.003, + "loss": 4.0925, + "step": 10481 + }, + { + "epoch": 0.10482, + "grad_norm": 1.018037071948851, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 10482 + }, + { + "epoch": 0.10483, + "grad_norm": 1.066461444248902, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 10483 + }, + { + "epoch": 0.10484, + "grad_norm": 0.9886521379123554, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 10484 + }, + { + "epoch": 0.10485, + "grad_norm": 1.0740120057836673, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 10485 + }, + { + "epoch": 0.10486, + "grad_norm": 0.8817530264679182, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 10486 + }, + { + "epoch": 0.10487, + "grad_norm": 0.8910553662001134, + "learning_rate": 0.003, + "loss": 4.0982, + "step": 10487 + }, + { + "epoch": 0.10488, + "grad_norm": 0.8746810891914258, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 10488 + }, + { + "epoch": 0.10489, + "grad_norm": 0.8472573888806005, + "learning_rate": 0.003, + "loss": 4.0887, + "step": 10489 + }, + { + "epoch": 0.1049, + "grad_norm": 0.9910382414508804, + "learning_rate": 0.003, + "loss": 4.1011, + "step": 10490 + }, + { + "epoch": 0.10491, + "grad_norm": 1.01162082786497, + "learning_rate": 0.003, + "loss": 4.1054, + "step": 10491 + }, + { + "epoch": 0.10492, + "grad_norm": 0.844444720460471, + "learning_rate": 0.003, + "loss": 4.0864, + "step": 10492 + }, + { + "epoch": 0.10493, + "grad_norm": 0.6943610312470553, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 10493 + }, + { + "epoch": 0.10494, + "grad_norm": 0.6713154976944795, + "learning_rate": 0.003, + "loss": 4.049, + "step": 10494 + }, + { + "epoch": 0.10495, + "grad_norm": 0.7376294393469298, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 10495 + }, + { + "epoch": 0.10496, + "grad_norm": 0.8791441808868978, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 10496 + }, + { + "epoch": 0.10497, + "grad_norm": 0.9786244066463133, + "learning_rate": 0.003, + "loss": 4.07, + "step": 10497 + }, + { + "epoch": 0.10498, + "grad_norm": 1.0833488313073092, + "learning_rate": 0.003, + "loss": 4.1015, + "step": 10498 + }, + { + "epoch": 0.10499, + "grad_norm": 0.9847750423685478, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 10499 + }, + { + "epoch": 0.105, + "grad_norm": 0.9917332775495494, + "learning_rate": 0.003, + "loss": 4.1146, + "step": 10500 + }, + { + "epoch": 0.10501, + "grad_norm": 0.9857660160310968, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 10501 + }, + { + "epoch": 0.10502, + "grad_norm": 0.825403471380381, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 10502 + }, + { + "epoch": 0.10503, + "grad_norm": 0.8592557862170064, + "learning_rate": 0.003, + "loss": 4.1102, + "step": 10503 + }, + { + "epoch": 0.10504, + "grad_norm": 0.9743495827294194, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 10504 + }, + { + "epoch": 0.10505, + "grad_norm": 1.1866082918714995, + "learning_rate": 0.003, + "loss": 4.0887, + "step": 10505 + }, + { + "epoch": 0.10506, + "grad_norm": 0.8759455955619688, + "learning_rate": 0.003, + "loss": 4.0856, + "step": 10506 + }, + { + "epoch": 0.10507, + "grad_norm": 0.8029348261579611, + "learning_rate": 0.003, + "loss": 4.0878, + "step": 10507 + }, + { + "epoch": 0.10508, + "grad_norm": 0.8415015903570401, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 10508 + }, + { + "epoch": 0.10509, + "grad_norm": 0.8943308444316465, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 10509 + }, + { + "epoch": 0.1051, + "grad_norm": 0.948277994316701, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 10510 + }, + { + "epoch": 0.10511, + "grad_norm": 0.7991898777205432, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 10511 + }, + { + "epoch": 0.10512, + "grad_norm": 0.7265463914666755, + "learning_rate": 0.003, + "loss": 4.1095, + "step": 10512 + }, + { + "epoch": 0.10513, + "grad_norm": 0.7758307363200123, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 10513 + }, + { + "epoch": 0.10514, + "grad_norm": 0.9439220763930075, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 10514 + }, + { + "epoch": 0.10515, + "grad_norm": 1.368592122987331, + "learning_rate": 0.003, + "loss": 4.099, + "step": 10515 + }, + { + "epoch": 0.10516, + "grad_norm": 0.929310410876089, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 10516 + }, + { + "epoch": 0.10517, + "grad_norm": 0.8507144269808623, + "learning_rate": 0.003, + "loss": 4.1057, + "step": 10517 + }, + { + "epoch": 0.10518, + "grad_norm": 0.7766267375374853, + "learning_rate": 0.003, + "loss": 4.1056, + "step": 10518 + }, + { + "epoch": 0.10519, + "grad_norm": 0.9450489625893496, + "learning_rate": 0.003, + "loss": 4.06, + "step": 10519 + }, + { + "epoch": 0.1052, + "grad_norm": 1.1693435664719927, + "learning_rate": 0.003, + "loss": 4.1212, + "step": 10520 + }, + { + "epoch": 0.10521, + "grad_norm": 0.9344965222540701, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 10521 + }, + { + "epoch": 0.10522, + "grad_norm": 1.2377106404880893, + "learning_rate": 0.003, + "loss": 4.1057, + "step": 10522 + }, + { + "epoch": 0.10523, + "grad_norm": 0.8969852323766027, + "learning_rate": 0.003, + "loss": 4.1146, + "step": 10523 + }, + { + "epoch": 0.10524, + "grad_norm": 0.8281210125569962, + "learning_rate": 0.003, + "loss": 4.083, + "step": 10524 + }, + { + "epoch": 0.10525, + "grad_norm": 0.8971983432365932, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 10525 + }, + { + "epoch": 0.10526, + "grad_norm": 0.9111432088539093, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 10526 + }, + { + "epoch": 0.10527, + "grad_norm": 1.1654145861755343, + "learning_rate": 0.003, + "loss": 4.1187, + "step": 10527 + }, + { + "epoch": 0.10528, + "grad_norm": 1.0977279974437038, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 10528 + }, + { + "epoch": 0.10529, + "grad_norm": 0.923656390544108, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 10529 + }, + { + "epoch": 0.1053, + "grad_norm": 0.8855052042850045, + "learning_rate": 0.003, + "loss": 4.1046, + "step": 10530 + }, + { + "epoch": 0.10531, + "grad_norm": 0.907331433504738, + "learning_rate": 0.003, + "loss": 4.0857, + "step": 10531 + }, + { + "epoch": 0.10532, + "grad_norm": 1.1377959225821521, + "learning_rate": 0.003, + "loss": 4.107, + "step": 10532 + }, + { + "epoch": 0.10533, + "grad_norm": 1.1603318303156538, + "learning_rate": 0.003, + "loss": 4.1, + "step": 10533 + }, + { + "epoch": 0.10534, + "grad_norm": 0.986545402049569, + "learning_rate": 0.003, + "loss": 4.0992, + "step": 10534 + }, + { + "epoch": 0.10535, + "grad_norm": 0.9172417114476418, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 10535 + }, + { + "epoch": 0.10536, + "grad_norm": 0.9785960676924129, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 10536 + }, + { + "epoch": 0.10537, + "grad_norm": 1.221406063334458, + "learning_rate": 0.003, + "loss": 4.0977, + "step": 10537 + }, + { + "epoch": 0.10538, + "grad_norm": 0.8732318820906584, + "learning_rate": 0.003, + "loss": 4.0893, + "step": 10538 + }, + { + "epoch": 0.10539, + "grad_norm": 0.8111878439467994, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 10539 + }, + { + "epoch": 0.1054, + "grad_norm": 0.8747854757736134, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 10540 + }, + { + "epoch": 0.10541, + "grad_norm": 0.9673056524818361, + "learning_rate": 0.003, + "loss": 4.102, + "step": 10541 + }, + { + "epoch": 0.10542, + "grad_norm": 0.9091450018919921, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 10542 + }, + { + "epoch": 0.10543, + "grad_norm": 1.0658978674091468, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 10543 + }, + { + "epoch": 0.10544, + "grad_norm": 1.081196870599217, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 10544 + }, + { + "epoch": 0.10545, + "grad_norm": 0.9106989455457283, + "learning_rate": 0.003, + "loss": 4.1, + "step": 10545 + }, + { + "epoch": 0.10546, + "grad_norm": 0.8743031834677025, + "learning_rate": 0.003, + "loss": 4.1207, + "step": 10546 + }, + { + "epoch": 0.10547, + "grad_norm": 0.8497673998032744, + "learning_rate": 0.003, + "loss": 4.071, + "step": 10547 + }, + { + "epoch": 0.10548, + "grad_norm": 0.8643096991187423, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 10548 + }, + { + "epoch": 0.10549, + "grad_norm": 0.9070663739716069, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 10549 + }, + { + "epoch": 0.1055, + "grad_norm": 0.8156517137786884, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 10550 + }, + { + "epoch": 0.10551, + "grad_norm": 0.9549077537014024, + "learning_rate": 0.003, + "loss": 4.115, + "step": 10551 + }, + { + "epoch": 0.10552, + "grad_norm": 1.0364075053581565, + "learning_rate": 0.003, + "loss": 4.1154, + "step": 10552 + }, + { + "epoch": 0.10553, + "grad_norm": 1.091128726381891, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 10553 + }, + { + "epoch": 0.10554, + "grad_norm": 1.0378085314923546, + "learning_rate": 0.003, + "loss": 4.0932, + "step": 10554 + }, + { + "epoch": 0.10555, + "grad_norm": 1.0860650510822893, + "learning_rate": 0.003, + "loss": 4.1117, + "step": 10555 + }, + { + "epoch": 0.10556, + "grad_norm": 1.1651333543342988, + "learning_rate": 0.003, + "loss": 4.0976, + "step": 10556 + }, + { + "epoch": 0.10557, + "grad_norm": 0.9359356380921641, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 10557 + }, + { + "epoch": 0.10558, + "grad_norm": 0.8390473243103149, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 10558 + }, + { + "epoch": 0.10559, + "grad_norm": 0.8447835094213495, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 10559 + }, + { + "epoch": 0.1056, + "grad_norm": 0.9165593249294828, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 10560 + }, + { + "epoch": 0.10561, + "grad_norm": 1.0853886006987412, + "learning_rate": 0.003, + "loss": 4.0884, + "step": 10561 + }, + { + "epoch": 0.10562, + "grad_norm": 1.037805527104938, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 10562 + }, + { + "epoch": 0.10563, + "grad_norm": 0.9121352061455461, + "learning_rate": 0.003, + "loss": 4.1105, + "step": 10563 + }, + { + "epoch": 0.10564, + "grad_norm": 0.8806780757678176, + "learning_rate": 0.003, + "loss": 4.1095, + "step": 10564 + }, + { + "epoch": 0.10565, + "grad_norm": 0.958349154198751, + "learning_rate": 0.003, + "loss": 4.1003, + "step": 10565 + }, + { + "epoch": 0.10566, + "grad_norm": 1.0510689808949, + "learning_rate": 0.003, + "loss": 4.1083, + "step": 10566 + }, + { + "epoch": 0.10567, + "grad_norm": 1.101632590921785, + "learning_rate": 0.003, + "loss": 4.1299, + "step": 10567 + }, + { + "epoch": 0.10568, + "grad_norm": 0.9789829502427924, + "learning_rate": 0.003, + "loss": 4.1206, + "step": 10568 + }, + { + "epoch": 0.10569, + "grad_norm": 0.9628762446757926, + "learning_rate": 0.003, + "loss": 4.1176, + "step": 10569 + }, + { + "epoch": 0.1057, + "grad_norm": 1.0228136413844169, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 10570 + }, + { + "epoch": 0.10571, + "grad_norm": 0.9518432095930123, + "learning_rate": 0.003, + "loss": 4.0928, + "step": 10571 + }, + { + "epoch": 0.10572, + "grad_norm": 1.0371414780888817, + "learning_rate": 0.003, + "loss": 4.0939, + "step": 10572 + }, + { + "epoch": 0.10573, + "grad_norm": 1.1296551076433214, + "learning_rate": 0.003, + "loss": 4.087, + "step": 10573 + }, + { + "epoch": 0.10574, + "grad_norm": 0.851805966045636, + "learning_rate": 0.003, + "loss": 4.0968, + "step": 10574 + }, + { + "epoch": 0.10575, + "grad_norm": 0.7186828679005042, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 10575 + }, + { + "epoch": 0.10576, + "grad_norm": 0.6725215978480624, + "learning_rate": 0.003, + "loss": 4.1073, + "step": 10576 + }, + { + "epoch": 0.10577, + "grad_norm": 0.6753300523265235, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 10577 + }, + { + "epoch": 0.10578, + "grad_norm": 0.8036408306179271, + "learning_rate": 0.003, + "loss": 4.09, + "step": 10578 + }, + { + "epoch": 0.10579, + "grad_norm": 0.9002537432412724, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 10579 + }, + { + "epoch": 0.1058, + "grad_norm": 1.107764913263939, + "learning_rate": 0.003, + "loss": 4.1054, + "step": 10580 + }, + { + "epoch": 0.10581, + "grad_norm": 0.905526267014512, + "learning_rate": 0.003, + "loss": 4.1061, + "step": 10581 + }, + { + "epoch": 0.10582, + "grad_norm": 0.6907087088967377, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 10582 + }, + { + "epoch": 0.10583, + "grad_norm": 0.6296792245924143, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 10583 + }, + { + "epoch": 0.10584, + "grad_norm": 0.6539090026606424, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 10584 + }, + { + "epoch": 0.10585, + "grad_norm": 0.8111806964029049, + "learning_rate": 0.003, + "loss": 4.1213, + "step": 10585 + }, + { + "epoch": 0.10586, + "grad_norm": 1.0127027071377503, + "learning_rate": 0.003, + "loss": 4.105, + "step": 10586 + }, + { + "epoch": 0.10587, + "grad_norm": 1.1494298129264517, + "learning_rate": 0.003, + "loss": 4.1177, + "step": 10587 + }, + { + "epoch": 0.10588, + "grad_norm": 0.6780396988344141, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 10588 + }, + { + "epoch": 0.10589, + "grad_norm": 0.6150373271956999, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 10589 + }, + { + "epoch": 0.1059, + "grad_norm": 0.7807762029544607, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 10590 + }, + { + "epoch": 0.10591, + "grad_norm": 0.8480453842543368, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 10591 + }, + { + "epoch": 0.10592, + "grad_norm": 0.801628622740933, + "learning_rate": 0.003, + "loss": 4.1191, + "step": 10592 + }, + { + "epoch": 0.10593, + "grad_norm": 0.8190123204782375, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 10593 + }, + { + "epoch": 0.10594, + "grad_norm": 0.9698499863301865, + "learning_rate": 0.003, + "loss": 4.0902, + "step": 10594 + }, + { + "epoch": 0.10595, + "grad_norm": 1.129460586371907, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 10595 + }, + { + "epoch": 0.10596, + "grad_norm": 0.8161991218164444, + "learning_rate": 0.003, + "loss": 4.0919, + "step": 10596 + }, + { + "epoch": 0.10597, + "grad_norm": 0.803040540643654, + "learning_rate": 0.003, + "loss": 4.0857, + "step": 10597 + }, + { + "epoch": 0.10598, + "grad_norm": 0.8685108097563087, + "learning_rate": 0.003, + "loss": 4.0974, + "step": 10598 + }, + { + "epoch": 0.10599, + "grad_norm": 0.9320196209790328, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 10599 + }, + { + "epoch": 0.106, + "grad_norm": 1.313912186590697, + "learning_rate": 0.003, + "loss": 4.1009, + "step": 10600 + }, + { + "epoch": 0.10601, + "grad_norm": 0.9083618378877668, + "learning_rate": 0.003, + "loss": 4.0912, + "step": 10601 + }, + { + "epoch": 0.10602, + "grad_norm": 1.0545005203417737, + "learning_rate": 0.003, + "loss": 4.11, + "step": 10602 + }, + { + "epoch": 0.10603, + "grad_norm": 1.2577228567991576, + "learning_rate": 0.003, + "loss": 4.1057, + "step": 10603 + }, + { + "epoch": 0.10604, + "grad_norm": 0.9455043000594463, + "learning_rate": 0.003, + "loss": 4.1228, + "step": 10604 + }, + { + "epoch": 0.10605, + "grad_norm": 1.0849554300014912, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 10605 + }, + { + "epoch": 0.10606, + "grad_norm": 0.9371235768760167, + "learning_rate": 0.003, + "loss": 4.101, + "step": 10606 + }, + { + "epoch": 0.10607, + "grad_norm": 0.8203058885320677, + "learning_rate": 0.003, + "loss": 4.1176, + "step": 10607 + }, + { + "epoch": 0.10608, + "grad_norm": 0.803807023226867, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 10608 + }, + { + "epoch": 0.10609, + "grad_norm": 0.9345210059357764, + "learning_rate": 0.003, + "loss": 4.1016, + "step": 10609 + }, + { + "epoch": 0.1061, + "grad_norm": 0.9525951020979606, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 10610 + }, + { + "epoch": 0.10611, + "grad_norm": 1.0762099007845962, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 10611 + }, + { + "epoch": 0.10612, + "grad_norm": 0.9977954722169001, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 10612 + }, + { + "epoch": 0.10613, + "grad_norm": 0.9348875752302945, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 10613 + }, + { + "epoch": 0.10614, + "grad_norm": 0.7722532541999543, + "learning_rate": 0.003, + "loss": 4.0964, + "step": 10614 + }, + { + "epoch": 0.10615, + "grad_norm": 0.6552921627337641, + "learning_rate": 0.003, + "loss": 4.0917, + "step": 10615 + }, + { + "epoch": 0.10616, + "grad_norm": 0.7795081468191901, + "learning_rate": 0.003, + "loss": 4.1012, + "step": 10616 + }, + { + "epoch": 0.10617, + "grad_norm": 0.843493082241814, + "learning_rate": 0.003, + "loss": 4.1016, + "step": 10617 + }, + { + "epoch": 0.10618, + "grad_norm": 0.8949546997474781, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 10618 + }, + { + "epoch": 0.10619, + "grad_norm": 0.9473619238401316, + "learning_rate": 0.003, + "loss": 4.0926, + "step": 10619 + }, + { + "epoch": 0.1062, + "grad_norm": 1.0102310083744424, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 10620 + }, + { + "epoch": 0.10621, + "grad_norm": 1.0649077238647118, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 10621 + }, + { + "epoch": 0.10622, + "grad_norm": 1.0940424434616534, + "learning_rate": 0.003, + "loss": 4.1042, + "step": 10622 + }, + { + "epoch": 0.10623, + "grad_norm": 1.0159399127209308, + "learning_rate": 0.003, + "loss": 4.0986, + "step": 10623 + }, + { + "epoch": 0.10624, + "grad_norm": 0.8445082290324857, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 10624 + }, + { + "epoch": 0.10625, + "grad_norm": 0.8360768508939936, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 10625 + }, + { + "epoch": 0.10626, + "grad_norm": 0.8258445616955935, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 10626 + }, + { + "epoch": 0.10627, + "grad_norm": 0.8918278548623568, + "learning_rate": 0.003, + "loss": 4.0952, + "step": 10627 + }, + { + "epoch": 0.10628, + "grad_norm": 0.9465486056396464, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 10628 + }, + { + "epoch": 0.10629, + "grad_norm": 0.8596848611215834, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 10629 + }, + { + "epoch": 0.1063, + "grad_norm": 0.7958583078216956, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 10630 + }, + { + "epoch": 0.10631, + "grad_norm": 0.9437734345205957, + "learning_rate": 0.003, + "loss": 4.1025, + "step": 10631 + }, + { + "epoch": 0.10632, + "grad_norm": 1.125672533113425, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 10632 + }, + { + "epoch": 0.10633, + "grad_norm": 0.8478688703135243, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 10633 + }, + { + "epoch": 0.10634, + "grad_norm": 1.0250490827210883, + "learning_rate": 0.003, + "loss": 4.0878, + "step": 10634 + }, + { + "epoch": 0.10635, + "grad_norm": 1.212274257186467, + "learning_rate": 0.003, + "loss": 4.0873, + "step": 10635 + }, + { + "epoch": 0.10636, + "grad_norm": 0.8932117719274438, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 10636 + }, + { + "epoch": 0.10637, + "grad_norm": 0.8274509545758832, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 10637 + }, + { + "epoch": 0.10638, + "grad_norm": 0.9353624467982324, + "learning_rate": 0.003, + "loss": 4.0857, + "step": 10638 + }, + { + "epoch": 0.10639, + "grad_norm": 1.0261415340037192, + "learning_rate": 0.003, + "loss": 4.0919, + "step": 10639 + }, + { + "epoch": 0.1064, + "grad_norm": 0.9880751014413623, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 10640 + }, + { + "epoch": 0.10641, + "grad_norm": 0.9861882486870125, + "learning_rate": 0.003, + "loss": 4.09, + "step": 10641 + }, + { + "epoch": 0.10642, + "grad_norm": 1.0262170878970345, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 10642 + }, + { + "epoch": 0.10643, + "grad_norm": 0.8857991976630065, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 10643 + }, + { + "epoch": 0.10644, + "grad_norm": 1.0153694677006924, + "learning_rate": 0.003, + "loss": 4.11, + "step": 10644 + }, + { + "epoch": 0.10645, + "grad_norm": 1.1435198267560063, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 10645 + }, + { + "epoch": 0.10646, + "grad_norm": 0.9691749656288253, + "learning_rate": 0.003, + "loss": 4.0913, + "step": 10646 + }, + { + "epoch": 0.10647, + "grad_norm": 1.0119870743220447, + "learning_rate": 0.003, + "loss": 4.0936, + "step": 10647 + }, + { + "epoch": 0.10648, + "grad_norm": 1.031678764259754, + "learning_rate": 0.003, + "loss": 4.0964, + "step": 10648 + }, + { + "epoch": 0.10649, + "grad_norm": 1.156451311123743, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 10649 + }, + { + "epoch": 0.1065, + "grad_norm": 1.0147595439631112, + "learning_rate": 0.003, + "loss": 4.1247, + "step": 10650 + }, + { + "epoch": 0.10651, + "grad_norm": 0.9797495605414368, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 10651 + }, + { + "epoch": 0.10652, + "grad_norm": 1.1030833213226756, + "learning_rate": 0.003, + "loss": 4.0982, + "step": 10652 + }, + { + "epoch": 0.10653, + "grad_norm": 0.975468257299299, + "learning_rate": 0.003, + "loss": 4.0974, + "step": 10653 + }, + { + "epoch": 0.10654, + "grad_norm": 0.8430841627519919, + "learning_rate": 0.003, + "loss": 4.0878, + "step": 10654 + }, + { + "epoch": 0.10655, + "grad_norm": 0.7961010517128089, + "learning_rate": 0.003, + "loss": 4.0944, + "step": 10655 + }, + { + "epoch": 0.10656, + "grad_norm": 0.7729391376707063, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 10656 + }, + { + "epoch": 0.10657, + "grad_norm": 0.7466933021045143, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 10657 + }, + { + "epoch": 0.10658, + "grad_norm": 0.7965663729907493, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 10658 + }, + { + "epoch": 0.10659, + "grad_norm": 0.8486766185635851, + "learning_rate": 0.003, + "loss": 4.0938, + "step": 10659 + }, + { + "epoch": 0.1066, + "grad_norm": 1.0521528583270086, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 10660 + }, + { + "epoch": 0.10661, + "grad_norm": 0.9798629763990518, + "learning_rate": 0.003, + "loss": 4.0992, + "step": 10661 + }, + { + "epoch": 0.10662, + "grad_norm": 1.1064668520172867, + "learning_rate": 0.003, + "loss": 4.0997, + "step": 10662 + }, + { + "epoch": 0.10663, + "grad_norm": 0.9207179845696153, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 10663 + }, + { + "epoch": 0.10664, + "grad_norm": 1.1553619763086682, + "learning_rate": 0.003, + "loss": 4.083, + "step": 10664 + }, + { + "epoch": 0.10665, + "grad_norm": 1.1419994851384185, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 10665 + }, + { + "epoch": 0.10666, + "grad_norm": 0.9130965550906798, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 10666 + }, + { + "epoch": 0.10667, + "grad_norm": 0.8977043380629977, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 10667 + }, + { + "epoch": 0.10668, + "grad_norm": 0.9264443009413197, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 10668 + }, + { + "epoch": 0.10669, + "grad_norm": 0.8266204379731151, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 10669 + }, + { + "epoch": 0.1067, + "grad_norm": 0.7213639746396845, + "learning_rate": 0.003, + "loss": 4.069, + "step": 10670 + }, + { + "epoch": 0.10671, + "grad_norm": 0.7134447271930482, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 10671 + }, + { + "epoch": 0.10672, + "grad_norm": 0.6867444632508992, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 10672 + }, + { + "epoch": 0.10673, + "grad_norm": 0.7014911305889922, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 10673 + }, + { + "epoch": 0.10674, + "grad_norm": 0.7677036935722186, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 10674 + }, + { + "epoch": 0.10675, + "grad_norm": 0.8352256127391506, + "learning_rate": 0.003, + "loss": 4.1042, + "step": 10675 + }, + { + "epoch": 0.10676, + "grad_norm": 1.0395719524062639, + "learning_rate": 0.003, + "loss": 4.0999, + "step": 10676 + }, + { + "epoch": 0.10677, + "grad_norm": 1.4854226122012888, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 10677 + }, + { + "epoch": 0.10678, + "grad_norm": 0.8142405936191625, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 10678 + }, + { + "epoch": 0.10679, + "grad_norm": 0.7729647044289278, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 10679 + }, + { + "epoch": 0.1068, + "grad_norm": 0.7941577783509005, + "learning_rate": 0.003, + "loss": 4.1063, + "step": 10680 + }, + { + "epoch": 0.10681, + "grad_norm": 0.9550299945103695, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 10681 + }, + { + "epoch": 0.10682, + "grad_norm": 0.998536377419294, + "learning_rate": 0.003, + "loss": 4.1077, + "step": 10682 + }, + { + "epoch": 0.10683, + "grad_norm": 0.9522402792343388, + "learning_rate": 0.003, + "loss": 4.0929, + "step": 10683 + }, + { + "epoch": 0.10684, + "grad_norm": 1.0122740938258505, + "learning_rate": 0.003, + "loss": 4.0958, + "step": 10684 + }, + { + "epoch": 0.10685, + "grad_norm": 1.0888355735910868, + "learning_rate": 0.003, + "loss": 4.1078, + "step": 10685 + }, + { + "epoch": 0.10686, + "grad_norm": 1.0652264099184143, + "learning_rate": 0.003, + "loss": 4.0978, + "step": 10686 + }, + { + "epoch": 0.10687, + "grad_norm": 1.0995102570172877, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 10687 + }, + { + "epoch": 0.10688, + "grad_norm": 1.0299122386891209, + "learning_rate": 0.003, + "loss": 4.0947, + "step": 10688 + }, + { + "epoch": 0.10689, + "grad_norm": 1.0030091209564866, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 10689 + }, + { + "epoch": 0.1069, + "grad_norm": 0.990356574268804, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 10690 + }, + { + "epoch": 0.10691, + "grad_norm": 0.9398546824637826, + "learning_rate": 0.003, + "loss": 4.1249, + "step": 10691 + }, + { + "epoch": 0.10692, + "grad_norm": 0.9638160893265136, + "learning_rate": 0.003, + "loss": 4.0996, + "step": 10692 + }, + { + "epoch": 0.10693, + "grad_norm": 1.0655252548413157, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 10693 + }, + { + "epoch": 0.10694, + "grad_norm": 0.8583429002229724, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 10694 + }, + { + "epoch": 0.10695, + "grad_norm": 0.8671418991546258, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 10695 + }, + { + "epoch": 0.10696, + "grad_norm": 0.9751457826865401, + "learning_rate": 0.003, + "loss": 4.1036, + "step": 10696 + }, + { + "epoch": 0.10697, + "grad_norm": 1.129034039626081, + "learning_rate": 0.003, + "loss": 4.107, + "step": 10697 + }, + { + "epoch": 0.10698, + "grad_norm": 0.9231382290438525, + "learning_rate": 0.003, + "loss": 4.1096, + "step": 10698 + }, + { + "epoch": 0.10699, + "grad_norm": 0.7612738566135506, + "learning_rate": 0.003, + "loss": 4.1216, + "step": 10699 + }, + { + "epoch": 0.107, + "grad_norm": 0.7093808064208661, + "learning_rate": 0.003, + "loss": 4.1105, + "step": 10700 + }, + { + "epoch": 0.10701, + "grad_norm": 0.657871914805737, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 10701 + }, + { + "epoch": 0.10702, + "grad_norm": 0.7909818692392661, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 10702 + }, + { + "epoch": 0.10703, + "grad_norm": 0.9064990133899663, + "learning_rate": 0.003, + "loss": 4.0982, + "step": 10703 + }, + { + "epoch": 0.10704, + "grad_norm": 1.059987161455331, + "learning_rate": 0.003, + "loss": 4.0898, + "step": 10704 + }, + { + "epoch": 0.10705, + "grad_norm": 1.1073294155078057, + "learning_rate": 0.003, + "loss": 4.0906, + "step": 10705 + }, + { + "epoch": 0.10706, + "grad_norm": 1.1084454955003236, + "learning_rate": 0.003, + "loss": 4.1005, + "step": 10706 + }, + { + "epoch": 0.10707, + "grad_norm": 0.8052651141329381, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 10707 + }, + { + "epoch": 0.10708, + "grad_norm": 0.8122106956015966, + "learning_rate": 0.003, + "loss": 4.0955, + "step": 10708 + }, + { + "epoch": 0.10709, + "grad_norm": 0.7885911422265968, + "learning_rate": 0.003, + "loss": 4.0973, + "step": 10709 + }, + { + "epoch": 0.1071, + "grad_norm": 0.9347597363617023, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 10710 + }, + { + "epoch": 0.10711, + "grad_norm": 1.0326106708252634, + "learning_rate": 0.003, + "loss": 4.0932, + "step": 10711 + }, + { + "epoch": 0.10712, + "grad_norm": 0.9943220248924097, + "learning_rate": 0.003, + "loss": 4.1104, + "step": 10712 + }, + { + "epoch": 0.10713, + "grad_norm": 0.8758343118565433, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 10713 + }, + { + "epoch": 0.10714, + "grad_norm": 0.8996641882279112, + "learning_rate": 0.003, + "loss": 4.1019, + "step": 10714 + }, + { + "epoch": 0.10715, + "grad_norm": 0.8862354744851694, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 10715 + }, + { + "epoch": 0.10716, + "grad_norm": 0.9002849550098625, + "learning_rate": 0.003, + "loss": 4.0996, + "step": 10716 + }, + { + "epoch": 0.10717, + "grad_norm": 0.8701313999337191, + "learning_rate": 0.003, + "loss": 4.092, + "step": 10717 + }, + { + "epoch": 0.10718, + "grad_norm": 0.9094943954854358, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 10718 + }, + { + "epoch": 0.10719, + "grad_norm": 0.9248436405747751, + "learning_rate": 0.003, + "loss": 4.1078, + "step": 10719 + }, + { + "epoch": 0.1072, + "grad_norm": 0.9160629184943875, + "learning_rate": 0.003, + "loss": 4.1251, + "step": 10720 + }, + { + "epoch": 0.10721, + "grad_norm": 1.1302162367363386, + "learning_rate": 0.003, + "loss": 4.1031, + "step": 10721 + }, + { + "epoch": 0.10722, + "grad_norm": 1.0688684416461092, + "learning_rate": 0.003, + "loss": 4.079, + "step": 10722 + }, + { + "epoch": 0.10723, + "grad_norm": 1.0633959377310542, + "learning_rate": 0.003, + "loss": 4.0941, + "step": 10723 + }, + { + "epoch": 0.10724, + "grad_norm": 0.9561958465742785, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 10724 + }, + { + "epoch": 0.10725, + "grad_norm": 0.9384265593904522, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 10725 + }, + { + "epoch": 0.10726, + "grad_norm": 0.899338268451886, + "learning_rate": 0.003, + "loss": 4.1095, + "step": 10726 + }, + { + "epoch": 0.10727, + "grad_norm": 0.9033602972489382, + "learning_rate": 0.003, + "loss": 4.1023, + "step": 10727 + }, + { + "epoch": 0.10728, + "grad_norm": 1.0178656794477339, + "learning_rate": 0.003, + "loss": 4.1171, + "step": 10728 + }, + { + "epoch": 0.10729, + "grad_norm": 1.0346921032937306, + "learning_rate": 0.003, + "loss": 4.0884, + "step": 10729 + }, + { + "epoch": 0.1073, + "grad_norm": 1.1570033282543726, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 10730 + }, + { + "epoch": 0.10731, + "grad_norm": 1.115174208842778, + "learning_rate": 0.003, + "loss": 4.0913, + "step": 10731 + }, + { + "epoch": 0.10732, + "grad_norm": 0.9982390765213424, + "learning_rate": 0.003, + "loss": 4.1073, + "step": 10732 + }, + { + "epoch": 0.10733, + "grad_norm": 1.148712295957351, + "learning_rate": 0.003, + "loss": 4.1125, + "step": 10733 + }, + { + "epoch": 0.10734, + "grad_norm": 0.9579540214332398, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 10734 + }, + { + "epoch": 0.10735, + "grad_norm": 1.0146855926830516, + "learning_rate": 0.003, + "loss": 4.1021, + "step": 10735 + }, + { + "epoch": 0.10736, + "grad_norm": 0.9742736541350404, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 10736 + }, + { + "epoch": 0.10737, + "grad_norm": 1.1404196655002237, + "learning_rate": 0.003, + "loss": 4.1016, + "step": 10737 + }, + { + "epoch": 0.10738, + "grad_norm": 0.933862597181587, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 10738 + }, + { + "epoch": 0.10739, + "grad_norm": 1.1088758822089204, + "learning_rate": 0.003, + "loss": 4.1015, + "step": 10739 + }, + { + "epoch": 0.1074, + "grad_norm": 1.0826806088118508, + "learning_rate": 0.003, + "loss": 4.0819, + "step": 10740 + }, + { + "epoch": 0.10741, + "grad_norm": 1.1615657148358984, + "learning_rate": 0.003, + "loss": 4.109, + "step": 10741 + }, + { + "epoch": 0.10742, + "grad_norm": 0.9748942411557058, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 10742 + }, + { + "epoch": 0.10743, + "grad_norm": 0.8591129336319034, + "learning_rate": 0.003, + "loss": 4.1092, + "step": 10743 + }, + { + "epoch": 0.10744, + "grad_norm": 0.68465842365011, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 10744 + }, + { + "epoch": 0.10745, + "grad_norm": 0.735963818043883, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 10745 + }, + { + "epoch": 0.10746, + "grad_norm": 0.7032873760724191, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 10746 + }, + { + "epoch": 0.10747, + "grad_norm": 0.8084149294400564, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 10747 + }, + { + "epoch": 0.10748, + "grad_norm": 1.0000248940159007, + "learning_rate": 0.003, + "loss": 4.1127, + "step": 10748 + }, + { + "epoch": 0.10749, + "grad_norm": 1.1865668409298504, + "learning_rate": 0.003, + "loss": 4.0993, + "step": 10749 + }, + { + "epoch": 0.1075, + "grad_norm": 0.8813887536453154, + "learning_rate": 0.003, + "loss": 4.101, + "step": 10750 + }, + { + "epoch": 0.10751, + "grad_norm": 0.7136999649728583, + "learning_rate": 0.003, + "loss": 4.1106, + "step": 10751 + }, + { + "epoch": 0.10752, + "grad_norm": 0.6285408218121631, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 10752 + }, + { + "epoch": 0.10753, + "grad_norm": 0.6256897266060606, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 10753 + }, + { + "epoch": 0.10754, + "grad_norm": 0.6869578345219896, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 10754 + }, + { + "epoch": 0.10755, + "grad_norm": 0.8698374354194014, + "learning_rate": 0.003, + "loss": 4.0984, + "step": 10755 + }, + { + "epoch": 0.10756, + "grad_norm": 0.9721107858575821, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 10756 + }, + { + "epoch": 0.10757, + "grad_norm": 1.2652471059106425, + "learning_rate": 0.003, + "loss": 4.0965, + "step": 10757 + }, + { + "epoch": 0.10758, + "grad_norm": 0.8051147018673998, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 10758 + }, + { + "epoch": 0.10759, + "grad_norm": 0.7887840797107059, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 10759 + }, + { + "epoch": 0.1076, + "grad_norm": 0.8756361057881464, + "learning_rate": 0.003, + "loss": 4.0917, + "step": 10760 + }, + { + "epoch": 0.10761, + "grad_norm": 0.8653144884210894, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 10761 + }, + { + "epoch": 0.10762, + "grad_norm": 1.001096932225531, + "learning_rate": 0.003, + "loss": 4.0982, + "step": 10762 + }, + { + "epoch": 0.10763, + "grad_norm": 1.2186220085471016, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 10763 + }, + { + "epoch": 0.10764, + "grad_norm": 0.9032971227920743, + "learning_rate": 0.003, + "loss": 4.091, + "step": 10764 + }, + { + "epoch": 0.10765, + "grad_norm": 0.7649265310167783, + "learning_rate": 0.003, + "loss": 4.08, + "step": 10765 + }, + { + "epoch": 0.10766, + "grad_norm": 0.7902275111508261, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 10766 + }, + { + "epoch": 0.10767, + "grad_norm": 0.7502679344823279, + "learning_rate": 0.003, + "loss": 4.074, + "step": 10767 + }, + { + "epoch": 0.10768, + "grad_norm": 0.8838195129661887, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 10768 + }, + { + "epoch": 0.10769, + "grad_norm": 1.0134134056123545, + "learning_rate": 0.003, + "loss": 4.0848, + "step": 10769 + }, + { + "epoch": 0.1077, + "grad_norm": 1.0986657093578556, + "learning_rate": 0.003, + "loss": 4.0959, + "step": 10770 + }, + { + "epoch": 0.10771, + "grad_norm": 0.9340339798670153, + "learning_rate": 0.003, + "loss": 4.097, + "step": 10771 + }, + { + "epoch": 0.10772, + "grad_norm": 0.8029309577541978, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 10772 + }, + { + "epoch": 0.10773, + "grad_norm": 0.7588033130832516, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 10773 + }, + { + "epoch": 0.10774, + "grad_norm": 0.7630196837242262, + "learning_rate": 0.003, + "loss": 4.0918, + "step": 10774 + }, + { + "epoch": 0.10775, + "grad_norm": 0.8215505719441673, + "learning_rate": 0.003, + "loss": 4.0942, + "step": 10775 + }, + { + "epoch": 0.10776, + "grad_norm": 0.9014502375445345, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 10776 + }, + { + "epoch": 0.10777, + "grad_norm": 0.8847353524677961, + "learning_rate": 0.003, + "loss": 4.0915, + "step": 10777 + }, + { + "epoch": 0.10778, + "grad_norm": 1.01748890463613, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 10778 + }, + { + "epoch": 0.10779, + "grad_norm": 1.2809884957311795, + "learning_rate": 0.003, + "loss": 4.073, + "step": 10779 + }, + { + "epoch": 0.1078, + "grad_norm": 0.7660630057398302, + "learning_rate": 0.003, + "loss": 4.1017, + "step": 10780 + }, + { + "epoch": 0.10781, + "grad_norm": 0.8149022369528812, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 10781 + }, + { + "epoch": 0.10782, + "grad_norm": 0.8050481083784805, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 10782 + }, + { + "epoch": 0.10783, + "grad_norm": 0.9537092641547571, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 10783 + }, + { + "epoch": 0.10784, + "grad_norm": 1.3966941530227983, + "learning_rate": 0.003, + "loss": 4.1182, + "step": 10784 + }, + { + "epoch": 0.10785, + "grad_norm": 0.8280524695857797, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 10785 + }, + { + "epoch": 0.10786, + "grad_norm": 0.778339206683131, + "learning_rate": 0.003, + "loss": 4.1001, + "step": 10786 + }, + { + "epoch": 0.10787, + "grad_norm": 0.671623911614919, + "learning_rate": 0.003, + "loss": 4.06, + "step": 10787 + }, + { + "epoch": 0.10788, + "grad_norm": 0.6036263181837579, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 10788 + }, + { + "epoch": 0.10789, + "grad_norm": 0.6528200096554009, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 10789 + }, + { + "epoch": 0.1079, + "grad_norm": 0.6416303971290929, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 10790 + }, + { + "epoch": 0.10791, + "grad_norm": 0.6548304017193016, + "learning_rate": 0.003, + "loss": 4.0961, + "step": 10791 + }, + { + "epoch": 0.10792, + "grad_norm": 0.7270914157479781, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 10792 + }, + { + "epoch": 0.10793, + "grad_norm": 1.0369299116804704, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 10793 + }, + { + "epoch": 0.10794, + "grad_norm": 1.384215703067732, + "learning_rate": 0.003, + "loss": 4.1117, + "step": 10794 + }, + { + "epoch": 0.10795, + "grad_norm": 0.7369176628094001, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 10795 + }, + { + "epoch": 0.10796, + "grad_norm": 0.7599422151878855, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 10796 + }, + { + "epoch": 0.10797, + "grad_norm": 0.8406436486957485, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 10797 + }, + { + "epoch": 0.10798, + "grad_norm": 0.9233274754105422, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 10798 + }, + { + "epoch": 0.10799, + "grad_norm": 1.0308525804747453, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 10799 + }, + { + "epoch": 0.108, + "grad_norm": 1.3733713740733506, + "learning_rate": 0.003, + "loss": 4.1006, + "step": 10800 + }, + { + "epoch": 0.10801, + "grad_norm": 0.7777407820185481, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 10801 + }, + { + "epoch": 0.10802, + "grad_norm": 0.7512742383039922, + "learning_rate": 0.003, + "loss": 4.0819, + "step": 10802 + }, + { + "epoch": 0.10803, + "grad_norm": 0.8642876455095708, + "learning_rate": 0.003, + "loss": 4.0991, + "step": 10803 + }, + { + "epoch": 0.10804, + "grad_norm": 0.9156517760838031, + "learning_rate": 0.003, + "loss": 4.0902, + "step": 10804 + }, + { + "epoch": 0.10805, + "grad_norm": 1.006782190223208, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 10805 + }, + { + "epoch": 0.10806, + "grad_norm": 0.9276082245161695, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 10806 + }, + { + "epoch": 0.10807, + "grad_norm": 0.8800970139315261, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 10807 + }, + { + "epoch": 0.10808, + "grad_norm": 0.8708434858064211, + "learning_rate": 0.003, + "loss": 4.1651, + "step": 10808 + }, + { + "epoch": 0.10809, + "grad_norm": 0.9562426016218708, + "learning_rate": 0.003, + "loss": 4.0933, + "step": 10809 + }, + { + "epoch": 0.1081, + "grad_norm": 1.1304946703846293, + "learning_rate": 0.003, + "loss": 4.0941, + "step": 10810 + }, + { + "epoch": 0.10811, + "grad_norm": 1.1728312749962047, + "learning_rate": 0.003, + "loss": 4.1229, + "step": 10811 + }, + { + "epoch": 0.10812, + "grad_norm": 1.004768520862092, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 10812 + }, + { + "epoch": 0.10813, + "grad_norm": 1.0389055453445937, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 10813 + }, + { + "epoch": 0.10814, + "grad_norm": 0.9288635138435398, + "learning_rate": 0.003, + "loss": 4.087, + "step": 10814 + }, + { + "epoch": 0.10815, + "grad_norm": 0.9600826909473025, + "learning_rate": 0.003, + "loss": 4.0956, + "step": 10815 + }, + { + "epoch": 0.10816, + "grad_norm": 0.9117913190313875, + "learning_rate": 0.003, + "loss": 4.0933, + "step": 10816 + }, + { + "epoch": 0.10817, + "grad_norm": 0.8936748628766625, + "learning_rate": 0.003, + "loss": 4.0873, + "step": 10817 + }, + { + "epoch": 0.10818, + "grad_norm": 0.8099817305585081, + "learning_rate": 0.003, + "loss": 4.0876, + "step": 10818 + }, + { + "epoch": 0.10819, + "grad_norm": 0.8103738281842612, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 10819 + }, + { + "epoch": 0.1082, + "grad_norm": 0.9227431088994197, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 10820 + }, + { + "epoch": 0.10821, + "grad_norm": 1.1531314792393335, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 10821 + }, + { + "epoch": 0.10822, + "grad_norm": 0.9768833174780258, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 10822 + }, + { + "epoch": 0.10823, + "grad_norm": 1.175142689741492, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 10823 + }, + { + "epoch": 0.10824, + "grad_norm": 1.0031723507911412, + "learning_rate": 0.003, + "loss": 4.0961, + "step": 10824 + }, + { + "epoch": 0.10825, + "grad_norm": 1.1588790873269992, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 10825 + }, + { + "epoch": 0.10826, + "grad_norm": 0.9303880393221259, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 10826 + }, + { + "epoch": 0.10827, + "grad_norm": 0.9258937470294385, + "learning_rate": 0.003, + "loss": 4.1185, + "step": 10827 + }, + { + "epoch": 0.10828, + "grad_norm": 0.929102258716846, + "learning_rate": 0.003, + "loss": 4.1247, + "step": 10828 + }, + { + "epoch": 0.10829, + "grad_norm": 1.0480497265689674, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 10829 + }, + { + "epoch": 0.1083, + "grad_norm": 1.036344976500489, + "learning_rate": 0.003, + "loss": 4.1063, + "step": 10830 + }, + { + "epoch": 0.10831, + "grad_norm": 0.9565655898099835, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 10831 + }, + { + "epoch": 0.10832, + "grad_norm": 0.9852948446291745, + "learning_rate": 0.003, + "loss": 4.1097, + "step": 10832 + }, + { + "epoch": 0.10833, + "grad_norm": 1.0349994727604397, + "learning_rate": 0.003, + "loss": 4.086, + "step": 10833 + }, + { + "epoch": 0.10834, + "grad_norm": 1.0891444771300862, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 10834 + }, + { + "epoch": 0.10835, + "grad_norm": 0.8994044236099675, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 10835 + }, + { + "epoch": 0.10836, + "grad_norm": 0.8860686028312054, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 10836 + }, + { + "epoch": 0.10837, + "grad_norm": 0.783324452511906, + "learning_rate": 0.003, + "loss": 4.0874, + "step": 10837 + }, + { + "epoch": 0.10838, + "grad_norm": 0.932607571087924, + "learning_rate": 0.003, + "loss": 4.108, + "step": 10838 + }, + { + "epoch": 0.10839, + "grad_norm": 1.0762597191955876, + "learning_rate": 0.003, + "loss": 4.0936, + "step": 10839 + }, + { + "epoch": 0.1084, + "grad_norm": 0.9429579433409142, + "learning_rate": 0.003, + "loss": 4.1079, + "step": 10840 + }, + { + "epoch": 0.10841, + "grad_norm": 1.060255728387347, + "learning_rate": 0.003, + "loss": 4.0969, + "step": 10841 + }, + { + "epoch": 0.10842, + "grad_norm": 1.0319932664205176, + "learning_rate": 0.003, + "loss": 4.1047, + "step": 10842 + }, + { + "epoch": 0.10843, + "grad_norm": 0.9923719329184173, + "learning_rate": 0.003, + "loss": 4.0919, + "step": 10843 + }, + { + "epoch": 0.10844, + "grad_norm": 1.096871373126558, + "learning_rate": 0.003, + "loss": 4.1242, + "step": 10844 + }, + { + "epoch": 0.10845, + "grad_norm": 0.9846887727736532, + "learning_rate": 0.003, + "loss": 4.1129, + "step": 10845 + }, + { + "epoch": 0.10846, + "grad_norm": 0.8997544751921531, + "learning_rate": 0.003, + "loss": 4.0857, + "step": 10846 + }, + { + "epoch": 0.10847, + "grad_norm": 0.7977026926624953, + "learning_rate": 0.003, + "loss": 4.1074, + "step": 10847 + }, + { + "epoch": 0.10848, + "grad_norm": 0.937450359498046, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 10848 + }, + { + "epoch": 0.10849, + "grad_norm": 0.997335411021704, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 10849 + }, + { + "epoch": 0.1085, + "grad_norm": 1.1229669164175435, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 10850 + }, + { + "epoch": 0.10851, + "grad_norm": 0.846249426059861, + "learning_rate": 0.003, + "loss": 4.0927, + "step": 10851 + }, + { + "epoch": 0.10852, + "grad_norm": 0.8127237148247968, + "learning_rate": 0.003, + "loss": 4.097, + "step": 10852 + }, + { + "epoch": 0.10853, + "grad_norm": 0.9408573603419037, + "learning_rate": 0.003, + "loss": 4.1019, + "step": 10853 + }, + { + "epoch": 0.10854, + "grad_norm": 1.1499651810096763, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 10854 + }, + { + "epoch": 0.10855, + "grad_norm": 0.9069374409047449, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 10855 + }, + { + "epoch": 0.10856, + "grad_norm": 0.7653439488895566, + "learning_rate": 0.003, + "loss": 4.0884, + "step": 10856 + }, + { + "epoch": 0.10857, + "grad_norm": 0.7420154910670483, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 10857 + }, + { + "epoch": 0.10858, + "grad_norm": 0.8248792388786437, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 10858 + }, + { + "epoch": 0.10859, + "grad_norm": 1.0317702943710716, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 10859 + }, + { + "epoch": 0.1086, + "grad_norm": 0.952696102617975, + "learning_rate": 0.003, + "loss": 4.075, + "step": 10860 + }, + { + "epoch": 0.10861, + "grad_norm": 0.9250394865749145, + "learning_rate": 0.003, + "loss": 4.1017, + "step": 10861 + }, + { + "epoch": 0.10862, + "grad_norm": 1.1066668942508229, + "learning_rate": 0.003, + "loss": 4.1302, + "step": 10862 + }, + { + "epoch": 0.10863, + "grad_norm": 1.2103493561457483, + "learning_rate": 0.003, + "loss": 4.1038, + "step": 10863 + }, + { + "epoch": 0.10864, + "grad_norm": 0.787585391123894, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 10864 + }, + { + "epoch": 0.10865, + "grad_norm": 0.747357217765881, + "learning_rate": 0.003, + "loss": 4.09, + "step": 10865 + }, + { + "epoch": 0.10866, + "grad_norm": 0.8829908595103317, + "learning_rate": 0.003, + "loss": 4.0949, + "step": 10866 + }, + { + "epoch": 0.10867, + "grad_norm": 1.1374425781623256, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 10867 + }, + { + "epoch": 0.10868, + "grad_norm": 1.0349040766366795, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 10868 + }, + { + "epoch": 0.10869, + "grad_norm": 0.9978899178201878, + "learning_rate": 0.003, + "loss": 4.0902, + "step": 10869 + }, + { + "epoch": 0.1087, + "grad_norm": 0.7980503818023075, + "learning_rate": 0.003, + "loss": 4.0927, + "step": 10870 + }, + { + "epoch": 0.10871, + "grad_norm": 0.7920407245773502, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 10871 + }, + { + "epoch": 0.10872, + "grad_norm": 0.8794239510162324, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 10872 + }, + { + "epoch": 0.10873, + "grad_norm": 0.8509288292495303, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 10873 + }, + { + "epoch": 0.10874, + "grad_norm": 0.7927261085671014, + "learning_rate": 0.003, + "loss": 4.077, + "step": 10874 + }, + { + "epoch": 0.10875, + "grad_norm": 0.8946979702312902, + "learning_rate": 0.003, + "loss": 4.103, + "step": 10875 + }, + { + "epoch": 0.10876, + "grad_norm": 0.8353680132471624, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 10876 + }, + { + "epoch": 0.10877, + "grad_norm": 0.8765867777088605, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 10877 + }, + { + "epoch": 0.10878, + "grad_norm": 1.0511537570582092, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 10878 + }, + { + "epoch": 0.10879, + "grad_norm": 1.2123668245882337, + "learning_rate": 0.003, + "loss": 4.092, + "step": 10879 + }, + { + "epoch": 0.1088, + "grad_norm": 1.041668046072083, + "learning_rate": 0.003, + "loss": 4.078, + "step": 10880 + }, + { + "epoch": 0.10881, + "grad_norm": 0.9712918727966118, + "learning_rate": 0.003, + "loss": 4.061, + "step": 10881 + }, + { + "epoch": 0.10882, + "grad_norm": 0.8829471767605723, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 10882 + }, + { + "epoch": 0.10883, + "grad_norm": 0.8550075135741931, + "learning_rate": 0.003, + "loss": 4.0979, + "step": 10883 + }, + { + "epoch": 0.10884, + "grad_norm": 0.9456984111537783, + "learning_rate": 0.003, + "loss": 4.0856, + "step": 10884 + }, + { + "epoch": 0.10885, + "grad_norm": 1.1129123192171337, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 10885 + }, + { + "epoch": 0.10886, + "grad_norm": 0.9964431259487623, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 10886 + }, + { + "epoch": 0.10887, + "grad_norm": 1.0408179046043333, + "learning_rate": 0.003, + "loss": 4.1111, + "step": 10887 + }, + { + "epoch": 0.10888, + "grad_norm": 1.006069997919354, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 10888 + }, + { + "epoch": 0.10889, + "grad_norm": 0.9847900562619722, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 10889 + }, + { + "epoch": 0.1089, + "grad_norm": 1.028320827150604, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 10890 + }, + { + "epoch": 0.10891, + "grad_norm": 0.9694947050831249, + "learning_rate": 0.003, + "loss": 4.0954, + "step": 10891 + }, + { + "epoch": 0.10892, + "grad_norm": 0.9887471686666744, + "learning_rate": 0.003, + "loss": 4.0929, + "step": 10892 + }, + { + "epoch": 0.10893, + "grad_norm": 1.1362493728590952, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 10893 + }, + { + "epoch": 0.10894, + "grad_norm": 0.8684659337577851, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 10894 + }, + { + "epoch": 0.10895, + "grad_norm": 0.8056421791403896, + "learning_rate": 0.003, + "loss": 4.1088, + "step": 10895 + }, + { + "epoch": 0.10896, + "grad_norm": 0.6524868407555793, + "learning_rate": 0.003, + "loss": 4.0857, + "step": 10896 + }, + { + "epoch": 0.10897, + "grad_norm": 0.5888273067576061, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 10897 + }, + { + "epoch": 0.10898, + "grad_norm": 0.6882398845141253, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 10898 + }, + { + "epoch": 0.10899, + "grad_norm": 0.7843648418141738, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 10899 + }, + { + "epoch": 0.109, + "grad_norm": 1.0047359388324504, + "learning_rate": 0.003, + "loss": 4.099, + "step": 10900 + }, + { + "epoch": 0.10901, + "grad_norm": 1.3655813247527946, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 10901 + }, + { + "epoch": 0.10902, + "grad_norm": 0.7058363978292598, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 10902 + }, + { + "epoch": 0.10903, + "grad_norm": 0.7683366849180756, + "learning_rate": 0.003, + "loss": 4.088, + "step": 10903 + }, + { + "epoch": 0.10904, + "grad_norm": 0.9399839184976495, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 10904 + }, + { + "epoch": 0.10905, + "grad_norm": 1.0993930789753907, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 10905 + }, + { + "epoch": 0.10906, + "grad_norm": 0.8520354249043842, + "learning_rate": 0.003, + "loss": 4.104, + "step": 10906 + }, + { + "epoch": 0.10907, + "grad_norm": 0.8826930777771175, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 10907 + }, + { + "epoch": 0.10908, + "grad_norm": 0.9528476184163581, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 10908 + }, + { + "epoch": 0.10909, + "grad_norm": 0.8668525472343837, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 10909 + }, + { + "epoch": 0.1091, + "grad_norm": 0.8352256892038756, + "learning_rate": 0.003, + "loss": 4.082, + "step": 10910 + }, + { + "epoch": 0.10911, + "grad_norm": 1.025931276097786, + "learning_rate": 0.003, + "loss": 4.0907, + "step": 10911 + }, + { + "epoch": 0.10912, + "grad_norm": 1.2733964343963418, + "learning_rate": 0.003, + "loss": 4.0881, + "step": 10912 + }, + { + "epoch": 0.10913, + "grad_norm": 0.7871613837364045, + "learning_rate": 0.003, + "loss": 4.1008, + "step": 10913 + }, + { + "epoch": 0.10914, + "grad_norm": 0.666707821006116, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 10914 + }, + { + "epoch": 0.10915, + "grad_norm": 0.7105166812984488, + "learning_rate": 0.003, + "loss": 4.079, + "step": 10915 + }, + { + "epoch": 0.10916, + "grad_norm": 0.8971189004384993, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 10916 + }, + { + "epoch": 0.10917, + "grad_norm": 1.039296938528312, + "learning_rate": 0.003, + "loss": 4.1028, + "step": 10917 + }, + { + "epoch": 0.10918, + "grad_norm": 0.9920137477536368, + "learning_rate": 0.003, + "loss": 4.1004, + "step": 10918 + }, + { + "epoch": 0.10919, + "grad_norm": 1.1877251011329901, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 10919 + }, + { + "epoch": 0.1092, + "grad_norm": 1.013317185290994, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 10920 + }, + { + "epoch": 0.10921, + "grad_norm": 1.0145736214361518, + "learning_rate": 0.003, + "loss": 4.1031, + "step": 10921 + }, + { + "epoch": 0.10922, + "grad_norm": 1.1317832535082821, + "learning_rate": 0.003, + "loss": 4.0978, + "step": 10922 + }, + { + "epoch": 0.10923, + "grad_norm": 1.0827398756097473, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 10923 + }, + { + "epoch": 0.10924, + "grad_norm": 0.923238594839259, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 10924 + }, + { + "epoch": 0.10925, + "grad_norm": 0.8227562811330069, + "learning_rate": 0.003, + "loss": 4.0931, + "step": 10925 + }, + { + "epoch": 0.10926, + "grad_norm": 0.7649983136272475, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 10926 + }, + { + "epoch": 0.10927, + "grad_norm": 0.6900790644043666, + "learning_rate": 0.003, + "loss": 4.0886, + "step": 10927 + }, + { + "epoch": 0.10928, + "grad_norm": 0.6132030166402541, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 10928 + }, + { + "epoch": 0.10929, + "grad_norm": 0.6035814171512837, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 10929 + }, + { + "epoch": 0.1093, + "grad_norm": 0.6427431795590084, + "learning_rate": 0.003, + "loss": 4.0893, + "step": 10930 + }, + { + "epoch": 0.10931, + "grad_norm": 0.7061120077448015, + "learning_rate": 0.003, + "loss": 4.0888, + "step": 10931 + }, + { + "epoch": 0.10932, + "grad_norm": 0.7388653184706705, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 10932 + }, + { + "epoch": 0.10933, + "grad_norm": 0.7768950073325869, + "learning_rate": 0.003, + "loss": 4.0933, + "step": 10933 + }, + { + "epoch": 0.10934, + "grad_norm": 0.8345740858188827, + "learning_rate": 0.003, + "loss": 4.1197, + "step": 10934 + }, + { + "epoch": 0.10935, + "grad_norm": 1.040327480375917, + "learning_rate": 0.003, + "loss": 4.095, + "step": 10935 + }, + { + "epoch": 0.10936, + "grad_norm": 1.4485576309662618, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 10936 + }, + { + "epoch": 0.10937, + "grad_norm": 0.8573582167655448, + "learning_rate": 0.003, + "loss": 4.1039, + "step": 10937 + }, + { + "epoch": 0.10938, + "grad_norm": 0.9474959826862678, + "learning_rate": 0.003, + "loss": 4.105, + "step": 10938 + }, + { + "epoch": 0.10939, + "grad_norm": 1.106781413477435, + "learning_rate": 0.003, + "loss": 4.0907, + "step": 10939 + }, + { + "epoch": 0.1094, + "grad_norm": 1.0996236402570079, + "learning_rate": 0.003, + "loss": 4.1285, + "step": 10940 + }, + { + "epoch": 0.10941, + "grad_norm": 0.8607805805154458, + "learning_rate": 0.003, + "loss": 4.1089, + "step": 10941 + }, + { + "epoch": 0.10942, + "grad_norm": 0.9439785387240215, + "learning_rate": 0.003, + "loss": 4.106, + "step": 10942 + }, + { + "epoch": 0.10943, + "grad_norm": 1.0354065609966294, + "learning_rate": 0.003, + "loss": 4.0884, + "step": 10943 + }, + { + "epoch": 0.10944, + "grad_norm": 0.948319873180743, + "learning_rate": 0.003, + "loss": 4.0994, + "step": 10944 + }, + { + "epoch": 0.10945, + "grad_norm": 1.0953808334086788, + "learning_rate": 0.003, + "loss": 4.1017, + "step": 10945 + }, + { + "epoch": 0.10946, + "grad_norm": 1.0632699149558704, + "learning_rate": 0.003, + "loss": 4.1018, + "step": 10946 + }, + { + "epoch": 0.10947, + "grad_norm": 1.3382798322330052, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 10947 + }, + { + "epoch": 0.10948, + "grad_norm": 0.8377152474539921, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 10948 + }, + { + "epoch": 0.10949, + "grad_norm": 0.7342418285709524, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 10949 + }, + { + "epoch": 0.1095, + "grad_norm": 0.6862244442697666, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 10950 + }, + { + "epoch": 0.10951, + "grad_norm": 0.643394937539552, + "learning_rate": 0.003, + "loss": 4.064, + "step": 10951 + }, + { + "epoch": 0.10952, + "grad_norm": 0.6052894428544429, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 10952 + }, + { + "epoch": 0.10953, + "grad_norm": 0.6696715717531783, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 10953 + }, + { + "epoch": 0.10954, + "grad_norm": 0.8222393584303349, + "learning_rate": 0.003, + "loss": 4.1087, + "step": 10954 + }, + { + "epoch": 0.10955, + "grad_norm": 1.0689573660145186, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 10955 + }, + { + "epoch": 0.10956, + "grad_norm": 1.3874261261611496, + "learning_rate": 0.003, + "loss": 4.1061, + "step": 10956 + }, + { + "epoch": 0.10957, + "grad_norm": 0.9163575778279299, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 10957 + }, + { + "epoch": 0.10958, + "grad_norm": 0.754609707118034, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 10958 + }, + { + "epoch": 0.10959, + "grad_norm": 0.7037295999822283, + "learning_rate": 0.003, + "loss": 4.0878, + "step": 10959 + }, + { + "epoch": 0.1096, + "grad_norm": 0.8613954326312055, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 10960 + }, + { + "epoch": 0.10961, + "grad_norm": 1.0593249340625663, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 10961 + }, + { + "epoch": 0.10962, + "grad_norm": 1.2797578189333763, + "learning_rate": 0.003, + "loss": 4.1089, + "step": 10962 + }, + { + "epoch": 0.10963, + "grad_norm": 0.7864541877096417, + "learning_rate": 0.003, + "loss": 4.0931, + "step": 10963 + }, + { + "epoch": 0.10964, + "grad_norm": 0.6978724807726525, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 10964 + }, + { + "epoch": 0.10965, + "grad_norm": 0.7187864289443937, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 10965 + }, + { + "epoch": 0.10966, + "grad_norm": 0.7724151950485852, + "learning_rate": 0.003, + "loss": 4.075, + "step": 10966 + }, + { + "epoch": 0.10967, + "grad_norm": 0.9099722418261509, + "learning_rate": 0.003, + "loss": 4.0967, + "step": 10967 + }, + { + "epoch": 0.10968, + "grad_norm": 1.0808764062773821, + "learning_rate": 0.003, + "loss": 4.107, + "step": 10968 + }, + { + "epoch": 0.10969, + "grad_norm": 0.9671692578700375, + "learning_rate": 0.003, + "loss": 4.1028, + "step": 10969 + }, + { + "epoch": 0.1097, + "grad_norm": 0.9753230819970405, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 10970 + }, + { + "epoch": 0.10971, + "grad_norm": 1.1627616976665547, + "learning_rate": 0.003, + "loss": 4.1058, + "step": 10971 + }, + { + "epoch": 0.10972, + "grad_norm": 0.8544407067883085, + "learning_rate": 0.003, + "loss": 4.1238, + "step": 10972 + }, + { + "epoch": 0.10973, + "grad_norm": 0.8332612683027971, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 10973 + }, + { + "epoch": 0.10974, + "grad_norm": 0.9338340109940729, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 10974 + }, + { + "epoch": 0.10975, + "grad_norm": 1.129357217193588, + "learning_rate": 0.003, + "loss": 4.0986, + "step": 10975 + }, + { + "epoch": 0.10976, + "grad_norm": 1.1298185545722819, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 10976 + }, + { + "epoch": 0.10977, + "grad_norm": 0.921379294676142, + "learning_rate": 0.003, + "loss": 4.0888, + "step": 10977 + }, + { + "epoch": 0.10978, + "grad_norm": 0.8920657450968826, + "learning_rate": 0.003, + "loss": 4.0971, + "step": 10978 + }, + { + "epoch": 0.10979, + "grad_norm": 0.8158802576014977, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 10979 + }, + { + "epoch": 0.1098, + "grad_norm": 0.6844655800412194, + "learning_rate": 0.003, + "loss": 4.071, + "step": 10980 + }, + { + "epoch": 0.10981, + "grad_norm": 0.7125024538991588, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 10981 + }, + { + "epoch": 0.10982, + "grad_norm": 0.8280715988209589, + "learning_rate": 0.003, + "loss": 4.057, + "step": 10982 + }, + { + "epoch": 0.10983, + "grad_norm": 0.9459628221512237, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 10983 + }, + { + "epoch": 0.10984, + "grad_norm": 1.327797457277312, + "learning_rate": 0.003, + "loss": 4.1174, + "step": 10984 + }, + { + "epoch": 0.10985, + "grad_norm": 0.9552175912523253, + "learning_rate": 0.003, + "loss": 4.0922, + "step": 10985 + }, + { + "epoch": 0.10986, + "grad_norm": 1.1781542520269463, + "learning_rate": 0.003, + "loss": 4.0985, + "step": 10986 + }, + { + "epoch": 0.10987, + "grad_norm": 0.9250515259297767, + "learning_rate": 0.003, + "loss": 4.1063, + "step": 10987 + }, + { + "epoch": 0.10988, + "grad_norm": 0.9713876518703021, + "learning_rate": 0.003, + "loss": 4.0989, + "step": 10988 + }, + { + "epoch": 0.10989, + "grad_norm": 1.054620062319649, + "learning_rate": 0.003, + "loss": 4.0919, + "step": 10989 + }, + { + "epoch": 0.1099, + "grad_norm": 1.0437729852616782, + "learning_rate": 0.003, + "loss": 4.0898, + "step": 10990 + }, + { + "epoch": 0.10991, + "grad_norm": 1.3225827159424122, + "learning_rate": 0.003, + "loss": 4.1015, + "step": 10991 + }, + { + "epoch": 0.10992, + "grad_norm": 1.0162821935814053, + "learning_rate": 0.003, + "loss": 4.0886, + "step": 10992 + }, + { + "epoch": 0.10993, + "grad_norm": 0.8925078076727114, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 10993 + }, + { + "epoch": 0.10994, + "grad_norm": 0.8755239360700177, + "learning_rate": 0.003, + "loss": 4.1263, + "step": 10994 + }, + { + "epoch": 0.10995, + "grad_norm": 0.9637995285050126, + "learning_rate": 0.003, + "loss": 4.0991, + "step": 10995 + }, + { + "epoch": 0.10996, + "grad_norm": 1.1358840166513098, + "learning_rate": 0.003, + "loss": 4.0954, + "step": 10996 + }, + { + "epoch": 0.10997, + "grad_norm": 0.9021719056480201, + "learning_rate": 0.003, + "loss": 4.1183, + "step": 10997 + }, + { + "epoch": 0.10998, + "grad_norm": 0.8373733639266345, + "learning_rate": 0.003, + "loss": 4.1207, + "step": 10998 + }, + { + "epoch": 0.10999, + "grad_norm": 0.9207340181431739, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 10999 + }, + { + "epoch": 0.11, + "grad_norm": 1.156758575475058, + "learning_rate": 0.003, + "loss": 4.1216, + "step": 11000 + }, + { + "epoch": 0.11001, + "grad_norm": 1.096152978128535, + "learning_rate": 0.003, + "loss": 4.1161, + "step": 11001 + }, + { + "epoch": 0.11002, + "grad_norm": 0.7761665995523702, + "learning_rate": 0.003, + "loss": 4.0952, + "step": 11002 + }, + { + "epoch": 0.11003, + "grad_norm": 0.689949179996817, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 11003 + }, + { + "epoch": 0.11004, + "grad_norm": 0.9491308507448749, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 11004 + }, + { + "epoch": 0.11005, + "grad_norm": 1.351291764481279, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 11005 + }, + { + "epoch": 0.11006, + "grad_norm": 0.8491924026368354, + "learning_rate": 0.003, + "loss": 4.0983, + "step": 11006 + }, + { + "epoch": 0.11007, + "grad_norm": 0.7961046968789953, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 11007 + }, + { + "epoch": 0.11008, + "grad_norm": 0.7069454453128994, + "learning_rate": 0.003, + "loss": 4.0944, + "step": 11008 + }, + { + "epoch": 0.11009, + "grad_norm": 0.6713027966756064, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 11009 + }, + { + "epoch": 0.1101, + "grad_norm": 0.7580935808366495, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 11010 + }, + { + "epoch": 0.11011, + "grad_norm": 0.8614513279135587, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 11011 + }, + { + "epoch": 0.11012, + "grad_norm": 0.9918113988484457, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 11012 + }, + { + "epoch": 0.11013, + "grad_norm": 1.0426262798114065, + "learning_rate": 0.003, + "loss": 4.0928, + "step": 11013 + }, + { + "epoch": 0.11014, + "grad_norm": 1.106008110717952, + "learning_rate": 0.003, + "loss": 4.0964, + "step": 11014 + }, + { + "epoch": 0.11015, + "grad_norm": 0.7991285697856759, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 11015 + }, + { + "epoch": 0.11016, + "grad_norm": 0.8223541871893375, + "learning_rate": 0.003, + "loss": 4.0856, + "step": 11016 + }, + { + "epoch": 0.11017, + "grad_norm": 0.9155722679630341, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 11017 + }, + { + "epoch": 0.11018, + "grad_norm": 0.8951822172603995, + "learning_rate": 0.003, + "loss": 4.1041, + "step": 11018 + }, + { + "epoch": 0.11019, + "grad_norm": 0.9673750713778863, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 11019 + }, + { + "epoch": 0.1102, + "grad_norm": 1.2188841511739732, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 11020 + }, + { + "epoch": 0.11021, + "grad_norm": 1.0049622835492242, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 11021 + }, + { + "epoch": 0.11022, + "grad_norm": 1.3258342387152022, + "learning_rate": 0.003, + "loss": 4.1105, + "step": 11022 + }, + { + "epoch": 0.11023, + "grad_norm": 0.7985389136408283, + "learning_rate": 0.003, + "loss": 4.1044, + "step": 11023 + }, + { + "epoch": 0.11024, + "grad_norm": 0.6470329556565424, + "learning_rate": 0.003, + "loss": 4.102, + "step": 11024 + }, + { + "epoch": 0.11025, + "grad_norm": 0.8558692970871703, + "learning_rate": 0.003, + "loss": 4.086, + "step": 11025 + }, + { + "epoch": 0.11026, + "grad_norm": 1.129985866662115, + "learning_rate": 0.003, + "loss": 4.086, + "step": 11026 + }, + { + "epoch": 0.11027, + "grad_norm": 1.1261394509385085, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 11027 + }, + { + "epoch": 0.11028, + "grad_norm": 1.0243735261897322, + "learning_rate": 0.003, + "loss": 4.084, + "step": 11028 + }, + { + "epoch": 0.11029, + "grad_norm": 0.9580662586961145, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 11029 + }, + { + "epoch": 0.1103, + "grad_norm": 0.9395783295496979, + "learning_rate": 0.003, + "loss": 4.0864, + "step": 11030 + }, + { + "epoch": 0.11031, + "grad_norm": 1.0369664734472004, + "learning_rate": 0.003, + "loss": 4.0887, + "step": 11031 + }, + { + "epoch": 0.11032, + "grad_norm": 1.011213942969309, + "learning_rate": 0.003, + "loss": 4.1008, + "step": 11032 + }, + { + "epoch": 0.11033, + "grad_norm": 0.8977693764866891, + "learning_rate": 0.003, + "loss": 4.0848, + "step": 11033 + }, + { + "epoch": 0.11034, + "grad_norm": 0.8882910151720783, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 11034 + }, + { + "epoch": 0.11035, + "grad_norm": 0.8972998903609543, + "learning_rate": 0.003, + "loss": 4.0985, + "step": 11035 + }, + { + "epoch": 0.11036, + "grad_norm": 0.9257301614344922, + "learning_rate": 0.003, + "loss": 4.1106, + "step": 11036 + }, + { + "epoch": 0.11037, + "grad_norm": 1.1414986147545914, + "learning_rate": 0.003, + "loss": 4.067, + "step": 11037 + }, + { + "epoch": 0.11038, + "grad_norm": 1.0081977832150641, + "learning_rate": 0.003, + "loss": 4.086, + "step": 11038 + }, + { + "epoch": 0.11039, + "grad_norm": 0.873201423247635, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 11039 + }, + { + "epoch": 0.1104, + "grad_norm": 0.8122033193164157, + "learning_rate": 0.003, + "loss": 4.1185, + "step": 11040 + }, + { + "epoch": 0.11041, + "grad_norm": 0.826302230637232, + "learning_rate": 0.003, + "loss": 4.0999, + "step": 11041 + }, + { + "epoch": 0.11042, + "grad_norm": 0.9975041209196024, + "learning_rate": 0.003, + "loss": 4.1113, + "step": 11042 + }, + { + "epoch": 0.11043, + "grad_norm": 1.1955454796200229, + "learning_rate": 0.003, + "loss": 4.1018, + "step": 11043 + }, + { + "epoch": 0.11044, + "grad_norm": 1.033576359849705, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 11044 + }, + { + "epoch": 0.11045, + "grad_norm": 1.2096398727470892, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 11045 + }, + { + "epoch": 0.11046, + "grad_norm": 0.9982758494376769, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 11046 + }, + { + "epoch": 0.11047, + "grad_norm": 1.0825385454864418, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 11047 + }, + { + "epoch": 0.11048, + "grad_norm": 0.993014400878651, + "learning_rate": 0.003, + "loss": 4.0891, + "step": 11048 + }, + { + "epoch": 0.11049, + "grad_norm": 1.05610698852147, + "learning_rate": 0.003, + "loss": 4.083, + "step": 11049 + }, + { + "epoch": 0.1105, + "grad_norm": 0.9097141940835776, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 11050 + }, + { + "epoch": 0.11051, + "grad_norm": 0.9272944375903635, + "learning_rate": 0.003, + "loss": 4.1023, + "step": 11051 + }, + { + "epoch": 0.11052, + "grad_norm": 1.001127810944245, + "learning_rate": 0.003, + "loss": 4.1006, + "step": 11052 + }, + { + "epoch": 0.11053, + "grad_norm": 1.0241929199820359, + "learning_rate": 0.003, + "loss": 4.0947, + "step": 11053 + }, + { + "epoch": 0.11054, + "grad_norm": 0.9559808250799029, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 11054 + }, + { + "epoch": 0.11055, + "grad_norm": 0.9400081710996248, + "learning_rate": 0.003, + "loss": 4.0936, + "step": 11055 + }, + { + "epoch": 0.11056, + "grad_norm": 0.8604009859626313, + "learning_rate": 0.003, + "loss": 4.0908, + "step": 11056 + }, + { + "epoch": 0.11057, + "grad_norm": 0.9217234695196611, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 11057 + }, + { + "epoch": 0.11058, + "grad_norm": 0.784947298848566, + "learning_rate": 0.003, + "loss": 4.1065, + "step": 11058 + }, + { + "epoch": 0.11059, + "grad_norm": 0.7460028735549842, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 11059 + }, + { + "epoch": 0.1106, + "grad_norm": 0.8787715138736634, + "learning_rate": 0.003, + "loss": 4.0955, + "step": 11060 + }, + { + "epoch": 0.11061, + "grad_norm": 1.0200169146337545, + "learning_rate": 0.003, + "loss": 4.1109, + "step": 11061 + }, + { + "epoch": 0.11062, + "grad_norm": 1.124959366104199, + "learning_rate": 0.003, + "loss": 4.1066, + "step": 11062 + }, + { + "epoch": 0.11063, + "grad_norm": 0.916201677239895, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 11063 + }, + { + "epoch": 0.11064, + "grad_norm": 0.9315380977712383, + "learning_rate": 0.003, + "loss": 4.0871, + "step": 11064 + }, + { + "epoch": 0.11065, + "grad_norm": 1.209899748436028, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 11065 + }, + { + "epoch": 0.11066, + "grad_norm": 0.9294400994131821, + "learning_rate": 0.003, + "loss": 4.103, + "step": 11066 + }, + { + "epoch": 0.11067, + "grad_norm": 0.991633154017647, + "learning_rate": 0.003, + "loss": 4.1076, + "step": 11067 + }, + { + "epoch": 0.11068, + "grad_norm": 1.0714268184959448, + "learning_rate": 0.003, + "loss": 4.084, + "step": 11068 + }, + { + "epoch": 0.11069, + "grad_norm": 0.8195547563673611, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 11069 + }, + { + "epoch": 0.1107, + "grad_norm": 0.732814068880551, + "learning_rate": 0.003, + "loss": 4.086, + "step": 11070 + }, + { + "epoch": 0.11071, + "grad_norm": 0.6572446226663663, + "learning_rate": 0.003, + "loss": 4.084, + "step": 11071 + }, + { + "epoch": 0.11072, + "grad_norm": 0.7018960301779674, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 11072 + }, + { + "epoch": 0.11073, + "grad_norm": 0.7759815253031658, + "learning_rate": 0.003, + "loss": 4.1101, + "step": 11073 + }, + { + "epoch": 0.11074, + "grad_norm": 0.9564886919505228, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 11074 + }, + { + "epoch": 0.11075, + "grad_norm": 1.0494440334116135, + "learning_rate": 0.003, + "loss": 4.1047, + "step": 11075 + }, + { + "epoch": 0.11076, + "grad_norm": 0.9723087924510206, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 11076 + }, + { + "epoch": 0.11077, + "grad_norm": 1.0639655612576544, + "learning_rate": 0.003, + "loss": 4.1356, + "step": 11077 + }, + { + "epoch": 0.11078, + "grad_norm": 0.917849200277183, + "learning_rate": 0.003, + "loss": 4.0982, + "step": 11078 + }, + { + "epoch": 0.11079, + "grad_norm": 0.9054234037593348, + "learning_rate": 0.003, + "loss": 4.0819, + "step": 11079 + }, + { + "epoch": 0.1108, + "grad_norm": 0.7629989303746657, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 11080 + }, + { + "epoch": 0.11081, + "grad_norm": 0.7575568421942913, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 11081 + }, + { + "epoch": 0.11082, + "grad_norm": 0.8518139910145223, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 11082 + }, + { + "epoch": 0.11083, + "grad_norm": 0.9585594201890845, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 11083 + }, + { + "epoch": 0.11084, + "grad_norm": 1.085609944922328, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 11084 + }, + { + "epoch": 0.11085, + "grad_norm": 0.9101695845832788, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 11085 + }, + { + "epoch": 0.11086, + "grad_norm": 0.9028776625453008, + "learning_rate": 0.003, + "loss": 4.1075, + "step": 11086 + }, + { + "epoch": 0.11087, + "grad_norm": 0.9692671256140618, + "learning_rate": 0.003, + "loss": 4.0908, + "step": 11087 + }, + { + "epoch": 0.11088, + "grad_norm": 1.0627491720509683, + "learning_rate": 0.003, + "loss": 4.0857, + "step": 11088 + }, + { + "epoch": 0.11089, + "grad_norm": 1.2029278135112018, + "learning_rate": 0.003, + "loss": 4.0857, + "step": 11089 + }, + { + "epoch": 0.1109, + "grad_norm": 1.0287197405343258, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 11090 + }, + { + "epoch": 0.11091, + "grad_norm": 0.9245506300339037, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 11091 + }, + { + "epoch": 0.11092, + "grad_norm": 0.827029030829062, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 11092 + }, + { + "epoch": 0.11093, + "grad_norm": 0.8557287516168338, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 11093 + }, + { + "epoch": 0.11094, + "grad_norm": 0.9368118396658807, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 11094 + }, + { + "epoch": 0.11095, + "grad_norm": 1.0782586608083278, + "learning_rate": 0.003, + "loss": 4.1075, + "step": 11095 + }, + { + "epoch": 0.11096, + "grad_norm": 1.1023521572285724, + "learning_rate": 0.003, + "loss": 4.0982, + "step": 11096 + }, + { + "epoch": 0.11097, + "grad_norm": 1.0015734120904822, + "learning_rate": 0.003, + "loss": 4.1058, + "step": 11097 + }, + { + "epoch": 0.11098, + "grad_norm": 0.9600311849967388, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 11098 + }, + { + "epoch": 0.11099, + "grad_norm": 0.9941734619190143, + "learning_rate": 0.003, + "loss": 4.1225, + "step": 11099 + }, + { + "epoch": 0.111, + "grad_norm": 1.031837046406522, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 11100 + }, + { + "epoch": 0.11101, + "grad_norm": 0.8898820041859903, + "learning_rate": 0.003, + "loss": 4.0997, + "step": 11101 + }, + { + "epoch": 0.11102, + "grad_norm": 0.9479145957708425, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 11102 + }, + { + "epoch": 0.11103, + "grad_norm": 1.1956894043237827, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 11103 + }, + { + "epoch": 0.11104, + "grad_norm": 1.1494620571282435, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 11104 + }, + { + "epoch": 0.11105, + "grad_norm": 0.8707004090478873, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 11105 + }, + { + "epoch": 0.11106, + "grad_norm": 0.7349433464146315, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 11106 + }, + { + "epoch": 0.11107, + "grad_norm": 0.7220130094362489, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 11107 + }, + { + "epoch": 0.11108, + "grad_norm": 0.8051272196670756, + "learning_rate": 0.003, + "loss": 4.0987, + "step": 11108 + }, + { + "epoch": 0.11109, + "grad_norm": 0.9808992243255188, + "learning_rate": 0.003, + "loss": 4.1001, + "step": 11109 + }, + { + "epoch": 0.1111, + "grad_norm": 1.0945755747191372, + "learning_rate": 0.003, + "loss": 4.0976, + "step": 11110 + }, + { + "epoch": 0.11111, + "grad_norm": 0.8727803729560786, + "learning_rate": 0.003, + "loss": 4.0886, + "step": 11111 + }, + { + "epoch": 0.11112, + "grad_norm": 0.8545111085125617, + "learning_rate": 0.003, + "loss": 4.1081, + "step": 11112 + }, + { + "epoch": 0.11113, + "grad_norm": 0.7525233196938497, + "learning_rate": 0.003, + "loss": 4.1037, + "step": 11113 + }, + { + "epoch": 0.11114, + "grad_norm": 0.7910674126145393, + "learning_rate": 0.003, + "loss": 4.1182, + "step": 11114 + }, + { + "epoch": 0.11115, + "grad_norm": 0.8276312813704546, + "learning_rate": 0.003, + "loss": 4.0928, + "step": 11115 + }, + { + "epoch": 0.11116, + "grad_norm": 0.8147373907521616, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 11116 + }, + { + "epoch": 0.11117, + "grad_norm": 0.9207749154660626, + "learning_rate": 0.003, + "loss": 4.1071, + "step": 11117 + }, + { + "epoch": 0.11118, + "grad_norm": 1.0646327377759863, + "learning_rate": 0.003, + "loss": 4.077, + "step": 11118 + }, + { + "epoch": 0.11119, + "grad_norm": 1.1470454455346635, + "learning_rate": 0.003, + "loss": 4.1058, + "step": 11119 + }, + { + "epoch": 0.1112, + "grad_norm": 1.066140383289869, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 11120 + }, + { + "epoch": 0.11121, + "grad_norm": 0.9753401528065198, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 11121 + }, + { + "epoch": 0.11122, + "grad_norm": 0.8824408253438378, + "learning_rate": 0.003, + "loss": 4.1017, + "step": 11122 + }, + { + "epoch": 0.11123, + "grad_norm": 0.8317866745727764, + "learning_rate": 0.003, + "loss": 4.1039, + "step": 11123 + }, + { + "epoch": 0.11124, + "grad_norm": 0.9072928965650875, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 11124 + }, + { + "epoch": 0.11125, + "grad_norm": 0.9776849866686539, + "learning_rate": 0.003, + "loss": 4.1033, + "step": 11125 + }, + { + "epoch": 0.11126, + "grad_norm": 0.9845125508131952, + "learning_rate": 0.003, + "loss": 4.0878, + "step": 11126 + }, + { + "epoch": 0.11127, + "grad_norm": 0.9824110324657134, + "learning_rate": 0.003, + "loss": 4.0981, + "step": 11127 + }, + { + "epoch": 0.11128, + "grad_norm": 0.8546452510028201, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 11128 + }, + { + "epoch": 0.11129, + "grad_norm": 0.7586995390901379, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 11129 + }, + { + "epoch": 0.1113, + "grad_norm": 0.743251925158843, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 11130 + }, + { + "epoch": 0.11131, + "grad_norm": 0.8256420380724693, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 11131 + }, + { + "epoch": 0.11132, + "grad_norm": 1.2716154139962272, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 11132 + }, + { + "epoch": 0.11133, + "grad_norm": 1.0250601208578096, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 11133 + }, + { + "epoch": 0.11134, + "grad_norm": 0.9182703221225448, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 11134 + }, + { + "epoch": 0.11135, + "grad_norm": 0.8536892279414858, + "learning_rate": 0.003, + "loss": 4.085, + "step": 11135 + }, + { + "epoch": 0.11136, + "grad_norm": 0.8370368018329062, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 11136 + }, + { + "epoch": 0.11137, + "grad_norm": 0.8815792247909126, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 11137 + }, + { + "epoch": 0.11138, + "grad_norm": 0.9413435484987607, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 11138 + }, + { + "epoch": 0.11139, + "grad_norm": 1.0501462882930455, + "learning_rate": 0.003, + "loss": 4.1123, + "step": 11139 + }, + { + "epoch": 0.1114, + "grad_norm": 1.1774714576520517, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 11140 + }, + { + "epoch": 0.11141, + "grad_norm": 1.017372622427802, + "learning_rate": 0.003, + "loss": 4.0954, + "step": 11141 + }, + { + "epoch": 0.11142, + "grad_norm": 1.224238579952709, + "learning_rate": 0.003, + "loss": 4.0959, + "step": 11142 + }, + { + "epoch": 0.11143, + "grad_norm": 0.8519935794215302, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 11143 + }, + { + "epoch": 0.11144, + "grad_norm": 0.7095676589447457, + "learning_rate": 0.003, + "loss": 4.055, + "step": 11144 + }, + { + "epoch": 0.11145, + "grad_norm": 0.7164301726390853, + "learning_rate": 0.003, + "loss": 4.089, + "step": 11145 + }, + { + "epoch": 0.11146, + "grad_norm": 0.6906332051513319, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 11146 + }, + { + "epoch": 0.11147, + "grad_norm": 0.6015106604374845, + "learning_rate": 0.003, + "loss": 4.0991, + "step": 11147 + }, + { + "epoch": 0.11148, + "grad_norm": 0.7515293653070366, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 11148 + }, + { + "epoch": 0.11149, + "grad_norm": 0.8885554651852801, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 11149 + }, + { + "epoch": 0.1115, + "grad_norm": 0.7759497489859205, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 11150 + }, + { + "epoch": 0.11151, + "grad_norm": 0.6483512160935777, + "learning_rate": 0.003, + "loss": 4.0971, + "step": 11151 + }, + { + "epoch": 0.11152, + "grad_norm": 0.7011964153090885, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 11152 + }, + { + "epoch": 0.11153, + "grad_norm": 0.7282038016959226, + "learning_rate": 0.003, + "loss": 4.0919, + "step": 11153 + }, + { + "epoch": 0.11154, + "grad_norm": 0.7917007424276977, + "learning_rate": 0.003, + "loss": 4.0878, + "step": 11154 + }, + { + "epoch": 0.11155, + "grad_norm": 0.878985280797932, + "learning_rate": 0.003, + "loss": 4.0966, + "step": 11155 + }, + { + "epoch": 0.11156, + "grad_norm": 0.83019118876844, + "learning_rate": 0.003, + "loss": 4.0912, + "step": 11156 + }, + { + "epoch": 0.11157, + "grad_norm": 0.9199654211196693, + "learning_rate": 0.003, + "loss": 4.0994, + "step": 11157 + }, + { + "epoch": 0.11158, + "grad_norm": 0.9831062112346727, + "learning_rate": 0.003, + "loss": 4.0989, + "step": 11158 + }, + { + "epoch": 0.11159, + "grad_norm": 1.0046237547961085, + "learning_rate": 0.003, + "loss": 4.09, + "step": 11159 + }, + { + "epoch": 0.1116, + "grad_norm": 1.6781460929712169, + "learning_rate": 0.003, + "loss": 4.1121, + "step": 11160 + }, + { + "epoch": 0.11161, + "grad_norm": 0.8239089792845657, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 11161 + }, + { + "epoch": 0.11162, + "grad_norm": 0.8296662673388867, + "learning_rate": 0.003, + "loss": 4.0918, + "step": 11162 + }, + { + "epoch": 0.11163, + "grad_norm": 0.8492855848775818, + "learning_rate": 0.003, + "loss": 4.0855, + "step": 11163 + }, + { + "epoch": 0.11164, + "grad_norm": 1.047270252622794, + "learning_rate": 0.003, + "loss": 4.0938, + "step": 11164 + }, + { + "epoch": 0.11165, + "grad_norm": 1.1537056578577085, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 11165 + }, + { + "epoch": 0.11166, + "grad_norm": 1.205755564694277, + "learning_rate": 0.003, + "loss": 4.0961, + "step": 11166 + }, + { + "epoch": 0.11167, + "grad_norm": 0.9597401281950495, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 11167 + }, + { + "epoch": 0.11168, + "grad_norm": 1.0093287521717043, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 11168 + }, + { + "epoch": 0.11169, + "grad_norm": 1.0370057524742466, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 11169 + }, + { + "epoch": 0.1117, + "grad_norm": 0.9659616133974819, + "learning_rate": 0.003, + "loss": 4.0891, + "step": 11170 + }, + { + "epoch": 0.11171, + "grad_norm": 0.8141280923210972, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 11171 + }, + { + "epoch": 0.11172, + "grad_norm": 0.701815654364921, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 11172 + }, + { + "epoch": 0.11173, + "grad_norm": 0.8490439959715378, + "learning_rate": 0.003, + "loss": 4.089, + "step": 11173 + }, + { + "epoch": 0.11174, + "grad_norm": 1.0310146465685464, + "learning_rate": 0.003, + "loss": 4.0958, + "step": 11174 + }, + { + "epoch": 0.11175, + "grad_norm": 1.0831182574442229, + "learning_rate": 0.003, + "loss": 4.0995, + "step": 11175 + }, + { + "epoch": 0.11176, + "grad_norm": 0.876956894639053, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 11176 + }, + { + "epoch": 0.11177, + "grad_norm": 0.9504101923631635, + "learning_rate": 0.003, + "loss": 4.1066, + "step": 11177 + }, + { + "epoch": 0.11178, + "grad_norm": 1.0818154179432284, + "learning_rate": 0.003, + "loss": 4.1009, + "step": 11178 + }, + { + "epoch": 0.11179, + "grad_norm": 1.075686129918887, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 11179 + }, + { + "epoch": 0.1118, + "grad_norm": 1.036389969747734, + "learning_rate": 0.003, + "loss": 4.093, + "step": 11180 + }, + { + "epoch": 0.11181, + "grad_norm": 1.01691383276777, + "learning_rate": 0.003, + "loss": 4.0972, + "step": 11181 + }, + { + "epoch": 0.11182, + "grad_norm": 1.0202530279280912, + "learning_rate": 0.003, + "loss": 4.0854, + "step": 11182 + }, + { + "epoch": 0.11183, + "grad_norm": 0.8439854248177495, + "learning_rate": 0.003, + "loss": 4.0899, + "step": 11183 + }, + { + "epoch": 0.11184, + "grad_norm": 0.7958930318402421, + "learning_rate": 0.003, + "loss": 4.0848, + "step": 11184 + }, + { + "epoch": 0.11185, + "grad_norm": 0.818000632842024, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 11185 + }, + { + "epoch": 0.11186, + "grad_norm": 0.9195256337290058, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 11186 + }, + { + "epoch": 0.11187, + "grad_norm": 0.9520595393055282, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 11187 + }, + { + "epoch": 0.11188, + "grad_norm": 1.1990533316635237, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 11188 + }, + { + "epoch": 0.11189, + "grad_norm": 0.9279507267566511, + "learning_rate": 0.003, + "loss": 4.1187, + "step": 11189 + }, + { + "epoch": 0.1119, + "grad_norm": 1.0295022127453026, + "learning_rate": 0.003, + "loss": 4.1159, + "step": 11190 + }, + { + "epoch": 0.11191, + "grad_norm": 1.2918720023903827, + "learning_rate": 0.003, + "loss": 4.0884, + "step": 11191 + }, + { + "epoch": 0.11192, + "grad_norm": 0.8620085355272923, + "learning_rate": 0.003, + "loss": 4.117, + "step": 11192 + }, + { + "epoch": 0.11193, + "grad_norm": 0.850503691303309, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 11193 + }, + { + "epoch": 0.11194, + "grad_norm": 0.9842451207661724, + "learning_rate": 0.003, + "loss": 4.1214, + "step": 11194 + }, + { + "epoch": 0.11195, + "grad_norm": 1.1435391410059392, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 11195 + }, + { + "epoch": 0.11196, + "grad_norm": 0.9661746808057262, + "learning_rate": 0.003, + "loss": 4.086, + "step": 11196 + }, + { + "epoch": 0.11197, + "grad_norm": 0.8900387842254533, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 11197 + }, + { + "epoch": 0.11198, + "grad_norm": 0.9616682457603061, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 11198 + }, + { + "epoch": 0.11199, + "grad_norm": 1.0554999824468345, + "learning_rate": 0.003, + "loss": 4.0953, + "step": 11199 + }, + { + "epoch": 0.112, + "grad_norm": 0.9818338289171278, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 11200 + }, + { + "epoch": 0.11201, + "grad_norm": 1.0161821512255864, + "learning_rate": 0.003, + "loss": 4.0922, + "step": 11201 + }, + { + "epoch": 0.11202, + "grad_norm": 0.9359170112295755, + "learning_rate": 0.003, + "loss": 4.097, + "step": 11202 + }, + { + "epoch": 0.11203, + "grad_norm": 0.9823011417476806, + "learning_rate": 0.003, + "loss": 4.0872, + "step": 11203 + }, + { + "epoch": 0.11204, + "grad_norm": 1.181337339377003, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 11204 + }, + { + "epoch": 0.11205, + "grad_norm": 1.0053284348661442, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 11205 + }, + { + "epoch": 0.11206, + "grad_norm": 0.9091754223757886, + "learning_rate": 0.003, + "loss": 4.1045, + "step": 11206 + }, + { + "epoch": 0.11207, + "grad_norm": 0.7939160685325812, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 11207 + }, + { + "epoch": 0.11208, + "grad_norm": 0.8794809243816446, + "learning_rate": 0.003, + "loss": 4.0946, + "step": 11208 + }, + { + "epoch": 0.11209, + "grad_norm": 1.0333215572020131, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 11209 + }, + { + "epoch": 0.1121, + "grad_norm": 1.1994810730399934, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 11210 + }, + { + "epoch": 0.11211, + "grad_norm": 0.7883069246746688, + "learning_rate": 0.003, + "loss": 4.1022, + "step": 11211 + }, + { + "epoch": 0.11212, + "grad_norm": 0.7543570572393807, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 11212 + }, + { + "epoch": 0.11213, + "grad_norm": 0.8481686867054684, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 11213 + }, + { + "epoch": 0.11214, + "grad_norm": 0.9924534739189692, + "learning_rate": 0.003, + "loss": 4.1149, + "step": 11214 + }, + { + "epoch": 0.11215, + "grad_norm": 1.125249168991632, + "learning_rate": 0.003, + "loss": 4.1189, + "step": 11215 + }, + { + "epoch": 0.11216, + "grad_norm": 0.8808051402234568, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 11216 + }, + { + "epoch": 0.11217, + "grad_norm": 0.8593832132500224, + "learning_rate": 0.003, + "loss": 4.086, + "step": 11217 + }, + { + "epoch": 0.11218, + "grad_norm": 0.8656083994690128, + "learning_rate": 0.003, + "loss": 4.0871, + "step": 11218 + }, + { + "epoch": 0.11219, + "grad_norm": 0.8905514752993446, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 11219 + }, + { + "epoch": 0.1122, + "grad_norm": 0.9715848000848435, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 11220 + }, + { + "epoch": 0.11221, + "grad_norm": 1.1914373039489818, + "learning_rate": 0.003, + "loss": 4.0871, + "step": 11221 + }, + { + "epoch": 0.11222, + "grad_norm": 0.8530264998775443, + "learning_rate": 0.003, + "loss": 4.0986, + "step": 11222 + }, + { + "epoch": 0.11223, + "grad_norm": 0.7347545304595516, + "learning_rate": 0.003, + "loss": 4.067, + "step": 11223 + }, + { + "epoch": 0.11224, + "grad_norm": 0.750705211858614, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 11224 + }, + { + "epoch": 0.11225, + "grad_norm": 0.8867525754250005, + "learning_rate": 0.003, + "loss": 4.1034, + "step": 11225 + }, + { + "epoch": 0.11226, + "grad_norm": 0.9374517910999828, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 11226 + }, + { + "epoch": 0.11227, + "grad_norm": 0.9607613637427005, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 11227 + }, + { + "epoch": 0.11228, + "grad_norm": 0.9643436641157891, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 11228 + }, + { + "epoch": 0.11229, + "grad_norm": 1.0115729568689629, + "learning_rate": 0.003, + "loss": 4.0917, + "step": 11229 + }, + { + "epoch": 0.1123, + "grad_norm": 1.0408717793318836, + "learning_rate": 0.003, + "loss": 4.0919, + "step": 11230 + }, + { + "epoch": 0.11231, + "grad_norm": 1.2489296722268408, + "learning_rate": 0.003, + "loss": 4.0961, + "step": 11231 + }, + { + "epoch": 0.11232, + "grad_norm": 0.8673853435344643, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 11232 + }, + { + "epoch": 0.11233, + "grad_norm": 0.8802782564432414, + "learning_rate": 0.003, + "loss": 4.0998, + "step": 11233 + }, + { + "epoch": 0.11234, + "grad_norm": 1.0222286845710342, + "learning_rate": 0.003, + "loss": 4.0902, + "step": 11234 + }, + { + "epoch": 0.11235, + "grad_norm": 0.9577007436286179, + "learning_rate": 0.003, + "loss": 4.0998, + "step": 11235 + }, + { + "epoch": 0.11236, + "grad_norm": 0.9259032049990754, + "learning_rate": 0.003, + "loss": 4.1107, + "step": 11236 + }, + { + "epoch": 0.11237, + "grad_norm": 1.1147536011785937, + "learning_rate": 0.003, + "loss": 4.103, + "step": 11237 + }, + { + "epoch": 0.11238, + "grad_norm": 1.0452217935723533, + "learning_rate": 0.003, + "loss": 4.0819, + "step": 11238 + }, + { + "epoch": 0.11239, + "grad_norm": 1.0457120522396022, + "learning_rate": 0.003, + "loss": 4.1057, + "step": 11239 + }, + { + "epoch": 0.1124, + "grad_norm": 0.9025136454149996, + "learning_rate": 0.003, + "loss": 4.0972, + "step": 11240 + }, + { + "epoch": 0.11241, + "grad_norm": 0.9584513544083916, + "learning_rate": 0.003, + "loss": 4.1008, + "step": 11241 + }, + { + "epoch": 0.11242, + "grad_norm": 1.2685967758152246, + "learning_rate": 0.003, + "loss": 4.1025, + "step": 11242 + }, + { + "epoch": 0.11243, + "grad_norm": 1.0776059828641649, + "learning_rate": 0.003, + "loss": 4.0996, + "step": 11243 + }, + { + "epoch": 0.11244, + "grad_norm": 0.9911185705164456, + "learning_rate": 0.003, + "loss": 4.086, + "step": 11244 + }, + { + "epoch": 0.11245, + "grad_norm": 0.9714354847877079, + "learning_rate": 0.003, + "loss": 4.1074, + "step": 11245 + }, + { + "epoch": 0.11246, + "grad_norm": 0.9372015201404659, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 11246 + }, + { + "epoch": 0.11247, + "grad_norm": 1.0163869573111552, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 11247 + }, + { + "epoch": 0.11248, + "grad_norm": 1.130476125649323, + "learning_rate": 0.003, + "loss": 4.1136, + "step": 11248 + }, + { + "epoch": 0.11249, + "grad_norm": 0.8328062348317273, + "learning_rate": 0.003, + "loss": 4.0918, + "step": 11249 + }, + { + "epoch": 0.1125, + "grad_norm": 0.7594154975093197, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 11250 + }, + { + "epoch": 0.11251, + "grad_norm": 0.6432880572255884, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 11251 + }, + { + "epoch": 0.11252, + "grad_norm": 0.5801556104879025, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 11252 + }, + { + "epoch": 0.11253, + "grad_norm": 0.5898763961934012, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 11253 + }, + { + "epoch": 0.11254, + "grad_norm": 0.661646277785929, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 11254 + }, + { + "epoch": 0.11255, + "grad_norm": 0.681730134691339, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 11255 + }, + { + "epoch": 0.11256, + "grad_norm": 0.8231289545451754, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 11256 + }, + { + "epoch": 0.11257, + "grad_norm": 1.1003877906640407, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 11257 + }, + { + "epoch": 0.11258, + "grad_norm": 1.066142746207035, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 11258 + }, + { + "epoch": 0.11259, + "grad_norm": 0.8410469765772541, + "learning_rate": 0.003, + "loss": 4.1091, + "step": 11259 + }, + { + "epoch": 0.1126, + "grad_norm": 0.7300542841806832, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 11260 + }, + { + "epoch": 0.11261, + "grad_norm": 0.6920098913160495, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 11261 + }, + { + "epoch": 0.11262, + "grad_norm": 0.8167238016291801, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 11262 + }, + { + "epoch": 0.11263, + "grad_norm": 0.9448198862049916, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 11263 + }, + { + "epoch": 0.11264, + "grad_norm": 1.0544394171734364, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 11264 + }, + { + "epoch": 0.11265, + "grad_norm": 0.9150889249129266, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 11265 + }, + { + "epoch": 0.11266, + "grad_norm": 0.9237953331218364, + "learning_rate": 0.003, + "loss": 4.0966, + "step": 11266 + }, + { + "epoch": 0.11267, + "grad_norm": 1.0050453550872565, + "learning_rate": 0.003, + "loss": 4.1067, + "step": 11267 + }, + { + "epoch": 0.11268, + "grad_norm": 1.2859785985527485, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 11268 + }, + { + "epoch": 0.11269, + "grad_norm": 0.9122590715356339, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 11269 + }, + { + "epoch": 0.1127, + "grad_norm": 0.9558042330502723, + "learning_rate": 0.003, + "loss": 4.0858, + "step": 11270 + }, + { + "epoch": 0.11271, + "grad_norm": 1.0527887872500457, + "learning_rate": 0.003, + "loss": 4.0939, + "step": 11271 + }, + { + "epoch": 0.11272, + "grad_norm": 1.0070978718364427, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 11272 + }, + { + "epoch": 0.11273, + "grad_norm": 1.0277622452587913, + "learning_rate": 0.003, + "loss": 4.1028, + "step": 11273 + }, + { + "epoch": 0.11274, + "grad_norm": 1.1885703695652448, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 11274 + }, + { + "epoch": 0.11275, + "grad_norm": 0.9334884678780844, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 11275 + }, + { + "epoch": 0.11276, + "grad_norm": 1.1095264751425142, + "learning_rate": 0.003, + "loss": 4.1021, + "step": 11276 + }, + { + "epoch": 0.11277, + "grad_norm": 1.015069743997468, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 11277 + }, + { + "epoch": 0.11278, + "grad_norm": 1.0865072757256489, + "learning_rate": 0.003, + "loss": 4.1104, + "step": 11278 + }, + { + "epoch": 0.11279, + "grad_norm": 1.0427625043821893, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 11279 + }, + { + "epoch": 0.1128, + "grad_norm": 1.0366036232683093, + "learning_rate": 0.003, + "loss": 4.0937, + "step": 11280 + }, + { + "epoch": 0.11281, + "grad_norm": 1.2332435995820707, + "learning_rate": 0.003, + "loss": 4.0963, + "step": 11281 + }, + { + "epoch": 0.11282, + "grad_norm": 0.8781237713469318, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 11282 + }, + { + "epoch": 0.11283, + "grad_norm": 0.7113596057521735, + "learning_rate": 0.003, + "loss": 4.0871, + "step": 11283 + }, + { + "epoch": 0.11284, + "grad_norm": 0.6629655362864838, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 11284 + }, + { + "epoch": 0.11285, + "grad_norm": 0.7312431220312264, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 11285 + }, + { + "epoch": 0.11286, + "grad_norm": 0.8614641565188512, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 11286 + }, + { + "epoch": 0.11287, + "grad_norm": 1.0720243221454566, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 11287 + }, + { + "epoch": 0.11288, + "grad_norm": 1.0768498447554693, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 11288 + }, + { + "epoch": 0.11289, + "grad_norm": 0.9562956504171958, + "learning_rate": 0.003, + "loss": 4.1035, + "step": 11289 + }, + { + "epoch": 0.1129, + "grad_norm": 0.8975837040285087, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 11290 + }, + { + "epoch": 0.11291, + "grad_norm": 0.8772329733044721, + "learning_rate": 0.003, + "loss": 4.1031, + "step": 11291 + }, + { + "epoch": 0.11292, + "grad_norm": 0.9237793432668792, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 11292 + }, + { + "epoch": 0.11293, + "grad_norm": 0.9544936281980384, + "learning_rate": 0.003, + "loss": 4.1134, + "step": 11293 + }, + { + "epoch": 0.11294, + "grad_norm": 0.9980191517632092, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 11294 + }, + { + "epoch": 0.11295, + "grad_norm": 0.9946280646980549, + "learning_rate": 0.003, + "loss": 4.1009, + "step": 11295 + }, + { + "epoch": 0.11296, + "grad_norm": 1.0166999232856886, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 11296 + }, + { + "epoch": 0.11297, + "grad_norm": 1.252901644843653, + "learning_rate": 0.003, + "loss": 4.1194, + "step": 11297 + }, + { + "epoch": 0.11298, + "grad_norm": 1.2542159971053828, + "learning_rate": 0.003, + "loss": 4.1011, + "step": 11298 + }, + { + "epoch": 0.11299, + "grad_norm": 0.8969353560535341, + "learning_rate": 0.003, + "loss": 4.0915, + "step": 11299 + }, + { + "epoch": 0.113, + "grad_norm": 0.8019689741939366, + "learning_rate": 0.003, + "loss": 4.1079, + "step": 11300 + }, + { + "epoch": 0.11301, + "grad_norm": 0.9187654560832896, + "learning_rate": 0.003, + "loss": 4.074, + "step": 11301 + }, + { + "epoch": 0.11302, + "grad_norm": 1.0738786682441583, + "learning_rate": 0.003, + "loss": 4.0856, + "step": 11302 + }, + { + "epoch": 0.11303, + "grad_norm": 1.2720466168291542, + "learning_rate": 0.003, + "loss": 4.099, + "step": 11303 + }, + { + "epoch": 0.11304, + "grad_norm": 0.9247862095757325, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 11304 + }, + { + "epoch": 0.11305, + "grad_norm": 0.8804893034135447, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 11305 + }, + { + "epoch": 0.11306, + "grad_norm": 0.9297122788749244, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 11306 + }, + { + "epoch": 0.11307, + "grad_norm": 0.9141518584049458, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 11307 + }, + { + "epoch": 0.11308, + "grad_norm": 0.9459673401878037, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 11308 + }, + { + "epoch": 0.11309, + "grad_norm": 0.7938031288765369, + "learning_rate": 0.003, + "loss": 4.0928, + "step": 11309 + }, + { + "epoch": 0.1131, + "grad_norm": 0.8600966414484259, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 11310 + }, + { + "epoch": 0.11311, + "grad_norm": 1.131992506472778, + "learning_rate": 0.003, + "loss": 4.0975, + "step": 11311 + }, + { + "epoch": 0.11312, + "grad_norm": 1.0998667941262545, + "learning_rate": 0.003, + "loss": 4.093, + "step": 11312 + }, + { + "epoch": 0.11313, + "grad_norm": 1.1546813775010696, + "learning_rate": 0.003, + "loss": 4.073, + "step": 11313 + }, + { + "epoch": 0.11314, + "grad_norm": 0.9564661153179174, + "learning_rate": 0.003, + "loss": 4.1003, + "step": 11314 + }, + { + "epoch": 0.11315, + "grad_norm": 0.9235701690333917, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 11315 + }, + { + "epoch": 0.11316, + "grad_norm": 0.8171681251937607, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 11316 + }, + { + "epoch": 0.11317, + "grad_norm": 0.7980292698558737, + "learning_rate": 0.003, + "loss": 4.0898, + "step": 11317 + }, + { + "epoch": 0.11318, + "grad_norm": 0.8049828479399856, + "learning_rate": 0.003, + "loss": 4.0934, + "step": 11318 + }, + { + "epoch": 0.11319, + "grad_norm": 0.821945912749478, + "learning_rate": 0.003, + "loss": 4.0925, + "step": 11319 + }, + { + "epoch": 0.1132, + "grad_norm": 0.8828989398679872, + "learning_rate": 0.003, + "loss": 4.0858, + "step": 11320 + }, + { + "epoch": 0.11321, + "grad_norm": 1.1131728681770399, + "learning_rate": 0.003, + "loss": 4.056, + "step": 11321 + }, + { + "epoch": 0.11322, + "grad_norm": 1.0686405500235516, + "learning_rate": 0.003, + "loss": 4.0898, + "step": 11322 + }, + { + "epoch": 0.11323, + "grad_norm": 0.8382093656335687, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 11323 + }, + { + "epoch": 0.11324, + "grad_norm": 0.6905316802691976, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 11324 + }, + { + "epoch": 0.11325, + "grad_norm": 0.7090986408673777, + "learning_rate": 0.003, + "loss": 4.0927, + "step": 11325 + }, + { + "epoch": 0.11326, + "grad_norm": 0.6530856366159342, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 11326 + }, + { + "epoch": 0.11327, + "grad_norm": 0.8149030513693352, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 11327 + }, + { + "epoch": 0.11328, + "grad_norm": 0.9751843006370889, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 11328 + }, + { + "epoch": 0.11329, + "grad_norm": 1.2899448221196481, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 11329 + }, + { + "epoch": 0.1133, + "grad_norm": 0.6815142182372425, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 11330 + }, + { + "epoch": 0.11331, + "grad_norm": 0.6849324012109173, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 11331 + }, + { + "epoch": 0.11332, + "grad_norm": 0.8522810918212895, + "learning_rate": 0.003, + "loss": 4.0938, + "step": 11332 + }, + { + "epoch": 0.11333, + "grad_norm": 1.088066908302685, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 11333 + }, + { + "epoch": 0.11334, + "grad_norm": 1.0671323778572612, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 11334 + }, + { + "epoch": 0.11335, + "grad_norm": 0.9177305377276991, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 11335 + }, + { + "epoch": 0.11336, + "grad_norm": 0.8736559199944975, + "learning_rate": 0.003, + "loss": 4.06, + "step": 11336 + }, + { + "epoch": 0.11337, + "grad_norm": 0.8659670925847458, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 11337 + }, + { + "epoch": 0.11338, + "grad_norm": 1.0089898442936824, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 11338 + }, + { + "epoch": 0.11339, + "grad_norm": 1.1949425818273396, + "learning_rate": 0.003, + "loss": 4.1243, + "step": 11339 + }, + { + "epoch": 0.1134, + "grad_norm": 1.0240293391997215, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 11340 + }, + { + "epoch": 0.11341, + "grad_norm": 0.9319549261075856, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 11341 + }, + { + "epoch": 0.11342, + "grad_norm": 0.8066652500811496, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 11342 + }, + { + "epoch": 0.11343, + "grad_norm": 0.7764478238281299, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 11343 + }, + { + "epoch": 0.11344, + "grad_norm": 0.7516072792569712, + "learning_rate": 0.003, + "loss": 4.0947, + "step": 11344 + }, + { + "epoch": 0.11345, + "grad_norm": 0.7597631588688307, + "learning_rate": 0.003, + "loss": 4.1007, + "step": 11345 + }, + { + "epoch": 0.11346, + "grad_norm": 0.8168664630486651, + "learning_rate": 0.003, + "loss": 4.0878, + "step": 11346 + }, + { + "epoch": 0.11347, + "grad_norm": 0.839845032340911, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 11347 + }, + { + "epoch": 0.11348, + "grad_norm": 0.8918825023781319, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 11348 + }, + { + "epoch": 0.11349, + "grad_norm": 0.9782875464246246, + "learning_rate": 0.003, + "loss": 4.0918, + "step": 11349 + }, + { + "epoch": 0.1135, + "grad_norm": 1.1481421199909936, + "learning_rate": 0.003, + "loss": 4.1161, + "step": 11350 + }, + { + "epoch": 0.11351, + "grad_norm": 0.8538312029474874, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 11351 + }, + { + "epoch": 0.11352, + "grad_norm": 0.9029518922349964, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 11352 + }, + { + "epoch": 0.11353, + "grad_norm": 1.1541970567038795, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 11353 + }, + { + "epoch": 0.11354, + "grad_norm": 1.0361094346281055, + "learning_rate": 0.003, + "loss": 4.0854, + "step": 11354 + }, + { + "epoch": 0.11355, + "grad_norm": 0.969803899227432, + "learning_rate": 0.003, + "loss": 4.1251, + "step": 11355 + }, + { + "epoch": 0.11356, + "grad_norm": 1.0666699843328562, + "learning_rate": 0.003, + "loss": 4.0952, + "step": 11356 + }, + { + "epoch": 0.11357, + "grad_norm": 1.1206954373787135, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 11357 + }, + { + "epoch": 0.11358, + "grad_norm": 1.1251672627121379, + "learning_rate": 0.003, + "loss": 4.0955, + "step": 11358 + }, + { + "epoch": 0.11359, + "grad_norm": 1.339618229076221, + "learning_rate": 0.003, + "loss": 4.1217, + "step": 11359 + }, + { + "epoch": 0.1136, + "grad_norm": 0.8436650872178116, + "learning_rate": 0.003, + "loss": 4.0946, + "step": 11360 + }, + { + "epoch": 0.11361, + "grad_norm": 0.816578825245825, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 11361 + }, + { + "epoch": 0.11362, + "grad_norm": 0.9699465302740039, + "learning_rate": 0.003, + "loss": 4.0933, + "step": 11362 + }, + { + "epoch": 0.11363, + "grad_norm": 1.1700911074277514, + "learning_rate": 0.003, + "loss": 4.0874, + "step": 11363 + }, + { + "epoch": 0.11364, + "grad_norm": 0.9645715422091756, + "learning_rate": 0.003, + "loss": 4.1055, + "step": 11364 + }, + { + "epoch": 0.11365, + "grad_norm": 1.1033697417791446, + "learning_rate": 0.003, + "loss": 4.092, + "step": 11365 + }, + { + "epoch": 0.11366, + "grad_norm": 1.1499839584912717, + "learning_rate": 0.003, + "loss": 4.091, + "step": 11366 + }, + { + "epoch": 0.11367, + "grad_norm": 1.022281538935548, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 11367 + }, + { + "epoch": 0.11368, + "grad_norm": 0.8574694338027685, + "learning_rate": 0.003, + "loss": 4.1034, + "step": 11368 + }, + { + "epoch": 0.11369, + "grad_norm": 0.7340439204477529, + "learning_rate": 0.003, + "loss": 4.0969, + "step": 11369 + }, + { + "epoch": 0.1137, + "grad_norm": 0.8249840662522727, + "learning_rate": 0.003, + "loss": 4.1081, + "step": 11370 + }, + { + "epoch": 0.11371, + "grad_norm": 0.9384425932880045, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 11371 + }, + { + "epoch": 0.11372, + "grad_norm": 0.8534238410866253, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 11372 + }, + { + "epoch": 0.11373, + "grad_norm": 0.7533718501739974, + "learning_rate": 0.003, + "loss": 4.0893, + "step": 11373 + }, + { + "epoch": 0.11374, + "grad_norm": 0.8375274624061597, + "learning_rate": 0.003, + "loss": 4.0887, + "step": 11374 + }, + { + "epoch": 0.11375, + "grad_norm": 0.9291619536590049, + "learning_rate": 0.003, + "loss": 4.0996, + "step": 11375 + }, + { + "epoch": 0.11376, + "grad_norm": 1.193378603756856, + "learning_rate": 0.003, + "loss": 4.1292, + "step": 11376 + }, + { + "epoch": 0.11377, + "grad_norm": 0.9569155917676204, + "learning_rate": 0.003, + "loss": 4.0887, + "step": 11377 + }, + { + "epoch": 0.11378, + "grad_norm": 0.8384472754269915, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 11378 + }, + { + "epoch": 0.11379, + "grad_norm": 0.8428721966848975, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 11379 + }, + { + "epoch": 0.1138, + "grad_norm": 1.0001736983507425, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 11380 + }, + { + "epoch": 0.11381, + "grad_norm": 1.1757986644881853, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 11381 + }, + { + "epoch": 0.11382, + "grad_norm": 0.9185144508690771, + "learning_rate": 0.003, + "loss": 4.0978, + "step": 11382 + }, + { + "epoch": 0.11383, + "grad_norm": 0.9915830438911467, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 11383 + }, + { + "epoch": 0.11384, + "grad_norm": 1.2155553787844495, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 11384 + }, + { + "epoch": 0.11385, + "grad_norm": 0.8931408549828402, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 11385 + }, + { + "epoch": 0.11386, + "grad_norm": 0.9840076006363396, + "learning_rate": 0.003, + "loss": 4.1031, + "step": 11386 + }, + { + "epoch": 0.11387, + "grad_norm": 1.024771890304262, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 11387 + }, + { + "epoch": 0.11388, + "grad_norm": 0.8906019814686371, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 11388 + }, + { + "epoch": 0.11389, + "grad_norm": 1.0501462857608732, + "learning_rate": 0.003, + "loss": 4.091, + "step": 11389 + }, + { + "epoch": 0.1139, + "grad_norm": 1.1588334538788612, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 11390 + }, + { + "epoch": 0.11391, + "grad_norm": 1.2252505740264634, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 11391 + }, + { + "epoch": 0.11392, + "grad_norm": 0.8540844717049403, + "learning_rate": 0.003, + "loss": 4.0969, + "step": 11392 + }, + { + "epoch": 0.11393, + "grad_norm": 0.8300836818086524, + "learning_rate": 0.003, + "loss": 4.054, + "step": 11393 + }, + { + "epoch": 0.11394, + "grad_norm": 0.7438237419594534, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 11394 + }, + { + "epoch": 0.11395, + "grad_norm": 0.7469026461348584, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 11395 + }, + { + "epoch": 0.11396, + "grad_norm": 0.8165886638611114, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 11396 + }, + { + "epoch": 0.11397, + "grad_norm": 0.8361491409729935, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 11397 + }, + { + "epoch": 0.11398, + "grad_norm": 0.825462307206807, + "learning_rate": 0.003, + "loss": 4.056, + "step": 11398 + }, + { + "epoch": 0.11399, + "grad_norm": 0.9663549960329568, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 11399 + }, + { + "epoch": 0.114, + "grad_norm": 1.085446246406364, + "learning_rate": 0.003, + "loss": 4.1107, + "step": 11400 + }, + { + "epoch": 0.11401, + "grad_norm": 0.8225600810467596, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 11401 + }, + { + "epoch": 0.11402, + "grad_norm": 0.8650788357618279, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 11402 + }, + { + "epoch": 0.11403, + "grad_norm": 0.9239380297018212, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 11403 + }, + { + "epoch": 0.11404, + "grad_norm": 0.9138576659307862, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 11404 + }, + { + "epoch": 0.11405, + "grad_norm": 1.1063060925751604, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 11405 + }, + { + "epoch": 0.11406, + "grad_norm": 0.968280228145328, + "learning_rate": 0.003, + "loss": 4.1029, + "step": 11406 + }, + { + "epoch": 0.11407, + "grad_norm": 0.9538305687565618, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 11407 + }, + { + "epoch": 0.11408, + "grad_norm": 0.7297628056191903, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 11408 + }, + { + "epoch": 0.11409, + "grad_norm": 0.6794541608118937, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 11409 + }, + { + "epoch": 0.1141, + "grad_norm": 0.8461841265618812, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 11410 + }, + { + "epoch": 0.11411, + "grad_norm": 1.118380834510809, + "learning_rate": 0.003, + "loss": 4.0981, + "step": 11411 + }, + { + "epoch": 0.11412, + "grad_norm": 1.2923879811334953, + "learning_rate": 0.003, + "loss": 4.1008, + "step": 11412 + }, + { + "epoch": 0.11413, + "grad_norm": 0.6946740680802306, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 11413 + }, + { + "epoch": 0.11414, + "grad_norm": 0.7430848348099186, + "learning_rate": 0.003, + "loss": 4.0855, + "step": 11414 + }, + { + "epoch": 0.11415, + "grad_norm": 0.7860608148791273, + "learning_rate": 0.003, + "loss": 4.082, + "step": 11415 + }, + { + "epoch": 0.11416, + "grad_norm": 0.8728779172892188, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 11416 + }, + { + "epoch": 0.11417, + "grad_norm": 1.019730576565571, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 11417 + }, + { + "epoch": 0.11418, + "grad_norm": 1.0496436719293236, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 11418 + }, + { + "epoch": 0.11419, + "grad_norm": 1.0324561585061647, + "learning_rate": 0.003, + "loss": 4.061, + "step": 11419 + }, + { + "epoch": 0.1142, + "grad_norm": 1.001424020139074, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 11420 + }, + { + "epoch": 0.11421, + "grad_norm": 0.987857728191177, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 11421 + }, + { + "epoch": 0.11422, + "grad_norm": 1.1071344521032516, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 11422 + }, + { + "epoch": 0.11423, + "grad_norm": 0.9579591906698591, + "learning_rate": 0.003, + "loss": 4.1156, + "step": 11423 + }, + { + "epoch": 0.11424, + "grad_norm": 0.9475056784569227, + "learning_rate": 0.003, + "loss": 4.1067, + "step": 11424 + }, + { + "epoch": 0.11425, + "grad_norm": 1.1603054331304683, + "learning_rate": 0.003, + "loss": 4.0976, + "step": 11425 + }, + { + "epoch": 0.11426, + "grad_norm": 0.8154176089945117, + "learning_rate": 0.003, + "loss": 4.108, + "step": 11426 + }, + { + "epoch": 0.11427, + "grad_norm": 0.765386931556148, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 11427 + }, + { + "epoch": 0.11428, + "grad_norm": 0.7678835907808844, + "learning_rate": 0.003, + "loss": 4.088, + "step": 11428 + }, + { + "epoch": 0.11429, + "grad_norm": 0.7769096202970582, + "learning_rate": 0.003, + "loss": 4.1117, + "step": 11429 + }, + { + "epoch": 0.1143, + "grad_norm": 0.8545025547585546, + "learning_rate": 0.003, + "loss": 4.102, + "step": 11430 + }, + { + "epoch": 0.11431, + "grad_norm": 0.8935962019008151, + "learning_rate": 0.003, + "loss": 4.1039, + "step": 11431 + }, + { + "epoch": 0.11432, + "grad_norm": 0.9802613618327947, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 11432 + }, + { + "epoch": 0.11433, + "grad_norm": 1.1867678311178003, + "learning_rate": 0.003, + "loss": 4.1302, + "step": 11433 + }, + { + "epoch": 0.11434, + "grad_norm": 1.0545194355890246, + "learning_rate": 0.003, + "loss": 4.1193, + "step": 11434 + }, + { + "epoch": 0.11435, + "grad_norm": 0.9506265292961794, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 11435 + }, + { + "epoch": 0.11436, + "grad_norm": 0.9313382350049672, + "learning_rate": 0.003, + "loss": 4.0977, + "step": 11436 + }, + { + "epoch": 0.11437, + "grad_norm": 1.0339764264241789, + "learning_rate": 0.003, + "loss": 4.052, + "step": 11437 + }, + { + "epoch": 0.11438, + "grad_norm": 1.1202442430792667, + "learning_rate": 0.003, + "loss": 4.0936, + "step": 11438 + }, + { + "epoch": 0.11439, + "grad_norm": 0.9876079422602098, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 11439 + }, + { + "epoch": 0.1144, + "grad_norm": 1.1524557203852144, + "learning_rate": 0.003, + "loss": 4.107, + "step": 11440 + }, + { + "epoch": 0.11441, + "grad_norm": 0.9540244457244769, + "learning_rate": 0.003, + "loss": 4.0845, + "step": 11441 + }, + { + "epoch": 0.11442, + "grad_norm": 1.0614163626325366, + "learning_rate": 0.003, + "loss": 4.0905, + "step": 11442 + }, + { + "epoch": 0.11443, + "grad_norm": 1.0122836759112739, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 11443 + }, + { + "epoch": 0.11444, + "grad_norm": 1.133004807403538, + "learning_rate": 0.003, + "loss": 4.079, + "step": 11444 + }, + { + "epoch": 0.11445, + "grad_norm": 0.8950620595833293, + "learning_rate": 0.003, + "loss": 4.1, + "step": 11445 + }, + { + "epoch": 0.11446, + "grad_norm": 0.8796973095109482, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 11446 + }, + { + "epoch": 0.11447, + "grad_norm": 0.9297671855717798, + "learning_rate": 0.003, + "loss": 4.1243, + "step": 11447 + }, + { + "epoch": 0.11448, + "grad_norm": 1.05504119303505, + "learning_rate": 0.003, + "loss": 4.1186, + "step": 11448 + }, + { + "epoch": 0.11449, + "grad_norm": 1.2846502740831718, + "learning_rate": 0.003, + "loss": 4.0898, + "step": 11449 + }, + { + "epoch": 0.1145, + "grad_norm": 0.934340145750696, + "learning_rate": 0.003, + "loss": 4.1032, + "step": 11450 + }, + { + "epoch": 0.11451, + "grad_norm": 1.0288360407601005, + "learning_rate": 0.003, + "loss": 4.0981, + "step": 11451 + }, + { + "epoch": 0.11452, + "grad_norm": 1.2524626367597984, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 11452 + }, + { + "epoch": 0.11453, + "grad_norm": 0.9098476026210716, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 11453 + }, + { + "epoch": 0.11454, + "grad_norm": 0.8256569669994234, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 11454 + }, + { + "epoch": 0.11455, + "grad_norm": 0.8265838429546257, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 11455 + }, + { + "epoch": 0.11456, + "grad_norm": 0.8490140785699342, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 11456 + }, + { + "epoch": 0.11457, + "grad_norm": 0.9451917142614737, + "learning_rate": 0.003, + "loss": 4.071, + "step": 11457 + }, + { + "epoch": 0.11458, + "grad_norm": 0.9770852066234127, + "learning_rate": 0.003, + "loss": 4.099, + "step": 11458 + }, + { + "epoch": 0.11459, + "grad_norm": 1.1766152810737593, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 11459 + }, + { + "epoch": 0.1146, + "grad_norm": 1.012136545008195, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 11460 + }, + { + "epoch": 0.11461, + "grad_norm": 0.9046788223219674, + "learning_rate": 0.003, + "loss": 4.0941, + "step": 11461 + }, + { + "epoch": 0.11462, + "grad_norm": 0.9469036715869649, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 11462 + }, + { + "epoch": 0.11463, + "grad_norm": 0.8692538905281171, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 11463 + }, + { + "epoch": 0.11464, + "grad_norm": 0.9223742888820864, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 11464 + }, + { + "epoch": 0.11465, + "grad_norm": 1.0594334990087941, + "learning_rate": 0.003, + "loss": 4.0944, + "step": 11465 + }, + { + "epoch": 0.11466, + "grad_norm": 1.0764773989522403, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 11466 + }, + { + "epoch": 0.11467, + "grad_norm": 1.003026095915086, + "learning_rate": 0.003, + "loss": 4.1014, + "step": 11467 + }, + { + "epoch": 0.11468, + "grad_norm": 0.9085736027141201, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 11468 + }, + { + "epoch": 0.11469, + "grad_norm": 0.9594923586544988, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 11469 + }, + { + "epoch": 0.1147, + "grad_norm": 0.9684051846051075, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 11470 + }, + { + "epoch": 0.11471, + "grad_norm": 1.179497515156821, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 11471 + }, + { + "epoch": 0.11472, + "grad_norm": 1.0287213450312322, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 11472 + }, + { + "epoch": 0.11473, + "grad_norm": 1.0286647663629818, + "learning_rate": 0.003, + "loss": 4.0873, + "step": 11473 + }, + { + "epoch": 0.11474, + "grad_norm": 1.1779849167486032, + "learning_rate": 0.003, + "loss": 4.0913, + "step": 11474 + }, + { + "epoch": 0.11475, + "grad_norm": 0.7873801679854139, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 11475 + }, + { + "epoch": 0.11476, + "grad_norm": 0.6434928266843207, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 11476 + }, + { + "epoch": 0.11477, + "grad_norm": 0.7631265962468616, + "learning_rate": 0.003, + "loss": 4.1108, + "step": 11477 + }, + { + "epoch": 0.11478, + "grad_norm": 0.8722811108714592, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 11478 + }, + { + "epoch": 0.11479, + "grad_norm": 1.047933802467063, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 11479 + }, + { + "epoch": 0.1148, + "grad_norm": 1.1227581817691066, + "learning_rate": 0.003, + "loss": 4.1124, + "step": 11480 + }, + { + "epoch": 0.11481, + "grad_norm": 0.8718555512794138, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 11481 + }, + { + "epoch": 0.11482, + "grad_norm": 0.8830360419850367, + "learning_rate": 0.003, + "loss": 4.1173, + "step": 11482 + }, + { + "epoch": 0.11483, + "grad_norm": 0.8780163619900201, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 11483 + }, + { + "epoch": 0.11484, + "grad_norm": 0.8512093771436473, + "learning_rate": 0.003, + "loss": 4.0854, + "step": 11484 + }, + { + "epoch": 0.11485, + "grad_norm": 0.9316835410853554, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 11485 + }, + { + "epoch": 0.11486, + "grad_norm": 0.8859852875702744, + "learning_rate": 0.003, + "loss": 4.0949, + "step": 11486 + }, + { + "epoch": 0.11487, + "grad_norm": 0.8436891654343818, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 11487 + }, + { + "epoch": 0.11488, + "grad_norm": 0.8522795800576273, + "learning_rate": 0.003, + "loss": 4.0974, + "step": 11488 + }, + { + "epoch": 0.11489, + "grad_norm": 1.0176237416407627, + "learning_rate": 0.003, + "loss": 4.0948, + "step": 11489 + }, + { + "epoch": 0.1149, + "grad_norm": 1.0896024394783417, + "learning_rate": 0.003, + "loss": 4.097, + "step": 11490 + }, + { + "epoch": 0.11491, + "grad_norm": 1.0475759853786468, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 11491 + }, + { + "epoch": 0.11492, + "grad_norm": 1.0842594829523204, + "learning_rate": 0.003, + "loss": 4.0955, + "step": 11492 + }, + { + "epoch": 0.11493, + "grad_norm": 1.0823064632938242, + "learning_rate": 0.003, + "loss": 4.0919, + "step": 11493 + }, + { + "epoch": 0.11494, + "grad_norm": 0.9039198907129657, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 11494 + }, + { + "epoch": 0.11495, + "grad_norm": 0.8708725248769316, + "learning_rate": 0.003, + "loss": 4.0864, + "step": 11495 + }, + { + "epoch": 0.11496, + "grad_norm": 0.8114445499649313, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 11496 + }, + { + "epoch": 0.11497, + "grad_norm": 0.7888974760721567, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 11497 + }, + { + "epoch": 0.11498, + "grad_norm": 0.8457299783842269, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 11498 + }, + { + "epoch": 0.11499, + "grad_norm": 1.0197694534790769, + "learning_rate": 0.003, + "loss": 4.0864, + "step": 11499 + }, + { + "epoch": 0.115, + "grad_norm": 1.1657118803273694, + "learning_rate": 0.003, + "loss": 4.111, + "step": 11500 + }, + { + "epoch": 0.11501, + "grad_norm": 0.6549729788620798, + "learning_rate": 0.003, + "loss": 4.1001, + "step": 11501 + }, + { + "epoch": 0.11502, + "grad_norm": 0.6535280539427838, + "learning_rate": 0.003, + "loss": 4.0927, + "step": 11502 + }, + { + "epoch": 0.11503, + "grad_norm": 0.7478946173387384, + "learning_rate": 0.003, + "loss": 4.0878, + "step": 11503 + }, + { + "epoch": 0.11504, + "grad_norm": 0.8720503742113124, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 11504 + }, + { + "epoch": 0.11505, + "grad_norm": 1.1622348817457944, + "learning_rate": 0.003, + "loss": 4.083, + "step": 11505 + }, + { + "epoch": 0.11506, + "grad_norm": 1.0079561786662994, + "learning_rate": 0.003, + "loss": 4.1074, + "step": 11506 + }, + { + "epoch": 0.11507, + "grad_norm": 0.9558534852131281, + "learning_rate": 0.003, + "loss": 4.0996, + "step": 11507 + }, + { + "epoch": 0.11508, + "grad_norm": 0.9553584811601563, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 11508 + }, + { + "epoch": 0.11509, + "grad_norm": 0.9312107206336283, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 11509 + }, + { + "epoch": 0.1151, + "grad_norm": 0.9043498632427543, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 11510 + }, + { + "epoch": 0.11511, + "grad_norm": 0.8905550956708166, + "learning_rate": 0.003, + "loss": 4.057, + "step": 11511 + }, + { + "epoch": 0.11512, + "grad_norm": 1.0253529056295685, + "learning_rate": 0.003, + "loss": 4.066, + "step": 11512 + }, + { + "epoch": 0.11513, + "grad_norm": 1.2433403036348158, + "learning_rate": 0.003, + "loss": 4.0977, + "step": 11513 + }, + { + "epoch": 0.11514, + "grad_norm": 0.7961056605153084, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 11514 + }, + { + "epoch": 0.11515, + "grad_norm": 0.703644445537716, + "learning_rate": 0.003, + "loss": 4.1128, + "step": 11515 + }, + { + "epoch": 0.11516, + "grad_norm": 0.7474407757129923, + "learning_rate": 0.003, + "loss": 4.0965, + "step": 11516 + }, + { + "epoch": 0.11517, + "grad_norm": 0.7576090295161401, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 11517 + }, + { + "epoch": 0.11518, + "grad_norm": 0.8501610807245347, + "learning_rate": 0.003, + "loss": 4.1051, + "step": 11518 + }, + { + "epoch": 0.11519, + "grad_norm": 0.8814875943793059, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 11519 + }, + { + "epoch": 0.1152, + "grad_norm": 0.9191749870935285, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 11520 + }, + { + "epoch": 0.11521, + "grad_norm": 1.0152365326359403, + "learning_rate": 0.003, + "loss": 4.079, + "step": 11521 + }, + { + "epoch": 0.11522, + "grad_norm": 1.0480897680305556, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 11522 + }, + { + "epoch": 0.11523, + "grad_norm": 0.9201974878334627, + "learning_rate": 0.003, + "loss": 4.0874, + "step": 11523 + }, + { + "epoch": 0.11524, + "grad_norm": 1.0175122134322983, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 11524 + }, + { + "epoch": 0.11525, + "grad_norm": 1.071275225866126, + "learning_rate": 0.003, + "loss": 4.1049, + "step": 11525 + }, + { + "epoch": 0.11526, + "grad_norm": 0.9470829240829498, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 11526 + }, + { + "epoch": 0.11527, + "grad_norm": 0.9612424261574787, + "learning_rate": 0.003, + "loss": 4.0965, + "step": 11527 + }, + { + "epoch": 0.11528, + "grad_norm": 0.8507918072523917, + "learning_rate": 0.003, + "loss": 4.069, + "step": 11528 + }, + { + "epoch": 0.11529, + "grad_norm": 0.880257056591339, + "learning_rate": 0.003, + "loss": 4.0977, + "step": 11529 + }, + { + "epoch": 0.1153, + "grad_norm": 1.051614222843063, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 11530 + }, + { + "epoch": 0.11531, + "grad_norm": 1.1048889775654231, + "learning_rate": 0.003, + "loss": 4.1014, + "step": 11531 + }, + { + "epoch": 0.11532, + "grad_norm": 1.1574452989833892, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 11532 + }, + { + "epoch": 0.11533, + "grad_norm": 0.8166874396190745, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 11533 + }, + { + "epoch": 0.11534, + "grad_norm": 0.7202513129345395, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 11534 + }, + { + "epoch": 0.11535, + "grad_norm": 0.8502365037608013, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 11535 + }, + { + "epoch": 0.11536, + "grad_norm": 1.0393220529485125, + "learning_rate": 0.003, + "loss": 4.079, + "step": 11536 + }, + { + "epoch": 0.11537, + "grad_norm": 0.9509490996908745, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 11537 + }, + { + "epoch": 0.11538, + "grad_norm": 1.0603356950221907, + "learning_rate": 0.003, + "loss": 4.1102, + "step": 11538 + }, + { + "epoch": 0.11539, + "grad_norm": 1.0191480904904706, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 11539 + }, + { + "epoch": 0.1154, + "grad_norm": 1.009810186098952, + "learning_rate": 0.003, + "loss": 4.1101, + "step": 11540 + }, + { + "epoch": 0.11541, + "grad_norm": 0.9239715608978077, + "learning_rate": 0.003, + "loss": 4.1053, + "step": 11541 + }, + { + "epoch": 0.11542, + "grad_norm": 0.9827307337828401, + "learning_rate": 0.003, + "loss": 4.1116, + "step": 11542 + }, + { + "epoch": 0.11543, + "grad_norm": 0.8677126283587316, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 11543 + }, + { + "epoch": 0.11544, + "grad_norm": 1.026534024757355, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 11544 + }, + { + "epoch": 0.11545, + "grad_norm": 1.080616525341122, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 11545 + }, + { + "epoch": 0.11546, + "grad_norm": 1.0479462801145076, + "learning_rate": 0.003, + "loss": 4.1052, + "step": 11546 + }, + { + "epoch": 0.11547, + "grad_norm": 0.9394979815350923, + "learning_rate": 0.003, + "loss": 4.074, + "step": 11547 + }, + { + "epoch": 0.11548, + "grad_norm": 1.1582570817386242, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 11548 + }, + { + "epoch": 0.11549, + "grad_norm": 1.0409310187827758, + "learning_rate": 0.003, + "loss": 4.1027, + "step": 11549 + }, + { + "epoch": 0.1155, + "grad_norm": 1.345642209163468, + "learning_rate": 0.003, + "loss": 4.083, + "step": 11550 + }, + { + "epoch": 0.11551, + "grad_norm": 1.0702365640638092, + "learning_rate": 0.003, + "loss": 4.0973, + "step": 11551 + }, + { + "epoch": 0.11552, + "grad_norm": 1.0989237866075894, + "learning_rate": 0.003, + "loss": 4.0949, + "step": 11552 + }, + { + "epoch": 0.11553, + "grad_norm": 1.052216610676183, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 11553 + }, + { + "epoch": 0.11554, + "grad_norm": 1.0182733114131393, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 11554 + }, + { + "epoch": 0.11555, + "grad_norm": 1.0621004302103396, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 11555 + }, + { + "epoch": 0.11556, + "grad_norm": 1.0789236838349332, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 11556 + }, + { + "epoch": 0.11557, + "grad_norm": 1.0391799206136105, + "learning_rate": 0.003, + "loss": 4.085, + "step": 11557 + }, + { + "epoch": 0.11558, + "grad_norm": 1.0543315623939544, + "learning_rate": 0.003, + "loss": 4.0936, + "step": 11558 + }, + { + "epoch": 0.11559, + "grad_norm": 0.9348539777061666, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 11559 + }, + { + "epoch": 0.1156, + "grad_norm": 0.8339354722449975, + "learning_rate": 0.003, + "loss": 4.109, + "step": 11560 + }, + { + "epoch": 0.11561, + "grad_norm": 0.8935953294560726, + "learning_rate": 0.003, + "loss": 4.0868, + "step": 11561 + }, + { + "epoch": 0.11562, + "grad_norm": 1.064490500773631, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 11562 + }, + { + "epoch": 0.11563, + "grad_norm": 0.9435095146945477, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 11563 + }, + { + "epoch": 0.11564, + "grad_norm": 1.1486956161503288, + "learning_rate": 0.003, + "loss": 4.0854, + "step": 11564 + }, + { + "epoch": 0.11565, + "grad_norm": 0.8401893866818503, + "learning_rate": 0.003, + "loss": 4.0934, + "step": 11565 + }, + { + "epoch": 0.11566, + "grad_norm": 0.9255708515196747, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 11566 + }, + { + "epoch": 0.11567, + "grad_norm": 0.9918224386862184, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 11567 + }, + { + "epoch": 0.11568, + "grad_norm": 1.0540696756537682, + "learning_rate": 0.003, + "loss": 4.0848, + "step": 11568 + }, + { + "epoch": 0.11569, + "grad_norm": 0.8085046113499036, + "learning_rate": 0.003, + "loss": 4.0845, + "step": 11569 + }, + { + "epoch": 0.1157, + "grad_norm": 0.7807593914411979, + "learning_rate": 0.003, + "loss": 4.043, + "step": 11570 + }, + { + "epoch": 0.11571, + "grad_norm": 0.7338646701334941, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 11571 + }, + { + "epoch": 0.11572, + "grad_norm": 0.6583547278243067, + "learning_rate": 0.003, + "loss": 4.034, + "step": 11572 + }, + { + "epoch": 0.11573, + "grad_norm": 0.6603257054462395, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 11573 + }, + { + "epoch": 0.11574, + "grad_norm": 0.6699182756541149, + "learning_rate": 0.003, + "loss": 4.042, + "step": 11574 + }, + { + "epoch": 0.11575, + "grad_norm": 0.7352017549256189, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 11575 + }, + { + "epoch": 0.11576, + "grad_norm": 0.9900203814727313, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 11576 + }, + { + "epoch": 0.11577, + "grad_norm": 1.1856169647495152, + "learning_rate": 0.003, + "loss": 4.067, + "step": 11577 + }, + { + "epoch": 0.11578, + "grad_norm": 0.9192338750666296, + "learning_rate": 0.003, + "loss": 4.0983, + "step": 11578 + }, + { + "epoch": 0.11579, + "grad_norm": 0.9281304021462402, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 11579 + }, + { + "epoch": 0.1158, + "grad_norm": 0.8340084794128628, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 11580 + }, + { + "epoch": 0.11581, + "grad_norm": 0.8898799288754836, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 11581 + }, + { + "epoch": 0.11582, + "grad_norm": 0.9534845699789593, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 11582 + }, + { + "epoch": 0.11583, + "grad_norm": 1.1805983417623036, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 11583 + }, + { + "epoch": 0.11584, + "grad_norm": 1.0961472332698379, + "learning_rate": 0.003, + "loss": 4.0977, + "step": 11584 + }, + { + "epoch": 0.11585, + "grad_norm": 0.9198007624098695, + "learning_rate": 0.003, + "loss": 4.1108, + "step": 11585 + }, + { + "epoch": 0.11586, + "grad_norm": 0.8703708319021776, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 11586 + }, + { + "epoch": 0.11587, + "grad_norm": 0.9833243534610767, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 11587 + }, + { + "epoch": 0.11588, + "grad_norm": 1.0299790002365121, + "learning_rate": 0.003, + "loss": 4.0936, + "step": 11588 + }, + { + "epoch": 0.11589, + "grad_norm": 1.1446978761879756, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 11589 + }, + { + "epoch": 0.1159, + "grad_norm": 0.8749806945220427, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 11590 + }, + { + "epoch": 0.11591, + "grad_norm": 0.8640639014381449, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 11591 + }, + { + "epoch": 0.11592, + "grad_norm": 0.85913763338769, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 11592 + }, + { + "epoch": 0.11593, + "grad_norm": 0.8999337947177478, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 11593 + }, + { + "epoch": 0.11594, + "grad_norm": 0.9970721457694773, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 11594 + }, + { + "epoch": 0.11595, + "grad_norm": 1.2158045376163171, + "learning_rate": 0.003, + "loss": 4.1069, + "step": 11595 + }, + { + "epoch": 0.11596, + "grad_norm": 1.0305611727064, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 11596 + }, + { + "epoch": 0.11597, + "grad_norm": 0.9984279858668237, + "learning_rate": 0.003, + "loss": 4.1057, + "step": 11597 + }, + { + "epoch": 0.11598, + "grad_norm": 1.1348582164366783, + "learning_rate": 0.003, + "loss": 4.0855, + "step": 11598 + }, + { + "epoch": 0.11599, + "grad_norm": 0.8746807984576714, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 11599 + }, + { + "epoch": 0.116, + "grad_norm": 0.8623777105577025, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 11600 + }, + { + "epoch": 0.11601, + "grad_norm": 0.8983163479503808, + "learning_rate": 0.003, + "loss": 4.1046, + "step": 11601 + }, + { + "epoch": 0.11602, + "grad_norm": 0.9141569756375708, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 11602 + }, + { + "epoch": 0.11603, + "grad_norm": 0.8872815544710486, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 11603 + }, + { + "epoch": 0.11604, + "grad_norm": 0.9731698893154298, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 11604 + }, + { + "epoch": 0.11605, + "grad_norm": 1.2806382155481886, + "learning_rate": 0.003, + "loss": 4.0901, + "step": 11605 + }, + { + "epoch": 0.11606, + "grad_norm": 0.9556546692816773, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 11606 + }, + { + "epoch": 0.11607, + "grad_norm": 1.0299841279541766, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 11607 + }, + { + "epoch": 0.11608, + "grad_norm": 0.9455467524529496, + "learning_rate": 0.003, + "loss": 4.0999, + "step": 11608 + }, + { + "epoch": 0.11609, + "grad_norm": 0.9172970410574666, + "learning_rate": 0.003, + "loss": 4.1168, + "step": 11609 + }, + { + "epoch": 0.1161, + "grad_norm": 0.9344648134439203, + "learning_rate": 0.003, + "loss": 4.104, + "step": 11610 + }, + { + "epoch": 0.11611, + "grad_norm": 0.8593872475530971, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 11611 + }, + { + "epoch": 0.11612, + "grad_norm": 0.8581385544021632, + "learning_rate": 0.003, + "loss": 4.1087, + "step": 11612 + }, + { + "epoch": 0.11613, + "grad_norm": 0.9061875456410564, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 11613 + }, + { + "epoch": 0.11614, + "grad_norm": 0.8433345715324523, + "learning_rate": 0.003, + "loss": 4.0876, + "step": 11614 + }, + { + "epoch": 0.11615, + "grad_norm": 0.7389356829290649, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 11615 + }, + { + "epoch": 0.11616, + "grad_norm": 0.914493865631304, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 11616 + }, + { + "epoch": 0.11617, + "grad_norm": 1.2861195651108321, + "learning_rate": 0.003, + "loss": 4.1027, + "step": 11617 + }, + { + "epoch": 0.11618, + "grad_norm": 1.0266448145828724, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 11618 + }, + { + "epoch": 0.11619, + "grad_norm": 1.0960732659883317, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 11619 + }, + { + "epoch": 0.1162, + "grad_norm": 1.0523715774209146, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 11620 + }, + { + "epoch": 0.11621, + "grad_norm": 1.1217591635695605, + "learning_rate": 0.003, + "loss": 4.1041, + "step": 11621 + }, + { + "epoch": 0.11622, + "grad_norm": 0.8853860467856381, + "learning_rate": 0.003, + "loss": 4.0962, + "step": 11622 + }, + { + "epoch": 0.11623, + "grad_norm": 0.889851985608964, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 11623 + }, + { + "epoch": 0.11624, + "grad_norm": 0.9692408266244152, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 11624 + }, + { + "epoch": 0.11625, + "grad_norm": 0.9763083072174051, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 11625 + }, + { + "epoch": 0.11626, + "grad_norm": 1.112554579184562, + "learning_rate": 0.003, + "loss": 4.077, + "step": 11626 + }, + { + "epoch": 0.11627, + "grad_norm": 0.9217989120322748, + "learning_rate": 0.003, + "loss": 4.068, + "step": 11627 + }, + { + "epoch": 0.11628, + "grad_norm": 0.9889567056106386, + "learning_rate": 0.003, + "loss": 4.1218, + "step": 11628 + }, + { + "epoch": 0.11629, + "grad_norm": 1.1500187454364905, + "learning_rate": 0.003, + "loss": 4.082, + "step": 11629 + }, + { + "epoch": 0.1163, + "grad_norm": 0.8636209769781427, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 11630 + }, + { + "epoch": 0.11631, + "grad_norm": 0.7537144391299775, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 11631 + }, + { + "epoch": 0.11632, + "grad_norm": 0.7643748836248855, + "learning_rate": 0.003, + "loss": 4.074, + "step": 11632 + }, + { + "epoch": 0.11633, + "grad_norm": 0.7540731887948658, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 11633 + }, + { + "epoch": 0.11634, + "grad_norm": 0.8357609782481921, + "learning_rate": 0.003, + "loss": 4.0918, + "step": 11634 + }, + { + "epoch": 0.11635, + "grad_norm": 0.9336792954245737, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 11635 + }, + { + "epoch": 0.11636, + "grad_norm": 1.062114629627697, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 11636 + }, + { + "epoch": 0.11637, + "grad_norm": 1.1056447943974068, + "learning_rate": 0.003, + "loss": 4.084, + "step": 11637 + }, + { + "epoch": 0.11638, + "grad_norm": 1.0409242052405585, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 11638 + }, + { + "epoch": 0.11639, + "grad_norm": 0.9389397766942058, + "learning_rate": 0.003, + "loss": 4.06, + "step": 11639 + }, + { + "epoch": 0.1164, + "grad_norm": 0.7833156716392007, + "learning_rate": 0.003, + "loss": 4.1037, + "step": 11640 + }, + { + "epoch": 0.11641, + "grad_norm": 0.8157300424221584, + "learning_rate": 0.003, + "loss": 4.0873, + "step": 11641 + }, + { + "epoch": 0.11642, + "grad_norm": 0.7778610146990635, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 11642 + }, + { + "epoch": 0.11643, + "grad_norm": 0.8163501526320557, + "learning_rate": 0.003, + "loss": 4.0936, + "step": 11643 + }, + { + "epoch": 0.11644, + "grad_norm": 0.862526918888475, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 11644 + }, + { + "epoch": 0.11645, + "grad_norm": 0.8231016906921058, + "learning_rate": 0.003, + "loss": 4.0963, + "step": 11645 + }, + { + "epoch": 0.11646, + "grad_norm": 0.7872052664642931, + "learning_rate": 0.003, + "loss": 4.099, + "step": 11646 + }, + { + "epoch": 0.11647, + "grad_norm": 0.9572602262709535, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 11647 + }, + { + "epoch": 0.11648, + "grad_norm": 1.1777494741940657, + "learning_rate": 0.003, + "loss": 4.0944, + "step": 11648 + }, + { + "epoch": 0.11649, + "grad_norm": 1.0926580340103613, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 11649 + }, + { + "epoch": 0.1165, + "grad_norm": 0.8233082757938179, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 11650 + }, + { + "epoch": 0.11651, + "grad_norm": 0.7246989572014204, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 11651 + }, + { + "epoch": 0.11652, + "grad_norm": 0.781514705230594, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 11652 + }, + { + "epoch": 0.11653, + "grad_norm": 0.8081090591575679, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 11653 + }, + { + "epoch": 0.11654, + "grad_norm": 0.8789062458255206, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 11654 + }, + { + "epoch": 0.11655, + "grad_norm": 0.9652482265517969, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 11655 + }, + { + "epoch": 0.11656, + "grad_norm": 1.0138733273276577, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 11656 + }, + { + "epoch": 0.11657, + "grad_norm": 0.909823560812409, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 11657 + }, + { + "epoch": 0.11658, + "grad_norm": 1.1318895350408789, + "learning_rate": 0.003, + "loss": 4.1066, + "step": 11658 + }, + { + "epoch": 0.11659, + "grad_norm": 0.999315445176952, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 11659 + }, + { + "epoch": 0.1166, + "grad_norm": 1.0681231138396488, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 11660 + }, + { + "epoch": 0.11661, + "grad_norm": 0.7984412900062778, + "learning_rate": 0.003, + "loss": 4.0935, + "step": 11661 + }, + { + "epoch": 0.11662, + "grad_norm": 0.9315940478280379, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 11662 + }, + { + "epoch": 0.11663, + "grad_norm": 0.960180737266701, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 11663 + }, + { + "epoch": 0.11664, + "grad_norm": 1.2031133125922895, + "learning_rate": 0.003, + "loss": 4.1105, + "step": 11664 + }, + { + "epoch": 0.11665, + "grad_norm": 1.1830702874081243, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 11665 + }, + { + "epoch": 0.11666, + "grad_norm": 0.9753619892572791, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 11666 + }, + { + "epoch": 0.11667, + "grad_norm": 1.0875911726738154, + "learning_rate": 0.003, + "loss": 4.1196, + "step": 11667 + }, + { + "epoch": 0.11668, + "grad_norm": 0.8676095536295696, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 11668 + }, + { + "epoch": 0.11669, + "grad_norm": 0.6994624424259369, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 11669 + }, + { + "epoch": 0.1167, + "grad_norm": 1.0011566524839903, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 11670 + }, + { + "epoch": 0.11671, + "grad_norm": 1.3808893515895955, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 11671 + }, + { + "epoch": 0.11672, + "grad_norm": 0.7186120679230394, + "learning_rate": 0.003, + "loss": 4.1075, + "step": 11672 + }, + { + "epoch": 0.11673, + "grad_norm": 0.7369011397093376, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 11673 + }, + { + "epoch": 0.11674, + "grad_norm": 0.9942245339134483, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 11674 + }, + { + "epoch": 0.11675, + "grad_norm": 1.0349780516751126, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 11675 + }, + { + "epoch": 0.11676, + "grad_norm": 0.9051438671096496, + "learning_rate": 0.003, + "loss": 4.0855, + "step": 11676 + }, + { + "epoch": 0.11677, + "grad_norm": 1.0108772384110651, + "learning_rate": 0.003, + "loss": 4.1005, + "step": 11677 + }, + { + "epoch": 0.11678, + "grad_norm": 1.131044572365773, + "learning_rate": 0.003, + "loss": 4.1015, + "step": 11678 + }, + { + "epoch": 0.11679, + "grad_norm": 0.8009657271371924, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 11679 + }, + { + "epoch": 0.1168, + "grad_norm": 0.8562042182798109, + "learning_rate": 0.003, + "loss": 4.108, + "step": 11680 + }, + { + "epoch": 0.11681, + "grad_norm": 0.8292823977560939, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 11681 + }, + { + "epoch": 0.11682, + "grad_norm": 0.8845882820486737, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 11682 + }, + { + "epoch": 0.11683, + "grad_norm": 0.9452669016402371, + "learning_rate": 0.003, + "loss": 4.1108, + "step": 11683 + }, + { + "epoch": 0.11684, + "grad_norm": 1.040914644238037, + "learning_rate": 0.003, + "loss": 4.0906, + "step": 11684 + }, + { + "epoch": 0.11685, + "grad_norm": 0.9994069611564993, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 11685 + }, + { + "epoch": 0.11686, + "grad_norm": 1.0052218282597227, + "learning_rate": 0.003, + "loss": 4.0933, + "step": 11686 + }, + { + "epoch": 0.11687, + "grad_norm": 0.9692084000722573, + "learning_rate": 0.003, + "loss": 4.1001, + "step": 11687 + }, + { + "epoch": 0.11688, + "grad_norm": 1.0186002526996951, + "learning_rate": 0.003, + "loss": 4.1013, + "step": 11688 + }, + { + "epoch": 0.11689, + "grad_norm": 1.1991453426044019, + "learning_rate": 0.003, + "loss": 4.1018, + "step": 11689 + }, + { + "epoch": 0.1169, + "grad_norm": 0.9260554518373962, + "learning_rate": 0.003, + "loss": 4.0891, + "step": 11690 + }, + { + "epoch": 0.11691, + "grad_norm": 0.9799157950211934, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 11691 + }, + { + "epoch": 0.11692, + "grad_norm": 1.0519221601999824, + "learning_rate": 0.003, + "loss": 4.08, + "step": 11692 + }, + { + "epoch": 0.11693, + "grad_norm": 1.1018705025468218, + "learning_rate": 0.003, + "loss": 4.095, + "step": 11693 + }, + { + "epoch": 0.11694, + "grad_norm": 1.2486510645682192, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 11694 + }, + { + "epoch": 0.11695, + "grad_norm": 0.9604694520527333, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 11695 + }, + { + "epoch": 0.11696, + "grad_norm": 0.8932987616778585, + "learning_rate": 0.003, + "loss": 4.0901, + "step": 11696 + }, + { + "epoch": 0.11697, + "grad_norm": 1.0446669906234554, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 11697 + }, + { + "epoch": 0.11698, + "grad_norm": 1.1625523386568448, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 11698 + }, + { + "epoch": 0.11699, + "grad_norm": 0.9451921987549693, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 11699 + }, + { + "epoch": 0.117, + "grad_norm": 0.9599959106137281, + "learning_rate": 0.003, + "loss": 4.0898, + "step": 11700 + }, + { + "epoch": 0.11701, + "grad_norm": 0.8421494861862855, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 11701 + }, + { + "epoch": 0.11702, + "grad_norm": 0.8482002011762976, + "learning_rate": 0.003, + "loss": 4.0868, + "step": 11702 + }, + { + "epoch": 0.11703, + "grad_norm": 0.9084783721357652, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 11703 + }, + { + "epoch": 0.11704, + "grad_norm": 0.854002897175805, + "learning_rate": 0.003, + "loss": 4.1057, + "step": 11704 + }, + { + "epoch": 0.11705, + "grad_norm": 0.8468442557851493, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 11705 + }, + { + "epoch": 0.11706, + "grad_norm": 0.9740618372350455, + "learning_rate": 0.003, + "loss": 4.103, + "step": 11706 + }, + { + "epoch": 0.11707, + "grad_norm": 1.3764541532621752, + "learning_rate": 0.003, + "loss": 4.097, + "step": 11707 + }, + { + "epoch": 0.11708, + "grad_norm": 0.7924923324150779, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 11708 + }, + { + "epoch": 0.11709, + "grad_norm": 0.7192614977953711, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 11709 + }, + { + "epoch": 0.1171, + "grad_norm": 0.6860040054964368, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 11710 + }, + { + "epoch": 0.11711, + "grad_norm": 0.6831465888682684, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 11711 + }, + { + "epoch": 0.11712, + "grad_norm": 0.7097048121995502, + "learning_rate": 0.003, + "loss": 4.1147, + "step": 11712 + }, + { + "epoch": 0.11713, + "grad_norm": 0.9253353303540238, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 11713 + }, + { + "epoch": 0.11714, + "grad_norm": 1.2673557229012302, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 11714 + }, + { + "epoch": 0.11715, + "grad_norm": 0.9404443189762333, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 11715 + }, + { + "epoch": 0.11716, + "grad_norm": 0.9602361128801753, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 11716 + }, + { + "epoch": 0.11717, + "grad_norm": 0.9233200097105245, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 11717 + }, + { + "epoch": 0.11718, + "grad_norm": 0.84998124387827, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 11718 + }, + { + "epoch": 0.11719, + "grad_norm": 0.9744487591510355, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 11719 + }, + { + "epoch": 0.1172, + "grad_norm": 1.0052411592007757, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 11720 + }, + { + "epoch": 0.11721, + "grad_norm": 1.1986615088412775, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 11721 + }, + { + "epoch": 0.11722, + "grad_norm": 0.834885509124728, + "learning_rate": 0.003, + "loss": 4.093, + "step": 11722 + }, + { + "epoch": 0.11723, + "grad_norm": 0.7455210396714101, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 11723 + }, + { + "epoch": 0.11724, + "grad_norm": 0.7808216255901206, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 11724 + }, + { + "epoch": 0.11725, + "grad_norm": 0.831970588825829, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 11725 + }, + { + "epoch": 0.11726, + "grad_norm": 0.9664189763580419, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 11726 + }, + { + "epoch": 0.11727, + "grad_norm": 0.9565006157453467, + "learning_rate": 0.003, + "loss": 4.1035, + "step": 11727 + }, + { + "epoch": 0.11728, + "grad_norm": 1.0452043048344577, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 11728 + }, + { + "epoch": 0.11729, + "grad_norm": 1.151651788158293, + "learning_rate": 0.003, + "loss": 4.0968, + "step": 11729 + }, + { + "epoch": 0.1173, + "grad_norm": 0.8620302175860334, + "learning_rate": 0.003, + "loss": 4.0893, + "step": 11730 + }, + { + "epoch": 0.11731, + "grad_norm": 0.9120677582124169, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 11731 + }, + { + "epoch": 0.11732, + "grad_norm": 1.0186869574978659, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 11732 + }, + { + "epoch": 0.11733, + "grad_norm": 1.2212130711571987, + "learning_rate": 0.003, + "loss": 4.0963, + "step": 11733 + }, + { + "epoch": 0.11734, + "grad_norm": 0.8704910356479687, + "learning_rate": 0.003, + "loss": 4.0888, + "step": 11734 + }, + { + "epoch": 0.11735, + "grad_norm": 0.716386703699676, + "learning_rate": 0.003, + "loss": 4.067, + "step": 11735 + }, + { + "epoch": 0.11736, + "grad_norm": 0.862054154400894, + "learning_rate": 0.003, + "loss": 4.0887, + "step": 11736 + }, + { + "epoch": 0.11737, + "grad_norm": 0.9340061554209025, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 11737 + }, + { + "epoch": 0.11738, + "grad_norm": 0.9268944260310923, + "learning_rate": 0.003, + "loss": 4.099, + "step": 11738 + }, + { + "epoch": 0.11739, + "grad_norm": 0.8565835054988687, + "learning_rate": 0.003, + "loss": 4.0923, + "step": 11739 + }, + { + "epoch": 0.1174, + "grad_norm": 0.7659299074020319, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 11740 + }, + { + "epoch": 0.11741, + "grad_norm": 0.8408421839734038, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 11741 + }, + { + "epoch": 0.11742, + "grad_norm": 1.1456952392284279, + "learning_rate": 0.003, + "loss": 4.1096, + "step": 11742 + }, + { + "epoch": 0.11743, + "grad_norm": 1.0466103319968232, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 11743 + }, + { + "epoch": 0.11744, + "grad_norm": 1.1198578709558302, + "learning_rate": 0.003, + "loss": 4.0858, + "step": 11744 + }, + { + "epoch": 0.11745, + "grad_norm": 0.9572295462577893, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 11745 + }, + { + "epoch": 0.11746, + "grad_norm": 0.9441716197699002, + "learning_rate": 0.003, + "loss": 4.059, + "step": 11746 + }, + { + "epoch": 0.11747, + "grad_norm": 1.0657129327797676, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 11747 + }, + { + "epoch": 0.11748, + "grad_norm": 1.2405082040579574, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 11748 + }, + { + "epoch": 0.11749, + "grad_norm": 0.861512859871639, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 11749 + }, + { + "epoch": 0.1175, + "grad_norm": 0.9271767658258799, + "learning_rate": 0.003, + "loss": 4.1022, + "step": 11750 + }, + { + "epoch": 0.11751, + "grad_norm": 0.9606921390650458, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 11751 + }, + { + "epoch": 0.11752, + "grad_norm": 0.9289264480467805, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 11752 + }, + { + "epoch": 0.11753, + "grad_norm": 1.0297406836877407, + "learning_rate": 0.003, + "loss": 4.1125, + "step": 11753 + }, + { + "epoch": 0.11754, + "grad_norm": 0.9193174361650066, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 11754 + }, + { + "epoch": 0.11755, + "grad_norm": 0.7600814464843886, + "learning_rate": 0.003, + "loss": 4.1004, + "step": 11755 + }, + { + "epoch": 0.11756, + "grad_norm": 0.9396516153219138, + "learning_rate": 0.003, + "loss": 4.0943, + "step": 11756 + }, + { + "epoch": 0.11757, + "grad_norm": 1.350786960905117, + "learning_rate": 0.003, + "loss": 4.0872, + "step": 11757 + }, + { + "epoch": 0.11758, + "grad_norm": 1.0213665895439015, + "learning_rate": 0.003, + "loss": 4.1104, + "step": 11758 + }, + { + "epoch": 0.11759, + "grad_norm": 0.9648225625356928, + "learning_rate": 0.003, + "loss": 4.1025, + "step": 11759 + }, + { + "epoch": 0.1176, + "grad_norm": 0.9601981162701113, + "learning_rate": 0.003, + "loss": 4.0874, + "step": 11760 + }, + { + "epoch": 0.11761, + "grad_norm": 1.13228914251621, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 11761 + }, + { + "epoch": 0.11762, + "grad_norm": 1.0064921946637726, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 11762 + }, + { + "epoch": 0.11763, + "grad_norm": 0.9910125341236696, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 11763 + }, + { + "epoch": 0.11764, + "grad_norm": 0.9758589853549087, + "learning_rate": 0.003, + "loss": 4.0912, + "step": 11764 + }, + { + "epoch": 0.11765, + "grad_norm": 1.0085277402815251, + "learning_rate": 0.003, + "loss": 4.0973, + "step": 11765 + }, + { + "epoch": 0.11766, + "grad_norm": 0.8700176268224782, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 11766 + }, + { + "epoch": 0.11767, + "grad_norm": 0.7883380408417915, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 11767 + }, + { + "epoch": 0.11768, + "grad_norm": 0.7121251423739637, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 11768 + }, + { + "epoch": 0.11769, + "grad_norm": 0.6509492978711283, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 11769 + }, + { + "epoch": 0.1177, + "grad_norm": 0.7338132929766362, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 11770 + }, + { + "epoch": 0.11771, + "grad_norm": 0.7865181334941894, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 11771 + }, + { + "epoch": 0.11772, + "grad_norm": 0.7911894976172725, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 11772 + }, + { + "epoch": 0.11773, + "grad_norm": 0.9800014737485845, + "learning_rate": 0.003, + "loss": 4.0845, + "step": 11773 + }, + { + "epoch": 0.11774, + "grad_norm": 1.4119542343425957, + "learning_rate": 0.003, + "loss": 4.085, + "step": 11774 + }, + { + "epoch": 0.11775, + "grad_norm": 0.7064059984401361, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 11775 + }, + { + "epoch": 0.11776, + "grad_norm": 0.7019357211397205, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 11776 + }, + { + "epoch": 0.11777, + "grad_norm": 0.8245763386471162, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 11777 + }, + { + "epoch": 0.11778, + "grad_norm": 0.9559225306152273, + "learning_rate": 0.003, + "loss": 4.096, + "step": 11778 + }, + { + "epoch": 0.11779, + "grad_norm": 1.2431370344902348, + "learning_rate": 0.003, + "loss": 4.09, + "step": 11779 + }, + { + "epoch": 0.1178, + "grad_norm": 1.0234570974106127, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 11780 + }, + { + "epoch": 0.11781, + "grad_norm": 1.107978831394946, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 11781 + }, + { + "epoch": 0.11782, + "grad_norm": 0.9974327679465759, + "learning_rate": 0.003, + "loss": 4.0964, + "step": 11782 + }, + { + "epoch": 0.11783, + "grad_norm": 1.0989691582351766, + "learning_rate": 0.003, + "loss": 4.1111, + "step": 11783 + }, + { + "epoch": 0.11784, + "grad_norm": 0.8564870741976472, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 11784 + }, + { + "epoch": 0.11785, + "grad_norm": 0.7709951391677018, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 11785 + }, + { + "epoch": 0.11786, + "grad_norm": 0.7677580280160918, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 11786 + }, + { + "epoch": 0.11787, + "grad_norm": 0.7982381770269343, + "learning_rate": 0.003, + "loss": 4.0991, + "step": 11787 + }, + { + "epoch": 0.11788, + "grad_norm": 0.8729342594944264, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 11788 + }, + { + "epoch": 0.11789, + "grad_norm": 0.9017881453661968, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 11789 + }, + { + "epoch": 0.1179, + "grad_norm": 1.1034364918430402, + "learning_rate": 0.003, + "loss": 4.0946, + "step": 11790 + }, + { + "epoch": 0.11791, + "grad_norm": 1.2026109653841168, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 11791 + }, + { + "epoch": 0.11792, + "grad_norm": 1.0145614208990714, + "learning_rate": 0.003, + "loss": 4.0931, + "step": 11792 + }, + { + "epoch": 0.11793, + "grad_norm": 0.9799695950363377, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 11793 + }, + { + "epoch": 0.11794, + "grad_norm": 0.8522836880521207, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 11794 + }, + { + "epoch": 0.11795, + "grad_norm": 0.9255333540557957, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 11795 + }, + { + "epoch": 0.11796, + "grad_norm": 1.1320182124380607, + "learning_rate": 0.003, + "loss": 4.1052, + "step": 11796 + }, + { + "epoch": 0.11797, + "grad_norm": 1.0489000052450246, + "learning_rate": 0.003, + "loss": 4.068, + "step": 11797 + }, + { + "epoch": 0.11798, + "grad_norm": 0.9767291362339403, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 11798 + }, + { + "epoch": 0.11799, + "grad_norm": 0.9561900566264225, + "learning_rate": 0.003, + "loss": 4.1104, + "step": 11799 + }, + { + "epoch": 0.118, + "grad_norm": 1.0458463458344696, + "learning_rate": 0.003, + "loss": 4.0941, + "step": 11800 + }, + { + "epoch": 0.11801, + "grad_norm": 0.9425659186231714, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 11801 + }, + { + "epoch": 0.11802, + "grad_norm": 0.9644888779150168, + "learning_rate": 0.003, + "loss": 4.0819, + "step": 11802 + }, + { + "epoch": 0.11803, + "grad_norm": 1.18398864894741, + "learning_rate": 0.003, + "loss": 4.1115, + "step": 11803 + }, + { + "epoch": 0.11804, + "grad_norm": 0.9095643898725806, + "learning_rate": 0.003, + "loss": 4.095, + "step": 11804 + }, + { + "epoch": 0.11805, + "grad_norm": 1.218416131436913, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 11805 + }, + { + "epoch": 0.11806, + "grad_norm": 0.9915983603343994, + "learning_rate": 0.003, + "loss": 4.0864, + "step": 11806 + }, + { + "epoch": 0.11807, + "grad_norm": 0.961179071462202, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 11807 + }, + { + "epoch": 0.11808, + "grad_norm": 0.80600437144899, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 11808 + }, + { + "epoch": 0.11809, + "grad_norm": 0.8974692921381201, + "learning_rate": 0.003, + "loss": 4.0981, + "step": 11809 + }, + { + "epoch": 0.1181, + "grad_norm": 0.9861909793318101, + "learning_rate": 0.003, + "loss": 4.079, + "step": 11810 + }, + { + "epoch": 0.11811, + "grad_norm": 0.929247210122778, + "learning_rate": 0.003, + "loss": 4.0975, + "step": 11811 + }, + { + "epoch": 0.11812, + "grad_norm": 0.9101504883654837, + "learning_rate": 0.003, + "loss": 4.086, + "step": 11812 + }, + { + "epoch": 0.11813, + "grad_norm": 0.8598204082234447, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 11813 + }, + { + "epoch": 0.11814, + "grad_norm": 0.7395576767988027, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 11814 + }, + { + "epoch": 0.11815, + "grad_norm": 0.7549332561536323, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 11815 + }, + { + "epoch": 0.11816, + "grad_norm": 0.8150240492518723, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 11816 + }, + { + "epoch": 0.11817, + "grad_norm": 0.954021639055784, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 11817 + }, + { + "epoch": 0.11818, + "grad_norm": 1.1664618790548895, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 11818 + }, + { + "epoch": 0.11819, + "grad_norm": 0.9707893518246115, + "learning_rate": 0.003, + "loss": 4.051, + "step": 11819 + }, + { + "epoch": 0.1182, + "grad_norm": 0.8591857013140873, + "learning_rate": 0.003, + "loss": 4.0893, + "step": 11820 + }, + { + "epoch": 0.11821, + "grad_norm": 0.8675002824499801, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 11821 + }, + { + "epoch": 0.11822, + "grad_norm": 1.0184120165543393, + "learning_rate": 0.003, + "loss": 4.09, + "step": 11822 + }, + { + "epoch": 0.11823, + "grad_norm": 1.2044815201580288, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 11823 + }, + { + "epoch": 0.11824, + "grad_norm": 0.8428010828090874, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 11824 + }, + { + "epoch": 0.11825, + "grad_norm": 0.8312565614433316, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 11825 + }, + { + "epoch": 0.11826, + "grad_norm": 0.9478220716194203, + "learning_rate": 0.003, + "loss": 4.091, + "step": 11826 + }, + { + "epoch": 0.11827, + "grad_norm": 1.2063579450920796, + "learning_rate": 0.003, + "loss": 4.1129, + "step": 11827 + }, + { + "epoch": 0.11828, + "grad_norm": 0.9371570792025025, + "learning_rate": 0.003, + "loss": 4.1037, + "step": 11828 + }, + { + "epoch": 0.11829, + "grad_norm": 1.2968878749084483, + "learning_rate": 0.003, + "loss": 4.1175, + "step": 11829 + }, + { + "epoch": 0.1183, + "grad_norm": 0.9893112107215147, + "learning_rate": 0.003, + "loss": 4.0952, + "step": 11830 + }, + { + "epoch": 0.11831, + "grad_norm": 0.9939980875584855, + "learning_rate": 0.003, + "loss": 4.0908, + "step": 11831 + }, + { + "epoch": 0.11832, + "grad_norm": 1.0542745737979595, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 11832 + }, + { + "epoch": 0.11833, + "grad_norm": 1.0357041410059662, + "learning_rate": 0.003, + "loss": 4.1197, + "step": 11833 + }, + { + "epoch": 0.11834, + "grad_norm": 0.9182665827969304, + "learning_rate": 0.003, + "loss": 4.1132, + "step": 11834 + }, + { + "epoch": 0.11835, + "grad_norm": 0.9752084302361046, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 11835 + }, + { + "epoch": 0.11836, + "grad_norm": 0.9565595030603186, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 11836 + }, + { + "epoch": 0.11837, + "grad_norm": 0.9466473570755174, + "learning_rate": 0.003, + "loss": 4.095, + "step": 11837 + }, + { + "epoch": 0.11838, + "grad_norm": 0.8734801789105325, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 11838 + }, + { + "epoch": 0.11839, + "grad_norm": 0.9134602757743311, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 11839 + }, + { + "epoch": 0.1184, + "grad_norm": 0.8868267467804094, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 11840 + }, + { + "epoch": 0.11841, + "grad_norm": 0.9143243118335659, + "learning_rate": 0.003, + "loss": 4.0915, + "step": 11841 + }, + { + "epoch": 0.11842, + "grad_norm": 0.8866107011082458, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 11842 + }, + { + "epoch": 0.11843, + "grad_norm": 0.971301276845872, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 11843 + }, + { + "epoch": 0.11844, + "grad_norm": 1.0302043856665823, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 11844 + }, + { + "epoch": 0.11845, + "grad_norm": 0.9366543287157849, + "learning_rate": 0.003, + "loss": 4.0994, + "step": 11845 + }, + { + "epoch": 0.11846, + "grad_norm": 0.8090307068613266, + "learning_rate": 0.003, + "loss": 4.0871, + "step": 11846 + }, + { + "epoch": 0.11847, + "grad_norm": 0.8709051593478219, + "learning_rate": 0.003, + "loss": 4.0954, + "step": 11847 + }, + { + "epoch": 0.11848, + "grad_norm": 1.1198334891638153, + "learning_rate": 0.003, + "loss": 4.1131, + "step": 11848 + }, + { + "epoch": 0.11849, + "grad_norm": 1.2284178104387797, + "learning_rate": 0.003, + "loss": 4.1103, + "step": 11849 + }, + { + "epoch": 0.1185, + "grad_norm": 0.975317633599507, + "learning_rate": 0.003, + "loss": 4.0919, + "step": 11850 + }, + { + "epoch": 0.11851, + "grad_norm": 0.9803338702585738, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 11851 + }, + { + "epoch": 0.11852, + "grad_norm": 1.0822202411831328, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 11852 + }, + { + "epoch": 0.11853, + "grad_norm": 1.0596306406722125, + "learning_rate": 0.003, + "loss": 4.0933, + "step": 11853 + }, + { + "epoch": 0.11854, + "grad_norm": 1.1215129742485612, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 11854 + }, + { + "epoch": 0.11855, + "grad_norm": 0.8948634542835897, + "learning_rate": 0.003, + "loss": 4.1016, + "step": 11855 + }, + { + "epoch": 0.11856, + "grad_norm": 0.9204860362028344, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 11856 + }, + { + "epoch": 0.11857, + "grad_norm": 1.0093525122404978, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 11857 + }, + { + "epoch": 0.11858, + "grad_norm": 1.2932757170581461, + "learning_rate": 0.003, + "loss": 4.1238, + "step": 11858 + }, + { + "epoch": 0.11859, + "grad_norm": 0.7056224796745593, + "learning_rate": 0.003, + "loss": 4.0906, + "step": 11859 + }, + { + "epoch": 0.1186, + "grad_norm": 0.7019652160849805, + "learning_rate": 0.003, + "loss": 4.079, + "step": 11860 + }, + { + "epoch": 0.11861, + "grad_norm": 0.7220589529508009, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 11861 + }, + { + "epoch": 0.11862, + "grad_norm": 1.071201582243624, + "learning_rate": 0.003, + "loss": 4.0985, + "step": 11862 + }, + { + "epoch": 0.11863, + "grad_norm": 1.1442848796523826, + "learning_rate": 0.003, + "loss": 4.105, + "step": 11863 + }, + { + "epoch": 0.11864, + "grad_norm": 0.689528856319085, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 11864 + }, + { + "epoch": 0.11865, + "grad_norm": 0.6239446507863784, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 11865 + }, + { + "epoch": 0.11866, + "grad_norm": 0.6408670640576777, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 11866 + }, + { + "epoch": 0.11867, + "grad_norm": 0.7507217589358482, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 11867 + }, + { + "epoch": 0.11868, + "grad_norm": 0.9341844379106476, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 11868 + }, + { + "epoch": 0.11869, + "grad_norm": 1.0706689406347305, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 11869 + }, + { + "epoch": 0.1187, + "grad_norm": 1.0148253486423675, + "learning_rate": 0.003, + "loss": 4.0872, + "step": 11870 + }, + { + "epoch": 0.11871, + "grad_norm": 0.9512489118954868, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 11871 + }, + { + "epoch": 0.11872, + "grad_norm": 0.8870624159293142, + "learning_rate": 0.003, + "loss": 4.0967, + "step": 11872 + }, + { + "epoch": 0.11873, + "grad_norm": 0.8891866190736758, + "learning_rate": 0.003, + "loss": 4.063, + "step": 11873 + }, + { + "epoch": 0.11874, + "grad_norm": 0.9744992991647309, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 11874 + }, + { + "epoch": 0.11875, + "grad_norm": 0.92499230406142, + "learning_rate": 0.003, + "loss": 4.0856, + "step": 11875 + }, + { + "epoch": 0.11876, + "grad_norm": 0.9512308670986421, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 11876 + }, + { + "epoch": 0.11877, + "grad_norm": 0.8623634646358607, + "learning_rate": 0.003, + "loss": 4.1028, + "step": 11877 + }, + { + "epoch": 0.11878, + "grad_norm": 0.9426323013061633, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 11878 + }, + { + "epoch": 0.11879, + "grad_norm": 0.9735676528705841, + "learning_rate": 0.003, + "loss": 4.095, + "step": 11879 + }, + { + "epoch": 0.1188, + "grad_norm": 1.0532677027925437, + "learning_rate": 0.003, + "loss": 4.0848, + "step": 11880 + }, + { + "epoch": 0.11881, + "grad_norm": 1.0468942058136443, + "learning_rate": 0.003, + "loss": 4.08, + "step": 11881 + }, + { + "epoch": 0.11882, + "grad_norm": 0.9059920114878208, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 11882 + }, + { + "epoch": 0.11883, + "grad_norm": 0.8583149337356998, + "learning_rate": 0.003, + "loss": 4.0995, + "step": 11883 + }, + { + "epoch": 0.11884, + "grad_norm": 0.8457121655745958, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 11884 + }, + { + "epoch": 0.11885, + "grad_norm": 0.9853524382582202, + "learning_rate": 0.003, + "loss": 4.1104, + "step": 11885 + }, + { + "epoch": 0.11886, + "grad_norm": 1.128390019673183, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 11886 + }, + { + "epoch": 0.11887, + "grad_norm": 0.9500783386683729, + "learning_rate": 0.003, + "loss": 4.061, + "step": 11887 + }, + { + "epoch": 0.11888, + "grad_norm": 1.0776486675735164, + "learning_rate": 0.003, + "loss": 4.1128, + "step": 11888 + }, + { + "epoch": 0.11889, + "grad_norm": 1.14897726652806, + "learning_rate": 0.003, + "loss": 4.0994, + "step": 11889 + }, + { + "epoch": 0.1189, + "grad_norm": 0.8759222816710914, + "learning_rate": 0.003, + "loss": 4.1027, + "step": 11890 + }, + { + "epoch": 0.11891, + "grad_norm": 0.7811247901950734, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 11891 + }, + { + "epoch": 0.11892, + "grad_norm": 0.9036550629941046, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 11892 + }, + { + "epoch": 0.11893, + "grad_norm": 1.085727926790284, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 11893 + }, + { + "epoch": 0.11894, + "grad_norm": 1.041692521496345, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 11894 + }, + { + "epoch": 0.11895, + "grad_norm": 0.8801440793483963, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 11895 + }, + { + "epoch": 0.11896, + "grad_norm": 0.8993046294194258, + "learning_rate": 0.003, + "loss": 4.0988, + "step": 11896 + }, + { + "epoch": 0.11897, + "grad_norm": 0.8972755702491334, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 11897 + }, + { + "epoch": 0.11898, + "grad_norm": 0.9469716603955483, + "learning_rate": 0.003, + "loss": 4.0996, + "step": 11898 + }, + { + "epoch": 0.11899, + "grad_norm": 1.0248800881124134, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 11899 + }, + { + "epoch": 0.119, + "grad_norm": 1.0914072752343296, + "learning_rate": 0.003, + "loss": 4.092, + "step": 11900 + }, + { + "epoch": 0.11901, + "grad_norm": 0.9792597103397298, + "learning_rate": 0.003, + "loss": 4.1027, + "step": 11901 + }, + { + "epoch": 0.11902, + "grad_norm": 1.1296736731682393, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 11902 + }, + { + "epoch": 0.11903, + "grad_norm": 0.8869032364352385, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 11903 + }, + { + "epoch": 0.11904, + "grad_norm": 0.8458668908013703, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 11904 + }, + { + "epoch": 0.11905, + "grad_norm": 0.9082804188896054, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 11905 + }, + { + "epoch": 0.11906, + "grad_norm": 0.9929555345462653, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 11906 + }, + { + "epoch": 0.11907, + "grad_norm": 1.0997075532136633, + "learning_rate": 0.003, + "loss": 4.0959, + "step": 11907 + }, + { + "epoch": 0.11908, + "grad_norm": 0.9203295659333136, + "learning_rate": 0.003, + "loss": 4.0969, + "step": 11908 + }, + { + "epoch": 0.11909, + "grad_norm": 1.088198546231743, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 11909 + }, + { + "epoch": 0.1191, + "grad_norm": 1.011940359323603, + "learning_rate": 0.003, + "loss": 4.0985, + "step": 11910 + }, + { + "epoch": 0.11911, + "grad_norm": 0.9103515458950083, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 11911 + }, + { + "epoch": 0.11912, + "grad_norm": 0.9195423685147066, + "learning_rate": 0.003, + "loss": 4.1208, + "step": 11912 + }, + { + "epoch": 0.11913, + "grad_norm": 0.9989018114539654, + "learning_rate": 0.003, + "loss": 4.1031, + "step": 11913 + }, + { + "epoch": 0.11914, + "grad_norm": 1.3075467906520217, + "learning_rate": 0.003, + "loss": 4.1156, + "step": 11914 + }, + { + "epoch": 0.11915, + "grad_norm": 0.8002103113809294, + "learning_rate": 0.003, + "loss": 4.0935, + "step": 11915 + }, + { + "epoch": 0.11916, + "grad_norm": 0.8576247896838758, + "learning_rate": 0.003, + "loss": 4.1058, + "step": 11916 + }, + { + "epoch": 0.11917, + "grad_norm": 0.8619218703543767, + "learning_rate": 0.003, + "loss": 4.0998, + "step": 11917 + }, + { + "epoch": 0.11918, + "grad_norm": 0.92064791043721, + "learning_rate": 0.003, + "loss": 4.0902, + "step": 11918 + }, + { + "epoch": 0.11919, + "grad_norm": 1.0752821205865917, + "learning_rate": 0.003, + "loss": 4.1022, + "step": 11919 + }, + { + "epoch": 0.1192, + "grad_norm": 0.9566371828463166, + "learning_rate": 0.003, + "loss": 4.0958, + "step": 11920 + }, + { + "epoch": 0.11921, + "grad_norm": 1.2450854148990922, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 11921 + }, + { + "epoch": 0.11922, + "grad_norm": 1.0503984617636044, + "learning_rate": 0.003, + "loss": 4.109, + "step": 11922 + }, + { + "epoch": 0.11923, + "grad_norm": 0.8806693630737407, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 11923 + }, + { + "epoch": 0.11924, + "grad_norm": 0.9726670496280382, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 11924 + }, + { + "epoch": 0.11925, + "grad_norm": 1.0663939646396499, + "learning_rate": 0.003, + "loss": 4.1076, + "step": 11925 + }, + { + "epoch": 0.11926, + "grad_norm": 0.9803547302414916, + "learning_rate": 0.003, + "loss": 4.0915, + "step": 11926 + }, + { + "epoch": 0.11927, + "grad_norm": 1.0531494552114178, + "learning_rate": 0.003, + "loss": 4.106, + "step": 11927 + }, + { + "epoch": 0.11928, + "grad_norm": 1.0238581220226066, + "learning_rate": 0.003, + "loss": 4.0873, + "step": 11928 + }, + { + "epoch": 0.11929, + "grad_norm": 0.8723079890121401, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 11929 + }, + { + "epoch": 0.1193, + "grad_norm": 0.8642083734209672, + "learning_rate": 0.003, + "loss": 4.0927, + "step": 11930 + }, + { + "epoch": 0.11931, + "grad_norm": 0.9115063943149905, + "learning_rate": 0.003, + "loss": 4.0819, + "step": 11931 + }, + { + "epoch": 0.11932, + "grad_norm": 0.9058993130164262, + "learning_rate": 0.003, + "loss": 4.1028, + "step": 11932 + }, + { + "epoch": 0.11933, + "grad_norm": 0.8326909815916078, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 11933 + }, + { + "epoch": 0.11934, + "grad_norm": 1.0188912113383914, + "learning_rate": 0.003, + "loss": 4.1095, + "step": 11934 + }, + { + "epoch": 0.11935, + "grad_norm": 1.0321452998030938, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 11935 + }, + { + "epoch": 0.11936, + "grad_norm": 1.0683212281284595, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 11936 + }, + { + "epoch": 0.11937, + "grad_norm": 1.3486434197500694, + "learning_rate": 0.003, + "loss": 4.1066, + "step": 11937 + }, + { + "epoch": 0.11938, + "grad_norm": 1.0307883913478488, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 11938 + }, + { + "epoch": 0.11939, + "grad_norm": 0.8778357420959993, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 11939 + }, + { + "epoch": 0.1194, + "grad_norm": 0.7893986663788695, + "learning_rate": 0.003, + "loss": 4.096, + "step": 11940 + }, + { + "epoch": 0.11941, + "grad_norm": 0.8254525362256458, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 11941 + }, + { + "epoch": 0.11942, + "grad_norm": 1.0086878067606946, + "learning_rate": 0.003, + "loss": 4.1059, + "step": 11942 + }, + { + "epoch": 0.11943, + "grad_norm": 1.2379923462647893, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 11943 + }, + { + "epoch": 0.11944, + "grad_norm": 0.767463581359972, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 11944 + }, + { + "epoch": 0.11945, + "grad_norm": 0.7338543339519061, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 11945 + }, + { + "epoch": 0.11946, + "grad_norm": 0.8682066749679745, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 11946 + }, + { + "epoch": 0.11947, + "grad_norm": 1.0724253338017942, + "learning_rate": 0.003, + "loss": 4.0994, + "step": 11947 + }, + { + "epoch": 0.11948, + "grad_norm": 1.0715203817820578, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 11948 + }, + { + "epoch": 0.11949, + "grad_norm": 0.9378444925190891, + "learning_rate": 0.003, + "loss": 4.1021, + "step": 11949 + }, + { + "epoch": 0.1195, + "grad_norm": 0.9350710252906247, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 11950 + }, + { + "epoch": 0.11951, + "grad_norm": 0.8734969757733143, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 11951 + }, + { + "epoch": 0.11952, + "grad_norm": 0.7529543123234806, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 11952 + }, + { + "epoch": 0.11953, + "grad_norm": 0.8643603482164126, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 11953 + }, + { + "epoch": 0.11954, + "grad_norm": 0.9710523229591159, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 11954 + }, + { + "epoch": 0.11955, + "grad_norm": 1.0927361963223938, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 11955 + }, + { + "epoch": 0.11956, + "grad_norm": 1.063304186885839, + "learning_rate": 0.003, + "loss": 4.0981, + "step": 11956 + }, + { + "epoch": 0.11957, + "grad_norm": 0.9244757518500899, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 11957 + }, + { + "epoch": 0.11958, + "grad_norm": 0.7988338865198208, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 11958 + }, + { + "epoch": 0.11959, + "grad_norm": 0.7275171779935972, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 11959 + }, + { + "epoch": 0.1196, + "grad_norm": 0.7765755598378262, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 11960 + }, + { + "epoch": 0.11961, + "grad_norm": 0.968100664626869, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 11961 + }, + { + "epoch": 0.11962, + "grad_norm": 1.2671133204283838, + "learning_rate": 0.003, + "loss": 4.1071, + "step": 11962 + }, + { + "epoch": 0.11963, + "grad_norm": 0.7544551507727536, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 11963 + }, + { + "epoch": 0.11964, + "grad_norm": 0.6241425904009411, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 11964 + }, + { + "epoch": 0.11965, + "grad_norm": 0.6387182391702654, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 11965 + }, + { + "epoch": 0.11966, + "grad_norm": 0.5812768096392469, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 11966 + }, + { + "epoch": 0.11967, + "grad_norm": 0.644667139236145, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 11967 + }, + { + "epoch": 0.11968, + "grad_norm": 0.7521249304218822, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 11968 + }, + { + "epoch": 0.11969, + "grad_norm": 0.8527341614041221, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 11969 + }, + { + "epoch": 0.1197, + "grad_norm": 0.9075140461657324, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 11970 + }, + { + "epoch": 0.11971, + "grad_norm": 1.0378457787723592, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 11971 + }, + { + "epoch": 0.11972, + "grad_norm": 1.1752130844340605, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 11972 + }, + { + "epoch": 0.11973, + "grad_norm": 1.0142498096983241, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 11973 + }, + { + "epoch": 0.11974, + "grad_norm": 1.1854619498914203, + "learning_rate": 0.003, + "loss": 4.062, + "step": 11974 + }, + { + "epoch": 0.11975, + "grad_norm": 1.0289773926307388, + "learning_rate": 0.003, + "loss": 4.063, + "step": 11975 + }, + { + "epoch": 0.11976, + "grad_norm": 1.0223582360777832, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 11976 + }, + { + "epoch": 0.11977, + "grad_norm": 0.8691558208987089, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 11977 + }, + { + "epoch": 0.11978, + "grad_norm": 0.9349579008470975, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 11978 + }, + { + "epoch": 0.11979, + "grad_norm": 1.1103977978184796, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 11979 + }, + { + "epoch": 0.1198, + "grad_norm": 1.098749769403459, + "learning_rate": 0.003, + "loss": 4.1105, + "step": 11980 + }, + { + "epoch": 0.11981, + "grad_norm": 1.0956451982569382, + "learning_rate": 0.003, + "loss": 4.1027, + "step": 11981 + }, + { + "epoch": 0.11982, + "grad_norm": 0.9591894182593492, + "learning_rate": 0.003, + "loss": 4.084, + "step": 11982 + }, + { + "epoch": 0.11983, + "grad_norm": 0.810203554841054, + "learning_rate": 0.003, + "loss": 4.0967, + "step": 11983 + }, + { + "epoch": 0.11984, + "grad_norm": 0.7597069908684222, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 11984 + }, + { + "epoch": 0.11985, + "grad_norm": 0.9184468653827235, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 11985 + }, + { + "epoch": 0.11986, + "grad_norm": 1.173177003666149, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 11986 + }, + { + "epoch": 0.11987, + "grad_norm": 0.8330849699905769, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 11987 + }, + { + "epoch": 0.11988, + "grad_norm": 0.7756203318593652, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 11988 + }, + { + "epoch": 0.11989, + "grad_norm": 0.8885210110071797, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 11989 + }, + { + "epoch": 0.1199, + "grad_norm": 1.0899710849399094, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 11990 + }, + { + "epoch": 0.11991, + "grad_norm": 1.0139270332750296, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 11991 + }, + { + "epoch": 0.11992, + "grad_norm": 0.9429273867943814, + "learning_rate": 0.003, + "loss": 4.1113, + "step": 11992 + }, + { + "epoch": 0.11993, + "grad_norm": 0.848073248178421, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 11993 + }, + { + "epoch": 0.11994, + "grad_norm": 0.8172158573020303, + "learning_rate": 0.003, + "loss": 4.0906, + "step": 11994 + }, + { + "epoch": 0.11995, + "grad_norm": 1.0577149995209103, + "learning_rate": 0.003, + "loss": 4.0898, + "step": 11995 + }, + { + "epoch": 0.11996, + "grad_norm": 1.0964232595215726, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 11996 + }, + { + "epoch": 0.11997, + "grad_norm": 1.2708702674073558, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 11997 + }, + { + "epoch": 0.11998, + "grad_norm": 1.0234048709297088, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 11998 + }, + { + "epoch": 0.11999, + "grad_norm": 1.088866098026185, + "learning_rate": 0.003, + "loss": 4.1041, + "step": 11999 + }, + { + "epoch": 0.12, + "grad_norm": 1.0050167202441624, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 12000 + }, + { + "epoch": 0.12001, + "grad_norm": 1.0886222441077127, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 12001 + }, + { + "epoch": 0.12002, + "grad_norm": 0.9151132927908924, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 12002 + }, + { + "epoch": 0.12003, + "grad_norm": 0.8563090723410711, + "learning_rate": 0.003, + "loss": 4.067, + "step": 12003 + }, + { + "epoch": 0.12004, + "grad_norm": 0.9036251169815563, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 12004 + }, + { + "epoch": 0.12005, + "grad_norm": 1.0369940300147027, + "learning_rate": 0.003, + "loss": 4.069, + "step": 12005 + }, + { + "epoch": 0.12006, + "grad_norm": 1.0132925987275758, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 12006 + }, + { + "epoch": 0.12007, + "grad_norm": 1.0421536488393857, + "learning_rate": 0.003, + "loss": 4.0819, + "step": 12007 + }, + { + "epoch": 0.12008, + "grad_norm": 1.0452623328137522, + "learning_rate": 0.003, + "loss": 4.112, + "step": 12008 + }, + { + "epoch": 0.12009, + "grad_norm": 0.9532328692268379, + "learning_rate": 0.003, + "loss": 4.0887, + "step": 12009 + }, + { + "epoch": 0.1201, + "grad_norm": 1.0487436151603846, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 12010 + }, + { + "epoch": 0.12011, + "grad_norm": 1.0152076110919481, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 12011 + }, + { + "epoch": 0.12012, + "grad_norm": 1.3342061087567287, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 12012 + }, + { + "epoch": 0.12013, + "grad_norm": 0.80311580533487, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 12013 + }, + { + "epoch": 0.12014, + "grad_norm": 0.7570848458182093, + "learning_rate": 0.003, + "loss": 4.071, + "step": 12014 + }, + { + "epoch": 0.12015, + "grad_norm": 0.7758610867739811, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 12015 + }, + { + "epoch": 0.12016, + "grad_norm": 0.9572357849085585, + "learning_rate": 0.003, + "loss": 4.0959, + "step": 12016 + }, + { + "epoch": 0.12017, + "grad_norm": 1.123760324524743, + "learning_rate": 0.003, + "loss": 4.1071, + "step": 12017 + }, + { + "epoch": 0.12018, + "grad_norm": 0.8369414066767921, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 12018 + }, + { + "epoch": 0.12019, + "grad_norm": 0.7986715281706961, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 12019 + }, + { + "epoch": 0.1202, + "grad_norm": 0.9357823982546265, + "learning_rate": 0.003, + "loss": 4.0955, + "step": 12020 + }, + { + "epoch": 0.12021, + "grad_norm": 1.0404823466524749, + "learning_rate": 0.003, + "loss": 4.0972, + "step": 12021 + }, + { + "epoch": 0.12022, + "grad_norm": 1.3413388212565611, + "learning_rate": 0.003, + "loss": 4.058, + "step": 12022 + }, + { + "epoch": 0.12023, + "grad_norm": 0.7488781655197063, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 12023 + }, + { + "epoch": 0.12024, + "grad_norm": 0.6291103030086327, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 12024 + }, + { + "epoch": 0.12025, + "grad_norm": 0.6249527400393582, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 12025 + }, + { + "epoch": 0.12026, + "grad_norm": 0.7364166346869837, + "learning_rate": 0.003, + "loss": 4.089, + "step": 12026 + }, + { + "epoch": 0.12027, + "grad_norm": 1.0602541407156427, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 12027 + }, + { + "epoch": 0.12028, + "grad_norm": 1.1216964746518712, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 12028 + }, + { + "epoch": 0.12029, + "grad_norm": 1.106116597582524, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 12029 + }, + { + "epoch": 0.1203, + "grad_norm": 1.0344749816296157, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 12030 + }, + { + "epoch": 0.12031, + "grad_norm": 1.0699492651388485, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 12031 + }, + { + "epoch": 0.12032, + "grad_norm": 0.9191005857264559, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 12032 + }, + { + "epoch": 0.12033, + "grad_norm": 0.9117539056182536, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 12033 + }, + { + "epoch": 0.12034, + "grad_norm": 1.1544872348261042, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 12034 + }, + { + "epoch": 0.12035, + "grad_norm": 1.0298050692368725, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 12035 + }, + { + "epoch": 0.12036, + "grad_norm": 1.0564921619229795, + "learning_rate": 0.003, + "loss": 4.0961, + "step": 12036 + }, + { + "epoch": 0.12037, + "grad_norm": 0.9785055344296542, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 12037 + }, + { + "epoch": 0.12038, + "grad_norm": 1.0411441499062488, + "learning_rate": 0.003, + "loss": 4.0931, + "step": 12038 + }, + { + "epoch": 0.12039, + "grad_norm": 0.8519044922336462, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 12039 + }, + { + "epoch": 0.1204, + "grad_norm": 0.7763435958484959, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 12040 + }, + { + "epoch": 0.12041, + "grad_norm": 0.750923044490003, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 12041 + }, + { + "epoch": 0.12042, + "grad_norm": 0.6817056101641928, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 12042 + }, + { + "epoch": 0.12043, + "grad_norm": 0.8084538289926416, + "learning_rate": 0.003, + "loss": 4.092, + "step": 12043 + }, + { + "epoch": 0.12044, + "grad_norm": 1.0137912462514658, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 12044 + }, + { + "epoch": 0.12045, + "grad_norm": 1.0728779412989322, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 12045 + }, + { + "epoch": 0.12046, + "grad_norm": 1.1410888637543153, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 12046 + }, + { + "epoch": 0.12047, + "grad_norm": 0.8830036467334388, + "learning_rate": 0.003, + "loss": 4.0959, + "step": 12047 + }, + { + "epoch": 0.12048, + "grad_norm": 0.7765900059530938, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 12048 + }, + { + "epoch": 0.12049, + "grad_norm": 0.830303944877207, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 12049 + }, + { + "epoch": 0.1205, + "grad_norm": 1.0288087238594172, + "learning_rate": 0.003, + "loss": 4.1053, + "step": 12050 + }, + { + "epoch": 0.12051, + "grad_norm": 1.254857214141687, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 12051 + }, + { + "epoch": 0.12052, + "grad_norm": 0.8665441092557049, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 12052 + }, + { + "epoch": 0.12053, + "grad_norm": 0.8961630511302968, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 12053 + }, + { + "epoch": 0.12054, + "grad_norm": 1.0146001753895153, + "learning_rate": 0.003, + "loss": 4.1018, + "step": 12054 + }, + { + "epoch": 0.12055, + "grad_norm": 1.0250118194818576, + "learning_rate": 0.003, + "loss": 4.1021, + "step": 12055 + }, + { + "epoch": 0.12056, + "grad_norm": 0.9718907311538125, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 12056 + }, + { + "epoch": 0.12057, + "grad_norm": 0.8325270754049835, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 12057 + }, + { + "epoch": 0.12058, + "grad_norm": 0.8475250669151635, + "learning_rate": 0.003, + "loss": 4.1027, + "step": 12058 + }, + { + "epoch": 0.12059, + "grad_norm": 0.8808906339030855, + "learning_rate": 0.003, + "loss": 4.1005, + "step": 12059 + }, + { + "epoch": 0.1206, + "grad_norm": 1.0497651174555434, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 12060 + }, + { + "epoch": 0.12061, + "grad_norm": 1.4111626116618436, + "learning_rate": 0.003, + "loss": 4.1155, + "step": 12061 + }, + { + "epoch": 0.12062, + "grad_norm": 0.7481992039351482, + "learning_rate": 0.003, + "loss": 4.06, + "step": 12062 + }, + { + "epoch": 0.12063, + "grad_norm": 0.8275600615495862, + "learning_rate": 0.003, + "loss": 4.0881, + "step": 12063 + }, + { + "epoch": 0.12064, + "grad_norm": 0.88678149550897, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 12064 + }, + { + "epoch": 0.12065, + "grad_norm": 0.9847133773059498, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 12065 + }, + { + "epoch": 0.12066, + "grad_norm": 1.2170462796409873, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 12066 + }, + { + "epoch": 0.12067, + "grad_norm": 1.079046628165063, + "learning_rate": 0.003, + "loss": 4.1135, + "step": 12067 + }, + { + "epoch": 0.12068, + "grad_norm": 1.0070433217978965, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 12068 + }, + { + "epoch": 0.12069, + "grad_norm": 1.0532196624412653, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 12069 + }, + { + "epoch": 0.1207, + "grad_norm": 0.8536569170599837, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 12070 + }, + { + "epoch": 0.12071, + "grad_norm": 0.8007973289038062, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 12071 + }, + { + "epoch": 0.12072, + "grad_norm": 0.8344438780484387, + "learning_rate": 0.003, + "loss": 4.1193, + "step": 12072 + }, + { + "epoch": 0.12073, + "grad_norm": 0.8156059363790338, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 12073 + }, + { + "epoch": 0.12074, + "grad_norm": 0.9133438454933838, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 12074 + }, + { + "epoch": 0.12075, + "grad_norm": 1.0085410206433918, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 12075 + }, + { + "epoch": 0.12076, + "grad_norm": 1.3468346003581002, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 12076 + }, + { + "epoch": 0.12077, + "grad_norm": 0.8700294654172432, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 12077 + }, + { + "epoch": 0.12078, + "grad_norm": 0.9342903026322471, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 12078 + }, + { + "epoch": 0.12079, + "grad_norm": 1.2264745422712207, + "learning_rate": 0.003, + "loss": 4.084, + "step": 12079 + }, + { + "epoch": 0.1208, + "grad_norm": 1.0728590610442097, + "learning_rate": 0.003, + "loss": 4.093, + "step": 12080 + }, + { + "epoch": 0.12081, + "grad_norm": 1.0146669303883087, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 12081 + }, + { + "epoch": 0.12082, + "grad_norm": 1.1870907884008515, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 12082 + }, + { + "epoch": 0.12083, + "grad_norm": 1.03956415189551, + "learning_rate": 0.003, + "loss": 4.068, + "step": 12083 + }, + { + "epoch": 0.12084, + "grad_norm": 0.8734525893399809, + "learning_rate": 0.003, + "loss": 4.0878, + "step": 12084 + }, + { + "epoch": 0.12085, + "grad_norm": 0.9115229661250986, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 12085 + }, + { + "epoch": 0.12086, + "grad_norm": 1.0309211786000259, + "learning_rate": 0.003, + "loss": 4.1016, + "step": 12086 + }, + { + "epoch": 0.12087, + "grad_norm": 1.1336131331505506, + "learning_rate": 0.003, + "loss": 4.1237, + "step": 12087 + }, + { + "epoch": 0.12088, + "grad_norm": 0.9455225773504679, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 12088 + }, + { + "epoch": 0.12089, + "grad_norm": 0.9728631386365678, + "learning_rate": 0.003, + "loss": 4.0999, + "step": 12089 + }, + { + "epoch": 0.1209, + "grad_norm": 1.0925049462478296, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 12090 + }, + { + "epoch": 0.12091, + "grad_norm": 1.05418386267452, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 12091 + }, + { + "epoch": 0.12092, + "grad_norm": 0.9676937175646085, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 12092 + }, + { + "epoch": 0.12093, + "grad_norm": 1.026209440751513, + "learning_rate": 0.003, + "loss": 4.1025, + "step": 12093 + }, + { + "epoch": 0.12094, + "grad_norm": 1.0894069079160287, + "learning_rate": 0.003, + "loss": 4.102, + "step": 12094 + }, + { + "epoch": 0.12095, + "grad_norm": 0.768137817595937, + "learning_rate": 0.003, + "loss": 4.0913, + "step": 12095 + }, + { + "epoch": 0.12096, + "grad_norm": 0.7646119968214725, + "learning_rate": 0.003, + "loss": 4.077, + "step": 12096 + }, + { + "epoch": 0.12097, + "grad_norm": 0.7159209236728756, + "learning_rate": 0.003, + "loss": 4.1041, + "step": 12097 + }, + { + "epoch": 0.12098, + "grad_norm": 0.857701592687128, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 12098 + }, + { + "epoch": 0.12099, + "grad_norm": 1.0218539208496251, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 12099 + }, + { + "epoch": 0.121, + "grad_norm": 1.2201685295722273, + "learning_rate": 0.003, + "loss": 4.071, + "step": 12100 + }, + { + "epoch": 0.12101, + "grad_norm": 0.8562643143502402, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 12101 + }, + { + "epoch": 0.12102, + "grad_norm": 0.8831313448539596, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 12102 + }, + { + "epoch": 0.12103, + "grad_norm": 0.8718571462321375, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 12103 + }, + { + "epoch": 0.12104, + "grad_norm": 0.8563130804224043, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 12104 + }, + { + "epoch": 0.12105, + "grad_norm": 0.7860252279203831, + "learning_rate": 0.003, + "loss": 4.07, + "step": 12105 + }, + { + "epoch": 0.12106, + "grad_norm": 0.8633403384810032, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 12106 + }, + { + "epoch": 0.12107, + "grad_norm": 0.8896186061865161, + "learning_rate": 0.003, + "loss": 4.1066, + "step": 12107 + }, + { + "epoch": 0.12108, + "grad_norm": 0.8274817226107964, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 12108 + }, + { + "epoch": 0.12109, + "grad_norm": 0.7797420395620885, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 12109 + }, + { + "epoch": 0.1211, + "grad_norm": 0.8352449005978791, + "learning_rate": 0.003, + "loss": 4.0916, + "step": 12110 + }, + { + "epoch": 0.12111, + "grad_norm": 1.0253289303945123, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 12111 + }, + { + "epoch": 0.12112, + "grad_norm": 1.4046477640241342, + "learning_rate": 0.003, + "loss": 4.1092, + "step": 12112 + }, + { + "epoch": 0.12113, + "grad_norm": 0.8373469364134555, + "learning_rate": 0.003, + "loss": 4.0912, + "step": 12113 + }, + { + "epoch": 0.12114, + "grad_norm": 0.8703326093969567, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 12114 + }, + { + "epoch": 0.12115, + "grad_norm": 0.8549558635110874, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 12115 + }, + { + "epoch": 0.12116, + "grad_norm": 0.8924462346090277, + "learning_rate": 0.003, + "loss": 4.1075, + "step": 12116 + }, + { + "epoch": 0.12117, + "grad_norm": 0.8632615043475872, + "learning_rate": 0.003, + "loss": 4.0998, + "step": 12117 + }, + { + "epoch": 0.12118, + "grad_norm": 1.0340785656680664, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 12118 + }, + { + "epoch": 0.12119, + "grad_norm": 1.2358514816283641, + "learning_rate": 0.003, + "loss": 4.1016, + "step": 12119 + }, + { + "epoch": 0.1212, + "grad_norm": 0.9675692337405871, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 12120 + }, + { + "epoch": 0.12121, + "grad_norm": 0.953139489921699, + "learning_rate": 0.003, + "loss": 4.0946, + "step": 12121 + }, + { + "epoch": 0.12122, + "grad_norm": 0.8632864628589392, + "learning_rate": 0.003, + "loss": 4.1008, + "step": 12122 + }, + { + "epoch": 0.12123, + "grad_norm": 0.9487844741960313, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 12123 + }, + { + "epoch": 0.12124, + "grad_norm": 1.0787729044749825, + "learning_rate": 0.003, + "loss": 4.084, + "step": 12124 + }, + { + "epoch": 0.12125, + "grad_norm": 0.9011279195663301, + "learning_rate": 0.003, + "loss": 4.0979, + "step": 12125 + }, + { + "epoch": 0.12126, + "grad_norm": 0.795410444609945, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 12126 + }, + { + "epoch": 0.12127, + "grad_norm": 0.8604626814702176, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 12127 + }, + { + "epoch": 0.12128, + "grad_norm": 1.0390578534960944, + "learning_rate": 0.003, + "loss": 4.1065, + "step": 12128 + }, + { + "epoch": 0.12129, + "grad_norm": 1.055497634384526, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 12129 + }, + { + "epoch": 0.1213, + "grad_norm": 1.0199299894129454, + "learning_rate": 0.003, + "loss": 4.0971, + "step": 12130 + }, + { + "epoch": 0.12131, + "grad_norm": 1.0772972860247136, + "learning_rate": 0.003, + "loss": 4.1032, + "step": 12131 + }, + { + "epoch": 0.12132, + "grad_norm": 1.074543209703453, + "learning_rate": 0.003, + "loss": 4.0953, + "step": 12132 + }, + { + "epoch": 0.12133, + "grad_norm": 0.8758251960366356, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 12133 + }, + { + "epoch": 0.12134, + "grad_norm": 0.8587643056416874, + "learning_rate": 0.003, + "loss": 4.0991, + "step": 12134 + }, + { + "epoch": 0.12135, + "grad_norm": 0.8051949773823324, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 12135 + }, + { + "epoch": 0.12136, + "grad_norm": 0.7700792429203638, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 12136 + }, + { + "epoch": 0.12137, + "grad_norm": 0.8037515800090013, + "learning_rate": 0.003, + "loss": 4.1139, + "step": 12137 + }, + { + "epoch": 0.12138, + "grad_norm": 0.9105057835302114, + "learning_rate": 0.003, + "loss": 4.1008, + "step": 12138 + }, + { + "epoch": 0.12139, + "grad_norm": 1.0502741567589453, + "learning_rate": 0.003, + "loss": 4.1016, + "step": 12139 + }, + { + "epoch": 0.1214, + "grad_norm": 1.1904614354368586, + "learning_rate": 0.003, + "loss": 4.106, + "step": 12140 + }, + { + "epoch": 0.12141, + "grad_norm": 0.8149895756258312, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 12141 + }, + { + "epoch": 0.12142, + "grad_norm": 0.7173406713849418, + "learning_rate": 0.003, + "loss": 4.085, + "step": 12142 + }, + { + "epoch": 0.12143, + "grad_norm": 0.7056149023343944, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 12143 + }, + { + "epoch": 0.12144, + "grad_norm": 0.7133932456938574, + "learning_rate": 0.003, + "loss": 4.0994, + "step": 12144 + }, + { + "epoch": 0.12145, + "grad_norm": 0.8708235909751089, + "learning_rate": 0.003, + "loss": 4.0946, + "step": 12145 + }, + { + "epoch": 0.12146, + "grad_norm": 1.0937616521892746, + "learning_rate": 0.003, + "loss": 4.0977, + "step": 12146 + }, + { + "epoch": 0.12147, + "grad_norm": 1.1068087058157368, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 12147 + }, + { + "epoch": 0.12148, + "grad_norm": 0.8434310354167307, + "learning_rate": 0.003, + "loss": 4.0918, + "step": 12148 + }, + { + "epoch": 0.12149, + "grad_norm": 0.7797969610776352, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 12149 + }, + { + "epoch": 0.1215, + "grad_norm": 0.8270181818547324, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 12150 + }, + { + "epoch": 0.12151, + "grad_norm": 1.0582536403784881, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 12151 + }, + { + "epoch": 0.12152, + "grad_norm": 1.02539167412742, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 12152 + }, + { + "epoch": 0.12153, + "grad_norm": 1.054280535043877, + "learning_rate": 0.003, + "loss": 4.1131, + "step": 12153 + }, + { + "epoch": 0.12154, + "grad_norm": 1.037503862180227, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 12154 + }, + { + "epoch": 0.12155, + "grad_norm": 0.9680994446998121, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 12155 + }, + { + "epoch": 0.12156, + "grad_norm": 0.7959915257353438, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 12156 + }, + { + "epoch": 0.12157, + "grad_norm": 0.8293771278842691, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 12157 + }, + { + "epoch": 0.12158, + "grad_norm": 1.0822693030682717, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 12158 + }, + { + "epoch": 0.12159, + "grad_norm": 1.0244830967179477, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 12159 + }, + { + "epoch": 0.1216, + "grad_norm": 0.800885672874286, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 12160 + }, + { + "epoch": 0.12161, + "grad_norm": 0.8936574572318592, + "learning_rate": 0.003, + "loss": 4.0871, + "step": 12161 + }, + { + "epoch": 0.12162, + "grad_norm": 0.8591280788178586, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 12162 + }, + { + "epoch": 0.12163, + "grad_norm": 0.8614825797125524, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 12163 + }, + { + "epoch": 0.12164, + "grad_norm": 0.915239837379017, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 12164 + }, + { + "epoch": 0.12165, + "grad_norm": 1.0208040550653712, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 12165 + }, + { + "epoch": 0.12166, + "grad_norm": 1.1510922171015554, + "learning_rate": 0.003, + "loss": 4.0943, + "step": 12166 + }, + { + "epoch": 0.12167, + "grad_norm": 0.8530181416913774, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 12167 + }, + { + "epoch": 0.12168, + "grad_norm": 0.9567316446039046, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 12168 + }, + { + "epoch": 0.12169, + "grad_norm": 1.3757804570087662, + "learning_rate": 0.003, + "loss": 4.1102, + "step": 12169 + }, + { + "epoch": 0.1217, + "grad_norm": 0.8153783280120215, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 12170 + }, + { + "epoch": 0.12171, + "grad_norm": 0.8914256001976482, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 12171 + }, + { + "epoch": 0.12172, + "grad_norm": 1.0715704778878445, + "learning_rate": 0.003, + "loss": 4.1004, + "step": 12172 + }, + { + "epoch": 0.12173, + "grad_norm": 1.338985193135708, + "learning_rate": 0.003, + "loss": 4.0934, + "step": 12173 + }, + { + "epoch": 0.12174, + "grad_norm": 0.847911300034504, + "learning_rate": 0.003, + "loss": 4.1058, + "step": 12174 + }, + { + "epoch": 0.12175, + "grad_norm": 0.8344378382976485, + "learning_rate": 0.003, + "loss": 4.0947, + "step": 12175 + }, + { + "epoch": 0.12176, + "grad_norm": 0.7936274729405293, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 12176 + }, + { + "epoch": 0.12177, + "grad_norm": 0.8890890857136268, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 12177 + }, + { + "epoch": 0.12178, + "grad_norm": 0.9533903170624753, + "learning_rate": 0.003, + "loss": 4.0963, + "step": 12178 + }, + { + "epoch": 0.12179, + "grad_norm": 0.9804814908367452, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 12179 + }, + { + "epoch": 0.1218, + "grad_norm": 1.0810653472006062, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 12180 + }, + { + "epoch": 0.12181, + "grad_norm": 1.09011485992177, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 12181 + }, + { + "epoch": 0.12182, + "grad_norm": 1.151057756407471, + "learning_rate": 0.003, + "loss": 4.0913, + "step": 12182 + }, + { + "epoch": 0.12183, + "grad_norm": 0.9624963089924442, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 12183 + }, + { + "epoch": 0.12184, + "grad_norm": 0.9896810546077532, + "learning_rate": 0.003, + "loss": 4.0854, + "step": 12184 + }, + { + "epoch": 0.12185, + "grad_norm": 1.0710706986027476, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 12185 + }, + { + "epoch": 0.12186, + "grad_norm": 0.8689036815150581, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 12186 + }, + { + "epoch": 0.12187, + "grad_norm": 0.8533646558914463, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 12187 + }, + { + "epoch": 0.12188, + "grad_norm": 0.8442786285991916, + "learning_rate": 0.003, + "loss": 4.0856, + "step": 12188 + }, + { + "epoch": 0.12189, + "grad_norm": 0.7258320269376084, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 12189 + }, + { + "epoch": 0.1219, + "grad_norm": 0.7493007111714942, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 12190 + }, + { + "epoch": 0.12191, + "grad_norm": 0.8126260820448451, + "learning_rate": 0.003, + "loss": 4.0977, + "step": 12191 + }, + { + "epoch": 0.12192, + "grad_norm": 0.8094901202289994, + "learning_rate": 0.003, + "loss": 4.0931, + "step": 12192 + }, + { + "epoch": 0.12193, + "grad_norm": 0.8661535420117265, + "learning_rate": 0.003, + "loss": 4.105, + "step": 12193 + }, + { + "epoch": 0.12194, + "grad_norm": 0.9932294938197798, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 12194 + }, + { + "epoch": 0.12195, + "grad_norm": 1.1877046195989915, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 12195 + }, + { + "epoch": 0.12196, + "grad_norm": 0.977874497575908, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 12196 + }, + { + "epoch": 0.12197, + "grad_norm": 0.9832582186492963, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 12197 + }, + { + "epoch": 0.12198, + "grad_norm": 1.11137238631857, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 12198 + }, + { + "epoch": 0.12199, + "grad_norm": 1.1493894035482843, + "learning_rate": 0.003, + "loss": 4.0966, + "step": 12199 + }, + { + "epoch": 0.122, + "grad_norm": 0.9764898412591378, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 12200 + }, + { + "epoch": 0.12201, + "grad_norm": 0.7957969030315891, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 12201 + }, + { + "epoch": 0.12202, + "grad_norm": 0.7169666505283141, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 12202 + }, + { + "epoch": 0.12203, + "grad_norm": 0.9042095799104531, + "learning_rate": 0.003, + "loss": 4.0967, + "step": 12203 + }, + { + "epoch": 0.12204, + "grad_norm": 1.2040182593600277, + "learning_rate": 0.003, + "loss": 4.1087, + "step": 12204 + }, + { + "epoch": 0.12205, + "grad_norm": 1.0021875784105758, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 12205 + }, + { + "epoch": 0.12206, + "grad_norm": 1.1193843788562738, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 12206 + }, + { + "epoch": 0.12207, + "grad_norm": 0.9568468636954698, + "learning_rate": 0.003, + "loss": 4.1225, + "step": 12207 + }, + { + "epoch": 0.12208, + "grad_norm": 1.2108633453131405, + "learning_rate": 0.003, + "loss": 4.1014, + "step": 12208 + }, + { + "epoch": 0.12209, + "grad_norm": 0.9013616347545388, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 12209 + }, + { + "epoch": 0.1221, + "grad_norm": 0.7583743037343729, + "learning_rate": 0.003, + "loss": 4.075, + "step": 12210 + }, + { + "epoch": 0.12211, + "grad_norm": 0.854484142253613, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 12211 + }, + { + "epoch": 0.12212, + "grad_norm": 0.8648448002284472, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 12212 + }, + { + "epoch": 0.12213, + "grad_norm": 0.9922574474562244, + "learning_rate": 0.003, + "loss": 4.0907, + "step": 12213 + }, + { + "epoch": 0.12214, + "grad_norm": 1.1961839655026492, + "learning_rate": 0.003, + "loss": 4.0931, + "step": 12214 + }, + { + "epoch": 0.12215, + "grad_norm": 0.9720054291028379, + "learning_rate": 0.003, + "loss": 4.1045, + "step": 12215 + }, + { + "epoch": 0.12216, + "grad_norm": 0.9573012828567232, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 12216 + }, + { + "epoch": 0.12217, + "grad_norm": 0.9020601348240719, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 12217 + }, + { + "epoch": 0.12218, + "grad_norm": 0.8549878172086276, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 12218 + }, + { + "epoch": 0.12219, + "grad_norm": 0.9187850244117227, + "learning_rate": 0.003, + "loss": 4.1038, + "step": 12219 + }, + { + "epoch": 0.1222, + "grad_norm": 0.9243833200004632, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 12220 + }, + { + "epoch": 0.12221, + "grad_norm": 1.0602967292132086, + "learning_rate": 0.003, + "loss": 4.091, + "step": 12221 + }, + { + "epoch": 0.12222, + "grad_norm": 0.9570973024725998, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 12222 + }, + { + "epoch": 0.12223, + "grad_norm": 1.0670567448752062, + "learning_rate": 0.003, + "loss": 4.0947, + "step": 12223 + }, + { + "epoch": 0.12224, + "grad_norm": 0.9757119829737123, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 12224 + }, + { + "epoch": 0.12225, + "grad_norm": 1.0605725184568129, + "learning_rate": 0.003, + "loss": 4.0905, + "step": 12225 + }, + { + "epoch": 0.12226, + "grad_norm": 1.094605169120047, + "learning_rate": 0.003, + "loss": 4.0935, + "step": 12226 + }, + { + "epoch": 0.12227, + "grad_norm": 0.937618419970385, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 12227 + }, + { + "epoch": 0.12228, + "grad_norm": 0.9032461084194371, + "learning_rate": 0.003, + "loss": 4.048, + "step": 12228 + }, + { + "epoch": 0.12229, + "grad_norm": 0.9697170532787076, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 12229 + }, + { + "epoch": 0.1223, + "grad_norm": 1.0105320205545736, + "learning_rate": 0.003, + "loss": 4.1099, + "step": 12230 + }, + { + "epoch": 0.12231, + "grad_norm": 1.139161710799362, + "learning_rate": 0.003, + "loss": 4.086, + "step": 12231 + }, + { + "epoch": 0.12232, + "grad_norm": 0.7821848752129718, + "learning_rate": 0.003, + "loss": 4.072, + "step": 12232 + }, + { + "epoch": 0.12233, + "grad_norm": 0.8023912162816685, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 12233 + }, + { + "epoch": 0.12234, + "grad_norm": 0.956616085749843, + "learning_rate": 0.003, + "loss": 4.1061, + "step": 12234 + }, + { + "epoch": 0.12235, + "grad_norm": 1.283200726795276, + "learning_rate": 0.003, + "loss": 4.073, + "step": 12235 + }, + { + "epoch": 0.12236, + "grad_norm": 0.7424860234635055, + "learning_rate": 0.003, + "loss": 4.1001, + "step": 12236 + }, + { + "epoch": 0.12237, + "grad_norm": 0.742929815197606, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 12237 + }, + { + "epoch": 0.12238, + "grad_norm": 0.8407333581296079, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 12238 + }, + { + "epoch": 0.12239, + "grad_norm": 0.957707898235125, + "learning_rate": 0.003, + "loss": 4.0918, + "step": 12239 + }, + { + "epoch": 0.1224, + "grad_norm": 1.1463936004212105, + "learning_rate": 0.003, + "loss": 4.101, + "step": 12240 + }, + { + "epoch": 0.12241, + "grad_norm": 0.8263727676898824, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 12241 + }, + { + "epoch": 0.12242, + "grad_norm": 0.6837352027047172, + "learning_rate": 0.003, + "loss": 4.077, + "step": 12242 + }, + { + "epoch": 0.12243, + "grad_norm": 0.7529887748185018, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 12243 + }, + { + "epoch": 0.12244, + "grad_norm": 0.826014021940816, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 12244 + }, + { + "epoch": 0.12245, + "grad_norm": 1.2524161035219004, + "learning_rate": 0.003, + "loss": 4.1069, + "step": 12245 + }, + { + "epoch": 0.12246, + "grad_norm": 0.9342066702650124, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 12246 + }, + { + "epoch": 0.12247, + "grad_norm": 0.8199906854707102, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 12247 + }, + { + "epoch": 0.12248, + "grad_norm": 0.7599921908260181, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 12248 + }, + { + "epoch": 0.12249, + "grad_norm": 0.8356440637371829, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 12249 + }, + { + "epoch": 0.1225, + "grad_norm": 0.8457528060731644, + "learning_rate": 0.003, + "loss": 4.0944, + "step": 12250 + }, + { + "epoch": 0.12251, + "grad_norm": 0.8694262874795654, + "learning_rate": 0.003, + "loss": 4.0976, + "step": 12251 + }, + { + "epoch": 0.12252, + "grad_norm": 1.0646473768193867, + "learning_rate": 0.003, + "loss": 4.0874, + "step": 12252 + }, + { + "epoch": 0.12253, + "grad_norm": 1.0714434386407747, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 12253 + }, + { + "epoch": 0.12254, + "grad_norm": 0.9573494951825994, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 12254 + }, + { + "epoch": 0.12255, + "grad_norm": 1.1051589997595892, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 12255 + }, + { + "epoch": 0.12256, + "grad_norm": 1.061774014425716, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 12256 + }, + { + "epoch": 0.12257, + "grad_norm": 0.9176082108977487, + "learning_rate": 0.003, + "loss": 4.1037, + "step": 12257 + }, + { + "epoch": 0.12258, + "grad_norm": 0.9026623457007, + "learning_rate": 0.003, + "loss": 4.069, + "step": 12258 + }, + { + "epoch": 0.12259, + "grad_norm": 0.9593006637959552, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 12259 + }, + { + "epoch": 0.1226, + "grad_norm": 1.0097669182369946, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 12260 + }, + { + "epoch": 0.12261, + "grad_norm": 1.0074266226750492, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 12261 + }, + { + "epoch": 0.12262, + "grad_norm": 0.9816437763872888, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 12262 + }, + { + "epoch": 0.12263, + "grad_norm": 1.1668893367216095, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 12263 + }, + { + "epoch": 0.12264, + "grad_norm": 0.9510792002071541, + "learning_rate": 0.003, + "loss": 4.0929, + "step": 12264 + }, + { + "epoch": 0.12265, + "grad_norm": 0.8496029214305116, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 12265 + }, + { + "epoch": 0.12266, + "grad_norm": 0.8128217592982511, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 12266 + }, + { + "epoch": 0.12267, + "grad_norm": 0.7973054854598843, + "learning_rate": 0.003, + "loss": 4.0929, + "step": 12267 + }, + { + "epoch": 0.12268, + "grad_norm": 1.008379647775467, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 12268 + }, + { + "epoch": 0.12269, + "grad_norm": 0.895656836292017, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 12269 + }, + { + "epoch": 0.1227, + "grad_norm": 0.8910694524793511, + "learning_rate": 0.003, + "loss": 4.0868, + "step": 12270 + }, + { + "epoch": 0.12271, + "grad_norm": 0.9937971228768797, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 12271 + }, + { + "epoch": 0.12272, + "grad_norm": 1.2982402691332013, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 12272 + }, + { + "epoch": 0.12273, + "grad_norm": 0.7684533861324893, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 12273 + }, + { + "epoch": 0.12274, + "grad_norm": 0.7096185024693779, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 12274 + }, + { + "epoch": 0.12275, + "grad_norm": 0.8588489150838406, + "learning_rate": 0.003, + "loss": 4.1016, + "step": 12275 + }, + { + "epoch": 0.12276, + "grad_norm": 0.9794550680334978, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 12276 + }, + { + "epoch": 0.12277, + "grad_norm": 0.9599279001739706, + "learning_rate": 0.003, + "loss": 4.0915, + "step": 12277 + }, + { + "epoch": 0.12278, + "grad_norm": 1.0201898133589173, + "learning_rate": 0.003, + "loss": 4.0968, + "step": 12278 + }, + { + "epoch": 0.12279, + "grad_norm": 1.061752564114531, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 12279 + }, + { + "epoch": 0.1228, + "grad_norm": 0.9622282622070477, + "learning_rate": 0.003, + "loss": 4.091, + "step": 12280 + }, + { + "epoch": 0.12281, + "grad_norm": 0.991311699106332, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 12281 + }, + { + "epoch": 0.12282, + "grad_norm": 1.258580175679114, + "learning_rate": 0.003, + "loss": 4.124, + "step": 12282 + }, + { + "epoch": 0.12283, + "grad_norm": 0.9447786299133054, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 12283 + }, + { + "epoch": 0.12284, + "grad_norm": 0.9836768239051336, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 12284 + }, + { + "epoch": 0.12285, + "grad_norm": 0.9691142082096343, + "learning_rate": 0.003, + "loss": 4.0952, + "step": 12285 + }, + { + "epoch": 0.12286, + "grad_norm": 1.0089760843948727, + "learning_rate": 0.003, + "loss": 4.0946, + "step": 12286 + }, + { + "epoch": 0.12287, + "grad_norm": 1.1110604764835863, + "learning_rate": 0.003, + "loss": 4.0997, + "step": 12287 + }, + { + "epoch": 0.12288, + "grad_norm": 1.0717731094613792, + "learning_rate": 0.003, + "loss": 4.109, + "step": 12288 + }, + { + "epoch": 0.12289, + "grad_norm": 1.190903254212888, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 12289 + }, + { + "epoch": 0.1229, + "grad_norm": 0.8561543915128847, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 12290 + }, + { + "epoch": 0.12291, + "grad_norm": 0.834838125086023, + "learning_rate": 0.003, + "loss": 4.0857, + "step": 12291 + }, + { + "epoch": 0.12292, + "grad_norm": 0.7837569454460322, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 12292 + }, + { + "epoch": 0.12293, + "grad_norm": 0.8257986436107122, + "learning_rate": 0.003, + "loss": 4.0954, + "step": 12293 + }, + { + "epoch": 0.12294, + "grad_norm": 0.9890351540497304, + "learning_rate": 0.003, + "loss": 4.0983, + "step": 12294 + }, + { + "epoch": 0.12295, + "grad_norm": 1.1879562331038451, + "learning_rate": 0.003, + "loss": 4.1039, + "step": 12295 + }, + { + "epoch": 0.12296, + "grad_norm": 1.0293880709391436, + "learning_rate": 0.003, + "loss": 4.0903, + "step": 12296 + }, + { + "epoch": 0.12297, + "grad_norm": 1.158199502735727, + "learning_rate": 0.003, + "loss": 4.082, + "step": 12297 + }, + { + "epoch": 0.12298, + "grad_norm": 0.9659108813053136, + "learning_rate": 0.003, + "loss": 4.0977, + "step": 12298 + }, + { + "epoch": 0.12299, + "grad_norm": 1.0442358357289654, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 12299 + }, + { + "epoch": 0.123, + "grad_norm": 1.226969204259032, + "learning_rate": 0.003, + "loss": 4.1004, + "step": 12300 + }, + { + "epoch": 0.12301, + "grad_norm": 0.7745303080572817, + "learning_rate": 0.003, + "loss": 4.1036, + "step": 12301 + }, + { + "epoch": 0.12302, + "grad_norm": 0.6633446010527331, + "learning_rate": 0.003, + "loss": 4.0901, + "step": 12302 + }, + { + "epoch": 0.12303, + "grad_norm": 0.6267122944606794, + "learning_rate": 0.003, + "loss": 4.1399, + "step": 12303 + }, + { + "epoch": 0.12304, + "grad_norm": 0.776882459637047, + "learning_rate": 0.003, + "loss": 4.057, + "step": 12304 + }, + { + "epoch": 0.12305, + "grad_norm": 1.0838739177866228, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 12305 + }, + { + "epoch": 0.12306, + "grad_norm": 1.2175331352571799, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 12306 + }, + { + "epoch": 0.12307, + "grad_norm": 0.7882595623445435, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 12307 + }, + { + "epoch": 0.12308, + "grad_norm": 0.7283665662086227, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 12308 + }, + { + "epoch": 0.12309, + "grad_norm": 0.8772273244564339, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 12309 + }, + { + "epoch": 0.1231, + "grad_norm": 1.1182512514780965, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 12310 + }, + { + "epoch": 0.12311, + "grad_norm": 1.086541535392881, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 12311 + }, + { + "epoch": 0.12312, + "grad_norm": 0.918833337170936, + "learning_rate": 0.003, + "loss": 4.0947, + "step": 12312 + }, + { + "epoch": 0.12313, + "grad_norm": 0.9832388191763729, + "learning_rate": 0.003, + "loss": 4.1047, + "step": 12313 + }, + { + "epoch": 0.12314, + "grad_norm": 0.9372937403296949, + "learning_rate": 0.003, + "loss": 4.075, + "step": 12314 + }, + { + "epoch": 0.12315, + "grad_norm": 0.886243034346975, + "learning_rate": 0.003, + "loss": 4.1019, + "step": 12315 + }, + { + "epoch": 0.12316, + "grad_norm": 0.9561223610573798, + "learning_rate": 0.003, + "loss": 4.0912, + "step": 12316 + }, + { + "epoch": 0.12317, + "grad_norm": 1.056014899923094, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 12317 + }, + { + "epoch": 0.12318, + "grad_norm": 0.8337062959052595, + "learning_rate": 0.003, + "loss": 4.0994, + "step": 12318 + }, + { + "epoch": 0.12319, + "grad_norm": 0.9417512640871896, + "learning_rate": 0.003, + "loss": 4.087, + "step": 12319 + }, + { + "epoch": 0.1232, + "grad_norm": 0.9675895873880896, + "learning_rate": 0.003, + "loss": 4.1085, + "step": 12320 + }, + { + "epoch": 0.12321, + "grad_norm": 0.9977918062034861, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 12321 + }, + { + "epoch": 0.12322, + "grad_norm": 1.0933383087626054, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 12322 + }, + { + "epoch": 0.12323, + "grad_norm": 0.9386774181113615, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 12323 + }, + { + "epoch": 0.12324, + "grad_norm": 1.250430835140608, + "learning_rate": 0.003, + "loss": 4.081, + "step": 12324 + }, + { + "epoch": 0.12325, + "grad_norm": 0.9487980370711807, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 12325 + }, + { + "epoch": 0.12326, + "grad_norm": 1.0751603798827223, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 12326 + }, + { + "epoch": 0.12327, + "grad_norm": 1.2801778921389038, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 12327 + }, + { + "epoch": 0.12328, + "grad_norm": 1.0431968563424647, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 12328 + }, + { + "epoch": 0.12329, + "grad_norm": 0.8136430379954736, + "learning_rate": 0.003, + "loss": 4.0985, + "step": 12329 + }, + { + "epoch": 0.1233, + "grad_norm": 0.709681805565897, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 12330 + }, + { + "epoch": 0.12331, + "grad_norm": 0.8872455396864791, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 12331 + }, + { + "epoch": 0.12332, + "grad_norm": 1.0469277592723474, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 12332 + }, + { + "epoch": 0.12333, + "grad_norm": 0.9927148850791283, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 12333 + }, + { + "epoch": 0.12334, + "grad_norm": 1.0588678018662647, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 12334 + }, + { + "epoch": 0.12335, + "grad_norm": 0.9978916604285389, + "learning_rate": 0.003, + "loss": 4.083, + "step": 12335 + }, + { + "epoch": 0.12336, + "grad_norm": 1.0160926775589822, + "learning_rate": 0.003, + "loss": 4.084, + "step": 12336 + }, + { + "epoch": 0.12337, + "grad_norm": 1.1798101629517939, + "learning_rate": 0.003, + "loss": 4.0872, + "step": 12337 + }, + { + "epoch": 0.12338, + "grad_norm": 0.7598460859506974, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 12338 + }, + { + "epoch": 0.12339, + "grad_norm": 0.7767275051034408, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 12339 + }, + { + "epoch": 0.1234, + "grad_norm": 0.7352023418044594, + "learning_rate": 0.003, + "loss": 4.1024, + "step": 12340 + }, + { + "epoch": 0.12341, + "grad_norm": 0.9770352091346023, + "learning_rate": 0.003, + "loss": 4.0932, + "step": 12341 + }, + { + "epoch": 0.12342, + "grad_norm": 1.2983160180085074, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 12342 + }, + { + "epoch": 0.12343, + "grad_norm": 0.8660686759200508, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 12343 + }, + { + "epoch": 0.12344, + "grad_norm": 0.9455833661382927, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 12344 + }, + { + "epoch": 0.12345, + "grad_norm": 0.9323401019496114, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 12345 + }, + { + "epoch": 0.12346, + "grad_norm": 0.7469391433040231, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 12346 + }, + { + "epoch": 0.12347, + "grad_norm": 0.781751347897037, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 12347 + }, + { + "epoch": 0.12348, + "grad_norm": 0.8240849239894326, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 12348 + }, + { + "epoch": 0.12349, + "grad_norm": 0.9546002970854781, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 12349 + }, + { + "epoch": 0.1235, + "grad_norm": 0.9530112722886457, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 12350 + }, + { + "epoch": 0.12351, + "grad_norm": 0.9423712080204297, + "learning_rate": 0.003, + "loss": 4.0919, + "step": 12351 + }, + { + "epoch": 0.12352, + "grad_norm": 0.9179300920133866, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 12352 + }, + { + "epoch": 0.12353, + "grad_norm": 0.9788831729480706, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 12353 + }, + { + "epoch": 0.12354, + "grad_norm": 1.183767579885599, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 12354 + }, + { + "epoch": 0.12355, + "grad_norm": 1.0830156094340704, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 12355 + }, + { + "epoch": 0.12356, + "grad_norm": 1.1067972368788623, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 12356 + }, + { + "epoch": 0.12357, + "grad_norm": 0.9701622977641973, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 12357 + }, + { + "epoch": 0.12358, + "grad_norm": 1.1002819354273858, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 12358 + }, + { + "epoch": 0.12359, + "grad_norm": 0.963370820639486, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 12359 + }, + { + "epoch": 0.1236, + "grad_norm": 0.9192534252553537, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 12360 + }, + { + "epoch": 0.12361, + "grad_norm": 0.9268681696335273, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 12361 + }, + { + "epoch": 0.12362, + "grad_norm": 1.0145457060213716, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 12362 + }, + { + "epoch": 0.12363, + "grad_norm": 1.093217959685353, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 12363 + }, + { + "epoch": 0.12364, + "grad_norm": 0.9389100727988615, + "learning_rate": 0.003, + "loss": 4.07, + "step": 12364 + }, + { + "epoch": 0.12365, + "grad_norm": 0.9296427177726995, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 12365 + }, + { + "epoch": 0.12366, + "grad_norm": 0.9298330269189934, + "learning_rate": 0.003, + "loss": 4.1315, + "step": 12366 + }, + { + "epoch": 0.12367, + "grad_norm": 1.1017403180535674, + "learning_rate": 0.003, + "loss": 4.1038, + "step": 12367 + }, + { + "epoch": 0.12368, + "grad_norm": 1.0916868626409408, + "learning_rate": 0.003, + "loss": 4.0893, + "step": 12368 + }, + { + "epoch": 0.12369, + "grad_norm": 0.8902628928702246, + "learning_rate": 0.003, + "loss": 4.0994, + "step": 12369 + }, + { + "epoch": 0.1237, + "grad_norm": 0.8212448613736568, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 12370 + }, + { + "epoch": 0.12371, + "grad_norm": 0.9054580886364202, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 12371 + }, + { + "epoch": 0.12372, + "grad_norm": 0.9396042450044189, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 12372 + }, + { + "epoch": 0.12373, + "grad_norm": 0.9390933983811288, + "learning_rate": 0.003, + "loss": 4.1, + "step": 12373 + }, + { + "epoch": 0.12374, + "grad_norm": 1.1693463451852468, + "learning_rate": 0.003, + "loss": 4.0956, + "step": 12374 + }, + { + "epoch": 0.12375, + "grad_norm": 0.863341734329042, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 12375 + }, + { + "epoch": 0.12376, + "grad_norm": 0.9618916891880944, + "learning_rate": 0.003, + "loss": 4.104, + "step": 12376 + }, + { + "epoch": 0.12377, + "grad_norm": 1.1126375134746849, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 12377 + }, + { + "epoch": 0.12378, + "grad_norm": 1.2980499196087685, + "learning_rate": 0.003, + "loss": 4.1173, + "step": 12378 + }, + { + "epoch": 0.12379, + "grad_norm": 0.9132467632155551, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 12379 + }, + { + "epoch": 0.1238, + "grad_norm": 0.8705804603518105, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 12380 + }, + { + "epoch": 0.12381, + "grad_norm": 0.7447103480976989, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 12381 + }, + { + "epoch": 0.12382, + "grad_norm": 0.7475890297423701, + "learning_rate": 0.003, + "loss": 4.071, + "step": 12382 + }, + { + "epoch": 0.12383, + "grad_norm": 0.7462646736050489, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 12383 + }, + { + "epoch": 0.12384, + "grad_norm": 0.7208135512599897, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 12384 + }, + { + "epoch": 0.12385, + "grad_norm": 0.7772464193855457, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 12385 + }, + { + "epoch": 0.12386, + "grad_norm": 1.0532202339193975, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 12386 + }, + { + "epoch": 0.12387, + "grad_norm": 1.2643841706239745, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 12387 + }, + { + "epoch": 0.12388, + "grad_norm": 0.8757530436910678, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 12388 + }, + { + "epoch": 0.12389, + "grad_norm": 0.7614682588762037, + "learning_rate": 0.003, + "loss": 4.0984, + "step": 12389 + }, + { + "epoch": 0.1239, + "grad_norm": 0.6817872351645037, + "learning_rate": 0.003, + "loss": 4.1128, + "step": 12390 + }, + { + "epoch": 0.12391, + "grad_norm": 0.6926286312434362, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 12391 + }, + { + "epoch": 0.12392, + "grad_norm": 0.6483908570205241, + "learning_rate": 0.003, + "loss": 4.062, + "step": 12392 + }, + { + "epoch": 0.12393, + "grad_norm": 0.5750842761160463, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 12393 + }, + { + "epoch": 0.12394, + "grad_norm": 0.6065432484056245, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 12394 + }, + { + "epoch": 0.12395, + "grad_norm": 0.6068824660787948, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 12395 + }, + { + "epoch": 0.12396, + "grad_norm": 0.6552268290947781, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 12396 + }, + { + "epoch": 0.12397, + "grad_norm": 0.8076846227142461, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 12397 + }, + { + "epoch": 0.12398, + "grad_norm": 1.212072440113358, + "learning_rate": 0.003, + "loss": 4.1043, + "step": 12398 + }, + { + "epoch": 0.12399, + "grad_norm": 1.023986304820749, + "learning_rate": 0.003, + "loss": 4.1039, + "step": 12399 + }, + { + "epoch": 0.124, + "grad_norm": 1.0115354230557836, + "learning_rate": 0.003, + "loss": 4.1156, + "step": 12400 + }, + { + "epoch": 0.12401, + "grad_norm": 1.165129343793754, + "learning_rate": 0.003, + "loss": 4.056, + "step": 12401 + }, + { + "epoch": 0.12402, + "grad_norm": 0.961487910874386, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 12402 + }, + { + "epoch": 0.12403, + "grad_norm": 1.0670785474847024, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 12403 + }, + { + "epoch": 0.12404, + "grad_norm": 1.1237151866936028, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 12404 + }, + { + "epoch": 0.12405, + "grad_norm": 1.1000091221859538, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 12405 + }, + { + "epoch": 0.12406, + "grad_norm": 0.8457814279310029, + "learning_rate": 0.003, + "loss": 4.066, + "step": 12406 + }, + { + "epoch": 0.12407, + "grad_norm": 0.8384918677838943, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 12407 + }, + { + "epoch": 0.12408, + "grad_norm": 0.8763793274468791, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 12408 + }, + { + "epoch": 0.12409, + "grad_norm": 0.8662227732981806, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 12409 + }, + { + "epoch": 0.1241, + "grad_norm": 1.0802203776017565, + "learning_rate": 0.003, + "loss": 4.0938, + "step": 12410 + }, + { + "epoch": 0.12411, + "grad_norm": 1.1639505599550386, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 12411 + }, + { + "epoch": 0.12412, + "grad_norm": 0.8079535752501046, + "learning_rate": 0.003, + "loss": 4.049, + "step": 12412 + }, + { + "epoch": 0.12413, + "grad_norm": 0.9651242131182983, + "learning_rate": 0.003, + "loss": 4.1021, + "step": 12413 + }, + { + "epoch": 0.12414, + "grad_norm": 1.0101629758673436, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 12414 + }, + { + "epoch": 0.12415, + "grad_norm": 1.2021002166958954, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 12415 + }, + { + "epoch": 0.12416, + "grad_norm": 1.057741710995768, + "learning_rate": 0.003, + "loss": 4.0997, + "step": 12416 + }, + { + "epoch": 0.12417, + "grad_norm": 1.0779511025502209, + "learning_rate": 0.003, + "loss": 4.0968, + "step": 12417 + }, + { + "epoch": 0.12418, + "grad_norm": 0.8855498967356863, + "learning_rate": 0.003, + "loss": 4.0868, + "step": 12418 + }, + { + "epoch": 0.12419, + "grad_norm": 0.9967318621791877, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 12419 + }, + { + "epoch": 0.1242, + "grad_norm": 1.1424010739254968, + "learning_rate": 0.003, + "loss": 4.059, + "step": 12420 + }, + { + "epoch": 0.12421, + "grad_norm": 1.1063599072651649, + "learning_rate": 0.003, + "loss": 4.108, + "step": 12421 + }, + { + "epoch": 0.12422, + "grad_norm": 1.1434639622188416, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 12422 + }, + { + "epoch": 0.12423, + "grad_norm": 0.8138073732378587, + "learning_rate": 0.003, + "loss": 4.0901, + "step": 12423 + }, + { + "epoch": 0.12424, + "grad_norm": 0.8185307101152125, + "learning_rate": 0.003, + "loss": 4.041, + "step": 12424 + }, + { + "epoch": 0.12425, + "grad_norm": 0.9198583027147832, + "learning_rate": 0.003, + "loss": 4.07, + "step": 12425 + }, + { + "epoch": 0.12426, + "grad_norm": 1.0429474741632971, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 12426 + }, + { + "epoch": 0.12427, + "grad_norm": 0.9656271908229809, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 12427 + }, + { + "epoch": 0.12428, + "grad_norm": 0.9103784267109477, + "learning_rate": 0.003, + "loss": 4.0979, + "step": 12428 + }, + { + "epoch": 0.12429, + "grad_norm": 0.8836694349953166, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 12429 + }, + { + "epoch": 0.1243, + "grad_norm": 0.8199481798303857, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 12430 + }, + { + "epoch": 0.12431, + "grad_norm": 0.9196425610745478, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 12431 + }, + { + "epoch": 0.12432, + "grad_norm": 1.0761451449506965, + "learning_rate": 0.003, + "loss": 4.0994, + "step": 12432 + }, + { + "epoch": 0.12433, + "grad_norm": 1.0016302818900802, + "learning_rate": 0.003, + "loss": 4.0992, + "step": 12433 + }, + { + "epoch": 0.12434, + "grad_norm": 1.0621016788599498, + "learning_rate": 0.003, + "loss": 4.1198, + "step": 12434 + }, + { + "epoch": 0.12435, + "grad_norm": 1.1285102803417262, + "learning_rate": 0.003, + "loss": 4.0886, + "step": 12435 + }, + { + "epoch": 0.12436, + "grad_norm": 1.1079668184363443, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 12436 + }, + { + "epoch": 0.12437, + "grad_norm": 1.1540313002628149, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 12437 + }, + { + "epoch": 0.12438, + "grad_norm": 1.0536174878293607, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 12438 + }, + { + "epoch": 0.12439, + "grad_norm": 1.0878611376231289, + "learning_rate": 0.003, + "loss": 4.093, + "step": 12439 + }, + { + "epoch": 0.1244, + "grad_norm": 1.067929959721543, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 12440 + }, + { + "epoch": 0.12441, + "grad_norm": 0.8780190337092976, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 12441 + }, + { + "epoch": 0.12442, + "grad_norm": 0.9167319059679551, + "learning_rate": 0.003, + "loss": 4.0893, + "step": 12442 + }, + { + "epoch": 0.12443, + "grad_norm": 0.9084157239820466, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 12443 + }, + { + "epoch": 0.12444, + "grad_norm": 0.8819426975009469, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 12444 + }, + { + "epoch": 0.12445, + "grad_norm": 0.9215057869734726, + "learning_rate": 0.003, + "loss": 4.0923, + "step": 12445 + }, + { + "epoch": 0.12446, + "grad_norm": 0.946732492025596, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 12446 + }, + { + "epoch": 0.12447, + "grad_norm": 0.9077277042652085, + "learning_rate": 0.003, + "loss": 4.1039, + "step": 12447 + }, + { + "epoch": 0.12448, + "grad_norm": 0.9496515352177012, + "learning_rate": 0.003, + "loss": 4.0976, + "step": 12448 + }, + { + "epoch": 0.12449, + "grad_norm": 1.023614479672272, + "learning_rate": 0.003, + "loss": 4.1241, + "step": 12449 + }, + { + "epoch": 0.1245, + "grad_norm": 1.0232740472209256, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 12450 + }, + { + "epoch": 0.12451, + "grad_norm": 1.0433322851299396, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 12451 + }, + { + "epoch": 0.12452, + "grad_norm": 1.1592014880204855, + "learning_rate": 0.003, + "loss": 4.1056, + "step": 12452 + }, + { + "epoch": 0.12453, + "grad_norm": 1.0275527604502308, + "learning_rate": 0.003, + "loss": 4.1053, + "step": 12453 + }, + { + "epoch": 0.12454, + "grad_norm": 1.1490655916551646, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 12454 + }, + { + "epoch": 0.12455, + "grad_norm": 1.003952968275617, + "learning_rate": 0.003, + "loss": 4.092, + "step": 12455 + }, + { + "epoch": 0.12456, + "grad_norm": 1.2590151212562803, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 12456 + }, + { + "epoch": 0.12457, + "grad_norm": 0.8802352583688532, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 12457 + }, + { + "epoch": 0.12458, + "grad_norm": 0.7500633906238446, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 12458 + }, + { + "epoch": 0.12459, + "grad_norm": 0.7455453195652659, + "learning_rate": 0.003, + "loss": 4.0966, + "step": 12459 + }, + { + "epoch": 0.1246, + "grad_norm": 0.7895719581040442, + "learning_rate": 0.003, + "loss": 4.1028, + "step": 12460 + }, + { + "epoch": 0.12461, + "grad_norm": 0.8120586901219388, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 12461 + }, + { + "epoch": 0.12462, + "grad_norm": 0.8177351662577136, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 12462 + }, + { + "epoch": 0.12463, + "grad_norm": 0.8560342572280737, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 12463 + }, + { + "epoch": 0.12464, + "grad_norm": 0.8395142161379134, + "learning_rate": 0.003, + "loss": 4.0893, + "step": 12464 + }, + { + "epoch": 0.12465, + "grad_norm": 0.9744105035971836, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 12465 + }, + { + "epoch": 0.12466, + "grad_norm": 1.1426280521145489, + "learning_rate": 0.003, + "loss": 4.1043, + "step": 12466 + }, + { + "epoch": 0.12467, + "grad_norm": 1.1326091102138858, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 12467 + }, + { + "epoch": 0.12468, + "grad_norm": 1.2612232760354865, + "learning_rate": 0.003, + "loss": 4.1069, + "step": 12468 + }, + { + "epoch": 0.12469, + "grad_norm": 0.8171693611015508, + "learning_rate": 0.003, + "loss": 4.059, + "step": 12469 + }, + { + "epoch": 0.1247, + "grad_norm": 0.7457832937675054, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 12470 + }, + { + "epoch": 0.12471, + "grad_norm": 0.7985372587966096, + "learning_rate": 0.003, + "loss": 4.0938, + "step": 12471 + }, + { + "epoch": 0.12472, + "grad_norm": 1.0211061220833, + "learning_rate": 0.003, + "loss": 4.0932, + "step": 12472 + }, + { + "epoch": 0.12473, + "grad_norm": 1.3726131860312414, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 12473 + }, + { + "epoch": 0.12474, + "grad_norm": 0.8457629146545124, + "learning_rate": 0.003, + "loss": 4.0873, + "step": 12474 + }, + { + "epoch": 0.12475, + "grad_norm": 0.7072389787492686, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 12475 + }, + { + "epoch": 0.12476, + "grad_norm": 0.8137661418092231, + "learning_rate": 0.003, + "loss": 4.1008, + "step": 12476 + }, + { + "epoch": 0.12477, + "grad_norm": 0.9217456366863254, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 12477 + }, + { + "epoch": 0.12478, + "grad_norm": 0.9812146327470057, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 12478 + }, + { + "epoch": 0.12479, + "grad_norm": 1.042986401881006, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 12479 + }, + { + "epoch": 0.1248, + "grad_norm": 0.8807411946580567, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 12480 + }, + { + "epoch": 0.12481, + "grad_norm": 0.8926462579502399, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 12481 + }, + { + "epoch": 0.12482, + "grad_norm": 1.0745427028004215, + "learning_rate": 0.003, + "loss": 4.0963, + "step": 12482 + }, + { + "epoch": 0.12483, + "grad_norm": 0.9929214487423865, + "learning_rate": 0.003, + "loss": 4.0986, + "step": 12483 + }, + { + "epoch": 0.12484, + "grad_norm": 0.9617781128077857, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 12484 + }, + { + "epoch": 0.12485, + "grad_norm": 0.8621816145149558, + "learning_rate": 0.003, + "loss": 4.0964, + "step": 12485 + }, + { + "epoch": 0.12486, + "grad_norm": 0.9037828404066399, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 12486 + }, + { + "epoch": 0.12487, + "grad_norm": 0.8679503105900107, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 12487 + }, + { + "epoch": 0.12488, + "grad_norm": 0.8598421443830003, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 12488 + }, + { + "epoch": 0.12489, + "grad_norm": 0.8650167407226419, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 12489 + }, + { + "epoch": 0.1249, + "grad_norm": 0.9485391624841892, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 12490 + }, + { + "epoch": 0.12491, + "grad_norm": 1.0914840318512178, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 12491 + }, + { + "epoch": 0.12492, + "grad_norm": 1.4196625390039204, + "learning_rate": 0.003, + "loss": 4.1057, + "step": 12492 + }, + { + "epoch": 0.12493, + "grad_norm": 0.8689361394534119, + "learning_rate": 0.003, + "loss": 4.0923, + "step": 12493 + }, + { + "epoch": 0.12494, + "grad_norm": 0.7529938697553933, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 12494 + }, + { + "epoch": 0.12495, + "grad_norm": 0.7869397679263302, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 12495 + }, + { + "epoch": 0.12496, + "grad_norm": 0.9354648250248937, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 12496 + }, + { + "epoch": 0.12497, + "grad_norm": 1.01137018920271, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 12497 + }, + { + "epoch": 0.12498, + "grad_norm": 1.051321584027189, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 12498 + }, + { + "epoch": 0.12499, + "grad_norm": 1.219799634017792, + "learning_rate": 0.003, + "loss": 4.0908, + "step": 12499 + }, + { + "epoch": 0.125, + "grad_norm": 0.9772052744068769, + "learning_rate": 0.003, + "loss": 4.1248, + "step": 12500 + }, + { + "epoch": 0.12501, + "grad_norm": 0.9023999983682449, + "learning_rate": 0.003, + "loss": 4.0945, + "step": 12501 + }, + { + "epoch": 0.12502, + "grad_norm": 0.8819061715691086, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 12502 + }, + { + "epoch": 0.12503, + "grad_norm": 0.9152110705443846, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 12503 + }, + { + "epoch": 0.12504, + "grad_norm": 0.9233537023380912, + "learning_rate": 0.003, + "loss": 4.083, + "step": 12504 + }, + { + "epoch": 0.12505, + "grad_norm": 0.9371286046139368, + "learning_rate": 0.003, + "loss": 4.0972, + "step": 12505 + }, + { + "epoch": 0.12506, + "grad_norm": 0.9842674964271714, + "learning_rate": 0.003, + "loss": 4.0856, + "step": 12506 + }, + { + "epoch": 0.12507, + "grad_norm": 0.9621168769983468, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 12507 + }, + { + "epoch": 0.12508, + "grad_norm": 0.9601103363092106, + "learning_rate": 0.003, + "loss": 4.08, + "step": 12508 + }, + { + "epoch": 0.12509, + "grad_norm": 1.013607126040525, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 12509 + }, + { + "epoch": 0.1251, + "grad_norm": 0.9894387369048696, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 12510 + }, + { + "epoch": 0.12511, + "grad_norm": 1.1777299248607307, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 12511 + }, + { + "epoch": 0.12512, + "grad_norm": 0.991445203764389, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 12512 + }, + { + "epoch": 0.12513, + "grad_norm": 0.9272080382281512, + "learning_rate": 0.003, + "loss": 4.0902, + "step": 12513 + }, + { + "epoch": 0.12514, + "grad_norm": 0.8295463935128615, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 12514 + }, + { + "epoch": 0.12515, + "grad_norm": 0.8142091850489257, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 12515 + }, + { + "epoch": 0.12516, + "grad_norm": 0.903511317581231, + "learning_rate": 0.003, + "loss": 4.0902, + "step": 12516 + }, + { + "epoch": 0.12517, + "grad_norm": 0.8060839127684248, + "learning_rate": 0.003, + "loss": 4.046, + "step": 12517 + }, + { + "epoch": 0.12518, + "grad_norm": 0.8318313264044653, + "learning_rate": 0.003, + "loss": 4.0958, + "step": 12518 + }, + { + "epoch": 0.12519, + "grad_norm": 0.8751459550107998, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 12519 + }, + { + "epoch": 0.1252, + "grad_norm": 1.0550188149120407, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 12520 + }, + { + "epoch": 0.12521, + "grad_norm": 1.1612599220876851, + "learning_rate": 0.003, + "loss": 4.0891, + "step": 12521 + }, + { + "epoch": 0.12522, + "grad_norm": 1.0472280592086465, + "learning_rate": 0.003, + "loss": 4.1198, + "step": 12522 + }, + { + "epoch": 0.12523, + "grad_norm": 1.1972820941746982, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 12523 + }, + { + "epoch": 0.12524, + "grad_norm": 0.9153940956168188, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 12524 + }, + { + "epoch": 0.12525, + "grad_norm": 0.8081300560111551, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 12525 + }, + { + "epoch": 0.12526, + "grad_norm": 0.903137227284903, + "learning_rate": 0.003, + "loss": 4.1052, + "step": 12526 + }, + { + "epoch": 0.12527, + "grad_norm": 0.9799312352354196, + "learning_rate": 0.003, + "loss": 4.1036, + "step": 12527 + }, + { + "epoch": 0.12528, + "grad_norm": 1.0055150664551715, + "learning_rate": 0.003, + "loss": 4.071, + "step": 12528 + }, + { + "epoch": 0.12529, + "grad_norm": 1.0278268288110866, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 12529 + }, + { + "epoch": 0.1253, + "grad_norm": 1.07602526626192, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 12530 + }, + { + "epoch": 0.12531, + "grad_norm": 1.1492762305204314, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 12531 + }, + { + "epoch": 0.12532, + "grad_norm": 0.9352272294009008, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 12532 + }, + { + "epoch": 0.12533, + "grad_norm": 0.8892177275978362, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 12533 + }, + { + "epoch": 0.12534, + "grad_norm": 1.0114800696680504, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 12534 + }, + { + "epoch": 0.12535, + "grad_norm": 1.0768018193623285, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 12535 + }, + { + "epoch": 0.12536, + "grad_norm": 0.9340025323695215, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 12536 + }, + { + "epoch": 0.12537, + "grad_norm": 0.8950988866336271, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 12537 + }, + { + "epoch": 0.12538, + "grad_norm": 0.9962498740816474, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 12538 + }, + { + "epoch": 0.12539, + "grad_norm": 1.2074334729709642, + "learning_rate": 0.003, + "loss": 4.113, + "step": 12539 + }, + { + "epoch": 0.1254, + "grad_norm": 0.9376791249148332, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 12540 + }, + { + "epoch": 0.12541, + "grad_norm": 0.9146861153969978, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 12541 + }, + { + "epoch": 0.12542, + "grad_norm": 0.7905394101265425, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 12542 + }, + { + "epoch": 0.12543, + "grad_norm": 0.7036995126135881, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 12543 + }, + { + "epoch": 0.12544, + "grad_norm": 0.6894012178600143, + "learning_rate": 0.003, + "loss": 4.107, + "step": 12544 + }, + { + "epoch": 0.12545, + "grad_norm": 0.7310534836339984, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 12545 + }, + { + "epoch": 0.12546, + "grad_norm": 1.0959427606998433, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 12546 + }, + { + "epoch": 0.12547, + "grad_norm": 1.3604499655224955, + "learning_rate": 0.003, + "loss": 4.1074, + "step": 12547 + }, + { + "epoch": 0.12548, + "grad_norm": 0.7142929200760227, + "learning_rate": 0.003, + "loss": 4.056, + "step": 12548 + }, + { + "epoch": 0.12549, + "grad_norm": 0.8370239050106169, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 12549 + }, + { + "epoch": 0.1255, + "grad_norm": 0.9792968007088423, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 12550 + }, + { + "epoch": 0.12551, + "grad_norm": 1.104038287909731, + "learning_rate": 0.003, + "loss": 4.1006, + "step": 12551 + }, + { + "epoch": 0.12552, + "grad_norm": 0.9109075912595407, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 12552 + }, + { + "epoch": 0.12553, + "grad_norm": 0.8076178172439208, + "learning_rate": 0.003, + "loss": 4.0908, + "step": 12553 + }, + { + "epoch": 0.12554, + "grad_norm": 0.8022634207297645, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 12554 + }, + { + "epoch": 0.12555, + "grad_norm": 0.7734066695111126, + "learning_rate": 0.003, + "loss": 4.0881, + "step": 12555 + }, + { + "epoch": 0.12556, + "grad_norm": 0.8108340027762065, + "learning_rate": 0.003, + "loss": 4.1059, + "step": 12556 + }, + { + "epoch": 0.12557, + "grad_norm": 0.8584409249242587, + "learning_rate": 0.003, + "loss": 4.0984, + "step": 12557 + }, + { + "epoch": 0.12558, + "grad_norm": 1.044999918908341, + "learning_rate": 0.003, + "loss": 4.0893, + "step": 12558 + }, + { + "epoch": 0.12559, + "grad_norm": 0.9789633376197848, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 12559 + }, + { + "epoch": 0.1256, + "grad_norm": 1.3231054069150685, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 12560 + }, + { + "epoch": 0.12561, + "grad_norm": 0.992704818498501, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 12561 + }, + { + "epoch": 0.12562, + "grad_norm": 0.9161534410318197, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 12562 + }, + { + "epoch": 0.12563, + "grad_norm": 0.991923147814018, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 12563 + }, + { + "epoch": 0.12564, + "grad_norm": 1.3963552579951515, + "learning_rate": 0.003, + "loss": 4.094, + "step": 12564 + }, + { + "epoch": 0.12565, + "grad_norm": 0.7736482160217792, + "learning_rate": 0.003, + "loss": 4.0886, + "step": 12565 + }, + { + "epoch": 0.12566, + "grad_norm": 0.6730352692878543, + "learning_rate": 0.003, + "loss": 4.088, + "step": 12566 + }, + { + "epoch": 0.12567, + "grad_norm": 0.9138881530304057, + "learning_rate": 0.003, + "loss": 4.1221, + "step": 12567 + }, + { + "epoch": 0.12568, + "grad_norm": 1.0066971053242915, + "learning_rate": 0.003, + "loss": 4.079, + "step": 12568 + }, + { + "epoch": 0.12569, + "grad_norm": 0.9073915207792339, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 12569 + }, + { + "epoch": 0.1257, + "grad_norm": 1.110940820163659, + "learning_rate": 0.003, + "loss": 4.1286, + "step": 12570 + }, + { + "epoch": 0.12571, + "grad_norm": 1.0905772922630095, + "learning_rate": 0.003, + "loss": 4.0967, + "step": 12571 + }, + { + "epoch": 0.12572, + "grad_norm": 0.9239137626440753, + "learning_rate": 0.003, + "loss": 4.0934, + "step": 12572 + }, + { + "epoch": 0.12573, + "grad_norm": 1.008223099174297, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 12573 + }, + { + "epoch": 0.12574, + "grad_norm": 1.1094893588120287, + "learning_rate": 0.003, + "loss": 4.096, + "step": 12574 + }, + { + "epoch": 0.12575, + "grad_norm": 0.81664357335465, + "learning_rate": 0.003, + "loss": 4.1163, + "step": 12575 + }, + { + "epoch": 0.12576, + "grad_norm": 0.8090241203803483, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 12576 + }, + { + "epoch": 0.12577, + "grad_norm": 0.7466922203095001, + "learning_rate": 0.003, + "loss": 4.064, + "step": 12577 + }, + { + "epoch": 0.12578, + "grad_norm": 0.8619938430766506, + "learning_rate": 0.003, + "loss": 4.076, + "step": 12578 + }, + { + "epoch": 0.12579, + "grad_norm": 0.8649739581041872, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 12579 + }, + { + "epoch": 0.1258, + "grad_norm": 0.7952715416823467, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 12580 + }, + { + "epoch": 0.12581, + "grad_norm": 0.9203260850046064, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 12581 + }, + { + "epoch": 0.12582, + "grad_norm": 1.0940378528829267, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 12582 + }, + { + "epoch": 0.12583, + "grad_norm": 1.013508026724092, + "learning_rate": 0.003, + "loss": 4.0971, + "step": 12583 + }, + { + "epoch": 0.12584, + "grad_norm": 0.9817936235910812, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 12584 + }, + { + "epoch": 0.12585, + "grad_norm": 1.1074696528899641, + "learning_rate": 0.003, + "loss": 4.109, + "step": 12585 + }, + { + "epoch": 0.12586, + "grad_norm": 1.1544542617642237, + "learning_rate": 0.003, + "loss": 4.0884, + "step": 12586 + }, + { + "epoch": 0.12587, + "grad_norm": 1.0054702814038652, + "learning_rate": 0.003, + "loss": 4.1036, + "step": 12587 + }, + { + "epoch": 0.12588, + "grad_norm": 0.8908177609969784, + "learning_rate": 0.003, + "loss": 4.066, + "step": 12588 + }, + { + "epoch": 0.12589, + "grad_norm": 0.9309703497491816, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 12589 + }, + { + "epoch": 0.1259, + "grad_norm": 0.8839426054934597, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 12590 + }, + { + "epoch": 0.12591, + "grad_norm": 0.8763529926173601, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 12591 + }, + { + "epoch": 0.12592, + "grad_norm": 0.863359311124618, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 12592 + }, + { + "epoch": 0.12593, + "grad_norm": 0.9144515411094422, + "learning_rate": 0.003, + "loss": 4.0873, + "step": 12593 + }, + { + "epoch": 0.12594, + "grad_norm": 0.841680707940564, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 12594 + }, + { + "epoch": 0.12595, + "grad_norm": 0.8061699368688643, + "learning_rate": 0.003, + "loss": 4.106, + "step": 12595 + }, + { + "epoch": 0.12596, + "grad_norm": 0.8843148213816681, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 12596 + }, + { + "epoch": 0.12597, + "grad_norm": 1.1989234844916787, + "learning_rate": 0.003, + "loss": 4.095, + "step": 12597 + }, + { + "epoch": 0.12598, + "grad_norm": 1.0967375221416913, + "learning_rate": 0.003, + "loss": 4.099, + "step": 12598 + }, + { + "epoch": 0.12599, + "grad_norm": 1.0462043180463803, + "learning_rate": 0.003, + "loss": 4.1041, + "step": 12599 + }, + { + "epoch": 0.126, + "grad_norm": 0.8935883430707421, + "learning_rate": 0.003, + "loss": 4.098, + "step": 12600 + }, + { + "epoch": 0.12601, + "grad_norm": 0.7271404222141363, + "learning_rate": 0.003, + "loss": 4.1112, + "step": 12601 + }, + { + "epoch": 0.12602, + "grad_norm": 0.7074201290060389, + "learning_rate": 0.003, + "loss": 4.0881, + "step": 12602 + }, + { + "epoch": 0.12603, + "grad_norm": 0.780807399825936, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 12603 + }, + { + "epoch": 0.12604, + "grad_norm": 0.9465997083003611, + "learning_rate": 0.003, + "loss": 4.1036, + "step": 12604 + }, + { + "epoch": 0.12605, + "grad_norm": 1.1043466399933166, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 12605 + }, + { + "epoch": 0.12606, + "grad_norm": 0.9621190173083083, + "learning_rate": 0.003, + "loss": 4.0901, + "step": 12606 + }, + { + "epoch": 0.12607, + "grad_norm": 1.0857920403184445, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 12607 + }, + { + "epoch": 0.12608, + "grad_norm": 1.1417704643449234, + "learning_rate": 0.003, + "loss": 4.072, + "step": 12608 + }, + { + "epoch": 0.12609, + "grad_norm": 0.9421103437113187, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 12609 + }, + { + "epoch": 0.1261, + "grad_norm": 0.9063025054066826, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 12610 + }, + { + "epoch": 0.12611, + "grad_norm": 0.7994198370371652, + "learning_rate": 0.003, + "loss": 4.0886, + "step": 12611 + }, + { + "epoch": 0.12612, + "grad_norm": 0.7636175243338881, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 12612 + }, + { + "epoch": 0.12613, + "grad_norm": 0.799281492649268, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 12613 + }, + { + "epoch": 0.12614, + "grad_norm": 0.8178924898260137, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 12614 + }, + { + "epoch": 0.12615, + "grad_norm": 0.9821261702866314, + "learning_rate": 0.003, + "loss": 4.0997, + "step": 12615 + }, + { + "epoch": 0.12616, + "grad_norm": 1.0407967648130723, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 12616 + }, + { + "epoch": 0.12617, + "grad_norm": 0.9433578320202893, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 12617 + }, + { + "epoch": 0.12618, + "grad_norm": 1.2938051044828642, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 12618 + }, + { + "epoch": 0.12619, + "grad_norm": 1.0592485642128842, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 12619 + }, + { + "epoch": 0.1262, + "grad_norm": 1.0122176591892478, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 12620 + }, + { + "epoch": 0.12621, + "grad_norm": 0.993994417632814, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 12621 + }, + { + "epoch": 0.12622, + "grad_norm": 1.037415957652766, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 12622 + }, + { + "epoch": 0.12623, + "grad_norm": 1.0838694110653007, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 12623 + }, + { + "epoch": 0.12624, + "grad_norm": 1.0380841838109072, + "learning_rate": 0.003, + "loss": 4.072, + "step": 12624 + }, + { + "epoch": 0.12625, + "grad_norm": 0.8908092542888468, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 12625 + }, + { + "epoch": 0.12626, + "grad_norm": 0.8194798640155146, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 12626 + }, + { + "epoch": 0.12627, + "grad_norm": 0.885492355477465, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 12627 + }, + { + "epoch": 0.12628, + "grad_norm": 0.862019047255476, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 12628 + }, + { + "epoch": 0.12629, + "grad_norm": 0.9215122599051112, + "learning_rate": 0.003, + "loss": 4.108, + "step": 12629 + }, + { + "epoch": 0.1263, + "grad_norm": 1.1265674107002008, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 12630 + }, + { + "epoch": 0.12631, + "grad_norm": 1.0485199449736506, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 12631 + }, + { + "epoch": 0.12632, + "grad_norm": 1.0894839892604955, + "learning_rate": 0.003, + "loss": 4.0931, + "step": 12632 + }, + { + "epoch": 0.12633, + "grad_norm": 1.1189766374351253, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 12633 + }, + { + "epoch": 0.12634, + "grad_norm": 0.9268084466731022, + "learning_rate": 0.003, + "loss": 4.0917, + "step": 12634 + }, + { + "epoch": 0.12635, + "grad_norm": 0.9278968417556661, + "learning_rate": 0.003, + "loss": 4.1201, + "step": 12635 + }, + { + "epoch": 0.12636, + "grad_norm": 1.3389324431542038, + "learning_rate": 0.003, + "loss": 4.0949, + "step": 12636 + }, + { + "epoch": 0.12637, + "grad_norm": 0.9294559895377964, + "learning_rate": 0.003, + "loss": 4.09, + "step": 12637 + }, + { + "epoch": 0.12638, + "grad_norm": 0.8852226422305655, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 12638 + }, + { + "epoch": 0.12639, + "grad_norm": 0.8123617472148847, + "learning_rate": 0.003, + "loss": 4.0958, + "step": 12639 + }, + { + "epoch": 0.1264, + "grad_norm": 0.8274232685598667, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 12640 + }, + { + "epoch": 0.12641, + "grad_norm": 1.0160057558439604, + "learning_rate": 0.003, + "loss": 4.0868, + "step": 12641 + }, + { + "epoch": 0.12642, + "grad_norm": 1.4294597248342236, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 12642 + }, + { + "epoch": 0.12643, + "grad_norm": 0.7172178499370168, + "learning_rate": 0.003, + "loss": 4.1013, + "step": 12643 + }, + { + "epoch": 0.12644, + "grad_norm": 0.676427999371077, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 12644 + }, + { + "epoch": 0.12645, + "grad_norm": 0.6411201053982815, + "learning_rate": 0.003, + "loss": 4.086, + "step": 12645 + }, + { + "epoch": 0.12646, + "grad_norm": 0.7283599681113574, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 12646 + }, + { + "epoch": 0.12647, + "grad_norm": 0.8768763862896161, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 12647 + }, + { + "epoch": 0.12648, + "grad_norm": 1.2749329584954063, + "learning_rate": 0.003, + "loss": 4.1172, + "step": 12648 + }, + { + "epoch": 0.12649, + "grad_norm": 0.9588635250383875, + "learning_rate": 0.003, + "loss": 4.054, + "step": 12649 + }, + { + "epoch": 0.1265, + "grad_norm": 0.8115486373152372, + "learning_rate": 0.003, + "loss": 4.1052, + "step": 12650 + }, + { + "epoch": 0.12651, + "grad_norm": 0.8247367512555528, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 12651 + }, + { + "epoch": 0.12652, + "grad_norm": 0.8618508247762805, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 12652 + }, + { + "epoch": 0.12653, + "grad_norm": 0.8859775741303744, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 12653 + }, + { + "epoch": 0.12654, + "grad_norm": 0.8943774352099432, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 12654 + }, + { + "epoch": 0.12655, + "grad_norm": 0.980223612318296, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 12655 + }, + { + "epoch": 0.12656, + "grad_norm": 1.1747202556156335, + "learning_rate": 0.003, + "loss": 4.1116, + "step": 12656 + }, + { + "epoch": 0.12657, + "grad_norm": 0.9732896792705065, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 12657 + }, + { + "epoch": 0.12658, + "grad_norm": 1.0145979944692503, + "learning_rate": 0.003, + "loss": 4.1034, + "step": 12658 + }, + { + "epoch": 0.12659, + "grad_norm": 1.0846034070522488, + "learning_rate": 0.003, + "loss": 4.053, + "step": 12659 + }, + { + "epoch": 0.1266, + "grad_norm": 1.127012054134465, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 12660 + }, + { + "epoch": 0.12661, + "grad_norm": 1.0492148040950495, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 12661 + }, + { + "epoch": 0.12662, + "grad_norm": 1.1968072144048836, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 12662 + }, + { + "epoch": 0.12663, + "grad_norm": 0.8038196979267652, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 12663 + }, + { + "epoch": 0.12664, + "grad_norm": 0.7526547203974964, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 12664 + }, + { + "epoch": 0.12665, + "grad_norm": 0.6616812099441614, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 12665 + }, + { + "epoch": 0.12666, + "grad_norm": 0.6111424790479509, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 12666 + }, + { + "epoch": 0.12667, + "grad_norm": 0.6665844263344736, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 12667 + }, + { + "epoch": 0.12668, + "grad_norm": 0.7412367724583749, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 12668 + }, + { + "epoch": 0.12669, + "grad_norm": 1.173417528820116, + "learning_rate": 0.003, + "loss": 4.1018, + "step": 12669 + }, + { + "epoch": 0.1267, + "grad_norm": 1.1755252901089872, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 12670 + }, + { + "epoch": 0.12671, + "grad_norm": 0.6691423481414988, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 12671 + }, + { + "epoch": 0.12672, + "grad_norm": 0.7731342483030905, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 12672 + }, + { + "epoch": 0.12673, + "grad_norm": 1.0755482755225818, + "learning_rate": 0.003, + "loss": 4.0893, + "step": 12673 + }, + { + "epoch": 0.12674, + "grad_norm": 1.2491602805619382, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 12674 + }, + { + "epoch": 0.12675, + "grad_norm": 0.9597672533435666, + "learning_rate": 0.003, + "loss": 4.0966, + "step": 12675 + }, + { + "epoch": 0.12676, + "grad_norm": 0.8590862792028856, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 12676 + }, + { + "epoch": 0.12677, + "grad_norm": 0.9761892578397884, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 12677 + }, + { + "epoch": 0.12678, + "grad_norm": 0.9227174449210696, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 12678 + }, + { + "epoch": 0.12679, + "grad_norm": 0.9187415631097591, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 12679 + }, + { + "epoch": 0.1268, + "grad_norm": 1.0246369627426888, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 12680 + }, + { + "epoch": 0.12681, + "grad_norm": 0.9736167198020269, + "learning_rate": 0.003, + "loss": 4.0955, + "step": 12681 + }, + { + "epoch": 0.12682, + "grad_norm": 0.9394744230967796, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 12682 + }, + { + "epoch": 0.12683, + "grad_norm": 0.8729589384396369, + "learning_rate": 0.003, + "loss": 4.0948, + "step": 12683 + }, + { + "epoch": 0.12684, + "grad_norm": 0.7774329344440306, + "learning_rate": 0.003, + "loss": 4.0958, + "step": 12684 + }, + { + "epoch": 0.12685, + "grad_norm": 0.8444647598547264, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 12685 + }, + { + "epoch": 0.12686, + "grad_norm": 0.9747599369136335, + "learning_rate": 0.003, + "loss": 4.0934, + "step": 12686 + }, + { + "epoch": 0.12687, + "grad_norm": 1.2148296561098226, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 12687 + }, + { + "epoch": 0.12688, + "grad_norm": 1.0050545067754602, + "learning_rate": 0.003, + "loss": 4.079, + "step": 12688 + }, + { + "epoch": 0.12689, + "grad_norm": 1.095960549729426, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 12689 + }, + { + "epoch": 0.1269, + "grad_norm": 1.0266522244095666, + "learning_rate": 0.003, + "loss": 4.071, + "step": 12690 + }, + { + "epoch": 0.12691, + "grad_norm": 1.0870336727058552, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 12691 + }, + { + "epoch": 0.12692, + "grad_norm": 1.2292824332108863, + "learning_rate": 0.003, + "loss": 4.1033, + "step": 12692 + }, + { + "epoch": 0.12693, + "grad_norm": 1.7400003525678518, + "learning_rate": 0.003, + "loss": 4.085, + "step": 12693 + }, + { + "epoch": 0.12694, + "grad_norm": 0.7380506838849168, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 12694 + }, + { + "epoch": 0.12695, + "grad_norm": 0.9728499264489139, + "learning_rate": 0.003, + "loss": 4.0988, + "step": 12695 + }, + { + "epoch": 0.12696, + "grad_norm": 1.1390261146005374, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 12696 + }, + { + "epoch": 0.12697, + "grad_norm": 1.0047118405929174, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 12697 + }, + { + "epoch": 0.12698, + "grad_norm": 1.160753357579373, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 12698 + }, + { + "epoch": 0.12699, + "grad_norm": 1.0980311215874454, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 12699 + }, + { + "epoch": 0.127, + "grad_norm": 0.9333469948399583, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 12700 + }, + { + "epoch": 0.12701, + "grad_norm": 1.2273734117655681, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 12701 + }, + { + "epoch": 0.12702, + "grad_norm": 0.943660424299364, + "learning_rate": 0.003, + "loss": 4.0913, + "step": 12702 + }, + { + "epoch": 0.12703, + "grad_norm": 1.1041814455713508, + "learning_rate": 0.003, + "loss": 4.0983, + "step": 12703 + }, + { + "epoch": 0.12704, + "grad_norm": 1.0338757493150266, + "learning_rate": 0.003, + "loss": 4.1063, + "step": 12704 + }, + { + "epoch": 0.12705, + "grad_norm": 0.9022360017116657, + "learning_rate": 0.003, + "loss": 4.0937, + "step": 12705 + }, + { + "epoch": 0.12706, + "grad_norm": 0.883521101535422, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 12706 + }, + { + "epoch": 0.12707, + "grad_norm": 0.8681998552528973, + "learning_rate": 0.003, + "loss": 4.0963, + "step": 12707 + }, + { + "epoch": 0.12708, + "grad_norm": 0.9428056816064104, + "learning_rate": 0.003, + "loss": 4.1078, + "step": 12708 + }, + { + "epoch": 0.12709, + "grad_norm": 1.087648806102441, + "learning_rate": 0.003, + "loss": 4.1001, + "step": 12709 + }, + { + "epoch": 0.1271, + "grad_norm": 0.9096923507613008, + "learning_rate": 0.003, + "loss": 4.0935, + "step": 12710 + }, + { + "epoch": 0.12711, + "grad_norm": 0.7974710381298415, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 12711 + }, + { + "epoch": 0.12712, + "grad_norm": 0.8936566240907319, + "learning_rate": 0.003, + "loss": 4.07, + "step": 12712 + }, + { + "epoch": 0.12713, + "grad_norm": 0.9330136754328376, + "learning_rate": 0.003, + "loss": 4.085, + "step": 12713 + }, + { + "epoch": 0.12714, + "grad_norm": 0.9510515292164203, + "learning_rate": 0.003, + "loss": 4.1058, + "step": 12714 + }, + { + "epoch": 0.12715, + "grad_norm": 1.053900359320174, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 12715 + }, + { + "epoch": 0.12716, + "grad_norm": 1.0506495651534915, + "learning_rate": 0.003, + "loss": 4.0905, + "step": 12716 + }, + { + "epoch": 0.12717, + "grad_norm": 0.9606720400266202, + "learning_rate": 0.003, + "loss": 4.1, + "step": 12717 + }, + { + "epoch": 0.12718, + "grad_norm": 1.0452039035514802, + "learning_rate": 0.003, + "loss": 4.0857, + "step": 12718 + }, + { + "epoch": 0.12719, + "grad_norm": 1.282804481030621, + "learning_rate": 0.003, + "loss": 4.1131, + "step": 12719 + }, + { + "epoch": 0.1272, + "grad_norm": 0.8678974621917694, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 12720 + }, + { + "epoch": 0.12721, + "grad_norm": 0.7483390476658757, + "learning_rate": 0.003, + "loss": 4.0954, + "step": 12721 + }, + { + "epoch": 0.12722, + "grad_norm": 0.6834628002837406, + "learning_rate": 0.003, + "loss": 4.0884, + "step": 12722 + }, + { + "epoch": 0.12723, + "grad_norm": 0.6915446962845225, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 12723 + }, + { + "epoch": 0.12724, + "grad_norm": 0.6915478039403316, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 12724 + }, + { + "epoch": 0.12725, + "grad_norm": 0.720653105262162, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 12725 + }, + { + "epoch": 0.12726, + "grad_norm": 0.8823087428054952, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 12726 + }, + { + "epoch": 0.12727, + "grad_norm": 1.13878918352502, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 12727 + }, + { + "epoch": 0.12728, + "grad_norm": 1.0623288851876986, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 12728 + }, + { + "epoch": 0.12729, + "grad_norm": 1.0342793131773047, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 12729 + }, + { + "epoch": 0.1273, + "grad_norm": 1.2518008242588525, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 12730 + }, + { + "epoch": 0.12731, + "grad_norm": 0.8867952045733, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 12731 + }, + { + "epoch": 0.12732, + "grad_norm": 0.7714444820266095, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 12732 + }, + { + "epoch": 0.12733, + "grad_norm": 0.6679650865693566, + "learning_rate": 0.003, + "loss": 4.0876, + "step": 12733 + }, + { + "epoch": 0.12734, + "grad_norm": 0.762830700091002, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 12734 + }, + { + "epoch": 0.12735, + "grad_norm": 0.7113841048189926, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 12735 + }, + { + "epoch": 0.12736, + "grad_norm": 0.5862257632552317, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 12736 + }, + { + "epoch": 0.12737, + "grad_norm": 0.6120870224768767, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 12737 + }, + { + "epoch": 0.12738, + "grad_norm": 0.6106981973476457, + "learning_rate": 0.003, + "loss": 4.079, + "step": 12738 + }, + { + "epoch": 0.12739, + "grad_norm": 0.7632193712436027, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 12739 + }, + { + "epoch": 0.1274, + "grad_norm": 0.9875735282030758, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 12740 + }, + { + "epoch": 0.12741, + "grad_norm": 1.4834980634291643, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 12741 + }, + { + "epoch": 0.12742, + "grad_norm": 0.7679248889135363, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 12742 + }, + { + "epoch": 0.12743, + "grad_norm": 0.8062478105687914, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 12743 + }, + { + "epoch": 0.12744, + "grad_norm": 0.80926254377738, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 12744 + }, + { + "epoch": 0.12745, + "grad_norm": 0.86631563610372, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 12745 + }, + { + "epoch": 0.12746, + "grad_norm": 1.139058874131467, + "learning_rate": 0.003, + "loss": 4.1043, + "step": 12746 + }, + { + "epoch": 0.12747, + "grad_norm": 1.0225688398798827, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 12747 + }, + { + "epoch": 0.12748, + "grad_norm": 1.241245519984615, + "learning_rate": 0.003, + "loss": 4.1035, + "step": 12748 + }, + { + "epoch": 0.12749, + "grad_norm": 0.9236187756461689, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 12749 + }, + { + "epoch": 0.1275, + "grad_norm": 0.9329801041725276, + "learning_rate": 0.003, + "loss": 4.089, + "step": 12750 + }, + { + "epoch": 0.12751, + "grad_norm": 1.0591896340489753, + "learning_rate": 0.003, + "loss": 4.1032, + "step": 12751 + }, + { + "epoch": 0.12752, + "grad_norm": 1.2259293843197325, + "learning_rate": 0.003, + "loss": 4.0979, + "step": 12752 + }, + { + "epoch": 0.12753, + "grad_norm": 0.7733677218624588, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 12753 + }, + { + "epoch": 0.12754, + "grad_norm": 0.925000752304408, + "learning_rate": 0.003, + "loss": 4.1099, + "step": 12754 + }, + { + "epoch": 0.12755, + "grad_norm": 1.0096593748241867, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 12755 + }, + { + "epoch": 0.12756, + "grad_norm": 1.1261872941544642, + "learning_rate": 0.003, + "loss": 4.1032, + "step": 12756 + }, + { + "epoch": 0.12757, + "grad_norm": 0.938716704223264, + "learning_rate": 0.003, + "loss": 4.0888, + "step": 12757 + }, + { + "epoch": 0.12758, + "grad_norm": 0.977359487405289, + "learning_rate": 0.003, + "loss": 4.082, + "step": 12758 + }, + { + "epoch": 0.12759, + "grad_norm": 1.3547393839590267, + "learning_rate": 0.003, + "loss": 4.0922, + "step": 12759 + }, + { + "epoch": 0.1276, + "grad_norm": 1.135480268778075, + "learning_rate": 0.003, + "loss": 4.1012, + "step": 12760 + }, + { + "epoch": 0.12761, + "grad_norm": 1.0772758740560622, + "learning_rate": 0.003, + "loss": 4.0874, + "step": 12761 + }, + { + "epoch": 0.12762, + "grad_norm": 1.1061662908879843, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 12762 + }, + { + "epoch": 0.12763, + "grad_norm": 1.0201383129633217, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 12763 + }, + { + "epoch": 0.12764, + "grad_norm": 1.0407465695626197, + "learning_rate": 0.003, + "loss": 4.1088, + "step": 12764 + }, + { + "epoch": 0.12765, + "grad_norm": 0.8481863619593831, + "learning_rate": 0.003, + "loss": 4.0872, + "step": 12765 + }, + { + "epoch": 0.12766, + "grad_norm": 0.8716884709382724, + "learning_rate": 0.003, + "loss": 4.091, + "step": 12766 + }, + { + "epoch": 0.12767, + "grad_norm": 0.9549564226121605, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 12767 + }, + { + "epoch": 0.12768, + "grad_norm": 0.9715026484826433, + "learning_rate": 0.003, + "loss": 4.1037, + "step": 12768 + }, + { + "epoch": 0.12769, + "grad_norm": 1.0191539424859173, + "learning_rate": 0.003, + "loss": 4.085, + "step": 12769 + }, + { + "epoch": 0.1277, + "grad_norm": 1.1311761646480707, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 12770 + }, + { + "epoch": 0.12771, + "grad_norm": 1.062936765081298, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 12771 + }, + { + "epoch": 0.12772, + "grad_norm": 0.8916600272547799, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 12772 + }, + { + "epoch": 0.12773, + "grad_norm": 0.8207313213161954, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 12773 + }, + { + "epoch": 0.12774, + "grad_norm": 0.994170631073144, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 12774 + }, + { + "epoch": 0.12775, + "grad_norm": 0.9687880005021929, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 12775 + }, + { + "epoch": 0.12776, + "grad_norm": 1.019870464356583, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 12776 + }, + { + "epoch": 0.12777, + "grad_norm": 1.2305653049479717, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 12777 + }, + { + "epoch": 0.12778, + "grad_norm": 0.9077127695730837, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 12778 + }, + { + "epoch": 0.12779, + "grad_norm": 0.814451652830171, + "learning_rate": 0.003, + "loss": 4.0973, + "step": 12779 + }, + { + "epoch": 0.1278, + "grad_norm": 0.7972746426376943, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 12780 + }, + { + "epoch": 0.12781, + "grad_norm": 0.8831065216030554, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 12781 + }, + { + "epoch": 0.12782, + "grad_norm": 1.0171369662215797, + "learning_rate": 0.003, + "loss": 4.0955, + "step": 12782 + }, + { + "epoch": 0.12783, + "grad_norm": 1.0834304864407187, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 12783 + }, + { + "epoch": 0.12784, + "grad_norm": 0.9681797502111038, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 12784 + }, + { + "epoch": 0.12785, + "grad_norm": 1.1790776843414201, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 12785 + }, + { + "epoch": 0.12786, + "grad_norm": 1.0009356616251308, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 12786 + }, + { + "epoch": 0.12787, + "grad_norm": 1.240243895126111, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 12787 + }, + { + "epoch": 0.12788, + "grad_norm": 1.0451278862708107, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 12788 + }, + { + "epoch": 0.12789, + "grad_norm": 1.0137045951545032, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 12789 + }, + { + "epoch": 0.1279, + "grad_norm": 0.923587376343357, + "learning_rate": 0.003, + "loss": 4.0994, + "step": 12790 + }, + { + "epoch": 0.12791, + "grad_norm": 0.9861616504392239, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 12791 + }, + { + "epoch": 0.12792, + "grad_norm": 1.1472679227632907, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 12792 + }, + { + "epoch": 0.12793, + "grad_norm": 0.9546721671432666, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 12793 + }, + { + "epoch": 0.12794, + "grad_norm": 0.9083022364420442, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 12794 + }, + { + "epoch": 0.12795, + "grad_norm": 0.932131523244964, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 12795 + }, + { + "epoch": 0.12796, + "grad_norm": 0.8771845328073569, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 12796 + }, + { + "epoch": 0.12797, + "grad_norm": 0.8763831858336958, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 12797 + }, + { + "epoch": 0.12798, + "grad_norm": 0.9607522382537547, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 12798 + }, + { + "epoch": 0.12799, + "grad_norm": 1.0533558611162774, + "learning_rate": 0.003, + "loss": 4.0939, + "step": 12799 + }, + { + "epoch": 0.128, + "grad_norm": 0.9890908125921698, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 12800 + }, + { + "epoch": 0.12801, + "grad_norm": 1.0126880321566158, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 12801 + }, + { + "epoch": 0.12802, + "grad_norm": 1.1248360295545319, + "learning_rate": 0.003, + "loss": 4.1217, + "step": 12802 + }, + { + "epoch": 0.12803, + "grad_norm": 0.9680941416305064, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 12803 + }, + { + "epoch": 0.12804, + "grad_norm": 0.9982786180253801, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 12804 + }, + { + "epoch": 0.12805, + "grad_norm": 1.033314060460122, + "learning_rate": 0.003, + "loss": 4.0886, + "step": 12805 + }, + { + "epoch": 0.12806, + "grad_norm": 1.0519364554659985, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 12806 + }, + { + "epoch": 0.12807, + "grad_norm": 0.8216285116811861, + "learning_rate": 0.003, + "loss": 4.0929, + "step": 12807 + }, + { + "epoch": 0.12808, + "grad_norm": 0.8081310454060191, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 12808 + }, + { + "epoch": 0.12809, + "grad_norm": 0.7847526329976181, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 12809 + }, + { + "epoch": 0.1281, + "grad_norm": 0.8910310078511067, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 12810 + }, + { + "epoch": 0.12811, + "grad_norm": 0.9282318909369796, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 12811 + }, + { + "epoch": 0.12812, + "grad_norm": 1.0820055514547602, + "learning_rate": 0.003, + "loss": 4.0902, + "step": 12812 + }, + { + "epoch": 0.12813, + "grad_norm": 1.109318470155335, + "learning_rate": 0.003, + "loss": 4.0953, + "step": 12813 + }, + { + "epoch": 0.12814, + "grad_norm": 1.055508976607161, + "learning_rate": 0.003, + "loss": 4.1023, + "step": 12814 + }, + { + "epoch": 0.12815, + "grad_norm": 0.9384291038090219, + "learning_rate": 0.003, + "loss": 4.052, + "step": 12815 + }, + { + "epoch": 0.12816, + "grad_norm": 1.1073656003870405, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 12816 + }, + { + "epoch": 0.12817, + "grad_norm": 1.0255052315012416, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 12817 + }, + { + "epoch": 0.12818, + "grad_norm": 0.9779910027019095, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 12818 + }, + { + "epoch": 0.12819, + "grad_norm": 1.085698400085855, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 12819 + }, + { + "epoch": 0.1282, + "grad_norm": 1.0016327779460978, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 12820 + }, + { + "epoch": 0.12821, + "grad_norm": 0.9443105739350274, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 12821 + }, + { + "epoch": 0.12822, + "grad_norm": 0.9007356598193502, + "learning_rate": 0.003, + "loss": 4.1065, + "step": 12822 + }, + { + "epoch": 0.12823, + "grad_norm": 0.8439232090271862, + "learning_rate": 0.003, + "loss": 4.0937, + "step": 12823 + }, + { + "epoch": 0.12824, + "grad_norm": 0.932670066004475, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 12824 + }, + { + "epoch": 0.12825, + "grad_norm": 1.3005527641938726, + "learning_rate": 0.003, + "loss": 4.0916, + "step": 12825 + }, + { + "epoch": 0.12826, + "grad_norm": 0.8320530963547931, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 12826 + }, + { + "epoch": 0.12827, + "grad_norm": 0.8486681954951634, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 12827 + }, + { + "epoch": 0.12828, + "grad_norm": 0.8162761088363923, + "learning_rate": 0.003, + "loss": 4.1078, + "step": 12828 + }, + { + "epoch": 0.12829, + "grad_norm": 0.8555308082679565, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 12829 + }, + { + "epoch": 0.1283, + "grad_norm": 0.8002340794601452, + "learning_rate": 0.003, + "loss": 4.0855, + "step": 12830 + }, + { + "epoch": 0.12831, + "grad_norm": 0.8237691300557687, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 12831 + }, + { + "epoch": 0.12832, + "grad_norm": 1.0012675840674676, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 12832 + }, + { + "epoch": 0.12833, + "grad_norm": 1.0527445885962299, + "learning_rate": 0.003, + "loss": 4.1223, + "step": 12833 + }, + { + "epoch": 0.12834, + "grad_norm": 1.1024562988891253, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 12834 + }, + { + "epoch": 0.12835, + "grad_norm": 0.8237766576558605, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 12835 + }, + { + "epoch": 0.12836, + "grad_norm": 0.80386797496053, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 12836 + }, + { + "epoch": 0.12837, + "grad_norm": 0.8897326341181536, + "learning_rate": 0.003, + "loss": 4.1029, + "step": 12837 + }, + { + "epoch": 0.12838, + "grad_norm": 1.051147363324553, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 12838 + }, + { + "epoch": 0.12839, + "grad_norm": 1.2235225329956996, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 12839 + }, + { + "epoch": 0.1284, + "grad_norm": 0.9178983854172994, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 12840 + }, + { + "epoch": 0.12841, + "grad_norm": 0.9343039058451948, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 12841 + }, + { + "epoch": 0.12842, + "grad_norm": 1.0603885733009581, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 12842 + }, + { + "epoch": 0.12843, + "grad_norm": 1.0584108689555154, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 12843 + }, + { + "epoch": 0.12844, + "grad_norm": 0.7788102997291951, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 12844 + }, + { + "epoch": 0.12845, + "grad_norm": 0.7455172823210588, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 12845 + }, + { + "epoch": 0.12846, + "grad_norm": 0.7384957108754986, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 12846 + }, + { + "epoch": 0.12847, + "grad_norm": 0.7082468603892804, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 12847 + }, + { + "epoch": 0.12848, + "grad_norm": 0.7861042694189743, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 12848 + }, + { + "epoch": 0.12849, + "grad_norm": 0.7312737491653428, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 12849 + }, + { + "epoch": 0.1285, + "grad_norm": 0.7141568279103787, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 12850 + }, + { + "epoch": 0.12851, + "grad_norm": 0.8971592203284804, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 12851 + }, + { + "epoch": 0.12852, + "grad_norm": 1.2350715770720802, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 12852 + }, + { + "epoch": 0.12853, + "grad_norm": 0.8284268967366795, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 12853 + }, + { + "epoch": 0.12854, + "grad_norm": 0.6598764112170777, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 12854 + }, + { + "epoch": 0.12855, + "grad_norm": 0.6666902491456704, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 12855 + }, + { + "epoch": 0.12856, + "grad_norm": 0.7228659205740492, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 12856 + }, + { + "epoch": 0.12857, + "grad_norm": 0.8815194357293987, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 12857 + }, + { + "epoch": 0.12858, + "grad_norm": 1.0935672156381766, + "learning_rate": 0.003, + "loss": 4.0881, + "step": 12858 + }, + { + "epoch": 0.12859, + "grad_norm": 1.1510124333767726, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 12859 + }, + { + "epoch": 0.1286, + "grad_norm": 0.9949807869035089, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 12860 + }, + { + "epoch": 0.12861, + "grad_norm": 0.9334997122840311, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 12861 + }, + { + "epoch": 0.12862, + "grad_norm": 0.9525245918213804, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 12862 + }, + { + "epoch": 0.12863, + "grad_norm": 1.0915300333537923, + "learning_rate": 0.003, + "loss": 4.091, + "step": 12863 + }, + { + "epoch": 0.12864, + "grad_norm": 0.9074920970171881, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 12864 + }, + { + "epoch": 0.12865, + "grad_norm": 1.0067178122193556, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 12865 + }, + { + "epoch": 0.12866, + "grad_norm": 1.1146638664811221, + "learning_rate": 0.003, + "loss": 4.0938, + "step": 12866 + }, + { + "epoch": 0.12867, + "grad_norm": 1.0583556791484592, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 12867 + }, + { + "epoch": 0.12868, + "grad_norm": 1.1447139655853786, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 12868 + }, + { + "epoch": 0.12869, + "grad_norm": 1.1074499566576235, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 12869 + }, + { + "epoch": 0.1287, + "grad_norm": 1.058741934577846, + "learning_rate": 0.003, + "loss": 4.086, + "step": 12870 + }, + { + "epoch": 0.12871, + "grad_norm": 0.8740978740281415, + "learning_rate": 0.003, + "loss": 4.114, + "step": 12871 + }, + { + "epoch": 0.12872, + "grad_norm": 0.7768189437780134, + "learning_rate": 0.003, + "loss": 4.063, + "step": 12872 + }, + { + "epoch": 0.12873, + "grad_norm": 0.8789353883793785, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 12873 + }, + { + "epoch": 0.12874, + "grad_norm": 1.0639212704419332, + "learning_rate": 0.003, + "loss": 4.0941, + "step": 12874 + }, + { + "epoch": 0.12875, + "grad_norm": 0.8727750894973382, + "learning_rate": 0.003, + "loss": 4.0979, + "step": 12875 + }, + { + "epoch": 0.12876, + "grad_norm": 0.8727781513925823, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 12876 + }, + { + "epoch": 0.12877, + "grad_norm": 0.924561186434694, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 12877 + }, + { + "epoch": 0.12878, + "grad_norm": 1.1523203250836032, + "learning_rate": 0.003, + "loss": 4.0955, + "step": 12878 + }, + { + "epoch": 0.12879, + "grad_norm": 1.048705752039162, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 12879 + }, + { + "epoch": 0.1288, + "grad_norm": 1.1179956809885336, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 12880 + }, + { + "epoch": 0.12881, + "grad_norm": 1.0864768833213954, + "learning_rate": 0.003, + "loss": 4.0868, + "step": 12881 + }, + { + "epoch": 0.12882, + "grad_norm": 0.9394253701043804, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 12882 + }, + { + "epoch": 0.12883, + "grad_norm": 0.8804642321496315, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 12883 + }, + { + "epoch": 0.12884, + "grad_norm": 1.0724216407496103, + "learning_rate": 0.003, + "loss": 4.0884, + "step": 12884 + }, + { + "epoch": 0.12885, + "grad_norm": 1.062649754413077, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 12885 + }, + { + "epoch": 0.12886, + "grad_norm": 1.0645639222070433, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 12886 + }, + { + "epoch": 0.12887, + "grad_norm": 0.9375413595216402, + "learning_rate": 0.003, + "loss": 4.0945, + "step": 12887 + }, + { + "epoch": 0.12888, + "grad_norm": 1.0327048603814917, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 12888 + }, + { + "epoch": 0.12889, + "grad_norm": 1.094887237487474, + "learning_rate": 0.003, + "loss": 4.1049, + "step": 12889 + }, + { + "epoch": 0.1289, + "grad_norm": 0.8367449649907898, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 12890 + }, + { + "epoch": 0.12891, + "grad_norm": 0.7695447272172267, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 12891 + }, + { + "epoch": 0.12892, + "grad_norm": 0.8554920070434391, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 12892 + }, + { + "epoch": 0.12893, + "grad_norm": 1.0531570968116326, + "learning_rate": 0.003, + "loss": 4.1252, + "step": 12893 + }, + { + "epoch": 0.12894, + "grad_norm": 1.2202431842098136, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 12894 + }, + { + "epoch": 0.12895, + "grad_norm": 0.9694818722741242, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 12895 + }, + { + "epoch": 0.12896, + "grad_norm": 0.9749350963537002, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 12896 + }, + { + "epoch": 0.12897, + "grad_norm": 0.9169916760518184, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 12897 + }, + { + "epoch": 0.12898, + "grad_norm": 0.8953679985978902, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 12898 + }, + { + "epoch": 0.12899, + "grad_norm": 0.9408184091399411, + "learning_rate": 0.003, + "loss": 4.1023, + "step": 12899 + }, + { + "epoch": 0.129, + "grad_norm": 1.2043018074360279, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 12900 + }, + { + "epoch": 0.12901, + "grad_norm": 1.0856163351565085, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 12901 + }, + { + "epoch": 0.12902, + "grad_norm": 0.8473805761633323, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 12902 + }, + { + "epoch": 0.12903, + "grad_norm": 0.8241270609641437, + "learning_rate": 0.003, + "loss": 4.0982, + "step": 12903 + }, + { + "epoch": 0.12904, + "grad_norm": 0.8960165641934851, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 12904 + }, + { + "epoch": 0.12905, + "grad_norm": 0.8962753445080539, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 12905 + }, + { + "epoch": 0.12906, + "grad_norm": 0.9552856711547891, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 12906 + }, + { + "epoch": 0.12907, + "grad_norm": 1.1452287746883705, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 12907 + }, + { + "epoch": 0.12908, + "grad_norm": 1.0604234569352189, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 12908 + }, + { + "epoch": 0.12909, + "grad_norm": 1.011753863819008, + "learning_rate": 0.003, + "loss": 4.086, + "step": 12909 + }, + { + "epoch": 0.1291, + "grad_norm": 1.0445008003516258, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 12910 + }, + { + "epoch": 0.12911, + "grad_norm": 0.9940525795456636, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 12911 + }, + { + "epoch": 0.12912, + "grad_norm": 0.9599273814899095, + "learning_rate": 0.003, + "loss": 4.052, + "step": 12912 + }, + { + "epoch": 0.12913, + "grad_norm": 0.9625679687031744, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 12913 + }, + { + "epoch": 0.12914, + "grad_norm": 1.0300420701255906, + "learning_rate": 0.003, + "loss": 4.0945, + "step": 12914 + }, + { + "epoch": 0.12915, + "grad_norm": 1.0543017414193143, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 12915 + }, + { + "epoch": 0.12916, + "grad_norm": 0.9345049795079063, + "learning_rate": 0.003, + "loss": 4.0983, + "step": 12916 + }, + { + "epoch": 0.12917, + "grad_norm": 1.0291412147193604, + "learning_rate": 0.003, + "loss": 4.0925, + "step": 12917 + }, + { + "epoch": 0.12918, + "grad_norm": 0.8627476523284539, + "learning_rate": 0.003, + "loss": 4.1057, + "step": 12918 + }, + { + "epoch": 0.12919, + "grad_norm": 0.8752566608605731, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 12919 + }, + { + "epoch": 0.1292, + "grad_norm": 1.0145844859352484, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 12920 + }, + { + "epoch": 0.12921, + "grad_norm": 1.256524453344014, + "learning_rate": 0.003, + "loss": 4.0934, + "step": 12921 + }, + { + "epoch": 0.12922, + "grad_norm": 1.13769702773458, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 12922 + }, + { + "epoch": 0.12923, + "grad_norm": 1.0916995509382057, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 12923 + }, + { + "epoch": 0.12924, + "grad_norm": 0.9618122302946851, + "learning_rate": 0.003, + "loss": 4.0975, + "step": 12924 + }, + { + "epoch": 0.12925, + "grad_norm": 0.9330659357294928, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 12925 + }, + { + "epoch": 0.12926, + "grad_norm": 0.9411280689811344, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 12926 + }, + { + "epoch": 0.12927, + "grad_norm": 0.9117135005368102, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 12927 + }, + { + "epoch": 0.12928, + "grad_norm": 0.8013434462061317, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 12928 + }, + { + "epoch": 0.12929, + "grad_norm": 0.7864320913448217, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 12929 + }, + { + "epoch": 0.1293, + "grad_norm": 0.7080516925603385, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 12930 + }, + { + "epoch": 0.12931, + "grad_norm": 0.6945638875189782, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 12931 + }, + { + "epoch": 0.12932, + "grad_norm": 0.7840440411305253, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 12932 + }, + { + "epoch": 0.12933, + "grad_norm": 0.9997461888422159, + "learning_rate": 0.003, + "loss": 4.0907, + "step": 12933 + }, + { + "epoch": 0.12934, + "grad_norm": 1.2041333992689591, + "learning_rate": 0.003, + "loss": 4.0976, + "step": 12934 + }, + { + "epoch": 0.12935, + "grad_norm": 0.8190784007675626, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 12935 + }, + { + "epoch": 0.12936, + "grad_norm": 1.0008505009060236, + "learning_rate": 0.003, + "loss": 4.0948, + "step": 12936 + }, + { + "epoch": 0.12937, + "grad_norm": 1.4691550606109318, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 12937 + }, + { + "epoch": 0.12938, + "grad_norm": 0.9458879164424308, + "learning_rate": 0.003, + "loss": 4.1036, + "step": 12938 + }, + { + "epoch": 0.12939, + "grad_norm": 0.8644702954005661, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 12939 + }, + { + "epoch": 0.1294, + "grad_norm": 1.048984355848527, + "learning_rate": 0.003, + "loss": 4.084, + "step": 12940 + }, + { + "epoch": 0.12941, + "grad_norm": 1.4713281666360352, + "learning_rate": 0.003, + "loss": 4.087, + "step": 12941 + }, + { + "epoch": 0.12942, + "grad_norm": 0.7121087815183335, + "learning_rate": 0.003, + "loss": 4.069, + "step": 12942 + }, + { + "epoch": 0.12943, + "grad_norm": 0.7993204893156545, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 12943 + }, + { + "epoch": 0.12944, + "grad_norm": 0.7719930464481353, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 12944 + }, + { + "epoch": 0.12945, + "grad_norm": 0.8460365770501128, + "learning_rate": 0.003, + "loss": 4.0905, + "step": 12945 + }, + { + "epoch": 0.12946, + "grad_norm": 1.0008603080334533, + "learning_rate": 0.003, + "loss": 4.1062, + "step": 12946 + }, + { + "epoch": 0.12947, + "grad_norm": 1.0694993370533366, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 12947 + }, + { + "epoch": 0.12948, + "grad_norm": 1.0582559072839413, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 12948 + }, + { + "epoch": 0.12949, + "grad_norm": 0.9923637186992238, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 12949 + }, + { + "epoch": 0.1295, + "grad_norm": 0.8925614996758512, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 12950 + }, + { + "epoch": 0.12951, + "grad_norm": 0.9362584457030526, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 12951 + }, + { + "epoch": 0.12952, + "grad_norm": 1.017716123977998, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 12952 + }, + { + "epoch": 0.12953, + "grad_norm": 1.0748519910017815, + "learning_rate": 0.003, + "loss": 4.1011, + "step": 12953 + }, + { + "epoch": 0.12954, + "grad_norm": 0.9739139897629142, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 12954 + }, + { + "epoch": 0.12955, + "grad_norm": 1.089355342593469, + "learning_rate": 0.003, + "loss": 4.1043, + "step": 12955 + }, + { + "epoch": 0.12956, + "grad_norm": 1.0092956584324633, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 12956 + }, + { + "epoch": 0.12957, + "grad_norm": 1.1179031890479054, + "learning_rate": 0.003, + "loss": 4.0819, + "step": 12957 + }, + { + "epoch": 0.12958, + "grad_norm": 0.801443252496362, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 12958 + }, + { + "epoch": 0.12959, + "grad_norm": 0.9023149698160762, + "learning_rate": 0.003, + "loss": 4.1051, + "step": 12959 + }, + { + "epoch": 0.1296, + "grad_norm": 1.2225521145107703, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 12960 + }, + { + "epoch": 0.12961, + "grad_norm": 1.2465067821107625, + "learning_rate": 0.003, + "loss": 4.1177, + "step": 12961 + }, + { + "epoch": 0.12962, + "grad_norm": 0.8700461344554798, + "learning_rate": 0.003, + "loss": 4.0928, + "step": 12962 + }, + { + "epoch": 0.12963, + "grad_norm": 0.7722305778048842, + "learning_rate": 0.003, + "loss": 4.0854, + "step": 12963 + }, + { + "epoch": 0.12964, + "grad_norm": 0.7417488843203212, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 12964 + }, + { + "epoch": 0.12965, + "grad_norm": 0.8444082555666117, + "learning_rate": 0.003, + "loss": 4.058, + "step": 12965 + }, + { + "epoch": 0.12966, + "grad_norm": 1.0360686043902652, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 12966 + }, + { + "epoch": 0.12967, + "grad_norm": 1.2462733805614623, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 12967 + }, + { + "epoch": 0.12968, + "grad_norm": 0.760456939213373, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 12968 + }, + { + "epoch": 0.12969, + "grad_norm": 0.7701545122660179, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 12969 + }, + { + "epoch": 0.1297, + "grad_norm": 0.8041935952379368, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 12970 + }, + { + "epoch": 0.12971, + "grad_norm": 0.8569313820716046, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 12971 + }, + { + "epoch": 0.12972, + "grad_norm": 0.9526575824668777, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 12972 + }, + { + "epoch": 0.12973, + "grad_norm": 0.9852953723035094, + "learning_rate": 0.003, + "loss": 4.1118, + "step": 12973 + }, + { + "epoch": 0.12974, + "grad_norm": 0.885235021917454, + "learning_rate": 0.003, + "loss": 4.076, + "step": 12974 + }, + { + "epoch": 0.12975, + "grad_norm": 0.8416699508068156, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 12975 + }, + { + "epoch": 0.12976, + "grad_norm": 0.7622107558572573, + "learning_rate": 0.003, + "loss": 4.0887, + "step": 12976 + }, + { + "epoch": 0.12977, + "grad_norm": 0.75563744918989, + "learning_rate": 0.003, + "loss": 4.0931, + "step": 12977 + }, + { + "epoch": 0.12978, + "grad_norm": 0.9715555467394393, + "learning_rate": 0.003, + "loss": 4.06, + "step": 12978 + }, + { + "epoch": 0.12979, + "grad_norm": 1.3945054338239402, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 12979 + }, + { + "epoch": 0.1298, + "grad_norm": 0.9587856310357558, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 12980 + }, + { + "epoch": 0.12981, + "grad_norm": 0.9767100108670385, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 12981 + }, + { + "epoch": 0.12982, + "grad_norm": 0.8947214577566055, + "learning_rate": 0.003, + "loss": 4.1081, + "step": 12982 + }, + { + "epoch": 0.12983, + "grad_norm": 1.0500377401798038, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 12983 + }, + { + "epoch": 0.12984, + "grad_norm": 1.1771539622530411, + "learning_rate": 0.003, + "loss": 4.0878, + "step": 12984 + }, + { + "epoch": 0.12985, + "grad_norm": 1.0177421055570401, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 12985 + }, + { + "epoch": 0.12986, + "grad_norm": 0.9843267956713773, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 12986 + }, + { + "epoch": 0.12987, + "grad_norm": 0.9917900865130522, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 12987 + }, + { + "epoch": 0.12988, + "grad_norm": 1.0186220305166747, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 12988 + }, + { + "epoch": 0.12989, + "grad_norm": 0.9464795416743305, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 12989 + }, + { + "epoch": 0.1299, + "grad_norm": 1.1585509112262458, + "learning_rate": 0.003, + "loss": 4.1055, + "step": 12990 + }, + { + "epoch": 0.12991, + "grad_norm": 1.1443633196221934, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 12991 + }, + { + "epoch": 0.12992, + "grad_norm": 0.9902552616103755, + "learning_rate": 0.003, + "loss": 4.0992, + "step": 12992 + }, + { + "epoch": 0.12993, + "grad_norm": 0.9440526500500426, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 12993 + }, + { + "epoch": 0.12994, + "grad_norm": 0.8392942011118482, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 12994 + }, + { + "epoch": 0.12995, + "grad_norm": 0.911235037133273, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 12995 + }, + { + "epoch": 0.12996, + "grad_norm": 1.152276126164641, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 12996 + }, + { + "epoch": 0.12997, + "grad_norm": 0.8990109906292347, + "learning_rate": 0.003, + "loss": 4.0907, + "step": 12997 + }, + { + "epoch": 0.12998, + "grad_norm": 0.9507106041081811, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 12998 + }, + { + "epoch": 0.12999, + "grad_norm": 1.0575585067020632, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 12999 + }, + { + "epoch": 0.13, + "grad_norm": 0.9872077733573356, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 13000 + }, + { + "epoch": 0.13001, + "grad_norm": 0.9785487441033178, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 13001 + }, + { + "epoch": 0.13002, + "grad_norm": 1.1010172767966278, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 13002 + }, + { + "epoch": 0.13003, + "grad_norm": 1.0698609660548601, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 13003 + }, + { + "epoch": 0.13004, + "grad_norm": 0.9425442173619252, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 13004 + }, + { + "epoch": 0.13005, + "grad_norm": 0.8495334540654886, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 13005 + }, + { + "epoch": 0.13006, + "grad_norm": 0.7817570315106067, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 13006 + }, + { + "epoch": 0.13007, + "grad_norm": 0.9132794553476841, + "learning_rate": 0.003, + "loss": 4.095, + "step": 13007 + }, + { + "epoch": 0.13008, + "grad_norm": 1.0254484820912007, + "learning_rate": 0.003, + "loss": 4.0984, + "step": 13008 + }, + { + "epoch": 0.13009, + "grad_norm": 0.9463460312730713, + "learning_rate": 0.003, + "loss": 4.083, + "step": 13009 + }, + { + "epoch": 0.1301, + "grad_norm": 0.9511224906192721, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 13010 + }, + { + "epoch": 0.13011, + "grad_norm": 0.9040682198534105, + "learning_rate": 0.003, + "loss": 4.086, + "step": 13011 + }, + { + "epoch": 0.13012, + "grad_norm": 0.929617295280915, + "learning_rate": 0.003, + "loss": 4.0986, + "step": 13012 + }, + { + "epoch": 0.13013, + "grad_norm": 1.0928902780182903, + "learning_rate": 0.003, + "loss": 4.1062, + "step": 13013 + }, + { + "epoch": 0.13014, + "grad_norm": 1.0232193840964559, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 13014 + }, + { + "epoch": 0.13015, + "grad_norm": 0.8734915948389282, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 13015 + }, + { + "epoch": 0.13016, + "grad_norm": 0.6388259908205345, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 13016 + }, + { + "epoch": 0.13017, + "grad_norm": 0.652337639069181, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 13017 + }, + { + "epoch": 0.13018, + "grad_norm": 0.8827256107869914, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 13018 + }, + { + "epoch": 0.13019, + "grad_norm": 1.3371709299442436, + "learning_rate": 0.003, + "loss": 4.0987, + "step": 13019 + }, + { + "epoch": 0.1302, + "grad_norm": 0.8164348904344245, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 13020 + }, + { + "epoch": 0.13021, + "grad_norm": 0.8407396391666754, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 13021 + }, + { + "epoch": 0.13022, + "grad_norm": 0.9886831721517524, + "learning_rate": 0.003, + "loss": 4.0928, + "step": 13022 + }, + { + "epoch": 0.13023, + "grad_norm": 1.1737798050736679, + "learning_rate": 0.003, + "loss": 4.0929, + "step": 13023 + }, + { + "epoch": 0.13024, + "grad_norm": 0.7789314634527666, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 13024 + }, + { + "epoch": 0.13025, + "grad_norm": 0.9520075407496126, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 13025 + }, + { + "epoch": 0.13026, + "grad_norm": 1.0458570650535366, + "learning_rate": 0.003, + "loss": 4.0989, + "step": 13026 + }, + { + "epoch": 0.13027, + "grad_norm": 0.7605890131429558, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 13027 + }, + { + "epoch": 0.13028, + "grad_norm": 0.9208367956504226, + "learning_rate": 0.003, + "loss": 4.105, + "step": 13028 + }, + { + "epoch": 0.13029, + "grad_norm": 1.0286038755878266, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 13029 + }, + { + "epoch": 0.1303, + "grad_norm": 1.0846563030625815, + "learning_rate": 0.003, + "loss": 4.1069, + "step": 13030 + }, + { + "epoch": 0.13031, + "grad_norm": 1.1581095273087845, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 13031 + }, + { + "epoch": 0.13032, + "grad_norm": 0.9425126396088283, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 13032 + }, + { + "epoch": 0.13033, + "grad_norm": 1.0217411686065399, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 13033 + }, + { + "epoch": 0.13034, + "grad_norm": 1.1551093953325848, + "learning_rate": 0.003, + "loss": 4.1101, + "step": 13034 + }, + { + "epoch": 0.13035, + "grad_norm": 0.8313713956070581, + "learning_rate": 0.003, + "loss": 4.0952, + "step": 13035 + }, + { + "epoch": 0.13036, + "grad_norm": 0.9236200605193985, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 13036 + }, + { + "epoch": 0.13037, + "grad_norm": 1.0555437553598797, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 13037 + }, + { + "epoch": 0.13038, + "grad_norm": 0.9663628420492938, + "learning_rate": 0.003, + "loss": 4.1018, + "step": 13038 + }, + { + "epoch": 0.13039, + "grad_norm": 0.968892804210825, + "learning_rate": 0.003, + "loss": 4.1051, + "step": 13039 + }, + { + "epoch": 0.1304, + "grad_norm": 1.041473024829074, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 13040 + }, + { + "epoch": 0.13041, + "grad_norm": 1.0612379508989764, + "learning_rate": 0.003, + "loss": 4.0854, + "step": 13041 + }, + { + "epoch": 0.13042, + "grad_norm": 1.1133253654183222, + "learning_rate": 0.003, + "loss": 4.0939, + "step": 13042 + }, + { + "epoch": 0.13043, + "grad_norm": 0.8609206423575635, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 13043 + }, + { + "epoch": 0.13044, + "grad_norm": 0.8752291817618052, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 13044 + }, + { + "epoch": 0.13045, + "grad_norm": 0.9362390396957331, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 13045 + }, + { + "epoch": 0.13046, + "grad_norm": 1.193409995888056, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 13046 + }, + { + "epoch": 0.13047, + "grad_norm": 1.0329625065014676, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 13047 + }, + { + "epoch": 0.13048, + "grad_norm": 0.9972663761213365, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 13048 + }, + { + "epoch": 0.13049, + "grad_norm": 1.0136582219072596, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 13049 + }, + { + "epoch": 0.1305, + "grad_norm": 0.9919773122206952, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 13050 + }, + { + "epoch": 0.13051, + "grad_norm": 0.981858798954194, + "learning_rate": 0.003, + "loss": 4.1033, + "step": 13051 + }, + { + "epoch": 0.13052, + "grad_norm": 0.9019372768421601, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 13052 + }, + { + "epoch": 0.13053, + "grad_norm": 0.7857066949562682, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 13053 + }, + { + "epoch": 0.13054, + "grad_norm": 0.8380410833965226, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 13054 + }, + { + "epoch": 0.13055, + "grad_norm": 0.9999804228110576, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 13055 + }, + { + "epoch": 0.13056, + "grad_norm": 1.2461833350633658, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 13056 + }, + { + "epoch": 0.13057, + "grad_norm": 0.7576564440126881, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 13057 + }, + { + "epoch": 0.13058, + "grad_norm": 0.6804465757128236, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 13058 + }, + { + "epoch": 0.13059, + "grad_norm": 0.7745936541589206, + "learning_rate": 0.003, + "loss": 4.1018, + "step": 13059 + }, + { + "epoch": 0.1306, + "grad_norm": 0.9669753059697394, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 13060 + }, + { + "epoch": 0.13061, + "grad_norm": 1.292313224854643, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 13061 + }, + { + "epoch": 0.13062, + "grad_norm": 0.8876199876868209, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 13062 + }, + { + "epoch": 0.13063, + "grad_norm": 0.9129927859818836, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 13063 + }, + { + "epoch": 0.13064, + "grad_norm": 1.0436860186741799, + "learning_rate": 0.003, + "loss": 4.1165, + "step": 13064 + }, + { + "epoch": 0.13065, + "grad_norm": 0.9580153904402132, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 13065 + }, + { + "epoch": 0.13066, + "grad_norm": 0.9013530416518812, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 13066 + }, + { + "epoch": 0.13067, + "grad_norm": 0.9276867401563169, + "learning_rate": 0.003, + "loss": 4.1303, + "step": 13067 + }, + { + "epoch": 0.13068, + "grad_norm": 1.041349862439267, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 13068 + }, + { + "epoch": 0.13069, + "grad_norm": 1.180182381237877, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 13069 + }, + { + "epoch": 0.1307, + "grad_norm": 0.93583756196018, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 13070 + }, + { + "epoch": 0.13071, + "grad_norm": 1.0069848236321042, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 13071 + }, + { + "epoch": 0.13072, + "grad_norm": 1.2903091099084212, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 13072 + }, + { + "epoch": 0.13073, + "grad_norm": 0.8432558804152402, + "learning_rate": 0.003, + "loss": 4.071, + "step": 13073 + }, + { + "epoch": 0.13074, + "grad_norm": 0.7610114466911876, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 13074 + }, + { + "epoch": 0.13075, + "grad_norm": 0.7328672539273223, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 13075 + }, + { + "epoch": 0.13076, + "grad_norm": 0.7423371562075278, + "learning_rate": 0.003, + "loss": 4.088, + "step": 13076 + }, + { + "epoch": 0.13077, + "grad_norm": 0.7917978288996428, + "learning_rate": 0.003, + "loss": 4.076, + "step": 13077 + }, + { + "epoch": 0.13078, + "grad_norm": 0.8492753928764781, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 13078 + }, + { + "epoch": 0.13079, + "grad_norm": 0.9637520597733715, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 13079 + }, + { + "epoch": 0.1308, + "grad_norm": 1.1285528213394325, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 13080 + }, + { + "epoch": 0.13081, + "grad_norm": 1.0596085868239427, + "learning_rate": 0.003, + "loss": 4.0872, + "step": 13081 + }, + { + "epoch": 0.13082, + "grad_norm": 0.944166472281693, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 13082 + }, + { + "epoch": 0.13083, + "grad_norm": 0.9958649397884562, + "learning_rate": 0.003, + "loss": 4.069, + "step": 13083 + }, + { + "epoch": 0.13084, + "grad_norm": 1.1737244744792246, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 13084 + }, + { + "epoch": 0.13085, + "grad_norm": 1.0179964670230222, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 13085 + }, + { + "epoch": 0.13086, + "grad_norm": 0.9898855256904289, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 13086 + }, + { + "epoch": 0.13087, + "grad_norm": 0.9921177059455311, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 13087 + }, + { + "epoch": 0.13088, + "grad_norm": 0.9463496624965083, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 13088 + }, + { + "epoch": 0.13089, + "grad_norm": 0.9237936365246425, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 13089 + }, + { + "epoch": 0.1309, + "grad_norm": 0.9969373427574315, + "learning_rate": 0.003, + "loss": 4.058, + "step": 13090 + }, + { + "epoch": 0.13091, + "grad_norm": 0.9150956615422726, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 13091 + }, + { + "epoch": 0.13092, + "grad_norm": 0.9660793188823901, + "learning_rate": 0.003, + "loss": 4.0819, + "step": 13092 + }, + { + "epoch": 0.13093, + "grad_norm": 1.086050730377896, + "learning_rate": 0.003, + "loss": 4.0873, + "step": 13093 + }, + { + "epoch": 0.13094, + "grad_norm": 1.0790772575820935, + "learning_rate": 0.003, + "loss": 4.089, + "step": 13094 + }, + { + "epoch": 0.13095, + "grad_norm": 0.8713473250896735, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 13095 + }, + { + "epoch": 0.13096, + "grad_norm": 0.8833956066271142, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 13096 + }, + { + "epoch": 0.13097, + "grad_norm": 0.9214225443797921, + "learning_rate": 0.003, + "loss": 4.1043, + "step": 13097 + }, + { + "epoch": 0.13098, + "grad_norm": 0.9495852382984056, + "learning_rate": 0.003, + "loss": 4.0878, + "step": 13098 + }, + { + "epoch": 0.13099, + "grad_norm": 0.961444857656626, + "learning_rate": 0.003, + "loss": 4.0891, + "step": 13099 + }, + { + "epoch": 0.131, + "grad_norm": 0.8957934329530338, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 13100 + }, + { + "epoch": 0.13101, + "grad_norm": 0.8977686723252247, + "learning_rate": 0.003, + "loss": 4.078, + "step": 13101 + }, + { + "epoch": 0.13102, + "grad_norm": 1.0015050014762186, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 13102 + }, + { + "epoch": 0.13103, + "grad_norm": 1.1447668601885452, + "learning_rate": 0.003, + "loss": 4.101, + "step": 13103 + }, + { + "epoch": 0.13104, + "grad_norm": 1.1064794265813467, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 13104 + }, + { + "epoch": 0.13105, + "grad_norm": 0.9158238403212685, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 13105 + }, + { + "epoch": 0.13106, + "grad_norm": 0.8355307666165467, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 13106 + }, + { + "epoch": 0.13107, + "grad_norm": 0.7896870513868581, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 13107 + }, + { + "epoch": 0.13108, + "grad_norm": 0.8324418347231821, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 13108 + }, + { + "epoch": 0.13109, + "grad_norm": 0.8490971237080954, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 13109 + }, + { + "epoch": 0.1311, + "grad_norm": 0.7572539057344594, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 13110 + }, + { + "epoch": 0.13111, + "grad_norm": 0.775452357123476, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 13111 + }, + { + "epoch": 0.13112, + "grad_norm": 0.8401299759420662, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 13112 + }, + { + "epoch": 0.13113, + "grad_norm": 0.9229274804068884, + "learning_rate": 0.003, + "loss": 4.1076, + "step": 13113 + }, + { + "epoch": 0.13114, + "grad_norm": 1.2532681577118807, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 13114 + }, + { + "epoch": 0.13115, + "grad_norm": 1.0392192809212526, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 13115 + }, + { + "epoch": 0.13116, + "grad_norm": 0.9639449665899528, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 13116 + }, + { + "epoch": 0.13117, + "grad_norm": 0.934938387341886, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 13117 + }, + { + "epoch": 0.13118, + "grad_norm": 1.0469920486145603, + "learning_rate": 0.003, + "loss": 4.1024, + "step": 13118 + }, + { + "epoch": 0.13119, + "grad_norm": 1.0171310803783347, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 13119 + }, + { + "epoch": 0.1312, + "grad_norm": 1.1151947006811234, + "learning_rate": 0.003, + "loss": 4.0943, + "step": 13120 + }, + { + "epoch": 0.13121, + "grad_norm": 0.884724966316226, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 13121 + }, + { + "epoch": 0.13122, + "grad_norm": 0.8986014413373032, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 13122 + }, + { + "epoch": 0.13123, + "grad_norm": 0.9593088033059508, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 13123 + }, + { + "epoch": 0.13124, + "grad_norm": 0.8478284001697868, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 13124 + }, + { + "epoch": 0.13125, + "grad_norm": 0.7031880131637845, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 13125 + }, + { + "epoch": 0.13126, + "grad_norm": 0.6840388518023653, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 13126 + }, + { + "epoch": 0.13127, + "grad_norm": 0.6854571944200158, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 13127 + }, + { + "epoch": 0.13128, + "grad_norm": 0.8329320957889822, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 13128 + }, + { + "epoch": 0.13129, + "grad_norm": 1.2356909677832044, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 13129 + }, + { + "epoch": 0.1313, + "grad_norm": 1.111237033340105, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 13130 + }, + { + "epoch": 0.13131, + "grad_norm": 0.8814344117852264, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 13131 + }, + { + "epoch": 0.13132, + "grad_norm": 0.9671820829857136, + "learning_rate": 0.003, + "loss": 4.0907, + "step": 13132 + }, + { + "epoch": 0.13133, + "grad_norm": 1.023395272073073, + "learning_rate": 0.003, + "loss": 4.1115, + "step": 13133 + }, + { + "epoch": 0.13134, + "grad_norm": 0.9768542030724278, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 13134 + }, + { + "epoch": 0.13135, + "grad_norm": 1.091963108482382, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 13135 + }, + { + "epoch": 0.13136, + "grad_norm": 1.1298741348852466, + "learning_rate": 0.003, + "loss": 4.0931, + "step": 13136 + }, + { + "epoch": 0.13137, + "grad_norm": 1.0480596828742936, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 13137 + }, + { + "epoch": 0.13138, + "grad_norm": 0.9441539045689533, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 13138 + }, + { + "epoch": 0.13139, + "grad_norm": 1.0066296877968466, + "learning_rate": 0.003, + "loss": 4.0973, + "step": 13139 + }, + { + "epoch": 0.1314, + "grad_norm": 1.0109662321600608, + "learning_rate": 0.003, + "loss": 4.1039, + "step": 13140 + }, + { + "epoch": 0.13141, + "grad_norm": 1.0643944326908674, + "learning_rate": 0.003, + "loss": 4.0984, + "step": 13141 + }, + { + "epoch": 0.13142, + "grad_norm": 0.944641040650044, + "learning_rate": 0.003, + "loss": 4.117, + "step": 13142 + }, + { + "epoch": 0.13143, + "grad_norm": 0.9870786822533083, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 13143 + }, + { + "epoch": 0.13144, + "grad_norm": 0.8971194827742611, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 13144 + }, + { + "epoch": 0.13145, + "grad_norm": 0.8983633995322297, + "learning_rate": 0.003, + "loss": 4.061, + "step": 13145 + }, + { + "epoch": 0.13146, + "grad_norm": 1.0702901315895497, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 13146 + }, + { + "epoch": 0.13147, + "grad_norm": 0.9715834251837631, + "learning_rate": 0.003, + "loss": 4.0949, + "step": 13147 + }, + { + "epoch": 0.13148, + "grad_norm": 1.2524950171791185, + "learning_rate": 0.003, + "loss": 4.1083, + "step": 13148 + }, + { + "epoch": 0.13149, + "grad_norm": 1.1263278070711589, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 13149 + }, + { + "epoch": 0.1315, + "grad_norm": 1.004923675418753, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 13150 + }, + { + "epoch": 0.13151, + "grad_norm": 1.1177033305767101, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 13151 + }, + { + "epoch": 0.13152, + "grad_norm": 0.9117610513910346, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 13152 + }, + { + "epoch": 0.13153, + "grad_norm": 0.8002012052113183, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 13153 + }, + { + "epoch": 0.13154, + "grad_norm": 0.8040158837949339, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 13154 + }, + { + "epoch": 0.13155, + "grad_norm": 0.9373031695523488, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 13155 + }, + { + "epoch": 0.13156, + "grad_norm": 1.146412853986717, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 13156 + }, + { + "epoch": 0.13157, + "grad_norm": 0.8699389236663466, + "learning_rate": 0.003, + "loss": 4.1152, + "step": 13157 + }, + { + "epoch": 0.13158, + "grad_norm": 0.9778267286624542, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 13158 + }, + { + "epoch": 0.13159, + "grad_norm": 1.0136594116476791, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 13159 + }, + { + "epoch": 0.1316, + "grad_norm": 1.1581267408433948, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 13160 + }, + { + "epoch": 0.13161, + "grad_norm": 1.1082482319156306, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 13161 + }, + { + "epoch": 0.13162, + "grad_norm": 0.9360646781106823, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 13162 + }, + { + "epoch": 0.13163, + "grad_norm": 0.9134962438554866, + "learning_rate": 0.003, + "loss": 4.1181, + "step": 13163 + }, + { + "epoch": 0.13164, + "grad_norm": 0.8877003879790046, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 13164 + }, + { + "epoch": 0.13165, + "grad_norm": 0.8937372457943279, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 13165 + }, + { + "epoch": 0.13166, + "grad_norm": 1.0248902201638521, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 13166 + }, + { + "epoch": 0.13167, + "grad_norm": 1.0484623193590072, + "learning_rate": 0.003, + "loss": 4.0942, + "step": 13167 + }, + { + "epoch": 0.13168, + "grad_norm": 0.9112340901062946, + "learning_rate": 0.003, + "loss": 4.1012, + "step": 13168 + }, + { + "epoch": 0.13169, + "grad_norm": 0.9371373377960317, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 13169 + }, + { + "epoch": 0.1317, + "grad_norm": 1.034574551338795, + "learning_rate": 0.003, + "loss": 4.091, + "step": 13170 + }, + { + "epoch": 0.13171, + "grad_norm": 1.06320518663373, + "learning_rate": 0.003, + "loss": 4.1081, + "step": 13171 + }, + { + "epoch": 0.13172, + "grad_norm": 1.0630505718209204, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 13172 + }, + { + "epoch": 0.13173, + "grad_norm": 0.8395997689323885, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 13173 + }, + { + "epoch": 0.13174, + "grad_norm": 0.9828761599552335, + "learning_rate": 0.003, + "loss": 4.106, + "step": 13174 + }, + { + "epoch": 0.13175, + "grad_norm": 1.1357585278657651, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 13175 + }, + { + "epoch": 0.13176, + "grad_norm": 1.0334748445349797, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 13176 + }, + { + "epoch": 0.13177, + "grad_norm": 1.0691971167747971, + "learning_rate": 0.003, + "loss": 4.0881, + "step": 13177 + }, + { + "epoch": 0.13178, + "grad_norm": 1.074438851556422, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 13178 + }, + { + "epoch": 0.13179, + "grad_norm": 0.9769631671460902, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 13179 + }, + { + "epoch": 0.1318, + "grad_norm": 0.9731857249969333, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 13180 + }, + { + "epoch": 0.13181, + "grad_norm": 0.8648096774103434, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 13181 + }, + { + "epoch": 0.13182, + "grad_norm": 0.8135811556888828, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 13182 + }, + { + "epoch": 0.13183, + "grad_norm": 0.8332300799550404, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 13183 + }, + { + "epoch": 0.13184, + "grad_norm": 1.0913175142839042, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 13184 + }, + { + "epoch": 0.13185, + "grad_norm": 1.2409678226566652, + "learning_rate": 0.003, + "loss": 4.1001, + "step": 13185 + }, + { + "epoch": 0.13186, + "grad_norm": 0.8289954445670332, + "learning_rate": 0.003, + "loss": 4.0888, + "step": 13186 + }, + { + "epoch": 0.13187, + "grad_norm": 0.8442882839475516, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 13187 + }, + { + "epoch": 0.13188, + "grad_norm": 0.8661065090242185, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 13188 + }, + { + "epoch": 0.13189, + "grad_norm": 0.718811535315877, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 13189 + }, + { + "epoch": 0.1319, + "grad_norm": 0.7756808233147338, + "learning_rate": 0.003, + "loss": 4.1044, + "step": 13190 + }, + { + "epoch": 0.13191, + "grad_norm": 0.83787766630841, + "learning_rate": 0.003, + "loss": 4.0971, + "step": 13191 + }, + { + "epoch": 0.13192, + "grad_norm": 0.9367556933750846, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 13192 + }, + { + "epoch": 0.13193, + "grad_norm": 1.1587818206896392, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 13193 + }, + { + "epoch": 0.13194, + "grad_norm": 0.8287117704106864, + "learning_rate": 0.003, + "loss": 4.072, + "step": 13194 + }, + { + "epoch": 0.13195, + "grad_norm": 0.8305114174867917, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 13195 + }, + { + "epoch": 0.13196, + "grad_norm": 0.8589421064477466, + "learning_rate": 0.003, + "loss": 4.031, + "step": 13196 + }, + { + "epoch": 0.13197, + "grad_norm": 0.9735893841689806, + "learning_rate": 0.003, + "loss": 4.0901, + "step": 13197 + }, + { + "epoch": 0.13198, + "grad_norm": 0.9964332537878365, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 13198 + }, + { + "epoch": 0.13199, + "grad_norm": 1.150720421494735, + "learning_rate": 0.003, + "loss": 4.0891, + "step": 13199 + }, + { + "epoch": 0.132, + "grad_norm": 0.875952359893899, + "learning_rate": 0.003, + "loss": 4.0874, + "step": 13200 + }, + { + "epoch": 0.13201, + "grad_norm": 0.9954252486283677, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 13201 + }, + { + "epoch": 0.13202, + "grad_norm": 0.9563259506027956, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 13202 + }, + { + "epoch": 0.13203, + "grad_norm": 1.0725835818001632, + "learning_rate": 0.003, + "loss": 4.0916, + "step": 13203 + }, + { + "epoch": 0.13204, + "grad_norm": 1.0015935627486392, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 13204 + }, + { + "epoch": 0.13205, + "grad_norm": 0.9837397051930121, + "learning_rate": 0.003, + "loss": 4.0888, + "step": 13205 + }, + { + "epoch": 0.13206, + "grad_norm": 1.0739978834960404, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 13206 + }, + { + "epoch": 0.13207, + "grad_norm": 1.1985187158541095, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 13207 + }, + { + "epoch": 0.13208, + "grad_norm": 0.9678089702241485, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 13208 + }, + { + "epoch": 0.13209, + "grad_norm": 0.9986044618691747, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 13209 + }, + { + "epoch": 0.1321, + "grad_norm": 0.9865706910160242, + "learning_rate": 0.003, + "loss": 4.1143, + "step": 13210 + }, + { + "epoch": 0.13211, + "grad_norm": 0.9916216383444284, + "learning_rate": 0.003, + "loss": 4.0944, + "step": 13211 + }, + { + "epoch": 0.13212, + "grad_norm": 0.9823308592075966, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 13212 + }, + { + "epoch": 0.13213, + "grad_norm": 0.8058609999776402, + "learning_rate": 0.003, + "loss": 4.0902, + "step": 13213 + }, + { + "epoch": 0.13214, + "grad_norm": 0.6905179195042105, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 13214 + }, + { + "epoch": 0.13215, + "grad_norm": 0.7108919419017576, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 13215 + }, + { + "epoch": 0.13216, + "grad_norm": 0.7120611290593309, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 13216 + }, + { + "epoch": 0.13217, + "grad_norm": 0.7534009013522089, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 13217 + }, + { + "epoch": 0.13218, + "grad_norm": 0.8163245462699045, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 13218 + }, + { + "epoch": 0.13219, + "grad_norm": 0.9435420619864289, + "learning_rate": 0.003, + "loss": 4.1265, + "step": 13219 + }, + { + "epoch": 0.1322, + "grad_norm": 1.140360847855267, + "learning_rate": 0.003, + "loss": 4.1147, + "step": 13220 + }, + { + "epoch": 0.13221, + "grad_norm": 0.888388947955735, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 13221 + }, + { + "epoch": 0.13222, + "grad_norm": 0.9658777240254408, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 13222 + }, + { + "epoch": 0.13223, + "grad_norm": 1.162500644347836, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 13223 + }, + { + "epoch": 0.13224, + "grad_norm": 0.9041397260936085, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 13224 + }, + { + "epoch": 0.13225, + "grad_norm": 1.0859354261488787, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 13225 + }, + { + "epoch": 0.13226, + "grad_norm": 1.1574919782046347, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 13226 + }, + { + "epoch": 0.13227, + "grad_norm": 0.9247096897926775, + "learning_rate": 0.003, + "loss": 4.0927, + "step": 13227 + }, + { + "epoch": 0.13228, + "grad_norm": 0.86229111424133, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 13228 + }, + { + "epoch": 0.13229, + "grad_norm": 0.847560257243791, + "learning_rate": 0.003, + "loss": 4.0874, + "step": 13229 + }, + { + "epoch": 0.1323, + "grad_norm": 0.8311750204841981, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 13230 + }, + { + "epoch": 0.13231, + "grad_norm": 1.0844257597005231, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 13231 + }, + { + "epoch": 0.13232, + "grad_norm": 1.2922856401845064, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 13232 + }, + { + "epoch": 0.13233, + "grad_norm": 0.9368522396658217, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 13233 + }, + { + "epoch": 0.13234, + "grad_norm": 1.0004863047334178, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 13234 + }, + { + "epoch": 0.13235, + "grad_norm": 1.0373205714588634, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 13235 + }, + { + "epoch": 0.13236, + "grad_norm": 0.8434203473892394, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 13236 + }, + { + "epoch": 0.13237, + "grad_norm": 0.7184174140462061, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 13237 + }, + { + "epoch": 0.13238, + "grad_norm": 0.7434152316833053, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 13238 + }, + { + "epoch": 0.13239, + "grad_norm": 0.9039218301772489, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 13239 + }, + { + "epoch": 0.1324, + "grad_norm": 1.0137774538726745, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 13240 + }, + { + "epoch": 0.13241, + "grad_norm": 1.1601737932982843, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 13241 + }, + { + "epoch": 0.13242, + "grad_norm": 0.8843485769629014, + "learning_rate": 0.003, + "loss": 4.101, + "step": 13242 + }, + { + "epoch": 0.13243, + "grad_norm": 0.9178358573900722, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 13243 + }, + { + "epoch": 0.13244, + "grad_norm": 1.341189378864638, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 13244 + }, + { + "epoch": 0.13245, + "grad_norm": 1.0472287768221853, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 13245 + }, + { + "epoch": 0.13246, + "grad_norm": 1.1388993727405217, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 13246 + }, + { + "epoch": 0.13247, + "grad_norm": 1.2089923077236926, + "learning_rate": 0.003, + "loss": 4.1043, + "step": 13247 + }, + { + "epoch": 0.13248, + "grad_norm": 0.9655493938395205, + "learning_rate": 0.003, + "loss": 4.1026, + "step": 13248 + }, + { + "epoch": 0.13249, + "grad_norm": 0.8773258975913082, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 13249 + }, + { + "epoch": 0.1325, + "grad_norm": 0.9071781383333825, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 13250 + }, + { + "epoch": 0.13251, + "grad_norm": 0.8732550659090038, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 13251 + }, + { + "epoch": 0.13252, + "grad_norm": 0.9263174740349942, + "learning_rate": 0.003, + "loss": 4.0922, + "step": 13252 + }, + { + "epoch": 0.13253, + "grad_norm": 0.993174334434721, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 13253 + }, + { + "epoch": 0.13254, + "grad_norm": 0.995053301010391, + "learning_rate": 0.003, + "loss": 4.1074, + "step": 13254 + }, + { + "epoch": 0.13255, + "grad_norm": 1.055215362447066, + "learning_rate": 0.003, + "loss": 4.0988, + "step": 13255 + }, + { + "epoch": 0.13256, + "grad_norm": 1.0917630183359468, + "learning_rate": 0.003, + "loss": 4.0962, + "step": 13256 + }, + { + "epoch": 0.13257, + "grad_norm": 0.9793608044370317, + "learning_rate": 0.003, + "loss": 4.0888, + "step": 13257 + }, + { + "epoch": 0.13258, + "grad_norm": 0.9481561830715173, + "learning_rate": 0.003, + "loss": 4.082, + "step": 13258 + }, + { + "epoch": 0.13259, + "grad_norm": 0.9572004505884641, + "learning_rate": 0.003, + "loss": 4.104, + "step": 13259 + }, + { + "epoch": 0.1326, + "grad_norm": 1.209374982142481, + "learning_rate": 0.003, + "loss": 4.0936, + "step": 13260 + }, + { + "epoch": 0.13261, + "grad_norm": 1.1504132809973286, + "learning_rate": 0.003, + "loss": 4.0888, + "step": 13261 + }, + { + "epoch": 0.13262, + "grad_norm": 0.9513902908380574, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 13262 + }, + { + "epoch": 0.13263, + "grad_norm": 0.8730836242735304, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 13263 + }, + { + "epoch": 0.13264, + "grad_norm": 0.6766072519021229, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 13264 + }, + { + "epoch": 0.13265, + "grad_norm": 0.5746449091561792, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 13265 + }, + { + "epoch": 0.13266, + "grad_norm": 0.6415557553815574, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 13266 + }, + { + "epoch": 0.13267, + "grad_norm": 0.5748073407271075, + "learning_rate": 0.003, + "loss": 4.071, + "step": 13267 + }, + { + "epoch": 0.13268, + "grad_norm": 0.5985903830415835, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 13268 + }, + { + "epoch": 0.13269, + "grad_norm": 0.654069400633266, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 13269 + }, + { + "epoch": 0.1327, + "grad_norm": 0.7702848716087565, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 13270 + }, + { + "epoch": 0.13271, + "grad_norm": 0.842644588053912, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 13271 + }, + { + "epoch": 0.13272, + "grad_norm": 0.8994642542307124, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 13272 + }, + { + "epoch": 0.13273, + "grad_norm": 0.9729953630190503, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 13273 + }, + { + "epoch": 0.13274, + "grad_norm": 1.0284308320720505, + "learning_rate": 0.003, + "loss": 4.102, + "step": 13274 + }, + { + "epoch": 0.13275, + "grad_norm": 1.1323132833306477, + "learning_rate": 0.003, + "loss": 4.0952, + "step": 13275 + }, + { + "epoch": 0.13276, + "grad_norm": 0.9581172426334902, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 13276 + }, + { + "epoch": 0.13277, + "grad_norm": 0.9936339189366039, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 13277 + }, + { + "epoch": 0.13278, + "grad_norm": 1.3155569896073163, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 13278 + }, + { + "epoch": 0.13279, + "grad_norm": 0.924090519405608, + "learning_rate": 0.003, + "loss": 4.0933, + "step": 13279 + }, + { + "epoch": 0.1328, + "grad_norm": 0.9790904106075842, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 13280 + }, + { + "epoch": 0.13281, + "grad_norm": 1.1311781070997495, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 13281 + }, + { + "epoch": 0.13282, + "grad_norm": 0.8604962773523706, + "learning_rate": 0.003, + "loss": 4.0966, + "step": 13282 + }, + { + "epoch": 0.13283, + "grad_norm": 0.7730626935626278, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 13283 + }, + { + "epoch": 0.13284, + "grad_norm": 0.8088614099480749, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 13284 + }, + { + "epoch": 0.13285, + "grad_norm": 0.9512360712199578, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 13285 + }, + { + "epoch": 0.13286, + "grad_norm": 1.0167161497591282, + "learning_rate": 0.003, + "loss": 4.1087, + "step": 13286 + }, + { + "epoch": 0.13287, + "grad_norm": 1.061557091271832, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 13287 + }, + { + "epoch": 0.13288, + "grad_norm": 1.0627911533250975, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 13288 + }, + { + "epoch": 0.13289, + "grad_norm": 1.0048690746103652, + "learning_rate": 0.003, + "loss": 4.1059, + "step": 13289 + }, + { + "epoch": 0.1329, + "grad_norm": 1.0438052180974682, + "learning_rate": 0.003, + "loss": 4.1062, + "step": 13290 + }, + { + "epoch": 0.13291, + "grad_norm": 1.0419060669277742, + "learning_rate": 0.003, + "loss": 4.085, + "step": 13291 + }, + { + "epoch": 0.13292, + "grad_norm": 0.934043659202913, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 13292 + }, + { + "epoch": 0.13293, + "grad_norm": 0.9786640056741102, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 13293 + }, + { + "epoch": 0.13294, + "grad_norm": 1.0558775951661954, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 13294 + }, + { + "epoch": 0.13295, + "grad_norm": 0.9634584997545496, + "learning_rate": 0.003, + "loss": 4.0873, + "step": 13295 + }, + { + "epoch": 0.13296, + "grad_norm": 1.0133561645224145, + "learning_rate": 0.003, + "loss": 4.1152, + "step": 13296 + }, + { + "epoch": 0.13297, + "grad_norm": 1.1951735490232744, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 13297 + }, + { + "epoch": 0.13298, + "grad_norm": 0.8212065453599408, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 13298 + }, + { + "epoch": 0.13299, + "grad_norm": 0.7013678928904146, + "learning_rate": 0.003, + "loss": 4.1224, + "step": 13299 + }, + { + "epoch": 0.133, + "grad_norm": 0.7354365193332403, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 13300 + }, + { + "epoch": 0.13301, + "grad_norm": 0.7683448120275547, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 13301 + }, + { + "epoch": 0.13302, + "grad_norm": 0.8195995669562756, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 13302 + }, + { + "epoch": 0.13303, + "grad_norm": 0.9852910588379463, + "learning_rate": 0.003, + "loss": 4.0941, + "step": 13303 + }, + { + "epoch": 0.13304, + "grad_norm": 1.1719008115246252, + "learning_rate": 0.003, + "loss": 4.0944, + "step": 13304 + }, + { + "epoch": 0.13305, + "grad_norm": 1.006637035029354, + "learning_rate": 0.003, + "loss": 4.1057, + "step": 13305 + }, + { + "epoch": 0.13306, + "grad_norm": 1.1049975429049455, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 13306 + }, + { + "epoch": 0.13307, + "grad_norm": 0.8698321446151928, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 13307 + }, + { + "epoch": 0.13308, + "grad_norm": 0.8236360656258656, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 13308 + }, + { + "epoch": 0.13309, + "grad_norm": 0.9004750483393915, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 13309 + }, + { + "epoch": 0.1331, + "grad_norm": 1.1394111471601216, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 13310 + }, + { + "epoch": 0.13311, + "grad_norm": 0.9436963358912399, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 13311 + }, + { + "epoch": 0.13312, + "grad_norm": 0.9778251156819693, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 13312 + }, + { + "epoch": 0.13313, + "grad_norm": 1.183222919779824, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 13313 + }, + { + "epoch": 0.13314, + "grad_norm": 0.9233108843902137, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 13314 + }, + { + "epoch": 0.13315, + "grad_norm": 0.8593357635257551, + "learning_rate": 0.003, + "loss": 4.0902, + "step": 13315 + }, + { + "epoch": 0.13316, + "grad_norm": 0.838121489407412, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 13316 + }, + { + "epoch": 0.13317, + "grad_norm": 0.9044991124637792, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 13317 + }, + { + "epoch": 0.13318, + "grad_norm": 0.9010375087559165, + "learning_rate": 0.003, + "loss": 4.079, + "step": 13318 + }, + { + "epoch": 0.13319, + "grad_norm": 0.7704620252799368, + "learning_rate": 0.003, + "loss": 4.0935, + "step": 13319 + }, + { + "epoch": 0.1332, + "grad_norm": 0.7653645068877163, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 13320 + }, + { + "epoch": 0.13321, + "grad_norm": 0.7648906542727318, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 13321 + }, + { + "epoch": 0.13322, + "grad_norm": 0.8722468929644939, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 13322 + }, + { + "epoch": 0.13323, + "grad_norm": 1.0710192266538265, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 13323 + }, + { + "epoch": 0.13324, + "grad_norm": 1.1579064884138774, + "learning_rate": 0.003, + "loss": 4.0926, + "step": 13324 + }, + { + "epoch": 0.13325, + "grad_norm": 0.929037537476983, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 13325 + }, + { + "epoch": 0.13326, + "grad_norm": 0.7862165915003471, + "learning_rate": 0.003, + "loss": 4.1011, + "step": 13326 + }, + { + "epoch": 0.13327, + "grad_norm": 0.8654931436439418, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 13327 + }, + { + "epoch": 0.13328, + "grad_norm": 0.9781721854442568, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 13328 + }, + { + "epoch": 0.13329, + "grad_norm": 1.0602578966459246, + "learning_rate": 0.003, + "loss": 4.101, + "step": 13329 + }, + { + "epoch": 0.1333, + "grad_norm": 1.0310756864940578, + "learning_rate": 0.003, + "loss": 4.101, + "step": 13330 + }, + { + "epoch": 0.13331, + "grad_norm": 0.9668623183975973, + "learning_rate": 0.003, + "loss": 4.0858, + "step": 13331 + }, + { + "epoch": 0.13332, + "grad_norm": 0.8255027975232511, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 13332 + }, + { + "epoch": 0.13333, + "grad_norm": 0.8951924915431422, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 13333 + }, + { + "epoch": 0.13334, + "grad_norm": 1.006936074318859, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 13334 + }, + { + "epoch": 0.13335, + "grad_norm": 1.0761834273487745, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 13335 + }, + { + "epoch": 0.13336, + "grad_norm": 0.997553433401841, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 13336 + }, + { + "epoch": 0.13337, + "grad_norm": 1.0966349599187795, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 13337 + }, + { + "epoch": 0.13338, + "grad_norm": 1.0673305938452706, + "learning_rate": 0.003, + "loss": 4.1176, + "step": 13338 + }, + { + "epoch": 0.13339, + "grad_norm": 1.2002296845597484, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 13339 + }, + { + "epoch": 0.1334, + "grad_norm": 0.9014765403388033, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 13340 + }, + { + "epoch": 0.13341, + "grad_norm": 0.759883775044865, + "learning_rate": 0.003, + "loss": 4.1144, + "step": 13341 + }, + { + "epoch": 0.13342, + "grad_norm": 0.664655615833606, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 13342 + }, + { + "epoch": 0.13343, + "grad_norm": 0.7715316632136893, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 13343 + }, + { + "epoch": 0.13344, + "grad_norm": 0.8779180507235521, + "learning_rate": 0.003, + "loss": 4.1088, + "step": 13344 + }, + { + "epoch": 0.13345, + "grad_norm": 1.158875710747508, + "learning_rate": 0.003, + "loss": 4.1208, + "step": 13345 + }, + { + "epoch": 0.13346, + "grad_norm": 0.9967558032816967, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 13346 + }, + { + "epoch": 0.13347, + "grad_norm": 1.0726959081200256, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 13347 + }, + { + "epoch": 0.13348, + "grad_norm": 0.8483208599708412, + "learning_rate": 0.003, + "loss": 4.0952, + "step": 13348 + }, + { + "epoch": 0.13349, + "grad_norm": 0.760504083191273, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 13349 + }, + { + "epoch": 0.1335, + "grad_norm": 0.7374874491820916, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 13350 + }, + { + "epoch": 0.13351, + "grad_norm": 0.7757261205088355, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 13351 + }, + { + "epoch": 0.13352, + "grad_norm": 1.0852487061541132, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 13352 + }, + { + "epoch": 0.13353, + "grad_norm": 1.2653717572732779, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 13353 + }, + { + "epoch": 0.13354, + "grad_norm": 0.8531127853236398, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 13354 + }, + { + "epoch": 0.13355, + "grad_norm": 0.8168414411956694, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 13355 + }, + { + "epoch": 0.13356, + "grad_norm": 0.8721310955009899, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 13356 + }, + { + "epoch": 0.13357, + "grad_norm": 0.8691467870344313, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 13357 + }, + { + "epoch": 0.13358, + "grad_norm": 0.83343537223158, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 13358 + }, + { + "epoch": 0.13359, + "grad_norm": 1.0035818778251735, + "learning_rate": 0.003, + "loss": 4.093, + "step": 13359 + }, + { + "epoch": 0.1336, + "grad_norm": 1.4334054442790312, + "learning_rate": 0.003, + "loss": 4.0919, + "step": 13360 + }, + { + "epoch": 0.13361, + "grad_norm": 0.7414273046940859, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 13361 + }, + { + "epoch": 0.13362, + "grad_norm": 0.7078761081862197, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 13362 + }, + { + "epoch": 0.13363, + "grad_norm": 0.8049930138016516, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 13363 + }, + { + "epoch": 0.13364, + "grad_norm": 0.8030005269768123, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 13364 + }, + { + "epoch": 0.13365, + "grad_norm": 0.8875868034541989, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 13365 + }, + { + "epoch": 0.13366, + "grad_norm": 0.9390275021725378, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 13366 + }, + { + "epoch": 0.13367, + "grad_norm": 0.9793821654316988, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 13367 + }, + { + "epoch": 0.13368, + "grad_norm": 1.3707857877316363, + "learning_rate": 0.003, + "loss": 4.1068, + "step": 13368 + }, + { + "epoch": 0.13369, + "grad_norm": 1.1075505267683585, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 13369 + }, + { + "epoch": 0.1337, + "grad_norm": 1.0065114476940287, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 13370 + }, + { + "epoch": 0.13371, + "grad_norm": 1.178034786355604, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 13371 + }, + { + "epoch": 0.13372, + "grad_norm": 0.8923441765594062, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 13372 + }, + { + "epoch": 0.13373, + "grad_norm": 0.8475050616973431, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 13373 + }, + { + "epoch": 0.13374, + "grad_norm": 0.8635503958941468, + "learning_rate": 0.003, + "loss": 4.1033, + "step": 13374 + }, + { + "epoch": 0.13375, + "grad_norm": 1.0393070407270935, + "learning_rate": 0.003, + "loss": 4.056, + "step": 13375 + }, + { + "epoch": 0.13376, + "grad_norm": 1.089829908136605, + "learning_rate": 0.003, + "loss": 4.0944, + "step": 13376 + }, + { + "epoch": 0.13377, + "grad_norm": 1.0205148226842657, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 13377 + }, + { + "epoch": 0.13378, + "grad_norm": 0.9547179135443791, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 13378 + }, + { + "epoch": 0.13379, + "grad_norm": 0.8353793622201402, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 13379 + }, + { + "epoch": 0.1338, + "grad_norm": 0.8155884485745533, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 13380 + }, + { + "epoch": 0.13381, + "grad_norm": 0.8352367832452048, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 13381 + }, + { + "epoch": 0.13382, + "grad_norm": 0.6818805586312642, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 13382 + }, + { + "epoch": 0.13383, + "grad_norm": 0.7173534006991076, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 13383 + }, + { + "epoch": 0.13384, + "grad_norm": 0.685162974458706, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 13384 + }, + { + "epoch": 0.13385, + "grad_norm": 0.9417969774034519, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 13385 + }, + { + "epoch": 0.13386, + "grad_norm": 1.3639952674818814, + "learning_rate": 0.003, + "loss": 4.0916, + "step": 13386 + }, + { + "epoch": 0.13387, + "grad_norm": 0.7476407405175547, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 13387 + }, + { + "epoch": 0.13388, + "grad_norm": 0.7517452766090448, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 13388 + }, + { + "epoch": 0.13389, + "grad_norm": 0.8650876576445697, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 13389 + }, + { + "epoch": 0.1339, + "grad_norm": 0.9242581975296174, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 13390 + }, + { + "epoch": 0.13391, + "grad_norm": 0.8798262317323962, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 13391 + }, + { + "epoch": 0.13392, + "grad_norm": 0.9228688749800353, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 13392 + }, + { + "epoch": 0.13393, + "grad_norm": 1.160999814969636, + "learning_rate": 0.003, + "loss": 4.0996, + "step": 13393 + }, + { + "epoch": 0.13394, + "grad_norm": 0.9580353225260979, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 13394 + }, + { + "epoch": 0.13395, + "grad_norm": 1.0812014354204096, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 13395 + }, + { + "epoch": 0.13396, + "grad_norm": 0.8366134640146347, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 13396 + }, + { + "epoch": 0.13397, + "grad_norm": 1.0499679849883803, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 13397 + }, + { + "epoch": 0.13398, + "grad_norm": 1.3452545378308736, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 13398 + }, + { + "epoch": 0.13399, + "grad_norm": 0.8074878010970381, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 13399 + }, + { + "epoch": 0.134, + "grad_norm": 0.9411545816309467, + "learning_rate": 0.003, + "loss": 4.106, + "step": 13400 + }, + { + "epoch": 0.13401, + "grad_norm": 1.0882061106654408, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 13401 + }, + { + "epoch": 0.13402, + "grad_norm": 1.3633804499711086, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 13402 + }, + { + "epoch": 0.13403, + "grad_norm": 0.9811953538382596, + "learning_rate": 0.003, + "loss": 4.1164, + "step": 13403 + }, + { + "epoch": 0.13404, + "grad_norm": 1.1265599386159733, + "learning_rate": 0.003, + "loss": 4.1121, + "step": 13404 + }, + { + "epoch": 0.13405, + "grad_norm": 1.0461968569044715, + "learning_rate": 0.003, + "loss": 4.074, + "step": 13405 + }, + { + "epoch": 0.13406, + "grad_norm": 0.9639351650168391, + "learning_rate": 0.003, + "loss": 4.0912, + "step": 13406 + }, + { + "epoch": 0.13407, + "grad_norm": 1.0476918262188137, + "learning_rate": 0.003, + "loss": 4.0937, + "step": 13407 + }, + { + "epoch": 0.13408, + "grad_norm": 1.0366662856573792, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 13408 + }, + { + "epoch": 0.13409, + "grad_norm": 1.0880416862405284, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 13409 + }, + { + "epoch": 0.1341, + "grad_norm": 1.1005099609031432, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 13410 + }, + { + "epoch": 0.13411, + "grad_norm": 1.0439133456839813, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 13411 + }, + { + "epoch": 0.13412, + "grad_norm": 0.9010147504941389, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 13412 + }, + { + "epoch": 0.13413, + "grad_norm": 0.9134035415226582, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 13413 + }, + { + "epoch": 0.13414, + "grad_norm": 0.886166532785208, + "learning_rate": 0.003, + "loss": 4.0975, + "step": 13414 + }, + { + "epoch": 0.13415, + "grad_norm": 0.9168065123701771, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 13415 + }, + { + "epoch": 0.13416, + "grad_norm": 1.153769898830424, + "learning_rate": 0.003, + "loss": 4.0871, + "step": 13416 + }, + { + "epoch": 0.13417, + "grad_norm": 0.8981233228464326, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 13417 + }, + { + "epoch": 0.13418, + "grad_norm": 0.8624398690566797, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 13418 + }, + { + "epoch": 0.13419, + "grad_norm": 1.0108438090909917, + "learning_rate": 0.003, + "loss": 4.0978, + "step": 13419 + }, + { + "epoch": 0.1342, + "grad_norm": 0.9937150923503353, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 13420 + }, + { + "epoch": 0.13421, + "grad_norm": 0.9830132951936665, + "learning_rate": 0.003, + "loss": 4.0975, + "step": 13421 + }, + { + "epoch": 0.13422, + "grad_norm": 0.9494757126047056, + "learning_rate": 0.003, + "loss": 4.1092, + "step": 13422 + }, + { + "epoch": 0.13423, + "grad_norm": 0.9506102495082208, + "learning_rate": 0.003, + "loss": 4.0881, + "step": 13423 + }, + { + "epoch": 0.13424, + "grad_norm": 1.2735645053801916, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 13424 + }, + { + "epoch": 0.13425, + "grad_norm": 1.0611208588967973, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 13425 + }, + { + "epoch": 0.13426, + "grad_norm": 0.8989972695896012, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 13426 + }, + { + "epoch": 0.13427, + "grad_norm": 1.0180947681504986, + "learning_rate": 0.003, + "loss": 4.0945, + "step": 13427 + }, + { + "epoch": 0.13428, + "grad_norm": 1.2125105924796709, + "learning_rate": 0.003, + "loss": 4.1074, + "step": 13428 + }, + { + "epoch": 0.13429, + "grad_norm": 0.8423863012889894, + "learning_rate": 0.003, + "loss": 4.0977, + "step": 13429 + }, + { + "epoch": 0.1343, + "grad_norm": 0.7919311307700649, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 13430 + }, + { + "epoch": 0.13431, + "grad_norm": 0.961793218644166, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 13431 + }, + { + "epoch": 0.13432, + "grad_norm": 1.2387453306243899, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 13432 + }, + { + "epoch": 0.13433, + "grad_norm": 0.9169518708542719, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 13433 + }, + { + "epoch": 0.13434, + "grad_norm": 0.947105297518013, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 13434 + }, + { + "epoch": 0.13435, + "grad_norm": 0.9839090874391867, + "learning_rate": 0.003, + "loss": 4.1109, + "step": 13435 + }, + { + "epoch": 0.13436, + "grad_norm": 1.0444241754536847, + "learning_rate": 0.003, + "loss": 4.103, + "step": 13436 + }, + { + "epoch": 0.13437, + "grad_norm": 1.0218727231717633, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 13437 + }, + { + "epoch": 0.13438, + "grad_norm": 1.166519262508102, + "learning_rate": 0.003, + "loss": 4.0953, + "step": 13438 + }, + { + "epoch": 0.13439, + "grad_norm": 0.7908929218410646, + "learning_rate": 0.003, + "loss": 4.0898, + "step": 13439 + }, + { + "epoch": 0.1344, + "grad_norm": 0.705917103205681, + "learning_rate": 0.003, + "loss": 4.0973, + "step": 13440 + }, + { + "epoch": 0.13441, + "grad_norm": 0.6838396270610559, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 13441 + }, + { + "epoch": 0.13442, + "grad_norm": 0.7352015052366501, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 13442 + }, + { + "epoch": 0.13443, + "grad_norm": 0.9349659441431727, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 13443 + }, + { + "epoch": 0.13444, + "grad_norm": 1.25945865283335, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 13444 + }, + { + "epoch": 0.13445, + "grad_norm": 0.9413036656679339, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 13445 + }, + { + "epoch": 0.13446, + "grad_norm": 0.8179413839817955, + "learning_rate": 0.003, + "loss": 4.0868, + "step": 13446 + }, + { + "epoch": 0.13447, + "grad_norm": 0.9918442592703888, + "learning_rate": 0.003, + "loss": 4.114, + "step": 13447 + }, + { + "epoch": 0.13448, + "grad_norm": 1.2373065469861737, + "learning_rate": 0.003, + "loss": 4.0918, + "step": 13448 + }, + { + "epoch": 0.13449, + "grad_norm": 0.9584489654305866, + "learning_rate": 0.003, + "loss": 4.0858, + "step": 13449 + }, + { + "epoch": 0.1345, + "grad_norm": 1.1909459136786413, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 13450 + }, + { + "epoch": 0.13451, + "grad_norm": 0.9831683010495688, + "learning_rate": 0.003, + "loss": 4.075, + "step": 13451 + }, + { + "epoch": 0.13452, + "grad_norm": 0.919751350475958, + "learning_rate": 0.003, + "loss": 4.067, + "step": 13452 + }, + { + "epoch": 0.13453, + "grad_norm": 0.8944767240887079, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 13453 + }, + { + "epoch": 0.13454, + "grad_norm": 1.0205203683401989, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 13454 + }, + { + "epoch": 0.13455, + "grad_norm": 1.1845969475631215, + "learning_rate": 0.003, + "loss": 4.1048, + "step": 13455 + }, + { + "epoch": 0.13456, + "grad_norm": 0.7845340551858228, + "learning_rate": 0.003, + "loss": 4.0933, + "step": 13456 + }, + { + "epoch": 0.13457, + "grad_norm": 0.6924883534994194, + "learning_rate": 0.003, + "loss": 4.049, + "step": 13457 + }, + { + "epoch": 0.13458, + "grad_norm": 0.8044456668064189, + "learning_rate": 0.003, + "loss": 4.0995, + "step": 13458 + }, + { + "epoch": 0.13459, + "grad_norm": 0.882249813678147, + "learning_rate": 0.003, + "loss": 4.0978, + "step": 13459 + }, + { + "epoch": 0.1346, + "grad_norm": 0.7660785398447458, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 13460 + }, + { + "epoch": 0.13461, + "grad_norm": 0.7497250213468858, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 13461 + }, + { + "epoch": 0.13462, + "grad_norm": 0.9416272058405345, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 13462 + }, + { + "epoch": 0.13463, + "grad_norm": 1.1909351153459253, + "learning_rate": 0.003, + "loss": 4.0923, + "step": 13463 + }, + { + "epoch": 0.13464, + "grad_norm": 1.0552344021648177, + "learning_rate": 0.003, + "loss": 4.0931, + "step": 13464 + }, + { + "epoch": 0.13465, + "grad_norm": 0.8805616886318614, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 13465 + }, + { + "epoch": 0.13466, + "grad_norm": 0.9071092904952632, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 13466 + }, + { + "epoch": 0.13467, + "grad_norm": 0.981498872004427, + "learning_rate": 0.003, + "loss": 4.1004, + "step": 13467 + }, + { + "epoch": 0.13468, + "grad_norm": 1.1061070019176091, + "learning_rate": 0.003, + "loss": 4.0855, + "step": 13468 + }, + { + "epoch": 0.13469, + "grad_norm": 0.9114096684026349, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 13469 + }, + { + "epoch": 0.1347, + "grad_norm": 1.044914485044345, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 13470 + }, + { + "epoch": 0.13471, + "grad_norm": 1.11363121472758, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 13471 + }, + { + "epoch": 0.13472, + "grad_norm": 0.9572118768238856, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 13472 + }, + { + "epoch": 0.13473, + "grad_norm": 0.9671314677153261, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 13473 + }, + { + "epoch": 0.13474, + "grad_norm": 1.139525101252777, + "learning_rate": 0.003, + "loss": 4.0893, + "step": 13474 + }, + { + "epoch": 0.13475, + "grad_norm": 0.9721699359801134, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 13475 + }, + { + "epoch": 0.13476, + "grad_norm": 0.9383326552468131, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 13476 + }, + { + "epoch": 0.13477, + "grad_norm": 1.0931290867313288, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 13477 + }, + { + "epoch": 0.13478, + "grad_norm": 1.126490789440189, + "learning_rate": 0.003, + "loss": 4.0942, + "step": 13478 + }, + { + "epoch": 0.13479, + "grad_norm": 1.0316971169284692, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 13479 + }, + { + "epoch": 0.1348, + "grad_norm": 1.0532126938522322, + "learning_rate": 0.003, + "loss": 4.0979, + "step": 13480 + }, + { + "epoch": 0.13481, + "grad_norm": 0.9483779559587425, + "learning_rate": 0.003, + "loss": 4.0974, + "step": 13481 + }, + { + "epoch": 0.13482, + "grad_norm": 0.9013033974498498, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 13482 + }, + { + "epoch": 0.13483, + "grad_norm": 0.9435170158122912, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 13483 + }, + { + "epoch": 0.13484, + "grad_norm": 1.2109010243254403, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 13484 + }, + { + "epoch": 0.13485, + "grad_norm": 1.0514417119993889, + "learning_rate": 0.003, + "loss": 4.1021, + "step": 13485 + }, + { + "epoch": 0.13486, + "grad_norm": 1.0057246531087394, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 13486 + }, + { + "epoch": 0.13487, + "grad_norm": 1.1471260095240534, + "learning_rate": 0.003, + "loss": 4.073, + "step": 13487 + }, + { + "epoch": 0.13488, + "grad_norm": 1.009518901355293, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 13488 + }, + { + "epoch": 0.13489, + "grad_norm": 0.9649318916413948, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 13489 + }, + { + "epoch": 0.1349, + "grad_norm": 0.9041949841760637, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 13490 + }, + { + "epoch": 0.13491, + "grad_norm": 1.0466063920776612, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 13491 + }, + { + "epoch": 0.13492, + "grad_norm": 1.2736774275263576, + "learning_rate": 0.003, + "loss": 4.0989, + "step": 13492 + }, + { + "epoch": 0.13493, + "grad_norm": 0.8108026482981228, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 13493 + }, + { + "epoch": 0.13494, + "grad_norm": 0.7436894595678959, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 13494 + }, + { + "epoch": 0.13495, + "grad_norm": 0.7774525452175587, + "learning_rate": 0.003, + "loss": 4.1101, + "step": 13495 + }, + { + "epoch": 0.13496, + "grad_norm": 0.90715633897908, + "learning_rate": 0.003, + "loss": 4.0991, + "step": 13496 + }, + { + "epoch": 0.13497, + "grad_norm": 1.368198117701012, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 13497 + }, + { + "epoch": 0.13498, + "grad_norm": 0.9465797907617106, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 13498 + }, + { + "epoch": 0.13499, + "grad_norm": 0.8522559151082583, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 13499 + }, + { + "epoch": 0.135, + "grad_norm": 0.9030860229787252, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 13500 + }, + { + "epoch": 0.13501, + "grad_norm": 0.8348847972512554, + "learning_rate": 0.003, + "loss": 4.094, + "step": 13501 + }, + { + "epoch": 0.13502, + "grad_norm": 0.9660341487320527, + "learning_rate": 0.003, + "loss": 4.0881, + "step": 13502 + }, + { + "epoch": 0.13503, + "grad_norm": 1.1629037817247843, + "learning_rate": 0.003, + "loss": 4.073, + "step": 13503 + }, + { + "epoch": 0.13504, + "grad_norm": 1.0569291824920555, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 13504 + }, + { + "epoch": 0.13505, + "grad_norm": 0.9554270062835452, + "learning_rate": 0.003, + "loss": 4.0917, + "step": 13505 + }, + { + "epoch": 0.13506, + "grad_norm": 0.8983670965587544, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 13506 + }, + { + "epoch": 0.13507, + "grad_norm": 0.8880349328679263, + "learning_rate": 0.003, + "loss": 4.0923, + "step": 13507 + }, + { + "epoch": 0.13508, + "grad_norm": 0.8140918181261108, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 13508 + }, + { + "epoch": 0.13509, + "grad_norm": 0.7404239764798114, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 13509 + }, + { + "epoch": 0.1351, + "grad_norm": 0.7235077371144683, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 13510 + }, + { + "epoch": 0.13511, + "grad_norm": 0.8484639991439143, + "learning_rate": 0.003, + "loss": 4.073, + "step": 13511 + }, + { + "epoch": 0.13512, + "grad_norm": 1.0751067890615134, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 13512 + }, + { + "epoch": 0.13513, + "grad_norm": 1.1828451835377958, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 13513 + }, + { + "epoch": 0.13514, + "grad_norm": 1.0080567387639703, + "learning_rate": 0.003, + "loss": 4.0919, + "step": 13514 + }, + { + "epoch": 0.13515, + "grad_norm": 1.0223522698244532, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 13515 + }, + { + "epoch": 0.13516, + "grad_norm": 0.8124072913007042, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 13516 + }, + { + "epoch": 0.13517, + "grad_norm": 0.8246049566847677, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 13517 + }, + { + "epoch": 0.13518, + "grad_norm": 0.9279865162803025, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 13518 + }, + { + "epoch": 0.13519, + "grad_norm": 1.0790165091703794, + "learning_rate": 0.003, + "loss": 4.0913, + "step": 13519 + }, + { + "epoch": 0.1352, + "grad_norm": 1.1179321203702368, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 13520 + }, + { + "epoch": 0.13521, + "grad_norm": 0.9709573822441261, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 13521 + }, + { + "epoch": 0.13522, + "grad_norm": 0.9845407951473611, + "learning_rate": 0.003, + "loss": 4.0923, + "step": 13522 + }, + { + "epoch": 0.13523, + "grad_norm": 1.0083537037565933, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 13523 + }, + { + "epoch": 0.13524, + "grad_norm": 1.129512198168827, + "learning_rate": 0.003, + "loss": 4.0939, + "step": 13524 + }, + { + "epoch": 0.13525, + "grad_norm": 0.9021943826692351, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 13525 + }, + { + "epoch": 0.13526, + "grad_norm": 0.8989819054898855, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 13526 + }, + { + "epoch": 0.13527, + "grad_norm": 1.0152893508384229, + "learning_rate": 0.003, + "loss": 4.076, + "step": 13527 + }, + { + "epoch": 0.13528, + "grad_norm": 0.9485921128846078, + "learning_rate": 0.003, + "loss": 4.074, + "step": 13528 + }, + { + "epoch": 0.13529, + "grad_norm": 1.0925652778608455, + "learning_rate": 0.003, + "loss": 4.1139, + "step": 13529 + }, + { + "epoch": 0.1353, + "grad_norm": 1.1033399300532505, + "learning_rate": 0.003, + "loss": 4.1005, + "step": 13530 + }, + { + "epoch": 0.13531, + "grad_norm": 1.020634427871289, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 13531 + }, + { + "epoch": 0.13532, + "grad_norm": 1.1485154410785172, + "learning_rate": 0.003, + "loss": 4.101, + "step": 13532 + }, + { + "epoch": 0.13533, + "grad_norm": 0.9831601636428202, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 13533 + }, + { + "epoch": 0.13534, + "grad_norm": 1.2898717838424105, + "learning_rate": 0.003, + "loss": 4.1106, + "step": 13534 + }, + { + "epoch": 0.13535, + "grad_norm": 0.9319133850640476, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 13535 + }, + { + "epoch": 0.13536, + "grad_norm": 0.8246426969364398, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 13536 + }, + { + "epoch": 0.13537, + "grad_norm": 0.8412648299395936, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 13537 + }, + { + "epoch": 0.13538, + "grad_norm": 0.7477705562927869, + "learning_rate": 0.003, + "loss": 4.0922, + "step": 13538 + }, + { + "epoch": 0.13539, + "grad_norm": 0.7521452285844823, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 13539 + }, + { + "epoch": 0.1354, + "grad_norm": 0.8093846307896786, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 13540 + }, + { + "epoch": 0.13541, + "grad_norm": 0.9863536018733797, + "learning_rate": 0.003, + "loss": 4.1044, + "step": 13541 + }, + { + "epoch": 0.13542, + "grad_norm": 1.2226420116285923, + "learning_rate": 0.003, + "loss": 4.099, + "step": 13542 + }, + { + "epoch": 0.13543, + "grad_norm": 0.9608851041545622, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 13543 + }, + { + "epoch": 0.13544, + "grad_norm": 0.9423622515019298, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 13544 + }, + { + "epoch": 0.13545, + "grad_norm": 0.9642772848953194, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 13545 + }, + { + "epoch": 0.13546, + "grad_norm": 1.1213654030667557, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 13546 + }, + { + "epoch": 0.13547, + "grad_norm": 0.9058722975656578, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 13547 + }, + { + "epoch": 0.13548, + "grad_norm": 0.8148991420495703, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 13548 + }, + { + "epoch": 0.13549, + "grad_norm": 0.8258021860654478, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 13549 + }, + { + "epoch": 0.1355, + "grad_norm": 0.9231496618596597, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 13550 + }, + { + "epoch": 0.13551, + "grad_norm": 1.1939130391278587, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 13551 + }, + { + "epoch": 0.13552, + "grad_norm": 1.1044468764017839, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 13552 + }, + { + "epoch": 0.13553, + "grad_norm": 0.8275375866515994, + "learning_rate": 0.003, + "loss": 4.1086, + "step": 13553 + }, + { + "epoch": 0.13554, + "grad_norm": 0.8075126243869627, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 13554 + }, + { + "epoch": 0.13555, + "grad_norm": 0.8815250197228948, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 13555 + }, + { + "epoch": 0.13556, + "grad_norm": 0.9504417351249175, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 13556 + }, + { + "epoch": 0.13557, + "grad_norm": 0.9095744933545749, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 13557 + }, + { + "epoch": 0.13558, + "grad_norm": 1.0056985388783073, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 13558 + }, + { + "epoch": 0.13559, + "grad_norm": 1.1080001626881524, + "learning_rate": 0.003, + "loss": 4.0977, + "step": 13559 + }, + { + "epoch": 0.1356, + "grad_norm": 0.9758331560617178, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 13560 + }, + { + "epoch": 0.13561, + "grad_norm": 1.126569322550907, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 13561 + }, + { + "epoch": 0.13562, + "grad_norm": 1.3349649633316043, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 13562 + }, + { + "epoch": 0.13563, + "grad_norm": 0.8803142889043307, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 13563 + }, + { + "epoch": 0.13564, + "grad_norm": 0.8663478896901837, + "learning_rate": 0.003, + "loss": 4.0966, + "step": 13564 + }, + { + "epoch": 0.13565, + "grad_norm": 0.9265201013331548, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 13565 + }, + { + "epoch": 0.13566, + "grad_norm": 1.030849199978633, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 13566 + }, + { + "epoch": 0.13567, + "grad_norm": 0.8854278876742893, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 13567 + }, + { + "epoch": 0.13568, + "grad_norm": 1.0157029730469802, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 13568 + }, + { + "epoch": 0.13569, + "grad_norm": 1.133128728693916, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 13569 + }, + { + "epoch": 0.1357, + "grad_norm": 0.9981040685719773, + "learning_rate": 0.003, + "loss": 4.085, + "step": 13570 + }, + { + "epoch": 0.13571, + "grad_norm": 1.1575818997164589, + "learning_rate": 0.003, + "loss": 4.0884, + "step": 13571 + }, + { + "epoch": 0.13572, + "grad_norm": 0.9094788068782816, + "learning_rate": 0.003, + "loss": 4.0871, + "step": 13572 + }, + { + "epoch": 0.13573, + "grad_norm": 0.8302588310833722, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 13573 + }, + { + "epoch": 0.13574, + "grad_norm": 0.9214769034388204, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 13574 + }, + { + "epoch": 0.13575, + "grad_norm": 1.0637243648943842, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 13575 + }, + { + "epoch": 0.13576, + "grad_norm": 0.9611403613371343, + "learning_rate": 0.003, + "loss": 4.0856, + "step": 13576 + }, + { + "epoch": 0.13577, + "grad_norm": 0.9628613905766384, + "learning_rate": 0.003, + "loss": 4.0936, + "step": 13577 + }, + { + "epoch": 0.13578, + "grad_norm": 1.2703236335614123, + "learning_rate": 0.003, + "loss": 4.1032, + "step": 13578 + }, + { + "epoch": 0.13579, + "grad_norm": 1.010585696475506, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 13579 + }, + { + "epoch": 0.1358, + "grad_norm": 1.0812387358098323, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 13580 + }, + { + "epoch": 0.13581, + "grad_norm": 0.9571419337458652, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 13581 + }, + { + "epoch": 0.13582, + "grad_norm": 1.0478290565774575, + "learning_rate": 0.003, + "loss": 4.0985, + "step": 13582 + }, + { + "epoch": 0.13583, + "grad_norm": 1.0080171132102091, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 13583 + }, + { + "epoch": 0.13584, + "grad_norm": 1.1229461349474312, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 13584 + }, + { + "epoch": 0.13585, + "grad_norm": 0.8937053157383335, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 13585 + }, + { + "epoch": 0.13586, + "grad_norm": 0.8796283806050472, + "learning_rate": 0.003, + "loss": 4.1193, + "step": 13586 + }, + { + "epoch": 0.13587, + "grad_norm": 1.0766292315375208, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 13587 + }, + { + "epoch": 0.13588, + "grad_norm": 1.1846995669485805, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 13588 + }, + { + "epoch": 0.13589, + "grad_norm": 0.7902173933184293, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 13589 + }, + { + "epoch": 0.1359, + "grad_norm": 0.6925960220064842, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 13590 + }, + { + "epoch": 0.13591, + "grad_norm": 0.6589221691528248, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 13591 + }, + { + "epoch": 0.13592, + "grad_norm": 0.5642711151068701, + "learning_rate": 0.003, + "loss": 4.058, + "step": 13592 + }, + { + "epoch": 0.13593, + "grad_norm": 0.5740948450847282, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 13593 + }, + { + "epoch": 0.13594, + "grad_norm": 0.5938015150027022, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 13594 + }, + { + "epoch": 0.13595, + "grad_norm": 0.6790898492881097, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 13595 + }, + { + "epoch": 0.13596, + "grad_norm": 0.6530141869175367, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 13596 + }, + { + "epoch": 0.13597, + "grad_norm": 0.6484511541963425, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 13597 + }, + { + "epoch": 0.13598, + "grad_norm": 0.6947691268160227, + "learning_rate": 0.003, + "loss": 4.1051, + "step": 13598 + }, + { + "epoch": 0.13599, + "grad_norm": 0.8291178230166169, + "learning_rate": 0.003, + "loss": 4.088, + "step": 13599 + }, + { + "epoch": 0.136, + "grad_norm": 1.21269263365347, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 13600 + }, + { + "epoch": 0.13601, + "grad_norm": 1.0755779322941, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 13601 + }, + { + "epoch": 0.13602, + "grad_norm": 1.0513035231232533, + "learning_rate": 0.003, + "loss": 4.081, + "step": 13602 + }, + { + "epoch": 0.13603, + "grad_norm": 1.1959430007348033, + "learning_rate": 0.003, + "loss": 4.0881, + "step": 13603 + }, + { + "epoch": 0.13604, + "grad_norm": 0.8605390357184035, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 13604 + }, + { + "epoch": 0.13605, + "grad_norm": 0.7227260204118963, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 13605 + }, + { + "epoch": 0.13606, + "grad_norm": 0.8056097574724497, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 13606 + }, + { + "epoch": 0.13607, + "grad_norm": 1.0148776257512635, + "learning_rate": 0.003, + "loss": 4.1046, + "step": 13607 + }, + { + "epoch": 0.13608, + "grad_norm": 1.0992112774470033, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 13608 + }, + { + "epoch": 0.13609, + "grad_norm": 0.9917031454694268, + "learning_rate": 0.003, + "loss": 4.052, + "step": 13609 + }, + { + "epoch": 0.1361, + "grad_norm": 1.378262176256691, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 13610 + }, + { + "epoch": 0.13611, + "grad_norm": 0.8862647169143499, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 13611 + }, + { + "epoch": 0.13612, + "grad_norm": 0.7197693414654428, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 13612 + }, + { + "epoch": 0.13613, + "grad_norm": 0.7773939977828908, + "learning_rate": 0.003, + "loss": 4.073, + "step": 13613 + }, + { + "epoch": 0.13614, + "grad_norm": 0.9869637991975421, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 13614 + }, + { + "epoch": 0.13615, + "grad_norm": 1.163804447606934, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 13615 + }, + { + "epoch": 0.13616, + "grad_norm": 0.848015729591127, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 13616 + }, + { + "epoch": 0.13617, + "grad_norm": 0.8061572098320081, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 13617 + }, + { + "epoch": 0.13618, + "grad_norm": 0.9708438619549377, + "learning_rate": 0.003, + "loss": 4.098, + "step": 13618 + }, + { + "epoch": 0.13619, + "grad_norm": 1.2380625075951288, + "learning_rate": 0.003, + "loss": 4.0995, + "step": 13619 + }, + { + "epoch": 0.1362, + "grad_norm": 0.7950350471684355, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 13620 + }, + { + "epoch": 0.13621, + "grad_norm": 0.7453401587273695, + "learning_rate": 0.003, + "loss": 4.064, + "step": 13621 + }, + { + "epoch": 0.13622, + "grad_norm": 0.7830420891595247, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 13622 + }, + { + "epoch": 0.13623, + "grad_norm": 0.8942862861267185, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 13623 + }, + { + "epoch": 0.13624, + "grad_norm": 1.1046555061514043, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 13624 + }, + { + "epoch": 0.13625, + "grad_norm": 1.1606610218376399, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 13625 + }, + { + "epoch": 0.13626, + "grad_norm": 0.9924295139615238, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 13626 + }, + { + "epoch": 0.13627, + "grad_norm": 1.252379694808365, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 13627 + }, + { + "epoch": 0.13628, + "grad_norm": 0.7644243936339519, + "learning_rate": 0.003, + "loss": 4.079, + "step": 13628 + }, + { + "epoch": 0.13629, + "grad_norm": 0.6559970907604247, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 13629 + }, + { + "epoch": 0.1363, + "grad_norm": 0.8017806160493137, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 13630 + }, + { + "epoch": 0.13631, + "grad_norm": 1.0671301101637478, + "learning_rate": 0.003, + "loss": 4.065, + "step": 13631 + }, + { + "epoch": 0.13632, + "grad_norm": 1.1838990600625954, + "learning_rate": 0.003, + "loss": 4.1067, + "step": 13632 + }, + { + "epoch": 0.13633, + "grad_norm": 0.9038288971421625, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 13633 + }, + { + "epoch": 0.13634, + "grad_norm": 0.8839040399389617, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 13634 + }, + { + "epoch": 0.13635, + "grad_norm": 0.8946512236601442, + "learning_rate": 0.003, + "loss": 4.0912, + "step": 13635 + }, + { + "epoch": 0.13636, + "grad_norm": 0.9587942110764102, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 13636 + }, + { + "epoch": 0.13637, + "grad_norm": 0.9408280478634411, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 13637 + }, + { + "epoch": 0.13638, + "grad_norm": 0.9886822552257566, + "learning_rate": 0.003, + "loss": 4.1004, + "step": 13638 + }, + { + "epoch": 0.13639, + "grad_norm": 1.2293674236840542, + "learning_rate": 0.003, + "loss": 4.0923, + "step": 13639 + }, + { + "epoch": 0.1364, + "grad_norm": 0.9505147997416329, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 13640 + }, + { + "epoch": 0.13641, + "grad_norm": 0.999321721108603, + "learning_rate": 0.003, + "loss": 4.1063, + "step": 13641 + }, + { + "epoch": 0.13642, + "grad_norm": 1.192786577302841, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 13642 + }, + { + "epoch": 0.13643, + "grad_norm": 1.1668246929060908, + "learning_rate": 0.003, + "loss": 4.0969, + "step": 13643 + }, + { + "epoch": 0.13644, + "grad_norm": 1.2004615003052568, + "learning_rate": 0.003, + "loss": 4.059, + "step": 13644 + }, + { + "epoch": 0.13645, + "grad_norm": 0.8492438217762301, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 13645 + }, + { + "epoch": 0.13646, + "grad_norm": 0.682104770401569, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 13646 + }, + { + "epoch": 0.13647, + "grad_norm": 0.7610827745110301, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 13647 + }, + { + "epoch": 0.13648, + "grad_norm": 0.8941801093292976, + "learning_rate": 0.003, + "loss": 4.043, + "step": 13648 + }, + { + "epoch": 0.13649, + "grad_norm": 0.9819117889518406, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 13649 + }, + { + "epoch": 0.1365, + "grad_norm": 0.9594079222258514, + "learning_rate": 0.003, + "loss": 4.0854, + "step": 13650 + }, + { + "epoch": 0.13651, + "grad_norm": 1.0562580512817925, + "learning_rate": 0.003, + "loss": 4.0926, + "step": 13651 + }, + { + "epoch": 0.13652, + "grad_norm": 1.2125355274614926, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 13652 + }, + { + "epoch": 0.13653, + "grad_norm": 0.8964197042040143, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 13653 + }, + { + "epoch": 0.13654, + "grad_norm": 0.777737324210047, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 13654 + }, + { + "epoch": 0.13655, + "grad_norm": 0.7329112195224557, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 13655 + }, + { + "epoch": 0.13656, + "grad_norm": 0.7533468066028345, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 13656 + }, + { + "epoch": 0.13657, + "grad_norm": 0.9546724986937072, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 13657 + }, + { + "epoch": 0.13658, + "grad_norm": 1.0283740561585122, + "learning_rate": 0.003, + "loss": 4.1104, + "step": 13658 + }, + { + "epoch": 0.13659, + "grad_norm": 0.9178849265280306, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 13659 + }, + { + "epoch": 0.1366, + "grad_norm": 1.232649197897835, + "learning_rate": 0.003, + "loss": 4.1168, + "step": 13660 + }, + { + "epoch": 0.13661, + "grad_norm": 1.1385372522714916, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 13661 + }, + { + "epoch": 0.13662, + "grad_norm": 1.0090094950683128, + "learning_rate": 0.003, + "loss": 4.0915, + "step": 13662 + }, + { + "epoch": 0.13663, + "grad_norm": 1.1607918610626795, + "learning_rate": 0.003, + "loss": 4.0932, + "step": 13663 + }, + { + "epoch": 0.13664, + "grad_norm": 0.9395459510116684, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 13664 + }, + { + "epoch": 0.13665, + "grad_norm": 0.973866182474934, + "learning_rate": 0.003, + "loss": 4.0972, + "step": 13665 + }, + { + "epoch": 0.13666, + "grad_norm": 1.0907246962382176, + "learning_rate": 0.003, + "loss": 4.0955, + "step": 13666 + }, + { + "epoch": 0.13667, + "grad_norm": 1.03111390428082, + "learning_rate": 0.003, + "loss": 4.0919, + "step": 13667 + }, + { + "epoch": 0.13668, + "grad_norm": 1.1755424517938953, + "learning_rate": 0.003, + "loss": 4.0949, + "step": 13668 + }, + { + "epoch": 0.13669, + "grad_norm": 1.0002130660642194, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 13669 + }, + { + "epoch": 0.1367, + "grad_norm": 1.027615824689353, + "learning_rate": 0.003, + "loss": 4.0878, + "step": 13670 + }, + { + "epoch": 0.13671, + "grad_norm": 0.8101226686312777, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 13671 + }, + { + "epoch": 0.13672, + "grad_norm": 0.8000888494913125, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 13672 + }, + { + "epoch": 0.13673, + "grad_norm": 0.7840791293133723, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 13673 + }, + { + "epoch": 0.13674, + "grad_norm": 0.9354805374244475, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 13674 + }, + { + "epoch": 0.13675, + "grad_norm": 1.087339671126237, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 13675 + }, + { + "epoch": 0.13676, + "grad_norm": 1.0082945451543506, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 13676 + }, + { + "epoch": 0.13677, + "grad_norm": 1.0895604714191223, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 13677 + }, + { + "epoch": 0.13678, + "grad_norm": 0.8669270233066877, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 13678 + }, + { + "epoch": 0.13679, + "grad_norm": 0.8521374938785631, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 13679 + }, + { + "epoch": 0.1368, + "grad_norm": 0.8609769324367942, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 13680 + }, + { + "epoch": 0.13681, + "grad_norm": 1.0436893255432997, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 13681 + }, + { + "epoch": 0.13682, + "grad_norm": 1.2422715577778671, + "learning_rate": 0.003, + "loss": 4.1033, + "step": 13682 + }, + { + "epoch": 0.13683, + "grad_norm": 0.8857607482959572, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 13683 + }, + { + "epoch": 0.13684, + "grad_norm": 0.7944327830012379, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 13684 + }, + { + "epoch": 0.13685, + "grad_norm": 0.8514518206527368, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 13685 + }, + { + "epoch": 0.13686, + "grad_norm": 0.8949210911962465, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 13686 + }, + { + "epoch": 0.13687, + "grad_norm": 0.8907789499563097, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 13687 + }, + { + "epoch": 0.13688, + "grad_norm": 1.0320938589789124, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 13688 + }, + { + "epoch": 0.13689, + "grad_norm": 1.0840950039536872, + "learning_rate": 0.003, + "loss": 4.0996, + "step": 13689 + }, + { + "epoch": 0.1369, + "grad_norm": 0.9339772103054526, + "learning_rate": 0.003, + "loss": 4.0957, + "step": 13690 + }, + { + "epoch": 0.13691, + "grad_norm": 1.059542604588264, + "learning_rate": 0.003, + "loss": 4.0955, + "step": 13691 + }, + { + "epoch": 0.13692, + "grad_norm": 1.3309406242428878, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 13692 + }, + { + "epoch": 0.13693, + "grad_norm": 1.0859677753414343, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 13693 + }, + { + "epoch": 0.13694, + "grad_norm": 1.0089401785696337, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 13694 + }, + { + "epoch": 0.13695, + "grad_norm": 0.9918676725523683, + "learning_rate": 0.003, + "loss": 4.082, + "step": 13695 + }, + { + "epoch": 0.13696, + "grad_norm": 0.8838843728710856, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 13696 + }, + { + "epoch": 0.13697, + "grad_norm": 1.0035375195397744, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 13697 + }, + { + "epoch": 0.13698, + "grad_norm": 0.9999706193693773, + "learning_rate": 0.003, + "loss": 4.069, + "step": 13698 + }, + { + "epoch": 0.13699, + "grad_norm": 1.0341959255512199, + "learning_rate": 0.003, + "loss": 4.0933, + "step": 13699 + }, + { + "epoch": 0.137, + "grad_norm": 0.9285759529163252, + "learning_rate": 0.003, + "loss": 4.067, + "step": 13700 + }, + { + "epoch": 0.13701, + "grad_norm": 0.9826917523347732, + "learning_rate": 0.003, + "loss": 4.1085, + "step": 13701 + }, + { + "epoch": 0.13702, + "grad_norm": 1.071216697170529, + "learning_rate": 0.003, + "loss": 4.068, + "step": 13702 + }, + { + "epoch": 0.13703, + "grad_norm": 0.9684902986545548, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 13703 + }, + { + "epoch": 0.13704, + "grad_norm": 0.8657648532607887, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 13704 + }, + { + "epoch": 0.13705, + "grad_norm": 0.761177557587298, + "learning_rate": 0.003, + "loss": 4.08, + "step": 13705 + }, + { + "epoch": 0.13706, + "grad_norm": 0.8681681846415135, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 13706 + }, + { + "epoch": 0.13707, + "grad_norm": 0.8606171143587017, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 13707 + }, + { + "epoch": 0.13708, + "grad_norm": 0.9418939330526587, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 13708 + }, + { + "epoch": 0.13709, + "grad_norm": 1.2866046866449894, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 13709 + }, + { + "epoch": 0.1371, + "grad_norm": 0.8744904283668338, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 13710 + }, + { + "epoch": 0.13711, + "grad_norm": 0.9218635832357963, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 13711 + }, + { + "epoch": 0.13712, + "grad_norm": 0.9661569594800131, + "learning_rate": 0.003, + "loss": 4.055, + "step": 13712 + }, + { + "epoch": 0.13713, + "grad_norm": 1.104979268635213, + "learning_rate": 0.003, + "loss": 4.081, + "step": 13713 + }, + { + "epoch": 0.13714, + "grad_norm": 0.9920439956066915, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 13714 + }, + { + "epoch": 0.13715, + "grad_norm": 1.0436967087961082, + "learning_rate": 0.003, + "loss": 4.0874, + "step": 13715 + }, + { + "epoch": 0.13716, + "grad_norm": 0.9176523355345482, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 13716 + }, + { + "epoch": 0.13717, + "grad_norm": 0.9458885771889878, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 13717 + }, + { + "epoch": 0.13718, + "grad_norm": 1.0775826640405215, + "learning_rate": 0.003, + "loss": 4.0936, + "step": 13718 + }, + { + "epoch": 0.13719, + "grad_norm": 0.9461455874490248, + "learning_rate": 0.003, + "loss": 4.071, + "step": 13719 + }, + { + "epoch": 0.1372, + "grad_norm": 1.224772511108044, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 13720 + }, + { + "epoch": 0.13721, + "grad_norm": 1.074666842565267, + "learning_rate": 0.003, + "loss": 4.071, + "step": 13721 + }, + { + "epoch": 0.13722, + "grad_norm": 0.9373690970373129, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 13722 + }, + { + "epoch": 0.13723, + "grad_norm": 0.9348061363017073, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 13723 + }, + { + "epoch": 0.13724, + "grad_norm": 0.9389700970863364, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 13724 + }, + { + "epoch": 0.13725, + "grad_norm": 0.9892854075383015, + "learning_rate": 0.003, + "loss": 4.0996, + "step": 13725 + }, + { + "epoch": 0.13726, + "grad_norm": 0.9709066550360923, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 13726 + }, + { + "epoch": 0.13727, + "grad_norm": 1.0619083013039234, + "learning_rate": 0.003, + "loss": 4.0929, + "step": 13727 + }, + { + "epoch": 0.13728, + "grad_norm": 1.122958148800062, + "learning_rate": 0.003, + "loss": 4.102, + "step": 13728 + }, + { + "epoch": 0.13729, + "grad_norm": 0.9288054430854779, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 13729 + }, + { + "epoch": 0.1373, + "grad_norm": 1.047849457580226, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 13730 + }, + { + "epoch": 0.13731, + "grad_norm": 1.1467696425417673, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 13731 + }, + { + "epoch": 0.13732, + "grad_norm": 0.8597451642666474, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 13732 + }, + { + "epoch": 0.13733, + "grad_norm": 0.970555688109015, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 13733 + }, + { + "epoch": 0.13734, + "grad_norm": 1.1169845170663177, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 13734 + }, + { + "epoch": 0.13735, + "grad_norm": 0.904513670681962, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 13735 + }, + { + "epoch": 0.13736, + "grad_norm": 0.824037223895435, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 13736 + }, + { + "epoch": 0.13737, + "grad_norm": 0.8719749391828395, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 13737 + }, + { + "epoch": 0.13738, + "grad_norm": 0.8791017159954068, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 13738 + }, + { + "epoch": 0.13739, + "grad_norm": 1.0972238203019815, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 13739 + }, + { + "epoch": 0.1374, + "grad_norm": 1.0649232406600586, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 13740 + }, + { + "epoch": 0.13741, + "grad_norm": 1.192456475210352, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 13741 + }, + { + "epoch": 0.13742, + "grad_norm": 0.8286062352730182, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 13742 + }, + { + "epoch": 0.13743, + "grad_norm": 0.8793634566613651, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 13743 + }, + { + "epoch": 0.13744, + "grad_norm": 0.9125827472998448, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 13744 + }, + { + "epoch": 0.13745, + "grad_norm": 0.9204852071679613, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 13745 + }, + { + "epoch": 0.13746, + "grad_norm": 1.0359840514287635, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 13746 + }, + { + "epoch": 0.13747, + "grad_norm": 1.0240296303309406, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 13747 + }, + { + "epoch": 0.13748, + "grad_norm": 1.091688079443003, + "learning_rate": 0.003, + "loss": 4.1146, + "step": 13748 + }, + { + "epoch": 0.13749, + "grad_norm": 1.0789629032852017, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 13749 + }, + { + "epoch": 0.1375, + "grad_norm": 0.9088039647185775, + "learning_rate": 0.003, + "loss": 4.0906, + "step": 13750 + }, + { + "epoch": 0.13751, + "grad_norm": 0.9342988548470396, + "learning_rate": 0.003, + "loss": 4.1022, + "step": 13751 + }, + { + "epoch": 0.13752, + "grad_norm": 0.9149838236864942, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 13752 + }, + { + "epoch": 0.13753, + "grad_norm": 0.8043574627191314, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 13753 + }, + { + "epoch": 0.13754, + "grad_norm": 0.7353599980698797, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 13754 + }, + { + "epoch": 0.13755, + "grad_norm": 0.7758556195652425, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 13755 + }, + { + "epoch": 0.13756, + "grad_norm": 0.9679815218742019, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 13756 + }, + { + "epoch": 0.13757, + "grad_norm": 1.0910111969108989, + "learning_rate": 0.003, + "loss": 4.071, + "step": 13757 + }, + { + "epoch": 0.13758, + "grad_norm": 1.1012270277721994, + "learning_rate": 0.003, + "loss": 4.0868, + "step": 13758 + }, + { + "epoch": 0.13759, + "grad_norm": 1.0865732772430325, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 13759 + }, + { + "epoch": 0.1376, + "grad_norm": 1.101088695916349, + "learning_rate": 0.003, + "loss": 4.099, + "step": 13760 + }, + { + "epoch": 0.13761, + "grad_norm": 1.084850916242147, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 13761 + }, + { + "epoch": 0.13762, + "grad_norm": 0.8752994984855265, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 13762 + }, + { + "epoch": 0.13763, + "grad_norm": 0.8074786488856138, + "learning_rate": 0.003, + "loss": 4.1102, + "step": 13763 + }, + { + "epoch": 0.13764, + "grad_norm": 0.8164219883084545, + "learning_rate": 0.003, + "loss": 4.1018, + "step": 13764 + }, + { + "epoch": 0.13765, + "grad_norm": 0.8123354295969643, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 13765 + }, + { + "epoch": 0.13766, + "grad_norm": 0.7220713139637559, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 13766 + }, + { + "epoch": 0.13767, + "grad_norm": 0.7717034773681029, + "learning_rate": 0.003, + "loss": 4.056, + "step": 13767 + }, + { + "epoch": 0.13768, + "grad_norm": 0.7658771879719891, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 13768 + }, + { + "epoch": 0.13769, + "grad_norm": 0.7295026187413807, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 13769 + }, + { + "epoch": 0.1377, + "grad_norm": 0.812334856709315, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 13770 + }, + { + "epoch": 0.13771, + "grad_norm": 0.8976616850214401, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 13771 + }, + { + "epoch": 0.13772, + "grad_norm": 1.307944527078481, + "learning_rate": 0.003, + "loss": 4.0915, + "step": 13772 + }, + { + "epoch": 0.13773, + "grad_norm": 1.0630790920947895, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 13773 + }, + { + "epoch": 0.13774, + "grad_norm": 0.9821187792293821, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 13774 + }, + { + "epoch": 0.13775, + "grad_norm": 1.177010524417492, + "learning_rate": 0.003, + "loss": 4.1008, + "step": 13775 + }, + { + "epoch": 0.13776, + "grad_norm": 1.0454139898135226, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 13776 + }, + { + "epoch": 0.13777, + "grad_norm": 0.9981366136139526, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 13777 + }, + { + "epoch": 0.13778, + "grad_norm": 0.9854920699473189, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 13778 + }, + { + "epoch": 0.13779, + "grad_norm": 1.182138083255626, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 13779 + }, + { + "epoch": 0.1378, + "grad_norm": 1.0986187006360761, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 13780 + }, + { + "epoch": 0.13781, + "grad_norm": 0.9339597683460923, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 13781 + }, + { + "epoch": 0.13782, + "grad_norm": 0.9906508702867922, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 13782 + }, + { + "epoch": 0.13783, + "grad_norm": 0.9811335120051732, + "learning_rate": 0.003, + "loss": 4.1167, + "step": 13783 + }, + { + "epoch": 0.13784, + "grad_norm": 1.0038962994701235, + "learning_rate": 0.003, + "loss": 4.0893, + "step": 13784 + }, + { + "epoch": 0.13785, + "grad_norm": 1.19636908421218, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 13785 + }, + { + "epoch": 0.13786, + "grad_norm": 0.8777376992040019, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 13786 + }, + { + "epoch": 0.13787, + "grad_norm": 0.8162112108878308, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 13787 + }, + { + "epoch": 0.13788, + "grad_norm": 0.7044755269672827, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 13788 + }, + { + "epoch": 0.13789, + "grad_norm": 0.6785728982922626, + "learning_rate": 0.003, + "loss": 4.0923, + "step": 13789 + }, + { + "epoch": 0.1379, + "grad_norm": 0.6751432051104621, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 13790 + }, + { + "epoch": 0.13791, + "grad_norm": 0.6520302847403976, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 13791 + }, + { + "epoch": 0.13792, + "grad_norm": 0.6771508185766564, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 13792 + }, + { + "epoch": 0.13793, + "grad_norm": 0.6779003292927163, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 13793 + }, + { + "epoch": 0.13794, + "grad_norm": 0.8139222977953493, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 13794 + }, + { + "epoch": 0.13795, + "grad_norm": 1.1516018013544476, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 13795 + }, + { + "epoch": 0.13796, + "grad_norm": 1.1826870883354157, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 13796 + }, + { + "epoch": 0.13797, + "grad_norm": 0.8523779124886801, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 13797 + }, + { + "epoch": 0.13798, + "grad_norm": 0.8271937947746741, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 13798 + }, + { + "epoch": 0.13799, + "grad_norm": 0.8852314835909663, + "learning_rate": 0.003, + "loss": 4.062, + "step": 13799 + }, + { + "epoch": 0.138, + "grad_norm": 1.2523237567046503, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 13800 + }, + { + "epoch": 0.13801, + "grad_norm": 1.011836311133381, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 13801 + }, + { + "epoch": 0.13802, + "grad_norm": 1.0170283105449962, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 13802 + }, + { + "epoch": 0.13803, + "grad_norm": 1.0318708385359519, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 13803 + }, + { + "epoch": 0.13804, + "grad_norm": 1.0180344436396247, + "learning_rate": 0.003, + "loss": 4.1011, + "step": 13804 + }, + { + "epoch": 0.13805, + "grad_norm": 0.9489328227159523, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 13805 + }, + { + "epoch": 0.13806, + "grad_norm": 0.9560998905252485, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 13806 + }, + { + "epoch": 0.13807, + "grad_norm": 1.134995029847103, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 13807 + }, + { + "epoch": 0.13808, + "grad_norm": 0.9963484976857472, + "learning_rate": 0.003, + "loss": 4.0886, + "step": 13808 + }, + { + "epoch": 0.13809, + "grad_norm": 1.0394524960619864, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 13809 + }, + { + "epoch": 0.1381, + "grad_norm": 1.201219908964545, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 13810 + }, + { + "epoch": 0.13811, + "grad_norm": 0.9276778934492497, + "learning_rate": 0.003, + "loss": 4.0994, + "step": 13811 + }, + { + "epoch": 0.13812, + "grad_norm": 1.0046255730460418, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 13812 + }, + { + "epoch": 0.13813, + "grad_norm": 0.9269252941535688, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 13813 + }, + { + "epoch": 0.13814, + "grad_norm": 1.1257480276085072, + "learning_rate": 0.003, + "loss": 4.0985, + "step": 13814 + }, + { + "epoch": 0.13815, + "grad_norm": 1.1379199347957243, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 13815 + }, + { + "epoch": 0.13816, + "grad_norm": 1.0787006513488797, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 13816 + }, + { + "epoch": 0.13817, + "grad_norm": 0.9379546565282292, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 13817 + }, + { + "epoch": 0.13818, + "grad_norm": 0.9677615732020445, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 13818 + }, + { + "epoch": 0.13819, + "grad_norm": 0.9466718315429485, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 13819 + }, + { + "epoch": 0.1382, + "grad_norm": 0.8742912458007804, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 13820 + }, + { + "epoch": 0.13821, + "grad_norm": 0.8806775580646011, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 13821 + }, + { + "epoch": 0.13822, + "grad_norm": 1.005569143186965, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 13822 + }, + { + "epoch": 0.13823, + "grad_norm": 1.3764020049002155, + "learning_rate": 0.003, + "loss": 4.073, + "step": 13823 + }, + { + "epoch": 0.13824, + "grad_norm": 0.8746273754354214, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 13824 + }, + { + "epoch": 0.13825, + "grad_norm": 0.9132584941963134, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 13825 + }, + { + "epoch": 0.13826, + "grad_norm": 1.1529580409356441, + "learning_rate": 0.003, + "loss": 4.0891, + "step": 13826 + }, + { + "epoch": 0.13827, + "grad_norm": 1.0233588682814123, + "learning_rate": 0.003, + "loss": 4.0941, + "step": 13827 + }, + { + "epoch": 0.13828, + "grad_norm": 0.91270806928929, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 13828 + }, + { + "epoch": 0.13829, + "grad_norm": 0.9475174108957283, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 13829 + }, + { + "epoch": 0.1383, + "grad_norm": 1.3111237342711508, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 13830 + }, + { + "epoch": 0.13831, + "grad_norm": 0.9304888366515388, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 13831 + }, + { + "epoch": 0.13832, + "grad_norm": 0.8881317978100025, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 13832 + }, + { + "epoch": 0.13833, + "grad_norm": 1.0996851181766896, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 13833 + }, + { + "epoch": 0.13834, + "grad_norm": 1.0769795668370241, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 13834 + }, + { + "epoch": 0.13835, + "grad_norm": 0.956165054246714, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 13835 + }, + { + "epoch": 0.13836, + "grad_norm": 0.9265113925555668, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 13836 + }, + { + "epoch": 0.13837, + "grad_norm": 0.8378333180630521, + "learning_rate": 0.003, + "loss": 4.083, + "step": 13837 + }, + { + "epoch": 0.13838, + "grad_norm": 0.7390344825816991, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 13838 + }, + { + "epoch": 0.13839, + "grad_norm": 0.8043809599658622, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 13839 + }, + { + "epoch": 0.1384, + "grad_norm": 1.0214322885779643, + "learning_rate": 0.003, + "loss": 4.0857, + "step": 13840 + }, + { + "epoch": 0.13841, + "grad_norm": 1.1548768433285097, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 13841 + }, + { + "epoch": 0.13842, + "grad_norm": 0.8919595324924937, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 13842 + }, + { + "epoch": 0.13843, + "grad_norm": 1.0511937682726746, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 13843 + }, + { + "epoch": 0.13844, + "grad_norm": 1.3356408038758667, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 13844 + }, + { + "epoch": 0.13845, + "grad_norm": 0.7968974284511724, + "learning_rate": 0.003, + "loss": 4.081, + "step": 13845 + }, + { + "epoch": 0.13846, + "grad_norm": 0.7133760223690724, + "learning_rate": 0.003, + "loss": 4.0929, + "step": 13846 + }, + { + "epoch": 0.13847, + "grad_norm": 0.7500581498876953, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 13847 + }, + { + "epoch": 0.13848, + "grad_norm": 1.0075442475073457, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 13848 + }, + { + "epoch": 0.13849, + "grad_norm": 1.2239935973877418, + "learning_rate": 0.003, + "loss": 4.083, + "step": 13849 + }, + { + "epoch": 0.1385, + "grad_norm": 0.8528571440147477, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 13850 + }, + { + "epoch": 0.13851, + "grad_norm": 0.9300359619972718, + "learning_rate": 0.003, + "loss": 4.059, + "step": 13851 + }, + { + "epoch": 0.13852, + "grad_norm": 0.9300047900455807, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 13852 + }, + { + "epoch": 0.13853, + "grad_norm": 0.9497177405217342, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 13853 + }, + { + "epoch": 0.13854, + "grad_norm": 0.9699847484464827, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 13854 + }, + { + "epoch": 0.13855, + "grad_norm": 1.1076800982742767, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 13855 + }, + { + "epoch": 0.13856, + "grad_norm": 1.1422078183668556, + "learning_rate": 0.003, + "loss": 4.102, + "step": 13856 + }, + { + "epoch": 0.13857, + "grad_norm": 0.8705627065222977, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 13857 + }, + { + "epoch": 0.13858, + "grad_norm": 0.7908893474796466, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 13858 + }, + { + "epoch": 0.13859, + "grad_norm": 0.7643695980156917, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 13859 + }, + { + "epoch": 0.1386, + "grad_norm": 1.104379925206656, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 13860 + }, + { + "epoch": 0.13861, + "grad_norm": 1.303501704601149, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 13861 + }, + { + "epoch": 0.13862, + "grad_norm": 0.7448944443356489, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 13862 + }, + { + "epoch": 0.13863, + "grad_norm": 0.8177099409196833, + "learning_rate": 0.003, + "loss": 4.1122, + "step": 13863 + }, + { + "epoch": 0.13864, + "grad_norm": 0.8989081348225098, + "learning_rate": 0.003, + "loss": 4.0983, + "step": 13864 + }, + { + "epoch": 0.13865, + "grad_norm": 1.0615834820411814, + "learning_rate": 0.003, + "loss": 4.0956, + "step": 13865 + }, + { + "epoch": 0.13866, + "grad_norm": 1.0342939298421876, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 13866 + }, + { + "epoch": 0.13867, + "grad_norm": 0.9919681118469447, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 13867 + }, + { + "epoch": 0.13868, + "grad_norm": 1.0955521495185805, + "learning_rate": 0.003, + "loss": 4.0908, + "step": 13868 + }, + { + "epoch": 0.13869, + "grad_norm": 1.0522991643633715, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 13869 + }, + { + "epoch": 0.1387, + "grad_norm": 0.910147200240396, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 13870 + }, + { + "epoch": 0.13871, + "grad_norm": 1.04943933796681, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 13871 + }, + { + "epoch": 0.13872, + "grad_norm": 1.3017650728611816, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 13872 + }, + { + "epoch": 0.13873, + "grad_norm": 0.9061022240277243, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 13873 + }, + { + "epoch": 0.13874, + "grad_norm": 0.8414154393804512, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 13874 + }, + { + "epoch": 0.13875, + "grad_norm": 0.8454869674885852, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 13875 + }, + { + "epoch": 0.13876, + "grad_norm": 1.0261250073559691, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 13876 + }, + { + "epoch": 0.13877, + "grad_norm": 1.2183336487362644, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 13877 + }, + { + "epoch": 0.13878, + "grad_norm": 0.9579376846504901, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 13878 + }, + { + "epoch": 0.13879, + "grad_norm": 0.9362561428380821, + "learning_rate": 0.003, + "loss": 4.099, + "step": 13879 + }, + { + "epoch": 0.1388, + "grad_norm": 1.133594718617753, + "learning_rate": 0.003, + "loss": 4.0937, + "step": 13880 + }, + { + "epoch": 0.13881, + "grad_norm": 0.8810449008331018, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 13881 + }, + { + "epoch": 0.13882, + "grad_norm": 0.9238196832915456, + "learning_rate": 0.003, + "loss": 4.073, + "step": 13882 + }, + { + "epoch": 0.13883, + "grad_norm": 0.8673099253294131, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 13883 + }, + { + "epoch": 0.13884, + "grad_norm": 0.9508802396679225, + "learning_rate": 0.003, + "loss": 4.0989, + "step": 13884 + }, + { + "epoch": 0.13885, + "grad_norm": 1.0317987047647532, + "learning_rate": 0.003, + "loss": 4.1199, + "step": 13885 + }, + { + "epoch": 0.13886, + "grad_norm": 0.9944194508807207, + "learning_rate": 0.003, + "loss": 4.053, + "step": 13886 + }, + { + "epoch": 0.13887, + "grad_norm": 1.0961310806773963, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 13887 + }, + { + "epoch": 0.13888, + "grad_norm": 0.927758565973871, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 13888 + }, + { + "epoch": 0.13889, + "grad_norm": 0.9903881451333022, + "learning_rate": 0.003, + "loss": 4.0902, + "step": 13889 + }, + { + "epoch": 0.1389, + "grad_norm": 1.2474983455684536, + "learning_rate": 0.003, + "loss": 4.0934, + "step": 13890 + }, + { + "epoch": 0.13891, + "grad_norm": 1.0722005967825978, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 13891 + }, + { + "epoch": 0.13892, + "grad_norm": 1.0511081867851924, + "learning_rate": 0.003, + "loss": 4.0922, + "step": 13892 + }, + { + "epoch": 0.13893, + "grad_norm": 1.0069571939486044, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 13893 + }, + { + "epoch": 0.13894, + "grad_norm": 1.0259016310681863, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 13894 + }, + { + "epoch": 0.13895, + "grad_norm": 0.984678912415564, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 13895 + }, + { + "epoch": 0.13896, + "grad_norm": 0.9115222475944047, + "learning_rate": 0.003, + "loss": 4.094, + "step": 13896 + }, + { + "epoch": 0.13897, + "grad_norm": 0.8862894294174112, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 13897 + }, + { + "epoch": 0.13898, + "grad_norm": 0.8602826529427388, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 13898 + }, + { + "epoch": 0.13899, + "grad_norm": 0.9105980050996468, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 13899 + }, + { + "epoch": 0.139, + "grad_norm": 1.063161496060695, + "learning_rate": 0.003, + "loss": 4.1103, + "step": 13900 + }, + { + "epoch": 0.13901, + "grad_norm": 0.9167703187643819, + "learning_rate": 0.003, + "loss": 4.0912, + "step": 13901 + }, + { + "epoch": 0.13902, + "grad_norm": 1.0368851935144288, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 13902 + }, + { + "epoch": 0.13903, + "grad_norm": 1.153762789989774, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 13903 + }, + { + "epoch": 0.13904, + "grad_norm": 1.0305641425839815, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 13904 + }, + { + "epoch": 0.13905, + "grad_norm": 0.9988723627520049, + "learning_rate": 0.003, + "loss": 4.0965, + "step": 13905 + }, + { + "epoch": 0.13906, + "grad_norm": 0.9888250844590237, + "learning_rate": 0.003, + "loss": 4.04, + "step": 13906 + }, + { + "epoch": 0.13907, + "grad_norm": 1.0709028712171553, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 13907 + }, + { + "epoch": 0.13908, + "grad_norm": 1.0911956551560584, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 13908 + }, + { + "epoch": 0.13909, + "grad_norm": 0.9737518349976756, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 13909 + }, + { + "epoch": 0.1391, + "grad_norm": 0.9497090768773532, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 13910 + }, + { + "epoch": 0.13911, + "grad_norm": 0.9531782972039226, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 13911 + }, + { + "epoch": 0.13912, + "grad_norm": 1.0203697172099597, + "learning_rate": 0.003, + "loss": 4.0887, + "step": 13912 + }, + { + "epoch": 0.13913, + "grad_norm": 0.9447436857169799, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 13913 + }, + { + "epoch": 0.13914, + "grad_norm": 0.8627013337532834, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 13914 + }, + { + "epoch": 0.13915, + "grad_norm": 0.7569187855500866, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 13915 + }, + { + "epoch": 0.13916, + "grad_norm": 0.8018446076822271, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 13916 + }, + { + "epoch": 0.13917, + "grad_norm": 0.7773172237161118, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 13917 + }, + { + "epoch": 0.13918, + "grad_norm": 0.8267373792271568, + "learning_rate": 0.003, + "loss": 4.0953, + "step": 13918 + }, + { + "epoch": 0.13919, + "grad_norm": 0.7978953054271283, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 13919 + }, + { + "epoch": 0.1392, + "grad_norm": 0.743389138011358, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 13920 + }, + { + "epoch": 0.13921, + "grad_norm": 0.7808474497963047, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 13921 + }, + { + "epoch": 0.13922, + "grad_norm": 0.8928355881927448, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 13922 + }, + { + "epoch": 0.13923, + "grad_norm": 0.921161527031958, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 13923 + }, + { + "epoch": 0.13924, + "grad_norm": 0.8164167563531662, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 13924 + }, + { + "epoch": 0.13925, + "grad_norm": 0.7719401035360982, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 13925 + }, + { + "epoch": 0.13926, + "grad_norm": 0.881248602068898, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 13926 + }, + { + "epoch": 0.13927, + "grad_norm": 1.3933016847674236, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 13927 + }, + { + "epoch": 0.13928, + "grad_norm": 0.8165123156113304, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 13928 + }, + { + "epoch": 0.13929, + "grad_norm": 0.7456281001951336, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 13929 + }, + { + "epoch": 0.1393, + "grad_norm": 0.8615404098171816, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 13930 + }, + { + "epoch": 0.13931, + "grad_norm": 1.0222644810090813, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 13931 + }, + { + "epoch": 0.13932, + "grad_norm": 1.137297878677271, + "learning_rate": 0.003, + "loss": 4.036, + "step": 13932 + }, + { + "epoch": 0.13933, + "grad_norm": 1.202580326938281, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 13933 + }, + { + "epoch": 0.13934, + "grad_norm": 1.0005551670842654, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 13934 + }, + { + "epoch": 0.13935, + "grad_norm": 1.0144155392703311, + "learning_rate": 0.003, + "loss": 4.075, + "step": 13935 + }, + { + "epoch": 0.13936, + "grad_norm": 0.9680942995358661, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 13936 + }, + { + "epoch": 0.13937, + "grad_norm": 1.0034525695147511, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 13937 + }, + { + "epoch": 0.13938, + "grad_norm": 1.151781293262244, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 13938 + }, + { + "epoch": 0.13939, + "grad_norm": 0.8314698655396626, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 13939 + }, + { + "epoch": 0.1394, + "grad_norm": 0.9043134972041872, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 13940 + }, + { + "epoch": 0.13941, + "grad_norm": 1.154771200086865, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 13941 + }, + { + "epoch": 0.13942, + "grad_norm": 1.1277376072251033, + "learning_rate": 0.003, + "loss": 4.089, + "step": 13942 + }, + { + "epoch": 0.13943, + "grad_norm": 1.070041030051616, + "learning_rate": 0.003, + "loss": 4.125, + "step": 13943 + }, + { + "epoch": 0.13944, + "grad_norm": 1.082594871307821, + "learning_rate": 0.003, + "loss": 4.0986, + "step": 13944 + }, + { + "epoch": 0.13945, + "grad_norm": 0.9886771824102993, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 13945 + }, + { + "epoch": 0.13946, + "grad_norm": 1.2470128988326026, + "learning_rate": 0.003, + "loss": 4.1008, + "step": 13946 + }, + { + "epoch": 0.13947, + "grad_norm": 0.8895915465442784, + "learning_rate": 0.003, + "loss": 4.0913, + "step": 13947 + }, + { + "epoch": 0.13948, + "grad_norm": 0.9816092901135318, + "learning_rate": 0.003, + "loss": 4.0906, + "step": 13948 + }, + { + "epoch": 0.13949, + "grad_norm": 1.1018696868440705, + "learning_rate": 0.003, + "loss": 4.1024, + "step": 13949 + }, + { + "epoch": 0.1395, + "grad_norm": 0.9111393780880268, + "learning_rate": 0.003, + "loss": 4.0998, + "step": 13950 + }, + { + "epoch": 0.13951, + "grad_norm": 1.0865598086680883, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 13951 + }, + { + "epoch": 0.13952, + "grad_norm": 1.0755511888711977, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 13952 + }, + { + "epoch": 0.13953, + "grad_norm": 0.9911857465908889, + "learning_rate": 0.003, + "loss": 4.075, + "step": 13953 + }, + { + "epoch": 0.13954, + "grad_norm": 0.873232608028177, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 13954 + }, + { + "epoch": 0.13955, + "grad_norm": 0.7832048973553214, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 13955 + }, + { + "epoch": 0.13956, + "grad_norm": 0.8655839973306254, + "learning_rate": 0.003, + "loss": 4.0871, + "step": 13956 + }, + { + "epoch": 0.13957, + "grad_norm": 1.0160986244899548, + "learning_rate": 0.003, + "loss": 4.0962, + "step": 13957 + }, + { + "epoch": 0.13958, + "grad_norm": 1.0581098478725928, + "learning_rate": 0.003, + "loss": 4.0856, + "step": 13958 + }, + { + "epoch": 0.13959, + "grad_norm": 0.8564470544930965, + "learning_rate": 0.003, + "loss": 4.075, + "step": 13959 + }, + { + "epoch": 0.1396, + "grad_norm": 0.869767241063261, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 13960 + }, + { + "epoch": 0.13961, + "grad_norm": 0.9410726976607848, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 13961 + }, + { + "epoch": 0.13962, + "grad_norm": 0.8885373750544098, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 13962 + }, + { + "epoch": 0.13963, + "grad_norm": 0.940869496008005, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 13963 + }, + { + "epoch": 0.13964, + "grad_norm": 1.0193150843426817, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 13964 + }, + { + "epoch": 0.13965, + "grad_norm": 1.0113614456938034, + "learning_rate": 0.003, + "loss": 4.042, + "step": 13965 + }, + { + "epoch": 0.13966, + "grad_norm": 0.9926091108494736, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 13966 + }, + { + "epoch": 0.13967, + "grad_norm": 1.0551820560070766, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 13967 + }, + { + "epoch": 0.13968, + "grad_norm": 0.9744481295414718, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 13968 + }, + { + "epoch": 0.13969, + "grad_norm": 1.0953522367850683, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 13969 + }, + { + "epoch": 0.1397, + "grad_norm": 0.9531980614432011, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 13970 + }, + { + "epoch": 0.13971, + "grad_norm": 1.1050595738245683, + "learning_rate": 0.003, + "loss": 4.096, + "step": 13971 + }, + { + "epoch": 0.13972, + "grad_norm": 0.9760275188656535, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 13972 + }, + { + "epoch": 0.13973, + "grad_norm": 1.0998007144737876, + "learning_rate": 0.003, + "loss": 4.096, + "step": 13973 + }, + { + "epoch": 0.13974, + "grad_norm": 1.0116032790322933, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 13974 + }, + { + "epoch": 0.13975, + "grad_norm": 1.0452738584990722, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 13975 + }, + { + "epoch": 0.13976, + "grad_norm": 0.9220618249769322, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 13976 + }, + { + "epoch": 0.13977, + "grad_norm": 0.7452491690918652, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 13977 + }, + { + "epoch": 0.13978, + "grad_norm": 0.7437695380520812, + "learning_rate": 0.003, + "loss": 4.1263, + "step": 13978 + }, + { + "epoch": 0.13979, + "grad_norm": 0.763591710885012, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 13979 + }, + { + "epoch": 0.1398, + "grad_norm": 1.0244131193342163, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 13980 + }, + { + "epoch": 0.13981, + "grad_norm": 1.477408093407531, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 13981 + }, + { + "epoch": 0.13982, + "grad_norm": 0.780971610661029, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 13982 + }, + { + "epoch": 0.13983, + "grad_norm": 0.7956167958250066, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 13983 + }, + { + "epoch": 0.13984, + "grad_norm": 0.7091612576597078, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 13984 + }, + { + "epoch": 0.13985, + "grad_norm": 0.7129804856494933, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 13985 + }, + { + "epoch": 0.13986, + "grad_norm": 0.7167492022449294, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 13986 + }, + { + "epoch": 0.13987, + "grad_norm": 0.7600845370363315, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 13987 + }, + { + "epoch": 0.13988, + "grad_norm": 0.9485759099200272, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 13988 + }, + { + "epoch": 0.13989, + "grad_norm": 1.339778827782584, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 13989 + }, + { + "epoch": 0.1399, + "grad_norm": 0.9130200179827276, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 13990 + }, + { + "epoch": 0.13991, + "grad_norm": 1.0447331274298033, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 13991 + }, + { + "epoch": 0.13992, + "grad_norm": 0.9415999710350706, + "learning_rate": 0.003, + "loss": 4.076, + "step": 13992 + }, + { + "epoch": 0.13993, + "grad_norm": 1.00998492033337, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 13993 + }, + { + "epoch": 0.13994, + "grad_norm": 1.1590712185888212, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 13994 + }, + { + "epoch": 0.13995, + "grad_norm": 0.9114645659842159, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 13995 + }, + { + "epoch": 0.13996, + "grad_norm": 0.8088978130781801, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 13996 + }, + { + "epoch": 0.13997, + "grad_norm": 1.0168535318985774, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 13997 + }, + { + "epoch": 0.13998, + "grad_norm": 1.1350750717117926, + "learning_rate": 0.003, + "loss": 4.088, + "step": 13998 + }, + { + "epoch": 0.13999, + "grad_norm": 0.9767113338568917, + "learning_rate": 0.003, + "loss": 4.0878, + "step": 13999 + }, + { + "epoch": 0.14, + "grad_norm": 0.9781703017414717, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 14000 + }, + { + "epoch": 0.14001, + "grad_norm": 1.0568179300543583, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 14001 + }, + { + "epoch": 0.14002, + "grad_norm": 1.2077209917170695, + "learning_rate": 0.003, + "loss": 4.0958, + "step": 14002 + }, + { + "epoch": 0.14003, + "grad_norm": 0.9942763057336801, + "learning_rate": 0.003, + "loss": 4.0941, + "step": 14003 + }, + { + "epoch": 0.14004, + "grad_norm": 1.059243760045733, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 14004 + }, + { + "epoch": 0.14005, + "grad_norm": 0.8579537113622177, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 14005 + }, + { + "epoch": 0.14006, + "grad_norm": 0.853032341784671, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 14006 + }, + { + "epoch": 0.14007, + "grad_norm": 0.9274499568047793, + "learning_rate": 0.003, + "loss": 4.09, + "step": 14007 + }, + { + "epoch": 0.14008, + "grad_norm": 1.0609249707495503, + "learning_rate": 0.003, + "loss": 4.092, + "step": 14008 + }, + { + "epoch": 0.14009, + "grad_norm": 1.1811955366267206, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 14009 + }, + { + "epoch": 0.1401, + "grad_norm": 0.9634552516246274, + "learning_rate": 0.003, + "loss": 4.0878, + "step": 14010 + }, + { + "epoch": 0.14011, + "grad_norm": 0.9559479705068828, + "learning_rate": 0.003, + "loss": 4.054, + "step": 14011 + }, + { + "epoch": 0.14012, + "grad_norm": 0.9042455812738966, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 14012 + }, + { + "epoch": 0.14013, + "grad_norm": 0.9409060407776157, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 14013 + }, + { + "epoch": 0.14014, + "grad_norm": 1.0210014546015003, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 14014 + }, + { + "epoch": 0.14015, + "grad_norm": 1.0332329567802536, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 14015 + }, + { + "epoch": 0.14016, + "grad_norm": 0.9751483551377307, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 14016 + }, + { + "epoch": 0.14017, + "grad_norm": 1.093916255913977, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 14017 + }, + { + "epoch": 0.14018, + "grad_norm": 1.051842332087759, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 14018 + }, + { + "epoch": 0.14019, + "grad_norm": 1.0138239987034734, + "learning_rate": 0.003, + "loss": 4.089, + "step": 14019 + }, + { + "epoch": 0.1402, + "grad_norm": 0.918277140613066, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 14020 + }, + { + "epoch": 0.14021, + "grad_norm": 1.005969349886633, + "learning_rate": 0.003, + "loss": 4.0899, + "step": 14021 + }, + { + "epoch": 0.14022, + "grad_norm": 1.0542071309762444, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 14022 + }, + { + "epoch": 0.14023, + "grad_norm": 1.0192821015602123, + "learning_rate": 0.003, + "loss": 4.067, + "step": 14023 + }, + { + "epoch": 0.14024, + "grad_norm": 1.1760741718076926, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 14024 + }, + { + "epoch": 0.14025, + "grad_norm": 0.8844804459954531, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 14025 + }, + { + "epoch": 0.14026, + "grad_norm": 0.8682171401948858, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 14026 + }, + { + "epoch": 0.14027, + "grad_norm": 1.0418908891617698, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 14027 + }, + { + "epoch": 0.14028, + "grad_norm": 1.0104986854872506, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 14028 + }, + { + "epoch": 0.14029, + "grad_norm": 1.009663765738693, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 14029 + }, + { + "epoch": 0.1403, + "grad_norm": 1.0342710862445936, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 14030 + }, + { + "epoch": 0.14031, + "grad_norm": 1.0585909051221591, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 14031 + }, + { + "epoch": 0.14032, + "grad_norm": 1.0147588237807095, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 14032 + }, + { + "epoch": 0.14033, + "grad_norm": 1.0142594224253378, + "learning_rate": 0.003, + "loss": 4.0988, + "step": 14033 + }, + { + "epoch": 0.14034, + "grad_norm": 0.9604920282364194, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 14034 + }, + { + "epoch": 0.14035, + "grad_norm": 0.9918763238966519, + "learning_rate": 0.003, + "loss": 4.1049, + "step": 14035 + }, + { + "epoch": 0.14036, + "grad_norm": 1.0443627010801018, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 14036 + }, + { + "epoch": 0.14037, + "grad_norm": 0.9344303480953491, + "learning_rate": 0.003, + "loss": 4.1059, + "step": 14037 + }, + { + "epoch": 0.14038, + "grad_norm": 0.8840896389361425, + "learning_rate": 0.003, + "loss": 4.1115, + "step": 14038 + }, + { + "epoch": 0.14039, + "grad_norm": 0.9097590728184456, + "learning_rate": 0.003, + "loss": 4.099, + "step": 14039 + }, + { + "epoch": 0.1404, + "grad_norm": 0.8839354338990897, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 14040 + }, + { + "epoch": 0.14041, + "grad_norm": 0.895912199081318, + "learning_rate": 0.003, + "loss": 4.072, + "step": 14041 + }, + { + "epoch": 0.14042, + "grad_norm": 1.0368905320215813, + "learning_rate": 0.003, + "loss": 4.0872, + "step": 14042 + }, + { + "epoch": 0.14043, + "grad_norm": 1.2416892158755488, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 14043 + }, + { + "epoch": 0.14044, + "grad_norm": 1.0655308391782417, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 14044 + }, + { + "epoch": 0.14045, + "grad_norm": 1.014649858684877, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 14045 + }, + { + "epoch": 0.14046, + "grad_norm": 1.0128978428419437, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 14046 + }, + { + "epoch": 0.14047, + "grad_norm": 1.0148729177594262, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 14047 + }, + { + "epoch": 0.14048, + "grad_norm": 0.9672100844245592, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 14048 + }, + { + "epoch": 0.14049, + "grad_norm": 0.9143516306228309, + "learning_rate": 0.003, + "loss": 4.071, + "step": 14049 + }, + { + "epoch": 0.1405, + "grad_norm": 0.8702659427331704, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 14050 + }, + { + "epoch": 0.14051, + "grad_norm": 0.7766265932479794, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 14051 + }, + { + "epoch": 0.14052, + "grad_norm": 1.0531189285265372, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 14052 + }, + { + "epoch": 0.14053, + "grad_norm": 1.3386732178897682, + "learning_rate": 0.003, + "loss": 4.1037, + "step": 14053 + }, + { + "epoch": 0.14054, + "grad_norm": 0.8544466089220816, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 14054 + }, + { + "epoch": 0.14055, + "grad_norm": 0.8120859711850591, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 14055 + }, + { + "epoch": 0.14056, + "grad_norm": 0.899749496190138, + "learning_rate": 0.003, + "loss": 4.0908, + "step": 14056 + }, + { + "epoch": 0.14057, + "grad_norm": 1.0161334996094638, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 14057 + }, + { + "epoch": 0.14058, + "grad_norm": 1.118947755486378, + "learning_rate": 0.003, + "loss": 4.0974, + "step": 14058 + }, + { + "epoch": 0.14059, + "grad_norm": 0.854820269440219, + "learning_rate": 0.003, + "loss": 4.0881, + "step": 14059 + }, + { + "epoch": 0.1406, + "grad_norm": 0.9037965771039809, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 14060 + }, + { + "epoch": 0.14061, + "grad_norm": 0.9786544719277056, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 14061 + }, + { + "epoch": 0.14062, + "grad_norm": 1.0317444823058926, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 14062 + }, + { + "epoch": 0.14063, + "grad_norm": 1.1088249376897832, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 14063 + }, + { + "epoch": 0.14064, + "grad_norm": 1.0871762850407984, + "learning_rate": 0.003, + "loss": 4.0935, + "step": 14064 + }, + { + "epoch": 0.14065, + "grad_norm": 1.0549012767881405, + "learning_rate": 0.003, + "loss": 4.0878, + "step": 14065 + }, + { + "epoch": 0.14066, + "grad_norm": 0.9844743295681644, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 14066 + }, + { + "epoch": 0.14067, + "grad_norm": 1.023510733809489, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 14067 + }, + { + "epoch": 0.14068, + "grad_norm": 1.0371527376738436, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 14068 + }, + { + "epoch": 0.14069, + "grad_norm": 0.945386530896466, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 14069 + }, + { + "epoch": 0.1407, + "grad_norm": 0.9778551262289161, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 14070 + }, + { + "epoch": 0.14071, + "grad_norm": 0.9993163944879095, + "learning_rate": 0.003, + "loss": 4.0873, + "step": 14071 + }, + { + "epoch": 0.14072, + "grad_norm": 0.9609465301613052, + "learning_rate": 0.003, + "loss": 4.1002, + "step": 14072 + }, + { + "epoch": 0.14073, + "grad_norm": 0.9242909672794081, + "learning_rate": 0.003, + "loss": 4.0916, + "step": 14073 + }, + { + "epoch": 0.14074, + "grad_norm": 0.9025115065166386, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 14074 + }, + { + "epoch": 0.14075, + "grad_norm": 0.851359565045131, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 14075 + }, + { + "epoch": 0.14076, + "grad_norm": 0.7840762543077655, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 14076 + }, + { + "epoch": 0.14077, + "grad_norm": 0.8222470998492466, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 14077 + }, + { + "epoch": 0.14078, + "grad_norm": 0.7027830178386816, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 14078 + }, + { + "epoch": 0.14079, + "grad_norm": 0.6595895324432595, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 14079 + }, + { + "epoch": 0.1408, + "grad_norm": 0.6868235440195654, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 14080 + }, + { + "epoch": 0.14081, + "grad_norm": 0.7181145639340877, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 14081 + }, + { + "epoch": 0.14082, + "grad_norm": 0.7970238517162975, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 14082 + }, + { + "epoch": 0.14083, + "grad_norm": 1.0184599379546493, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 14083 + }, + { + "epoch": 0.14084, + "grad_norm": 1.285007947410615, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 14084 + }, + { + "epoch": 0.14085, + "grad_norm": 1.067547803950603, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 14085 + }, + { + "epoch": 0.14086, + "grad_norm": 1.181939954762276, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 14086 + }, + { + "epoch": 0.14087, + "grad_norm": 0.9252942965271072, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 14087 + }, + { + "epoch": 0.14088, + "grad_norm": 0.864403407919305, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 14088 + }, + { + "epoch": 0.14089, + "grad_norm": 0.8697682040171842, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 14089 + }, + { + "epoch": 0.1409, + "grad_norm": 0.9564099660022675, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 14090 + }, + { + "epoch": 0.14091, + "grad_norm": 1.1099060579686737, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 14091 + }, + { + "epoch": 0.14092, + "grad_norm": 1.0537091221596513, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 14092 + }, + { + "epoch": 0.14093, + "grad_norm": 1.121502142018702, + "learning_rate": 0.003, + "loss": 4.069, + "step": 14093 + }, + { + "epoch": 0.14094, + "grad_norm": 0.8823991398599901, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 14094 + }, + { + "epoch": 0.14095, + "grad_norm": 0.7743650402228722, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 14095 + }, + { + "epoch": 0.14096, + "grad_norm": 0.7158317692551045, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 14096 + }, + { + "epoch": 0.14097, + "grad_norm": 0.8272539063669545, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 14097 + }, + { + "epoch": 0.14098, + "grad_norm": 0.9686409775258437, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 14098 + }, + { + "epoch": 0.14099, + "grad_norm": 1.0444622963382162, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 14099 + }, + { + "epoch": 0.141, + "grad_norm": 1.0149562039648299, + "learning_rate": 0.003, + "loss": 4.1008, + "step": 14100 + }, + { + "epoch": 0.14101, + "grad_norm": 0.9675234604390469, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 14101 + }, + { + "epoch": 0.14102, + "grad_norm": 0.909123150054251, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 14102 + }, + { + "epoch": 0.14103, + "grad_norm": 0.8854412984507708, + "learning_rate": 0.003, + "loss": 4.079, + "step": 14103 + }, + { + "epoch": 0.14104, + "grad_norm": 0.9725639647661376, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 14104 + }, + { + "epoch": 0.14105, + "grad_norm": 1.1096207230822221, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 14105 + }, + { + "epoch": 0.14106, + "grad_norm": 1.0922784697608032, + "learning_rate": 0.003, + "loss": 4.075, + "step": 14106 + }, + { + "epoch": 0.14107, + "grad_norm": 0.9558184384507955, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 14107 + }, + { + "epoch": 0.14108, + "grad_norm": 0.9621337287920178, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 14108 + }, + { + "epoch": 0.14109, + "grad_norm": 1.035559075169336, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 14109 + }, + { + "epoch": 0.1411, + "grad_norm": 0.9419798651206963, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 14110 + }, + { + "epoch": 0.14111, + "grad_norm": 1.031656905838924, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 14111 + }, + { + "epoch": 0.14112, + "grad_norm": 1.0714244981498335, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 14112 + }, + { + "epoch": 0.14113, + "grad_norm": 1.020846902707669, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 14113 + }, + { + "epoch": 0.14114, + "grad_norm": 1.0703635598296604, + "learning_rate": 0.003, + "loss": 4.08, + "step": 14114 + }, + { + "epoch": 0.14115, + "grad_norm": 0.9912975132847905, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 14115 + }, + { + "epoch": 0.14116, + "grad_norm": 0.9966017307079775, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 14116 + }, + { + "epoch": 0.14117, + "grad_norm": 0.9240846860619585, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 14117 + }, + { + "epoch": 0.14118, + "grad_norm": 0.9262132284494319, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 14118 + }, + { + "epoch": 0.14119, + "grad_norm": 0.9007229397204254, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 14119 + }, + { + "epoch": 0.1412, + "grad_norm": 0.8203422649191234, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 14120 + }, + { + "epoch": 0.14121, + "grad_norm": 0.8236229019325196, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 14121 + }, + { + "epoch": 0.14122, + "grad_norm": 0.9778501812037531, + "learning_rate": 0.003, + "loss": 4.0977, + "step": 14122 + }, + { + "epoch": 0.14123, + "grad_norm": 1.0877914065162442, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 14123 + }, + { + "epoch": 0.14124, + "grad_norm": 0.8448633961428123, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 14124 + }, + { + "epoch": 0.14125, + "grad_norm": 0.803250264473479, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 14125 + }, + { + "epoch": 0.14126, + "grad_norm": 0.7462293721837611, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 14126 + }, + { + "epoch": 0.14127, + "grad_norm": 0.6343397441782072, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 14127 + }, + { + "epoch": 0.14128, + "grad_norm": 0.7807343484039558, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 14128 + }, + { + "epoch": 0.14129, + "grad_norm": 0.9863944680630348, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 14129 + }, + { + "epoch": 0.1413, + "grad_norm": 1.127534148321148, + "learning_rate": 0.003, + "loss": 4.0956, + "step": 14130 + }, + { + "epoch": 0.14131, + "grad_norm": 0.9908991122569281, + "learning_rate": 0.003, + "loss": 4.069, + "step": 14131 + }, + { + "epoch": 0.14132, + "grad_norm": 1.197819795531229, + "learning_rate": 0.003, + "loss": 4.1041, + "step": 14132 + }, + { + "epoch": 0.14133, + "grad_norm": 0.9360220406257987, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 14133 + }, + { + "epoch": 0.14134, + "grad_norm": 1.0819626512695029, + "learning_rate": 0.003, + "loss": 4.0999, + "step": 14134 + }, + { + "epoch": 0.14135, + "grad_norm": 0.9786323533808043, + "learning_rate": 0.003, + "loss": 4.0917, + "step": 14135 + }, + { + "epoch": 0.14136, + "grad_norm": 0.9475983224614479, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 14136 + }, + { + "epoch": 0.14137, + "grad_norm": 0.9523057680633161, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 14137 + }, + { + "epoch": 0.14138, + "grad_norm": 0.9953890510659464, + "learning_rate": 0.003, + "loss": 4.0819, + "step": 14138 + }, + { + "epoch": 0.14139, + "grad_norm": 1.098025442722754, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 14139 + }, + { + "epoch": 0.1414, + "grad_norm": 1.0702507297358692, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 14140 + }, + { + "epoch": 0.14141, + "grad_norm": 1.2628169615471436, + "learning_rate": 0.003, + "loss": 4.0854, + "step": 14141 + }, + { + "epoch": 0.14142, + "grad_norm": 0.9154039779523112, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 14142 + }, + { + "epoch": 0.14143, + "grad_norm": 0.8012789916344393, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 14143 + }, + { + "epoch": 0.14144, + "grad_norm": 0.7968051973425723, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 14144 + }, + { + "epoch": 0.14145, + "grad_norm": 0.7690442121297268, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 14145 + }, + { + "epoch": 0.14146, + "grad_norm": 0.8605712517287046, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 14146 + }, + { + "epoch": 0.14147, + "grad_norm": 0.9487088456588897, + "learning_rate": 0.003, + "loss": 4.081, + "step": 14147 + }, + { + "epoch": 0.14148, + "grad_norm": 1.0035505750105043, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 14148 + }, + { + "epoch": 0.14149, + "grad_norm": 1.1015947925393863, + "learning_rate": 0.003, + "loss": 4.0868, + "step": 14149 + }, + { + "epoch": 0.1415, + "grad_norm": 0.8831054155536091, + "learning_rate": 0.003, + "loss": 4.0966, + "step": 14150 + }, + { + "epoch": 0.14151, + "grad_norm": 0.8121773241938571, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 14151 + }, + { + "epoch": 0.14152, + "grad_norm": 0.741251186343512, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 14152 + }, + { + "epoch": 0.14153, + "grad_norm": 0.6535082293118707, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 14153 + }, + { + "epoch": 0.14154, + "grad_norm": 0.7488128814236085, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 14154 + }, + { + "epoch": 0.14155, + "grad_norm": 0.9669075717247493, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 14155 + }, + { + "epoch": 0.14156, + "grad_norm": 1.2217160071389768, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 14156 + }, + { + "epoch": 0.14157, + "grad_norm": 0.844289897464921, + "learning_rate": 0.003, + "loss": 4.0864, + "step": 14157 + }, + { + "epoch": 0.14158, + "grad_norm": 0.9286215865390686, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 14158 + }, + { + "epoch": 0.14159, + "grad_norm": 0.94900192779694, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 14159 + }, + { + "epoch": 0.1416, + "grad_norm": 1.011059192231509, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 14160 + }, + { + "epoch": 0.14161, + "grad_norm": 1.3481631896006976, + "learning_rate": 0.003, + "loss": 4.1245, + "step": 14161 + }, + { + "epoch": 0.14162, + "grad_norm": 0.9923270509533557, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 14162 + }, + { + "epoch": 0.14163, + "grad_norm": 1.1545598421630299, + "learning_rate": 0.003, + "loss": 4.075, + "step": 14163 + }, + { + "epoch": 0.14164, + "grad_norm": 1.0171213186396129, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 14164 + }, + { + "epoch": 0.14165, + "grad_norm": 1.0753377377688569, + "learning_rate": 0.003, + "loss": 4.0972, + "step": 14165 + }, + { + "epoch": 0.14166, + "grad_norm": 0.9980352192187116, + "learning_rate": 0.003, + "loss": 4.0937, + "step": 14166 + }, + { + "epoch": 0.14167, + "grad_norm": 1.0416339713222473, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 14167 + }, + { + "epoch": 0.14168, + "grad_norm": 1.0844378736643812, + "learning_rate": 0.003, + "loss": 4.0868, + "step": 14168 + }, + { + "epoch": 0.14169, + "grad_norm": 1.0261645611495915, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 14169 + }, + { + "epoch": 0.1417, + "grad_norm": 0.9667447490403226, + "learning_rate": 0.003, + "loss": 4.0855, + "step": 14170 + }, + { + "epoch": 0.14171, + "grad_norm": 1.011988449724314, + "learning_rate": 0.003, + "loss": 4.0888, + "step": 14171 + }, + { + "epoch": 0.14172, + "grad_norm": 1.0019800092626514, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 14172 + }, + { + "epoch": 0.14173, + "grad_norm": 0.7825143415489462, + "learning_rate": 0.003, + "loss": 4.0881, + "step": 14173 + }, + { + "epoch": 0.14174, + "grad_norm": 0.7374700422552599, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 14174 + }, + { + "epoch": 0.14175, + "grad_norm": 0.946470844261005, + "learning_rate": 0.003, + "loss": 4.0978, + "step": 14175 + }, + { + "epoch": 0.14176, + "grad_norm": 1.2929369011854794, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 14176 + }, + { + "epoch": 0.14177, + "grad_norm": 0.8939843231530757, + "learning_rate": 0.003, + "loss": 4.1137, + "step": 14177 + }, + { + "epoch": 0.14178, + "grad_norm": 0.9342643770015284, + "learning_rate": 0.003, + "loss": 4.043, + "step": 14178 + }, + { + "epoch": 0.14179, + "grad_norm": 1.033800015338059, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 14179 + }, + { + "epoch": 0.1418, + "grad_norm": 1.0743333040193197, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 14180 + }, + { + "epoch": 0.14181, + "grad_norm": 0.9594910902463681, + "learning_rate": 0.003, + "loss": 4.0845, + "step": 14181 + }, + { + "epoch": 0.14182, + "grad_norm": 0.9748560764291033, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 14182 + }, + { + "epoch": 0.14183, + "grad_norm": 1.2437375580838774, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 14183 + }, + { + "epoch": 0.14184, + "grad_norm": 1.0548635960026467, + "learning_rate": 0.003, + "loss": 4.0977, + "step": 14184 + }, + { + "epoch": 0.14185, + "grad_norm": 1.1385673451552012, + "learning_rate": 0.003, + "loss": 4.0932, + "step": 14185 + }, + { + "epoch": 0.14186, + "grad_norm": 0.9725871569432094, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 14186 + }, + { + "epoch": 0.14187, + "grad_norm": 0.8649221651736533, + "learning_rate": 0.003, + "loss": 4.088, + "step": 14187 + }, + { + "epoch": 0.14188, + "grad_norm": 0.8129549105780057, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 14188 + }, + { + "epoch": 0.14189, + "grad_norm": 0.749811838233913, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 14189 + }, + { + "epoch": 0.1419, + "grad_norm": 0.7538889235902039, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 14190 + }, + { + "epoch": 0.14191, + "grad_norm": 0.8181397442592956, + "learning_rate": 0.003, + "loss": 4.073, + "step": 14191 + }, + { + "epoch": 0.14192, + "grad_norm": 0.7705947024785165, + "learning_rate": 0.003, + "loss": 4.1173, + "step": 14192 + }, + { + "epoch": 0.14193, + "grad_norm": 0.8210590381121436, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 14193 + }, + { + "epoch": 0.14194, + "grad_norm": 0.8498607678894289, + "learning_rate": 0.003, + "loss": 4.1052, + "step": 14194 + }, + { + "epoch": 0.14195, + "grad_norm": 0.9582399260206752, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 14195 + }, + { + "epoch": 0.14196, + "grad_norm": 1.0445960566466224, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 14196 + }, + { + "epoch": 0.14197, + "grad_norm": 0.9549894698771688, + "learning_rate": 0.003, + "loss": 4.082, + "step": 14197 + }, + { + "epoch": 0.14198, + "grad_norm": 0.9251253965313724, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 14198 + }, + { + "epoch": 0.14199, + "grad_norm": 1.1468735137490733, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 14199 + }, + { + "epoch": 0.142, + "grad_norm": 1.1104135274215978, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 14200 + }, + { + "epoch": 0.14201, + "grad_norm": 1.059792216793346, + "learning_rate": 0.003, + "loss": 4.044, + "step": 14201 + }, + { + "epoch": 0.14202, + "grad_norm": 1.0018538825729457, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 14202 + }, + { + "epoch": 0.14203, + "grad_norm": 1.0642425602761003, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 14203 + }, + { + "epoch": 0.14204, + "grad_norm": 0.9769106660602661, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 14204 + }, + { + "epoch": 0.14205, + "grad_norm": 1.0959725070657629, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 14205 + }, + { + "epoch": 0.14206, + "grad_norm": 0.9465094657730181, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 14206 + }, + { + "epoch": 0.14207, + "grad_norm": 0.8647954617701136, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 14207 + }, + { + "epoch": 0.14208, + "grad_norm": 0.9644044177495231, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 14208 + }, + { + "epoch": 0.14209, + "grad_norm": 1.0216727223691575, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 14209 + }, + { + "epoch": 0.1421, + "grad_norm": 0.9766456733040981, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 14210 + }, + { + "epoch": 0.14211, + "grad_norm": 0.9567672366253372, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 14211 + }, + { + "epoch": 0.14212, + "grad_norm": 1.1053375045703002, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 14212 + }, + { + "epoch": 0.14213, + "grad_norm": 1.1591683249466502, + "learning_rate": 0.003, + "loss": 4.0884, + "step": 14213 + }, + { + "epoch": 0.14214, + "grad_norm": 1.061530709198683, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 14214 + }, + { + "epoch": 0.14215, + "grad_norm": 0.9867363700957806, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 14215 + }, + { + "epoch": 0.14216, + "grad_norm": 1.1318673865222753, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 14216 + }, + { + "epoch": 0.14217, + "grad_norm": 1.0155519662172816, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 14217 + }, + { + "epoch": 0.14218, + "grad_norm": 1.162288146727698, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 14218 + }, + { + "epoch": 0.14219, + "grad_norm": 0.9997006996773105, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 14219 + }, + { + "epoch": 0.1422, + "grad_norm": 1.0606951666030384, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 14220 + }, + { + "epoch": 0.14221, + "grad_norm": 1.0676373037653488, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 14221 + }, + { + "epoch": 0.14222, + "grad_norm": 0.8528516220752385, + "learning_rate": 0.003, + "loss": 4.0999, + "step": 14222 + }, + { + "epoch": 0.14223, + "grad_norm": 0.8395597665248518, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 14223 + }, + { + "epoch": 0.14224, + "grad_norm": 0.884317661594737, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 14224 + }, + { + "epoch": 0.14225, + "grad_norm": 0.983409135224199, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 14225 + }, + { + "epoch": 0.14226, + "grad_norm": 1.1399990431649805, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 14226 + }, + { + "epoch": 0.14227, + "grad_norm": 0.769150748867688, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 14227 + }, + { + "epoch": 0.14228, + "grad_norm": 0.6380785062301577, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 14228 + }, + { + "epoch": 0.14229, + "grad_norm": 0.6774278842721995, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 14229 + }, + { + "epoch": 0.1423, + "grad_norm": 0.7949269089082509, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 14230 + }, + { + "epoch": 0.14231, + "grad_norm": 0.9874395261030108, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 14231 + }, + { + "epoch": 0.14232, + "grad_norm": 1.1255727312767339, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 14232 + }, + { + "epoch": 0.14233, + "grad_norm": 0.8952354342133777, + "learning_rate": 0.003, + "loss": 4.1003, + "step": 14233 + }, + { + "epoch": 0.14234, + "grad_norm": 0.8409643060109887, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 14234 + }, + { + "epoch": 0.14235, + "grad_norm": 0.7385820329570673, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 14235 + }, + { + "epoch": 0.14236, + "grad_norm": 0.847149898180548, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 14236 + }, + { + "epoch": 0.14237, + "grad_norm": 0.976832745178921, + "learning_rate": 0.003, + "loss": 4.063, + "step": 14237 + }, + { + "epoch": 0.14238, + "grad_norm": 1.002793150223732, + "learning_rate": 0.003, + "loss": 4.0888, + "step": 14238 + }, + { + "epoch": 0.14239, + "grad_norm": 1.1121294713069256, + "learning_rate": 0.003, + "loss": 4.1104, + "step": 14239 + }, + { + "epoch": 0.1424, + "grad_norm": 1.0642758928150466, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 14240 + }, + { + "epoch": 0.14241, + "grad_norm": 0.9961251128832938, + "learning_rate": 0.003, + "loss": 4.1019, + "step": 14241 + }, + { + "epoch": 0.14242, + "grad_norm": 1.037694514092418, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 14242 + }, + { + "epoch": 0.14243, + "grad_norm": 0.9359883098304765, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 14243 + }, + { + "epoch": 0.14244, + "grad_norm": 0.9960538656239689, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 14244 + }, + { + "epoch": 0.14245, + "grad_norm": 1.1938908627690623, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 14245 + }, + { + "epoch": 0.14246, + "grad_norm": 1.1594343349254745, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 14246 + }, + { + "epoch": 0.14247, + "grad_norm": 0.9247259275193648, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 14247 + }, + { + "epoch": 0.14248, + "grad_norm": 1.025318518441329, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 14248 + }, + { + "epoch": 0.14249, + "grad_norm": 1.161000217482664, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 14249 + }, + { + "epoch": 0.1425, + "grad_norm": 0.8697162201802691, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 14250 + }, + { + "epoch": 0.14251, + "grad_norm": 0.801196596499568, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 14251 + }, + { + "epoch": 0.14252, + "grad_norm": 0.7245503610311297, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 14252 + }, + { + "epoch": 0.14253, + "grad_norm": 0.6947962080630692, + "learning_rate": 0.003, + "loss": 4.0946, + "step": 14253 + }, + { + "epoch": 0.14254, + "grad_norm": 0.7503927227040603, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 14254 + }, + { + "epoch": 0.14255, + "grad_norm": 0.8210038862656045, + "learning_rate": 0.003, + "loss": 4.057, + "step": 14255 + }, + { + "epoch": 0.14256, + "grad_norm": 0.8416004031984307, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 14256 + }, + { + "epoch": 0.14257, + "grad_norm": 0.9301859209238496, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 14257 + }, + { + "epoch": 0.14258, + "grad_norm": 0.9593645999953336, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 14258 + }, + { + "epoch": 0.14259, + "grad_norm": 0.8761228629745896, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 14259 + }, + { + "epoch": 0.1426, + "grad_norm": 1.0408447270987904, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 14260 + }, + { + "epoch": 0.14261, + "grad_norm": 1.096716448997523, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 14261 + }, + { + "epoch": 0.14262, + "grad_norm": 1.1841636732017307, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 14262 + }, + { + "epoch": 0.14263, + "grad_norm": 1.226486378870345, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 14263 + }, + { + "epoch": 0.14264, + "grad_norm": 0.972279759288991, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 14264 + }, + { + "epoch": 0.14265, + "grad_norm": 0.7901402263670024, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 14265 + }, + { + "epoch": 0.14266, + "grad_norm": 0.8365040905009538, + "learning_rate": 0.003, + "loss": 4.089, + "step": 14266 + }, + { + "epoch": 0.14267, + "grad_norm": 0.9625464312803685, + "learning_rate": 0.003, + "loss": 4.0936, + "step": 14267 + }, + { + "epoch": 0.14268, + "grad_norm": 1.0117710636727402, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 14268 + }, + { + "epoch": 0.14269, + "grad_norm": 1.0288358235225066, + "learning_rate": 0.003, + "loss": 4.0996, + "step": 14269 + }, + { + "epoch": 0.1427, + "grad_norm": 1.1761583735092824, + "learning_rate": 0.003, + "loss": 4.0908, + "step": 14270 + }, + { + "epoch": 0.14271, + "grad_norm": 0.8299507597201572, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 14271 + }, + { + "epoch": 0.14272, + "grad_norm": 0.8775059468405105, + "learning_rate": 0.003, + "loss": 4.0918, + "step": 14272 + }, + { + "epoch": 0.14273, + "grad_norm": 1.0450278375864777, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 14273 + }, + { + "epoch": 0.14274, + "grad_norm": 1.0585447297138586, + "learning_rate": 0.003, + "loss": 4.0857, + "step": 14274 + }, + { + "epoch": 0.14275, + "grad_norm": 1.1289399939522247, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 14275 + }, + { + "epoch": 0.14276, + "grad_norm": 0.9675938998914626, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 14276 + }, + { + "epoch": 0.14277, + "grad_norm": 0.9807168476049197, + "learning_rate": 0.003, + "loss": 4.0905, + "step": 14277 + }, + { + "epoch": 0.14278, + "grad_norm": 1.0405577849142709, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 14278 + }, + { + "epoch": 0.14279, + "grad_norm": 0.9352667434153391, + "learning_rate": 0.003, + "loss": 4.1014, + "step": 14279 + }, + { + "epoch": 0.1428, + "grad_norm": 1.00127698019752, + "learning_rate": 0.003, + "loss": 4.0953, + "step": 14280 + }, + { + "epoch": 0.14281, + "grad_norm": 1.0738948991271482, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 14281 + }, + { + "epoch": 0.14282, + "grad_norm": 0.9551436880100509, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 14282 + }, + { + "epoch": 0.14283, + "grad_norm": 0.864809199768195, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 14283 + }, + { + "epoch": 0.14284, + "grad_norm": 0.8823517279918525, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 14284 + }, + { + "epoch": 0.14285, + "grad_norm": 1.0025070289252958, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 14285 + }, + { + "epoch": 0.14286, + "grad_norm": 1.2838725601473926, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 14286 + }, + { + "epoch": 0.14287, + "grad_norm": 1.0354378989023456, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 14287 + }, + { + "epoch": 0.14288, + "grad_norm": 0.9681214962466057, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 14288 + }, + { + "epoch": 0.14289, + "grad_norm": 1.079204915043843, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 14289 + }, + { + "epoch": 0.1429, + "grad_norm": 0.9405912886618573, + "learning_rate": 0.003, + "loss": 4.041, + "step": 14290 + }, + { + "epoch": 0.14291, + "grad_norm": 0.8190550707768256, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 14291 + }, + { + "epoch": 0.14292, + "grad_norm": 0.7823907423268109, + "learning_rate": 0.003, + "loss": 4.069, + "step": 14292 + }, + { + "epoch": 0.14293, + "grad_norm": 0.6929634719908795, + "learning_rate": 0.003, + "loss": 4.04, + "step": 14293 + }, + { + "epoch": 0.14294, + "grad_norm": 0.6520132280929395, + "learning_rate": 0.003, + "loss": 4.0919, + "step": 14294 + }, + { + "epoch": 0.14295, + "grad_norm": 0.7216275884543363, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 14295 + }, + { + "epoch": 0.14296, + "grad_norm": 0.8042597561227761, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 14296 + }, + { + "epoch": 0.14297, + "grad_norm": 0.9252637750553753, + "learning_rate": 0.003, + "loss": 4.0855, + "step": 14297 + }, + { + "epoch": 0.14298, + "grad_norm": 1.11661852050436, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 14298 + }, + { + "epoch": 0.14299, + "grad_norm": 1.0414808456752196, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 14299 + }, + { + "epoch": 0.143, + "grad_norm": 1.232152060613595, + "learning_rate": 0.003, + "loss": 4.0972, + "step": 14300 + }, + { + "epoch": 0.14301, + "grad_norm": 0.8579087246700543, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 14301 + }, + { + "epoch": 0.14302, + "grad_norm": 0.739864730479044, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 14302 + }, + { + "epoch": 0.14303, + "grad_norm": 0.7674165324124369, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 14303 + }, + { + "epoch": 0.14304, + "grad_norm": 0.9906438130670885, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 14304 + }, + { + "epoch": 0.14305, + "grad_norm": 1.1808445104833238, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 14305 + }, + { + "epoch": 0.14306, + "grad_norm": 0.7469007128444103, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 14306 + }, + { + "epoch": 0.14307, + "grad_norm": 0.8708267970323613, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 14307 + }, + { + "epoch": 0.14308, + "grad_norm": 1.2194849041031937, + "learning_rate": 0.003, + "loss": 4.0976, + "step": 14308 + }, + { + "epoch": 0.14309, + "grad_norm": 0.9899655987765626, + "learning_rate": 0.003, + "loss": 4.0985, + "step": 14309 + }, + { + "epoch": 0.1431, + "grad_norm": 0.9574658803683402, + "learning_rate": 0.003, + "loss": 4.061, + "step": 14310 + }, + { + "epoch": 0.14311, + "grad_norm": 1.0197743109237534, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 14311 + }, + { + "epoch": 0.14312, + "grad_norm": 1.0404470562032275, + "learning_rate": 0.003, + "loss": 4.06, + "step": 14312 + }, + { + "epoch": 0.14313, + "grad_norm": 1.0678068357239305, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 14313 + }, + { + "epoch": 0.14314, + "grad_norm": 0.9673582402495701, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 14314 + }, + { + "epoch": 0.14315, + "grad_norm": 1.0313050046207342, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 14315 + }, + { + "epoch": 0.14316, + "grad_norm": 0.9731165238562522, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 14316 + }, + { + "epoch": 0.14317, + "grad_norm": 0.8933575669535071, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 14317 + }, + { + "epoch": 0.14318, + "grad_norm": 0.7764112299520958, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 14318 + }, + { + "epoch": 0.14319, + "grad_norm": 0.7901061215152928, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 14319 + }, + { + "epoch": 0.1432, + "grad_norm": 0.943717446430798, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 14320 + }, + { + "epoch": 0.14321, + "grad_norm": 1.1815852028723917, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 14321 + }, + { + "epoch": 0.14322, + "grad_norm": 0.8704461590304073, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 14322 + }, + { + "epoch": 0.14323, + "grad_norm": 0.8444179554546244, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 14323 + }, + { + "epoch": 0.14324, + "grad_norm": 1.0686982472102307, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 14324 + }, + { + "epoch": 0.14325, + "grad_norm": 1.535539068555881, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 14325 + }, + { + "epoch": 0.14326, + "grad_norm": 0.7697033853088325, + "learning_rate": 0.003, + "loss": 4.0942, + "step": 14326 + }, + { + "epoch": 0.14327, + "grad_norm": 0.798103382240663, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 14327 + }, + { + "epoch": 0.14328, + "grad_norm": 0.7890713227798192, + "learning_rate": 0.003, + "loss": 4.1144, + "step": 14328 + }, + { + "epoch": 0.14329, + "grad_norm": 0.7783203127598053, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 14329 + }, + { + "epoch": 0.1433, + "grad_norm": 0.9074092208687227, + "learning_rate": 0.003, + "loss": 4.0878, + "step": 14330 + }, + { + "epoch": 0.14331, + "grad_norm": 0.8996206419297509, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 14331 + }, + { + "epoch": 0.14332, + "grad_norm": 1.1243796328174156, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 14332 + }, + { + "epoch": 0.14333, + "grad_norm": 1.1819340231028863, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 14333 + }, + { + "epoch": 0.14334, + "grad_norm": 1.049990468736676, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 14334 + }, + { + "epoch": 0.14335, + "grad_norm": 0.971861088056968, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 14335 + }, + { + "epoch": 0.14336, + "grad_norm": 1.069946016064609, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 14336 + }, + { + "epoch": 0.14337, + "grad_norm": 1.0594680235823368, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 14337 + }, + { + "epoch": 0.14338, + "grad_norm": 1.160791662911144, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 14338 + }, + { + "epoch": 0.14339, + "grad_norm": 0.9314505014587761, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 14339 + }, + { + "epoch": 0.1434, + "grad_norm": 0.9542962341121145, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 14340 + }, + { + "epoch": 0.14341, + "grad_norm": 0.9787711878490403, + "learning_rate": 0.003, + "loss": 4.0985, + "step": 14341 + }, + { + "epoch": 0.14342, + "grad_norm": 1.0980859293627518, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 14342 + }, + { + "epoch": 0.14343, + "grad_norm": 1.0935348962699218, + "learning_rate": 0.003, + "loss": 4.1177, + "step": 14343 + }, + { + "epoch": 0.14344, + "grad_norm": 1.0254155082095509, + "learning_rate": 0.003, + "loss": 4.1147, + "step": 14344 + }, + { + "epoch": 0.14345, + "grad_norm": 1.0660138880848296, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 14345 + }, + { + "epoch": 0.14346, + "grad_norm": 0.9768509011199782, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 14346 + }, + { + "epoch": 0.14347, + "grad_norm": 1.0531394297520724, + "learning_rate": 0.003, + "loss": 4.097, + "step": 14347 + }, + { + "epoch": 0.14348, + "grad_norm": 0.9183223511201126, + "learning_rate": 0.003, + "loss": 4.0955, + "step": 14348 + }, + { + "epoch": 0.14349, + "grad_norm": 0.9711078742427085, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 14349 + }, + { + "epoch": 0.1435, + "grad_norm": 1.16256235160606, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 14350 + }, + { + "epoch": 0.14351, + "grad_norm": 0.9720899068305483, + "learning_rate": 0.003, + "loss": 4.084, + "step": 14351 + }, + { + "epoch": 0.14352, + "grad_norm": 0.9834990054186945, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 14352 + }, + { + "epoch": 0.14353, + "grad_norm": 1.1187911721478752, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 14353 + }, + { + "epoch": 0.14354, + "grad_norm": 0.9406210875655281, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 14354 + }, + { + "epoch": 0.14355, + "grad_norm": 0.9444705386552177, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 14355 + }, + { + "epoch": 0.14356, + "grad_norm": 1.0707287056681185, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 14356 + }, + { + "epoch": 0.14357, + "grad_norm": 1.0508735816894448, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 14357 + }, + { + "epoch": 0.14358, + "grad_norm": 1.2948640627053416, + "learning_rate": 0.003, + "loss": 4.061, + "step": 14358 + }, + { + "epoch": 0.14359, + "grad_norm": 0.8958831069782376, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 14359 + }, + { + "epoch": 0.1436, + "grad_norm": 0.7981066692759334, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 14360 + }, + { + "epoch": 0.14361, + "grad_norm": 0.8243097493340907, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 14361 + }, + { + "epoch": 0.14362, + "grad_norm": 0.8242673378787606, + "learning_rate": 0.003, + "loss": 4.1144, + "step": 14362 + }, + { + "epoch": 0.14363, + "grad_norm": 0.736927723695665, + "learning_rate": 0.003, + "loss": 4.061, + "step": 14363 + }, + { + "epoch": 0.14364, + "grad_norm": 0.7787313421624953, + "learning_rate": 0.003, + "loss": 4.0857, + "step": 14364 + }, + { + "epoch": 0.14365, + "grad_norm": 0.8160863789344391, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 14365 + }, + { + "epoch": 0.14366, + "grad_norm": 0.8620891241148487, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 14366 + }, + { + "epoch": 0.14367, + "grad_norm": 0.8794551313843108, + "learning_rate": 0.003, + "loss": 4.054, + "step": 14367 + }, + { + "epoch": 0.14368, + "grad_norm": 1.0002490260759087, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 14368 + }, + { + "epoch": 0.14369, + "grad_norm": 1.194849039971133, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 14369 + }, + { + "epoch": 0.1437, + "grad_norm": 1.1521321256197925, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 14370 + }, + { + "epoch": 0.14371, + "grad_norm": 0.9228474837388875, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 14371 + }, + { + "epoch": 0.14372, + "grad_norm": 1.1096299146762212, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 14372 + }, + { + "epoch": 0.14373, + "grad_norm": 1.0660503233006322, + "learning_rate": 0.003, + "loss": 4.092, + "step": 14373 + }, + { + "epoch": 0.14374, + "grad_norm": 0.8934732494938307, + "learning_rate": 0.003, + "loss": 4.079, + "step": 14374 + }, + { + "epoch": 0.14375, + "grad_norm": 0.8040609305376994, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 14375 + }, + { + "epoch": 0.14376, + "grad_norm": 0.8584131322992812, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 14376 + }, + { + "epoch": 0.14377, + "grad_norm": 0.9313320560391131, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 14377 + }, + { + "epoch": 0.14378, + "grad_norm": 1.278653069290951, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 14378 + }, + { + "epoch": 0.14379, + "grad_norm": 0.728366167387422, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 14379 + }, + { + "epoch": 0.1438, + "grad_norm": 0.6345846674975081, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 14380 + }, + { + "epoch": 0.14381, + "grad_norm": 0.6744235219812535, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 14381 + }, + { + "epoch": 0.14382, + "grad_norm": 0.8011324105583683, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 14382 + }, + { + "epoch": 0.14383, + "grad_norm": 0.8442588251905928, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 14383 + }, + { + "epoch": 0.14384, + "grad_norm": 0.881618117044, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 14384 + }, + { + "epoch": 0.14385, + "grad_norm": 0.9784807322973302, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 14385 + }, + { + "epoch": 0.14386, + "grad_norm": 1.0982501093157857, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 14386 + }, + { + "epoch": 0.14387, + "grad_norm": 1.1321803041402188, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 14387 + }, + { + "epoch": 0.14388, + "grad_norm": 0.9892618320685146, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 14388 + }, + { + "epoch": 0.14389, + "grad_norm": 1.0664814742357072, + "learning_rate": 0.003, + "loss": 4.1004, + "step": 14389 + }, + { + "epoch": 0.1439, + "grad_norm": 1.2608075239660734, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 14390 + }, + { + "epoch": 0.14391, + "grad_norm": 1.0894124464541353, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 14391 + }, + { + "epoch": 0.14392, + "grad_norm": 1.0775124371104892, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 14392 + }, + { + "epoch": 0.14393, + "grad_norm": 1.0923113837721887, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 14393 + }, + { + "epoch": 0.14394, + "grad_norm": 0.9836641993525376, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 14394 + }, + { + "epoch": 0.14395, + "grad_norm": 1.0358052462211582, + "learning_rate": 0.003, + "loss": 4.071, + "step": 14395 + }, + { + "epoch": 0.14396, + "grad_norm": 0.9549943521994575, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 14396 + }, + { + "epoch": 0.14397, + "grad_norm": 0.864019385782395, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 14397 + }, + { + "epoch": 0.14398, + "grad_norm": 0.8550713015356483, + "learning_rate": 0.003, + "loss": 4.059, + "step": 14398 + }, + { + "epoch": 0.14399, + "grad_norm": 1.087830840700268, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 14399 + }, + { + "epoch": 0.144, + "grad_norm": 1.2360787456485025, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 14400 + }, + { + "epoch": 0.14401, + "grad_norm": 0.7542010886384433, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 14401 + }, + { + "epoch": 0.14402, + "grad_norm": 0.8893387755605521, + "learning_rate": 0.003, + "loss": 4.0902, + "step": 14402 + }, + { + "epoch": 0.14403, + "grad_norm": 0.9287770963774216, + "learning_rate": 0.003, + "loss": 4.0915, + "step": 14403 + }, + { + "epoch": 0.14404, + "grad_norm": 0.9539507098990296, + "learning_rate": 0.003, + "loss": 4.0912, + "step": 14404 + }, + { + "epoch": 0.14405, + "grad_norm": 0.9053713583447572, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 14405 + }, + { + "epoch": 0.14406, + "grad_norm": 0.931409159707179, + "learning_rate": 0.003, + "loss": 4.0993, + "step": 14406 + }, + { + "epoch": 0.14407, + "grad_norm": 0.8928360545722142, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 14407 + }, + { + "epoch": 0.14408, + "grad_norm": 0.9632015436419991, + "learning_rate": 0.003, + "loss": 4.1012, + "step": 14408 + }, + { + "epoch": 0.14409, + "grad_norm": 1.241954492074289, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 14409 + }, + { + "epoch": 0.1441, + "grad_norm": 1.0494342399159942, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 14410 + }, + { + "epoch": 0.14411, + "grad_norm": 0.8927721181064726, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 14411 + }, + { + "epoch": 0.14412, + "grad_norm": 0.8822915661743962, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 14412 + }, + { + "epoch": 0.14413, + "grad_norm": 1.0421920317404036, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 14413 + }, + { + "epoch": 0.14414, + "grad_norm": 1.1909746130209655, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 14414 + }, + { + "epoch": 0.14415, + "grad_norm": 0.902161870825093, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 14415 + }, + { + "epoch": 0.14416, + "grad_norm": 0.7757454363934395, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 14416 + }, + { + "epoch": 0.14417, + "grad_norm": 0.7562758606013457, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 14417 + }, + { + "epoch": 0.14418, + "grad_norm": 0.7973803498365637, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 14418 + }, + { + "epoch": 0.14419, + "grad_norm": 0.8627639082914278, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 14419 + }, + { + "epoch": 0.1442, + "grad_norm": 0.7603259280620488, + "learning_rate": 0.003, + "loss": 4.082, + "step": 14420 + }, + { + "epoch": 0.14421, + "grad_norm": 0.7795340619438654, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 14421 + }, + { + "epoch": 0.14422, + "grad_norm": 0.7476597513614887, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 14422 + }, + { + "epoch": 0.14423, + "grad_norm": 0.767831402693119, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 14423 + }, + { + "epoch": 0.14424, + "grad_norm": 0.838094853986366, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 14424 + }, + { + "epoch": 0.14425, + "grad_norm": 0.9115932501719597, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 14425 + }, + { + "epoch": 0.14426, + "grad_norm": 1.0717394252589778, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 14426 + }, + { + "epoch": 0.14427, + "grad_norm": 1.072151386698525, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 14427 + }, + { + "epoch": 0.14428, + "grad_norm": 0.9752752560545125, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 14428 + }, + { + "epoch": 0.14429, + "grad_norm": 1.2197159185551836, + "learning_rate": 0.003, + "loss": 4.0987, + "step": 14429 + }, + { + "epoch": 0.1443, + "grad_norm": 1.0230537847847552, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 14430 + }, + { + "epoch": 0.14431, + "grad_norm": 0.9521307641744597, + "learning_rate": 0.003, + "loss": 4.0944, + "step": 14431 + }, + { + "epoch": 0.14432, + "grad_norm": 0.9262154601464844, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 14432 + }, + { + "epoch": 0.14433, + "grad_norm": 1.004903335291544, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 14433 + }, + { + "epoch": 0.14434, + "grad_norm": 1.061345719165486, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 14434 + }, + { + "epoch": 0.14435, + "grad_norm": 0.9021720869385274, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 14435 + }, + { + "epoch": 0.14436, + "grad_norm": 1.0460353996124319, + "learning_rate": 0.003, + "loss": 4.0995, + "step": 14436 + }, + { + "epoch": 0.14437, + "grad_norm": 1.121391958346411, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 14437 + }, + { + "epoch": 0.14438, + "grad_norm": 1.0197830285112484, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 14438 + }, + { + "epoch": 0.14439, + "grad_norm": 1.0587826211673452, + "learning_rate": 0.003, + "loss": 4.1126, + "step": 14439 + }, + { + "epoch": 0.1444, + "grad_norm": 1.0665392622319978, + "learning_rate": 0.003, + "loss": 4.0903, + "step": 14440 + }, + { + "epoch": 0.14441, + "grad_norm": 0.9343764733744462, + "learning_rate": 0.003, + "loss": 4.1022, + "step": 14441 + }, + { + "epoch": 0.14442, + "grad_norm": 1.193838854773833, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 14442 + }, + { + "epoch": 0.14443, + "grad_norm": 1.3010468264658868, + "learning_rate": 0.003, + "loss": 4.056, + "step": 14443 + }, + { + "epoch": 0.14444, + "grad_norm": 1.0050046532649979, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 14444 + }, + { + "epoch": 0.14445, + "grad_norm": 1.0875669975242583, + "learning_rate": 0.003, + "loss": 4.088, + "step": 14445 + }, + { + "epoch": 0.14446, + "grad_norm": 1.141378995945983, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 14446 + }, + { + "epoch": 0.14447, + "grad_norm": 0.8655669957567669, + "learning_rate": 0.003, + "loss": 4.0979, + "step": 14447 + }, + { + "epoch": 0.14448, + "grad_norm": 0.9754449939808217, + "learning_rate": 0.003, + "loss": 4.083, + "step": 14448 + }, + { + "epoch": 0.14449, + "grad_norm": 1.1330434656476305, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 14449 + }, + { + "epoch": 0.1445, + "grad_norm": 0.9137290154781871, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 14450 + }, + { + "epoch": 0.14451, + "grad_norm": 0.868378276922839, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 14451 + }, + { + "epoch": 0.14452, + "grad_norm": 0.908327332845315, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 14452 + }, + { + "epoch": 0.14453, + "grad_norm": 1.0208233274997625, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 14453 + }, + { + "epoch": 0.14454, + "grad_norm": 0.9709462994471744, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 14454 + }, + { + "epoch": 0.14455, + "grad_norm": 1.056811973119894, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 14455 + }, + { + "epoch": 0.14456, + "grad_norm": 0.9583648819909224, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 14456 + }, + { + "epoch": 0.14457, + "grad_norm": 0.9727061552735528, + "learning_rate": 0.003, + "loss": 4.0819, + "step": 14457 + }, + { + "epoch": 0.14458, + "grad_norm": 0.910593109735548, + "learning_rate": 0.003, + "loss": 4.063, + "step": 14458 + }, + { + "epoch": 0.14459, + "grad_norm": 0.9435032356881393, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 14459 + }, + { + "epoch": 0.1446, + "grad_norm": 0.910756431489279, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 14460 + }, + { + "epoch": 0.14461, + "grad_norm": 1.1662666343428796, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 14461 + }, + { + "epoch": 0.14462, + "grad_norm": 1.1709226210858563, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 14462 + }, + { + "epoch": 0.14463, + "grad_norm": 0.9114083786061561, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 14463 + }, + { + "epoch": 0.14464, + "grad_norm": 0.7557572019867024, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 14464 + }, + { + "epoch": 0.14465, + "grad_norm": 0.7357173606076157, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 14465 + }, + { + "epoch": 0.14466, + "grad_norm": 0.8879875914265742, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 14466 + }, + { + "epoch": 0.14467, + "grad_norm": 1.0596354040490559, + "learning_rate": 0.003, + "loss": 4.1181, + "step": 14467 + }, + { + "epoch": 0.14468, + "grad_norm": 0.9844245861296533, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 14468 + }, + { + "epoch": 0.14469, + "grad_norm": 1.1722470499658328, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 14469 + }, + { + "epoch": 0.1447, + "grad_norm": 0.8234904425433229, + "learning_rate": 0.003, + "loss": 4.0941, + "step": 14470 + }, + { + "epoch": 0.14471, + "grad_norm": 0.7091830192223889, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 14471 + }, + { + "epoch": 0.14472, + "grad_norm": 0.7424541560635294, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 14472 + }, + { + "epoch": 0.14473, + "grad_norm": 0.804622340616879, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 14473 + }, + { + "epoch": 0.14474, + "grad_norm": 0.8376453738872336, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 14474 + }, + { + "epoch": 0.14475, + "grad_norm": 0.7856820275838281, + "learning_rate": 0.003, + "loss": 4.0954, + "step": 14475 + }, + { + "epoch": 0.14476, + "grad_norm": 0.9621708791619894, + "learning_rate": 0.003, + "loss": 4.0966, + "step": 14476 + }, + { + "epoch": 0.14477, + "grad_norm": 1.3030634311556089, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 14477 + }, + { + "epoch": 0.14478, + "grad_norm": 0.6879507764912074, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 14478 + }, + { + "epoch": 0.14479, + "grad_norm": 0.6446377742248782, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 14479 + }, + { + "epoch": 0.1448, + "grad_norm": 0.5834001780166503, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 14480 + }, + { + "epoch": 0.14481, + "grad_norm": 0.7204075593850434, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 14481 + }, + { + "epoch": 0.14482, + "grad_norm": 0.8524031916788574, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 14482 + }, + { + "epoch": 0.14483, + "grad_norm": 1.0741397728217776, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 14483 + }, + { + "epoch": 0.14484, + "grad_norm": 1.0615120268106406, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 14484 + }, + { + "epoch": 0.14485, + "grad_norm": 1.2219343052124532, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 14485 + }, + { + "epoch": 0.14486, + "grad_norm": 1.0500259659638063, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 14486 + }, + { + "epoch": 0.14487, + "grad_norm": 1.0346237245948693, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 14487 + }, + { + "epoch": 0.14488, + "grad_norm": 0.9624753427758614, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 14488 + }, + { + "epoch": 0.14489, + "grad_norm": 0.9752289404154442, + "learning_rate": 0.003, + "loss": 4.067, + "step": 14489 + }, + { + "epoch": 0.1449, + "grad_norm": 1.063074867369594, + "learning_rate": 0.003, + "loss": 4.063, + "step": 14490 + }, + { + "epoch": 0.14491, + "grad_norm": 0.9821325144537693, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 14491 + }, + { + "epoch": 0.14492, + "grad_norm": 1.0328800731377976, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 14492 + }, + { + "epoch": 0.14493, + "grad_norm": 1.1052143165665387, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 14493 + }, + { + "epoch": 0.14494, + "grad_norm": 0.913595222602332, + "learning_rate": 0.003, + "loss": 4.0942, + "step": 14494 + }, + { + "epoch": 0.14495, + "grad_norm": 0.8958824243030956, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 14495 + }, + { + "epoch": 0.14496, + "grad_norm": 1.0927799519216035, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 14496 + }, + { + "epoch": 0.14497, + "grad_norm": 1.2542392457363059, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 14497 + }, + { + "epoch": 0.14498, + "grad_norm": 1.2421829279350025, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 14498 + }, + { + "epoch": 0.14499, + "grad_norm": 0.9276349731207228, + "learning_rate": 0.003, + "loss": 4.0908, + "step": 14499 + }, + { + "epoch": 0.145, + "grad_norm": 0.9693525076569746, + "learning_rate": 0.003, + "loss": 4.1117, + "step": 14500 + }, + { + "epoch": 0.14501, + "grad_norm": 0.9026620126474479, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 14501 + }, + { + "epoch": 0.14502, + "grad_norm": 0.8620125483903323, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 14502 + }, + { + "epoch": 0.14503, + "grad_norm": 0.85214811435518, + "learning_rate": 0.003, + "loss": 4.101, + "step": 14503 + }, + { + "epoch": 0.14504, + "grad_norm": 0.8820974634867096, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 14504 + }, + { + "epoch": 0.14505, + "grad_norm": 1.0103638901948166, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 14505 + }, + { + "epoch": 0.14506, + "grad_norm": 1.3039913917444566, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 14506 + }, + { + "epoch": 0.14507, + "grad_norm": 1.0288844552559637, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 14507 + }, + { + "epoch": 0.14508, + "grad_norm": 1.0245571314904105, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 14508 + }, + { + "epoch": 0.14509, + "grad_norm": 0.8860540136510928, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 14509 + }, + { + "epoch": 0.1451, + "grad_norm": 0.8639789696294304, + "learning_rate": 0.003, + "loss": 4.091, + "step": 14510 + }, + { + "epoch": 0.14511, + "grad_norm": 0.8741445845591167, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 14511 + }, + { + "epoch": 0.14512, + "grad_norm": 0.8773429109131458, + "learning_rate": 0.003, + "loss": 4.1163, + "step": 14512 + }, + { + "epoch": 0.14513, + "grad_norm": 0.9560432761905198, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 14513 + }, + { + "epoch": 0.14514, + "grad_norm": 0.9388316041908069, + "learning_rate": 0.003, + "loss": 4.096, + "step": 14514 + }, + { + "epoch": 0.14515, + "grad_norm": 0.9392244302371254, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 14515 + }, + { + "epoch": 0.14516, + "grad_norm": 0.8764603701874505, + "learning_rate": 0.003, + "loss": 4.0953, + "step": 14516 + }, + { + "epoch": 0.14517, + "grad_norm": 0.9027845710806471, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 14517 + }, + { + "epoch": 0.14518, + "grad_norm": 0.8583447146519886, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 14518 + }, + { + "epoch": 0.14519, + "grad_norm": 1.0216938687962345, + "learning_rate": 0.003, + "loss": 4.0918, + "step": 14519 + }, + { + "epoch": 0.1452, + "grad_norm": 1.09350788746558, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 14520 + }, + { + "epoch": 0.14521, + "grad_norm": 1.004356544572544, + "learning_rate": 0.003, + "loss": 4.079, + "step": 14521 + }, + { + "epoch": 0.14522, + "grad_norm": 1.1668560449192569, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 14522 + }, + { + "epoch": 0.14523, + "grad_norm": 1.1113420265750862, + "learning_rate": 0.003, + "loss": 4.0819, + "step": 14523 + }, + { + "epoch": 0.14524, + "grad_norm": 1.2809516934036884, + "learning_rate": 0.003, + "loss": 4.1086, + "step": 14524 + }, + { + "epoch": 0.14525, + "grad_norm": 0.7493415363343509, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 14525 + }, + { + "epoch": 0.14526, + "grad_norm": 0.7946633011443707, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 14526 + }, + { + "epoch": 0.14527, + "grad_norm": 1.2026551429341554, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 14527 + }, + { + "epoch": 0.14528, + "grad_norm": 1.1496554776376688, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 14528 + }, + { + "epoch": 0.14529, + "grad_norm": 1.08802058451073, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 14529 + }, + { + "epoch": 0.1453, + "grad_norm": 0.8556671647142027, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 14530 + }, + { + "epoch": 0.14531, + "grad_norm": 0.6982956159825934, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 14531 + }, + { + "epoch": 0.14532, + "grad_norm": 0.7370895744101993, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 14532 + }, + { + "epoch": 0.14533, + "grad_norm": 0.7542280461750118, + "learning_rate": 0.003, + "loss": 4.0917, + "step": 14533 + }, + { + "epoch": 0.14534, + "grad_norm": 0.8782329235331938, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 14534 + }, + { + "epoch": 0.14535, + "grad_norm": 1.1163279225219416, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 14535 + }, + { + "epoch": 0.14536, + "grad_norm": 1.1205994070347056, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 14536 + }, + { + "epoch": 0.14537, + "grad_norm": 1.0967747848668903, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 14537 + }, + { + "epoch": 0.14538, + "grad_norm": 0.9843171988455912, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 14538 + }, + { + "epoch": 0.14539, + "grad_norm": 1.026182691549459, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 14539 + }, + { + "epoch": 0.1454, + "grad_norm": 1.076448347730221, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 14540 + }, + { + "epoch": 0.14541, + "grad_norm": 0.9880631716718905, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 14541 + }, + { + "epoch": 0.14542, + "grad_norm": 1.0203854966794206, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 14542 + }, + { + "epoch": 0.14543, + "grad_norm": 0.9468378800155504, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 14543 + }, + { + "epoch": 0.14544, + "grad_norm": 0.8992110818164734, + "learning_rate": 0.003, + "loss": 4.0954, + "step": 14544 + }, + { + "epoch": 0.14545, + "grad_norm": 0.8034039069303637, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 14545 + }, + { + "epoch": 0.14546, + "grad_norm": 0.7679564331354067, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 14546 + }, + { + "epoch": 0.14547, + "grad_norm": 0.8340814180158435, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 14547 + }, + { + "epoch": 0.14548, + "grad_norm": 0.8441751662079061, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 14548 + }, + { + "epoch": 0.14549, + "grad_norm": 0.972745431121555, + "learning_rate": 0.003, + "loss": 4.0872, + "step": 14549 + }, + { + "epoch": 0.1455, + "grad_norm": 1.101808285271303, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 14550 + }, + { + "epoch": 0.14551, + "grad_norm": 1.0942359193173314, + "learning_rate": 0.003, + "loss": 4.077, + "step": 14551 + }, + { + "epoch": 0.14552, + "grad_norm": 0.8736882745635742, + "learning_rate": 0.003, + "loss": 4.095, + "step": 14552 + }, + { + "epoch": 0.14553, + "grad_norm": 0.8047923424479686, + "learning_rate": 0.003, + "loss": 4.089, + "step": 14553 + }, + { + "epoch": 0.14554, + "grad_norm": 0.8565758836759252, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 14554 + }, + { + "epoch": 0.14555, + "grad_norm": 0.9376816946036598, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 14555 + }, + { + "epoch": 0.14556, + "grad_norm": 1.0575057465233872, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 14556 + }, + { + "epoch": 0.14557, + "grad_norm": 1.2039242090937605, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 14557 + }, + { + "epoch": 0.14558, + "grad_norm": 0.9188800459556062, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 14558 + }, + { + "epoch": 0.14559, + "grad_norm": 0.831463054954382, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 14559 + }, + { + "epoch": 0.1456, + "grad_norm": 0.8085969780046193, + "learning_rate": 0.003, + "loss": 4.053, + "step": 14560 + }, + { + "epoch": 0.14561, + "grad_norm": 0.919754374628227, + "learning_rate": 0.003, + "loss": 4.071, + "step": 14561 + }, + { + "epoch": 0.14562, + "grad_norm": 0.980698998644735, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 14562 + }, + { + "epoch": 0.14563, + "grad_norm": 0.9008858791100862, + "learning_rate": 0.003, + "loss": 4.0901, + "step": 14563 + }, + { + "epoch": 0.14564, + "grad_norm": 0.8459256635252463, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 14564 + }, + { + "epoch": 0.14565, + "grad_norm": 0.9864092529802055, + "learning_rate": 0.003, + "loss": 4.0958, + "step": 14565 + }, + { + "epoch": 0.14566, + "grad_norm": 1.2346829759094111, + "learning_rate": 0.003, + "loss": 4.1073, + "step": 14566 + }, + { + "epoch": 0.14567, + "grad_norm": 1.0403457261715918, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 14567 + }, + { + "epoch": 0.14568, + "grad_norm": 1.1617243669733246, + "learning_rate": 0.003, + "loss": 4.0893, + "step": 14568 + }, + { + "epoch": 0.14569, + "grad_norm": 1.1232725133725079, + "learning_rate": 0.003, + "loss": 4.1087, + "step": 14569 + }, + { + "epoch": 0.1457, + "grad_norm": 1.032553255739852, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 14570 + }, + { + "epoch": 0.14571, + "grad_norm": 0.9123638597725476, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 14571 + }, + { + "epoch": 0.14572, + "grad_norm": 0.8866081073337518, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 14572 + }, + { + "epoch": 0.14573, + "grad_norm": 0.8559352262033345, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 14573 + }, + { + "epoch": 0.14574, + "grad_norm": 1.035312951781289, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 14574 + }, + { + "epoch": 0.14575, + "grad_norm": 1.149263873767304, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 14575 + }, + { + "epoch": 0.14576, + "grad_norm": 0.8812337019596362, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 14576 + }, + { + "epoch": 0.14577, + "grad_norm": 0.9105820725782182, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 14577 + }, + { + "epoch": 0.14578, + "grad_norm": 1.07018039199129, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 14578 + }, + { + "epoch": 0.14579, + "grad_norm": 0.8526583210330224, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 14579 + }, + { + "epoch": 0.1458, + "grad_norm": 1.014951175331579, + "learning_rate": 0.003, + "loss": 4.08, + "step": 14580 + }, + { + "epoch": 0.14581, + "grad_norm": 1.0394920710146827, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 14581 + }, + { + "epoch": 0.14582, + "grad_norm": 0.9819538936220432, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 14582 + }, + { + "epoch": 0.14583, + "grad_norm": 0.9480654402279213, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 14583 + }, + { + "epoch": 0.14584, + "grad_norm": 0.8352300764810974, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 14584 + }, + { + "epoch": 0.14585, + "grad_norm": 0.9184417043908921, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 14585 + }, + { + "epoch": 0.14586, + "grad_norm": 0.9888622419462517, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 14586 + }, + { + "epoch": 0.14587, + "grad_norm": 0.9874400444819094, + "learning_rate": 0.003, + "loss": 4.08, + "step": 14587 + }, + { + "epoch": 0.14588, + "grad_norm": 1.0973093096088142, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 14588 + }, + { + "epoch": 0.14589, + "grad_norm": 0.995480096398956, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 14589 + }, + { + "epoch": 0.1459, + "grad_norm": 1.100406300860888, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 14590 + }, + { + "epoch": 0.14591, + "grad_norm": 0.972385888526875, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 14591 + }, + { + "epoch": 0.14592, + "grad_norm": 1.027260843594216, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 14592 + }, + { + "epoch": 0.14593, + "grad_norm": 1.17019814375396, + "learning_rate": 0.003, + "loss": 4.1021, + "step": 14593 + }, + { + "epoch": 0.14594, + "grad_norm": 1.0389026431278598, + "learning_rate": 0.003, + "loss": 4.079, + "step": 14594 + }, + { + "epoch": 0.14595, + "grad_norm": 0.9458497177886221, + "learning_rate": 0.003, + "loss": 4.078, + "step": 14595 + }, + { + "epoch": 0.14596, + "grad_norm": 0.9266129729291167, + "learning_rate": 0.003, + "loss": 4.0819, + "step": 14596 + }, + { + "epoch": 0.14597, + "grad_norm": 1.1766060316390035, + "learning_rate": 0.003, + "loss": 4.1191, + "step": 14597 + }, + { + "epoch": 0.14598, + "grad_norm": 1.1152569070705782, + "learning_rate": 0.003, + "loss": 4.1283, + "step": 14598 + }, + { + "epoch": 0.14599, + "grad_norm": 1.1361282507184984, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 14599 + }, + { + "epoch": 0.146, + "grad_norm": 0.9133118909786805, + "learning_rate": 0.003, + "loss": 4.0933, + "step": 14600 + }, + { + "epoch": 0.14601, + "grad_norm": 0.946739595465057, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 14601 + }, + { + "epoch": 0.14602, + "grad_norm": 1.0133785890370177, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 14602 + }, + { + "epoch": 0.14603, + "grad_norm": 1.1451176362940498, + "learning_rate": 0.003, + "loss": 4.1036, + "step": 14603 + }, + { + "epoch": 0.14604, + "grad_norm": 0.9093332265121042, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 14604 + }, + { + "epoch": 0.14605, + "grad_norm": 1.0441648101688927, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 14605 + }, + { + "epoch": 0.14606, + "grad_norm": 1.079347330754128, + "learning_rate": 0.003, + "loss": 4.073, + "step": 14606 + }, + { + "epoch": 0.14607, + "grad_norm": 0.7801921147160009, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 14607 + }, + { + "epoch": 0.14608, + "grad_norm": 0.7686272733453338, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 14608 + }, + { + "epoch": 0.14609, + "grad_norm": 0.7556228201861274, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 14609 + }, + { + "epoch": 0.1461, + "grad_norm": 0.7872068735556745, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 14610 + }, + { + "epoch": 0.14611, + "grad_norm": 0.9129548567436679, + "learning_rate": 0.003, + "loss": 4.0937, + "step": 14611 + }, + { + "epoch": 0.14612, + "grad_norm": 1.027243526724431, + "learning_rate": 0.003, + "loss": 4.0902, + "step": 14612 + }, + { + "epoch": 0.14613, + "grad_norm": 0.9272973774051629, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 14613 + }, + { + "epoch": 0.14614, + "grad_norm": 0.9647586400476557, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 14614 + }, + { + "epoch": 0.14615, + "grad_norm": 1.1164402753816132, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 14615 + }, + { + "epoch": 0.14616, + "grad_norm": 1.0416308498025988, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 14616 + }, + { + "epoch": 0.14617, + "grad_norm": 1.1383560031666815, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 14617 + }, + { + "epoch": 0.14618, + "grad_norm": 1.0285469089577781, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 14618 + }, + { + "epoch": 0.14619, + "grad_norm": 1.049432105508515, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 14619 + }, + { + "epoch": 0.1462, + "grad_norm": 1.2051569037813643, + "learning_rate": 0.003, + "loss": 4.0949, + "step": 14620 + }, + { + "epoch": 0.14621, + "grad_norm": 1.0655131311373658, + "learning_rate": 0.003, + "loss": 4.1056, + "step": 14621 + }, + { + "epoch": 0.14622, + "grad_norm": 0.8500912462058322, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 14622 + }, + { + "epoch": 0.14623, + "grad_norm": 0.768430453139792, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 14623 + }, + { + "epoch": 0.14624, + "grad_norm": 0.8048371916620336, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 14624 + }, + { + "epoch": 0.14625, + "grad_norm": 0.7730268506792852, + "learning_rate": 0.003, + "loss": 4.0874, + "step": 14625 + }, + { + "epoch": 0.14626, + "grad_norm": 0.7974561633584192, + "learning_rate": 0.003, + "loss": 4.0903, + "step": 14626 + }, + { + "epoch": 0.14627, + "grad_norm": 0.8684715841985146, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 14627 + }, + { + "epoch": 0.14628, + "grad_norm": 0.8889739823146665, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 14628 + }, + { + "epoch": 0.14629, + "grad_norm": 0.8996365797996472, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 14629 + }, + { + "epoch": 0.1463, + "grad_norm": 0.9939673338505665, + "learning_rate": 0.003, + "loss": 4.1075, + "step": 14630 + }, + { + "epoch": 0.14631, + "grad_norm": 1.0490476681917833, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 14631 + }, + { + "epoch": 0.14632, + "grad_norm": 0.7403179983966585, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 14632 + }, + { + "epoch": 0.14633, + "grad_norm": 0.8231757039389991, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 14633 + }, + { + "epoch": 0.14634, + "grad_norm": 0.8668777171006161, + "learning_rate": 0.003, + "loss": 4.092, + "step": 14634 + }, + { + "epoch": 0.14635, + "grad_norm": 1.0574249025448612, + "learning_rate": 0.003, + "loss": 4.0912, + "step": 14635 + }, + { + "epoch": 0.14636, + "grad_norm": 1.4346487697424002, + "learning_rate": 0.003, + "loss": 4.0819, + "step": 14636 + }, + { + "epoch": 0.14637, + "grad_norm": 0.9194459685297682, + "learning_rate": 0.003, + "loss": 4.1017, + "step": 14637 + }, + { + "epoch": 0.14638, + "grad_norm": 0.9152727733837983, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 14638 + }, + { + "epoch": 0.14639, + "grad_norm": 0.9999676422215226, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 14639 + }, + { + "epoch": 0.1464, + "grad_norm": 1.4323446479550006, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 14640 + }, + { + "epoch": 0.14641, + "grad_norm": 0.9507600627354679, + "learning_rate": 0.003, + "loss": 4.089, + "step": 14641 + }, + { + "epoch": 0.14642, + "grad_norm": 1.009965243285751, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 14642 + }, + { + "epoch": 0.14643, + "grad_norm": 1.1353894920690872, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 14643 + }, + { + "epoch": 0.14644, + "grad_norm": 0.8052739990844088, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 14644 + }, + { + "epoch": 0.14645, + "grad_norm": 0.7551030263055885, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 14645 + }, + { + "epoch": 0.14646, + "grad_norm": 0.8040287109340135, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 14646 + }, + { + "epoch": 0.14647, + "grad_norm": 0.8566213638261437, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 14647 + }, + { + "epoch": 0.14648, + "grad_norm": 0.9491849621938803, + "learning_rate": 0.003, + "loss": 4.0873, + "step": 14648 + }, + { + "epoch": 0.14649, + "grad_norm": 1.1322974068420595, + "learning_rate": 0.003, + "loss": 4.0864, + "step": 14649 + }, + { + "epoch": 0.1465, + "grad_norm": 1.0481398170914695, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 14650 + }, + { + "epoch": 0.14651, + "grad_norm": 1.0931627084861275, + "learning_rate": 0.003, + "loss": 4.1051, + "step": 14651 + }, + { + "epoch": 0.14652, + "grad_norm": 0.9102401310999072, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 14652 + }, + { + "epoch": 0.14653, + "grad_norm": 0.8751975289370232, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 14653 + }, + { + "epoch": 0.14654, + "grad_norm": 0.9727555028030958, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 14654 + }, + { + "epoch": 0.14655, + "grad_norm": 1.1904553960170545, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 14655 + }, + { + "epoch": 0.14656, + "grad_norm": 0.9260854462681297, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 14656 + }, + { + "epoch": 0.14657, + "grad_norm": 0.9137080294704051, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 14657 + }, + { + "epoch": 0.14658, + "grad_norm": 0.9860494766158939, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 14658 + }, + { + "epoch": 0.14659, + "grad_norm": 1.2341609448928417, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 14659 + }, + { + "epoch": 0.1466, + "grad_norm": 0.8759899269957326, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 14660 + }, + { + "epoch": 0.14661, + "grad_norm": 0.874385178180722, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 14661 + }, + { + "epoch": 0.14662, + "grad_norm": 0.864304810979198, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 14662 + }, + { + "epoch": 0.14663, + "grad_norm": 0.9153340156306464, + "learning_rate": 0.003, + "loss": 4.0948, + "step": 14663 + }, + { + "epoch": 0.14664, + "grad_norm": 0.9608979675249911, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 14664 + }, + { + "epoch": 0.14665, + "grad_norm": 1.117834232418335, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 14665 + }, + { + "epoch": 0.14666, + "grad_norm": 1.058274079376159, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 14666 + }, + { + "epoch": 0.14667, + "grad_norm": 0.944743759929571, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 14667 + }, + { + "epoch": 0.14668, + "grad_norm": 0.8413310639454784, + "learning_rate": 0.003, + "loss": 4.0985, + "step": 14668 + }, + { + "epoch": 0.14669, + "grad_norm": 0.8989908570831676, + "learning_rate": 0.003, + "loss": 4.089, + "step": 14669 + }, + { + "epoch": 0.1467, + "grad_norm": 1.00957662611984, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 14670 + }, + { + "epoch": 0.14671, + "grad_norm": 1.0702640984141634, + "learning_rate": 0.003, + "loss": 4.0925, + "step": 14671 + }, + { + "epoch": 0.14672, + "grad_norm": 1.0083693630645487, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 14672 + }, + { + "epoch": 0.14673, + "grad_norm": 1.1388455067360217, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 14673 + }, + { + "epoch": 0.14674, + "grad_norm": 0.9797742546189854, + "learning_rate": 0.003, + "loss": 4.042, + "step": 14674 + }, + { + "epoch": 0.14675, + "grad_norm": 1.0976005906491766, + "learning_rate": 0.003, + "loss": 4.082, + "step": 14675 + }, + { + "epoch": 0.14676, + "grad_norm": 0.9586177656394848, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 14676 + }, + { + "epoch": 0.14677, + "grad_norm": 0.9402377279302978, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 14677 + }, + { + "epoch": 0.14678, + "grad_norm": 0.981540895285961, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 14678 + }, + { + "epoch": 0.14679, + "grad_norm": 1.031019032039138, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 14679 + }, + { + "epoch": 0.1468, + "grad_norm": 1.0829915457497756, + "learning_rate": 0.003, + "loss": 4.1044, + "step": 14680 + }, + { + "epoch": 0.14681, + "grad_norm": 1.125898717819587, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 14681 + }, + { + "epoch": 0.14682, + "grad_norm": 0.9550835912279904, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 14682 + }, + { + "epoch": 0.14683, + "grad_norm": 1.0004874135798187, + "learning_rate": 0.003, + "loss": 4.072, + "step": 14683 + }, + { + "epoch": 0.14684, + "grad_norm": 1.040383429680946, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 14684 + }, + { + "epoch": 0.14685, + "grad_norm": 0.8783284118302115, + "learning_rate": 0.003, + "loss": 4.0899, + "step": 14685 + }, + { + "epoch": 0.14686, + "grad_norm": 0.9588702498609908, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 14686 + }, + { + "epoch": 0.14687, + "grad_norm": 1.0749036999397628, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 14687 + }, + { + "epoch": 0.14688, + "grad_norm": 1.3805598993926453, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 14688 + }, + { + "epoch": 0.14689, + "grad_norm": 0.9580767852560518, + "learning_rate": 0.003, + "loss": 4.0881, + "step": 14689 + }, + { + "epoch": 0.1469, + "grad_norm": 0.8041465815825737, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 14690 + }, + { + "epoch": 0.14691, + "grad_norm": 0.6680531112834734, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 14691 + }, + { + "epoch": 0.14692, + "grad_norm": 0.702363403542044, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 14692 + }, + { + "epoch": 0.14693, + "grad_norm": 0.7188724627378696, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 14693 + }, + { + "epoch": 0.14694, + "grad_norm": 0.7810087892976746, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 14694 + }, + { + "epoch": 0.14695, + "grad_norm": 0.8583574173821925, + "learning_rate": 0.003, + "loss": 4.0962, + "step": 14695 + }, + { + "epoch": 0.14696, + "grad_norm": 0.8308532627104515, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 14696 + }, + { + "epoch": 0.14697, + "grad_norm": 0.9016521980102125, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 14697 + }, + { + "epoch": 0.14698, + "grad_norm": 0.9561244329026802, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 14698 + }, + { + "epoch": 0.14699, + "grad_norm": 1.301349981695908, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 14699 + }, + { + "epoch": 0.147, + "grad_norm": 1.0534153279771619, + "learning_rate": 0.003, + "loss": 4.0845, + "step": 14700 + }, + { + "epoch": 0.14701, + "grad_norm": 1.0614702388049975, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 14701 + }, + { + "epoch": 0.14702, + "grad_norm": 0.8450551875710314, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 14702 + }, + { + "epoch": 0.14703, + "grad_norm": 0.877548328024781, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 14703 + }, + { + "epoch": 0.14704, + "grad_norm": 0.8553921727700116, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 14704 + }, + { + "epoch": 0.14705, + "grad_norm": 0.906094132540581, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 14705 + }, + { + "epoch": 0.14706, + "grad_norm": 0.9813444171318209, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 14706 + }, + { + "epoch": 0.14707, + "grad_norm": 1.0452603708317547, + "learning_rate": 0.003, + "loss": 4.0874, + "step": 14707 + }, + { + "epoch": 0.14708, + "grad_norm": 0.8031693017004379, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 14708 + }, + { + "epoch": 0.14709, + "grad_norm": 0.7920464767752385, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 14709 + }, + { + "epoch": 0.1471, + "grad_norm": 1.0615540136469022, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 14710 + }, + { + "epoch": 0.14711, + "grad_norm": 1.496665049853003, + "learning_rate": 0.003, + "loss": 4.0901, + "step": 14711 + }, + { + "epoch": 0.14712, + "grad_norm": 0.6485456232399575, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 14712 + }, + { + "epoch": 0.14713, + "grad_norm": 0.898608380057148, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 14713 + }, + { + "epoch": 0.14714, + "grad_norm": 1.3771694993172834, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 14714 + }, + { + "epoch": 0.14715, + "grad_norm": 0.7904822286991244, + "learning_rate": 0.003, + "loss": 4.0934, + "step": 14715 + }, + { + "epoch": 0.14716, + "grad_norm": 0.7187052442871521, + "learning_rate": 0.003, + "loss": 4.0971, + "step": 14716 + }, + { + "epoch": 0.14717, + "grad_norm": 0.7764145120626375, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 14717 + }, + { + "epoch": 0.14718, + "grad_norm": 0.9763947464872922, + "learning_rate": 0.003, + "loss": 4.042, + "step": 14718 + }, + { + "epoch": 0.14719, + "grad_norm": 1.2364068909314294, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 14719 + }, + { + "epoch": 0.1472, + "grad_norm": 1.013205143688417, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 14720 + }, + { + "epoch": 0.14721, + "grad_norm": 1.0681169599751075, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 14721 + }, + { + "epoch": 0.14722, + "grad_norm": 0.9113413264060484, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 14722 + }, + { + "epoch": 0.14723, + "grad_norm": 0.8234637676693246, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 14723 + }, + { + "epoch": 0.14724, + "grad_norm": 0.8010711409876051, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 14724 + }, + { + "epoch": 0.14725, + "grad_norm": 0.8891501989399467, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 14725 + }, + { + "epoch": 0.14726, + "grad_norm": 0.906817740376405, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 14726 + }, + { + "epoch": 0.14727, + "grad_norm": 0.9406268619453289, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 14727 + }, + { + "epoch": 0.14728, + "grad_norm": 1.09379949045309, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 14728 + }, + { + "epoch": 0.14729, + "grad_norm": 1.1476609967560736, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 14729 + }, + { + "epoch": 0.1473, + "grad_norm": 1.098505649768357, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 14730 + }, + { + "epoch": 0.14731, + "grad_norm": 0.9040520027949577, + "learning_rate": 0.003, + "loss": 4.0893, + "step": 14731 + }, + { + "epoch": 0.14732, + "grad_norm": 1.0318052888534293, + "learning_rate": 0.003, + "loss": 4.0943, + "step": 14732 + }, + { + "epoch": 0.14733, + "grad_norm": 1.165706601911814, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 14733 + }, + { + "epoch": 0.14734, + "grad_norm": 1.063629158692407, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 14734 + }, + { + "epoch": 0.14735, + "grad_norm": 0.9208056645126291, + "learning_rate": 0.003, + "loss": 4.0855, + "step": 14735 + }, + { + "epoch": 0.14736, + "grad_norm": 0.9928269369601324, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 14736 + }, + { + "epoch": 0.14737, + "grad_norm": 1.1258576271440117, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 14737 + }, + { + "epoch": 0.14738, + "grad_norm": 1.0330810287257335, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 14738 + }, + { + "epoch": 0.14739, + "grad_norm": 1.0173917406238742, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 14739 + }, + { + "epoch": 0.1474, + "grad_norm": 1.018191907756927, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 14740 + }, + { + "epoch": 0.14741, + "grad_norm": 1.0088195108520883, + "learning_rate": 0.003, + "loss": 4.059, + "step": 14741 + }, + { + "epoch": 0.14742, + "grad_norm": 0.9629240041943994, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 14742 + }, + { + "epoch": 0.14743, + "grad_norm": 0.8761297429782801, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 14743 + }, + { + "epoch": 0.14744, + "grad_norm": 0.8813633835484943, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 14744 + }, + { + "epoch": 0.14745, + "grad_norm": 1.0609705073265316, + "learning_rate": 0.003, + "loss": 4.0943, + "step": 14745 + }, + { + "epoch": 0.14746, + "grad_norm": 1.0470211415706874, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 14746 + }, + { + "epoch": 0.14747, + "grad_norm": 0.9562757338367955, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 14747 + }, + { + "epoch": 0.14748, + "grad_norm": 1.0076900759884062, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 14748 + }, + { + "epoch": 0.14749, + "grad_norm": 1.083713345914116, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 14749 + }, + { + "epoch": 0.1475, + "grad_norm": 0.9917525387442542, + "learning_rate": 0.003, + "loss": 4.0872, + "step": 14750 + }, + { + "epoch": 0.14751, + "grad_norm": 1.1768182063883863, + "learning_rate": 0.003, + "loss": 4.061, + "step": 14751 + }, + { + "epoch": 0.14752, + "grad_norm": 0.8765029597927932, + "learning_rate": 0.003, + "loss": 4.1021, + "step": 14752 + }, + { + "epoch": 0.14753, + "grad_norm": 1.0013724793933005, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 14753 + }, + { + "epoch": 0.14754, + "grad_norm": 1.1302680934207117, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 14754 + }, + { + "epoch": 0.14755, + "grad_norm": 1.0830065895640413, + "learning_rate": 0.003, + "loss": 4.1077, + "step": 14755 + }, + { + "epoch": 0.14756, + "grad_norm": 1.1489362823119977, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 14756 + }, + { + "epoch": 0.14757, + "grad_norm": 0.956984926431225, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 14757 + }, + { + "epoch": 0.14758, + "grad_norm": 0.8676425918626922, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 14758 + }, + { + "epoch": 0.14759, + "grad_norm": 1.0477119297260997, + "learning_rate": 0.003, + "loss": 4.0908, + "step": 14759 + }, + { + "epoch": 0.1476, + "grad_norm": 0.9918224362431405, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 14760 + }, + { + "epoch": 0.14761, + "grad_norm": 1.1333930681066144, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 14761 + }, + { + "epoch": 0.14762, + "grad_norm": 0.9680167593193062, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 14762 + }, + { + "epoch": 0.14763, + "grad_norm": 1.1130987019665148, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 14763 + }, + { + "epoch": 0.14764, + "grad_norm": 0.9797484147128562, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 14764 + }, + { + "epoch": 0.14765, + "grad_norm": 0.9428857719427979, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 14765 + }, + { + "epoch": 0.14766, + "grad_norm": 0.9566357219800155, + "learning_rate": 0.003, + "loss": 4.0872, + "step": 14766 + }, + { + "epoch": 0.14767, + "grad_norm": 0.9615498387746793, + "learning_rate": 0.003, + "loss": 4.1099, + "step": 14767 + }, + { + "epoch": 0.14768, + "grad_norm": 0.9879744474822495, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 14768 + }, + { + "epoch": 0.14769, + "grad_norm": 1.0616455184772466, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 14769 + }, + { + "epoch": 0.1477, + "grad_norm": 1.3449785966380683, + "learning_rate": 0.003, + "loss": 4.1114, + "step": 14770 + }, + { + "epoch": 0.14771, + "grad_norm": 0.9672583608406822, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 14771 + }, + { + "epoch": 0.14772, + "grad_norm": 0.8593857441497724, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 14772 + }, + { + "epoch": 0.14773, + "grad_norm": 0.928830736300277, + "learning_rate": 0.003, + "loss": 4.0978, + "step": 14773 + }, + { + "epoch": 0.14774, + "grad_norm": 0.9420486092145699, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 14774 + }, + { + "epoch": 0.14775, + "grad_norm": 0.9069068365844426, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 14775 + }, + { + "epoch": 0.14776, + "grad_norm": 0.9869703971071639, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 14776 + }, + { + "epoch": 0.14777, + "grad_norm": 1.119031649502446, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 14777 + }, + { + "epoch": 0.14778, + "grad_norm": 0.8558686132260704, + "learning_rate": 0.003, + "loss": 4.0868, + "step": 14778 + }, + { + "epoch": 0.14779, + "grad_norm": 0.8099253156292258, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 14779 + }, + { + "epoch": 0.1478, + "grad_norm": 0.9290418240107634, + "learning_rate": 0.003, + "loss": 4.075, + "step": 14780 + }, + { + "epoch": 0.14781, + "grad_norm": 0.9925220570606771, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 14781 + }, + { + "epoch": 0.14782, + "grad_norm": 0.8858406019700646, + "learning_rate": 0.003, + "loss": 4.0982, + "step": 14782 + }, + { + "epoch": 0.14783, + "grad_norm": 0.9618912409064978, + "learning_rate": 0.003, + "loss": 4.0884, + "step": 14783 + }, + { + "epoch": 0.14784, + "grad_norm": 1.0293633288272688, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 14784 + }, + { + "epoch": 0.14785, + "grad_norm": 1.1816435836720276, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 14785 + }, + { + "epoch": 0.14786, + "grad_norm": 0.9621167876860154, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 14786 + }, + { + "epoch": 0.14787, + "grad_norm": 1.0695188498166237, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 14787 + }, + { + "epoch": 0.14788, + "grad_norm": 0.8957026693089948, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 14788 + }, + { + "epoch": 0.14789, + "grad_norm": 0.9188406828849088, + "learning_rate": 0.003, + "loss": 4.086, + "step": 14789 + }, + { + "epoch": 0.1479, + "grad_norm": 0.933633941615261, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 14790 + }, + { + "epoch": 0.14791, + "grad_norm": 1.1115477094919752, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 14791 + }, + { + "epoch": 0.14792, + "grad_norm": 0.92483766962112, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 14792 + }, + { + "epoch": 0.14793, + "grad_norm": 0.9478726588529544, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 14793 + }, + { + "epoch": 0.14794, + "grad_norm": 0.9757376297627024, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 14794 + }, + { + "epoch": 0.14795, + "grad_norm": 0.946300180852162, + "learning_rate": 0.003, + "loss": 4.1141, + "step": 14795 + }, + { + "epoch": 0.14796, + "grad_norm": 1.0957784226095915, + "learning_rate": 0.003, + "loss": 4.041, + "step": 14796 + }, + { + "epoch": 0.14797, + "grad_norm": 1.118405387846101, + "learning_rate": 0.003, + "loss": 4.081, + "step": 14797 + }, + { + "epoch": 0.14798, + "grad_norm": 0.9266617352007458, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 14798 + }, + { + "epoch": 0.14799, + "grad_norm": 0.8305779243379453, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 14799 + }, + { + "epoch": 0.148, + "grad_norm": 0.8536612594059925, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 14800 + }, + { + "epoch": 0.14801, + "grad_norm": 1.0726874223528344, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 14801 + }, + { + "epoch": 0.14802, + "grad_norm": 1.030212698218163, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 14802 + }, + { + "epoch": 0.14803, + "grad_norm": 0.8674880264398992, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 14803 + }, + { + "epoch": 0.14804, + "grad_norm": 0.7506200307329559, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 14804 + }, + { + "epoch": 0.14805, + "grad_norm": 0.7467308233990052, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 14805 + }, + { + "epoch": 0.14806, + "grad_norm": 0.8441701992241563, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 14806 + }, + { + "epoch": 0.14807, + "grad_norm": 1.0273540063268338, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 14807 + }, + { + "epoch": 0.14808, + "grad_norm": 1.2173271830941266, + "learning_rate": 0.003, + "loss": 4.0962, + "step": 14808 + }, + { + "epoch": 0.14809, + "grad_norm": 0.9366556285363833, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 14809 + }, + { + "epoch": 0.1481, + "grad_norm": 1.0396275545732043, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 14810 + }, + { + "epoch": 0.14811, + "grad_norm": 1.0372849173643646, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 14811 + }, + { + "epoch": 0.14812, + "grad_norm": 1.0281647744533158, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 14812 + }, + { + "epoch": 0.14813, + "grad_norm": 0.9407755871351523, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 14813 + }, + { + "epoch": 0.14814, + "grad_norm": 0.9318111946787239, + "learning_rate": 0.003, + "loss": 4.062, + "step": 14814 + }, + { + "epoch": 0.14815, + "grad_norm": 0.8325606519449019, + "learning_rate": 0.003, + "loss": 4.0981, + "step": 14815 + }, + { + "epoch": 0.14816, + "grad_norm": 0.8445187710164662, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 14816 + }, + { + "epoch": 0.14817, + "grad_norm": 0.8537571140503739, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 14817 + }, + { + "epoch": 0.14818, + "grad_norm": 0.93235886243547, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 14818 + }, + { + "epoch": 0.14819, + "grad_norm": 1.126438432950548, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 14819 + }, + { + "epoch": 0.1482, + "grad_norm": 0.9693993974737314, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 14820 + }, + { + "epoch": 0.14821, + "grad_norm": 0.9474458962069947, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 14821 + }, + { + "epoch": 0.14822, + "grad_norm": 1.0595030484534282, + "learning_rate": 0.003, + "loss": 4.037, + "step": 14822 + }, + { + "epoch": 0.14823, + "grad_norm": 1.1659180425140443, + "learning_rate": 0.003, + "loss": 4.051, + "step": 14823 + }, + { + "epoch": 0.14824, + "grad_norm": 0.9608310746854415, + "learning_rate": 0.003, + "loss": 4.0974, + "step": 14824 + }, + { + "epoch": 0.14825, + "grad_norm": 1.001168938759652, + "learning_rate": 0.003, + "loss": 4.067, + "step": 14825 + }, + { + "epoch": 0.14826, + "grad_norm": 1.0762352184434931, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 14826 + }, + { + "epoch": 0.14827, + "grad_norm": 0.8332488217923946, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 14827 + }, + { + "epoch": 0.14828, + "grad_norm": 0.7315636234085605, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 14828 + }, + { + "epoch": 0.14829, + "grad_norm": 0.8898036978033856, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 14829 + }, + { + "epoch": 0.1483, + "grad_norm": 1.0413483827709198, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 14830 + }, + { + "epoch": 0.14831, + "grad_norm": 1.1306410512922986, + "learning_rate": 0.003, + "loss": 4.0923, + "step": 14831 + }, + { + "epoch": 0.14832, + "grad_norm": 0.9558039056085215, + "learning_rate": 0.003, + "loss": 4.0907, + "step": 14832 + }, + { + "epoch": 0.14833, + "grad_norm": 1.0641490150479107, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 14833 + }, + { + "epoch": 0.14834, + "grad_norm": 1.0617012272106674, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 14834 + }, + { + "epoch": 0.14835, + "grad_norm": 0.9447693197678106, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 14835 + }, + { + "epoch": 0.14836, + "grad_norm": 0.8491557925302191, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 14836 + }, + { + "epoch": 0.14837, + "grad_norm": 0.8364139829222659, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 14837 + }, + { + "epoch": 0.14838, + "grad_norm": 0.9204018578090175, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 14838 + }, + { + "epoch": 0.14839, + "grad_norm": 0.8989030734758936, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 14839 + }, + { + "epoch": 0.1484, + "grad_norm": 1.0582913566229784, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 14840 + }, + { + "epoch": 0.14841, + "grad_norm": 1.227194339928111, + "learning_rate": 0.003, + "loss": 4.052, + "step": 14841 + }, + { + "epoch": 0.14842, + "grad_norm": 0.8041618616052012, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 14842 + }, + { + "epoch": 0.14843, + "grad_norm": 0.8074417658370009, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 14843 + }, + { + "epoch": 0.14844, + "grad_norm": 0.9395178300194551, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 14844 + }, + { + "epoch": 0.14845, + "grad_norm": 1.1859044941790913, + "learning_rate": 0.003, + "loss": 4.0855, + "step": 14845 + }, + { + "epoch": 0.14846, + "grad_norm": 1.07714933168914, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 14846 + }, + { + "epoch": 0.14847, + "grad_norm": 1.1379993530893004, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 14847 + }, + { + "epoch": 0.14848, + "grad_norm": 0.9067707599272397, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 14848 + }, + { + "epoch": 0.14849, + "grad_norm": 1.030852707944337, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 14849 + }, + { + "epoch": 0.1485, + "grad_norm": 1.2518289974383363, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 14850 + }, + { + "epoch": 0.14851, + "grad_norm": 0.8729004558698198, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 14851 + }, + { + "epoch": 0.14852, + "grad_norm": 0.7176457127164583, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 14852 + }, + { + "epoch": 0.14853, + "grad_norm": 0.8393915408490467, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 14853 + }, + { + "epoch": 0.14854, + "grad_norm": 0.9795126545453833, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 14854 + }, + { + "epoch": 0.14855, + "grad_norm": 1.0620283116154794, + "learning_rate": 0.003, + "loss": 4.0969, + "step": 14855 + }, + { + "epoch": 0.14856, + "grad_norm": 0.9719544610512548, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 14856 + }, + { + "epoch": 0.14857, + "grad_norm": 1.0933796604598358, + "learning_rate": 0.003, + "loss": 4.0937, + "step": 14857 + }, + { + "epoch": 0.14858, + "grad_norm": 0.9970133106340723, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 14858 + }, + { + "epoch": 0.14859, + "grad_norm": 0.8631917128730071, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 14859 + }, + { + "epoch": 0.1486, + "grad_norm": 0.8033160640209519, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 14860 + }, + { + "epoch": 0.14861, + "grad_norm": 0.8472293673406257, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 14861 + }, + { + "epoch": 0.14862, + "grad_norm": 0.8847932725025752, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 14862 + }, + { + "epoch": 0.14863, + "grad_norm": 0.9581873132256032, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 14863 + }, + { + "epoch": 0.14864, + "grad_norm": 0.8422242577203205, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 14864 + }, + { + "epoch": 0.14865, + "grad_norm": 0.9224463872495965, + "learning_rate": 0.003, + "loss": 4.0997, + "step": 14865 + }, + { + "epoch": 0.14866, + "grad_norm": 0.8462103663561696, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 14866 + }, + { + "epoch": 0.14867, + "grad_norm": 0.9066850019938976, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 14867 + }, + { + "epoch": 0.14868, + "grad_norm": 1.2307555587032057, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 14868 + }, + { + "epoch": 0.14869, + "grad_norm": 1.162318503890178, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 14869 + }, + { + "epoch": 0.1487, + "grad_norm": 0.9690019753776828, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 14870 + }, + { + "epoch": 0.14871, + "grad_norm": 1.031723463536081, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 14871 + }, + { + "epoch": 0.14872, + "grad_norm": 0.9898282367820462, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 14872 + }, + { + "epoch": 0.14873, + "grad_norm": 1.0422594275687842, + "learning_rate": 0.003, + "loss": 4.094, + "step": 14873 + }, + { + "epoch": 0.14874, + "grad_norm": 1.073081454003803, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 14874 + }, + { + "epoch": 0.14875, + "grad_norm": 0.8766575252838705, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 14875 + }, + { + "epoch": 0.14876, + "grad_norm": 0.8291487776319423, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 14876 + }, + { + "epoch": 0.14877, + "grad_norm": 0.8779437571984459, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 14877 + }, + { + "epoch": 0.14878, + "grad_norm": 0.9354809806807283, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 14878 + }, + { + "epoch": 0.14879, + "grad_norm": 1.0947738960548887, + "learning_rate": 0.003, + "loss": 4.0964, + "step": 14879 + }, + { + "epoch": 0.1488, + "grad_norm": 1.1992722448330988, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 14880 + }, + { + "epoch": 0.14881, + "grad_norm": 0.881872154776049, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 14881 + }, + { + "epoch": 0.14882, + "grad_norm": 0.9113240692293961, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 14882 + }, + { + "epoch": 0.14883, + "grad_norm": 1.003698145015082, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 14883 + }, + { + "epoch": 0.14884, + "grad_norm": 1.1864689300998574, + "learning_rate": 0.003, + "loss": 4.075, + "step": 14884 + }, + { + "epoch": 0.14885, + "grad_norm": 0.8810999875969996, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 14885 + }, + { + "epoch": 0.14886, + "grad_norm": 0.7933880419995626, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 14886 + }, + { + "epoch": 0.14887, + "grad_norm": 0.7445653553433971, + "learning_rate": 0.003, + "loss": 4.067, + "step": 14887 + }, + { + "epoch": 0.14888, + "grad_norm": 0.9069163822330671, + "learning_rate": 0.003, + "loss": 4.1066, + "step": 14888 + }, + { + "epoch": 0.14889, + "grad_norm": 0.9400485903951985, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 14889 + }, + { + "epoch": 0.1489, + "grad_norm": 0.8728909418128894, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 14890 + }, + { + "epoch": 0.14891, + "grad_norm": 1.0403765211143143, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 14891 + }, + { + "epoch": 0.14892, + "grad_norm": 1.218409220966335, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 14892 + }, + { + "epoch": 0.14893, + "grad_norm": 0.9521619247301835, + "learning_rate": 0.003, + "loss": 4.059, + "step": 14893 + }, + { + "epoch": 0.14894, + "grad_norm": 0.96385253419481, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 14894 + }, + { + "epoch": 0.14895, + "grad_norm": 1.0827216718016528, + "learning_rate": 0.003, + "loss": 4.0972, + "step": 14895 + }, + { + "epoch": 0.14896, + "grad_norm": 1.1054395263610406, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 14896 + }, + { + "epoch": 0.14897, + "grad_norm": 0.9301093162303585, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 14897 + }, + { + "epoch": 0.14898, + "grad_norm": 0.8466208618998217, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 14898 + }, + { + "epoch": 0.14899, + "grad_norm": 0.9152575587851234, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 14899 + }, + { + "epoch": 0.149, + "grad_norm": 0.9818973745890289, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 14900 + }, + { + "epoch": 0.14901, + "grad_norm": 1.0274464845490192, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 14901 + }, + { + "epoch": 0.14902, + "grad_norm": 1.2120290295991187, + "learning_rate": 0.003, + "loss": 4.0916, + "step": 14902 + }, + { + "epoch": 0.14903, + "grad_norm": 0.7466611427336209, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 14903 + }, + { + "epoch": 0.14904, + "grad_norm": 0.6683993512507248, + "learning_rate": 0.003, + "loss": 4.0912, + "step": 14904 + }, + { + "epoch": 0.14905, + "grad_norm": 0.7911871831096717, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 14905 + }, + { + "epoch": 0.14906, + "grad_norm": 0.9371615027931597, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 14906 + }, + { + "epoch": 0.14907, + "grad_norm": 1.3128508585214425, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 14907 + }, + { + "epoch": 0.14908, + "grad_norm": 0.6520824504568395, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 14908 + }, + { + "epoch": 0.14909, + "grad_norm": 0.7104747191959222, + "learning_rate": 0.003, + "loss": 4.068, + "step": 14909 + }, + { + "epoch": 0.1491, + "grad_norm": 0.8574574833937768, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 14910 + }, + { + "epoch": 0.14911, + "grad_norm": 0.9255308977473562, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 14911 + }, + { + "epoch": 0.14912, + "grad_norm": 0.9859243494968752, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 14912 + }, + { + "epoch": 0.14913, + "grad_norm": 1.1671118910786684, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 14913 + }, + { + "epoch": 0.14914, + "grad_norm": 0.8817409008080823, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 14914 + }, + { + "epoch": 0.14915, + "grad_norm": 0.9382739122289769, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 14915 + }, + { + "epoch": 0.14916, + "grad_norm": 1.028843930368404, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 14916 + }, + { + "epoch": 0.14917, + "grad_norm": 1.1668505101680564, + "learning_rate": 0.003, + "loss": 4.102, + "step": 14917 + }, + { + "epoch": 0.14918, + "grad_norm": 0.9742519657936379, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 14918 + }, + { + "epoch": 0.14919, + "grad_norm": 1.189992640204692, + "learning_rate": 0.003, + "loss": 4.1027, + "step": 14919 + }, + { + "epoch": 0.1492, + "grad_norm": 1.048193642761215, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 14920 + }, + { + "epoch": 0.14921, + "grad_norm": 1.2488330880926983, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 14921 + }, + { + "epoch": 0.14922, + "grad_norm": 0.90073184633448, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 14922 + }, + { + "epoch": 0.14923, + "grad_norm": 0.9255727125793031, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 14923 + }, + { + "epoch": 0.14924, + "grad_norm": 1.1428003582086084, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 14924 + }, + { + "epoch": 0.14925, + "grad_norm": 0.8436718913806494, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 14925 + }, + { + "epoch": 0.14926, + "grad_norm": 0.7390514869024759, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 14926 + }, + { + "epoch": 0.14927, + "grad_norm": 0.707249012796253, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 14927 + }, + { + "epoch": 0.14928, + "grad_norm": 0.6283210719000251, + "learning_rate": 0.003, + "loss": 4.0872, + "step": 14928 + }, + { + "epoch": 0.14929, + "grad_norm": 0.576298571851741, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 14929 + }, + { + "epoch": 0.1493, + "grad_norm": 0.5932547208683459, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 14930 + }, + { + "epoch": 0.14931, + "grad_norm": 0.6600729545943833, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 14931 + }, + { + "epoch": 0.14932, + "grad_norm": 0.8102008441993173, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 14932 + }, + { + "epoch": 0.14933, + "grad_norm": 0.9867614112820933, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 14933 + }, + { + "epoch": 0.14934, + "grad_norm": 1.1144135173647418, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 14934 + }, + { + "epoch": 0.14935, + "grad_norm": 1.038501064672105, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 14935 + }, + { + "epoch": 0.14936, + "grad_norm": 1.0925561335788347, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 14936 + }, + { + "epoch": 0.14937, + "grad_norm": 1.01988640715744, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 14937 + }, + { + "epoch": 0.14938, + "grad_norm": 0.9006562385873643, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 14938 + }, + { + "epoch": 0.14939, + "grad_norm": 1.1615576487779402, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 14939 + }, + { + "epoch": 0.1494, + "grad_norm": 1.1682781947623864, + "learning_rate": 0.003, + "loss": 4.073, + "step": 14940 + }, + { + "epoch": 0.14941, + "grad_norm": 1.1164583005568485, + "learning_rate": 0.003, + "loss": 4.1119, + "step": 14941 + }, + { + "epoch": 0.14942, + "grad_norm": 1.1750011905869642, + "learning_rate": 0.003, + "loss": 4.054, + "step": 14942 + }, + { + "epoch": 0.14943, + "grad_norm": 0.9841406167705045, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 14943 + }, + { + "epoch": 0.14944, + "grad_norm": 1.1164792841799636, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 14944 + }, + { + "epoch": 0.14945, + "grad_norm": 0.8804700471006347, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 14945 + }, + { + "epoch": 0.14946, + "grad_norm": 0.8598700999019337, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 14946 + }, + { + "epoch": 0.14947, + "grad_norm": 0.8295608862101799, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 14947 + }, + { + "epoch": 0.14948, + "grad_norm": 0.7815351620866495, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 14948 + }, + { + "epoch": 0.14949, + "grad_norm": 0.7868943153653798, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 14949 + }, + { + "epoch": 0.1495, + "grad_norm": 0.7768935546674706, + "learning_rate": 0.003, + "loss": 4.0848, + "step": 14950 + }, + { + "epoch": 0.14951, + "grad_norm": 0.8343353200517245, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 14951 + }, + { + "epoch": 0.14952, + "grad_norm": 0.9460518175566938, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 14952 + }, + { + "epoch": 0.14953, + "grad_norm": 1.1090106343418509, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 14953 + }, + { + "epoch": 0.14954, + "grad_norm": 1.1006551406343577, + "learning_rate": 0.003, + "loss": 4.1098, + "step": 14954 + }, + { + "epoch": 0.14955, + "grad_norm": 1.2529506735404645, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 14955 + }, + { + "epoch": 0.14956, + "grad_norm": 1.0140525144778822, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 14956 + }, + { + "epoch": 0.14957, + "grad_norm": 1.0921495207672811, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 14957 + }, + { + "epoch": 0.14958, + "grad_norm": 1.2459375044790735, + "learning_rate": 0.003, + "loss": 4.1035, + "step": 14958 + }, + { + "epoch": 0.14959, + "grad_norm": 0.8277616299795953, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 14959 + }, + { + "epoch": 0.1496, + "grad_norm": 0.8037915465398181, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 14960 + }, + { + "epoch": 0.14961, + "grad_norm": 0.8161694618639441, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 14961 + }, + { + "epoch": 0.14962, + "grad_norm": 0.8528318950797544, + "learning_rate": 0.003, + "loss": 4.059, + "step": 14962 + }, + { + "epoch": 0.14963, + "grad_norm": 0.9006272427248331, + "learning_rate": 0.003, + "loss": 4.069, + "step": 14963 + }, + { + "epoch": 0.14964, + "grad_norm": 0.98504197689359, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 14964 + }, + { + "epoch": 0.14965, + "grad_norm": 1.1290252654401167, + "learning_rate": 0.003, + "loss": 4.074, + "step": 14965 + }, + { + "epoch": 0.14966, + "grad_norm": 1.2675935278854713, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 14966 + }, + { + "epoch": 0.14967, + "grad_norm": 0.9516242221405834, + "learning_rate": 0.003, + "loss": 4.0986, + "step": 14967 + }, + { + "epoch": 0.14968, + "grad_norm": 0.9640976167389047, + "learning_rate": 0.003, + "loss": 4.079, + "step": 14968 + }, + { + "epoch": 0.14969, + "grad_norm": 1.0365736613967174, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 14969 + }, + { + "epoch": 0.1497, + "grad_norm": 1.063037600060497, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 14970 + }, + { + "epoch": 0.14971, + "grad_norm": 1.1120173723522138, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 14971 + }, + { + "epoch": 0.14972, + "grad_norm": 0.9414301431305664, + "learning_rate": 0.003, + "loss": 4.084, + "step": 14972 + }, + { + "epoch": 0.14973, + "grad_norm": 1.0464519552977811, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 14973 + }, + { + "epoch": 0.14974, + "grad_norm": 1.1931316330823598, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 14974 + }, + { + "epoch": 0.14975, + "grad_norm": 1.0078671266537655, + "learning_rate": 0.003, + "loss": 4.086, + "step": 14975 + }, + { + "epoch": 0.14976, + "grad_norm": 1.1968531987724693, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 14976 + }, + { + "epoch": 0.14977, + "grad_norm": 1.1560460152699668, + "learning_rate": 0.003, + "loss": 4.1029, + "step": 14977 + }, + { + "epoch": 0.14978, + "grad_norm": 0.9037387639695335, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 14978 + }, + { + "epoch": 0.14979, + "grad_norm": 0.957917208052118, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 14979 + }, + { + "epoch": 0.1498, + "grad_norm": 1.1331301112884007, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 14980 + }, + { + "epoch": 0.14981, + "grad_norm": 1.0016150493499458, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 14981 + }, + { + "epoch": 0.14982, + "grad_norm": 1.1360816214638219, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 14982 + }, + { + "epoch": 0.14983, + "grad_norm": 1.0279745578829336, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 14983 + }, + { + "epoch": 0.14984, + "grad_norm": 1.0506670365020176, + "learning_rate": 0.003, + "loss": 4.061, + "step": 14984 + }, + { + "epoch": 0.14985, + "grad_norm": 0.9418580459563878, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 14985 + }, + { + "epoch": 0.14986, + "grad_norm": 0.7582732804898829, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 14986 + }, + { + "epoch": 0.14987, + "grad_norm": 0.7065722372160369, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 14987 + }, + { + "epoch": 0.14988, + "grad_norm": 0.7403527953867218, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 14988 + }, + { + "epoch": 0.14989, + "grad_norm": 0.7941735683073315, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 14989 + }, + { + "epoch": 0.1499, + "grad_norm": 0.9986463946568017, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 14990 + }, + { + "epoch": 0.14991, + "grad_norm": 1.1068499011417936, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 14991 + }, + { + "epoch": 0.14992, + "grad_norm": 0.8254976921969661, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 14992 + }, + { + "epoch": 0.14993, + "grad_norm": 0.7972535765761616, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 14993 + }, + { + "epoch": 0.14994, + "grad_norm": 0.8176820242789884, + "learning_rate": 0.003, + "loss": 4.073, + "step": 14994 + }, + { + "epoch": 0.14995, + "grad_norm": 0.9082859872924608, + "learning_rate": 0.003, + "loss": 4.075, + "step": 14995 + }, + { + "epoch": 0.14996, + "grad_norm": 1.0161044574149738, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 14996 + }, + { + "epoch": 0.14997, + "grad_norm": 0.9932783023427505, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 14997 + }, + { + "epoch": 0.14998, + "grad_norm": 1.1483258068897348, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 14998 + }, + { + "epoch": 0.14999, + "grad_norm": 0.8948175763835226, + "learning_rate": 0.003, + "loss": 4.0908, + "step": 14999 + }, + { + "epoch": 0.15, + "grad_norm": 0.9104479834008374, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 15000 + }, + { + "epoch": 0.15001, + "grad_norm": 0.931319293277713, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 15001 + }, + { + "epoch": 0.15002, + "grad_norm": 0.8698790746399278, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 15002 + }, + { + "epoch": 0.15003, + "grad_norm": 0.9645310760897184, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 15003 + }, + { + "epoch": 0.15004, + "grad_norm": 1.2523120530500356, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 15004 + }, + { + "epoch": 0.15005, + "grad_norm": 0.9501051840827929, + "learning_rate": 0.003, + "loss": 4.096, + "step": 15005 + }, + { + "epoch": 0.15006, + "grad_norm": 1.1619701456069962, + "learning_rate": 0.003, + "loss": 4.1044, + "step": 15006 + }, + { + "epoch": 0.15007, + "grad_norm": 0.9487370066637558, + "learning_rate": 0.003, + "loss": 4.0996, + "step": 15007 + }, + { + "epoch": 0.15008, + "grad_norm": 1.0108245977178367, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 15008 + }, + { + "epoch": 0.15009, + "grad_norm": 1.0473259942156123, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 15009 + }, + { + "epoch": 0.1501, + "grad_norm": 1.1286111118764588, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 15010 + }, + { + "epoch": 0.15011, + "grad_norm": 0.926382738167326, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 15011 + }, + { + "epoch": 0.15012, + "grad_norm": 0.9791833454564608, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 15012 + }, + { + "epoch": 0.15013, + "grad_norm": 1.0979019106019359, + "learning_rate": 0.003, + "loss": 4.0918, + "step": 15013 + }, + { + "epoch": 0.15014, + "grad_norm": 0.9223740121269328, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 15014 + }, + { + "epoch": 0.15015, + "grad_norm": 0.9623854037837613, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 15015 + }, + { + "epoch": 0.15016, + "grad_norm": 0.9707245757305712, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 15016 + }, + { + "epoch": 0.15017, + "grad_norm": 1.0791194785614917, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 15017 + }, + { + "epoch": 0.15018, + "grad_norm": 1.09604945931361, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 15018 + }, + { + "epoch": 0.15019, + "grad_norm": 1.1392686245966401, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 15019 + }, + { + "epoch": 0.1502, + "grad_norm": 0.801786611003596, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 15020 + }, + { + "epoch": 0.15021, + "grad_norm": 0.8308415464817841, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 15021 + }, + { + "epoch": 0.15022, + "grad_norm": 0.871439336217148, + "learning_rate": 0.003, + "loss": 4.0962, + "step": 15022 + }, + { + "epoch": 0.15023, + "grad_norm": 0.9834313653969391, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 15023 + }, + { + "epoch": 0.15024, + "grad_norm": 1.035443047011185, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 15024 + }, + { + "epoch": 0.15025, + "grad_norm": 1.0591704089286167, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 15025 + }, + { + "epoch": 0.15026, + "grad_norm": 1.2177106743734125, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 15026 + }, + { + "epoch": 0.15027, + "grad_norm": 0.9142216625458663, + "learning_rate": 0.003, + "loss": 4.093, + "step": 15027 + }, + { + "epoch": 0.15028, + "grad_norm": 1.020411790717614, + "learning_rate": 0.003, + "loss": 4.0884, + "step": 15028 + }, + { + "epoch": 0.15029, + "grad_norm": 1.1283704224415978, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 15029 + }, + { + "epoch": 0.1503, + "grad_norm": 0.9806289299418071, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 15030 + }, + { + "epoch": 0.15031, + "grad_norm": 0.9941211615421099, + "learning_rate": 0.003, + "loss": 4.066, + "step": 15031 + }, + { + "epoch": 0.15032, + "grad_norm": 1.1866538862070592, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 15032 + }, + { + "epoch": 0.15033, + "grad_norm": 1.2389896605592816, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 15033 + }, + { + "epoch": 0.15034, + "grad_norm": 0.8867610761327838, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 15034 + }, + { + "epoch": 0.15035, + "grad_norm": 0.8906294325306437, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 15035 + }, + { + "epoch": 0.15036, + "grad_norm": 0.9549465553530418, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 15036 + }, + { + "epoch": 0.15037, + "grad_norm": 1.1594113963181591, + "learning_rate": 0.003, + "loss": 4.1069, + "step": 15037 + }, + { + "epoch": 0.15038, + "grad_norm": 0.9233709872345994, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 15038 + }, + { + "epoch": 0.15039, + "grad_norm": 0.9091266708064162, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 15039 + }, + { + "epoch": 0.1504, + "grad_norm": 0.8320390121046842, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 15040 + }, + { + "epoch": 0.15041, + "grad_norm": 0.7633752655645272, + "learning_rate": 0.003, + "loss": 4.1109, + "step": 15041 + }, + { + "epoch": 0.15042, + "grad_norm": 0.8448120221028312, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 15042 + }, + { + "epoch": 0.15043, + "grad_norm": 0.8582219844957641, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 15043 + }, + { + "epoch": 0.15044, + "grad_norm": 0.8845788613896037, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 15044 + }, + { + "epoch": 0.15045, + "grad_norm": 0.8591906216262435, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 15045 + }, + { + "epoch": 0.15046, + "grad_norm": 0.7937944826883396, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 15046 + }, + { + "epoch": 0.15047, + "grad_norm": 0.7268910106905411, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 15047 + }, + { + "epoch": 0.15048, + "grad_norm": 0.8739198337921111, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 15048 + }, + { + "epoch": 0.15049, + "grad_norm": 1.200598344071806, + "learning_rate": 0.003, + "loss": 4.0999, + "step": 15049 + }, + { + "epoch": 0.1505, + "grad_norm": 0.979187032423917, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 15050 + }, + { + "epoch": 0.15051, + "grad_norm": 1.1660542360592214, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 15051 + }, + { + "epoch": 0.15052, + "grad_norm": 0.9912110185394836, + "learning_rate": 0.003, + "loss": 4.087, + "step": 15052 + }, + { + "epoch": 0.15053, + "grad_norm": 1.1518287678337726, + "learning_rate": 0.003, + "loss": 4.0934, + "step": 15053 + }, + { + "epoch": 0.15054, + "grad_norm": 0.8664681502534709, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 15054 + }, + { + "epoch": 0.15055, + "grad_norm": 0.7896500405972969, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 15055 + }, + { + "epoch": 0.15056, + "grad_norm": 0.7392006810623563, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 15056 + }, + { + "epoch": 0.15057, + "grad_norm": 0.7963838256059327, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 15057 + }, + { + "epoch": 0.15058, + "grad_norm": 0.9014888686618083, + "learning_rate": 0.003, + "loss": 4.0923, + "step": 15058 + }, + { + "epoch": 0.15059, + "grad_norm": 1.1124501627109835, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 15059 + }, + { + "epoch": 0.1506, + "grad_norm": 0.8860071524915372, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 15060 + }, + { + "epoch": 0.15061, + "grad_norm": 0.8327758394906547, + "learning_rate": 0.003, + "loss": 4.0872, + "step": 15061 + }, + { + "epoch": 0.15062, + "grad_norm": 0.9960049448448539, + "learning_rate": 0.003, + "loss": 4.077, + "step": 15062 + }, + { + "epoch": 0.15063, + "grad_norm": 1.325775900261404, + "learning_rate": 0.003, + "loss": 4.0954, + "step": 15063 + }, + { + "epoch": 0.15064, + "grad_norm": 0.9344596952497445, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 15064 + }, + { + "epoch": 0.15065, + "grad_norm": 0.9140940808546324, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 15065 + }, + { + "epoch": 0.15066, + "grad_norm": 1.0282053321246554, + "learning_rate": 0.003, + "loss": 4.0935, + "step": 15066 + }, + { + "epoch": 0.15067, + "grad_norm": 1.0510844606874918, + "learning_rate": 0.003, + "loss": 4.1048, + "step": 15067 + }, + { + "epoch": 0.15068, + "grad_norm": 1.049557556047745, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 15068 + }, + { + "epoch": 0.15069, + "grad_norm": 0.9935590700445011, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 15069 + }, + { + "epoch": 0.1507, + "grad_norm": 0.921123933458733, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 15070 + }, + { + "epoch": 0.15071, + "grad_norm": 0.8235110229426629, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 15071 + }, + { + "epoch": 0.15072, + "grad_norm": 0.7269443430652391, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 15072 + }, + { + "epoch": 0.15073, + "grad_norm": 0.6857463234725705, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 15073 + }, + { + "epoch": 0.15074, + "grad_norm": 0.874992346339928, + "learning_rate": 0.003, + "loss": 4.076, + "step": 15074 + }, + { + "epoch": 0.15075, + "grad_norm": 1.087077191122076, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 15075 + }, + { + "epoch": 0.15076, + "grad_norm": 0.954673078474433, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 15076 + }, + { + "epoch": 0.15077, + "grad_norm": 1.0659981834801247, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 15077 + }, + { + "epoch": 0.15078, + "grad_norm": 1.160923063617543, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 15078 + }, + { + "epoch": 0.15079, + "grad_norm": 1.166270131956899, + "learning_rate": 0.003, + "loss": 4.0886, + "step": 15079 + }, + { + "epoch": 0.1508, + "grad_norm": 0.8785784200861291, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 15080 + }, + { + "epoch": 0.15081, + "grad_norm": 0.8374329030996557, + "learning_rate": 0.003, + "loss": 4.0986, + "step": 15081 + }, + { + "epoch": 0.15082, + "grad_norm": 0.8115379711475327, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 15082 + }, + { + "epoch": 0.15083, + "grad_norm": 0.8850290338014076, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 15083 + }, + { + "epoch": 0.15084, + "grad_norm": 0.8914452364774292, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 15084 + }, + { + "epoch": 0.15085, + "grad_norm": 0.8546023063386037, + "learning_rate": 0.003, + "loss": 4.0864, + "step": 15085 + }, + { + "epoch": 0.15086, + "grad_norm": 1.0966865612837757, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 15086 + }, + { + "epoch": 0.15087, + "grad_norm": 0.9164590263107982, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 15087 + }, + { + "epoch": 0.15088, + "grad_norm": 1.0902285341591487, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 15088 + }, + { + "epoch": 0.15089, + "grad_norm": 1.1736553420796854, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 15089 + }, + { + "epoch": 0.1509, + "grad_norm": 0.9456250072363903, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 15090 + }, + { + "epoch": 0.15091, + "grad_norm": 0.9350680185701131, + "learning_rate": 0.003, + "loss": 4.0932, + "step": 15091 + }, + { + "epoch": 0.15092, + "grad_norm": 1.058760791538963, + "learning_rate": 0.003, + "loss": 4.0966, + "step": 15092 + }, + { + "epoch": 0.15093, + "grad_norm": 1.1476019190901443, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 15093 + }, + { + "epoch": 0.15094, + "grad_norm": 0.9872034970857719, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 15094 + }, + { + "epoch": 0.15095, + "grad_norm": 1.0225269043395846, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 15095 + }, + { + "epoch": 0.15096, + "grad_norm": 1.0126780484647222, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 15096 + }, + { + "epoch": 0.15097, + "grad_norm": 0.9230274905865083, + "learning_rate": 0.003, + "loss": 4.072, + "step": 15097 + }, + { + "epoch": 0.15098, + "grad_norm": 0.8719234567601442, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 15098 + }, + { + "epoch": 0.15099, + "grad_norm": 1.0557217748010168, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 15099 + }, + { + "epoch": 0.151, + "grad_norm": 1.1102074391045602, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 15100 + }, + { + "epoch": 0.15101, + "grad_norm": 1.0370206174256693, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 15101 + }, + { + "epoch": 0.15102, + "grad_norm": 1.0737417775566704, + "learning_rate": 0.003, + "loss": 4.0923, + "step": 15102 + }, + { + "epoch": 0.15103, + "grad_norm": 0.9845312763978304, + "learning_rate": 0.003, + "loss": 4.083, + "step": 15103 + }, + { + "epoch": 0.15104, + "grad_norm": 1.1049336796221643, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 15104 + }, + { + "epoch": 0.15105, + "grad_norm": 1.2303680953029525, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 15105 + }, + { + "epoch": 0.15106, + "grad_norm": 0.9507957916099791, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 15106 + }, + { + "epoch": 0.15107, + "grad_norm": 0.9055127885017512, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 15107 + }, + { + "epoch": 0.15108, + "grad_norm": 0.9089067801855328, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 15108 + }, + { + "epoch": 0.15109, + "grad_norm": 1.0310685499273569, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 15109 + }, + { + "epoch": 0.1511, + "grad_norm": 1.243319933769187, + "learning_rate": 0.003, + "loss": 4.047, + "step": 15110 + }, + { + "epoch": 0.15111, + "grad_norm": 0.8872450433148288, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 15111 + }, + { + "epoch": 0.15112, + "grad_norm": 0.9452797677961886, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 15112 + }, + { + "epoch": 0.15113, + "grad_norm": 0.9886146686220433, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 15113 + }, + { + "epoch": 0.15114, + "grad_norm": 1.1235943610687726, + "learning_rate": 0.003, + "loss": 4.06, + "step": 15114 + }, + { + "epoch": 0.15115, + "grad_norm": 1.0038230124915115, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 15115 + }, + { + "epoch": 0.15116, + "grad_norm": 1.0830526959581306, + "learning_rate": 0.003, + "loss": 4.0858, + "step": 15116 + }, + { + "epoch": 0.15117, + "grad_norm": 0.9921990381889474, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 15117 + }, + { + "epoch": 0.15118, + "grad_norm": 1.0365285267045834, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 15118 + }, + { + "epoch": 0.15119, + "grad_norm": 0.8968368129301871, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 15119 + }, + { + "epoch": 0.1512, + "grad_norm": 0.8674233092888736, + "learning_rate": 0.003, + "loss": 4.0999, + "step": 15120 + }, + { + "epoch": 0.15121, + "grad_norm": 0.8238581586631851, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 15121 + }, + { + "epoch": 0.15122, + "grad_norm": 1.0053480701083943, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 15122 + }, + { + "epoch": 0.15123, + "grad_norm": 1.1700889945622388, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 15123 + }, + { + "epoch": 0.15124, + "grad_norm": 0.8763955530919195, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 15124 + }, + { + "epoch": 0.15125, + "grad_norm": 0.9167327119335428, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 15125 + }, + { + "epoch": 0.15126, + "grad_norm": 0.9559538328443866, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 15126 + }, + { + "epoch": 0.15127, + "grad_norm": 0.974389974326513, + "learning_rate": 0.003, + "loss": 4.0878, + "step": 15127 + }, + { + "epoch": 0.15128, + "grad_norm": 1.0820887135449986, + "learning_rate": 0.003, + "loss": 4.0864, + "step": 15128 + }, + { + "epoch": 0.15129, + "grad_norm": 1.1549927760511465, + "learning_rate": 0.003, + "loss": 4.0996, + "step": 15129 + }, + { + "epoch": 0.1513, + "grad_norm": 0.8755464661297412, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 15130 + }, + { + "epoch": 0.15131, + "grad_norm": 0.9061416609077914, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 15131 + }, + { + "epoch": 0.15132, + "grad_norm": 0.9198441494589286, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 15132 + }, + { + "epoch": 0.15133, + "grad_norm": 1.0513316727767488, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 15133 + }, + { + "epoch": 0.15134, + "grad_norm": 1.0164096791473425, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 15134 + }, + { + "epoch": 0.15135, + "grad_norm": 0.9520107911869734, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 15135 + }, + { + "epoch": 0.15136, + "grad_norm": 0.8980499440581962, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 15136 + }, + { + "epoch": 0.15137, + "grad_norm": 0.8739658695858102, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 15137 + }, + { + "epoch": 0.15138, + "grad_norm": 0.9210006399410474, + "learning_rate": 0.003, + "loss": 4.1052, + "step": 15138 + }, + { + "epoch": 0.15139, + "grad_norm": 0.9173765826246364, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 15139 + }, + { + "epoch": 0.1514, + "grad_norm": 1.0659873780897715, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 15140 + }, + { + "epoch": 0.15141, + "grad_norm": 0.9800383652575361, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 15141 + }, + { + "epoch": 0.15142, + "grad_norm": 0.906274468171165, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 15142 + }, + { + "epoch": 0.15143, + "grad_norm": 0.841450622940112, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 15143 + }, + { + "epoch": 0.15144, + "grad_norm": 0.8535243810797298, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 15144 + }, + { + "epoch": 0.15145, + "grad_norm": 1.1079494548200066, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 15145 + }, + { + "epoch": 0.15146, + "grad_norm": 1.1296433113086481, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 15146 + }, + { + "epoch": 0.15147, + "grad_norm": 1.00138393931858, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 15147 + }, + { + "epoch": 0.15148, + "grad_norm": 1.0566318308621163, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 15148 + }, + { + "epoch": 0.15149, + "grad_norm": 0.9755093048359352, + "learning_rate": 0.003, + "loss": 4.09, + "step": 15149 + }, + { + "epoch": 0.1515, + "grad_norm": 0.9784813301681908, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 15150 + }, + { + "epoch": 0.15151, + "grad_norm": 0.9193708550293552, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 15151 + }, + { + "epoch": 0.15152, + "grad_norm": 0.900727077400924, + "learning_rate": 0.003, + "loss": 4.1123, + "step": 15152 + }, + { + "epoch": 0.15153, + "grad_norm": 0.8778920987616318, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 15153 + }, + { + "epoch": 0.15154, + "grad_norm": 0.825552831526053, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 15154 + }, + { + "epoch": 0.15155, + "grad_norm": 0.9850018246443522, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 15155 + }, + { + "epoch": 0.15156, + "grad_norm": 1.1296434151007493, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 15156 + }, + { + "epoch": 0.15157, + "grad_norm": 1.0448776165885458, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 15157 + }, + { + "epoch": 0.15158, + "grad_norm": 1.2775680575482375, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 15158 + }, + { + "epoch": 0.15159, + "grad_norm": 0.8178979540518811, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 15159 + }, + { + "epoch": 0.1516, + "grad_norm": 0.768246064159124, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 15160 + }, + { + "epoch": 0.15161, + "grad_norm": 1.0405776187710276, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 15161 + }, + { + "epoch": 0.15162, + "grad_norm": 0.9823974820279163, + "learning_rate": 0.003, + "loss": 4.106, + "step": 15162 + }, + { + "epoch": 0.15163, + "grad_norm": 1.0002167840365008, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 15163 + }, + { + "epoch": 0.15164, + "grad_norm": 0.8941917182557578, + "learning_rate": 0.003, + "loss": 4.069, + "step": 15164 + }, + { + "epoch": 0.15165, + "grad_norm": 1.0472086981259003, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 15165 + }, + { + "epoch": 0.15166, + "grad_norm": 1.0178804253579137, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 15166 + }, + { + "epoch": 0.15167, + "grad_norm": 0.9682218394564448, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 15167 + }, + { + "epoch": 0.15168, + "grad_norm": 0.8827766169706042, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 15168 + }, + { + "epoch": 0.15169, + "grad_norm": 0.8367581337301362, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 15169 + }, + { + "epoch": 0.1517, + "grad_norm": 1.0839300602024455, + "learning_rate": 0.003, + "loss": 4.056, + "step": 15170 + }, + { + "epoch": 0.15171, + "grad_norm": 1.1726810819150073, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 15171 + }, + { + "epoch": 0.15172, + "grad_norm": 1.0517790683179402, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 15172 + }, + { + "epoch": 0.15173, + "grad_norm": 0.9583423199267139, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 15173 + }, + { + "epoch": 0.15174, + "grad_norm": 0.9543561218115898, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 15174 + }, + { + "epoch": 0.15175, + "grad_norm": 1.1568841190504215, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 15175 + }, + { + "epoch": 0.15176, + "grad_norm": 0.9835682724127945, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 15176 + }, + { + "epoch": 0.15177, + "grad_norm": 0.9879553477290558, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 15177 + }, + { + "epoch": 0.15178, + "grad_norm": 1.003463087559272, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 15178 + }, + { + "epoch": 0.15179, + "grad_norm": 1.152112026150714, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 15179 + }, + { + "epoch": 0.1518, + "grad_norm": 1.214590702520319, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 15180 + }, + { + "epoch": 0.15181, + "grad_norm": 0.7902880387457807, + "learning_rate": 0.003, + "loss": 4.059, + "step": 15181 + }, + { + "epoch": 0.15182, + "grad_norm": 0.8599493611875796, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 15182 + }, + { + "epoch": 0.15183, + "grad_norm": 1.1154265752151489, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 15183 + }, + { + "epoch": 0.15184, + "grad_norm": 1.0719797221662317, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 15184 + }, + { + "epoch": 0.15185, + "grad_norm": 1.0831147462352226, + "learning_rate": 0.003, + "loss": 4.085, + "step": 15185 + }, + { + "epoch": 0.15186, + "grad_norm": 0.897563382299871, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 15186 + }, + { + "epoch": 0.15187, + "grad_norm": 0.8399090851162266, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 15187 + }, + { + "epoch": 0.15188, + "grad_norm": 0.932687588550206, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 15188 + }, + { + "epoch": 0.15189, + "grad_norm": 1.0655341353823446, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 15189 + }, + { + "epoch": 0.1519, + "grad_norm": 1.1340121303676967, + "learning_rate": 0.003, + "loss": 4.0936, + "step": 15190 + }, + { + "epoch": 0.15191, + "grad_norm": 0.9867501320240778, + "learning_rate": 0.003, + "loss": 4.104, + "step": 15191 + }, + { + "epoch": 0.15192, + "grad_norm": 0.9320741225521991, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 15192 + }, + { + "epoch": 0.15193, + "grad_norm": 0.9922425463959831, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 15193 + }, + { + "epoch": 0.15194, + "grad_norm": 1.0608438721613602, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 15194 + }, + { + "epoch": 0.15195, + "grad_norm": 1.0991420254079054, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 15195 + }, + { + "epoch": 0.15196, + "grad_norm": 0.8805489450915264, + "learning_rate": 0.003, + "loss": 4.0872, + "step": 15196 + }, + { + "epoch": 0.15197, + "grad_norm": 0.9028227020885176, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 15197 + }, + { + "epoch": 0.15198, + "grad_norm": 0.9330038131491715, + "learning_rate": 0.003, + "loss": 4.086, + "step": 15198 + }, + { + "epoch": 0.15199, + "grad_norm": 1.0470036404955456, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 15199 + }, + { + "epoch": 0.152, + "grad_norm": 1.218296437063532, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 15200 + }, + { + "epoch": 0.15201, + "grad_norm": 1.0677472806179267, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 15201 + }, + { + "epoch": 0.15202, + "grad_norm": 1.0294090190749403, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 15202 + }, + { + "epoch": 0.15203, + "grad_norm": 1.1263465776770114, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 15203 + }, + { + "epoch": 0.15204, + "grad_norm": 0.6895194031345796, + "learning_rate": 0.003, + "loss": 4.0953, + "step": 15204 + }, + { + "epoch": 0.15205, + "grad_norm": 0.6870133158919096, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 15205 + }, + { + "epoch": 0.15206, + "grad_norm": 0.8117391565216929, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 15206 + }, + { + "epoch": 0.15207, + "grad_norm": 0.8127505677365994, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 15207 + }, + { + "epoch": 0.15208, + "grad_norm": 0.8364216963757153, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 15208 + }, + { + "epoch": 0.15209, + "grad_norm": 0.8145989227129922, + "learning_rate": 0.003, + "loss": 4.0873, + "step": 15209 + }, + { + "epoch": 0.1521, + "grad_norm": 0.9706231674398657, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 15210 + }, + { + "epoch": 0.15211, + "grad_norm": 1.2478563926278767, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 15211 + }, + { + "epoch": 0.15212, + "grad_norm": 0.7866509418359351, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 15212 + }, + { + "epoch": 0.15213, + "grad_norm": 0.6089696802783601, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 15213 + }, + { + "epoch": 0.15214, + "grad_norm": 0.650722604961699, + "learning_rate": 0.003, + "loss": 4.061, + "step": 15214 + }, + { + "epoch": 0.15215, + "grad_norm": 0.6512675917745587, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 15215 + }, + { + "epoch": 0.15216, + "grad_norm": 0.5604541408586048, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 15216 + }, + { + "epoch": 0.15217, + "grad_norm": 0.8405971915595959, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 15217 + }, + { + "epoch": 0.15218, + "grad_norm": 1.4241525705418698, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 15218 + }, + { + "epoch": 0.15219, + "grad_norm": 0.8945132778473138, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 15219 + }, + { + "epoch": 0.1522, + "grad_norm": 0.9391764208520771, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 15220 + }, + { + "epoch": 0.15221, + "grad_norm": 1.1362917028379544, + "learning_rate": 0.003, + "loss": 4.0969, + "step": 15221 + }, + { + "epoch": 0.15222, + "grad_norm": 0.9737452589111013, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 15222 + }, + { + "epoch": 0.15223, + "grad_norm": 0.9021594342565704, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 15223 + }, + { + "epoch": 0.15224, + "grad_norm": 0.8249102310145384, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 15224 + }, + { + "epoch": 0.15225, + "grad_norm": 0.8904424706013546, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 15225 + }, + { + "epoch": 0.15226, + "grad_norm": 0.9900369047792563, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 15226 + }, + { + "epoch": 0.15227, + "grad_norm": 1.0988499770487428, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 15227 + }, + { + "epoch": 0.15228, + "grad_norm": 1.186364803660202, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 15228 + }, + { + "epoch": 0.15229, + "grad_norm": 1.3780872240525257, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 15229 + }, + { + "epoch": 0.1523, + "grad_norm": 0.8825322728625716, + "learning_rate": 0.003, + "loss": 4.0936, + "step": 15230 + }, + { + "epoch": 0.15231, + "grad_norm": 0.90943567489747, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 15231 + }, + { + "epoch": 0.15232, + "grad_norm": 1.068002548418485, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 15232 + }, + { + "epoch": 0.15233, + "grad_norm": 1.1869662206066822, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 15233 + }, + { + "epoch": 0.15234, + "grad_norm": 0.9240622891549248, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 15234 + }, + { + "epoch": 0.15235, + "grad_norm": 1.128355767673242, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 15235 + }, + { + "epoch": 0.15236, + "grad_norm": 1.066717309876731, + "learning_rate": 0.003, + "loss": 4.0941, + "step": 15236 + }, + { + "epoch": 0.15237, + "grad_norm": 0.8635636235124252, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 15237 + }, + { + "epoch": 0.15238, + "grad_norm": 0.7124185694299446, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 15238 + }, + { + "epoch": 0.15239, + "grad_norm": 0.8667692657122733, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 15239 + }, + { + "epoch": 0.1524, + "grad_norm": 1.0790627490043403, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 15240 + }, + { + "epoch": 0.15241, + "grad_norm": 1.2174177319000792, + "learning_rate": 0.003, + "loss": 4.0989, + "step": 15241 + }, + { + "epoch": 0.15242, + "grad_norm": 0.7941050971580655, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 15242 + }, + { + "epoch": 0.15243, + "grad_norm": 0.6612669509255636, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 15243 + }, + { + "epoch": 0.15244, + "grad_norm": 0.7997536284873389, + "learning_rate": 0.003, + "loss": 4.082, + "step": 15244 + }, + { + "epoch": 0.15245, + "grad_norm": 0.9085473175226056, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 15245 + }, + { + "epoch": 0.15246, + "grad_norm": 0.9072832672660933, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 15246 + }, + { + "epoch": 0.15247, + "grad_norm": 1.0197083732816214, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 15247 + }, + { + "epoch": 0.15248, + "grad_norm": 0.933758392484516, + "learning_rate": 0.003, + "loss": 4.0928, + "step": 15248 + }, + { + "epoch": 0.15249, + "grad_norm": 0.9567391991390576, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 15249 + }, + { + "epoch": 0.1525, + "grad_norm": 0.8767416517698086, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 15250 + }, + { + "epoch": 0.15251, + "grad_norm": 0.8779479960984673, + "learning_rate": 0.003, + "loss": 4.1031, + "step": 15251 + }, + { + "epoch": 0.15252, + "grad_norm": 0.9086753470214751, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 15252 + }, + { + "epoch": 0.15253, + "grad_norm": 0.9209451399039114, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 15253 + }, + { + "epoch": 0.15254, + "grad_norm": 1.0580678792837188, + "learning_rate": 0.003, + "loss": 4.07, + "step": 15254 + }, + { + "epoch": 0.15255, + "grad_norm": 1.135753117186445, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 15255 + }, + { + "epoch": 0.15256, + "grad_norm": 1.1457704995713298, + "learning_rate": 0.003, + "loss": 4.069, + "step": 15256 + }, + { + "epoch": 0.15257, + "grad_norm": 1.0854907708500827, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 15257 + }, + { + "epoch": 0.15258, + "grad_norm": 1.0124498398893578, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 15258 + }, + { + "epoch": 0.15259, + "grad_norm": 0.9899233890784849, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 15259 + }, + { + "epoch": 0.1526, + "grad_norm": 1.05807073338885, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 15260 + }, + { + "epoch": 0.15261, + "grad_norm": 0.9526695704438424, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 15261 + }, + { + "epoch": 0.15262, + "grad_norm": 0.9821534397815488, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 15262 + }, + { + "epoch": 0.15263, + "grad_norm": 1.1187937248496844, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 15263 + }, + { + "epoch": 0.15264, + "grad_norm": 1.0250810529736074, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 15264 + }, + { + "epoch": 0.15265, + "grad_norm": 1.1021139564842417, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 15265 + }, + { + "epoch": 0.15266, + "grad_norm": 0.9706724468571734, + "learning_rate": 0.003, + "loss": 4.049, + "step": 15266 + }, + { + "epoch": 0.15267, + "grad_norm": 0.7998352572389079, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 15267 + }, + { + "epoch": 0.15268, + "grad_norm": 0.8532275048586067, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 15268 + }, + { + "epoch": 0.15269, + "grad_norm": 1.0149158104442335, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 15269 + }, + { + "epoch": 0.1527, + "grad_norm": 1.2384950294538573, + "learning_rate": 0.003, + "loss": 4.078, + "step": 15270 + }, + { + "epoch": 0.15271, + "grad_norm": 1.066278966110427, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 15271 + }, + { + "epoch": 0.15272, + "grad_norm": 1.0106358961904032, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 15272 + }, + { + "epoch": 0.15273, + "grad_norm": 1.1911351377354713, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 15273 + }, + { + "epoch": 0.15274, + "grad_norm": 1.2360668538099449, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 15274 + }, + { + "epoch": 0.15275, + "grad_norm": 0.9105469816259705, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 15275 + }, + { + "epoch": 0.15276, + "grad_norm": 0.7817966481046521, + "learning_rate": 0.003, + "loss": 4.0855, + "step": 15276 + }, + { + "epoch": 0.15277, + "grad_norm": 0.7853618355082954, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 15277 + }, + { + "epoch": 0.15278, + "grad_norm": 0.8502014857007658, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 15278 + }, + { + "epoch": 0.15279, + "grad_norm": 0.9025981408111081, + "learning_rate": 0.003, + "loss": 4.073, + "step": 15279 + }, + { + "epoch": 0.1528, + "grad_norm": 1.0299578769126079, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 15280 + }, + { + "epoch": 0.15281, + "grad_norm": 1.2338075919093996, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 15281 + }, + { + "epoch": 0.15282, + "grad_norm": 1.1578084763702126, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 15282 + }, + { + "epoch": 0.15283, + "grad_norm": 1.1953454215816077, + "learning_rate": 0.003, + "loss": 4.0923, + "step": 15283 + }, + { + "epoch": 0.15284, + "grad_norm": 0.8894587671691975, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 15284 + }, + { + "epoch": 0.15285, + "grad_norm": 0.9028712799998575, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 15285 + }, + { + "epoch": 0.15286, + "grad_norm": 1.0010709251799794, + "learning_rate": 0.003, + "loss": 4.074, + "step": 15286 + }, + { + "epoch": 0.15287, + "grad_norm": 1.0775357504468008, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 15287 + }, + { + "epoch": 0.15288, + "grad_norm": 0.9299308068508235, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 15288 + }, + { + "epoch": 0.15289, + "grad_norm": 1.0801992978117096, + "learning_rate": 0.003, + "loss": 4.049, + "step": 15289 + }, + { + "epoch": 0.1529, + "grad_norm": 1.022047015292239, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 15290 + }, + { + "epoch": 0.15291, + "grad_norm": 1.05610861514632, + "learning_rate": 0.003, + "loss": 4.0899, + "step": 15291 + }, + { + "epoch": 0.15292, + "grad_norm": 0.8990912773095489, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 15292 + }, + { + "epoch": 0.15293, + "grad_norm": 0.8916044723389804, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 15293 + }, + { + "epoch": 0.15294, + "grad_norm": 0.9365013150306947, + "learning_rate": 0.003, + "loss": 4.0848, + "step": 15294 + }, + { + "epoch": 0.15295, + "grad_norm": 0.8506647184913037, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 15295 + }, + { + "epoch": 0.15296, + "grad_norm": 0.9573454717764328, + "learning_rate": 0.003, + "loss": 4.0925, + "step": 15296 + }, + { + "epoch": 0.15297, + "grad_norm": 1.1534920614152682, + "learning_rate": 0.003, + "loss": 4.069, + "step": 15297 + }, + { + "epoch": 0.15298, + "grad_norm": 1.0434226777196847, + "learning_rate": 0.003, + "loss": 4.0899, + "step": 15298 + }, + { + "epoch": 0.15299, + "grad_norm": 1.1737319592718236, + "learning_rate": 0.003, + "loss": 4.0881, + "step": 15299 + }, + { + "epoch": 0.153, + "grad_norm": 0.9282452226758077, + "learning_rate": 0.003, + "loss": 4.0943, + "step": 15300 + }, + { + "epoch": 0.15301, + "grad_norm": 0.86306014789999, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 15301 + }, + { + "epoch": 0.15302, + "grad_norm": 0.9140534661334827, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 15302 + }, + { + "epoch": 0.15303, + "grad_norm": 0.9535982092968598, + "learning_rate": 0.003, + "loss": 4.0977, + "step": 15303 + }, + { + "epoch": 0.15304, + "grad_norm": 1.0751336795512656, + "learning_rate": 0.003, + "loss": 4.1082, + "step": 15304 + }, + { + "epoch": 0.15305, + "grad_norm": 1.0129913657666836, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 15305 + }, + { + "epoch": 0.15306, + "grad_norm": 0.9775835211100742, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 15306 + }, + { + "epoch": 0.15307, + "grad_norm": 0.9818899517111416, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 15307 + }, + { + "epoch": 0.15308, + "grad_norm": 1.0295236082191443, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 15308 + }, + { + "epoch": 0.15309, + "grad_norm": 1.1980153637863054, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 15309 + }, + { + "epoch": 0.1531, + "grad_norm": 0.9983999586903378, + "learning_rate": 0.003, + "loss": 4.0917, + "step": 15310 + }, + { + "epoch": 0.15311, + "grad_norm": 1.0100172269049852, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 15311 + }, + { + "epoch": 0.15312, + "grad_norm": 0.9576078911544427, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 15312 + }, + { + "epoch": 0.15313, + "grad_norm": 1.05439136409979, + "learning_rate": 0.003, + "loss": 4.0948, + "step": 15313 + }, + { + "epoch": 0.15314, + "grad_norm": 1.0366810933918715, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 15314 + }, + { + "epoch": 0.15315, + "grad_norm": 0.8387629630392678, + "learning_rate": 0.003, + "loss": 4.0845, + "step": 15315 + }, + { + "epoch": 0.15316, + "grad_norm": 0.9719655982905502, + "learning_rate": 0.003, + "loss": 4.0973, + "step": 15316 + }, + { + "epoch": 0.15317, + "grad_norm": 1.1475140612787782, + "learning_rate": 0.003, + "loss": 4.1039, + "step": 15317 + }, + { + "epoch": 0.15318, + "grad_norm": 0.8744573249259142, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 15318 + }, + { + "epoch": 0.15319, + "grad_norm": 0.9089119524461857, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 15319 + }, + { + "epoch": 0.1532, + "grad_norm": 1.1362367673629654, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 15320 + }, + { + "epoch": 0.15321, + "grad_norm": 1.3521008050822485, + "learning_rate": 0.003, + "loss": 4.0857, + "step": 15321 + }, + { + "epoch": 0.15322, + "grad_norm": 0.8514884215069218, + "learning_rate": 0.003, + "loss": 4.0932, + "step": 15322 + }, + { + "epoch": 0.15323, + "grad_norm": 0.7297987523344688, + "learning_rate": 0.003, + "loss": 4.081, + "step": 15323 + }, + { + "epoch": 0.15324, + "grad_norm": 0.778672333339039, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 15324 + }, + { + "epoch": 0.15325, + "grad_norm": 0.9980772509510342, + "learning_rate": 0.003, + "loss": 4.094, + "step": 15325 + }, + { + "epoch": 0.15326, + "grad_norm": 1.2394720167369941, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 15326 + }, + { + "epoch": 0.15327, + "grad_norm": 0.7441055636220688, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 15327 + }, + { + "epoch": 0.15328, + "grad_norm": 0.7291708337776174, + "learning_rate": 0.003, + "loss": 4.036, + "step": 15328 + }, + { + "epoch": 0.15329, + "grad_norm": 0.7849634261614127, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 15329 + }, + { + "epoch": 0.1533, + "grad_norm": 0.9576284052925993, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 15330 + }, + { + "epoch": 0.15331, + "grad_norm": 1.4537940335841768, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 15331 + }, + { + "epoch": 0.15332, + "grad_norm": 0.6967801200454877, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 15332 + }, + { + "epoch": 0.15333, + "grad_norm": 0.7811632907320885, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 15333 + }, + { + "epoch": 0.15334, + "grad_norm": 1.1584449540648905, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 15334 + }, + { + "epoch": 0.15335, + "grad_norm": 1.112616635589485, + "learning_rate": 0.003, + "loss": 4.1, + "step": 15335 + }, + { + "epoch": 0.15336, + "grad_norm": 1.084357415329987, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 15336 + }, + { + "epoch": 0.15337, + "grad_norm": 0.9506857583640346, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 15337 + }, + { + "epoch": 0.15338, + "grad_norm": 0.9251098918436385, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 15338 + }, + { + "epoch": 0.15339, + "grad_norm": 0.9906148217708848, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 15339 + }, + { + "epoch": 0.1534, + "grad_norm": 1.1049096847593285, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 15340 + }, + { + "epoch": 0.15341, + "grad_norm": 0.9352969880943619, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 15341 + }, + { + "epoch": 0.15342, + "grad_norm": 0.8973943985475014, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 15342 + }, + { + "epoch": 0.15343, + "grad_norm": 0.9744698160935995, + "learning_rate": 0.003, + "loss": 4.061, + "step": 15343 + }, + { + "epoch": 0.15344, + "grad_norm": 0.9216723051708908, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 15344 + }, + { + "epoch": 0.15345, + "grad_norm": 0.9986940239485664, + "learning_rate": 0.003, + "loss": 4.0905, + "step": 15345 + }, + { + "epoch": 0.15346, + "grad_norm": 1.1212947820081267, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 15346 + }, + { + "epoch": 0.15347, + "grad_norm": 0.9766008495018536, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 15347 + }, + { + "epoch": 0.15348, + "grad_norm": 0.9155959842081269, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 15348 + }, + { + "epoch": 0.15349, + "grad_norm": 0.8846141528809668, + "learning_rate": 0.003, + "loss": 4.072, + "step": 15349 + }, + { + "epoch": 0.1535, + "grad_norm": 0.8578476969984926, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 15350 + }, + { + "epoch": 0.15351, + "grad_norm": 0.8665623236818796, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 15351 + }, + { + "epoch": 0.15352, + "grad_norm": 0.9457505430416461, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 15352 + }, + { + "epoch": 0.15353, + "grad_norm": 0.9543671874352742, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 15353 + }, + { + "epoch": 0.15354, + "grad_norm": 0.86600930760884, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 15354 + }, + { + "epoch": 0.15355, + "grad_norm": 1.1552154240052899, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 15355 + }, + { + "epoch": 0.15356, + "grad_norm": 1.273191111852812, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 15356 + }, + { + "epoch": 0.15357, + "grad_norm": 0.852807800719506, + "learning_rate": 0.003, + "loss": 4.085, + "step": 15357 + }, + { + "epoch": 0.15358, + "grad_norm": 0.9077294334590712, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 15358 + }, + { + "epoch": 0.15359, + "grad_norm": 1.1243123429854551, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 15359 + }, + { + "epoch": 0.1536, + "grad_norm": 0.9132311055508577, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 15360 + }, + { + "epoch": 0.15361, + "grad_norm": 0.9712530512893895, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 15361 + }, + { + "epoch": 0.15362, + "grad_norm": 1.160516427315112, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 15362 + }, + { + "epoch": 0.15363, + "grad_norm": 0.9326386690545608, + "learning_rate": 0.003, + "loss": 4.1013, + "step": 15363 + }, + { + "epoch": 0.15364, + "grad_norm": 0.8407466904821066, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 15364 + }, + { + "epoch": 0.15365, + "grad_norm": 1.0032382505708344, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 15365 + }, + { + "epoch": 0.15366, + "grad_norm": 1.3261182161794047, + "learning_rate": 0.003, + "loss": 4.1101, + "step": 15366 + }, + { + "epoch": 0.15367, + "grad_norm": 0.9159133954505527, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 15367 + }, + { + "epoch": 0.15368, + "grad_norm": 1.0963464142858823, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 15368 + }, + { + "epoch": 0.15369, + "grad_norm": 1.220459587036576, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 15369 + }, + { + "epoch": 0.1537, + "grad_norm": 1.0219890440118864, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 15370 + }, + { + "epoch": 0.15371, + "grad_norm": 0.9791393825357693, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 15371 + }, + { + "epoch": 0.15372, + "grad_norm": 0.908759269287215, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 15372 + }, + { + "epoch": 0.15373, + "grad_norm": 1.056356980941746, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 15373 + }, + { + "epoch": 0.15374, + "grad_norm": 1.1201510970114081, + "learning_rate": 0.003, + "loss": 4.081, + "step": 15374 + }, + { + "epoch": 0.15375, + "grad_norm": 0.9606021508070215, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 15375 + }, + { + "epoch": 0.15376, + "grad_norm": 1.0102057547930416, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 15376 + }, + { + "epoch": 0.15377, + "grad_norm": 1.0565704321540779, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 15377 + }, + { + "epoch": 0.15378, + "grad_norm": 0.9732337113760198, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 15378 + }, + { + "epoch": 0.15379, + "grad_norm": 1.1534841008330299, + "learning_rate": 0.003, + "loss": 4.1061, + "step": 15379 + }, + { + "epoch": 0.1538, + "grad_norm": 0.8907770082231602, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 15380 + }, + { + "epoch": 0.15381, + "grad_norm": 0.8960301165327065, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 15381 + }, + { + "epoch": 0.15382, + "grad_norm": 0.9195813065320044, + "learning_rate": 0.003, + "loss": 4.0855, + "step": 15382 + }, + { + "epoch": 0.15383, + "grad_norm": 1.0647802420396173, + "learning_rate": 0.003, + "loss": 4.1072, + "step": 15383 + }, + { + "epoch": 0.15384, + "grad_norm": 0.9219050331833214, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 15384 + }, + { + "epoch": 0.15385, + "grad_norm": 0.9369312343643442, + "learning_rate": 0.003, + "loss": 4.0964, + "step": 15385 + }, + { + "epoch": 0.15386, + "grad_norm": 1.0126675873896418, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 15386 + }, + { + "epoch": 0.15387, + "grad_norm": 0.9904735382134243, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 15387 + }, + { + "epoch": 0.15388, + "grad_norm": 1.0488288889271749, + "learning_rate": 0.003, + "loss": 4.0941, + "step": 15388 + }, + { + "epoch": 0.15389, + "grad_norm": 1.1679992410812816, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 15389 + }, + { + "epoch": 0.1539, + "grad_norm": 1.0804071793349492, + "learning_rate": 0.003, + "loss": 4.1055, + "step": 15390 + }, + { + "epoch": 0.15391, + "grad_norm": 0.8723524232356551, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 15391 + }, + { + "epoch": 0.15392, + "grad_norm": 0.8780297107371089, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 15392 + }, + { + "epoch": 0.15393, + "grad_norm": 1.0268824966068109, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 15393 + }, + { + "epoch": 0.15394, + "grad_norm": 1.1288252529296903, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 15394 + }, + { + "epoch": 0.15395, + "grad_norm": 0.903943010431595, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 15395 + }, + { + "epoch": 0.15396, + "grad_norm": 0.9366776264624694, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 15396 + }, + { + "epoch": 0.15397, + "grad_norm": 0.810098056430841, + "learning_rate": 0.003, + "loss": 4.0983, + "step": 15397 + }, + { + "epoch": 0.15398, + "grad_norm": 0.8048630390979554, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 15398 + }, + { + "epoch": 0.15399, + "grad_norm": 0.903839764538714, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 15399 + }, + { + "epoch": 0.154, + "grad_norm": 0.9758652347138065, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 15400 + }, + { + "epoch": 0.15401, + "grad_norm": 0.8161203290970186, + "learning_rate": 0.003, + "loss": 4.067, + "step": 15401 + }, + { + "epoch": 0.15402, + "grad_norm": 0.7865094197902369, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 15402 + }, + { + "epoch": 0.15403, + "grad_norm": 0.9219511845311912, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 15403 + }, + { + "epoch": 0.15404, + "grad_norm": 1.2114717328959215, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 15404 + }, + { + "epoch": 0.15405, + "grad_norm": 1.0835760415536337, + "learning_rate": 0.003, + "loss": 4.1107, + "step": 15405 + }, + { + "epoch": 0.15406, + "grad_norm": 1.181116558185889, + "learning_rate": 0.003, + "loss": 4.062, + "step": 15406 + }, + { + "epoch": 0.15407, + "grad_norm": 0.8833166694482677, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 15407 + }, + { + "epoch": 0.15408, + "grad_norm": 0.841505087715076, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 15408 + }, + { + "epoch": 0.15409, + "grad_norm": 0.8949007399694598, + "learning_rate": 0.003, + "loss": 4.0938, + "step": 15409 + }, + { + "epoch": 0.1541, + "grad_norm": 0.918140741833193, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 15410 + }, + { + "epoch": 0.15411, + "grad_norm": 0.8805553638202924, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 15411 + }, + { + "epoch": 0.15412, + "grad_norm": 0.9124201738217915, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 15412 + }, + { + "epoch": 0.15413, + "grad_norm": 0.9222183292066466, + "learning_rate": 0.003, + "loss": 4.1038, + "step": 15413 + }, + { + "epoch": 0.15414, + "grad_norm": 1.041015283544823, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 15414 + }, + { + "epoch": 0.15415, + "grad_norm": 1.1360784092856422, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 15415 + }, + { + "epoch": 0.15416, + "grad_norm": 0.9805166509801004, + "learning_rate": 0.003, + "loss": 4.0886, + "step": 15416 + }, + { + "epoch": 0.15417, + "grad_norm": 1.1536695190292634, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 15417 + }, + { + "epoch": 0.15418, + "grad_norm": 0.9693517696411421, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 15418 + }, + { + "epoch": 0.15419, + "grad_norm": 0.8625184693514797, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 15419 + }, + { + "epoch": 0.1542, + "grad_norm": 0.8435725675422313, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 15420 + }, + { + "epoch": 0.15421, + "grad_norm": 0.9769653494678785, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 15421 + }, + { + "epoch": 0.15422, + "grad_norm": 1.1955814103831661, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 15422 + }, + { + "epoch": 0.15423, + "grad_norm": 0.9122588471340036, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 15423 + }, + { + "epoch": 0.15424, + "grad_norm": 0.9359750644660203, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 15424 + }, + { + "epoch": 0.15425, + "grad_norm": 0.98950858913703, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 15425 + }, + { + "epoch": 0.15426, + "grad_norm": 1.074344563157277, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 15426 + }, + { + "epoch": 0.15427, + "grad_norm": 0.9234935195137572, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 15427 + }, + { + "epoch": 0.15428, + "grad_norm": 0.8428937136537137, + "learning_rate": 0.003, + "loss": 4.0848, + "step": 15428 + }, + { + "epoch": 0.15429, + "grad_norm": 0.8589133082239022, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 15429 + }, + { + "epoch": 0.1543, + "grad_norm": 0.8126811175867564, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 15430 + }, + { + "epoch": 0.15431, + "grad_norm": 0.8877292170237443, + "learning_rate": 0.003, + "loss": 4.0917, + "step": 15431 + }, + { + "epoch": 0.15432, + "grad_norm": 1.0544103796448516, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 15432 + }, + { + "epoch": 0.15433, + "grad_norm": 1.0591434706084677, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 15433 + }, + { + "epoch": 0.15434, + "grad_norm": 1.0380190805539298, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 15434 + }, + { + "epoch": 0.15435, + "grad_norm": 0.9744891682597306, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 15435 + }, + { + "epoch": 0.15436, + "grad_norm": 0.8813186763636935, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 15436 + }, + { + "epoch": 0.15437, + "grad_norm": 0.8788289099915375, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 15437 + }, + { + "epoch": 0.15438, + "grad_norm": 0.8330340635990507, + "learning_rate": 0.003, + "loss": 4.094, + "step": 15438 + }, + { + "epoch": 0.15439, + "grad_norm": 0.8114241503431349, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 15439 + }, + { + "epoch": 0.1544, + "grad_norm": 0.9366586935257486, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 15440 + }, + { + "epoch": 0.15441, + "grad_norm": 1.1166711048496927, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 15441 + }, + { + "epoch": 0.15442, + "grad_norm": 0.917193898780145, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 15442 + }, + { + "epoch": 0.15443, + "grad_norm": 0.9703454785476748, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 15443 + }, + { + "epoch": 0.15444, + "grad_norm": 1.1778675473323017, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 15444 + }, + { + "epoch": 0.15445, + "grad_norm": 1.0539791058200725, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 15445 + }, + { + "epoch": 0.15446, + "grad_norm": 1.1598388101063284, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 15446 + }, + { + "epoch": 0.15447, + "grad_norm": 1.0374071350630174, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 15447 + }, + { + "epoch": 0.15448, + "grad_norm": 1.1091063480355123, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 15448 + }, + { + "epoch": 0.15449, + "grad_norm": 1.024267631156153, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 15449 + }, + { + "epoch": 0.1545, + "grad_norm": 1.022105535080398, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 15450 + }, + { + "epoch": 0.15451, + "grad_norm": 1.0875364584534721, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 15451 + }, + { + "epoch": 0.15452, + "grad_norm": 0.9215275300306409, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 15452 + }, + { + "epoch": 0.15453, + "grad_norm": 0.8978001207229455, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 15453 + }, + { + "epoch": 0.15454, + "grad_norm": 1.035466871117155, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 15454 + }, + { + "epoch": 0.15455, + "grad_norm": 1.174625152330979, + "learning_rate": 0.003, + "loss": 4.106, + "step": 15455 + }, + { + "epoch": 0.15456, + "grad_norm": 1.0283553066260123, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 15456 + }, + { + "epoch": 0.15457, + "grad_norm": 1.2287148665694012, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 15457 + }, + { + "epoch": 0.15458, + "grad_norm": 1.1364786217919942, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 15458 + }, + { + "epoch": 0.15459, + "grad_norm": 1.018120455159289, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 15459 + }, + { + "epoch": 0.1546, + "grad_norm": 1.111822144741687, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 15460 + }, + { + "epoch": 0.15461, + "grad_norm": 0.923763371179255, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 15461 + }, + { + "epoch": 0.15462, + "grad_norm": 0.8407406567847826, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 15462 + }, + { + "epoch": 0.15463, + "grad_norm": 0.8205550707802245, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 15463 + }, + { + "epoch": 0.15464, + "grad_norm": 0.8612777951636056, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 15464 + }, + { + "epoch": 0.15465, + "grad_norm": 0.8549000066057266, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 15465 + }, + { + "epoch": 0.15466, + "grad_norm": 0.9190986969849092, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 15466 + }, + { + "epoch": 0.15467, + "grad_norm": 0.883513412539122, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 15467 + }, + { + "epoch": 0.15468, + "grad_norm": 0.9312961350981269, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 15468 + }, + { + "epoch": 0.15469, + "grad_norm": 0.9136219198248873, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 15469 + }, + { + "epoch": 0.1547, + "grad_norm": 1.0305191111688252, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 15470 + }, + { + "epoch": 0.15471, + "grad_norm": 1.2095496338603433, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 15471 + }, + { + "epoch": 0.15472, + "grad_norm": 0.9477723442716905, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 15472 + }, + { + "epoch": 0.15473, + "grad_norm": 1.1758489163239902, + "learning_rate": 0.003, + "loss": 4.0986, + "step": 15473 + }, + { + "epoch": 0.15474, + "grad_norm": 1.0970583591322434, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 15474 + }, + { + "epoch": 0.15475, + "grad_norm": 0.9050250698153368, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 15475 + }, + { + "epoch": 0.15476, + "grad_norm": 1.0215399708305442, + "learning_rate": 0.003, + "loss": 4.0947, + "step": 15476 + }, + { + "epoch": 0.15477, + "grad_norm": 1.140823830229048, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 15477 + }, + { + "epoch": 0.15478, + "grad_norm": 0.8484231326142477, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 15478 + }, + { + "epoch": 0.15479, + "grad_norm": 0.8056268583193773, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 15479 + }, + { + "epoch": 0.1548, + "grad_norm": 0.7258919572004691, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 15480 + }, + { + "epoch": 0.15481, + "grad_norm": 0.7683009997760593, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 15481 + }, + { + "epoch": 0.15482, + "grad_norm": 0.9268624660012709, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 15482 + }, + { + "epoch": 0.15483, + "grad_norm": 1.2006913743967302, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 15483 + }, + { + "epoch": 0.15484, + "grad_norm": 0.780585591974057, + "learning_rate": 0.003, + "loss": 4.1, + "step": 15484 + }, + { + "epoch": 0.15485, + "grad_norm": 0.7841627518597984, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 15485 + }, + { + "epoch": 0.15486, + "grad_norm": 0.7874299393520852, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 15486 + }, + { + "epoch": 0.15487, + "grad_norm": 0.9112927270844453, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 15487 + }, + { + "epoch": 0.15488, + "grad_norm": 0.8344448964945357, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 15488 + }, + { + "epoch": 0.15489, + "grad_norm": 0.7773481743460254, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 15489 + }, + { + "epoch": 0.1549, + "grad_norm": 0.7201431092334954, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 15490 + }, + { + "epoch": 0.15491, + "grad_norm": 0.7396542441612265, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 15491 + }, + { + "epoch": 0.15492, + "grad_norm": 0.8331504995780095, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 15492 + }, + { + "epoch": 0.15493, + "grad_norm": 0.9716161740308155, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 15493 + }, + { + "epoch": 0.15494, + "grad_norm": 1.4194825332958338, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 15494 + }, + { + "epoch": 0.15495, + "grad_norm": 0.9745051062659559, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 15495 + }, + { + "epoch": 0.15496, + "grad_norm": 1.1380022261423877, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 15496 + }, + { + "epoch": 0.15497, + "grad_norm": 0.9492292277451831, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 15497 + }, + { + "epoch": 0.15498, + "grad_norm": 1.0785876040208742, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 15498 + }, + { + "epoch": 0.15499, + "grad_norm": 0.981168725995321, + "learning_rate": 0.003, + "loss": 4.06, + "step": 15499 + }, + { + "epoch": 0.155, + "grad_norm": 1.0667126333586598, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 15500 + }, + { + "epoch": 0.15501, + "grad_norm": 1.0954098078242576, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 15501 + }, + { + "epoch": 0.15502, + "grad_norm": 0.8872877341389628, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 15502 + }, + { + "epoch": 0.15503, + "grad_norm": 0.8143446512466253, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 15503 + }, + { + "epoch": 0.15504, + "grad_norm": 0.8574641578162346, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 15504 + }, + { + "epoch": 0.15505, + "grad_norm": 1.0249757120406207, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 15505 + }, + { + "epoch": 0.15506, + "grad_norm": 1.0893781636364817, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 15506 + }, + { + "epoch": 0.15507, + "grad_norm": 0.8688034827902502, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 15507 + }, + { + "epoch": 0.15508, + "grad_norm": 0.9415575515923393, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 15508 + }, + { + "epoch": 0.15509, + "grad_norm": 1.0149119132944726, + "learning_rate": 0.003, + "loss": 4.048, + "step": 15509 + }, + { + "epoch": 0.1551, + "grad_norm": 0.9200699152403218, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 15510 + }, + { + "epoch": 0.15511, + "grad_norm": 1.2142929186334814, + "learning_rate": 0.003, + "loss": 4.1094, + "step": 15511 + }, + { + "epoch": 0.15512, + "grad_norm": 1.0927606063741895, + "learning_rate": 0.003, + "loss": 4.1049, + "step": 15512 + }, + { + "epoch": 0.15513, + "grad_norm": 0.9313775802587372, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 15513 + }, + { + "epoch": 0.15514, + "grad_norm": 1.0439859961348552, + "learning_rate": 0.003, + "loss": 4.1098, + "step": 15514 + }, + { + "epoch": 0.15515, + "grad_norm": 1.1335963627177066, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 15515 + }, + { + "epoch": 0.15516, + "grad_norm": 1.2066111313771402, + "learning_rate": 0.003, + "loss": 4.0943, + "step": 15516 + }, + { + "epoch": 0.15517, + "grad_norm": 1.060713797943098, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 15517 + }, + { + "epoch": 0.15518, + "grad_norm": 1.3211331102982031, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 15518 + }, + { + "epoch": 0.15519, + "grad_norm": 1.0604578748391424, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 15519 + }, + { + "epoch": 0.1552, + "grad_norm": 0.9842507952937842, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 15520 + }, + { + "epoch": 0.15521, + "grad_norm": 1.063400642324669, + "learning_rate": 0.003, + "loss": 4.1028, + "step": 15521 + }, + { + "epoch": 0.15522, + "grad_norm": 1.0233792437550595, + "learning_rate": 0.003, + "loss": 4.0988, + "step": 15522 + }, + { + "epoch": 0.15523, + "grad_norm": 1.0642642320694087, + "learning_rate": 0.003, + "loss": 4.0901, + "step": 15523 + }, + { + "epoch": 0.15524, + "grad_norm": 0.9959299692728968, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 15524 + }, + { + "epoch": 0.15525, + "grad_norm": 1.0222072398055257, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 15525 + }, + { + "epoch": 0.15526, + "grad_norm": 1.1422284380680467, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 15526 + }, + { + "epoch": 0.15527, + "grad_norm": 1.172274888877104, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 15527 + }, + { + "epoch": 0.15528, + "grad_norm": 0.8367644007674544, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 15528 + }, + { + "epoch": 0.15529, + "grad_norm": 0.8532374516942376, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 15529 + }, + { + "epoch": 0.1553, + "grad_norm": 0.8065984511548429, + "learning_rate": 0.003, + "loss": 4.0952, + "step": 15530 + }, + { + "epoch": 0.15531, + "grad_norm": 0.7542840592192782, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 15531 + }, + { + "epoch": 0.15532, + "grad_norm": 0.7694208141340712, + "learning_rate": 0.003, + "loss": 4.0985, + "step": 15532 + }, + { + "epoch": 0.15533, + "grad_norm": 0.7458906193188874, + "learning_rate": 0.003, + "loss": 4.079, + "step": 15533 + }, + { + "epoch": 0.15534, + "grad_norm": 0.8268219367153778, + "learning_rate": 0.003, + "loss": 4.0887, + "step": 15534 + }, + { + "epoch": 0.15535, + "grad_norm": 0.8431401427569836, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 15535 + }, + { + "epoch": 0.15536, + "grad_norm": 0.972708654555106, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 15536 + }, + { + "epoch": 0.15537, + "grad_norm": 1.3860304206396736, + "learning_rate": 0.003, + "loss": 4.0968, + "step": 15537 + }, + { + "epoch": 0.15538, + "grad_norm": 0.8611568925733759, + "learning_rate": 0.003, + "loss": 4.061, + "step": 15538 + }, + { + "epoch": 0.15539, + "grad_norm": 0.885850609753444, + "learning_rate": 0.003, + "loss": 4.1018, + "step": 15539 + }, + { + "epoch": 0.1554, + "grad_norm": 0.8964197895334711, + "learning_rate": 0.003, + "loss": 4.1048, + "step": 15540 + }, + { + "epoch": 0.15541, + "grad_norm": 0.9353274443414468, + "learning_rate": 0.003, + "loss": 4.0933, + "step": 15541 + }, + { + "epoch": 0.15542, + "grad_norm": 0.9660532172030759, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 15542 + }, + { + "epoch": 0.15543, + "grad_norm": 1.0282469130396499, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 15543 + }, + { + "epoch": 0.15544, + "grad_norm": 0.9213450069041198, + "learning_rate": 0.003, + "loss": 4.1164, + "step": 15544 + }, + { + "epoch": 0.15545, + "grad_norm": 0.9118840799495626, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 15545 + }, + { + "epoch": 0.15546, + "grad_norm": 0.9063750575502594, + "learning_rate": 0.003, + "loss": 4.0947, + "step": 15546 + }, + { + "epoch": 0.15547, + "grad_norm": 0.9559735464173067, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 15547 + }, + { + "epoch": 0.15548, + "grad_norm": 1.0724841504794178, + "learning_rate": 0.003, + "loss": 4.0899, + "step": 15548 + }, + { + "epoch": 0.15549, + "grad_norm": 1.0916726946982145, + "learning_rate": 0.003, + "loss": 4.0933, + "step": 15549 + }, + { + "epoch": 0.1555, + "grad_norm": 1.0876851084367638, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 15550 + }, + { + "epoch": 0.15551, + "grad_norm": 1.029156927930153, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 15551 + }, + { + "epoch": 0.15552, + "grad_norm": 0.9769084625077543, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 15552 + }, + { + "epoch": 0.15553, + "grad_norm": 1.2630677613501473, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 15553 + }, + { + "epoch": 0.15554, + "grad_norm": 0.9935393973312288, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 15554 + }, + { + "epoch": 0.15555, + "grad_norm": 1.1182480634042322, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 15555 + }, + { + "epoch": 0.15556, + "grad_norm": 1.0028943921605844, + "learning_rate": 0.003, + "loss": 4.05, + "step": 15556 + }, + { + "epoch": 0.15557, + "grad_norm": 1.031650535918803, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 15557 + }, + { + "epoch": 0.15558, + "grad_norm": 0.8574996567006572, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 15558 + }, + { + "epoch": 0.15559, + "grad_norm": 0.8196269783827448, + "learning_rate": 0.003, + "loss": 4.058, + "step": 15559 + }, + { + "epoch": 0.1556, + "grad_norm": 0.9449173245445136, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 15560 + }, + { + "epoch": 0.15561, + "grad_norm": 1.0812195015960522, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 15561 + }, + { + "epoch": 0.15562, + "grad_norm": 0.9276659623865838, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 15562 + }, + { + "epoch": 0.15563, + "grad_norm": 0.8483703780399338, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 15563 + }, + { + "epoch": 0.15564, + "grad_norm": 1.0716685570120421, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 15564 + }, + { + "epoch": 0.15565, + "grad_norm": 1.299967868229868, + "learning_rate": 0.003, + "loss": 4.063, + "step": 15565 + }, + { + "epoch": 0.15566, + "grad_norm": 0.8799250742242631, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 15566 + }, + { + "epoch": 0.15567, + "grad_norm": 0.9437911066808506, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 15567 + }, + { + "epoch": 0.15568, + "grad_norm": 0.9106163988273723, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 15568 + }, + { + "epoch": 0.15569, + "grad_norm": 1.1263209510135577, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 15569 + }, + { + "epoch": 0.1557, + "grad_norm": 1.0083636991380474, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 15570 + }, + { + "epoch": 0.15571, + "grad_norm": 0.9972213580343325, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 15571 + }, + { + "epoch": 0.15572, + "grad_norm": 1.1597338555056727, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 15572 + }, + { + "epoch": 0.15573, + "grad_norm": 0.7808542350836801, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 15573 + }, + { + "epoch": 0.15574, + "grad_norm": 0.8409596027160338, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 15574 + }, + { + "epoch": 0.15575, + "grad_norm": 1.02099676767791, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 15575 + }, + { + "epoch": 0.15576, + "grad_norm": 1.1035902289708284, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 15576 + }, + { + "epoch": 0.15577, + "grad_norm": 0.9176682805721234, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 15577 + }, + { + "epoch": 0.15578, + "grad_norm": 0.9974674635466865, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 15578 + }, + { + "epoch": 0.15579, + "grad_norm": 1.200320904242632, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 15579 + }, + { + "epoch": 0.1558, + "grad_norm": 1.0353886537668897, + "learning_rate": 0.003, + "loss": 4.0961, + "step": 15580 + }, + { + "epoch": 0.15581, + "grad_norm": 1.1336398933825564, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 15581 + }, + { + "epoch": 0.15582, + "grad_norm": 0.7831394371132027, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 15582 + }, + { + "epoch": 0.15583, + "grad_norm": 0.6618318022082501, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 15583 + }, + { + "epoch": 0.15584, + "grad_norm": 0.7663745630037706, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 15584 + }, + { + "epoch": 0.15585, + "grad_norm": 1.0508543706363356, + "learning_rate": 0.003, + "loss": 4.0961, + "step": 15585 + }, + { + "epoch": 0.15586, + "grad_norm": 1.361916738617122, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 15586 + }, + { + "epoch": 0.15587, + "grad_norm": 0.7896716322850271, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 15587 + }, + { + "epoch": 0.15588, + "grad_norm": 0.758228694596777, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 15588 + }, + { + "epoch": 0.15589, + "grad_norm": 1.0083028224704782, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 15589 + }, + { + "epoch": 0.1559, + "grad_norm": 1.368247502179751, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 15590 + }, + { + "epoch": 0.15591, + "grad_norm": 0.9288627763665291, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 15591 + }, + { + "epoch": 0.15592, + "grad_norm": 0.8628065837429335, + "learning_rate": 0.003, + "loss": 4.0922, + "step": 15592 + }, + { + "epoch": 0.15593, + "grad_norm": 0.9147951765351533, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 15593 + }, + { + "epoch": 0.15594, + "grad_norm": 0.8663364725881738, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 15594 + }, + { + "epoch": 0.15595, + "grad_norm": 0.8289293952352027, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 15595 + }, + { + "epoch": 0.15596, + "grad_norm": 1.0239553546347369, + "learning_rate": 0.003, + "loss": 4.047, + "step": 15596 + }, + { + "epoch": 0.15597, + "grad_norm": 1.1656736244883892, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 15597 + }, + { + "epoch": 0.15598, + "grad_norm": 0.9810166594408865, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 15598 + }, + { + "epoch": 0.15599, + "grad_norm": 0.8973430277090985, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 15599 + }, + { + "epoch": 0.156, + "grad_norm": 0.9222944262566568, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 15600 + }, + { + "epoch": 0.15601, + "grad_norm": 0.9886092243411647, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 15601 + }, + { + "epoch": 0.15602, + "grad_norm": 1.2831529499440761, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 15602 + }, + { + "epoch": 0.15603, + "grad_norm": 0.7925999707308636, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 15603 + }, + { + "epoch": 0.15604, + "grad_norm": 0.7341546238098774, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 15604 + }, + { + "epoch": 0.15605, + "grad_norm": 1.0527606763944237, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 15605 + }, + { + "epoch": 0.15606, + "grad_norm": 1.1598580933619764, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 15606 + }, + { + "epoch": 0.15607, + "grad_norm": 0.7982284530842869, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 15607 + }, + { + "epoch": 0.15608, + "grad_norm": 0.7456188629210454, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 15608 + }, + { + "epoch": 0.15609, + "grad_norm": 0.7781499039867238, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 15609 + }, + { + "epoch": 0.1561, + "grad_norm": 0.8823057774008047, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 15610 + }, + { + "epoch": 0.15611, + "grad_norm": 0.9506956249491159, + "learning_rate": 0.003, + "loss": 4.079, + "step": 15611 + }, + { + "epoch": 0.15612, + "grad_norm": 1.2279114061392926, + "learning_rate": 0.003, + "loss": 4.052, + "step": 15612 + }, + { + "epoch": 0.15613, + "grad_norm": 1.0941329307046597, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 15613 + }, + { + "epoch": 0.15614, + "grad_norm": 1.1805575677144093, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 15614 + }, + { + "epoch": 0.15615, + "grad_norm": 0.8731237175142837, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 15615 + }, + { + "epoch": 0.15616, + "grad_norm": 0.8391889087046885, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 15616 + }, + { + "epoch": 0.15617, + "grad_norm": 0.720151360192541, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 15617 + }, + { + "epoch": 0.15618, + "grad_norm": 0.7623583775581732, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 15618 + }, + { + "epoch": 0.15619, + "grad_norm": 0.7184160832013416, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 15619 + }, + { + "epoch": 0.1562, + "grad_norm": 0.7419152795294724, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 15620 + }, + { + "epoch": 0.15621, + "grad_norm": 0.8553684418602074, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 15621 + }, + { + "epoch": 0.15622, + "grad_norm": 1.0887262428164741, + "learning_rate": 0.003, + "loss": 4.0893, + "step": 15622 + }, + { + "epoch": 0.15623, + "grad_norm": 1.0763302681935476, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 15623 + }, + { + "epoch": 0.15624, + "grad_norm": 1.1253645012283922, + "learning_rate": 0.003, + "loss": 4.0947, + "step": 15624 + }, + { + "epoch": 0.15625, + "grad_norm": 1.0326910088051524, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 15625 + }, + { + "epoch": 0.15626, + "grad_norm": 1.2105996590456285, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 15626 + }, + { + "epoch": 0.15627, + "grad_norm": 1.016848242198179, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 15627 + }, + { + "epoch": 0.15628, + "grad_norm": 1.2115349210127289, + "learning_rate": 0.003, + "loss": 4.049, + "step": 15628 + }, + { + "epoch": 0.15629, + "grad_norm": 0.9577991918662226, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 15629 + }, + { + "epoch": 0.1563, + "grad_norm": 1.1012582045092247, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 15630 + }, + { + "epoch": 0.15631, + "grad_norm": 1.0197708007585846, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 15631 + }, + { + "epoch": 0.15632, + "grad_norm": 1.0601986707886895, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 15632 + }, + { + "epoch": 0.15633, + "grad_norm": 0.9706409124723919, + "learning_rate": 0.003, + "loss": 4.0922, + "step": 15633 + }, + { + "epoch": 0.15634, + "grad_norm": 1.0662839751875277, + "learning_rate": 0.003, + "loss": 4.094, + "step": 15634 + }, + { + "epoch": 0.15635, + "grad_norm": 1.0824685283770963, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 15635 + }, + { + "epoch": 0.15636, + "grad_norm": 1.023988213606166, + "learning_rate": 0.003, + "loss": 4.069, + "step": 15636 + }, + { + "epoch": 0.15637, + "grad_norm": 1.084191255274759, + "learning_rate": 0.003, + "loss": 4.1015, + "step": 15637 + }, + { + "epoch": 0.15638, + "grad_norm": 0.884452878923361, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 15638 + }, + { + "epoch": 0.15639, + "grad_norm": 0.866114174844579, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 15639 + }, + { + "epoch": 0.1564, + "grad_norm": 1.0030452634829672, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 15640 + }, + { + "epoch": 0.15641, + "grad_norm": 1.0225658646843288, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 15641 + }, + { + "epoch": 0.15642, + "grad_norm": 1.0287841411077103, + "learning_rate": 0.003, + "loss": 4.0898, + "step": 15642 + }, + { + "epoch": 0.15643, + "grad_norm": 1.0381716518264774, + "learning_rate": 0.003, + "loss": 4.0965, + "step": 15643 + }, + { + "epoch": 0.15644, + "grad_norm": 1.3628938026913453, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 15644 + }, + { + "epoch": 0.15645, + "grad_norm": 0.849049953361216, + "learning_rate": 0.003, + "loss": 4.0907, + "step": 15645 + }, + { + "epoch": 0.15646, + "grad_norm": 0.7888509209199441, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 15646 + }, + { + "epoch": 0.15647, + "grad_norm": 0.9517556285470536, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 15647 + }, + { + "epoch": 0.15648, + "grad_norm": 1.0703557487837057, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 15648 + }, + { + "epoch": 0.15649, + "grad_norm": 1.0540058340176441, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 15649 + }, + { + "epoch": 0.1565, + "grad_norm": 0.8931083033657232, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 15650 + }, + { + "epoch": 0.15651, + "grad_norm": 0.7907580864763839, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 15651 + }, + { + "epoch": 0.15652, + "grad_norm": 0.7259843030666423, + "learning_rate": 0.003, + "loss": 4.0908, + "step": 15652 + }, + { + "epoch": 0.15653, + "grad_norm": 0.8589736845240897, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 15653 + }, + { + "epoch": 0.15654, + "grad_norm": 0.9725754499760675, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 15654 + }, + { + "epoch": 0.15655, + "grad_norm": 1.091288272629763, + "learning_rate": 0.003, + "loss": 4.1168, + "step": 15655 + }, + { + "epoch": 0.15656, + "grad_norm": 0.808197551886759, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 15656 + }, + { + "epoch": 0.15657, + "grad_norm": 0.7953074307201199, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 15657 + }, + { + "epoch": 0.15658, + "grad_norm": 0.8811138612615538, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 15658 + }, + { + "epoch": 0.15659, + "grad_norm": 0.8680413077808393, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 15659 + }, + { + "epoch": 0.1566, + "grad_norm": 1.081418438209925, + "learning_rate": 0.003, + "loss": 4.0886, + "step": 15660 + }, + { + "epoch": 0.15661, + "grad_norm": 1.3627123388723321, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 15661 + }, + { + "epoch": 0.15662, + "grad_norm": 1.0362785074009948, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 15662 + }, + { + "epoch": 0.15663, + "grad_norm": 1.2054088985143807, + "learning_rate": 0.003, + "loss": 4.072, + "step": 15663 + }, + { + "epoch": 0.15664, + "grad_norm": 1.0133508768222843, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 15664 + }, + { + "epoch": 0.15665, + "grad_norm": 0.9292342618200954, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 15665 + }, + { + "epoch": 0.15666, + "grad_norm": 0.9720688066662803, + "learning_rate": 0.003, + "loss": 4.058, + "step": 15666 + }, + { + "epoch": 0.15667, + "grad_norm": 1.2212922603396992, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 15667 + }, + { + "epoch": 0.15668, + "grad_norm": 0.9156469678705218, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 15668 + }, + { + "epoch": 0.15669, + "grad_norm": 0.9395771375048656, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 15669 + }, + { + "epoch": 0.1567, + "grad_norm": 1.11354458818089, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 15670 + }, + { + "epoch": 0.15671, + "grad_norm": 0.9645274204286521, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 15671 + }, + { + "epoch": 0.15672, + "grad_norm": 0.9673878396971555, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 15672 + }, + { + "epoch": 0.15673, + "grad_norm": 0.9807899645997915, + "learning_rate": 0.003, + "loss": 4.0908, + "step": 15673 + }, + { + "epoch": 0.15674, + "grad_norm": 1.008377398258221, + "learning_rate": 0.003, + "loss": 4.074, + "step": 15674 + }, + { + "epoch": 0.15675, + "grad_norm": 1.1075492494763919, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 15675 + }, + { + "epoch": 0.15676, + "grad_norm": 0.9766587466748744, + "learning_rate": 0.003, + "loss": 4.088, + "step": 15676 + }, + { + "epoch": 0.15677, + "grad_norm": 0.9499413781349715, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 15677 + }, + { + "epoch": 0.15678, + "grad_norm": 1.027524103150609, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 15678 + }, + { + "epoch": 0.15679, + "grad_norm": 0.9910420361988583, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 15679 + }, + { + "epoch": 0.1568, + "grad_norm": 1.0400830620531045, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 15680 + }, + { + "epoch": 0.15681, + "grad_norm": 0.9044354467750185, + "learning_rate": 0.003, + "loss": 4.1028, + "step": 15681 + }, + { + "epoch": 0.15682, + "grad_norm": 0.7913390476947003, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 15682 + }, + { + "epoch": 0.15683, + "grad_norm": 0.8685126582880774, + "learning_rate": 0.003, + "loss": 4.064, + "step": 15683 + }, + { + "epoch": 0.15684, + "grad_norm": 0.9026056875604399, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 15684 + }, + { + "epoch": 0.15685, + "grad_norm": 0.9855644938751627, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 15685 + }, + { + "epoch": 0.15686, + "grad_norm": 1.0146692737958032, + "learning_rate": 0.003, + "loss": 4.0963, + "step": 15686 + }, + { + "epoch": 0.15687, + "grad_norm": 1.1210776534892328, + "learning_rate": 0.003, + "loss": 4.0992, + "step": 15687 + }, + { + "epoch": 0.15688, + "grad_norm": 0.9217257311745487, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 15688 + }, + { + "epoch": 0.15689, + "grad_norm": 0.9977763610499195, + "learning_rate": 0.003, + "loss": 4.0985, + "step": 15689 + }, + { + "epoch": 0.1569, + "grad_norm": 1.2338128288610872, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 15690 + }, + { + "epoch": 0.15691, + "grad_norm": 1.1950713768196048, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 15691 + }, + { + "epoch": 0.15692, + "grad_norm": 1.220830453481974, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 15692 + }, + { + "epoch": 0.15693, + "grad_norm": 1.136218851969061, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 15693 + }, + { + "epoch": 0.15694, + "grad_norm": 0.884948615634766, + "learning_rate": 0.003, + "loss": 4.0899, + "step": 15694 + }, + { + "epoch": 0.15695, + "grad_norm": 0.8676457109321565, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 15695 + }, + { + "epoch": 0.15696, + "grad_norm": 0.8694110590849428, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 15696 + }, + { + "epoch": 0.15697, + "grad_norm": 0.9350222646409588, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 15697 + }, + { + "epoch": 0.15698, + "grad_norm": 1.3242872514504263, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 15698 + }, + { + "epoch": 0.15699, + "grad_norm": 0.8527533079068585, + "learning_rate": 0.003, + "loss": 4.1077, + "step": 15699 + }, + { + "epoch": 0.157, + "grad_norm": 0.6730975992087191, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 15700 + }, + { + "epoch": 0.15701, + "grad_norm": 0.7990138147786353, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 15701 + }, + { + "epoch": 0.15702, + "grad_norm": 0.9375421610035254, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 15702 + }, + { + "epoch": 0.15703, + "grad_norm": 1.0757931979785071, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 15703 + }, + { + "epoch": 0.15704, + "grad_norm": 1.027370317151545, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 15704 + }, + { + "epoch": 0.15705, + "grad_norm": 0.9708795349932001, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 15705 + }, + { + "epoch": 0.15706, + "grad_norm": 1.2041895536497635, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 15706 + }, + { + "epoch": 0.15707, + "grad_norm": 0.9409046903212518, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 15707 + }, + { + "epoch": 0.15708, + "grad_norm": 0.968894037296606, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 15708 + }, + { + "epoch": 0.15709, + "grad_norm": 1.0283730244609937, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 15709 + }, + { + "epoch": 0.1571, + "grad_norm": 1.1056247018722127, + "learning_rate": 0.003, + "loss": 4.0942, + "step": 15710 + }, + { + "epoch": 0.15711, + "grad_norm": 0.8083000414150083, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 15711 + }, + { + "epoch": 0.15712, + "grad_norm": 0.7492468676077301, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 15712 + }, + { + "epoch": 0.15713, + "grad_norm": 0.9663306932812636, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 15713 + }, + { + "epoch": 0.15714, + "grad_norm": 1.0300027377595962, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 15714 + }, + { + "epoch": 0.15715, + "grad_norm": 0.9135879494400162, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 15715 + }, + { + "epoch": 0.15716, + "grad_norm": 0.9293065800185889, + "learning_rate": 0.003, + "loss": 4.0943, + "step": 15716 + }, + { + "epoch": 0.15717, + "grad_norm": 1.1067556358362585, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 15717 + }, + { + "epoch": 0.15718, + "grad_norm": 1.0555009230781707, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 15718 + }, + { + "epoch": 0.15719, + "grad_norm": 1.31793945920362, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 15719 + }, + { + "epoch": 0.1572, + "grad_norm": 0.9743664628940539, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 15720 + }, + { + "epoch": 0.15721, + "grad_norm": 0.9304107571397547, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 15721 + }, + { + "epoch": 0.15722, + "grad_norm": 0.9372042765045638, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 15722 + }, + { + "epoch": 0.15723, + "grad_norm": 1.1249830936713836, + "learning_rate": 0.003, + "loss": 4.0959, + "step": 15723 + }, + { + "epoch": 0.15724, + "grad_norm": 1.1781422948112896, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 15724 + }, + { + "epoch": 0.15725, + "grad_norm": 0.7438843131582263, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 15725 + }, + { + "epoch": 0.15726, + "grad_norm": 0.6798463535338952, + "learning_rate": 0.003, + "loss": 4.041, + "step": 15726 + }, + { + "epoch": 0.15727, + "grad_norm": 0.7260163979586804, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 15727 + }, + { + "epoch": 0.15728, + "grad_norm": 0.8178872056606141, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 15728 + }, + { + "epoch": 0.15729, + "grad_norm": 1.0458419502983076, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 15729 + }, + { + "epoch": 0.1573, + "grad_norm": 0.9679393846439166, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 15730 + }, + { + "epoch": 0.15731, + "grad_norm": 1.0443561062948625, + "learning_rate": 0.003, + "loss": 4.058, + "step": 15731 + }, + { + "epoch": 0.15732, + "grad_norm": 0.9644431959211367, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 15732 + }, + { + "epoch": 0.15733, + "grad_norm": 1.1623789583343551, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 15733 + }, + { + "epoch": 0.15734, + "grad_norm": 1.042215773402187, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 15734 + }, + { + "epoch": 0.15735, + "grad_norm": 1.144712188604885, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 15735 + }, + { + "epoch": 0.15736, + "grad_norm": 1.130929720256891, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 15736 + }, + { + "epoch": 0.15737, + "grad_norm": 1.0324500304523194, + "learning_rate": 0.003, + "loss": 4.0899, + "step": 15737 + }, + { + "epoch": 0.15738, + "grad_norm": 1.0334595247008718, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 15738 + }, + { + "epoch": 0.15739, + "grad_norm": 0.8807346105845321, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 15739 + }, + { + "epoch": 0.1574, + "grad_norm": 0.8747275497308097, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 15740 + }, + { + "epoch": 0.15741, + "grad_norm": 0.9617298933731302, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 15741 + }, + { + "epoch": 0.15742, + "grad_norm": 1.18997436300891, + "learning_rate": 0.003, + "loss": 4.063, + "step": 15742 + }, + { + "epoch": 0.15743, + "grad_norm": 0.9510166852619638, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 15743 + }, + { + "epoch": 0.15744, + "grad_norm": 1.0090304296709263, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 15744 + }, + { + "epoch": 0.15745, + "grad_norm": 1.2557842709841145, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 15745 + }, + { + "epoch": 0.15746, + "grad_norm": 0.7623661558322513, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 15746 + }, + { + "epoch": 0.15747, + "grad_norm": 0.7500895930860231, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 15747 + }, + { + "epoch": 0.15748, + "grad_norm": 0.7705917581647724, + "learning_rate": 0.003, + "loss": 4.058, + "step": 15748 + }, + { + "epoch": 0.15749, + "grad_norm": 0.8537996257062574, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 15749 + }, + { + "epoch": 0.1575, + "grad_norm": 1.0589535761995443, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 15750 + }, + { + "epoch": 0.15751, + "grad_norm": 1.1403439387770784, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 15751 + }, + { + "epoch": 0.15752, + "grad_norm": 0.961519231017326, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 15752 + }, + { + "epoch": 0.15753, + "grad_norm": 1.057835548979928, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 15753 + }, + { + "epoch": 0.15754, + "grad_norm": 0.9669954453646754, + "learning_rate": 0.003, + "loss": 4.082, + "step": 15754 + }, + { + "epoch": 0.15755, + "grad_norm": 1.017303463928552, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 15755 + }, + { + "epoch": 0.15756, + "grad_norm": 1.0935053673695136, + "learning_rate": 0.003, + "loss": 4.0993, + "step": 15756 + }, + { + "epoch": 0.15757, + "grad_norm": 1.0361940703944021, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 15757 + }, + { + "epoch": 0.15758, + "grad_norm": 0.9952289730492886, + "learning_rate": 0.003, + "loss": 4.0845, + "step": 15758 + }, + { + "epoch": 0.15759, + "grad_norm": 1.0066734276066078, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 15759 + }, + { + "epoch": 0.1576, + "grad_norm": 0.9568075529311973, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 15760 + }, + { + "epoch": 0.15761, + "grad_norm": 1.0510949392879398, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 15761 + }, + { + "epoch": 0.15762, + "grad_norm": 0.9179918099887558, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 15762 + }, + { + "epoch": 0.15763, + "grad_norm": 1.0528519544967807, + "learning_rate": 0.003, + "loss": 4.081, + "step": 15763 + }, + { + "epoch": 0.15764, + "grad_norm": 0.9314183405024905, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 15764 + }, + { + "epoch": 0.15765, + "grad_norm": 1.1421817585694012, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 15765 + }, + { + "epoch": 0.15766, + "grad_norm": 0.9357957041256403, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 15766 + }, + { + "epoch": 0.15767, + "grad_norm": 0.8450911500413198, + "learning_rate": 0.003, + "loss": 4.0947, + "step": 15767 + }, + { + "epoch": 0.15768, + "grad_norm": 0.885310684358978, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 15768 + }, + { + "epoch": 0.15769, + "grad_norm": 0.9286420073564032, + "learning_rate": 0.003, + "loss": 4.092, + "step": 15769 + }, + { + "epoch": 0.1577, + "grad_norm": 1.2763146252094717, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 15770 + }, + { + "epoch": 0.15771, + "grad_norm": 1.0762261319216602, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 15771 + }, + { + "epoch": 0.15772, + "grad_norm": 1.2860977930431106, + "learning_rate": 0.003, + "loss": 4.1113, + "step": 15772 + }, + { + "epoch": 0.15773, + "grad_norm": 0.9463488460617837, + "learning_rate": 0.003, + "loss": 4.085, + "step": 15773 + }, + { + "epoch": 0.15774, + "grad_norm": 0.8575739330084644, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 15774 + }, + { + "epoch": 0.15775, + "grad_norm": 0.7875348023560395, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 15775 + }, + { + "epoch": 0.15776, + "grad_norm": 0.8315742664364427, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 15776 + }, + { + "epoch": 0.15777, + "grad_norm": 0.9376204449656194, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 15777 + }, + { + "epoch": 0.15778, + "grad_norm": 1.1213800168381438, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 15778 + }, + { + "epoch": 0.15779, + "grad_norm": 0.9677784450165104, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 15779 + }, + { + "epoch": 0.1578, + "grad_norm": 0.9142222126195715, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 15780 + }, + { + "epoch": 0.15781, + "grad_norm": 1.1567080203778966, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 15781 + }, + { + "epoch": 0.15782, + "grad_norm": 1.0875028258849753, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 15782 + }, + { + "epoch": 0.15783, + "grad_norm": 0.8349363541187321, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 15783 + }, + { + "epoch": 0.15784, + "grad_norm": 0.7686648575158785, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 15784 + }, + { + "epoch": 0.15785, + "grad_norm": 0.8943276468241085, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 15785 + }, + { + "epoch": 0.15786, + "grad_norm": 1.062281718566549, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 15786 + }, + { + "epoch": 0.15787, + "grad_norm": 0.8319443067668273, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 15787 + }, + { + "epoch": 0.15788, + "grad_norm": 0.7279383755061583, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 15788 + }, + { + "epoch": 0.15789, + "grad_norm": 0.6842291057272144, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 15789 + }, + { + "epoch": 0.1579, + "grad_norm": 0.889649664816208, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 15790 + }, + { + "epoch": 0.15791, + "grad_norm": 0.9433718994228649, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 15791 + }, + { + "epoch": 0.15792, + "grad_norm": 1.2760794870289827, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 15792 + }, + { + "epoch": 0.15793, + "grad_norm": 1.0198452407600855, + "learning_rate": 0.003, + "loss": 4.0974, + "step": 15793 + }, + { + "epoch": 0.15794, + "grad_norm": 1.0362867293003581, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 15794 + }, + { + "epoch": 0.15795, + "grad_norm": 0.9488995598961963, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 15795 + }, + { + "epoch": 0.15796, + "grad_norm": 0.9268979609616538, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 15796 + }, + { + "epoch": 0.15797, + "grad_norm": 0.8038649572019114, + "learning_rate": 0.003, + "loss": 4.032, + "step": 15797 + }, + { + "epoch": 0.15798, + "grad_norm": 0.762507683134303, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 15798 + }, + { + "epoch": 0.15799, + "grad_norm": 0.9032191186424204, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 15799 + }, + { + "epoch": 0.158, + "grad_norm": 0.9669972956953021, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 15800 + }, + { + "epoch": 0.15801, + "grad_norm": 1.3345952424214946, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 15801 + }, + { + "epoch": 0.15802, + "grad_norm": 0.9848101777849072, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 15802 + }, + { + "epoch": 0.15803, + "grad_norm": 1.0328147627956656, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 15803 + }, + { + "epoch": 0.15804, + "grad_norm": 1.0149973608231113, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 15804 + }, + { + "epoch": 0.15805, + "grad_norm": 1.0240679823748493, + "learning_rate": 0.003, + "loss": 4.0907, + "step": 15805 + }, + { + "epoch": 0.15806, + "grad_norm": 1.1724147609037696, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 15806 + }, + { + "epoch": 0.15807, + "grad_norm": 1.0255625089198295, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 15807 + }, + { + "epoch": 0.15808, + "grad_norm": 1.101597082505975, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 15808 + }, + { + "epoch": 0.15809, + "grad_norm": 1.0846818142365775, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 15809 + }, + { + "epoch": 0.1581, + "grad_norm": 1.0849905353295288, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 15810 + }, + { + "epoch": 0.15811, + "grad_norm": 0.9645177593761194, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 15811 + }, + { + "epoch": 0.15812, + "grad_norm": 1.0914209335906693, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 15812 + }, + { + "epoch": 0.15813, + "grad_norm": 0.8037435312580105, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 15813 + }, + { + "epoch": 0.15814, + "grad_norm": 0.7834191781529092, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 15814 + }, + { + "epoch": 0.15815, + "grad_norm": 0.8898876463013576, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 15815 + }, + { + "epoch": 0.15816, + "grad_norm": 0.9175277199919403, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 15816 + }, + { + "epoch": 0.15817, + "grad_norm": 1.0320798812583574, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 15817 + }, + { + "epoch": 0.15818, + "grad_norm": 1.0978875472040253, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 15818 + }, + { + "epoch": 0.15819, + "grad_norm": 1.0456955038809574, + "learning_rate": 0.003, + "loss": 4.035, + "step": 15819 + }, + { + "epoch": 0.1582, + "grad_norm": 0.9310402481145503, + "learning_rate": 0.003, + "loss": 4.108, + "step": 15820 + }, + { + "epoch": 0.15821, + "grad_norm": 1.0243177286728957, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 15821 + }, + { + "epoch": 0.15822, + "grad_norm": 1.2197741619217763, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 15822 + }, + { + "epoch": 0.15823, + "grad_norm": 1.0269730353444606, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 15823 + }, + { + "epoch": 0.15824, + "grad_norm": 1.0613680852123204, + "learning_rate": 0.003, + "loss": 4.0864, + "step": 15824 + }, + { + "epoch": 0.15825, + "grad_norm": 1.2120312331346703, + "learning_rate": 0.003, + "loss": 4.0884, + "step": 15825 + }, + { + "epoch": 0.15826, + "grad_norm": 0.9349935716109848, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 15826 + }, + { + "epoch": 0.15827, + "grad_norm": 1.153611406816111, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 15827 + }, + { + "epoch": 0.15828, + "grad_norm": 1.1436589917471967, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 15828 + }, + { + "epoch": 0.15829, + "grad_norm": 0.9446844193040163, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 15829 + }, + { + "epoch": 0.1583, + "grad_norm": 0.8766312388226903, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 15830 + }, + { + "epoch": 0.15831, + "grad_norm": 1.0178979329002729, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 15831 + }, + { + "epoch": 0.15832, + "grad_norm": 1.07019077539559, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 15832 + }, + { + "epoch": 0.15833, + "grad_norm": 0.9425624352762172, + "learning_rate": 0.003, + "loss": 4.05, + "step": 15833 + }, + { + "epoch": 0.15834, + "grad_norm": 0.9265469031967339, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 15834 + }, + { + "epoch": 0.15835, + "grad_norm": 0.9818334688839481, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 15835 + }, + { + "epoch": 0.15836, + "grad_norm": 1.016138841058393, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 15836 + }, + { + "epoch": 0.15837, + "grad_norm": 1.2291726935738074, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 15837 + }, + { + "epoch": 0.15838, + "grad_norm": 0.9731881590726368, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 15838 + }, + { + "epoch": 0.15839, + "grad_norm": 1.2277892701351694, + "learning_rate": 0.003, + "loss": 4.062, + "step": 15839 + }, + { + "epoch": 0.1584, + "grad_norm": 0.8332648553221847, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 15840 + }, + { + "epoch": 0.15841, + "grad_norm": 0.660597440236932, + "learning_rate": 0.003, + "loss": 4.078, + "step": 15841 + }, + { + "epoch": 0.15842, + "grad_norm": 0.6622948857371246, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 15842 + }, + { + "epoch": 0.15843, + "grad_norm": 0.7831661594371441, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 15843 + }, + { + "epoch": 0.15844, + "grad_norm": 1.0110412913472409, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 15844 + }, + { + "epoch": 0.15845, + "grad_norm": 1.0184771497706155, + "learning_rate": 0.003, + "loss": 4.0969, + "step": 15845 + }, + { + "epoch": 0.15846, + "grad_norm": 1.1689420928175727, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 15846 + }, + { + "epoch": 0.15847, + "grad_norm": 0.9787541772253365, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 15847 + }, + { + "epoch": 0.15848, + "grad_norm": 0.959062238982396, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 15848 + }, + { + "epoch": 0.15849, + "grad_norm": 1.088960491122121, + "learning_rate": 0.003, + "loss": 4.08, + "step": 15849 + }, + { + "epoch": 0.1585, + "grad_norm": 1.173521105191213, + "learning_rate": 0.003, + "loss": 4.0998, + "step": 15850 + }, + { + "epoch": 0.15851, + "grad_norm": 1.027353486856592, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 15851 + }, + { + "epoch": 0.15852, + "grad_norm": 1.0363199709092574, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 15852 + }, + { + "epoch": 0.15853, + "grad_norm": 1.1087545512359507, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 15853 + }, + { + "epoch": 0.15854, + "grad_norm": 0.9531629590053476, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 15854 + }, + { + "epoch": 0.15855, + "grad_norm": 1.1225231823178006, + "learning_rate": 0.003, + "loss": 4.074, + "step": 15855 + }, + { + "epoch": 0.15856, + "grad_norm": 0.7901576642093684, + "learning_rate": 0.003, + "loss": 4.093, + "step": 15856 + }, + { + "epoch": 0.15857, + "grad_norm": 0.7525410670842909, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 15857 + }, + { + "epoch": 0.15858, + "grad_norm": 0.7726569043785391, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 15858 + }, + { + "epoch": 0.15859, + "grad_norm": 0.7501135403672647, + "learning_rate": 0.003, + "loss": 4.089, + "step": 15859 + }, + { + "epoch": 0.1586, + "grad_norm": 0.900515722258503, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 15860 + }, + { + "epoch": 0.15861, + "grad_norm": 1.0780521854342682, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 15861 + }, + { + "epoch": 0.15862, + "grad_norm": 1.2771258591143386, + "learning_rate": 0.003, + "loss": 4.0965, + "step": 15862 + }, + { + "epoch": 0.15863, + "grad_norm": 0.7740518297166336, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 15863 + }, + { + "epoch": 0.15864, + "grad_norm": 0.6293479069911586, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 15864 + }, + { + "epoch": 0.15865, + "grad_norm": 0.7807268836886082, + "learning_rate": 0.003, + "loss": 4.052, + "step": 15865 + }, + { + "epoch": 0.15866, + "grad_norm": 1.0369531248747144, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 15866 + }, + { + "epoch": 0.15867, + "grad_norm": 1.4672612248839334, + "learning_rate": 0.003, + "loss": 4.076, + "step": 15867 + }, + { + "epoch": 0.15868, + "grad_norm": 0.6354764351569627, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 15868 + }, + { + "epoch": 0.15869, + "grad_norm": 0.9917013178430887, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 15869 + }, + { + "epoch": 0.1587, + "grad_norm": 1.4807568656774888, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 15870 + }, + { + "epoch": 0.15871, + "grad_norm": 0.5981508182826052, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 15871 + }, + { + "epoch": 0.15872, + "grad_norm": 0.9179768529180138, + "learning_rate": 0.003, + "loss": 4.0855, + "step": 15872 + }, + { + "epoch": 0.15873, + "grad_norm": 0.979618288549501, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 15873 + }, + { + "epoch": 0.15874, + "grad_norm": 1.0271030479679424, + "learning_rate": 0.003, + "loss": 4.051, + "step": 15874 + }, + { + "epoch": 0.15875, + "grad_norm": 0.9911897528459245, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 15875 + }, + { + "epoch": 0.15876, + "grad_norm": 1.0397043954996412, + "learning_rate": 0.003, + "loss": 4.08, + "step": 15876 + }, + { + "epoch": 0.15877, + "grad_norm": 1.0066601732701606, + "learning_rate": 0.003, + "loss": 4.0941, + "step": 15877 + }, + { + "epoch": 0.15878, + "grad_norm": 0.9801984274008906, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 15878 + }, + { + "epoch": 0.15879, + "grad_norm": 1.2345092391564532, + "learning_rate": 0.003, + "loss": 4.0868, + "step": 15879 + }, + { + "epoch": 0.1588, + "grad_norm": 0.8592867977188222, + "learning_rate": 0.003, + "loss": 4.0942, + "step": 15880 + }, + { + "epoch": 0.15881, + "grad_norm": 0.9672423007695633, + "learning_rate": 0.003, + "loss": 4.058, + "step": 15881 + }, + { + "epoch": 0.15882, + "grad_norm": 1.1017379666475386, + "learning_rate": 0.003, + "loss": 4.045, + "step": 15882 + }, + { + "epoch": 0.15883, + "grad_norm": 1.169208833579286, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 15883 + }, + { + "epoch": 0.15884, + "grad_norm": 0.8197387185644105, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 15884 + }, + { + "epoch": 0.15885, + "grad_norm": 0.8294938383560584, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 15885 + }, + { + "epoch": 0.15886, + "grad_norm": 0.8872420893616674, + "learning_rate": 0.003, + "loss": 4.0855, + "step": 15886 + }, + { + "epoch": 0.15887, + "grad_norm": 0.9345323642274654, + "learning_rate": 0.003, + "loss": 4.0899, + "step": 15887 + }, + { + "epoch": 0.15888, + "grad_norm": 0.9865918272910974, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 15888 + }, + { + "epoch": 0.15889, + "grad_norm": 0.9287839790885517, + "learning_rate": 0.003, + "loss": 4.0937, + "step": 15889 + }, + { + "epoch": 0.1589, + "grad_norm": 1.0248493308543711, + "learning_rate": 0.003, + "loss": 4.0978, + "step": 15890 + }, + { + "epoch": 0.15891, + "grad_norm": 1.0682635034258106, + "learning_rate": 0.003, + "loss": 4.1091, + "step": 15891 + }, + { + "epoch": 0.15892, + "grad_norm": 0.8863408931874633, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 15892 + }, + { + "epoch": 0.15893, + "grad_norm": 0.9759809138488861, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 15893 + }, + { + "epoch": 0.15894, + "grad_norm": 0.9077348777863288, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 15894 + }, + { + "epoch": 0.15895, + "grad_norm": 1.089241790915074, + "learning_rate": 0.003, + "loss": 4.0887, + "step": 15895 + }, + { + "epoch": 0.15896, + "grad_norm": 1.2249405942940117, + "learning_rate": 0.003, + "loss": 4.063, + "step": 15896 + }, + { + "epoch": 0.15897, + "grad_norm": 0.9682245345210783, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 15897 + }, + { + "epoch": 0.15898, + "grad_norm": 1.1437140635900005, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 15898 + }, + { + "epoch": 0.15899, + "grad_norm": 1.3762905085850343, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 15899 + }, + { + "epoch": 0.159, + "grad_norm": 0.731132982122956, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 15900 + }, + { + "epoch": 0.15901, + "grad_norm": 0.7604526330757749, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 15901 + }, + { + "epoch": 0.15902, + "grad_norm": 0.8091081218419411, + "learning_rate": 0.003, + "loss": 4.1119, + "step": 15902 + }, + { + "epoch": 0.15903, + "grad_norm": 0.9138836267696424, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 15903 + }, + { + "epoch": 0.15904, + "grad_norm": 1.1346670690369771, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 15904 + }, + { + "epoch": 0.15905, + "grad_norm": 1.2064058197300966, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 15905 + }, + { + "epoch": 0.15906, + "grad_norm": 1.034978947846275, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 15906 + }, + { + "epoch": 0.15907, + "grad_norm": 1.2014482222510476, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 15907 + }, + { + "epoch": 0.15908, + "grad_norm": 0.9171260578841384, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 15908 + }, + { + "epoch": 0.15909, + "grad_norm": 0.9100066824897164, + "learning_rate": 0.003, + "loss": 4.083, + "step": 15909 + }, + { + "epoch": 0.1591, + "grad_norm": 1.0687258590242457, + "learning_rate": 0.003, + "loss": 4.087, + "step": 15910 + }, + { + "epoch": 0.15911, + "grad_norm": 1.0913947740995993, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 15911 + }, + { + "epoch": 0.15912, + "grad_norm": 1.090062165217744, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 15912 + }, + { + "epoch": 0.15913, + "grad_norm": 1.0738799771317074, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 15913 + }, + { + "epoch": 0.15914, + "grad_norm": 1.035784906394474, + "learning_rate": 0.003, + "loss": 4.1012, + "step": 15914 + }, + { + "epoch": 0.15915, + "grad_norm": 1.0032653968162704, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 15915 + }, + { + "epoch": 0.15916, + "grad_norm": 1.050742339116215, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 15916 + }, + { + "epoch": 0.15917, + "grad_norm": 0.8863883282668868, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 15917 + }, + { + "epoch": 0.15918, + "grad_norm": 0.8572928474945728, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 15918 + }, + { + "epoch": 0.15919, + "grad_norm": 0.9079555836918002, + "learning_rate": 0.003, + "loss": 4.0928, + "step": 15919 + }, + { + "epoch": 0.1592, + "grad_norm": 0.8793314929642051, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 15920 + }, + { + "epoch": 0.15921, + "grad_norm": 0.9910321667306762, + "learning_rate": 0.003, + "loss": 4.108, + "step": 15921 + }, + { + "epoch": 0.15922, + "grad_norm": 1.0424631814600145, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 15922 + }, + { + "epoch": 0.15923, + "grad_norm": 1.094365679430473, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 15923 + }, + { + "epoch": 0.15924, + "grad_norm": 1.1452593843292889, + "learning_rate": 0.003, + "loss": 4.0857, + "step": 15924 + }, + { + "epoch": 0.15925, + "grad_norm": 0.9964435241349789, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 15925 + }, + { + "epoch": 0.15926, + "grad_norm": 1.274425905970494, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 15926 + }, + { + "epoch": 0.15927, + "grad_norm": 0.845431387367486, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 15927 + }, + { + "epoch": 0.15928, + "grad_norm": 0.8635356587753668, + "learning_rate": 0.003, + "loss": 4.065, + "step": 15928 + }, + { + "epoch": 0.15929, + "grad_norm": 0.879222108052293, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 15929 + }, + { + "epoch": 0.1593, + "grad_norm": 0.9334683026164846, + "learning_rate": 0.003, + "loss": 4.07, + "step": 15930 + }, + { + "epoch": 0.15931, + "grad_norm": 1.019650523280888, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 15931 + }, + { + "epoch": 0.15932, + "grad_norm": 1.013870152324945, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 15932 + }, + { + "epoch": 0.15933, + "grad_norm": 0.955116514767827, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 15933 + }, + { + "epoch": 0.15934, + "grad_norm": 0.8988227521847098, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 15934 + }, + { + "epoch": 0.15935, + "grad_norm": 0.8826087207406549, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 15935 + }, + { + "epoch": 0.15936, + "grad_norm": 0.9077341912064462, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 15936 + }, + { + "epoch": 0.15937, + "grad_norm": 0.8108373712642741, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 15937 + }, + { + "epoch": 0.15938, + "grad_norm": 0.834726575363103, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 15938 + }, + { + "epoch": 0.15939, + "grad_norm": 0.7890873718615576, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 15939 + }, + { + "epoch": 0.1594, + "grad_norm": 0.778186658781821, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 15940 + }, + { + "epoch": 0.15941, + "grad_norm": 1.0391703472247389, + "learning_rate": 0.003, + "loss": 4.051, + "step": 15941 + }, + { + "epoch": 0.15942, + "grad_norm": 1.2527067461502035, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 15942 + }, + { + "epoch": 0.15943, + "grad_norm": 0.7832094454223829, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 15943 + }, + { + "epoch": 0.15944, + "grad_norm": 0.8326353264512342, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 15944 + }, + { + "epoch": 0.15945, + "grad_norm": 0.9776536312829355, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 15945 + }, + { + "epoch": 0.15946, + "grad_norm": 1.1761604074350485, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 15946 + }, + { + "epoch": 0.15947, + "grad_norm": 0.9343342013679248, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 15947 + }, + { + "epoch": 0.15948, + "grad_norm": 0.9259487731601618, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 15948 + }, + { + "epoch": 0.15949, + "grad_norm": 0.9922446532931917, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 15949 + }, + { + "epoch": 0.1595, + "grad_norm": 1.179605698401649, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 15950 + }, + { + "epoch": 0.15951, + "grad_norm": 0.8406588244684183, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 15951 + }, + { + "epoch": 0.15952, + "grad_norm": 0.8798223253124555, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 15952 + }, + { + "epoch": 0.15953, + "grad_norm": 1.1021232703385526, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 15953 + }, + { + "epoch": 0.15954, + "grad_norm": 1.2792100484216868, + "learning_rate": 0.003, + "loss": 4.1052, + "step": 15954 + }, + { + "epoch": 0.15955, + "grad_norm": 0.8014784404972527, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 15955 + }, + { + "epoch": 0.15956, + "grad_norm": 0.6746475070962193, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 15956 + }, + { + "epoch": 0.15957, + "grad_norm": 0.7557848605040285, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 15957 + }, + { + "epoch": 0.15958, + "grad_norm": 0.8389016069554045, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 15958 + }, + { + "epoch": 0.15959, + "grad_norm": 0.9998976028923484, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 15959 + }, + { + "epoch": 0.1596, + "grad_norm": 1.2476274581087763, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 15960 + }, + { + "epoch": 0.15961, + "grad_norm": 0.9552920178140516, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 15961 + }, + { + "epoch": 0.15962, + "grad_norm": 0.9012025855120287, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 15962 + }, + { + "epoch": 0.15963, + "grad_norm": 0.9208548277308138, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 15963 + }, + { + "epoch": 0.15964, + "grad_norm": 1.2796933776608541, + "learning_rate": 0.003, + "loss": 4.047, + "step": 15964 + }, + { + "epoch": 0.15965, + "grad_norm": 1.1408199742147165, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 15965 + }, + { + "epoch": 0.15966, + "grad_norm": 0.83551653174545, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 15966 + }, + { + "epoch": 0.15967, + "grad_norm": 0.7495784380838688, + "learning_rate": 0.003, + "loss": 4.065, + "step": 15967 + }, + { + "epoch": 0.15968, + "grad_norm": 0.7545445149567002, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 15968 + }, + { + "epoch": 0.15969, + "grad_norm": 0.8772138480707413, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 15969 + }, + { + "epoch": 0.1597, + "grad_norm": 1.145988759237524, + "learning_rate": 0.003, + "loss": 4.087, + "step": 15970 + }, + { + "epoch": 0.15971, + "grad_norm": 1.1147724439739533, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 15971 + }, + { + "epoch": 0.15972, + "grad_norm": 0.9565193265598885, + "learning_rate": 0.003, + "loss": 4.0933, + "step": 15972 + }, + { + "epoch": 0.15973, + "grad_norm": 1.0874031271162286, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 15973 + }, + { + "epoch": 0.15974, + "grad_norm": 0.953459190891649, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 15974 + }, + { + "epoch": 0.15975, + "grad_norm": 0.8391390009546319, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 15975 + }, + { + "epoch": 0.15976, + "grad_norm": 0.7470859789257781, + "learning_rate": 0.003, + "loss": 4.063, + "step": 15976 + }, + { + "epoch": 0.15977, + "grad_norm": 0.8300962206393157, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 15977 + }, + { + "epoch": 0.15978, + "grad_norm": 0.8558871689758444, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 15978 + }, + { + "epoch": 0.15979, + "grad_norm": 0.9380296571241467, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 15979 + }, + { + "epoch": 0.1598, + "grad_norm": 1.2837338513044843, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 15980 + }, + { + "epoch": 0.15981, + "grad_norm": 0.9833419488580739, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 15981 + }, + { + "epoch": 0.15982, + "grad_norm": 0.9854131316496347, + "learning_rate": 0.003, + "loss": 4.0878, + "step": 15982 + }, + { + "epoch": 0.15983, + "grad_norm": 1.091659541554118, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 15983 + }, + { + "epoch": 0.15984, + "grad_norm": 1.059305037486897, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 15984 + }, + { + "epoch": 0.15985, + "grad_norm": 1.0226621621785588, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 15985 + }, + { + "epoch": 0.15986, + "grad_norm": 1.1160566441539626, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 15986 + }, + { + "epoch": 0.15987, + "grad_norm": 0.9887253367389169, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 15987 + }, + { + "epoch": 0.15988, + "grad_norm": 1.0079876653492712, + "learning_rate": 0.003, + "loss": 4.086, + "step": 15988 + }, + { + "epoch": 0.15989, + "grad_norm": 1.0171714010255504, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 15989 + }, + { + "epoch": 0.1599, + "grad_norm": 1.172802554818858, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 15990 + }, + { + "epoch": 0.15991, + "grad_norm": 1.137834438739498, + "learning_rate": 0.003, + "loss": 4.0922, + "step": 15991 + }, + { + "epoch": 0.15992, + "grad_norm": 1.041516207067733, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 15992 + }, + { + "epoch": 0.15993, + "grad_norm": 1.0546811946173797, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 15993 + }, + { + "epoch": 0.15994, + "grad_norm": 1.1335349568710917, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 15994 + }, + { + "epoch": 0.15995, + "grad_norm": 0.8923354587029558, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 15995 + }, + { + "epoch": 0.15996, + "grad_norm": 0.9073525266819159, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 15996 + }, + { + "epoch": 0.15997, + "grad_norm": 1.0873011090439746, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 15997 + }, + { + "epoch": 0.15998, + "grad_norm": 1.081014111309391, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 15998 + }, + { + "epoch": 0.15999, + "grad_norm": 0.944118348922414, + "learning_rate": 0.003, + "loss": 4.068, + "step": 15999 + }, + { + "epoch": 0.16, + "grad_norm": 0.8269606805411867, + "learning_rate": 0.003, + "loss": 4.082, + "step": 16000 + }, + { + "epoch": 0.16001, + "grad_norm": 1.0162140856913147, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 16001 + }, + { + "epoch": 0.16002, + "grad_norm": 1.2028674230587038, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 16002 + }, + { + "epoch": 0.16003, + "grad_norm": 1.074235594552825, + "learning_rate": 0.003, + "loss": 4.0991, + "step": 16003 + }, + { + "epoch": 0.16004, + "grad_norm": 0.8730882941891372, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 16004 + }, + { + "epoch": 0.16005, + "grad_norm": 0.738817555739496, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 16005 + }, + { + "epoch": 0.16006, + "grad_norm": 0.8148417778863171, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 16006 + }, + { + "epoch": 0.16007, + "grad_norm": 1.018096498714664, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 16007 + }, + { + "epoch": 0.16008, + "grad_norm": 1.3357683415261388, + "learning_rate": 0.003, + "loss": 4.0934, + "step": 16008 + }, + { + "epoch": 0.16009, + "grad_norm": 0.9325327938576854, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 16009 + }, + { + "epoch": 0.1601, + "grad_norm": 0.8559048471597575, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 16010 + }, + { + "epoch": 0.16011, + "grad_norm": 0.873430979210134, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 16011 + }, + { + "epoch": 0.16012, + "grad_norm": 1.0195359160467716, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 16012 + }, + { + "epoch": 0.16013, + "grad_norm": 1.3365241936306207, + "learning_rate": 0.003, + "loss": 4.106, + "step": 16013 + }, + { + "epoch": 0.16014, + "grad_norm": 0.9105149906484808, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 16014 + }, + { + "epoch": 0.16015, + "grad_norm": 1.0274370282192966, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 16015 + }, + { + "epoch": 0.16016, + "grad_norm": 1.0449500009623987, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 16016 + }, + { + "epoch": 0.16017, + "grad_norm": 1.0039711171541923, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 16017 + }, + { + "epoch": 0.16018, + "grad_norm": 0.8553220338834571, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 16018 + }, + { + "epoch": 0.16019, + "grad_norm": 0.8695955372130583, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 16019 + }, + { + "epoch": 0.1602, + "grad_norm": 0.9453896806024594, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 16020 + }, + { + "epoch": 0.16021, + "grad_norm": 1.1622814824798382, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 16021 + }, + { + "epoch": 0.16022, + "grad_norm": 1.0833819438480463, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 16022 + }, + { + "epoch": 0.16023, + "grad_norm": 0.7762269093165016, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 16023 + }, + { + "epoch": 0.16024, + "grad_norm": 0.6387498876607048, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 16024 + }, + { + "epoch": 0.16025, + "grad_norm": 0.6764882547085834, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 16025 + }, + { + "epoch": 0.16026, + "grad_norm": 0.884438628291668, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 16026 + }, + { + "epoch": 0.16027, + "grad_norm": 1.0613802737898816, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 16027 + }, + { + "epoch": 0.16028, + "grad_norm": 1.012875056459876, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 16028 + }, + { + "epoch": 0.16029, + "grad_norm": 1.204478446663581, + "learning_rate": 0.003, + "loss": 4.086, + "step": 16029 + }, + { + "epoch": 0.1603, + "grad_norm": 0.9518019822343159, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 16030 + }, + { + "epoch": 0.16031, + "grad_norm": 1.0173931173634527, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 16031 + }, + { + "epoch": 0.16032, + "grad_norm": 1.0410995057756685, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 16032 + }, + { + "epoch": 0.16033, + "grad_norm": 0.824633208409138, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 16033 + }, + { + "epoch": 0.16034, + "grad_norm": 0.8542885245729803, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 16034 + }, + { + "epoch": 0.16035, + "grad_norm": 0.8002734367419061, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 16035 + }, + { + "epoch": 0.16036, + "grad_norm": 0.819294051945251, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 16036 + }, + { + "epoch": 0.16037, + "grad_norm": 0.9273733408773418, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 16037 + }, + { + "epoch": 0.16038, + "grad_norm": 0.883440033480316, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 16038 + }, + { + "epoch": 0.16039, + "grad_norm": 0.9240300178208061, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 16039 + }, + { + "epoch": 0.1604, + "grad_norm": 0.9259174070489228, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 16040 + }, + { + "epoch": 0.16041, + "grad_norm": 1.0653341896361466, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 16041 + }, + { + "epoch": 0.16042, + "grad_norm": 1.0154984946999295, + "learning_rate": 0.003, + "loss": 4.0983, + "step": 16042 + }, + { + "epoch": 0.16043, + "grad_norm": 0.9472157467600227, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 16043 + }, + { + "epoch": 0.16044, + "grad_norm": 1.125029784998067, + "learning_rate": 0.003, + "loss": 4.0878, + "step": 16044 + }, + { + "epoch": 0.16045, + "grad_norm": 1.0355491196438407, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 16045 + }, + { + "epoch": 0.16046, + "grad_norm": 1.0018920603152246, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 16046 + }, + { + "epoch": 0.16047, + "grad_norm": 1.0429554042920302, + "learning_rate": 0.003, + "loss": 4.0935, + "step": 16047 + }, + { + "epoch": 0.16048, + "grad_norm": 0.9663336034455842, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 16048 + }, + { + "epoch": 0.16049, + "grad_norm": 1.5540250196736454, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 16049 + }, + { + "epoch": 0.1605, + "grad_norm": 0.9802063545436647, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 16050 + }, + { + "epoch": 0.16051, + "grad_norm": 1.178123809405368, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 16051 + }, + { + "epoch": 0.16052, + "grad_norm": 1.0754951690561978, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 16052 + }, + { + "epoch": 0.16053, + "grad_norm": 0.9375423759825882, + "learning_rate": 0.003, + "loss": 4.0901, + "step": 16053 + }, + { + "epoch": 0.16054, + "grad_norm": 0.9027724134416651, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 16054 + }, + { + "epoch": 0.16055, + "grad_norm": 0.9923908426106327, + "learning_rate": 0.003, + "loss": 4.0819, + "step": 16055 + }, + { + "epoch": 0.16056, + "grad_norm": 1.2763700542037637, + "learning_rate": 0.003, + "loss": 4.0932, + "step": 16056 + }, + { + "epoch": 0.16057, + "grad_norm": 0.7862279920872993, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 16057 + }, + { + "epoch": 0.16058, + "grad_norm": 0.6665669828397947, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 16058 + }, + { + "epoch": 0.16059, + "grad_norm": 0.7375894956269987, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 16059 + }, + { + "epoch": 0.1606, + "grad_norm": 0.8079550488192577, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 16060 + }, + { + "epoch": 0.16061, + "grad_norm": 0.8672653399081101, + "learning_rate": 0.003, + "loss": 4.0926, + "step": 16061 + }, + { + "epoch": 0.16062, + "grad_norm": 0.9248706941427756, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 16062 + }, + { + "epoch": 0.16063, + "grad_norm": 0.9907321763478665, + "learning_rate": 0.003, + "loss": 4.081, + "step": 16063 + }, + { + "epoch": 0.16064, + "grad_norm": 1.0404159215892577, + "learning_rate": 0.003, + "loss": 4.0955, + "step": 16064 + }, + { + "epoch": 0.16065, + "grad_norm": 1.1712680190092282, + "learning_rate": 0.003, + "loss": 4.068, + "step": 16065 + }, + { + "epoch": 0.16066, + "grad_norm": 0.9957042123077742, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 16066 + }, + { + "epoch": 0.16067, + "grad_norm": 1.2663639857343698, + "learning_rate": 0.003, + "loss": 4.0942, + "step": 16067 + }, + { + "epoch": 0.16068, + "grad_norm": 0.7792473369642737, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 16068 + }, + { + "epoch": 0.16069, + "grad_norm": 0.7074567358629517, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 16069 + }, + { + "epoch": 0.1607, + "grad_norm": 0.8898468162614108, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 16070 + }, + { + "epoch": 0.16071, + "grad_norm": 1.0602219988190484, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 16071 + }, + { + "epoch": 0.16072, + "grad_norm": 1.0553917278784992, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 16072 + }, + { + "epoch": 0.16073, + "grad_norm": 0.997187847824805, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 16073 + }, + { + "epoch": 0.16074, + "grad_norm": 0.9097074721645234, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 16074 + }, + { + "epoch": 0.16075, + "grad_norm": 0.9897113984066431, + "learning_rate": 0.003, + "loss": 4.0848, + "step": 16075 + }, + { + "epoch": 0.16076, + "grad_norm": 1.335795153998264, + "learning_rate": 0.003, + "loss": 4.0979, + "step": 16076 + }, + { + "epoch": 0.16077, + "grad_norm": 0.7690932500692094, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 16077 + }, + { + "epoch": 0.16078, + "grad_norm": 0.6038408915332641, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 16078 + }, + { + "epoch": 0.16079, + "grad_norm": 0.6077747253793363, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 16079 + }, + { + "epoch": 0.1608, + "grad_norm": 0.6504224944468463, + "learning_rate": 0.003, + "loss": 4.038, + "step": 16080 + }, + { + "epoch": 0.16081, + "grad_norm": 0.7654115206688827, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 16081 + }, + { + "epoch": 0.16082, + "grad_norm": 1.1072679635593854, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 16082 + }, + { + "epoch": 0.16083, + "grad_norm": 1.0884612090914298, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 16083 + }, + { + "epoch": 0.16084, + "grad_norm": 0.7870318702598812, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 16084 + }, + { + "epoch": 0.16085, + "grad_norm": 0.7870318926951514, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 16085 + }, + { + "epoch": 0.16086, + "grad_norm": 0.9361814766236227, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 16086 + }, + { + "epoch": 0.16087, + "grad_norm": 1.1678644935840568, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 16087 + }, + { + "epoch": 0.16088, + "grad_norm": 0.8591127482889848, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 16088 + }, + { + "epoch": 0.16089, + "grad_norm": 0.8969178009435688, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 16089 + }, + { + "epoch": 0.1609, + "grad_norm": 1.279108856170437, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 16090 + }, + { + "epoch": 0.16091, + "grad_norm": 1.1972977271863883, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 16091 + }, + { + "epoch": 0.16092, + "grad_norm": 0.7876575525337091, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 16092 + }, + { + "epoch": 0.16093, + "grad_norm": 0.8096535443062386, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 16093 + }, + { + "epoch": 0.16094, + "grad_norm": 0.930702011100421, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 16094 + }, + { + "epoch": 0.16095, + "grad_norm": 1.1073384472294623, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 16095 + }, + { + "epoch": 0.16096, + "grad_norm": 1.030894450438812, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 16096 + }, + { + "epoch": 0.16097, + "grad_norm": 1.1297220882476162, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 16097 + }, + { + "epoch": 0.16098, + "grad_norm": 1.0048329730337762, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 16098 + }, + { + "epoch": 0.16099, + "grad_norm": 1.1535021270460015, + "learning_rate": 0.003, + "loss": 4.079, + "step": 16099 + }, + { + "epoch": 0.161, + "grad_norm": 1.0600878993362524, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 16100 + }, + { + "epoch": 0.16101, + "grad_norm": 1.1746890640219785, + "learning_rate": 0.003, + "loss": 4.08, + "step": 16101 + }, + { + "epoch": 0.16102, + "grad_norm": 0.9125675191264994, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 16102 + }, + { + "epoch": 0.16103, + "grad_norm": 1.1224318920120804, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 16103 + }, + { + "epoch": 0.16104, + "grad_norm": 1.3369431825012918, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 16104 + }, + { + "epoch": 0.16105, + "grad_norm": 0.7237982765026987, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 16105 + }, + { + "epoch": 0.16106, + "grad_norm": 0.64650962647748, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 16106 + }, + { + "epoch": 0.16107, + "grad_norm": 0.9065173529534585, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 16107 + }, + { + "epoch": 0.16108, + "grad_norm": 1.114407382294119, + "learning_rate": 0.003, + "loss": 4.088, + "step": 16108 + }, + { + "epoch": 0.16109, + "grad_norm": 1.0721346541062686, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 16109 + }, + { + "epoch": 0.1611, + "grad_norm": 1.086705989189854, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 16110 + }, + { + "epoch": 0.16111, + "grad_norm": 0.9167370335222023, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 16111 + }, + { + "epoch": 0.16112, + "grad_norm": 1.0457710466304166, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 16112 + }, + { + "epoch": 0.16113, + "grad_norm": 1.1601869885960452, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 16113 + }, + { + "epoch": 0.16114, + "grad_norm": 0.8312721477553048, + "learning_rate": 0.003, + "loss": 4.06, + "step": 16114 + }, + { + "epoch": 0.16115, + "grad_norm": 0.8729894642460365, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 16115 + }, + { + "epoch": 0.16116, + "grad_norm": 1.0115575738660978, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 16116 + }, + { + "epoch": 0.16117, + "grad_norm": 1.0427899386032666, + "learning_rate": 0.003, + "loss": 4.0858, + "step": 16117 + }, + { + "epoch": 0.16118, + "grad_norm": 0.9133504404720045, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 16118 + }, + { + "epoch": 0.16119, + "grad_norm": 1.0052579182429844, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 16119 + }, + { + "epoch": 0.1612, + "grad_norm": 1.1100005580512666, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 16120 + }, + { + "epoch": 0.16121, + "grad_norm": 1.1144509722667733, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 16121 + }, + { + "epoch": 0.16122, + "grad_norm": 0.9683861434961853, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 16122 + }, + { + "epoch": 0.16123, + "grad_norm": 1.031546773725875, + "learning_rate": 0.003, + "loss": 4.0956, + "step": 16123 + }, + { + "epoch": 0.16124, + "grad_norm": 1.1549330009133982, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 16124 + }, + { + "epoch": 0.16125, + "grad_norm": 0.9664932738430825, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 16125 + }, + { + "epoch": 0.16126, + "grad_norm": 0.9720527727279671, + "learning_rate": 0.003, + "loss": 4.063, + "step": 16126 + }, + { + "epoch": 0.16127, + "grad_norm": 0.966930599231324, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 16127 + }, + { + "epoch": 0.16128, + "grad_norm": 1.0958020895354166, + "learning_rate": 0.003, + "loss": 4.0876, + "step": 16128 + }, + { + "epoch": 0.16129, + "grad_norm": 1.0466972772313536, + "learning_rate": 0.003, + "loss": 4.085, + "step": 16129 + }, + { + "epoch": 0.1613, + "grad_norm": 0.922043331082599, + "learning_rate": 0.003, + "loss": 4.0907, + "step": 16130 + }, + { + "epoch": 0.16131, + "grad_norm": 0.8661134633012201, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 16131 + }, + { + "epoch": 0.16132, + "grad_norm": 0.9047634819076963, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 16132 + }, + { + "epoch": 0.16133, + "grad_norm": 0.8592548094535108, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 16133 + }, + { + "epoch": 0.16134, + "grad_norm": 0.9387823394292516, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 16134 + }, + { + "epoch": 0.16135, + "grad_norm": 0.991155049821999, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 16135 + }, + { + "epoch": 0.16136, + "grad_norm": 1.0722540335606572, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 16136 + }, + { + "epoch": 0.16137, + "grad_norm": 1.0766674506157325, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 16137 + }, + { + "epoch": 0.16138, + "grad_norm": 1.2577004282581086, + "learning_rate": 0.003, + "loss": 4.0962, + "step": 16138 + }, + { + "epoch": 0.16139, + "grad_norm": 1.0802934519670024, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 16139 + }, + { + "epoch": 0.1614, + "grad_norm": 0.8595968854641071, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 16140 + }, + { + "epoch": 0.16141, + "grad_norm": 0.8811293846319718, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 16141 + }, + { + "epoch": 0.16142, + "grad_norm": 1.1272798564207678, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 16142 + }, + { + "epoch": 0.16143, + "grad_norm": 1.3853140718075236, + "learning_rate": 0.003, + "loss": 4.074, + "step": 16143 + }, + { + "epoch": 0.16144, + "grad_norm": 0.9104687691574788, + "learning_rate": 0.003, + "loss": 4.0989, + "step": 16144 + }, + { + "epoch": 0.16145, + "grad_norm": 1.0104209834712967, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 16145 + }, + { + "epoch": 0.16146, + "grad_norm": 1.0291308207917498, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 16146 + }, + { + "epoch": 0.16147, + "grad_norm": 0.968425181885951, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 16147 + }, + { + "epoch": 0.16148, + "grad_norm": 0.9302758922776154, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 16148 + }, + { + "epoch": 0.16149, + "grad_norm": 0.9656661611865287, + "learning_rate": 0.003, + "loss": 4.06, + "step": 16149 + }, + { + "epoch": 0.1615, + "grad_norm": 1.061281919270156, + "learning_rate": 0.003, + "loss": 4.0929, + "step": 16150 + }, + { + "epoch": 0.16151, + "grad_norm": 1.1881081963302287, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 16151 + }, + { + "epoch": 0.16152, + "grad_norm": 0.9648058488102659, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 16152 + }, + { + "epoch": 0.16153, + "grad_norm": 1.0207873249758936, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 16153 + }, + { + "epoch": 0.16154, + "grad_norm": 1.078419697336838, + "learning_rate": 0.003, + "loss": 4.1038, + "step": 16154 + }, + { + "epoch": 0.16155, + "grad_norm": 0.811978782354792, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 16155 + }, + { + "epoch": 0.16156, + "grad_norm": 0.5841430338318133, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 16156 + }, + { + "epoch": 0.16157, + "grad_norm": 0.585951921625642, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 16157 + }, + { + "epoch": 0.16158, + "grad_norm": 0.6403485757976987, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 16158 + }, + { + "epoch": 0.16159, + "grad_norm": 0.6504837708299556, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 16159 + }, + { + "epoch": 0.1616, + "grad_norm": 0.8900884624710059, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 16160 + }, + { + "epoch": 0.16161, + "grad_norm": 1.3062696129416094, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 16161 + }, + { + "epoch": 0.16162, + "grad_norm": 0.8667888964828143, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 16162 + }, + { + "epoch": 0.16163, + "grad_norm": 0.9362895378971742, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 16163 + }, + { + "epoch": 0.16164, + "grad_norm": 1.0868829388811367, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 16164 + }, + { + "epoch": 0.16165, + "grad_norm": 1.0170902723582416, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 16165 + }, + { + "epoch": 0.16166, + "grad_norm": 0.871373691487063, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 16166 + }, + { + "epoch": 0.16167, + "grad_norm": 0.8321159717587762, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 16167 + }, + { + "epoch": 0.16168, + "grad_norm": 0.8685368244213992, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 16168 + }, + { + "epoch": 0.16169, + "grad_norm": 0.9204275135772269, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 16169 + }, + { + "epoch": 0.1617, + "grad_norm": 1.2229054701522726, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 16170 + }, + { + "epoch": 0.16171, + "grad_norm": 0.9644818028910481, + "learning_rate": 0.003, + "loss": 4.074, + "step": 16171 + }, + { + "epoch": 0.16172, + "grad_norm": 1.042596189002793, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 16172 + }, + { + "epoch": 0.16173, + "grad_norm": 1.3344676331151222, + "learning_rate": 0.003, + "loss": 4.1049, + "step": 16173 + }, + { + "epoch": 0.16174, + "grad_norm": 1.0053273973575043, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 16174 + }, + { + "epoch": 0.16175, + "grad_norm": 0.8360045649583121, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 16175 + }, + { + "epoch": 0.16176, + "grad_norm": 0.8638840787553371, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 16176 + }, + { + "epoch": 0.16177, + "grad_norm": 1.0189206043111847, + "learning_rate": 0.003, + "loss": 4.08, + "step": 16177 + }, + { + "epoch": 0.16178, + "grad_norm": 0.8750121699348608, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 16178 + }, + { + "epoch": 0.16179, + "grad_norm": 0.9605152666828264, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 16179 + }, + { + "epoch": 0.1618, + "grad_norm": 1.027467505658457, + "learning_rate": 0.003, + "loss": 4.067, + "step": 16180 + }, + { + "epoch": 0.16181, + "grad_norm": 1.0321966980055173, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 16181 + }, + { + "epoch": 0.16182, + "grad_norm": 1.0839258766811057, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 16182 + }, + { + "epoch": 0.16183, + "grad_norm": 0.9939341277724547, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 16183 + }, + { + "epoch": 0.16184, + "grad_norm": 0.9383823381823536, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 16184 + }, + { + "epoch": 0.16185, + "grad_norm": 0.917244152512682, + "learning_rate": 0.003, + "loss": 4.0949, + "step": 16185 + }, + { + "epoch": 0.16186, + "grad_norm": 1.1065897554070332, + "learning_rate": 0.003, + "loss": 4.055, + "step": 16186 + }, + { + "epoch": 0.16187, + "grad_norm": 1.0477315557238176, + "learning_rate": 0.003, + "loss": 4.0874, + "step": 16187 + }, + { + "epoch": 0.16188, + "grad_norm": 1.0897501518786687, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 16188 + }, + { + "epoch": 0.16189, + "grad_norm": 0.9706817731582148, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 16189 + }, + { + "epoch": 0.1619, + "grad_norm": 0.9406552965876658, + "learning_rate": 0.003, + "loss": 4.032, + "step": 16190 + }, + { + "epoch": 0.16191, + "grad_norm": 0.9222830349305138, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 16191 + }, + { + "epoch": 0.16192, + "grad_norm": 1.1202713091124106, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 16192 + }, + { + "epoch": 0.16193, + "grad_norm": 1.2562326629252725, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 16193 + }, + { + "epoch": 0.16194, + "grad_norm": 0.8516632153266922, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 16194 + }, + { + "epoch": 0.16195, + "grad_norm": 0.7160775779553795, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 16195 + }, + { + "epoch": 0.16196, + "grad_norm": 0.849165559959625, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 16196 + }, + { + "epoch": 0.16197, + "grad_norm": 0.999992401567131, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 16197 + }, + { + "epoch": 0.16198, + "grad_norm": 1.2041422943308189, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 16198 + }, + { + "epoch": 0.16199, + "grad_norm": 1.096547840550011, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 16199 + }, + { + "epoch": 0.162, + "grad_norm": 0.9748746434516735, + "learning_rate": 0.003, + "loss": 4.096, + "step": 16200 + }, + { + "epoch": 0.16201, + "grad_norm": 0.9796769709360063, + "learning_rate": 0.003, + "loss": 4.081, + "step": 16201 + }, + { + "epoch": 0.16202, + "grad_norm": 1.1861877484687542, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 16202 + }, + { + "epoch": 0.16203, + "grad_norm": 1.1107033177126395, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 16203 + }, + { + "epoch": 0.16204, + "grad_norm": 0.9756892332783752, + "learning_rate": 0.003, + "loss": 4.0954, + "step": 16204 + }, + { + "epoch": 0.16205, + "grad_norm": 1.067865253907001, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 16205 + }, + { + "epoch": 0.16206, + "grad_norm": 1.0556572323898752, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 16206 + }, + { + "epoch": 0.16207, + "grad_norm": 0.9414708567735196, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 16207 + }, + { + "epoch": 0.16208, + "grad_norm": 0.890047691532058, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 16208 + }, + { + "epoch": 0.16209, + "grad_norm": 0.9469546330014931, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 16209 + }, + { + "epoch": 0.1621, + "grad_norm": 0.9902882957007725, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 16210 + }, + { + "epoch": 0.16211, + "grad_norm": 1.0193720170137295, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 16211 + }, + { + "epoch": 0.16212, + "grad_norm": 0.9758434435576765, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 16212 + }, + { + "epoch": 0.16213, + "grad_norm": 1.0545680631675929, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 16213 + }, + { + "epoch": 0.16214, + "grad_norm": 0.9994477323315079, + "learning_rate": 0.003, + "loss": 4.0946, + "step": 16214 + }, + { + "epoch": 0.16215, + "grad_norm": 1.0983222730361477, + "learning_rate": 0.003, + "loss": 4.1014, + "step": 16215 + }, + { + "epoch": 0.16216, + "grad_norm": 1.1330628096427495, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 16216 + }, + { + "epoch": 0.16217, + "grad_norm": 0.9592180190191573, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 16217 + }, + { + "epoch": 0.16218, + "grad_norm": 1.004932418560183, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 16218 + }, + { + "epoch": 0.16219, + "grad_norm": 1.0902216378546743, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 16219 + }, + { + "epoch": 0.1622, + "grad_norm": 0.9073338008385707, + "learning_rate": 0.003, + "loss": 4.056, + "step": 16220 + }, + { + "epoch": 0.16221, + "grad_norm": 0.9136266399896347, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 16221 + }, + { + "epoch": 0.16222, + "grad_norm": 0.8985680490724846, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 16222 + }, + { + "epoch": 0.16223, + "grad_norm": 0.9455239014139228, + "learning_rate": 0.003, + "loss": 4.0899, + "step": 16223 + }, + { + "epoch": 0.16224, + "grad_norm": 1.1686969216260128, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 16224 + }, + { + "epoch": 0.16225, + "grad_norm": 1.1189643004642584, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 16225 + }, + { + "epoch": 0.16226, + "grad_norm": 1.1556304412868068, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 16226 + }, + { + "epoch": 0.16227, + "grad_norm": 1.026581267043118, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 16227 + }, + { + "epoch": 0.16228, + "grad_norm": 1.1200044909043017, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 16228 + }, + { + "epoch": 0.16229, + "grad_norm": 1.1440431550063437, + "learning_rate": 0.003, + "loss": 4.091, + "step": 16229 + }, + { + "epoch": 0.1623, + "grad_norm": 0.8287797185997358, + "learning_rate": 0.003, + "loss": 4.041, + "step": 16230 + }, + { + "epoch": 0.16231, + "grad_norm": 0.7467336531607476, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 16231 + }, + { + "epoch": 0.16232, + "grad_norm": 0.9706606295139087, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 16232 + }, + { + "epoch": 0.16233, + "grad_norm": 1.1928834444667396, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 16233 + }, + { + "epoch": 0.16234, + "grad_norm": 1.08493231591982, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 16234 + }, + { + "epoch": 0.16235, + "grad_norm": 1.1013415725318487, + "learning_rate": 0.003, + "loss": 4.0886, + "step": 16235 + }, + { + "epoch": 0.16236, + "grad_norm": 1.0555611763970938, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 16236 + }, + { + "epoch": 0.16237, + "grad_norm": 0.7734129045449566, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 16237 + }, + { + "epoch": 0.16238, + "grad_norm": 0.7389545643015902, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 16238 + }, + { + "epoch": 0.16239, + "grad_norm": 0.9131776125600434, + "learning_rate": 0.003, + "loss": 4.087, + "step": 16239 + }, + { + "epoch": 0.1624, + "grad_norm": 1.237803128264424, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 16240 + }, + { + "epoch": 0.16241, + "grad_norm": 1.132326391141311, + "learning_rate": 0.003, + "loss": 4.1022, + "step": 16241 + }, + { + "epoch": 0.16242, + "grad_norm": 0.9480766142434579, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 16242 + }, + { + "epoch": 0.16243, + "grad_norm": 1.1162481888681581, + "learning_rate": 0.003, + "loss": 4.0856, + "step": 16243 + }, + { + "epoch": 0.16244, + "grad_norm": 0.9081196238343037, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 16244 + }, + { + "epoch": 0.16245, + "grad_norm": 0.9121933161874416, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 16245 + }, + { + "epoch": 0.16246, + "grad_norm": 1.067699219982869, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 16246 + }, + { + "epoch": 0.16247, + "grad_norm": 1.0798954955798545, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 16247 + }, + { + "epoch": 0.16248, + "grad_norm": 1.0063274545085876, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 16248 + }, + { + "epoch": 0.16249, + "grad_norm": 0.9513829482466851, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 16249 + }, + { + "epoch": 0.1625, + "grad_norm": 0.8651884274360327, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 16250 + }, + { + "epoch": 0.16251, + "grad_norm": 1.0333087839079207, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 16251 + }, + { + "epoch": 0.16252, + "grad_norm": 1.1452529270078984, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 16252 + }, + { + "epoch": 0.16253, + "grad_norm": 0.9692425098802695, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 16253 + }, + { + "epoch": 0.16254, + "grad_norm": 1.0528563263134527, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 16254 + }, + { + "epoch": 0.16255, + "grad_norm": 1.1880020040286836, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 16255 + }, + { + "epoch": 0.16256, + "grad_norm": 0.863031879015665, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 16256 + }, + { + "epoch": 0.16257, + "grad_norm": 0.8273131309996586, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 16257 + }, + { + "epoch": 0.16258, + "grad_norm": 0.8508033183575217, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 16258 + }, + { + "epoch": 0.16259, + "grad_norm": 0.9014841709299755, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 16259 + }, + { + "epoch": 0.1626, + "grad_norm": 0.9138066900783085, + "learning_rate": 0.003, + "loss": 4.063, + "step": 16260 + }, + { + "epoch": 0.16261, + "grad_norm": 0.8888205184278071, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 16261 + }, + { + "epoch": 0.16262, + "grad_norm": 0.9512816048478471, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 16262 + }, + { + "epoch": 0.16263, + "grad_norm": 1.2919176203386016, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 16263 + }, + { + "epoch": 0.16264, + "grad_norm": 0.8173538177414529, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 16264 + }, + { + "epoch": 0.16265, + "grad_norm": 0.8774026282848785, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 16265 + }, + { + "epoch": 0.16266, + "grad_norm": 0.9440605153624376, + "learning_rate": 0.003, + "loss": 4.075, + "step": 16266 + }, + { + "epoch": 0.16267, + "grad_norm": 1.1829659670761927, + "learning_rate": 0.003, + "loss": 4.063, + "step": 16267 + }, + { + "epoch": 0.16268, + "grad_norm": 1.0656373997534598, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 16268 + }, + { + "epoch": 0.16269, + "grad_norm": 1.170036007618956, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 16269 + }, + { + "epoch": 0.1627, + "grad_norm": 0.8781645298739575, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 16270 + }, + { + "epoch": 0.16271, + "grad_norm": 0.7702972067117295, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 16271 + }, + { + "epoch": 0.16272, + "grad_norm": 0.8184869645136134, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 16272 + }, + { + "epoch": 0.16273, + "grad_norm": 0.8623008996423699, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 16273 + }, + { + "epoch": 0.16274, + "grad_norm": 1.0371011048528689, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 16274 + }, + { + "epoch": 0.16275, + "grad_norm": 0.9527352897628232, + "learning_rate": 0.003, + "loss": 4.1026, + "step": 16275 + }, + { + "epoch": 0.16276, + "grad_norm": 0.8727970594367147, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 16276 + }, + { + "epoch": 0.16277, + "grad_norm": 0.8998421045497749, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 16277 + }, + { + "epoch": 0.16278, + "grad_norm": 0.9439412862134963, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 16278 + }, + { + "epoch": 0.16279, + "grad_norm": 1.1228977311910784, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 16279 + }, + { + "epoch": 0.1628, + "grad_norm": 1.3635137117299625, + "learning_rate": 0.003, + "loss": 4.0858, + "step": 16280 + }, + { + "epoch": 0.16281, + "grad_norm": 0.9095455457056646, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 16281 + }, + { + "epoch": 0.16282, + "grad_norm": 0.9967867203986784, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 16282 + }, + { + "epoch": 0.16283, + "grad_norm": 1.2411249438118888, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 16283 + }, + { + "epoch": 0.16284, + "grad_norm": 0.9263791823599612, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 16284 + }, + { + "epoch": 0.16285, + "grad_norm": 0.9330737760666263, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 16285 + }, + { + "epoch": 0.16286, + "grad_norm": 1.0107028941843568, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 16286 + }, + { + "epoch": 0.16287, + "grad_norm": 1.1311228223362337, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 16287 + }, + { + "epoch": 0.16288, + "grad_norm": 0.9574467390704557, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 16288 + }, + { + "epoch": 0.16289, + "grad_norm": 0.8896652468334297, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 16289 + }, + { + "epoch": 0.1629, + "grad_norm": 0.9624016369471585, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 16290 + }, + { + "epoch": 0.16291, + "grad_norm": 1.1523720210307173, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 16291 + }, + { + "epoch": 0.16292, + "grad_norm": 0.9517891299434359, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 16292 + }, + { + "epoch": 0.16293, + "grad_norm": 1.043014085095631, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 16293 + }, + { + "epoch": 0.16294, + "grad_norm": 1.1052102601102503, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 16294 + }, + { + "epoch": 0.16295, + "grad_norm": 1.0249111043800194, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 16295 + }, + { + "epoch": 0.16296, + "grad_norm": 0.8678939114749524, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 16296 + }, + { + "epoch": 0.16297, + "grad_norm": 0.8664357341661135, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 16297 + }, + { + "epoch": 0.16298, + "grad_norm": 0.9073300735767897, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 16298 + }, + { + "epoch": 0.16299, + "grad_norm": 0.9621614905945949, + "learning_rate": 0.003, + "loss": 4.0906, + "step": 16299 + }, + { + "epoch": 0.163, + "grad_norm": 1.1585366562247184, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 16300 + }, + { + "epoch": 0.16301, + "grad_norm": 1.0592176294024376, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 16301 + }, + { + "epoch": 0.16302, + "grad_norm": 1.184773418509133, + "learning_rate": 0.003, + "loss": 4.086, + "step": 16302 + }, + { + "epoch": 0.16303, + "grad_norm": 0.8730375600742831, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 16303 + }, + { + "epoch": 0.16304, + "grad_norm": 0.941813922298109, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 16304 + }, + { + "epoch": 0.16305, + "grad_norm": 0.9269585667300442, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 16305 + }, + { + "epoch": 0.16306, + "grad_norm": 0.9852987018998235, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 16306 + }, + { + "epoch": 0.16307, + "grad_norm": 1.1103509304895813, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 16307 + }, + { + "epoch": 0.16308, + "grad_norm": 0.9582877744402408, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 16308 + }, + { + "epoch": 0.16309, + "grad_norm": 0.8482369574569542, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 16309 + }, + { + "epoch": 0.1631, + "grad_norm": 0.7973327861154085, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 16310 + }, + { + "epoch": 0.16311, + "grad_norm": 0.9232288453288886, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 16311 + }, + { + "epoch": 0.16312, + "grad_norm": 0.9321955980430487, + "learning_rate": 0.003, + "loss": 4.1092, + "step": 16312 + }, + { + "epoch": 0.16313, + "grad_norm": 1.1261568692998045, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 16313 + }, + { + "epoch": 0.16314, + "grad_norm": 1.0842697938394592, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 16314 + }, + { + "epoch": 0.16315, + "grad_norm": 0.9431357428962619, + "learning_rate": 0.003, + "loss": 4.043, + "step": 16315 + }, + { + "epoch": 0.16316, + "grad_norm": 1.0716060539300458, + "learning_rate": 0.003, + "loss": 4.071, + "step": 16316 + }, + { + "epoch": 0.16317, + "grad_norm": 1.1337564497282882, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 16317 + }, + { + "epoch": 0.16318, + "grad_norm": 0.9098344065858708, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 16318 + }, + { + "epoch": 0.16319, + "grad_norm": 0.764733383661203, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 16319 + }, + { + "epoch": 0.1632, + "grad_norm": 0.989113368608259, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 16320 + }, + { + "epoch": 0.16321, + "grad_norm": 1.4221640200985568, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 16321 + }, + { + "epoch": 0.16322, + "grad_norm": 0.8060344183425783, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 16322 + }, + { + "epoch": 0.16323, + "grad_norm": 0.717067507045507, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 16323 + }, + { + "epoch": 0.16324, + "grad_norm": 0.902545989268112, + "learning_rate": 0.003, + "loss": 4.072, + "step": 16324 + }, + { + "epoch": 0.16325, + "grad_norm": 0.9778525894729883, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 16325 + }, + { + "epoch": 0.16326, + "grad_norm": 0.9729281256787194, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 16326 + }, + { + "epoch": 0.16327, + "grad_norm": 0.9944511999015029, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 16327 + }, + { + "epoch": 0.16328, + "grad_norm": 1.1269057276097976, + "learning_rate": 0.003, + "loss": 4.0933, + "step": 16328 + }, + { + "epoch": 0.16329, + "grad_norm": 0.9386231269781623, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 16329 + }, + { + "epoch": 0.1633, + "grad_norm": 1.0010991980529593, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 16330 + }, + { + "epoch": 0.16331, + "grad_norm": 1.1567276932841146, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 16331 + }, + { + "epoch": 0.16332, + "grad_norm": 0.918892986296925, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 16332 + }, + { + "epoch": 0.16333, + "grad_norm": 0.8885396418231442, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 16333 + }, + { + "epoch": 0.16334, + "grad_norm": 1.0731813429273338, + "learning_rate": 0.003, + "loss": 4.0945, + "step": 16334 + }, + { + "epoch": 0.16335, + "grad_norm": 1.1600565339099853, + "learning_rate": 0.003, + "loss": 4.1049, + "step": 16335 + }, + { + "epoch": 0.16336, + "grad_norm": 1.0446881917130995, + "learning_rate": 0.003, + "loss": 4.0908, + "step": 16336 + }, + { + "epoch": 0.16337, + "grad_norm": 1.0955783273564725, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 16337 + }, + { + "epoch": 0.16338, + "grad_norm": 1.0493879605496452, + "learning_rate": 0.003, + "loss": 4.0856, + "step": 16338 + }, + { + "epoch": 0.16339, + "grad_norm": 1.063424446144055, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 16339 + }, + { + "epoch": 0.1634, + "grad_norm": 1.057963878547314, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 16340 + }, + { + "epoch": 0.16341, + "grad_norm": 0.8365247098516165, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 16341 + }, + { + "epoch": 0.16342, + "grad_norm": 0.7980942021618586, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 16342 + }, + { + "epoch": 0.16343, + "grad_norm": 0.8389368759108288, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 16343 + }, + { + "epoch": 0.16344, + "grad_norm": 0.957878591739511, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 16344 + }, + { + "epoch": 0.16345, + "grad_norm": 1.1979360488150872, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 16345 + }, + { + "epoch": 0.16346, + "grad_norm": 0.9921886419235585, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 16346 + }, + { + "epoch": 0.16347, + "grad_norm": 1.1817075040965612, + "learning_rate": 0.003, + "loss": 4.0939, + "step": 16347 + }, + { + "epoch": 0.16348, + "grad_norm": 1.0151571284630665, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 16348 + }, + { + "epoch": 0.16349, + "grad_norm": 0.9467473115666001, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 16349 + }, + { + "epoch": 0.1635, + "grad_norm": 1.0146223192595536, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 16350 + }, + { + "epoch": 0.16351, + "grad_norm": 1.0033118857806405, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 16351 + }, + { + "epoch": 0.16352, + "grad_norm": 1.117343939722878, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 16352 + }, + { + "epoch": 0.16353, + "grad_norm": 1.1036136375171632, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 16353 + }, + { + "epoch": 0.16354, + "grad_norm": 0.99646635449294, + "learning_rate": 0.003, + "loss": 4.1054, + "step": 16354 + }, + { + "epoch": 0.16355, + "grad_norm": 1.227144176627965, + "learning_rate": 0.003, + "loss": 4.1056, + "step": 16355 + }, + { + "epoch": 0.16356, + "grad_norm": 1.0496861434469587, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 16356 + }, + { + "epoch": 0.16357, + "grad_norm": 1.1147169522842697, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 16357 + }, + { + "epoch": 0.16358, + "grad_norm": 0.9681949498501194, + "learning_rate": 0.003, + "loss": 4.0935, + "step": 16358 + }, + { + "epoch": 0.16359, + "grad_norm": 1.1658480195581276, + "learning_rate": 0.003, + "loss": 4.0845, + "step": 16359 + }, + { + "epoch": 0.1636, + "grad_norm": 1.1064128865989298, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 16360 + }, + { + "epoch": 0.16361, + "grad_norm": 1.1970777778186277, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 16361 + }, + { + "epoch": 0.16362, + "grad_norm": 0.8019608815185553, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 16362 + }, + { + "epoch": 0.16363, + "grad_norm": 0.7777764906370227, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 16363 + }, + { + "epoch": 0.16364, + "grad_norm": 0.8276430152098931, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 16364 + }, + { + "epoch": 0.16365, + "grad_norm": 0.8341397184283724, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 16365 + }, + { + "epoch": 0.16366, + "grad_norm": 0.7766562281629581, + "learning_rate": 0.003, + "loss": 4.071, + "step": 16366 + }, + { + "epoch": 0.16367, + "grad_norm": 0.8267143587507615, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 16367 + }, + { + "epoch": 0.16368, + "grad_norm": 0.9247083877881385, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 16368 + }, + { + "epoch": 0.16369, + "grad_norm": 1.150258953052074, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 16369 + }, + { + "epoch": 0.1637, + "grad_norm": 0.9014421652955448, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 16370 + }, + { + "epoch": 0.16371, + "grad_norm": 0.8959122638039856, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 16371 + }, + { + "epoch": 0.16372, + "grad_norm": 1.0214712646602857, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 16372 + }, + { + "epoch": 0.16373, + "grad_norm": 0.9063137085749958, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 16373 + }, + { + "epoch": 0.16374, + "grad_norm": 0.907444617215819, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 16374 + }, + { + "epoch": 0.16375, + "grad_norm": 1.1639215511595487, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 16375 + }, + { + "epoch": 0.16376, + "grad_norm": 1.0398165215559763, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 16376 + }, + { + "epoch": 0.16377, + "grad_norm": 0.9808303690725838, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 16377 + }, + { + "epoch": 0.16378, + "grad_norm": 1.3106964304277915, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 16378 + }, + { + "epoch": 0.16379, + "grad_norm": 0.8686625471574007, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 16379 + }, + { + "epoch": 0.1638, + "grad_norm": 0.6623647642625452, + "learning_rate": 0.003, + "loss": 4.059, + "step": 16380 + }, + { + "epoch": 0.16381, + "grad_norm": 0.6821626907077121, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 16381 + }, + { + "epoch": 0.16382, + "grad_norm": 0.7734896650737811, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 16382 + }, + { + "epoch": 0.16383, + "grad_norm": 1.011075700406094, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 16383 + }, + { + "epoch": 0.16384, + "grad_norm": 1.1707852398011138, + "learning_rate": 0.003, + "loss": 4.075, + "step": 16384 + }, + { + "epoch": 0.16385, + "grad_norm": 0.8143778009027428, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 16385 + }, + { + "epoch": 0.16386, + "grad_norm": 0.8447538661101776, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 16386 + }, + { + "epoch": 0.16387, + "grad_norm": 0.7571944137617708, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 16387 + }, + { + "epoch": 0.16388, + "grad_norm": 0.8111219139548167, + "learning_rate": 0.003, + "loss": 4.062, + "step": 16388 + }, + { + "epoch": 0.16389, + "grad_norm": 0.9220805478097504, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 16389 + }, + { + "epoch": 0.1639, + "grad_norm": 1.0250751580042508, + "learning_rate": 0.003, + "loss": 4.0905, + "step": 16390 + }, + { + "epoch": 0.16391, + "grad_norm": 1.2460607349036437, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 16391 + }, + { + "epoch": 0.16392, + "grad_norm": 0.9073005824781849, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 16392 + }, + { + "epoch": 0.16393, + "grad_norm": 1.136983107365839, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 16393 + }, + { + "epoch": 0.16394, + "grad_norm": 1.1653827233525922, + "learning_rate": 0.003, + "loss": 4.1051, + "step": 16394 + }, + { + "epoch": 0.16395, + "grad_norm": 1.080308789650068, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 16395 + }, + { + "epoch": 0.16396, + "grad_norm": 1.0133641726093063, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 16396 + }, + { + "epoch": 0.16397, + "grad_norm": 1.0389843328991022, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 16397 + }, + { + "epoch": 0.16398, + "grad_norm": 1.1439436315849214, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 16398 + }, + { + "epoch": 0.16399, + "grad_norm": 1.1114739639970634, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 16399 + }, + { + "epoch": 0.164, + "grad_norm": 0.8706043882850701, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 16400 + }, + { + "epoch": 0.16401, + "grad_norm": 0.8145519421242345, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 16401 + }, + { + "epoch": 0.16402, + "grad_norm": 0.7270659029769548, + "learning_rate": 0.003, + "loss": 4.0953, + "step": 16402 + }, + { + "epoch": 0.16403, + "grad_norm": 0.8051415307213541, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 16403 + }, + { + "epoch": 0.16404, + "grad_norm": 0.7932167670290939, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 16404 + }, + { + "epoch": 0.16405, + "grad_norm": 0.8271384276147279, + "learning_rate": 0.003, + "loss": 4.063, + "step": 16405 + }, + { + "epoch": 0.16406, + "grad_norm": 0.9335677116648193, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 16406 + }, + { + "epoch": 0.16407, + "grad_norm": 0.9423075440417622, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 16407 + }, + { + "epoch": 0.16408, + "grad_norm": 1.0805196157573864, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 16408 + }, + { + "epoch": 0.16409, + "grad_norm": 1.4057313831023857, + "learning_rate": 0.003, + "loss": 4.051, + "step": 16409 + }, + { + "epoch": 0.1641, + "grad_norm": 0.9581260077814677, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 16410 + }, + { + "epoch": 0.16411, + "grad_norm": 1.1324979069835843, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 16411 + }, + { + "epoch": 0.16412, + "grad_norm": 1.0494145456907402, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 16412 + }, + { + "epoch": 0.16413, + "grad_norm": 1.0645360194432658, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 16413 + }, + { + "epoch": 0.16414, + "grad_norm": 0.9177855235604141, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 16414 + }, + { + "epoch": 0.16415, + "grad_norm": 0.8998954553779329, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 16415 + }, + { + "epoch": 0.16416, + "grad_norm": 0.9915151595433971, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 16416 + }, + { + "epoch": 0.16417, + "grad_norm": 1.1092539469573264, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 16417 + }, + { + "epoch": 0.16418, + "grad_norm": 0.8843537025900154, + "learning_rate": 0.003, + "loss": 4.071, + "step": 16418 + }, + { + "epoch": 0.16419, + "grad_norm": 1.0833017074823947, + "learning_rate": 0.003, + "loss": 4.0898, + "step": 16419 + }, + { + "epoch": 0.1642, + "grad_norm": 1.0982661109217768, + "learning_rate": 0.003, + "loss": 4.1124, + "step": 16420 + }, + { + "epoch": 0.16421, + "grad_norm": 1.131486121507006, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 16421 + }, + { + "epoch": 0.16422, + "grad_norm": 0.8981996393231813, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 16422 + }, + { + "epoch": 0.16423, + "grad_norm": 1.0057711959692832, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 16423 + }, + { + "epoch": 0.16424, + "grad_norm": 1.1707306553810128, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 16424 + }, + { + "epoch": 0.16425, + "grad_norm": 1.0616284351659566, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 16425 + }, + { + "epoch": 0.16426, + "grad_norm": 0.9534166505928695, + "learning_rate": 0.003, + "loss": 4.0974, + "step": 16426 + }, + { + "epoch": 0.16427, + "grad_norm": 1.0814277109411952, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 16427 + }, + { + "epoch": 0.16428, + "grad_norm": 0.9757381574325722, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 16428 + }, + { + "epoch": 0.16429, + "grad_norm": 0.9191966897493564, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 16429 + }, + { + "epoch": 0.1643, + "grad_norm": 1.1990370233454235, + "learning_rate": 0.003, + "loss": 4.074, + "step": 16430 + }, + { + "epoch": 0.16431, + "grad_norm": 1.1102911689529398, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 16431 + }, + { + "epoch": 0.16432, + "grad_norm": 1.0959868024277635, + "learning_rate": 0.003, + "loss": 4.092, + "step": 16432 + }, + { + "epoch": 0.16433, + "grad_norm": 1.1256293232480539, + "learning_rate": 0.003, + "loss": 4.0902, + "step": 16433 + }, + { + "epoch": 0.16434, + "grad_norm": 0.9470639978134163, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 16434 + }, + { + "epoch": 0.16435, + "grad_norm": 1.21338169943296, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 16435 + }, + { + "epoch": 0.16436, + "grad_norm": 0.9138954431583326, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 16436 + }, + { + "epoch": 0.16437, + "grad_norm": 0.832309421246958, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 16437 + }, + { + "epoch": 0.16438, + "grad_norm": 0.7424344341010659, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 16438 + }, + { + "epoch": 0.16439, + "grad_norm": 0.7811089174584779, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 16439 + }, + { + "epoch": 0.1644, + "grad_norm": 0.9065795100615794, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 16440 + }, + { + "epoch": 0.16441, + "grad_norm": 0.892297885535587, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 16441 + }, + { + "epoch": 0.16442, + "grad_norm": 0.9065285606674014, + "learning_rate": 0.003, + "loss": 4.043, + "step": 16442 + }, + { + "epoch": 0.16443, + "grad_norm": 0.935434566593761, + "learning_rate": 0.003, + "loss": 4.0963, + "step": 16443 + }, + { + "epoch": 0.16444, + "grad_norm": 1.062561080559196, + "learning_rate": 0.003, + "loss": 4.0876, + "step": 16444 + }, + { + "epoch": 0.16445, + "grad_norm": 1.1591216162489975, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 16445 + }, + { + "epoch": 0.16446, + "grad_norm": 1.0189865642448739, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 16446 + }, + { + "epoch": 0.16447, + "grad_norm": 1.153922164687869, + "learning_rate": 0.003, + "loss": 4.0933, + "step": 16447 + }, + { + "epoch": 0.16448, + "grad_norm": 1.0521957544677183, + "learning_rate": 0.003, + "loss": 4.0891, + "step": 16448 + }, + { + "epoch": 0.16449, + "grad_norm": 1.0516150784188352, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 16449 + }, + { + "epoch": 0.1645, + "grad_norm": 0.9925373771511623, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 16450 + }, + { + "epoch": 0.16451, + "grad_norm": 1.2839748592896216, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 16451 + }, + { + "epoch": 0.16452, + "grad_norm": 0.783014127117882, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 16452 + }, + { + "epoch": 0.16453, + "grad_norm": 0.759691939357197, + "learning_rate": 0.003, + "loss": 4.083, + "step": 16453 + }, + { + "epoch": 0.16454, + "grad_norm": 0.7881937528650392, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 16454 + }, + { + "epoch": 0.16455, + "grad_norm": 0.6958994638819804, + "learning_rate": 0.003, + "loss": 4.0857, + "step": 16455 + }, + { + "epoch": 0.16456, + "grad_norm": 0.7216816995302018, + "learning_rate": 0.003, + "loss": 4.072, + "step": 16456 + }, + { + "epoch": 0.16457, + "grad_norm": 0.7175487318759499, + "learning_rate": 0.003, + "loss": 4.0874, + "step": 16457 + }, + { + "epoch": 0.16458, + "grad_norm": 0.7247869326531684, + "learning_rate": 0.003, + "loss": 4.053, + "step": 16458 + }, + { + "epoch": 0.16459, + "grad_norm": 0.6712250465945199, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 16459 + }, + { + "epoch": 0.1646, + "grad_norm": 0.7498029392201481, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 16460 + }, + { + "epoch": 0.16461, + "grad_norm": 0.8204551339614574, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 16461 + }, + { + "epoch": 0.16462, + "grad_norm": 0.9806102192176226, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 16462 + }, + { + "epoch": 0.16463, + "grad_norm": 1.0907096026877074, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 16463 + }, + { + "epoch": 0.16464, + "grad_norm": 0.9654752170722813, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 16464 + }, + { + "epoch": 0.16465, + "grad_norm": 1.0031196012618604, + "learning_rate": 0.003, + "loss": 4.0868, + "step": 16465 + }, + { + "epoch": 0.16466, + "grad_norm": 1.2792766434931824, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 16466 + }, + { + "epoch": 0.16467, + "grad_norm": 1.1636965847487744, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 16467 + }, + { + "epoch": 0.16468, + "grad_norm": 0.9782752762281298, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 16468 + }, + { + "epoch": 0.16469, + "grad_norm": 1.071496219851254, + "learning_rate": 0.003, + "loss": 4.0872, + "step": 16469 + }, + { + "epoch": 0.1647, + "grad_norm": 1.2550004639083363, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 16470 + }, + { + "epoch": 0.16471, + "grad_norm": 1.41392454991626, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 16471 + }, + { + "epoch": 0.16472, + "grad_norm": 0.7866787837482843, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 16472 + }, + { + "epoch": 0.16473, + "grad_norm": 0.797120795448815, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 16473 + }, + { + "epoch": 0.16474, + "grad_norm": 0.9396315665753611, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 16474 + }, + { + "epoch": 0.16475, + "grad_norm": 1.2281433619296347, + "learning_rate": 0.003, + "loss": 4.0941, + "step": 16475 + }, + { + "epoch": 0.16476, + "grad_norm": 0.9226260217959018, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 16476 + }, + { + "epoch": 0.16477, + "grad_norm": 1.107630562747444, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 16477 + }, + { + "epoch": 0.16478, + "grad_norm": 0.9572586946192534, + "learning_rate": 0.003, + "loss": 4.0939, + "step": 16478 + }, + { + "epoch": 0.16479, + "grad_norm": 1.0726950883800912, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 16479 + }, + { + "epoch": 0.1648, + "grad_norm": 1.1741457360320326, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 16480 + }, + { + "epoch": 0.16481, + "grad_norm": 0.8198320536528987, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 16481 + }, + { + "epoch": 0.16482, + "grad_norm": 0.8090782296319683, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 16482 + }, + { + "epoch": 0.16483, + "grad_norm": 0.7782579343682644, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 16483 + }, + { + "epoch": 0.16484, + "grad_norm": 0.7621437316802481, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 16484 + }, + { + "epoch": 0.16485, + "grad_norm": 0.872840821677062, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 16485 + }, + { + "epoch": 0.16486, + "grad_norm": 1.2166348719244713, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 16486 + }, + { + "epoch": 0.16487, + "grad_norm": 1.2909376631805496, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 16487 + }, + { + "epoch": 0.16488, + "grad_norm": 0.9088987707781695, + "learning_rate": 0.003, + "loss": 4.0931, + "step": 16488 + }, + { + "epoch": 0.16489, + "grad_norm": 0.9623862831326357, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 16489 + }, + { + "epoch": 0.1649, + "grad_norm": 1.1772353881245123, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 16490 + }, + { + "epoch": 0.16491, + "grad_norm": 1.0490703754196713, + "learning_rate": 0.003, + "loss": 4.048, + "step": 16491 + }, + { + "epoch": 0.16492, + "grad_norm": 1.0774770871176274, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 16492 + }, + { + "epoch": 0.16493, + "grad_norm": 1.046447269993178, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 16493 + }, + { + "epoch": 0.16494, + "grad_norm": 1.1606990721003372, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 16494 + }, + { + "epoch": 0.16495, + "grad_norm": 1.0680262518436368, + "learning_rate": 0.003, + "loss": 4.0992, + "step": 16495 + }, + { + "epoch": 0.16496, + "grad_norm": 0.9857558866577001, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 16496 + }, + { + "epoch": 0.16497, + "grad_norm": 0.9861643466612614, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 16497 + }, + { + "epoch": 0.16498, + "grad_norm": 1.178411185393598, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 16498 + }, + { + "epoch": 0.16499, + "grad_norm": 1.0426795787305672, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 16499 + }, + { + "epoch": 0.165, + "grad_norm": 1.1116833995627224, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 16500 + }, + { + "epoch": 0.16501, + "grad_norm": 1.0953207616767227, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 16501 + }, + { + "epoch": 0.16502, + "grad_norm": 0.8612591402390954, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 16502 + }, + { + "epoch": 0.16503, + "grad_norm": 0.8726171951561549, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 16503 + }, + { + "epoch": 0.16504, + "grad_norm": 0.8826139740695993, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 16504 + }, + { + "epoch": 0.16505, + "grad_norm": 1.0251839343838984, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 16505 + }, + { + "epoch": 0.16506, + "grad_norm": 1.211376131791072, + "learning_rate": 0.003, + "loss": 4.0966, + "step": 16506 + }, + { + "epoch": 0.16507, + "grad_norm": 0.916072357971208, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 16507 + }, + { + "epoch": 0.16508, + "grad_norm": 1.0062182789338099, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 16508 + }, + { + "epoch": 0.16509, + "grad_norm": 1.1562172594586548, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 16509 + }, + { + "epoch": 0.1651, + "grad_norm": 0.9748312393302015, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 16510 + }, + { + "epoch": 0.16511, + "grad_norm": 0.8326679508893662, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 16511 + }, + { + "epoch": 0.16512, + "grad_norm": 0.9406119671611223, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 16512 + }, + { + "epoch": 0.16513, + "grad_norm": 1.1010448744867094, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 16513 + }, + { + "epoch": 0.16514, + "grad_norm": 0.9206358896487637, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 16514 + }, + { + "epoch": 0.16515, + "grad_norm": 1.126646107244673, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 16515 + }, + { + "epoch": 0.16516, + "grad_norm": 0.9806369532450748, + "learning_rate": 0.003, + "loss": 4.058, + "step": 16516 + }, + { + "epoch": 0.16517, + "grad_norm": 1.0504319036303607, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 16517 + }, + { + "epoch": 0.16518, + "grad_norm": 0.9845074156501539, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 16518 + }, + { + "epoch": 0.16519, + "grad_norm": 1.157981415709716, + "learning_rate": 0.003, + "loss": 4.0874, + "step": 16519 + }, + { + "epoch": 0.1652, + "grad_norm": 1.0583786972611324, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 16520 + }, + { + "epoch": 0.16521, + "grad_norm": 1.1131173437907504, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 16521 + }, + { + "epoch": 0.16522, + "grad_norm": 0.8221876578103038, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 16522 + }, + { + "epoch": 0.16523, + "grad_norm": 0.702218323622442, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 16523 + }, + { + "epoch": 0.16524, + "grad_norm": 0.6538447803577517, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 16524 + }, + { + "epoch": 0.16525, + "grad_norm": 0.7842846376343932, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 16525 + }, + { + "epoch": 0.16526, + "grad_norm": 0.952893081418633, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 16526 + }, + { + "epoch": 0.16527, + "grad_norm": 1.2687541820269337, + "learning_rate": 0.003, + "loss": 4.0907, + "step": 16527 + }, + { + "epoch": 0.16528, + "grad_norm": 0.9001882340958044, + "learning_rate": 0.003, + "loss": 4.037, + "step": 16528 + }, + { + "epoch": 0.16529, + "grad_norm": 0.9870761863323391, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 16529 + }, + { + "epoch": 0.1653, + "grad_norm": 1.1692184344994727, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 16530 + }, + { + "epoch": 0.16531, + "grad_norm": 1.0368453234743829, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 16531 + }, + { + "epoch": 0.16532, + "grad_norm": 1.100235988156221, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 16532 + }, + { + "epoch": 0.16533, + "grad_norm": 1.0669765451799496, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 16533 + }, + { + "epoch": 0.16534, + "grad_norm": 1.091921136608918, + "learning_rate": 0.003, + "loss": 4.0872, + "step": 16534 + }, + { + "epoch": 0.16535, + "grad_norm": 1.0152736601996277, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 16535 + }, + { + "epoch": 0.16536, + "grad_norm": 1.1671428824791577, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 16536 + }, + { + "epoch": 0.16537, + "grad_norm": 1.1659172916161982, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 16537 + }, + { + "epoch": 0.16538, + "grad_norm": 0.9797549513637428, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 16538 + }, + { + "epoch": 0.16539, + "grad_norm": 1.0479549292970163, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 16539 + }, + { + "epoch": 0.1654, + "grad_norm": 1.049014214480001, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 16540 + }, + { + "epoch": 0.16541, + "grad_norm": 1.0398309970879558, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 16541 + }, + { + "epoch": 0.16542, + "grad_norm": 0.9909896700766191, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 16542 + }, + { + "epoch": 0.16543, + "grad_norm": 1.0469468447311614, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 16543 + }, + { + "epoch": 0.16544, + "grad_norm": 1.0481623404487392, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 16544 + }, + { + "epoch": 0.16545, + "grad_norm": 0.9858282133404034, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 16545 + }, + { + "epoch": 0.16546, + "grad_norm": 1.0117779896459889, + "learning_rate": 0.003, + "loss": 4.056, + "step": 16546 + }, + { + "epoch": 0.16547, + "grad_norm": 1.0050112259361548, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 16547 + }, + { + "epoch": 0.16548, + "grad_norm": 1.0558764895451618, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 16548 + }, + { + "epoch": 0.16549, + "grad_norm": 1.0530624119557994, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 16549 + }, + { + "epoch": 0.1655, + "grad_norm": 0.990642594436276, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 16550 + }, + { + "epoch": 0.16551, + "grad_norm": 1.12919321095025, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 16551 + }, + { + "epoch": 0.16552, + "grad_norm": 0.9804382332716368, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 16552 + }, + { + "epoch": 0.16553, + "grad_norm": 1.1431988774096116, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 16553 + }, + { + "epoch": 0.16554, + "grad_norm": 1.2648142419604074, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 16554 + }, + { + "epoch": 0.16555, + "grad_norm": 1.0620828851924469, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 16555 + }, + { + "epoch": 0.16556, + "grad_norm": 1.0388870980362406, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 16556 + }, + { + "epoch": 0.16557, + "grad_norm": 0.9877093287365224, + "learning_rate": 0.003, + "loss": 4.072, + "step": 16557 + }, + { + "epoch": 0.16558, + "grad_norm": 1.0313207334874341, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 16558 + }, + { + "epoch": 0.16559, + "grad_norm": 1.1853885182243893, + "learning_rate": 0.003, + "loss": 4.0918, + "step": 16559 + }, + { + "epoch": 0.1656, + "grad_norm": 0.8909004169543192, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 16560 + }, + { + "epoch": 0.16561, + "grad_norm": 0.8752037471141652, + "learning_rate": 0.003, + "loss": 4.093, + "step": 16561 + }, + { + "epoch": 0.16562, + "grad_norm": 0.9760727868903457, + "learning_rate": 0.003, + "loss": 4.0878, + "step": 16562 + }, + { + "epoch": 0.16563, + "grad_norm": 1.0726396939693692, + "learning_rate": 0.003, + "loss": 4.074, + "step": 16563 + }, + { + "epoch": 0.16564, + "grad_norm": 1.044061726358533, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 16564 + }, + { + "epoch": 0.16565, + "grad_norm": 1.3664474441380963, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 16565 + }, + { + "epoch": 0.16566, + "grad_norm": 0.8903763645937675, + "learning_rate": 0.003, + "loss": 4.103, + "step": 16566 + }, + { + "epoch": 0.16567, + "grad_norm": 0.9194319576790123, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 16567 + }, + { + "epoch": 0.16568, + "grad_norm": 0.8712822559317813, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 16568 + }, + { + "epoch": 0.16569, + "grad_norm": 0.9615586599160553, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 16569 + }, + { + "epoch": 0.1657, + "grad_norm": 1.1788690663159356, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 16570 + }, + { + "epoch": 0.16571, + "grad_norm": 0.869699307114246, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 16571 + }, + { + "epoch": 0.16572, + "grad_norm": 0.9763249020447643, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 16572 + }, + { + "epoch": 0.16573, + "grad_norm": 1.1382973105974474, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 16573 + }, + { + "epoch": 0.16574, + "grad_norm": 0.9725905593152545, + "learning_rate": 0.003, + "loss": 4.046, + "step": 16574 + }, + { + "epoch": 0.16575, + "grad_norm": 1.0510344940894694, + "learning_rate": 0.003, + "loss": 4.1018, + "step": 16575 + }, + { + "epoch": 0.16576, + "grad_norm": 1.193430231381592, + "learning_rate": 0.003, + "loss": 4.076, + "step": 16576 + }, + { + "epoch": 0.16577, + "grad_norm": 0.891287317479745, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 16577 + }, + { + "epoch": 0.16578, + "grad_norm": 0.8715435195533318, + "learning_rate": 0.003, + "loss": 4.067, + "step": 16578 + }, + { + "epoch": 0.16579, + "grad_norm": 0.922628419164927, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 16579 + }, + { + "epoch": 0.1658, + "grad_norm": 0.8549084047887486, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 16580 + }, + { + "epoch": 0.16581, + "grad_norm": 0.965725126726442, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 16581 + }, + { + "epoch": 0.16582, + "grad_norm": 1.0755935743114604, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 16582 + }, + { + "epoch": 0.16583, + "grad_norm": 0.9590465737915841, + "learning_rate": 0.003, + "loss": 4.073, + "step": 16583 + }, + { + "epoch": 0.16584, + "grad_norm": 0.990950236671472, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 16584 + }, + { + "epoch": 0.16585, + "grad_norm": 0.9934570820406158, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 16585 + }, + { + "epoch": 0.16586, + "grad_norm": 1.0638531523595982, + "learning_rate": 0.003, + "loss": 4.0858, + "step": 16586 + }, + { + "epoch": 0.16587, + "grad_norm": 0.9066100288845095, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 16587 + }, + { + "epoch": 0.16588, + "grad_norm": 0.9034953673261473, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 16588 + }, + { + "epoch": 0.16589, + "grad_norm": 1.0035407901851456, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 16589 + }, + { + "epoch": 0.1659, + "grad_norm": 1.1194120003804597, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 16590 + }, + { + "epoch": 0.16591, + "grad_norm": 0.9920276341496579, + "learning_rate": 0.003, + "loss": 4.1017, + "step": 16591 + }, + { + "epoch": 0.16592, + "grad_norm": 1.1800682359801609, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 16592 + }, + { + "epoch": 0.16593, + "grad_norm": 1.1845306454091367, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 16593 + }, + { + "epoch": 0.16594, + "grad_norm": 1.0149085007149032, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 16594 + }, + { + "epoch": 0.16595, + "grad_norm": 1.0531094030944645, + "learning_rate": 0.003, + "loss": 4.0945, + "step": 16595 + }, + { + "epoch": 0.16596, + "grad_norm": 0.861865285899016, + "learning_rate": 0.003, + "loss": 4.064, + "step": 16596 + }, + { + "epoch": 0.16597, + "grad_norm": 1.0401213515304892, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 16597 + }, + { + "epoch": 0.16598, + "grad_norm": 1.0531724853096573, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 16598 + }, + { + "epoch": 0.16599, + "grad_norm": 1.01488928282496, + "learning_rate": 0.003, + "loss": 4.0964, + "step": 16599 + }, + { + "epoch": 0.166, + "grad_norm": 1.2894703013947775, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 16600 + }, + { + "epoch": 0.16601, + "grad_norm": 0.9938731176768969, + "learning_rate": 0.003, + "loss": 4.1044, + "step": 16601 + }, + { + "epoch": 0.16602, + "grad_norm": 1.1908981083062715, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 16602 + }, + { + "epoch": 0.16603, + "grad_norm": 0.9392714260805681, + "learning_rate": 0.003, + "loss": 4.0989, + "step": 16603 + }, + { + "epoch": 0.16604, + "grad_norm": 0.9947326874700516, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 16604 + }, + { + "epoch": 0.16605, + "grad_norm": 1.2287753037254965, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 16605 + }, + { + "epoch": 0.16606, + "grad_norm": 0.9146224585006899, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 16606 + }, + { + "epoch": 0.16607, + "grad_norm": 0.8419735086030495, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 16607 + }, + { + "epoch": 0.16608, + "grad_norm": 0.8346721218557821, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 16608 + }, + { + "epoch": 0.16609, + "grad_norm": 0.8621923100589906, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 16609 + }, + { + "epoch": 0.1661, + "grad_norm": 0.8923808721564329, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 16610 + }, + { + "epoch": 0.16611, + "grad_norm": 0.9320318360974327, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 16611 + }, + { + "epoch": 0.16612, + "grad_norm": 1.1450179779046155, + "learning_rate": 0.003, + "loss": 4.063, + "step": 16612 + }, + { + "epoch": 0.16613, + "grad_norm": 1.0455562692006997, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 16613 + }, + { + "epoch": 0.16614, + "grad_norm": 1.0067766826007265, + "learning_rate": 0.003, + "loss": 4.083, + "step": 16614 + }, + { + "epoch": 0.16615, + "grad_norm": 0.995580188878611, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 16615 + }, + { + "epoch": 0.16616, + "grad_norm": 0.8757351743740567, + "learning_rate": 0.003, + "loss": 4.1142, + "step": 16616 + }, + { + "epoch": 0.16617, + "grad_norm": 0.8759232559403759, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 16617 + }, + { + "epoch": 0.16618, + "grad_norm": 1.0002003535259971, + "learning_rate": 0.003, + "loss": 4.074, + "step": 16618 + }, + { + "epoch": 0.16619, + "grad_norm": 1.0609785378157148, + "learning_rate": 0.003, + "loss": 4.062, + "step": 16619 + }, + { + "epoch": 0.1662, + "grad_norm": 0.9692980043329431, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 16620 + }, + { + "epoch": 0.16621, + "grad_norm": 1.3021079652257457, + "learning_rate": 0.003, + "loss": 4.0855, + "step": 16621 + }, + { + "epoch": 0.16622, + "grad_norm": 0.8124068560730148, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 16622 + }, + { + "epoch": 0.16623, + "grad_norm": 0.8562480592272761, + "learning_rate": 0.003, + "loss": 4.0995, + "step": 16623 + }, + { + "epoch": 0.16624, + "grad_norm": 1.0559352934928505, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 16624 + }, + { + "epoch": 0.16625, + "grad_norm": 1.2219037923822076, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 16625 + }, + { + "epoch": 0.16626, + "grad_norm": 0.952824944706206, + "learning_rate": 0.003, + "loss": 4.0901, + "step": 16626 + }, + { + "epoch": 0.16627, + "grad_norm": 1.0142860751649438, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 16627 + }, + { + "epoch": 0.16628, + "grad_norm": 1.2227364058143282, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 16628 + }, + { + "epoch": 0.16629, + "grad_norm": 0.7408540376083553, + "learning_rate": 0.003, + "loss": 4.1015, + "step": 16629 + }, + { + "epoch": 0.1663, + "grad_norm": 0.748086309388994, + "learning_rate": 0.003, + "loss": 4.047, + "step": 16630 + }, + { + "epoch": 0.16631, + "grad_norm": 0.8108339742876692, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 16631 + }, + { + "epoch": 0.16632, + "grad_norm": 0.9709574955235369, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 16632 + }, + { + "epoch": 0.16633, + "grad_norm": 1.1855398319730404, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 16633 + }, + { + "epoch": 0.16634, + "grad_norm": 0.8827665932988588, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 16634 + }, + { + "epoch": 0.16635, + "grad_norm": 0.8810013038443375, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 16635 + }, + { + "epoch": 0.16636, + "grad_norm": 0.8891090914255272, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 16636 + }, + { + "epoch": 0.16637, + "grad_norm": 0.7888977314051207, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 16637 + }, + { + "epoch": 0.16638, + "grad_norm": 0.6670099537692519, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 16638 + }, + { + "epoch": 0.16639, + "grad_norm": 0.6729654251381538, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 16639 + }, + { + "epoch": 0.1664, + "grad_norm": 0.7075613474836533, + "learning_rate": 0.003, + "loss": 4.082, + "step": 16640 + }, + { + "epoch": 0.16641, + "grad_norm": 0.8158283100609237, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 16641 + }, + { + "epoch": 0.16642, + "grad_norm": 1.0326662537066196, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 16642 + }, + { + "epoch": 0.16643, + "grad_norm": 1.3962251888611972, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 16643 + }, + { + "epoch": 0.16644, + "grad_norm": 0.7576714565054758, + "learning_rate": 0.003, + "loss": 4.081, + "step": 16644 + }, + { + "epoch": 0.16645, + "grad_norm": 0.6788795629893567, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 16645 + }, + { + "epoch": 0.16646, + "grad_norm": 0.8537246923495545, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 16646 + }, + { + "epoch": 0.16647, + "grad_norm": 1.083282033952763, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 16647 + }, + { + "epoch": 0.16648, + "grad_norm": 1.1667861205223728, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 16648 + }, + { + "epoch": 0.16649, + "grad_norm": 0.8588815419353215, + "learning_rate": 0.003, + "loss": 4.0878, + "step": 16649 + }, + { + "epoch": 0.1665, + "grad_norm": 0.8400465527663585, + "learning_rate": 0.003, + "loss": 4.072, + "step": 16650 + }, + { + "epoch": 0.16651, + "grad_norm": 1.0363438966407217, + "learning_rate": 0.003, + "loss": 4.041, + "step": 16651 + }, + { + "epoch": 0.16652, + "grad_norm": 1.0591154612659934, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 16652 + }, + { + "epoch": 0.16653, + "grad_norm": 0.8727278438666786, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 16653 + }, + { + "epoch": 0.16654, + "grad_norm": 1.0248770040144033, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 16654 + }, + { + "epoch": 0.16655, + "grad_norm": 1.3486826554401614, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 16655 + }, + { + "epoch": 0.16656, + "grad_norm": 0.7948528256903451, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 16656 + }, + { + "epoch": 0.16657, + "grad_norm": 0.9164500842441934, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 16657 + }, + { + "epoch": 0.16658, + "grad_norm": 1.3100750793135831, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 16658 + }, + { + "epoch": 0.16659, + "grad_norm": 1.10340497261793, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 16659 + }, + { + "epoch": 0.1666, + "grad_norm": 0.9190229814164225, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 16660 + }, + { + "epoch": 0.16661, + "grad_norm": 1.0097263537717052, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 16661 + }, + { + "epoch": 0.16662, + "grad_norm": 1.078595742046761, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 16662 + }, + { + "epoch": 0.16663, + "grad_norm": 1.0797891149812604, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 16663 + }, + { + "epoch": 0.16664, + "grad_norm": 1.1069921439946269, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 16664 + }, + { + "epoch": 0.16665, + "grad_norm": 1.0176078025683735, + "learning_rate": 0.003, + "loss": 4.0933, + "step": 16665 + }, + { + "epoch": 0.16666, + "grad_norm": 1.1552191696663092, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 16666 + }, + { + "epoch": 0.16667, + "grad_norm": 0.9324404262613838, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 16667 + }, + { + "epoch": 0.16668, + "grad_norm": 1.0019757794996023, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 16668 + }, + { + "epoch": 0.16669, + "grad_norm": 1.0991976045234721, + "learning_rate": 0.003, + "loss": 4.0923, + "step": 16669 + }, + { + "epoch": 0.1667, + "grad_norm": 1.0128849883183946, + "learning_rate": 0.003, + "loss": 4.054, + "step": 16670 + }, + { + "epoch": 0.16671, + "grad_norm": 1.0931849499937962, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 16671 + }, + { + "epoch": 0.16672, + "grad_norm": 0.9812348284640295, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 16672 + }, + { + "epoch": 0.16673, + "grad_norm": 0.9778592212759791, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 16673 + }, + { + "epoch": 0.16674, + "grad_norm": 1.0545120508109425, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 16674 + }, + { + "epoch": 0.16675, + "grad_norm": 1.1110630836832625, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 16675 + }, + { + "epoch": 0.16676, + "grad_norm": 1.0626452338852985, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 16676 + }, + { + "epoch": 0.16677, + "grad_norm": 1.0176436007781615, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 16677 + }, + { + "epoch": 0.16678, + "grad_norm": 0.9645399316578736, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 16678 + }, + { + "epoch": 0.16679, + "grad_norm": 0.9988207449972248, + "learning_rate": 0.003, + "loss": 4.081, + "step": 16679 + }, + { + "epoch": 0.1668, + "grad_norm": 1.134510799795786, + "learning_rate": 0.003, + "loss": 4.0868, + "step": 16680 + }, + { + "epoch": 0.16681, + "grad_norm": 1.100488697152193, + "learning_rate": 0.003, + "loss": 4.057, + "step": 16681 + }, + { + "epoch": 0.16682, + "grad_norm": 1.2389404130360666, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 16682 + }, + { + "epoch": 0.16683, + "grad_norm": 0.8388332581278795, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 16683 + }, + { + "epoch": 0.16684, + "grad_norm": 0.8269767204617885, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 16684 + }, + { + "epoch": 0.16685, + "grad_norm": 0.8157592038436304, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 16685 + }, + { + "epoch": 0.16686, + "grad_norm": 0.9765700745441275, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 16686 + }, + { + "epoch": 0.16687, + "grad_norm": 1.0794222687173145, + "learning_rate": 0.003, + "loss": 4.0868, + "step": 16687 + }, + { + "epoch": 0.16688, + "grad_norm": 1.0909077943788, + "learning_rate": 0.003, + "loss": 4.075, + "step": 16688 + }, + { + "epoch": 0.16689, + "grad_norm": 1.219824943155162, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 16689 + }, + { + "epoch": 0.1669, + "grad_norm": 1.048395137927214, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 16690 + }, + { + "epoch": 0.16691, + "grad_norm": 1.0663218203654112, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 16691 + }, + { + "epoch": 0.16692, + "grad_norm": 0.9334742427104825, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 16692 + }, + { + "epoch": 0.16693, + "grad_norm": 0.9557884791185572, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 16693 + }, + { + "epoch": 0.16694, + "grad_norm": 0.9896525420854605, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 16694 + }, + { + "epoch": 0.16695, + "grad_norm": 1.1147522396309812, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 16695 + }, + { + "epoch": 0.16696, + "grad_norm": 0.967545881632985, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 16696 + }, + { + "epoch": 0.16697, + "grad_norm": 1.1904185482959682, + "learning_rate": 0.003, + "loss": 4.071, + "step": 16697 + }, + { + "epoch": 0.16698, + "grad_norm": 1.2343019313505192, + "learning_rate": 0.003, + "loss": 4.079, + "step": 16698 + }, + { + "epoch": 0.16699, + "grad_norm": 0.8427455504837388, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 16699 + }, + { + "epoch": 0.167, + "grad_norm": 0.8225370876034956, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 16700 + }, + { + "epoch": 0.16701, + "grad_norm": 0.8085602442527429, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 16701 + }, + { + "epoch": 0.16702, + "grad_norm": 0.924180704332601, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 16702 + }, + { + "epoch": 0.16703, + "grad_norm": 1.2240879407061853, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 16703 + }, + { + "epoch": 0.16704, + "grad_norm": 0.9490689634896586, + "learning_rate": 0.003, + "loss": 4.0975, + "step": 16704 + }, + { + "epoch": 0.16705, + "grad_norm": 0.7958362558749051, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 16705 + }, + { + "epoch": 0.16706, + "grad_norm": 0.838292184631257, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 16706 + }, + { + "epoch": 0.16707, + "grad_norm": 0.8606987687685781, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 16707 + }, + { + "epoch": 0.16708, + "grad_norm": 0.9796685769144863, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 16708 + }, + { + "epoch": 0.16709, + "grad_norm": 1.096045077902185, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 16709 + }, + { + "epoch": 0.1671, + "grad_norm": 0.9309562495038861, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 16710 + }, + { + "epoch": 0.16711, + "grad_norm": 0.9743229206111286, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 16711 + }, + { + "epoch": 0.16712, + "grad_norm": 1.0100680797991972, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 16712 + }, + { + "epoch": 0.16713, + "grad_norm": 1.0450623116258029, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 16713 + }, + { + "epoch": 0.16714, + "grad_norm": 0.9891231322078815, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 16714 + }, + { + "epoch": 0.16715, + "grad_norm": 1.094996940419535, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 16715 + }, + { + "epoch": 0.16716, + "grad_norm": 0.9960874222566649, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 16716 + }, + { + "epoch": 0.16717, + "grad_norm": 0.9939629668166133, + "learning_rate": 0.003, + "loss": 4.064, + "step": 16717 + }, + { + "epoch": 0.16718, + "grad_norm": 1.1093330143272073, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 16718 + }, + { + "epoch": 0.16719, + "grad_norm": 1.1815942486345608, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 16719 + }, + { + "epoch": 0.1672, + "grad_norm": 0.9627437395261305, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 16720 + }, + { + "epoch": 0.16721, + "grad_norm": 1.0596581800863236, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 16721 + }, + { + "epoch": 0.16722, + "grad_norm": 1.1024638615750197, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 16722 + }, + { + "epoch": 0.16723, + "grad_norm": 0.7662074372073967, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 16723 + }, + { + "epoch": 0.16724, + "grad_norm": 0.6991167773339303, + "learning_rate": 0.003, + "loss": 4.0954, + "step": 16724 + }, + { + "epoch": 0.16725, + "grad_norm": 1.067051507524648, + "learning_rate": 0.003, + "loss": 4.0987, + "step": 16725 + }, + { + "epoch": 0.16726, + "grad_norm": 1.380038533360582, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 16726 + }, + { + "epoch": 0.16727, + "grad_norm": 0.7534299593543643, + "learning_rate": 0.003, + "loss": 4.052, + "step": 16727 + }, + { + "epoch": 0.16728, + "grad_norm": 0.854952743923592, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 16728 + }, + { + "epoch": 0.16729, + "grad_norm": 0.9084570019378497, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 16729 + }, + { + "epoch": 0.1673, + "grad_norm": 0.9372299684835902, + "learning_rate": 0.003, + "loss": 4.112, + "step": 16730 + }, + { + "epoch": 0.16731, + "grad_norm": 1.1084352635539205, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 16731 + }, + { + "epoch": 0.16732, + "grad_norm": 1.1095834659567618, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 16732 + }, + { + "epoch": 0.16733, + "grad_norm": 0.8875383627681049, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 16733 + }, + { + "epoch": 0.16734, + "grad_norm": 0.929753855875747, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 16734 + }, + { + "epoch": 0.16735, + "grad_norm": 1.1107584415970921, + "learning_rate": 0.003, + "loss": 4.1021, + "step": 16735 + }, + { + "epoch": 0.16736, + "grad_norm": 1.1716051442084694, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 16736 + }, + { + "epoch": 0.16737, + "grad_norm": 0.9385251204843446, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 16737 + }, + { + "epoch": 0.16738, + "grad_norm": 0.8831569413638083, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 16738 + }, + { + "epoch": 0.16739, + "grad_norm": 0.9201453441704882, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 16739 + }, + { + "epoch": 0.1674, + "grad_norm": 1.1280700125204504, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 16740 + }, + { + "epoch": 0.16741, + "grad_norm": 0.989007799828296, + "learning_rate": 0.003, + "loss": 4.0871, + "step": 16741 + }, + { + "epoch": 0.16742, + "grad_norm": 1.2139890833345763, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 16742 + }, + { + "epoch": 0.16743, + "grad_norm": 0.9008847404763056, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 16743 + }, + { + "epoch": 0.16744, + "grad_norm": 0.8444217770931906, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 16744 + }, + { + "epoch": 0.16745, + "grad_norm": 0.9518179415504797, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 16745 + }, + { + "epoch": 0.16746, + "grad_norm": 1.2246756147054043, + "learning_rate": 0.003, + "loss": 4.1048, + "step": 16746 + }, + { + "epoch": 0.16747, + "grad_norm": 0.9274369468795292, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 16747 + }, + { + "epoch": 0.16748, + "grad_norm": 1.0275999741983568, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 16748 + }, + { + "epoch": 0.16749, + "grad_norm": 1.0810520441426166, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 16749 + }, + { + "epoch": 0.1675, + "grad_norm": 0.8926220274232461, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 16750 + }, + { + "epoch": 0.16751, + "grad_norm": 0.9055468121801162, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 16751 + }, + { + "epoch": 0.16752, + "grad_norm": 0.9242556864175553, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 16752 + }, + { + "epoch": 0.16753, + "grad_norm": 0.9994850906480764, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 16753 + }, + { + "epoch": 0.16754, + "grad_norm": 1.1463146687531958, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 16754 + }, + { + "epoch": 0.16755, + "grad_norm": 1.0729710026051373, + "learning_rate": 0.003, + "loss": 4.0969, + "step": 16755 + }, + { + "epoch": 0.16756, + "grad_norm": 1.2149440665785458, + "learning_rate": 0.003, + "loss": 4.047, + "step": 16756 + }, + { + "epoch": 0.16757, + "grad_norm": 0.8954213495337094, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 16757 + }, + { + "epoch": 0.16758, + "grad_norm": 1.1594233543122676, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 16758 + }, + { + "epoch": 0.16759, + "grad_norm": 1.2013007238062559, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 16759 + }, + { + "epoch": 0.1676, + "grad_norm": 0.8510918717706047, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 16760 + }, + { + "epoch": 0.16761, + "grad_norm": 0.8291981212091329, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 16761 + }, + { + "epoch": 0.16762, + "grad_norm": 0.9649216281509695, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 16762 + }, + { + "epoch": 0.16763, + "grad_norm": 1.1066592884171385, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 16763 + }, + { + "epoch": 0.16764, + "grad_norm": 0.9035061892197336, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 16764 + }, + { + "epoch": 0.16765, + "grad_norm": 0.9535651751678924, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 16765 + }, + { + "epoch": 0.16766, + "grad_norm": 1.0381049373462796, + "learning_rate": 0.003, + "loss": 4.1111, + "step": 16766 + }, + { + "epoch": 0.16767, + "grad_norm": 1.1508602329919129, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 16767 + }, + { + "epoch": 0.16768, + "grad_norm": 0.8491666406389142, + "learning_rate": 0.003, + "loss": 4.0873, + "step": 16768 + }, + { + "epoch": 0.16769, + "grad_norm": 0.7653230261118272, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 16769 + }, + { + "epoch": 0.1677, + "grad_norm": 0.7182935795997769, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 16770 + }, + { + "epoch": 0.16771, + "grad_norm": 0.7755779054548725, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 16771 + }, + { + "epoch": 0.16772, + "grad_norm": 1.0784784755301788, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 16772 + }, + { + "epoch": 0.16773, + "grad_norm": 1.367010474191669, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 16773 + }, + { + "epoch": 0.16774, + "grad_norm": 0.8864699780565714, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 16774 + }, + { + "epoch": 0.16775, + "grad_norm": 0.8832301790265598, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 16775 + }, + { + "epoch": 0.16776, + "grad_norm": 0.9414468918404895, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 16776 + }, + { + "epoch": 0.16777, + "grad_norm": 1.1821884463244492, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 16777 + }, + { + "epoch": 0.16778, + "grad_norm": 1.0925762693902057, + "learning_rate": 0.003, + "loss": 4.0994, + "step": 16778 + }, + { + "epoch": 0.16779, + "grad_norm": 0.9129958670626298, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 16779 + }, + { + "epoch": 0.1678, + "grad_norm": 1.0384592287524574, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 16780 + }, + { + "epoch": 0.16781, + "grad_norm": 1.1554117312948697, + "learning_rate": 0.003, + "loss": 4.1149, + "step": 16781 + }, + { + "epoch": 0.16782, + "grad_norm": 0.8753135302691717, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 16782 + }, + { + "epoch": 0.16783, + "grad_norm": 0.867856359310337, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 16783 + }, + { + "epoch": 0.16784, + "grad_norm": 0.8613243802815497, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 16784 + }, + { + "epoch": 0.16785, + "grad_norm": 0.8723029476219484, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 16785 + }, + { + "epoch": 0.16786, + "grad_norm": 1.0515434223530777, + "learning_rate": 0.003, + "loss": 4.0938, + "step": 16786 + }, + { + "epoch": 0.16787, + "grad_norm": 1.24738919406831, + "learning_rate": 0.003, + "loss": 4.061, + "step": 16787 + }, + { + "epoch": 0.16788, + "grad_norm": 0.7116152926906005, + "learning_rate": 0.003, + "loss": 4.092, + "step": 16788 + }, + { + "epoch": 0.16789, + "grad_norm": 0.7360862782619644, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 16789 + }, + { + "epoch": 0.1679, + "grad_norm": 0.8830570335197504, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 16790 + }, + { + "epoch": 0.16791, + "grad_norm": 1.0491439299590968, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 16791 + }, + { + "epoch": 0.16792, + "grad_norm": 1.103257910368256, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 16792 + }, + { + "epoch": 0.16793, + "grad_norm": 1.0381692788924899, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 16793 + }, + { + "epoch": 0.16794, + "grad_norm": 0.963359773471874, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 16794 + }, + { + "epoch": 0.16795, + "grad_norm": 1.1084861242832829, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 16795 + }, + { + "epoch": 0.16796, + "grad_norm": 1.0334262479076088, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 16796 + }, + { + "epoch": 0.16797, + "grad_norm": 1.0359077775927403, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 16797 + }, + { + "epoch": 0.16798, + "grad_norm": 1.0856085817272356, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 16798 + }, + { + "epoch": 0.16799, + "grad_norm": 0.9447958689832883, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 16799 + }, + { + "epoch": 0.168, + "grad_norm": 0.991323960625156, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 16800 + }, + { + "epoch": 0.16801, + "grad_norm": 1.1062348301285594, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 16801 + }, + { + "epoch": 0.16802, + "grad_norm": 1.1914397748599566, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 16802 + }, + { + "epoch": 0.16803, + "grad_norm": 0.8802041767245744, + "learning_rate": 0.003, + "loss": 4.1019, + "step": 16803 + }, + { + "epoch": 0.16804, + "grad_norm": 0.874564795359864, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 16804 + }, + { + "epoch": 0.16805, + "grad_norm": 0.918094497372565, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 16805 + }, + { + "epoch": 0.16806, + "grad_norm": 1.0528466282813127, + "learning_rate": 0.003, + "loss": 4.0887, + "step": 16806 + }, + { + "epoch": 0.16807, + "grad_norm": 1.0602980201384666, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 16807 + }, + { + "epoch": 0.16808, + "grad_norm": 1.057008248252162, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 16808 + }, + { + "epoch": 0.16809, + "grad_norm": 0.9772067180797183, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 16809 + }, + { + "epoch": 0.1681, + "grad_norm": 0.9766607888163783, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 16810 + }, + { + "epoch": 0.16811, + "grad_norm": 1.0608021545838455, + "learning_rate": 0.003, + "loss": 4.0984, + "step": 16811 + }, + { + "epoch": 0.16812, + "grad_norm": 0.9825708099339856, + "learning_rate": 0.003, + "loss": 4.049, + "step": 16812 + }, + { + "epoch": 0.16813, + "grad_norm": 0.8954139831884007, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 16813 + }, + { + "epoch": 0.16814, + "grad_norm": 1.0624411148794228, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 16814 + }, + { + "epoch": 0.16815, + "grad_norm": 1.0623151411256138, + "learning_rate": 0.003, + "loss": 4.1123, + "step": 16815 + }, + { + "epoch": 0.16816, + "grad_norm": 1.154663677425857, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 16816 + }, + { + "epoch": 0.16817, + "grad_norm": 0.8078243070855937, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 16817 + }, + { + "epoch": 0.16818, + "grad_norm": 0.7233763840837986, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 16818 + }, + { + "epoch": 0.16819, + "grad_norm": 0.8100933417963719, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 16819 + }, + { + "epoch": 0.1682, + "grad_norm": 0.8931708994663681, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 16820 + }, + { + "epoch": 0.16821, + "grad_norm": 1.031444527643463, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 16821 + }, + { + "epoch": 0.16822, + "grad_norm": 1.069035380830388, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 16822 + }, + { + "epoch": 0.16823, + "grad_norm": 0.9807850093634775, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 16823 + }, + { + "epoch": 0.16824, + "grad_norm": 1.0268090960250698, + "learning_rate": 0.003, + "loss": 4.094, + "step": 16824 + }, + { + "epoch": 0.16825, + "grad_norm": 1.1430472672001022, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 16825 + }, + { + "epoch": 0.16826, + "grad_norm": 1.1197116020608455, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 16826 + }, + { + "epoch": 0.16827, + "grad_norm": 1.2355886666737792, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 16827 + }, + { + "epoch": 0.16828, + "grad_norm": 0.9682181051629203, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 16828 + }, + { + "epoch": 0.16829, + "grad_norm": 0.9797772146612209, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 16829 + }, + { + "epoch": 0.1683, + "grad_norm": 0.9430423767469833, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 16830 + }, + { + "epoch": 0.16831, + "grad_norm": 1.3240875842798738, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 16831 + }, + { + "epoch": 0.16832, + "grad_norm": 1.0146071006630921, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 16832 + }, + { + "epoch": 0.16833, + "grad_norm": 1.0077124748381046, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 16833 + }, + { + "epoch": 0.16834, + "grad_norm": 1.1315128959521843, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 16834 + }, + { + "epoch": 0.16835, + "grad_norm": 0.9528544560028561, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 16835 + }, + { + "epoch": 0.16836, + "grad_norm": 0.963440026502493, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 16836 + }, + { + "epoch": 0.16837, + "grad_norm": 1.05114540030502, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 16837 + }, + { + "epoch": 0.16838, + "grad_norm": 1.2658580609590082, + "learning_rate": 0.003, + "loss": 4.1015, + "step": 16838 + }, + { + "epoch": 0.16839, + "grad_norm": 0.8213634483873891, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 16839 + }, + { + "epoch": 0.1684, + "grad_norm": 0.6853845563182607, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 16840 + }, + { + "epoch": 0.16841, + "grad_norm": 0.5921861728564217, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 16841 + }, + { + "epoch": 0.16842, + "grad_norm": 0.7455191110217251, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 16842 + }, + { + "epoch": 0.16843, + "grad_norm": 0.8097125922049726, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 16843 + }, + { + "epoch": 0.16844, + "grad_norm": 0.9814349682420912, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 16844 + }, + { + "epoch": 0.16845, + "grad_norm": 1.3200588335380827, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 16845 + }, + { + "epoch": 0.16846, + "grad_norm": 0.7522648002956938, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 16846 + }, + { + "epoch": 0.16847, + "grad_norm": 0.6447994121086472, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 16847 + }, + { + "epoch": 0.16848, + "grad_norm": 0.6979487755645498, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 16848 + }, + { + "epoch": 0.16849, + "grad_norm": 0.6642464463902737, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 16849 + }, + { + "epoch": 0.1685, + "grad_norm": 0.8118266453225255, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 16850 + }, + { + "epoch": 0.16851, + "grad_norm": 1.1498104463086074, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 16851 + }, + { + "epoch": 0.16852, + "grad_norm": 1.243298381589849, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 16852 + }, + { + "epoch": 0.16853, + "grad_norm": 1.018464689367223, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 16853 + }, + { + "epoch": 0.16854, + "grad_norm": 1.1518953921390804, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 16854 + }, + { + "epoch": 0.16855, + "grad_norm": 0.9424296276892982, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 16855 + }, + { + "epoch": 0.16856, + "grad_norm": 0.9407003264277453, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 16856 + }, + { + "epoch": 0.16857, + "grad_norm": 0.9815893609942771, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 16857 + }, + { + "epoch": 0.16858, + "grad_norm": 1.0774052991271121, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 16858 + }, + { + "epoch": 0.16859, + "grad_norm": 1.1326013364628618, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 16859 + }, + { + "epoch": 0.1686, + "grad_norm": 1.03042554497815, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 16860 + }, + { + "epoch": 0.16861, + "grad_norm": 1.0512528132313843, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 16861 + }, + { + "epoch": 0.16862, + "grad_norm": 1.125161273104075, + "learning_rate": 0.003, + "loss": 4.065, + "step": 16862 + }, + { + "epoch": 0.16863, + "grad_norm": 0.9075523911573483, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 16863 + }, + { + "epoch": 0.16864, + "grad_norm": 1.1172815804819083, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 16864 + }, + { + "epoch": 0.16865, + "grad_norm": 0.9056209845908195, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 16865 + }, + { + "epoch": 0.16866, + "grad_norm": 1.1146038124408708, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 16866 + }, + { + "epoch": 0.16867, + "grad_norm": 1.3342383531049933, + "learning_rate": 0.003, + "loss": 4.06, + "step": 16867 + }, + { + "epoch": 0.16868, + "grad_norm": 0.8735724321676355, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 16868 + }, + { + "epoch": 0.16869, + "grad_norm": 1.0027333509387237, + "learning_rate": 0.003, + "loss": 4.061, + "step": 16869 + }, + { + "epoch": 0.1687, + "grad_norm": 1.1676197372151962, + "learning_rate": 0.003, + "loss": 4.0983, + "step": 16870 + }, + { + "epoch": 0.16871, + "grad_norm": 0.9480271036600267, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 16871 + }, + { + "epoch": 0.16872, + "grad_norm": 0.9362740826738024, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 16872 + }, + { + "epoch": 0.16873, + "grad_norm": 0.9739090667553453, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 16873 + }, + { + "epoch": 0.16874, + "grad_norm": 1.1570855541938996, + "learning_rate": 0.003, + "loss": 4.052, + "step": 16874 + }, + { + "epoch": 0.16875, + "grad_norm": 0.9151770197109351, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 16875 + }, + { + "epoch": 0.16876, + "grad_norm": 0.7771208025424089, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 16876 + }, + { + "epoch": 0.16877, + "grad_norm": 0.9168950149356871, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 16877 + }, + { + "epoch": 0.16878, + "grad_norm": 1.0179378827680612, + "learning_rate": 0.003, + "loss": 4.045, + "step": 16878 + }, + { + "epoch": 0.16879, + "grad_norm": 1.2073217015272835, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 16879 + }, + { + "epoch": 0.1688, + "grad_norm": 1.0748810221070781, + "learning_rate": 0.003, + "loss": 4.0905, + "step": 16880 + }, + { + "epoch": 0.16881, + "grad_norm": 1.1112265333256262, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 16881 + }, + { + "epoch": 0.16882, + "grad_norm": 0.9278885651225035, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 16882 + }, + { + "epoch": 0.16883, + "grad_norm": 0.9831887636699694, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 16883 + }, + { + "epoch": 0.16884, + "grad_norm": 1.2083800555172433, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 16884 + }, + { + "epoch": 0.16885, + "grad_norm": 0.9308206481374236, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 16885 + }, + { + "epoch": 0.16886, + "grad_norm": 0.8274211760417409, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 16886 + }, + { + "epoch": 0.16887, + "grad_norm": 1.0605769599388164, + "learning_rate": 0.003, + "loss": 4.0976, + "step": 16887 + }, + { + "epoch": 0.16888, + "grad_norm": 1.0291847513893337, + "learning_rate": 0.003, + "loss": 4.1024, + "step": 16888 + }, + { + "epoch": 0.16889, + "grad_norm": 1.1461132999378327, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 16889 + }, + { + "epoch": 0.1689, + "grad_norm": 0.9724390806059294, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 16890 + }, + { + "epoch": 0.16891, + "grad_norm": 1.243246848885493, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 16891 + }, + { + "epoch": 0.16892, + "grad_norm": 1.0308128402891974, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 16892 + }, + { + "epoch": 0.16893, + "grad_norm": 0.934837635519902, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 16893 + }, + { + "epoch": 0.16894, + "grad_norm": 1.1215956429440812, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 16894 + }, + { + "epoch": 0.16895, + "grad_norm": 1.1930054033599886, + "learning_rate": 0.003, + "loss": 4.0981, + "step": 16895 + }, + { + "epoch": 0.16896, + "grad_norm": 0.9838770753755465, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 16896 + }, + { + "epoch": 0.16897, + "grad_norm": 1.0574447139639978, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 16897 + }, + { + "epoch": 0.16898, + "grad_norm": 1.012386603974613, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 16898 + }, + { + "epoch": 0.16899, + "grad_norm": 1.0332232106079378, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 16899 + }, + { + "epoch": 0.169, + "grad_norm": 0.9601577892440551, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 16900 + }, + { + "epoch": 0.16901, + "grad_norm": 0.9338458032054076, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 16901 + }, + { + "epoch": 0.16902, + "grad_norm": 0.9422310577161109, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 16902 + }, + { + "epoch": 0.16903, + "grad_norm": 1.0085409540888532, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 16903 + }, + { + "epoch": 0.16904, + "grad_norm": 1.0159605024665364, + "learning_rate": 0.003, + "loss": 4.1129, + "step": 16904 + }, + { + "epoch": 0.16905, + "grad_norm": 1.3120812588859267, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 16905 + }, + { + "epoch": 0.16906, + "grad_norm": 0.890649735047688, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 16906 + }, + { + "epoch": 0.16907, + "grad_norm": 0.9574361303921428, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 16907 + }, + { + "epoch": 0.16908, + "grad_norm": 1.2795556089943634, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 16908 + }, + { + "epoch": 0.16909, + "grad_norm": 0.8565623291772639, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 16909 + }, + { + "epoch": 0.1691, + "grad_norm": 0.8352715829505836, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 16910 + }, + { + "epoch": 0.16911, + "grad_norm": 0.9651337394652272, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 16911 + }, + { + "epoch": 0.16912, + "grad_norm": 1.2614980385243706, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 16912 + }, + { + "epoch": 0.16913, + "grad_norm": 0.9949959133216311, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 16913 + }, + { + "epoch": 0.16914, + "grad_norm": 0.9310150290565476, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 16914 + }, + { + "epoch": 0.16915, + "grad_norm": 0.9095281574337111, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 16915 + }, + { + "epoch": 0.16916, + "grad_norm": 0.8698741782537451, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 16916 + }, + { + "epoch": 0.16917, + "grad_norm": 0.9510889129202285, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 16917 + }, + { + "epoch": 0.16918, + "grad_norm": 1.1102731139972755, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 16918 + }, + { + "epoch": 0.16919, + "grad_norm": 0.8312112934334217, + "learning_rate": 0.003, + "loss": 4.062, + "step": 16919 + }, + { + "epoch": 0.1692, + "grad_norm": 0.7815106169863348, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 16920 + }, + { + "epoch": 0.16921, + "grad_norm": 0.8480474286605658, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 16921 + }, + { + "epoch": 0.16922, + "grad_norm": 0.8539207836284356, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 16922 + }, + { + "epoch": 0.16923, + "grad_norm": 0.9105070877577716, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 16923 + }, + { + "epoch": 0.16924, + "grad_norm": 0.9691451177072139, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 16924 + }, + { + "epoch": 0.16925, + "grad_norm": 0.9565471710699693, + "learning_rate": 0.003, + "loss": 4.078, + "step": 16925 + }, + { + "epoch": 0.16926, + "grad_norm": 1.1063788406388348, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 16926 + }, + { + "epoch": 0.16927, + "grad_norm": 1.1886031611964314, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 16927 + }, + { + "epoch": 0.16928, + "grad_norm": 0.9846968911995546, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 16928 + }, + { + "epoch": 0.16929, + "grad_norm": 0.9272376946442058, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 16929 + }, + { + "epoch": 0.1693, + "grad_norm": 0.8893421757769405, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 16930 + }, + { + "epoch": 0.16931, + "grad_norm": 1.0130969141140551, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 16931 + }, + { + "epoch": 0.16932, + "grad_norm": 1.1381946504366818, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 16932 + }, + { + "epoch": 0.16933, + "grad_norm": 1.0849578288176962, + "learning_rate": 0.003, + "loss": 4.0854, + "step": 16933 + }, + { + "epoch": 0.16934, + "grad_norm": 1.2310064006647092, + "learning_rate": 0.003, + "loss": 4.0937, + "step": 16934 + }, + { + "epoch": 0.16935, + "grad_norm": 0.8972966413417564, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 16935 + }, + { + "epoch": 0.16936, + "grad_norm": 1.0788428074545415, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 16936 + }, + { + "epoch": 0.16937, + "grad_norm": 1.0662757150832387, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 16937 + }, + { + "epoch": 0.16938, + "grad_norm": 0.9493571072891561, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 16938 + }, + { + "epoch": 0.16939, + "grad_norm": 1.109168796021773, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 16939 + }, + { + "epoch": 0.1694, + "grad_norm": 1.1414088748307305, + "learning_rate": 0.003, + "loss": 4.0887, + "step": 16940 + }, + { + "epoch": 0.16941, + "grad_norm": 0.957389898631992, + "learning_rate": 0.003, + "loss": 4.0857, + "step": 16941 + }, + { + "epoch": 0.16942, + "grad_norm": 1.1853076409897931, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 16942 + }, + { + "epoch": 0.16943, + "grad_norm": 0.9738324831534609, + "learning_rate": 0.003, + "loss": 4.034, + "step": 16943 + }, + { + "epoch": 0.16944, + "grad_norm": 0.9709755381109622, + "learning_rate": 0.003, + "loss": 4.075, + "step": 16944 + }, + { + "epoch": 0.16945, + "grad_norm": 1.003769936832352, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 16945 + }, + { + "epoch": 0.16946, + "grad_norm": 1.2616171831505183, + "learning_rate": 0.003, + "loss": 4.0848, + "step": 16946 + }, + { + "epoch": 0.16947, + "grad_norm": 0.906927327286301, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 16947 + }, + { + "epoch": 0.16948, + "grad_norm": 1.0460360399666688, + "learning_rate": 0.003, + "loss": 4.08, + "step": 16948 + }, + { + "epoch": 0.16949, + "grad_norm": 1.0024263341234232, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 16949 + }, + { + "epoch": 0.1695, + "grad_norm": 1.007927260397112, + "learning_rate": 0.003, + "loss": 4.081, + "step": 16950 + }, + { + "epoch": 0.16951, + "grad_norm": 1.3323989907842386, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 16951 + }, + { + "epoch": 0.16952, + "grad_norm": 0.8727576775762045, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 16952 + }, + { + "epoch": 0.16953, + "grad_norm": 0.7894830638469373, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 16953 + }, + { + "epoch": 0.16954, + "grad_norm": 0.898842046151003, + "learning_rate": 0.003, + "loss": 4.1079, + "step": 16954 + }, + { + "epoch": 0.16955, + "grad_norm": 1.3510368868210914, + "learning_rate": 0.003, + "loss": 4.0929, + "step": 16955 + }, + { + "epoch": 0.16956, + "grad_norm": 0.9444552361958362, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 16956 + }, + { + "epoch": 0.16957, + "grad_norm": 1.0241536529459698, + "learning_rate": 0.003, + "loss": 4.0845, + "step": 16957 + }, + { + "epoch": 0.16958, + "grad_norm": 1.2070289480265366, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 16958 + }, + { + "epoch": 0.16959, + "grad_norm": 0.8195496990845997, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 16959 + }, + { + "epoch": 0.1696, + "grad_norm": 0.8805083884063785, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 16960 + }, + { + "epoch": 0.16961, + "grad_norm": 1.0497915812438796, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 16961 + }, + { + "epoch": 0.16962, + "grad_norm": 1.1381966861833035, + "learning_rate": 0.003, + "loss": 4.1052, + "step": 16962 + }, + { + "epoch": 0.16963, + "grad_norm": 0.9239114534694449, + "learning_rate": 0.003, + "loss": 4.09, + "step": 16963 + }, + { + "epoch": 0.16964, + "grad_norm": 1.1120466914741012, + "learning_rate": 0.003, + "loss": 4.0944, + "step": 16964 + }, + { + "epoch": 0.16965, + "grad_norm": 1.0536255156086367, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 16965 + }, + { + "epoch": 0.16966, + "grad_norm": 1.003407225574236, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 16966 + }, + { + "epoch": 0.16967, + "grad_norm": 1.0637741298474856, + "learning_rate": 0.003, + "loss": 4.068, + "step": 16967 + }, + { + "epoch": 0.16968, + "grad_norm": 1.047063784428621, + "learning_rate": 0.003, + "loss": 4.085, + "step": 16968 + }, + { + "epoch": 0.16969, + "grad_norm": 1.0802858469443242, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 16969 + }, + { + "epoch": 0.1697, + "grad_norm": 1.0497581667544262, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 16970 + }, + { + "epoch": 0.16971, + "grad_norm": 0.9414606148890946, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 16971 + }, + { + "epoch": 0.16972, + "grad_norm": 1.1170299432222692, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 16972 + }, + { + "epoch": 0.16973, + "grad_norm": 1.0766082497347986, + "learning_rate": 0.003, + "loss": 4.0864, + "step": 16973 + }, + { + "epoch": 0.16974, + "grad_norm": 0.971177833154046, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 16974 + }, + { + "epoch": 0.16975, + "grad_norm": 1.0091262759483555, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 16975 + }, + { + "epoch": 0.16976, + "grad_norm": 0.8777320338942457, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 16976 + }, + { + "epoch": 0.16977, + "grad_norm": 0.9230449519623944, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 16977 + }, + { + "epoch": 0.16978, + "grad_norm": 1.1365094966425633, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 16978 + }, + { + "epoch": 0.16979, + "grad_norm": 0.8920837690830092, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 16979 + }, + { + "epoch": 0.1698, + "grad_norm": 0.8482112156533284, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 16980 + }, + { + "epoch": 0.16981, + "grad_norm": 0.8988239598906194, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 16981 + }, + { + "epoch": 0.16982, + "grad_norm": 0.9182813792006643, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 16982 + }, + { + "epoch": 0.16983, + "grad_norm": 1.0526003288823398, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 16983 + }, + { + "epoch": 0.16984, + "grad_norm": 1.14484332899849, + "learning_rate": 0.003, + "loss": 4.0934, + "step": 16984 + }, + { + "epoch": 0.16985, + "grad_norm": 0.8572863326627641, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 16985 + }, + { + "epoch": 0.16986, + "grad_norm": 0.8165088118914843, + "learning_rate": 0.003, + "loss": 4.062, + "step": 16986 + }, + { + "epoch": 0.16987, + "grad_norm": 0.8647532475948965, + "learning_rate": 0.003, + "loss": 4.0985, + "step": 16987 + }, + { + "epoch": 0.16988, + "grad_norm": 0.9754417129048155, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 16988 + }, + { + "epoch": 0.16989, + "grad_norm": 1.317056029191013, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 16989 + }, + { + "epoch": 0.1699, + "grad_norm": 0.7884656523077743, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 16990 + }, + { + "epoch": 0.16991, + "grad_norm": 0.6747074639773813, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 16991 + }, + { + "epoch": 0.16992, + "grad_norm": 0.6947827620379723, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 16992 + }, + { + "epoch": 0.16993, + "grad_norm": 0.736273038791794, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 16993 + }, + { + "epoch": 0.16994, + "grad_norm": 0.9212350951810363, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 16994 + }, + { + "epoch": 0.16995, + "grad_norm": 1.0320385225499165, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 16995 + }, + { + "epoch": 0.16996, + "grad_norm": 1.1230714065219136, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 16996 + }, + { + "epoch": 0.16997, + "grad_norm": 1.0261445604311525, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 16997 + }, + { + "epoch": 0.16998, + "grad_norm": 1.055679231910794, + "learning_rate": 0.003, + "loss": 4.0854, + "step": 16998 + }, + { + "epoch": 0.16999, + "grad_norm": 1.0998618231294142, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 16999 + }, + { + "epoch": 0.17, + "grad_norm": 1.1655329952227764, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 17000 + }, + { + "epoch": 0.17001, + "grad_norm": 0.9612932226921341, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 17001 + }, + { + "epoch": 0.17002, + "grad_norm": 0.9000920315799099, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 17002 + }, + { + "epoch": 0.17003, + "grad_norm": 0.9213594390656381, + "learning_rate": 0.003, + "loss": 4.034, + "step": 17003 + }, + { + "epoch": 0.17004, + "grad_norm": 1.1295358274167493, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 17004 + }, + { + "epoch": 0.17005, + "grad_norm": 1.1653813133738868, + "learning_rate": 0.003, + "loss": 4.059, + "step": 17005 + }, + { + "epoch": 0.17006, + "grad_norm": 0.9310859969740729, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 17006 + }, + { + "epoch": 0.17007, + "grad_norm": 1.0655412204129018, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 17007 + }, + { + "epoch": 0.17008, + "grad_norm": 1.143489797703558, + "learning_rate": 0.003, + "loss": 4.0978, + "step": 17008 + }, + { + "epoch": 0.17009, + "grad_norm": 1.0266644230944915, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 17009 + }, + { + "epoch": 0.1701, + "grad_norm": 0.9816851492671904, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 17010 + }, + { + "epoch": 0.17011, + "grad_norm": 1.1085875717939988, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 17011 + }, + { + "epoch": 0.17012, + "grad_norm": 1.144955433284326, + "learning_rate": 0.003, + "loss": 4.0913, + "step": 17012 + }, + { + "epoch": 0.17013, + "grad_norm": 0.8687979014130185, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 17013 + }, + { + "epoch": 0.17014, + "grad_norm": 1.0027879547968, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 17014 + }, + { + "epoch": 0.17015, + "grad_norm": 1.3632475777054143, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 17015 + }, + { + "epoch": 0.17016, + "grad_norm": 0.9645686330297072, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 17016 + }, + { + "epoch": 0.17017, + "grad_norm": 1.156357762842491, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 17017 + }, + { + "epoch": 0.17018, + "grad_norm": 1.0662467673643754, + "learning_rate": 0.003, + "loss": 4.0898, + "step": 17018 + }, + { + "epoch": 0.17019, + "grad_norm": 1.0738803661093201, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 17019 + }, + { + "epoch": 0.1702, + "grad_norm": 0.9633129446683043, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 17020 + }, + { + "epoch": 0.17021, + "grad_norm": 1.0417070117637972, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 17021 + }, + { + "epoch": 0.17022, + "grad_norm": 1.1311883958998308, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 17022 + }, + { + "epoch": 0.17023, + "grad_norm": 0.7550485977785689, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 17023 + }, + { + "epoch": 0.17024, + "grad_norm": 0.6968840254996508, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 17024 + }, + { + "epoch": 0.17025, + "grad_norm": 0.7592513524058513, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 17025 + }, + { + "epoch": 0.17026, + "grad_norm": 0.8599585540448047, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 17026 + }, + { + "epoch": 0.17027, + "grad_norm": 1.0458860824335472, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 17027 + }, + { + "epoch": 0.17028, + "grad_norm": 1.1742993613564443, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 17028 + }, + { + "epoch": 0.17029, + "grad_norm": 0.8767028260817562, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 17029 + }, + { + "epoch": 0.1703, + "grad_norm": 0.9121111457394047, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 17030 + }, + { + "epoch": 0.17031, + "grad_norm": 0.9909112410974145, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 17031 + }, + { + "epoch": 0.17032, + "grad_norm": 1.131843799853556, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 17032 + }, + { + "epoch": 0.17033, + "grad_norm": 1.048593612163451, + "learning_rate": 0.003, + "loss": 4.063, + "step": 17033 + }, + { + "epoch": 0.17034, + "grad_norm": 1.1735984831908892, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 17034 + }, + { + "epoch": 0.17035, + "grad_norm": 0.9334068109717782, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 17035 + }, + { + "epoch": 0.17036, + "grad_norm": 0.9911031670888386, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 17036 + }, + { + "epoch": 0.17037, + "grad_norm": 1.2733875493741849, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 17037 + }, + { + "epoch": 0.17038, + "grad_norm": 0.9589845212250199, + "learning_rate": 0.003, + "loss": 4.062, + "step": 17038 + }, + { + "epoch": 0.17039, + "grad_norm": 0.9200520125757194, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 17039 + }, + { + "epoch": 0.1704, + "grad_norm": 0.9963600532069022, + "learning_rate": 0.003, + "loss": 4.0932, + "step": 17040 + }, + { + "epoch": 0.17041, + "grad_norm": 1.2525498571450993, + "learning_rate": 0.003, + "loss": 4.079, + "step": 17041 + }, + { + "epoch": 0.17042, + "grad_norm": 0.8152043363208599, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 17042 + }, + { + "epoch": 0.17043, + "grad_norm": 0.8088870195973014, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 17043 + }, + { + "epoch": 0.17044, + "grad_norm": 0.8767006075376047, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 17044 + }, + { + "epoch": 0.17045, + "grad_norm": 0.970081396543649, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 17045 + }, + { + "epoch": 0.17046, + "grad_norm": 1.00536731207844, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 17046 + }, + { + "epoch": 0.17047, + "grad_norm": 1.0982982685925178, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 17047 + }, + { + "epoch": 0.17048, + "grad_norm": 1.0540882363934976, + "learning_rate": 0.003, + "loss": 4.0986, + "step": 17048 + }, + { + "epoch": 0.17049, + "grad_norm": 1.0537041772969982, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 17049 + }, + { + "epoch": 0.1705, + "grad_norm": 1.2472645753388485, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 17050 + }, + { + "epoch": 0.17051, + "grad_norm": 1.0695607235644633, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 17051 + }, + { + "epoch": 0.17052, + "grad_norm": 0.8405652670477259, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 17052 + }, + { + "epoch": 0.17053, + "grad_norm": 0.8428535476221581, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 17053 + }, + { + "epoch": 0.17054, + "grad_norm": 1.0373845416530403, + "learning_rate": 0.003, + "loss": 4.0988, + "step": 17054 + }, + { + "epoch": 0.17055, + "grad_norm": 1.1104833096240985, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 17055 + }, + { + "epoch": 0.17056, + "grad_norm": 0.99130507093245, + "learning_rate": 0.003, + "loss": 4.063, + "step": 17056 + }, + { + "epoch": 0.17057, + "grad_norm": 1.063850418422872, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 17057 + }, + { + "epoch": 0.17058, + "grad_norm": 0.9925088238586706, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 17058 + }, + { + "epoch": 0.17059, + "grad_norm": 1.0097757211667546, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 17059 + }, + { + "epoch": 0.1706, + "grad_norm": 1.0302977028860663, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 17060 + }, + { + "epoch": 0.17061, + "grad_norm": 1.0893122949714833, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 17061 + }, + { + "epoch": 0.17062, + "grad_norm": 1.171433547142537, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 17062 + }, + { + "epoch": 0.17063, + "grad_norm": 0.7904143213364668, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 17063 + }, + { + "epoch": 0.17064, + "grad_norm": 0.7806878548002807, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 17064 + }, + { + "epoch": 0.17065, + "grad_norm": 0.7915276774916914, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 17065 + }, + { + "epoch": 0.17066, + "grad_norm": 0.9435753921494558, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 17066 + }, + { + "epoch": 0.17067, + "grad_norm": 1.1018680844244444, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 17067 + }, + { + "epoch": 0.17068, + "grad_norm": 1.2294174798441622, + "learning_rate": 0.003, + "loss": 4.0845, + "step": 17068 + }, + { + "epoch": 0.17069, + "grad_norm": 0.9930925493624421, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 17069 + }, + { + "epoch": 0.1707, + "grad_norm": 1.0044155964179775, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 17070 + }, + { + "epoch": 0.17071, + "grad_norm": 0.9829027933473973, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 17071 + }, + { + "epoch": 0.17072, + "grad_norm": 1.054358354271117, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 17072 + }, + { + "epoch": 0.17073, + "grad_norm": 1.1475499434317498, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 17073 + }, + { + "epoch": 0.17074, + "grad_norm": 0.9755538141504658, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 17074 + }, + { + "epoch": 0.17075, + "grad_norm": 1.2236217984995386, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 17075 + }, + { + "epoch": 0.17076, + "grad_norm": 1.0754987956046365, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 17076 + }, + { + "epoch": 0.17077, + "grad_norm": 1.1441829553756893, + "learning_rate": 0.003, + "loss": 4.0917, + "step": 17077 + }, + { + "epoch": 0.17078, + "grad_norm": 0.8636942430701885, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 17078 + }, + { + "epoch": 0.17079, + "grad_norm": 0.8530656416935916, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 17079 + }, + { + "epoch": 0.1708, + "grad_norm": 0.9933007045148369, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 17080 + }, + { + "epoch": 0.17081, + "grad_norm": 1.2820050535650034, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 17081 + }, + { + "epoch": 0.17082, + "grad_norm": 1.082513265289859, + "learning_rate": 0.003, + "loss": 4.1081, + "step": 17082 + }, + { + "epoch": 0.17083, + "grad_norm": 1.1549006824944454, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 17083 + }, + { + "epoch": 0.17084, + "grad_norm": 1.0636981211406051, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 17084 + }, + { + "epoch": 0.17085, + "grad_norm": 1.1208211284580962, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 17085 + }, + { + "epoch": 0.17086, + "grad_norm": 0.9484473452975615, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 17086 + }, + { + "epoch": 0.17087, + "grad_norm": 0.9678302467008759, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 17087 + }, + { + "epoch": 0.17088, + "grad_norm": 1.2879782877037143, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 17088 + }, + { + "epoch": 0.17089, + "grad_norm": 1.0466507691786753, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 17089 + }, + { + "epoch": 0.1709, + "grad_norm": 1.0883782357039746, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 17090 + }, + { + "epoch": 0.17091, + "grad_norm": 0.9624041692935246, + "learning_rate": 0.003, + "loss": 4.0992, + "step": 17091 + }, + { + "epoch": 0.17092, + "grad_norm": 1.0476049915627403, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 17092 + }, + { + "epoch": 0.17093, + "grad_norm": 1.023950684606274, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 17093 + }, + { + "epoch": 0.17094, + "grad_norm": 1.134951753033142, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 17094 + }, + { + "epoch": 0.17095, + "grad_norm": 1.1788514295899923, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 17095 + }, + { + "epoch": 0.17096, + "grad_norm": 1.114827044919604, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 17096 + }, + { + "epoch": 0.17097, + "grad_norm": 0.9379758935891594, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 17097 + }, + { + "epoch": 0.17098, + "grad_norm": 0.9552039584756868, + "learning_rate": 0.003, + "loss": 4.0918, + "step": 17098 + }, + { + "epoch": 0.17099, + "grad_norm": 0.9888508295556777, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 17099 + }, + { + "epoch": 0.171, + "grad_norm": 0.9315574202273936, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 17100 + }, + { + "epoch": 0.17101, + "grad_norm": 0.8840793492167259, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 17101 + }, + { + "epoch": 0.17102, + "grad_norm": 0.8230207893812931, + "learning_rate": 0.003, + "loss": 4.053, + "step": 17102 + }, + { + "epoch": 0.17103, + "grad_norm": 0.804542654459796, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 17103 + }, + { + "epoch": 0.17104, + "grad_norm": 0.8844166999780292, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 17104 + }, + { + "epoch": 0.17105, + "grad_norm": 1.025887523142765, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 17105 + }, + { + "epoch": 0.17106, + "grad_norm": 0.9580105340405675, + "learning_rate": 0.003, + "loss": 4.066, + "step": 17106 + }, + { + "epoch": 0.17107, + "grad_norm": 0.9708590654121104, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 17107 + }, + { + "epoch": 0.17108, + "grad_norm": 1.1692546265867165, + "learning_rate": 0.003, + "loss": 4.069, + "step": 17108 + }, + { + "epoch": 0.17109, + "grad_norm": 1.0046931690521228, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 17109 + }, + { + "epoch": 0.1711, + "grad_norm": 0.9739238046670415, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 17110 + }, + { + "epoch": 0.17111, + "grad_norm": 1.0675322139557897, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 17111 + }, + { + "epoch": 0.17112, + "grad_norm": 1.1311867240664282, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 17112 + }, + { + "epoch": 0.17113, + "grad_norm": 0.8612563249538273, + "learning_rate": 0.003, + "loss": 4.046, + "step": 17113 + }, + { + "epoch": 0.17114, + "grad_norm": 0.7474282426523464, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 17114 + }, + { + "epoch": 0.17115, + "grad_norm": 0.7444236541588136, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 17115 + }, + { + "epoch": 0.17116, + "grad_norm": 0.7675631254727495, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 17116 + }, + { + "epoch": 0.17117, + "grad_norm": 0.9341458366495701, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 17117 + }, + { + "epoch": 0.17118, + "grad_norm": 1.342731279039567, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 17118 + }, + { + "epoch": 0.17119, + "grad_norm": 0.8888490183821955, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 17119 + }, + { + "epoch": 0.1712, + "grad_norm": 0.8785980633015877, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 17120 + }, + { + "epoch": 0.17121, + "grad_norm": 0.8271204744742502, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 17121 + }, + { + "epoch": 0.17122, + "grad_norm": 1.0893178857713155, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 17122 + }, + { + "epoch": 0.17123, + "grad_norm": 1.0790245112228576, + "learning_rate": 0.003, + "loss": 4.058, + "step": 17123 + }, + { + "epoch": 0.17124, + "grad_norm": 1.0390544417732577, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 17124 + }, + { + "epoch": 0.17125, + "grad_norm": 1.0680299787118601, + "learning_rate": 0.003, + "loss": 4.0989, + "step": 17125 + }, + { + "epoch": 0.17126, + "grad_norm": 0.8959368150225251, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 17126 + }, + { + "epoch": 0.17127, + "grad_norm": 0.9029081741407197, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 17127 + }, + { + "epoch": 0.17128, + "grad_norm": 1.1005418785087984, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 17128 + }, + { + "epoch": 0.17129, + "grad_norm": 1.227755738556662, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 17129 + }, + { + "epoch": 0.1713, + "grad_norm": 0.8930105757517267, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 17130 + }, + { + "epoch": 0.17131, + "grad_norm": 0.9462391565337224, + "learning_rate": 0.003, + "loss": 4.052, + "step": 17131 + }, + { + "epoch": 0.17132, + "grad_norm": 1.1064612845471093, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 17132 + }, + { + "epoch": 0.17133, + "grad_norm": 1.1188013635852194, + "learning_rate": 0.003, + "loss": 4.0994, + "step": 17133 + }, + { + "epoch": 0.17134, + "grad_norm": 0.9363002842585174, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 17134 + }, + { + "epoch": 0.17135, + "grad_norm": 0.9489678341562948, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 17135 + }, + { + "epoch": 0.17136, + "grad_norm": 1.050221401326341, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 17136 + }, + { + "epoch": 0.17137, + "grad_norm": 1.0622341104039013, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 17137 + }, + { + "epoch": 0.17138, + "grad_norm": 1.0155810568070514, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 17138 + }, + { + "epoch": 0.17139, + "grad_norm": 1.2292390804474131, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 17139 + }, + { + "epoch": 0.1714, + "grad_norm": 0.863961097048021, + "learning_rate": 0.003, + "loss": 4.0941, + "step": 17140 + }, + { + "epoch": 0.17141, + "grad_norm": 0.9722801840188489, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 17141 + }, + { + "epoch": 0.17142, + "grad_norm": 1.0853630378803718, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 17142 + }, + { + "epoch": 0.17143, + "grad_norm": 1.122704506930695, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 17143 + }, + { + "epoch": 0.17144, + "grad_norm": 1.0481516128000807, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 17144 + }, + { + "epoch": 0.17145, + "grad_norm": 1.0045136623605473, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 17145 + }, + { + "epoch": 0.17146, + "grad_norm": 1.2933677053129453, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 17146 + }, + { + "epoch": 0.17147, + "grad_norm": 0.9268112487343603, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 17147 + }, + { + "epoch": 0.17148, + "grad_norm": 0.9063201930351329, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 17148 + }, + { + "epoch": 0.17149, + "grad_norm": 0.9831481336546762, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 17149 + }, + { + "epoch": 0.1715, + "grad_norm": 0.9257564212499901, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 17150 + }, + { + "epoch": 0.17151, + "grad_norm": 1.0402013131540702, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 17151 + }, + { + "epoch": 0.17152, + "grad_norm": 1.2696311564090137, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 17152 + }, + { + "epoch": 0.17153, + "grad_norm": 0.9228383895476694, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 17153 + }, + { + "epoch": 0.17154, + "grad_norm": 0.8756047249868955, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 17154 + }, + { + "epoch": 0.17155, + "grad_norm": 0.9255483707631438, + "learning_rate": 0.003, + "loss": 4.06, + "step": 17155 + }, + { + "epoch": 0.17156, + "grad_norm": 1.0128656678138324, + "learning_rate": 0.003, + "loss": 4.0935, + "step": 17156 + }, + { + "epoch": 0.17157, + "grad_norm": 1.1153609586444866, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 17157 + }, + { + "epoch": 0.17158, + "grad_norm": 1.0274785717558435, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 17158 + }, + { + "epoch": 0.17159, + "grad_norm": 1.139271791196328, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 17159 + }, + { + "epoch": 0.1716, + "grad_norm": 0.7723158269993754, + "learning_rate": 0.003, + "loss": 4.055, + "step": 17160 + }, + { + "epoch": 0.17161, + "grad_norm": 0.752605545040367, + "learning_rate": 0.003, + "loss": 4.0915, + "step": 17161 + }, + { + "epoch": 0.17162, + "grad_norm": 0.8166100343125613, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 17162 + }, + { + "epoch": 0.17163, + "grad_norm": 0.8936798807231556, + "learning_rate": 0.003, + "loss": 4.041, + "step": 17163 + }, + { + "epoch": 0.17164, + "grad_norm": 0.9154145349642477, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 17164 + }, + { + "epoch": 0.17165, + "grad_norm": 1.1327583741213283, + "learning_rate": 0.003, + "loss": 4.074, + "step": 17165 + }, + { + "epoch": 0.17166, + "grad_norm": 0.9331357208271185, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 17166 + }, + { + "epoch": 0.17167, + "grad_norm": 0.9467964356194164, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 17167 + }, + { + "epoch": 0.17168, + "grad_norm": 0.9382489475647661, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 17168 + }, + { + "epoch": 0.17169, + "grad_norm": 0.9870543068149238, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 17169 + }, + { + "epoch": 0.1717, + "grad_norm": 0.987522265166596, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 17170 + }, + { + "epoch": 0.17171, + "grad_norm": 0.9962977373903137, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 17171 + }, + { + "epoch": 0.17172, + "grad_norm": 1.29602221576462, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 17172 + }, + { + "epoch": 0.17173, + "grad_norm": 1.0900028502315877, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 17173 + }, + { + "epoch": 0.17174, + "grad_norm": 1.0442412551577036, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 17174 + }, + { + "epoch": 0.17175, + "grad_norm": 1.1023911290190684, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 17175 + }, + { + "epoch": 0.17176, + "grad_norm": 0.8737615155250994, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 17176 + }, + { + "epoch": 0.17177, + "grad_norm": 0.9565378019645701, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 17177 + }, + { + "epoch": 0.17178, + "grad_norm": 1.3564786735656318, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 17178 + }, + { + "epoch": 0.17179, + "grad_norm": 0.8800004739148151, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 17179 + }, + { + "epoch": 0.1718, + "grad_norm": 0.9157066814198301, + "learning_rate": 0.003, + "loss": 4.082, + "step": 17180 + }, + { + "epoch": 0.17181, + "grad_norm": 1.0554536999625397, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 17181 + }, + { + "epoch": 0.17182, + "grad_norm": 1.0218463723248616, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 17182 + }, + { + "epoch": 0.17183, + "grad_norm": 1.053457533387476, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 17183 + }, + { + "epoch": 0.17184, + "grad_norm": 1.1067109261185086, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 17184 + }, + { + "epoch": 0.17185, + "grad_norm": 1.3243671365734395, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 17185 + }, + { + "epoch": 0.17186, + "grad_norm": 0.8621315136750969, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 17186 + }, + { + "epoch": 0.17187, + "grad_norm": 0.7760921296669563, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 17187 + }, + { + "epoch": 0.17188, + "grad_norm": 0.8458072075768651, + "learning_rate": 0.003, + "loss": 4.0916, + "step": 17188 + }, + { + "epoch": 0.17189, + "grad_norm": 0.8216415966073443, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 17189 + }, + { + "epoch": 0.1719, + "grad_norm": 0.9698408181757641, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 17190 + }, + { + "epoch": 0.17191, + "grad_norm": 1.2007599445961525, + "learning_rate": 0.003, + "loss": 4.0994, + "step": 17191 + }, + { + "epoch": 0.17192, + "grad_norm": 0.8777559219440424, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 17192 + }, + { + "epoch": 0.17193, + "grad_norm": 0.8967970380775911, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 17193 + }, + { + "epoch": 0.17194, + "grad_norm": 0.9504608414112655, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 17194 + }, + { + "epoch": 0.17195, + "grad_norm": 0.9524172231807588, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 17195 + }, + { + "epoch": 0.17196, + "grad_norm": 1.0039289611732396, + "learning_rate": 0.003, + "loss": 4.048, + "step": 17196 + }, + { + "epoch": 0.17197, + "grad_norm": 1.0193458507818818, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 17197 + }, + { + "epoch": 0.17198, + "grad_norm": 1.0259098854133823, + "learning_rate": 0.003, + "loss": 4.1151, + "step": 17198 + }, + { + "epoch": 0.17199, + "grad_norm": 0.9222040554487325, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 17199 + }, + { + "epoch": 0.172, + "grad_norm": 0.8645827424364376, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 17200 + }, + { + "epoch": 0.17201, + "grad_norm": 0.9602326342002696, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 17201 + }, + { + "epoch": 0.17202, + "grad_norm": 1.1041672761890415, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 17202 + }, + { + "epoch": 0.17203, + "grad_norm": 0.9420710769467078, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 17203 + }, + { + "epoch": 0.17204, + "grad_norm": 1.2443445496953833, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 17204 + }, + { + "epoch": 0.17205, + "grad_norm": 0.8880420313424482, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 17205 + }, + { + "epoch": 0.17206, + "grad_norm": 1.0878149906816859, + "learning_rate": 0.003, + "loss": 4.0891, + "step": 17206 + }, + { + "epoch": 0.17207, + "grad_norm": 1.1931812422390757, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 17207 + }, + { + "epoch": 0.17208, + "grad_norm": 0.9811414998966448, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 17208 + }, + { + "epoch": 0.17209, + "grad_norm": 1.1211547932389252, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 17209 + }, + { + "epoch": 0.1721, + "grad_norm": 1.156367472851477, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 17210 + }, + { + "epoch": 0.17211, + "grad_norm": 1.0394105799115956, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 17211 + }, + { + "epoch": 0.17212, + "grad_norm": 0.8491206488423867, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 17212 + }, + { + "epoch": 0.17213, + "grad_norm": 0.8205635859303586, + "learning_rate": 0.003, + "loss": 4.0871, + "step": 17213 + }, + { + "epoch": 0.17214, + "grad_norm": 1.0271108121709813, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 17214 + }, + { + "epoch": 0.17215, + "grad_norm": 1.0228966457904398, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 17215 + }, + { + "epoch": 0.17216, + "grad_norm": 1.054942371275972, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 17216 + }, + { + "epoch": 0.17217, + "grad_norm": 1.0255006918603038, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 17217 + }, + { + "epoch": 0.17218, + "grad_norm": 1.0056981737547919, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 17218 + }, + { + "epoch": 0.17219, + "grad_norm": 1.0252612343144933, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 17219 + }, + { + "epoch": 0.1722, + "grad_norm": 1.1845876084641436, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 17220 + }, + { + "epoch": 0.17221, + "grad_norm": 0.99874627368797, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 17221 + }, + { + "epoch": 0.17222, + "grad_norm": 0.9268374934084799, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 17222 + }, + { + "epoch": 0.17223, + "grad_norm": 0.9557723450219542, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 17223 + }, + { + "epoch": 0.17224, + "grad_norm": 1.2164263302979814, + "learning_rate": 0.003, + "loss": 4.1017, + "step": 17224 + }, + { + "epoch": 0.17225, + "grad_norm": 1.031724781764682, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 17225 + }, + { + "epoch": 0.17226, + "grad_norm": 1.1099702206262598, + "learning_rate": 0.003, + "loss": 4.096, + "step": 17226 + }, + { + "epoch": 0.17227, + "grad_norm": 1.0857704716036223, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 17227 + }, + { + "epoch": 0.17228, + "grad_norm": 1.1101538131027293, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 17228 + }, + { + "epoch": 0.17229, + "grad_norm": 1.04331804263426, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 17229 + }, + { + "epoch": 0.1723, + "grad_norm": 1.0781870886476537, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 17230 + }, + { + "epoch": 0.17231, + "grad_norm": 0.915088880318123, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 17231 + }, + { + "epoch": 0.17232, + "grad_norm": 0.8396770675168571, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 17232 + }, + { + "epoch": 0.17233, + "grad_norm": 0.9152175771933209, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 17233 + }, + { + "epoch": 0.17234, + "grad_norm": 0.9547883171173379, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 17234 + }, + { + "epoch": 0.17235, + "grad_norm": 0.9206542424213469, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 17235 + }, + { + "epoch": 0.17236, + "grad_norm": 1.3514788970977905, + "learning_rate": 0.003, + "loss": 4.0994, + "step": 17236 + }, + { + "epoch": 0.17237, + "grad_norm": 0.9621115214826137, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 17237 + }, + { + "epoch": 0.17238, + "grad_norm": 0.9447327419579791, + "learning_rate": 0.003, + "loss": 4.0949, + "step": 17238 + }, + { + "epoch": 0.17239, + "grad_norm": 1.0681942110107845, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 17239 + }, + { + "epoch": 0.1724, + "grad_norm": 1.1775961977621687, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 17240 + }, + { + "epoch": 0.17241, + "grad_norm": 0.8189419147655096, + "learning_rate": 0.003, + "loss": 4.079, + "step": 17241 + }, + { + "epoch": 0.17242, + "grad_norm": 0.8680816013427626, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 17242 + }, + { + "epoch": 0.17243, + "grad_norm": 0.9347481557560829, + "learning_rate": 0.003, + "loss": 4.057, + "step": 17243 + }, + { + "epoch": 0.17244, + "grad_norm": 1.0144820025751151, + "learning_rate": 0.003, + "loss": 4.1009, + "step": 17244 + }, + { + "epoch": 0.17245, + "grad_norm": 1.1813822046549052, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 17245 + }, + { + "epoch": 0.17246, + "grad_norm": 0.9269958467884492, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 17246 + }, + { + "epoch": 0.17247, + "grad_norm": 0.8677175742211432, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 17247 + }, + { + "epoch": 0.17248, + "grad_norm": 1.0110072194972464, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 17248 + }, + { + "epoch": 0.17249, + "grad_norm": 1.1682392721811696, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 17249 + }, + { + "epoch": 0.1725, + "grad_norm": 0.9561255315330501, + "learning_rate": 0.003, + "loss": 4.0905, + "step": 17250 + }, + { + "epoch": 0.17251, + "grad_norm": 1.0049395385363562, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 17251 + }, + { + "epoch": 0.17252, + "grad_norm": 1.1590574268812828, + "learning_rate": 0.003, + "loss": 4.0927, + "step": 17252 + }, + { + "epoch": 0.17253, + "grad_norm": 1.0107464433856792, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 17253 + }, + { + "epoch": 0.17254, + "grad_norm": 1.1773826569090113, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 17254 + }, + { + "epoch": 0.17255, + "grad_norm": 0.996782028385972, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 17255 + }, + { + "epoch": 0.17256, + "grad_norm": 1.019203449366528, + "learning_rate": 0.003, + "loss": 4.066, + "step": 17256 + }, + { + "epoch": 0.17257, + "grad_norm": 1.1801273075784462, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 17257 + }, + { + "epoch": 0.17258, + "grad_norm": 0.9511632061536082, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 17258 + }, + { + "epoch": 0.17259, + "grad_norm": 1.0794317942651421, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 17259 + }, + { + "epoch": 0.1726, + "grad_norm": 0.846024067141986, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 17260 + }, + { + "epoch": 0.17261, + "grad_norm": 0.942216544306562, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 17261 + }, + { + "epoch": 0.17262, + "grad_norm": 0.9755655914549023, + "learning_rate": 0.003, + "loss": 4.06, + "step": 17262 + }, + { + "epoch": 0.17263, + "grad_norm": 1.0609950154121113, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 17263 + }, + { + "epoch": 0.17264, + "grad_norm": 1.2621699788189376, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 17264 + }, + { + "epoch": 0.17265, + "grad_norm": 0.7400189868536591, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 17265 + }, + { + "epoch": 0.17266, + "grad_norm": 0.69908886541335, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 17266 + }, + { + "epoch": 0.17267, + "grad_norm": 0.8294122767795425, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 17267 + }, + { + "epoch": 0.17268, + "grad_norm": 0.9579100752160787, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 17268 + }, + { + "epoch": 0.17269, + "grad_norm": 1.1216136705125384, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 17269 + }, + { + "epoch": 0.1727, + "grad_norm": 0.8827467549847806, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 17270 + }, + { + "epoch": 0.17271, + "grad_norm": 1.0222300722762185, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 17271 + }, + { + "epoch": 0.17272, + "grad_norm": 1.1468075724025866, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 17272 + }, + { + "epoch": 0.17273, + "grad_norm": 0.8776904864312329, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 17273 + }, + { + "epoch": 0.17274, + "grad_norm": 0.8634041563822494, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 17274 + }, + { + "epoch": 0.17275, + "grad_norm": 1.059536672562131, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 17275 + }, + { + "epoch": 0.17276, + "grad_norm": 0.9577052854596758, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 17276 + }, + { + "epoch": 0.17277, + "grad_norm": 0.9695294721035181, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 17277 + }, + { + "epoch": 0.17278, + "grad_norm": 1.009401444802055, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 17278 + }, + { + "epoch": 0.17279, + "grad_norm": 1.0895183490218532, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 17279 + }, + { + "epoch": 0.1728, + "grad_norm": 0.9524668541325774, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 17280 + }, + { + "epoch": 0.17281, + "grad_norm": 1.014085591249688, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 17281 + }, + { + "epoch": 0.17282, + "grad_norm": 1.2622348630524682, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 17282 + }, + { + "epoch": 0.17283, + "grad_norm": 1.1295284042885372, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 17283 + }, + { + "epoch": 0.17284, + "grad_norm": 0.9908986352905802, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 17284 + }, + { + "epoch": 0.17285, + "grad_norm": 1.0112449817130023, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 17285 + }, + { + "epoch": 0.17286, + "grad_norm": 1.1596312456637432, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 17286 + }, + { + "epoch": 0.17287, + "grad_norm": 1.0569091378734898, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 17287 + }, + { + "epoch": 0.17288, + "grad_norm": 1.0256299694555464, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 17288 + }, + { + "epoch": 0.17289, + "grad_norm": 0.887964143254142, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 17289 + }, + { + "epoch": 0.1729, + "grad_norm": 0.7556916799792978, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 17290 + }, + { + "epoch": 0.17291, + "grad_norm": 0.7116794482300961, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 17291 + }, + { + "epoch": 0.17292, + "grad_norm": 0.8864560980047371, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 17292 + }, + { + "epoch": 0.17293, + "grad_norm": 1.0564076798396658, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 17293 + }, + { + "epoch": 0.17294, + "grad_norm": 1.2936714873304813, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 17294 + }, + { + "epoch": 0.17295, + "grad_norm": 0.7946520559848832, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 17295 + }, + { + "epoch": 0.17296, + "grad_norm": 0.7676406910530207, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 17296 + }, + { + "epoch": 0.17297, + "grad_norm": 0.810002864953582, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 17297 + }, + { + "epoch": 0.17298, + "grad_norm": 0.8786335465347755, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 17298 + }, + { + "epoch": 0.17299, + "grad_norm": 0.9123623519971706, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 17299 + }, + { + "epoch": 0.173, + "grad_norm": 1.0225648865485466, + "learning_rate": 0.003, + "loss": 4.0893, + "step": 17300 + }, + { + "epoch": 0.17301, + "grad_norm": 0.9924731691948884, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 17301 + }, + { + "epoch": 0.17302, + "grad_norm": 1.206901678646934, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 17302 + }, + { + "epoch": 0.17303, + "grad_norm": 1.1562806868597792, + "learning_rate": 0.003, + "loss": 4.07, + "step": 17303 + }, + { + "epoch": 0.17304, + "grad_norm": 1.0997877701464753, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 17304 + }, + { + "epoch": 0.17305, + "grad_norm": 1.2625484054442548, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 17305 + }, + { + "epoch": 0.17306, + "grad_norm": 0.884071396044905, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 17306 + }, + { + "epoch": 0.17307, + "grad_norm": 0.9431257851158984, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 17307 + }, + { + "epoch": 0.17308, + "grad_norm": 1.230315838394911, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 17308 + }, + { + "epoch": 0.17309, + "grad_norm": 1.0486887945853598, + "learning_rate": 0.003, + "loss": 4.0899, + "step": 17309 + }, + { + "epoch": 0.1731, + "grad_norm": 1.0506830219240217, + "learning_rate": 0.003, + "loss": 4.0876, + "step": 17310 + }, + { + "epoch": 0.17311, + "grad_norm": 1.038723813097465, + "learning_rate": 0.003, + "loss": 4.0905, + "step": 17311 + }, + { + "epoch": 0.17312, + "grad_norm": 0.9917597151620848, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 17312 + }, + { + "epoch": 0.17313, + "grad_norm": 1.0334024333762346, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 17313 + }, + { + "epoch": 0.17314, + "grad_norm": 0.9065277870507201, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 17314 + }, + { + "epoch": 0.17315, + "grad_norm": 0.9846212674739533, + "learning_rate": 0.003, + "loss": 4.055, + "step": 17315 + }, + { + "epoch": 0.17316, + "grad_norm": 1.0236394876629982, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 17316 + }, + { + "epoch": 0.17317, + "grad_norm": 1.1553500771531662, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 17317 + }, + { + "epoch": 0.17318, + "grad_norm": 1.2303128794299505, + "learning_rate": 0.003, + "loss": 4.1004, + "step": 17318 + }, + { + "epoch": 0.17319, + "grad_norm": 0.8221270595610066, + "learning_rate": 0.003, + "loss": 4.043, + "step": 17319 + }, + { + "epoch": 0.1732, + "grad_norm": 0.7106796752628632, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 17320 + }, + { + "epoch": 0.17321, + "grad_norm": 0.7772468577680012, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 17321 + }, + { + "epoch": 0.17322, + "grad_norm": 0.8481199305366922, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 17322 + }, + { + "epoch": 0.17323, + "grad_norm": 0.9689187681984914, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 17323 + }, + { + "epoch": 0.17324, + "grad_norm": 1.1526405502476156, + "learning_rate": 0.003, + "loss": 4.0923, + "step": 17324 + }, + { + "epoch": 0.17325, + "grad_norm": 0.7822422148942367, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 17325 + }, + { + "epoch": 0.17326, + "grad_norm": 1.0171339475636036, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 17326 + }, + { + "epoch": 0.17327, + "grad_norm": 1.354111303393376, + "learning_rate": 0.003, + "loss": 4.1079, + "step": 17327 + }, + { + "epoch": 0.17328, + "grad_norm": 0.8992179536397418, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 17328 + }, + { + "epoch": 0.17329, + "grad_norm": 0.8054732762814114, + "learning_rate": 0.003, + "loss": 4.0854, + "step": 17329 + }, + { + "epoch": 0.1733, + "grad_norm": 0.9484772555922377, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 17330 + }, + { + "epoch": 0.17331, + "grad_norm": 1.0855720081089342, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 17331 + }, + { + "epoch": 0.17332, + "grad_norm": 0.9029114156572293, + "learning_rate": 0.003, + "loss": 4.0926, + "step": 17332 + }, + { + "epoch": 0.17333, + "grad_norm": 1.0058238317843218, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 17333 + }, + { + "epoch": 0.17334, + "grad_norm": 1.1337844324970352, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 17334 + }, + { + "epoch": 0.17335, + "grad_norm": 1.0862685725262557, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 17335 + }, + { + "epoch": 0.17336, + "grad_norm": 1.4698406767933623, + "learning_rate": 0.003, + "loss": 4.091, + "step": 17336 + }, + { + "epoch": 0.17337, + "grad_norm": 0.72100982932371, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 17337 + }, + { + "epoch": 0.17338, + "grad_norm": 0.8170015352081219, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 17338 + }, + { + "epoch": 0.17339, + "grad_norm": 1.3179004895213364, + "learning_rate": 0.003, + "loss": 4.0949, + "step": 17339 + }, + { + "epoch": 0.1734, + "grad_norm": 0.8448654213092218, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 17340 + }, + { + "epoch": 0.17341, + "grad_norm": 0.9276373697432436, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 17341 + }, + { + "epoch": 0.17342, + "grad_norm": 1.0830342011001926, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 17342 + }, + { + "epoch": 0.17343, + "grad_norm": 1.091848302916558, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 17343 + }, + { + "epoch": 0.17344, + "grad_norm": 1.0252905183921122, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 17344 + }, + { + "epoch": 0.17345, + "grad_norm": 1.1132555077428035, + "learning_rate": 0.003, + "loss": 4.063, + "step": 17345 + }, + { + "epoch": 0.17346, + "grad_norm": 1.0857865629763015, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 17346 + }, + { + "epoch": 0.17347, + "grad_norm": 1.1375354129640145, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 17347 + }, + { + "epoch": 0.17348, + "grad_norm": 1.0956890562324806, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 17348 + }, + { + "epoch": 0.17349, + "grad_norm": 1.0469158886360008, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 17349 + }, + { + "epoch": 0.1735, + "grad_norm": 0.9175800875146404, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 17350 + }, + { + "epoch": 0.17351, + "grad_norm": 0.8584943572614899, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 17351 + }, + { + "epoch": 0.17352, + "grad_norm": 0.7444349531639638, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 17352 + }, + { + "epoch": 0.17353, + "grad_norm": 0.8776957230418969, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 17353 + }, + { + "epoch": 0.17354, + "grad_norm": 1.145625501691808, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 17354 + }, + { + "epoch": 0.17355, + "grad_norm": 1.2105852005479336, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 17355 + }, + { + "epoch": 0.17356, + "grad_norm": 0.8178745125405998, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 17356 + }, + { + "epoch": 0.17357, + "grad_norm": 0.9108309793351036, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 17357 + }, + { + "epoch": 0.17358, + "grad_norm": 1.189053421049032, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 17358 + }, + { + "epoch": 0.17359, + "grad_norm": 0.9468792229084793, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 17359 + }, + { + "epoch": 0.1736, + "grad_norm": 1.098054687668747, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 17360 + }, + { + "epoch": 0.17361, + "grad_norm": 0.9086361986433631, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 17361 + }, + { + "epoch": 0.17362, + "grad_norm": 0.9782307471915842, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 17362 + }, + { + "epoch": 0.17363, + "grad_norm": 0.9901601391035325, + "learning_rate": 0.003, + "loss": 4.0917, + "step": 17363 + }, + { + "epoch": 0.17364, + "grad_norm": 1.0758577751462868, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 17364 + }, + { + "epoch": 0.17365, + "grad_norm": 1.0360928790175765, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 17365 + }, + { + "epoch": 0.17366, + "grad_norm": 1.1645101216826783, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 17366 + }, + { + "epoch": 0.17367, + "grad_norm": 1.0148337353013932, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 17367 + }, + { + "epoch": 0.17368, + "grad_norm": 1.0028596701038006, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 17368 + }, + { + "epoch": 0.17369, + "grad_norm": 0.9967669407302905, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 17369 + }, + { + "epoch": 0.1737, + "grad_norm": 1.0433476915796323, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 17370 + }, + { + "epoch": 0.17371, + "grad_norm": 1.070718613934026, + "learning_rate": 0.003, + "loss": 4.071, + "step": 17371 + }, + { + "epoch": 0.17372, + "grad_norm": 1.1105900956664416, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 17372 + }, + { + "epoch": 0.17373, + "grad_norm": 1.0781507494912694, + "learning_rate": 0.003, + "loss": 4.0858, + "step": 17373 + }, + { + "epoch": 0.17374, + "grad_norm": 0.9763675091047451, + "learning_rate": 0.003, + "loss": 4.0955, + "step": 17374 + }, + { + "epoch": 0.17375, + "grad_norm": 0.9983156065201204, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 17375 + }, + { + "epoch": 0.17376, + "grad_norm": 0.998818685427647, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 17376 + }, + { + "epoch": 0.17377, + "grad_norm": 1.0319691568906249, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 17377 + }, + { + "epoch": 0.17378, + "grad_norm": 1.0607378875823426, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 17378 + }, + { + "epoch": 0.17379, + "grad_norm": 0.9570106426751588, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 17379 + }, + { + "epoch": 0.1738, + "grad_norm": 0.9322175748386616, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 17380 + }, + { + "epoch": 0.17381, + "grad_norm": 0.9503430093296958, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 17381 + }, + { + "epoch": 0.17382, + "grad_norm": 1.1615834997993502, + "learning_rate": 0.003, + "loss": 4.0987, + "step": 17382 + }, + { + "epoch": 0.17383, + "grad_norm": 1.0258029144569443, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 17383 + }, + { + "epoch": 0.17384, + "grad_norm": 1.1409943670382854, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 17384 + }, + { + "epoch": 0.17385, + "grad_norm": 1.0240160505569038, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 17385 + }, + { + "epoch": 0.17386, + "grad_norm": 1.3379467648037928, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 17386 + }, + { + "epoch": 0.17387, + "grad_norm": 1.1069045123067187, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 17387 + }, + { + "epoch": 0.17388, + "grad_norm": 0.9391032169002569, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 17388 + }, + { + "epoch": 0.17389, + "grad_norm": 0.8465527619309497, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 17389 + }, + { + "epoch": 0.1739, + "grad_norm": 0.9115123592257924, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 17390 + }, + { + "epoch": 0.17391, + "grad_norm": 1.0267818937768358, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 17391 + }, + { + "epoch": 0.17392, + "grad_norm": 1.0840950777475424, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 17392 + }, + { + "epoch": 0.17393, + "grad_norm": 1.0720175208999065, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 17393 + }, + { + "epoch": 0.17394, + "grad_norm": 1.0861924479785203, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 17394 + }, + { + "epoch": 0.17395, + "grad_norm": 1.1684947012063378, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 17395 + }, + { + "epoch": 0.17396, + "grad_norm": 0.8729189983594349, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 17396 + }, + { + "epoch": 0.17397, + "grad_norm": 1.0084166128638665, + "learning_rate": 0.003, + "loss": 4.071, + "step": 17397 + }, + { + "epoch": 0.17398, + "grad_norm": 1.2085980242919379, + "learning_rate": 0.003, + "loss": 4.1103, + "step": 17398 + }, + { + "epoch": 0.17399, + "grad_norm": 0.8483244454336815, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 17399 + }, + { + "epoch": 0.174, + "grad_norm": 0.9487716731168339, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 17400 + }, + { + "epoch": 0.17401, + "grad_norm": 1.2550849590668514, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 17401 + }, + { + "epoch": 0.17402, + "grad_norm": 0.9832587861181368, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 17402 + }, + { + "epoch": 0.17403, + "grad_norm": 0.9449242987557578, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 17403 + }, + { + "epoch": 0.17404, + "grad_norm": 0.954583848562176, + "learning_rate": 0.003, + "loss": 4.062, + "step": 17404 + }, + { + "epoch": 0.17405, + "grad_norm": 1.3644166031252456, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 17405 + }, + { + "epoch": 0.17406, + "grad_norm": 0.8154326706745237, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 17406 + }, + { + "epoch": 0.17407, + "grad_norm": 0.7457657414502429, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 17407 + }, + { + "epoch": 0.17408, + "grad_norm": 0.7963566346436325, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 17408 + }, + { + "epoch": 0.17409, + "grad_norm": 0.9041902638188678, + "learning_rate": 0.003, + "loss": 4.077, + "step": 17409 + }, + { + "epoch": 0.1741, + "grad_norm": 0.9664802094530911, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 17410 + }, + { + "epoch": 0.17411, + "grad_norm": 1.0939018029784824, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 17411 + }, + { + "epoch": 0.17412, + "grad_norm": 1.0473154442891432, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 17412 + }, + { + "epoch": 0.17413, + "grad_norm": 1.0928623297877686, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 17413 + }, + { + "epoch": 0.17414, + "grad_norm": 0.9570883221060904, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 17414 + }, + { + "epoch": 0.17415, + "grad_norm": 0.9568934715111818, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 17415 + }, + { + "epoch": 0.17416, + "grad_norm": 1.3103298670821841, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 17416 + }, + { + "epoch": 0.17417, + "grad_norm": 0.8103512871053109, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 17417 + }, + { + "epoch": 0.17418, + "grad_norm": 0.7580757957459259, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 17418 + }, + { + "epoch": 0.17419, + "grad_norm": 0.749639732167045, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 17419 + }, + { + "epoch": 0.1742, + "grad_norm": 0.977760452716497, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 17420 + }, + { + "epoch": 0.17421, + "grad_norm": 1.330086505455746, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 17421 + }, + { + "epoch": 0.17422, + "grad_norm": 0.8629504666869762, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 17422 + }, + { + "epoch": 0.17423, + "grad_norm": 0.9872652670040969, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 17423 + }, + { + "epoch": 0.17424, + "grad_norm": 1.2689163832846964, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 17424 + }, + { + "epoch": 0.17425, + "grad_norm": 0.9710642473103328, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 17425 + }, + { + "epoch": 0.17426, + "grad_norm": 0.9229760854248968, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 17426 + }, + { + "epoch": 0.17427, + "grad_norm": 1.0342151775058164, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 17427 + }, + { + "epoch": 0.17428, + "grad_norm": 0.9224727170765185, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 17428 + }, + { + "epoch": 0.17429, + "grad_norm": 0.9446734128291358, + "learning_rate": 0.003, + "loss": 4.0978, + "step": 17429 + }, + { + "epoch": 0.1743, + "grad_norm": 1.1119508829480327, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 17430 + }, + { + "epoch": 0.17431, + "grad_norm": 1.057167072054624, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 17431 + }, + { + "epoch": 0.17432, + "grad_norm": 1.1827444189799694, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 17432 + }, + { + "epoch": 0.17433, + "grad_norm": 0.8420110056845417, + "learning_rate": 0.003, + "loss": 4.069, + "step": 17433 + }, + { + "epoch": 0.17434, + "grad_norm": 0.8455112477746531, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 17434 + }, + { + "epoch": 0.17435, + "grad_norm": 1.0463088373757377, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 17435 + }, + { + "epoch": 0.17436, + "grad_norm": 1.2351222856440298, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 17436 + }, + { + "epoch": 0.17437, + "grad_norm": 0.9388773475885295, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 17437 + }, + { + "epoch": 0.17438, + "grad_norm": 0.8815029122193192, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 17438 + }, + { + "epoch": 0.17439, + "grad_norm": 0.8966156210318048, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 17439 + }, + { + "epoch": 0.1744, + "grad_norm": 1.0203851476970787, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 17440 + }, + { + "epoch": 0.17441, + "grad_norm": 0.9996205784416918, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 17441 + }, + { + "epoch": 0.17442, + "grad_norm": 1.126059731506089, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 17442 + }, + { + "epoch": 0.17443, + "grad_norm": 0.9331936382268411, + "learning_rate": 0.003, + "loss": 4.074, + "step": 17443 + }, + { + "epoch": 0.17444, + "grad_norm": 1.0105496385476342, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 17444 + }, + { + "epoch": 0.17445, + "grad_norm": 1.0928204443257785, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 17445 + }, + { + "epoch": 0.17446, + "grad_norm": 1.1572820131863522, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 17446 + }, + { + "epoch": 0.17447, + "grad_norm": 1.0665237847510294, + "learning_rate": 0.003, + "loss": 4.054, + "step": 17447 + }, + { + "epoch": 0.17448, + "grad_norm": 1.0452936297722033, + "learning_rate": 0.003, + "loss": 4.0888, + "step": 17448 + }, + { + "epoch": 0.17449, + "grad_norm": 0.8934969129978804, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 17449 + }, + { + "epoch": 0.1745, + "grad_norm": 0.7654360385696133, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 17450 + }, + { + "epoch": 0.17451, + "grad_norm": 0.8138148612580318, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 17451 + }, + { + "epoch": 0.17452, + "grad_norm": 0.9796283087452445, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 17452 + }, + { + "epoch": 0.17453, + "grad_norm": 1.2661429953742989, + "learning_rate": 0.003, + "loss": 4.0898, + "step": 17453 + }, + { + "epoch": 0.17454, + "grad_norm": 0.9170812087056143, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 17454 + }, + { + "epoch": 0.17455, + "grad_norm": 0.97996724284638, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 17455 + }, + { + "epoch": 0.17456, + "grad_norm": 1.2287673793185825, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 17456 + }, + { + "epoch": 0.17457, + "grad_norm": 1.1354527564732044, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 17457 + }, + { + "epoch": 0.17458, + "grad_norm": 1.0129955036546674, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 17458 + }, + { + "epoch": 0.17459, + "grad_norm": 0.9737549741885916, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 17459 + }, + { + "epoch": 0.1746, + "grad_norm": 1.2585601199922931, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 17460 + }, + { + "epoch": 0.17461, + "grad_norm": 0.8894079986364414, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 17461 + }, + { + "epoch": 0.17462, + "grad_norm": 1.1018453590785036, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 17462 + }, + { + "epoch": 0.17463, + "grad_norm": 1.2249502784800037, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 17463 + }, + { + "epoch": 0.17464, + "grad_norm": 0.9035570748542914, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 17464 + }, + { + "epoch": 0.17465, + "grad_norm": 0.879633235118732, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 17465 + }, + { + "epoch": 0.17466, + "grad_norm": 1.0129199795763748, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 17466 + }, + { + "epoch": 0.17467, + "grad_norm": 1.0112181956444075, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 17467 + }, + { + "epoch": 0.17468, + "grad_norm": 0.9692654365233465, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 17468 + }, + { + "epoch": 0.17469, + "grad_norm": 1.0764108319012935, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 17469 + }, + { + "epoch": 0.1747, + "grad_norm": 1.0964680562754767, + "learning_rate": 0.003, + "loss": 4.0871, + "step": 17470 + }, + { + "epoch": 0.17471, + "grad_norm": 0.9871966617261495, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 17471 + }, + { + "epoch": 0.17472, + "grad_norm": 0.985886618613454, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 17472 + }, + { + "epoch": 0.17473, + "grad_norm": 1.0321017084840978, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 17473 + }, + { + "epoch": 0.17474, + "grad_norm": 1.136230782878352, + "learning_rate": 0.003, + "loss": 4.062, + "step": 17474 + }, + { + "epoch": 0.17475, + "grad_norm": 0.8086062416715832, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 17475 + }, + { + "epoch": 0.17476, + "grad_norm": 0.8572538040774352, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 17476 + }, + { + "epoch": 0.17477, + "grad_norm": 1.0351091030642354, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 17477 + }, + { + "epoch": 0.17478, + "grad_norm": 1.2124599927207744, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 17478 + }, + { + "epoch": 0.17479, + "grad_norm": 0.9822449266807828, + "learning_rate": 0.003, + "loss": 4.044, + "step": 17479 + }, + { + "epoch": 0.1748, + "grad_norm": 1.0302696534913642, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 17480 + }, + { + "epoch": 0.17481, + "grad_norm": 1.0832331158079342, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 17481 + }, + { + "epoch": 0.17482, + "grad_norm": 0.9017235724556893, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 17482 + }, + { + "epoch": 0.17483, + "grad_norm": 0.8496219401902695, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 17483 + }, + { + "epoch": 0.17484, + "grad_norm": 0.977480625030177, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 17484 + }, + { + "epoch": 0.17485, + "grad_norm": 0.9383827825555782, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 17485 + }, + { + "epoch": 0.17486, + "grad_norm": 1.0090821639471208, + "learning_rate": 0.003, + "loss": 4.1031, + "step": 17486 + }, + { + "epoch": 0.17487, + "grad_norm": 1.0792418193479596, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 17487 + }, + { + "epoch": 0.17488, + "grad_norm": 1.0094772426484677, + "learning_rate": 0.003, + "loss": 4.06, + "step": 17488 + }, + { + "epoch": 0.17489, + "grad_norm": 1.006605186665462, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 17489 + }, + { + "epoch": 0.1749, + "grad_norm": 1.0591535388186128, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 17490 + }, + { + "epoch": 0.17491, + "grad_norm": 1.3946439375800292, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 17491 + }, + { + "epoch": 0.17492, + "grad_norm": 0.7802234344945406, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 17492 + }, + { + "epoch": 0.17493, + "grad_norm": 0.8530768664884452, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 17493 + }, + { + "epoch": 0.17494, + "grad_norm": 0.996113870875562, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 17494 + }, + { + "epoch": 0.17495, + "grad_norm": 1.2501506680186198, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 17495 + }, + { + "epoch": 0.17496, + "grad_norm": 0.8363406876276396, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 17496 + }, + { + "epoch": 0.17497, + "grad_norm": 0.8821571831432515, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 17497 + }, + { + "epoch": 0.17498, + "grad_norm": 1.0018645592647684, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 17498 + }, + { + "epoch": 0.17499, + "grad_norm": 1.222270851642828, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 17499 + }, + { + "epoch": 0.175, + "grad_norm": 0.8666381106984253, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 17500 + }, + { + "epoch": 0.17501, + "grad_norm": 0.9727325583642625, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 17501 + }, + { + "epoch": 0.17502, + "grad_norm": 1.353149922157011, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 17502 + }, + { + "epoch": 0.17503, + "grad_norm": 0.9052186820907415, + "learning_rate": 0.003, + "loss": 4.0957, + "step": 17503 + }, + { + "epoch": 0.17504, + "grad_norm": 0.8323641203985677, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 17504 + }, + { + "epoch": 0.17505, + "grad_norm": 0.9457537232950727, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 17505 + }, + { + "epoch": 0.17506, + "grad_norm": 0.9286030558122577, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 17506 + }, + { + "epoch": 0.17507, + "grad_norm": 0.9459770126040448, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 17507 + }, + { + "epoch": 0.17508, + "grad_norm": 0.8538186293287007, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 17508 + }, + { + "epoch": 0.17509, + "grad_norm": 0.8998755338660392, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 17509 + }, + { + "epoch": 0.1751, + "grad_norm": 0.9250735361799428, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 17510 + }, + { + "epoch": 0.17511, + "grad_norm": 1.006523017467908, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 17511 + }, + { + "epoch": 0.17512, + "grad_norm": 1.098382187096705, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 17512 + }, + { + "epoch": 0.17513, + "grad_norm": 1.3114498740457738, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 17513 + }, + { + "epoch": 0.17514, + "grad_norm": 1.0789111291521345, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 17514 + }, + { + "epoch": 0.17515, + "grad_norm": 0.9932049526387396, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 17515 + }, + { + "epoch": 0.17516, + "grad_norm": 0.9809181062793424, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 17516 + }, + { + "epoch": 0.17517, + "grad_norm": 1.2384142766831892, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 17517 + }, + { + "epoch": 0.17518, + "grad_norm": 1.0566744292034602, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 17518 + }, + { + "epoch": 0.17519, + "grad_norm": 1.0439627691701772, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 17519 + }, + { + "epoch": 0.1752, + "grad_norm": 1.087108228097138, + "learning_rate": 0.003, + "loss": 4.07, + "step": 17520 + }, + { + "epoch": 0.17521, + "grad_norm": 1.0811612996416373, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 17521 + }, + { + "epoch": 0.17522, + "grad_norm": 1.0489945205180227, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 17522 + }, + { + "epoch": 0.17523, + "grad_norm": 1.0088391548090017, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 17523 + }, + { + "epoch": 0.17524, + "grad_norm": 1.3720702857014826, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 17524 + }, + { + "epoch": 0.17525, + "grad_norm": 0.7346741763004629, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 17525 + }, + { + "epoch": 0.17526, + "grad_norm": 0.7371430558909786, + "learning_rate": 0.003, + "loss": 4.052, + "step": 17526 + }, + { + "epoch": 0.17527, + "grad_norm": 1.0335250968706458, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 17527 + }, + { + "epoch": 0.17528, + "grad_norm": 1.2870242233826459, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 17528 + }, + { + "epoch": 0.17529, + "grad_norm": 1.0084961135866193, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 17529 + }, + { + "epoch": 0.1753, + "grad_norm": 0.9870555662671587, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 17530 + }, + { + "epoch": 0.17531, + "grad_norm": 0.9292897155818721, + "learning_rate": 0.003, + "loss": 4.088, + "step": 17531 + }, + { + "epoch": 0.17532, + "grad_norm": 0.9388423753806633, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 17532 + }, + { + "epoch": 0.17533, + "grad_norm": 1.0584388131545883, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 17533 + }, + { + "epoch": 0.17534, + "grad_norm": 1.13609865828893, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 17534 + }, + { + "epoch": 0.17535, + "grad_norm": 0.9329255381229412, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 17535 + }, + { + "epoch": 0.17536, + "grad_norm": 0.8518293033910412, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 17536 + }, + { + "epoch": 0.17537, + "grad_norm": 0.886062277606697, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 17537 + }, + { + "epoch": 0.17538, + "grad_norm": 1.0292325604531536, + "learning_rate": 0.003, + "loss": 4.0929, + "step": 17538 + }, + { + "epoch": 0.17539, + "grad_norm": 1.189765651502402, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 17539 + }, + { + "epoch": 0.1754, + "grad_norm": 0.8014526272008075, + "learning_rate": 0.003, + "loss": 4.0939, + "step": 17540 + }, + { + "epoch": 0.17541, + "grad_norm": 0.9373040581987747, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 17541 + }, + { + "epoch": 0.17542, + "grad_norm": 0.9671931557244087, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 17542 + }, + { + "epoch": 0.17543, + "grad_norm": 1.102005739349272, + "learning_rate": 0.003, + "loss": 4.0864, + "step": 17543 + }, + { + "epoch": 0.17544, + "grad_norm": 1.0212457186937425, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 17544 + }, + { + "epoch": 0.17545, + "grad_norm": 1.097007047821511, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 17545 + }, + { + "epoch": 0.17546, + "grad_norm": 1.2920259945803134, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 17546 + }, + { + "epoch": 0.17547, + "grad_norm": 0.9093152883813185, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 17547 + }, + { + "epoch": 0.17548, + "grad_norm": 0.872163951773715, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 17548 + }, + { + "epoch": 0.17549, + "grad_norm": 1.0825820771732502, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 17549 + }, + { + "epoch": 0.1755, + "grad_norm": 1.2291404499757652, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 17550 + }, + { + "epoch": 0.17551, + "grad_norm": 0.9608878099435652, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 17551 + }, + { + "epoch": 0.17552, + "grad_norm": 1.2212705514834234, + "learning_rate": 0.003, + "loss": 4.1053, + "step": 17552 + }, + { + "epoch": 0.17553, + "grad_norm": 0.9015728060435605, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 17553 + }, + { + "epoch": 0.17554, + "grad_norm": 1.0215740273404283, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 17554 + }, + { + "epoch": 0.17555, + "grad_norm": 1.3722045704050925, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 17555 + }, + { + "epoch": 0.17556, + "grad_norm": 0.6377539434609153, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 17556 + }, + { + "epoch": 0.17557, + "grad_norm": 0.774150100431813, + "learning_rate": 0.003, + "loss": 4.06, + "step": 17557 + }, + { + "epoch": 0.17558, + "grad_norm": 1.0602729285529124, + "learning_rate": 0.003, + "loss": 4.0947, + "step": 17558 + }, + { + "epoch": 0.17559, + "grad_norm": 1.1932011482600418, + "learning_rate": 0.003, + "loss": 4.0872, + "step": 17559 + }, + { + "epoch": 0.1756, + "grad_norm": 0.9390199055531998, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 17560 + }, + { + "epoch": 0.17561, + "grad_norm": 1.0353510699136599, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 17561 + }, + { + "epoch": 0.17562, + "grad_norm": 1.1649906107525778, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 17562 + }, + { + "epoch": 0.17563, + "grad_norm": 0.987056503854763, + "learning_rate": 0.003, + "loss": 4.058, + "step": 17563 + }, + { + "epoch": 0.17564, + "grad_norm": 1.0796211490737222, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 17564 + }, + { + "epoch": 0.17565, + "grad_norm": 1.0899544331737505, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 17565 + }, + { + "epoch": 0.17566, + "grad_norm": 0.992944636414507, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 17566 + }, + { + "epoch": 0.17567, + "grad_norm": 1.1401343876286918, + "learning_rate": 0.003, + "loss": 4.0845, + "step": 17567 + }, + { + "epoch": 0.17568, + "grad_norm": 0.971798945882637, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 17568 + }, + { + "epoch": 0.17569, + "grad_norm": 1.1905296916583286, + "learning_rate": 0.003, + "loss": 4.0874, + "step": 17569 + }, + { + "epoch": 0.1757, + "grad_norm": 1.0013268100201305, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 17570 + }, + { + "epoch": 0.17571, + "grad_norm": 1.0703912570805727, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 17571 + }, + { + "epoch": 0.17572, + "grad_norm": 0.9625918171431953, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 17572 + }, + { + "epoch": 0.17573, + "grad_norm": 0.7412609543066667, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 17573 + }, + { + "epoch": 0.17574, + "grad_norm": 0.7914573148618446, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 17574 + }, + { + "epoch": 0.17575, + "grad_norm": 1.0624093092449887, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 17575 + }, + { + "epoch": 0.17576, + "grad_norm": 1.2706088620347102, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 17576 + }, + { + "epoch": 0.17577, + "grad_norm": 0.9145055602538661, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 17577 + }, + { + "epoch": 0.17578, + "grad_norm": 0.9130660066642083, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 17578 + }, + { + "epoch": 0.17579, + "grad_norm": 0.9765551728246369, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 17579 + }, + { + "epoch": 0.1758, + "grad_norm": 1.1276303206040579, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 17580 + }, + { + "epoch": 0.17581, + "grad_norm": 0.9454902029144582, + "learning_rate": 0.003, + "loss": 4.0922, + "step": 17581 + }, + { + "epoch": 0.17582, + "grad_norm": 0.8299275914035231, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 17582 + }, + { + "epoch": 0.17583, + "grad_norm": 0.678462629481439, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 17583 + }, + { + "epoch": 0.17584, + "grad_norm": 0.7613480776188396, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 17584 + }, + { + "epoch": 0.17585, + "grad_norm": 0.7665262758139261, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 17585 + }, + { + "epoch": 0.17586, + "grad_norm": 0.8038868263865511, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 17586 + }, + { + "epoch": 0.17587, + "grad_norm": 0.9035483755998116, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 17587 + }, + { + "epoch": 0.17588, + "grad_norm": 1.2267233128500863, + "learning_rate": 0.003, + "loss": 4.1075, + "step": 17588 + }, + { + "epoch": 0.17589, + "grad_norm": 1.1480706308694582, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 17589 + }, + { + "epoch": 0.1759, + "grad_norm": 1.0042048357213051, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 17590 + }, + { + "epoch": 0.17591, + "grad_norm": 0.9294522475431819, + "learning_rate": 0.003, + "loss": 4.08, + "step": 17591 + }, + { + "epoch": 0.17592, + "grad_norm": 1.0362643967688903, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 17592 + }, + { + "epoch": 0.17593, + "grad_norm": 1.0321995617895257, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 17593 + }, + { + "epoch": 0.17594, + "grad_norm": 1.0386430315412396, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 17594 + }, + { + "epoch": 0.17595, + "grad_norm": 1.0678050476335301, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 17595 + }, + { + "epoch": 0.17596, + "grad_norm": 1.0624889861728988, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 17596 + }, + { + "epoch": 0.17597, + "grad_norm": 0.9901044754423722, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 17597 + }, + { + "epoch": 0.17598, + "grad_norm": 0.9302609848888366, + "learning_rate": 0.003, + "loss": 4.1018, + "step": 17598 + }, + { + "epoch": 0.17599, + "grad_norm": 0.9379433888505828, + "learning_rate": 0.003, + "loss": 4.1014, + "step": 17599 + }, + { + "epoch": 0.176, + "grad_norm": 0.9677222269973247, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 17600 + }, + { + "epoch": 0.17601, + "grad_norm": 0.8913119204725604, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 17601 + }, + { + "epoch": 0.17602, + "grad_norm": 0.9721587623169952, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 17602 + }, + { + "epoch": 0.17603, + "grad_norm": 1.1164976280496317, + "learning_rate": 0.003, + "loss": 4.07, + "step": 17603 + }, + { + "epoch": 0.17604, + "grad_norm": 1.1014450371377844, + "learning_rate": 0.003, + "loss": 4.067, + "step": 17604 + }, + { + "epoch": 0.17605, + "grad_norm": 1.0243577711509433, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 17605 + }, + { + "epoch": 0.17606, + "grad_norm": 1.4112824946532196, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 17606 + }, + { + "epoch": 0.17607, + "grad_norm": 0.909086466481729, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 17607 + }, + { + "epoch": 0.17608, + "grad_norm": 0.8941405452522573, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 17608 + }, + { + "epoch": 0.17609, + "grad_norm": 0.9002891241672091, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 17609 + }, + { + "epoch": 0.1761, + "grad_norm": 1.1846911865106737, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 17610 + }, + { + "epoch": 0.17611, + "grad_norm": 1.3847909770895408, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 17611 + }, + { + "epoch": 0.17612, + "grad_norm": 0.7872255177308692, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 17612 + }, + { + "epoch": 0.17613, + "grad_norm": 0.7621752191697989, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 17613 + }, + { + "epoch": 0.17614, + "grad_norm": 0.8753665906041146, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 17614 + }, + { + "epoch": 0.17615, + "grad_norm": 0.9097103983080875, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 17615 + }, + { + "epoch": 0.17616, + "grad_norm": 1.1185381656592095, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 17616 + }, + { + "epoch": 0.17617, + "grad_norm": 1.0690748462720623, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 17617 + }, + { + "epoch": 0.17618, + "grad_norm": 1.0601241647580704, + "learning_rate": 0.003, + "loss": 4.0884, + "step": 17618 + }, + { + "epoch": 0.17619, + "grad_norm": 1.247612804343138, + "learning_rate": 0.003, + "loss": 4.0981, + "step": 17619 + }, + { + "epoch": 0.1762, + "grad_norm": 0.7885733143015525, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 17620 + }, + { + "epoch": 0.17621, + "grad_norm": 0.7161479273189552, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 17621 + }, + { + "epoch": 0.17622, + "grad_norm": 0.7512040944459849, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 17622 + }, + { + "epoch": 0.17623, + "grad_norm": 0.8844730029656654, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 17623 + }, + { + "epoch": 0.17624, + "grad_norm": 0.9394838631909322, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 17624 + }, + { + "epoch": 0.17625, + "grad_norm": 1.1065916860522222, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 17625 + }, + { + "epoch": 0.17626, + "grad_norm": 0.9532897534176753, + "learning_rate": 0.003, + "loss": 4.045, + "step": 17626 + }, + { + "epoch": 0.17627, + "grad_norm": 1.1634528876241659, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 17627 + }, + { + "epoch": 0.17628, + "grad_norm": 1.129794557661065, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 17628 + }, + { + "epoch": 0.17629, + "grad_norm": 0.7636330711723168, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 17629 + }, + { + "epoch": 0.1763, + "grad_norm": 0.7024129093009343, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 17630 + }, + { + "epoch": 0.17631, + "grad_norm": 0.7176146286986166, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 17631 + }, + { + "epoch": 0.17632, + "grad_norm": 0.8296827249579345, + "learning_rate": 0.003, + "loss": 4.1105, + "step": 17632 + }, + { + "epoch": 0.17633, + "grad_norm": 0.9026637953044871, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 17633 + }, + { + "epoch": 0.17634, + "grad_norm": 0.9389993005925434, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 17634 + }, + { + "epoch": 0.17635, + "grad_norm": 1.2514870016906314, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 17635 + }, + { + "epoch": 0.17636, + "grad_norm": 1.291792456196519, + "learning_rate": 0.003, + "loss": 4.064, + "step": 17636 + }, + { + "epoch": 0.17637, + "grad_norm": 0.9093466700842752, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 17637 + }, + { + "epoch": 0.17638, + "grad_norm": 0.8304348743136708, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 17638 + }, + { + "epoch": 0.17639, + "grad_norm": 0.8429933865141075, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 17639 + }, + { + "epoch": 0.1764, + "grad_norm": 0.9945108851524345, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 17640 + }, + { + "epoch": 0.17641, + "grad_norm": 1.2174442638356227, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 17641 + }, + { + "epoch": 0.17642, + "grad_norm": 0.8646367544346092, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 17642 + }, + { + "epoch": 0.17643, + "grad_norm": 0.9460394026879193, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 17643 + }, + { + "epoch": 0.17644, + "grad_norm": 1.2641667456336232, + "learning_rate": 0.003, + "loss": 4.066, + "step": 17644 + }, + { + "epoch": 0.17645, + "grad_norm": 1.0773753888491973, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 17645 + }, + { + "epoch": 0.17646, + "grad_norm": 1.0069713211103928, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 17646 + }, + { + "epoch": 0.17647, + "grad_norm": 1.0811209824487356, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 17647 + }, + { + "epoch": 0.17648, + "grad_norm": 1.0829590680330914, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 17648 + }, + { + "epoch": 0.17649, + "grad_norm": 1.0459419260321705, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 17649 + }, + { + "epoch": 0.1765, + "grad_norm": 0.9625584670933655, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 17650 + }, + { + "epoch": 0.17651, + "grad_norm": 1.095464086638937, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 17651 + }, + { + "epoch": 0.17652, + "grad_norm": 0.9799642450227692, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 17652 + }, + { + "epoch": 0.17653, + "grad_norm": 1.0897354310240974, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 17653 + }, + { + "epoch": 0.17654, + "grad_norm": 1.1938858235269565, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 17654 + }, + { + "epoch": 0.17655, + "grad_norm": 0.981259696979905, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 17655 + }, + { + "epoch": 0.17656, + "grad_norm": 1.0272275456051927, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 17656 + }, + { + "epoch": 0.17657, + "grad_norm": 1.024848115620557, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 17657 + }, + { + "epoch": 0.17658, + "grad_norm": 1.0140438894077453, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 17658 + }, + { + "epoch": 0.17659, + "grad_norm": 1.165131459322621, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 17659 + }, + { + "epoch": 0.1766, + "grad_norm": 0.8285731505805449, + "learning_rate": 0.003, + "loss": 4.0945, + "step": 17660 + }, + { + "epoch": 0.17661, + "grad_norm": 0.8474947452505825, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 17661 + }, + { + "epoch": 0.17662, + "grad_norm": 0.8911940399106444, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 17662 + }, + { + "epoch": 0.17663, + "grad_norm": 0.9070082337130119, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 17663 + }, + { + "epoch": 0.17664, + "grad_norm": 1.0905682351973138, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 17664 + }, + { + "epoch": 0.17665, + "grad_norm": 1.0399870630540686, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 17665 + }, + { + "epoch": 0.17666, + "grad_norm": 1.0031453799861838, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 17666 + }, + { + "epoch": 0.17667, + "grad_norm": 1.174886526959176, + "learning_rate": 0.003, + "loss": 4.06, + "step": 17667 + }, + { + "epoch": 0.17668, + "grad_norm": 1.0844044307633838, + "learning_rate": 0.003, + "loss": 4.0987, + "step": 17668 + }, + { + "epoch": 0.17669, + "grad_norm": 1.5208501943438404, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 17669 + }, + { + "epoch": 0.1767, + "grad_norm": 0.8204302415614123, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 17670 + }, + { + "epoch": 0.17671, + "grad_norm": 0.748902511110535, + "learning_rate": 0.003, + "loss": 4.085, + "step": 17671 + }, + { + "epoch": 0.17672, + "grad_norm": 0.8019180167277407, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 17672 + }, + { + "epoch": 0.17673, + "grad_norm": 0.9846201851450112, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 17673 + }, + { + "epoch": 0.17674, + "grad_norm": 1.2130746894721371, + "learning_rate": 0.003, + "loss": 4.0881, + "step": 17674 + }, + { + "epoch": 0.17675, + "grad_norm": 1.0310652665941145, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 17675 + }, + { + "epoch": 0.17676, + "grad_norm": 1.070960888653492, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 17676 + }, + { + "epoch": 0.17677, + "grad_norm": 1.0369528995123871, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 17677 + }, + { + "epoch": 0.17678, + "grad_norm": 1.0777316493884248, + "learning_rate": 0.003, + "loss": 4.1022, + "step": 17678 + }, + { + "epoch": 0.17679, + "grad_norm": 1.062155392290248, + "learning_rate": 0.003, + "loss": 4.047, + "step": 17679 + }, + { + "epoch": 0.1768, + "grad_norm": 1.11994069350479, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 17680 + }, + { + "epoch": 0.17681, + "grad_norm": 0.9236703990400813, + "learning_rate": 0.003, + "loss": 4.0981, + "step": 17681 + }, + { + "epoch": 0.17682, + "grad_norm": 0.8809401745702806, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 17682 + }, + { + "epoch": 0.17683, + "grad_norm": 0.7598480985392145, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 17683 + }, + { + "epoch": 0.17684, + "grad_norm": 0.8414117209055333, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 17684 + }, + { + "epoch": 0.17685, + "grad_norm": 1.0059432913672828, + "learning_rate": 0.003, + "loss": 4.0858, + "step": 17685 + }, + { + "epoch": 0.17686, + "grad_norm": 1.2477655729733743, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 17686 + }, + { + "epoch": 0.17687, + "grad_norm": 0.9003341176618227, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 17687 + }, + { + "epoch": 0.17688, + "grad_norm": 0.7774941544212306, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 17688 + }, + { + "epoch": 0.17689, + "grad_norm": 0.8807379793416121, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 17689 + }, + { + "epoch": 0.1769, + "grad_norm": 1.0575875701830433, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 17690 + }, + { + "epoch": 0.17691, + "grad_norm": 1.141192079271236, + "learning_rate": 0.003, + "loss": 4.081, + "step": 17691 + }, + { + "epoch": 0.17692, + "grad_norm": 0.8937503147821798, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 17692 + }, + { + "epoch": 0.17693, + "grad_norm": 1.1351352710125346, + "learning_rate": 0.003, + "loss": 4.091, + "step": 17693 + }, + { + "epoch": 0.17694, + "grad_norm": 1.0180399488914296, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 17694 + }, + { + "epoch": 0.17695, + "grad_norm": 1.1445765852703722, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 17695 + }, + { + "epoch": 0.17696, + "grad_norm": 1.342358470358858, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 17696 + }, + { + "epoch": 0.17697, + "grad_norm": 0.81088956822231, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 17697 + }, + { + "epoch": 0.17698, + "grad_norm": 0.790464787518728, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 17698 + }, + { + "epoch": 0.17699, + "grad_norm": 0.8861242198885093, + "learning_rate": 0.003, + "loss": 4.054, + "step": 17699 + }, + { + "epoch": 0.177, + "grad_norm": 1.132799416801299, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 17700 + }, + { + "epoch": 0.17701, + "grad_norm": 1.0810401666736835, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 17701 + }, + { + "epoch": 0.17702, + "grad_norm": 1.0551936445221501, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 17702 + }, + { + "epoch": 0.17703, + "grad_norm": 1.10606367940861, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 17703 + }, + { + "epoch": 0.17704, + "grad_norm": 1.091587808812865, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 17704 + }, + { + "epoch": 0.17705, + "grad_norm": 1.0839834434227806, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 17705 + }, + { + "epoch": 0.17706, + "grad_norm": 0.8291290013235343, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 17706 + }, + { + "epoch": 0.17707, + "grad_norm": 0.8173539375183021, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 17707 + }, + { + "epoch": 0.17708, + "grad_norm": 0.9752838753689883, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 17708 + }, + { + "epoch": 0.17709, + "grad_norm": 1.2254185772864274, + "learning_rate": 0.003, + "loss": 4.1006, + "step": 17709 + }, + { + "epoch": 0.1771, + "grad_norm": 0.8492630584848739, + "learning_rate": 0.003, + "loss": 4.058, + "step": 17710 + }, + { + "epoch": 0.17711, + "grad_norm": 0.7708406170827706, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 17711 + }, + { + "epoch": 0.17712, + "grad_norm": 0.7807858835405597, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 17712 + }, + { + "epoch": 0.17713, + "grad_norm": 0.9814571919871725, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 17713 + }, + { + "epoch": 0.17714, + "grad_norm": 1.2387481750162297, + "learning_rate": 0.003, + "loss": 4.073, + "step": 17714 + }, + { + "epoch": 0.17715, + "grad_norm": 0.9017080492045172, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 17715 + }, + { + "epoch": 0.17716, + "grad_norm": 0.8443511103388479, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 17716 + }, + { + "epoch": 0.17717, + "grad_norm": 0.9140798118398156, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 17717 + }, + { + "epoch": 0.17718, + "grad_norm": 1.1251629848477094, + "learning_rate": 0.003, + "loss": 4.0948, + "step": 17718 + }, + { + "epoch": 0.17719, + "grad_norm": 1.101602402541681, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 17719 + }, + { + "epoch": 0.1772, + "grad_norm": 0.8944889884015044, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 17720 + }, + { + "epoch": 0.17721, + "grad_norm": 0.8891059108988009, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 17721 + }, + { + "epoch": 0.17722, + "grad_norm": 0.941304657536419, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 17722 + }, + { + "epoch": 0.17723, + "grad_norm": 1.018081572343701, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 17723 + }, + { + "epoch": 0.17724, + "grad_norm": 1.0351945769809814, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 17724 + }, + { + "epoch": 0.17725, + "grad_norm": 1.0036455029701492, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 17725 + }, + { + "epoch": 0.17726, + "grad_norm": 1.216831047290161, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 17726 + }, + { + "epoch": 0.17727, + "grad_norm": 1.241453752293488, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 17727 + }, + { + "epoch": 0.17728, + "grad_norm": 1.1742073466748046, + "learning_rate": 0.003, + "loss": 4.0872, + "step": 17728 + }, + { + "epoch": 0.17729, + "grad_norm": 1.0817718947060402, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 17729 + }, + { + "epoch": 0.1773, + "grad_norm": 1.1198637280477954, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 17730 + }, + { + "epoch": 0.17731, + "grad_norm": 0.862358664425527, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 17731 + }, + { + "epoch": 0.17732, + "grad_norm": 0.951968756152031, + "learning_rate": 0.003, + "loss": 4.0887, + "step": 17732 + }, + { + "epoch": 0.17733, + "grad_norm": 1.1275992307480718, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 17733 + }, + { + "epoch": 0.17734, + "grad_norm": 0.9318239458139564, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 17734 + }, + { + "epoch": 0.17735, + "grad_norm": 1.0686796120927529, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 17735 + }, + { + "epoch": 0.17736, + "grad_norm": 1.0865367449062062, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 17736 + }, + { + "epoch": 0.17737, + "grad_norm": 0.8961875742936343, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 17737 + }, + { + "epoch": 0.17738, + "grad_norm": 0.9388887030089387, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 17738 + }, + { + "epoch": 0.17739, + "grad_norm": 1.1135205012399243, + "learning_rate": 0.003, + "loss": 4.0978, + "step": 17739 + }, + { + "epoch": 0.1774, + "grad_norm": 1.035901096461221, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 17740 + }, + { + "epoch": 0.17741, + "grad_norm": 1.0882822184521188, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 17741 + }, + { + "epoch": 0.17742, + "grad_norm": 1.0346496357929307, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 17742 + }, + { + "epoch": 0.17743, + "grad_norm": 1.0338938213953968, + "learning_rate": 0.003, + "loss": 4.1076, + "step": 17743 + }, + { + "epoch": 0.17744, + "grad_norm": 1.2146600571131014, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 17744 + }, + { + "epoch": 0.17745, + "grad_norm": 1.0119874983202648, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 17745 + }, + { + "epoch": 0.17746, + "grad_norm": 1.2243174397301575, + "learning_rate": 0.003, + "loss": 4.0819, + "step": 17746 + }, + { + "epoch": 0.17747, + "grad_norm": 0.9029619055701144, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 17747 + }, + { + "epoch": 0.17748, + "grad_norm": 0.8695207764832109, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 17748 + }, + { + "epoch": 0.17749, + "grad_norm": 1.003786174359247, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 17749 + }, + { + "epoch": 0.1775, + "grad_norm": 1.0279342885587657, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 17750 + }, + { + "epoch": 0.17751, + "grad_norm": 1.1516588319766423, + "learning_rate": 0.003, + "loss": 4.0919, + "step": 17751 + }, + { + "epoch": 0.17752, + "grad_norm": 1.0021098157167105, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 17752 + }, + { + "epoch": 0.17753, + "grad_norm": 1.0577194004350186, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 17753 + }, + { + "epoch": 0.17754, + "grad_norm": 1.1436398299522854, + "learning_rate": 0.003, + "loss": 4.096, + "step": 17754 + }, + { + "epoch": 0.17755, + "grad_norm": 1.060653521299358, + "learning_rate": 0.003, + "loss": 4.079, + "step": 17755 + }, + { + "epoch": 0.17756, + "grad_norm": 1.1894465095389832, + "learning_rate": 0.003, + "loss": 4.044, + "step": 17756 + }, + { + "epoch": 0.17757, + "grad_norm": 0.8198346644060902, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 17757 + }, + { + "epoch": 0.17758, + "grad_norm": 0.8451798084455847, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 17758 + }, + { + "epoch": 0.17759, + "grad_norm": 1.0503274118216812, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 17759 + }, + { + "epoch": 0.1776, + "grad_norm": 1.124646780170803, + "learning_rate": 0.003, + "loss": 4.047, + "step": 17760 + }, + { + "epoch": 0.17761, + "grad_norm": 1.065143287806241, + "learning_rate": 0.003, + "loss": 4.065, + "step": 17761 + }, + { + "epoch": 0.17762, + "grad_norm": 1.2147469578570418, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 17762 + }, + { + "epoch": 0.17763, + "grad_norm": 0.8391632810239761, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 17763 + }, + { + "epoch": 0.17764, + "grad_norm": 0.7790993868311349, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 17764 + }, + { + "epoch": 0.17765, + "grad_norm": 1.038140499383883, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 17765 + }, + { + "epoch": 0.17766, + "grad_norm": 1.109700653179373, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 17766 + }, + { + "epoch": 0.17767, + "grad_norm": 0.9571903459326111, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 17767 + }, + { + "epoch": 0.17768, + "grad_norm": 1.154409539171532, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 17768 + }, + { + "epoch": 0.17769, + "grad_norm": 0.9948571566137628, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 17769 + }, + { + "epoch": 0.1777, + "grad_norm": 1.0619873005236888, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 17770 + }, + { + "epoch": 0.17771, + "grad_norm": 0.8736802106802213, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 17771 + }, + { + "epoch": 0.17772, + "grad_norm": 0.9890778642335937, + "learning_rate": 0.003, + "loss": 4.068, + "step": 17772 + }, + { + "epoch": 0.17773, + "grad_norm": 1.1336770101049556, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 17773 + }, + { + "epoch": 0.17774, + "grad_norm": 1.0262813926855707, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 17774 + }, + { + "epoch": 0.17775, + "grad_norm": 1.308342230325029, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 17775 + }, + { + "epoch": 0.17776, + "grad_norm": 0.9102542465251244, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 17776 + }, + { + "epoch": 0.17777, + "grad_norm": 0.8964306997657877, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 17777 + }, + { + "epoch": 0.17778, + "grad_norm": 0.9326358693967879, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 17778 + }, + { + "epoch": 0.17779, + "grad_norm": 1.0254421870200858, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 17779 + }, + { + "epoch": 0.1778, + "grad_norm": 1.1757002711414615, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 17780 + }, + { + "epoch": 0.17781, + "grad_norm": 1.1871005162257249, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 17781 + }, + { + "epoch": 0.17782, + "grad_norm": 1.0293389878995411, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 17782 + }, + { + "epoch": 0.17783, + "grad_norm": 1.0249281860602624, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 17783 + }, + { + "epoch": 0.17784, + "grad_norm": 1.0752679684394768, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 17784 + }, + { + "epoch": 0.17785, + "grad_norm": 1.0121368088327796, + "learning_rate": 0.003, + "loss": 4.048, + "step": 17785 + }, + { + "epoch": 0.17786, + "grad_norm": 0.9315479073309232, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 17786 + }, + { + "epoch": 0.17787, + "grad_norm": 0.8922626930764127, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 17787 + }, + { + "epoch": 0.17788, + "grad_norm": 0.9925985120307392, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 17788 + }, + { + "epoch": 0.17789, + "grad_norm": 1.106031317088377, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 17789 + }, + { + "epoch": 0.1779, + "grad_norm": 0.9720298596996036, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 17790 + }, + { + "epoch": 0.17791, + "grad_norm": 1.077286049770686, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 17791 + }, + { + "epoch": 0.17792, + "grad_norm": 1.1655651850044308, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 17792 + }, + { + "epoch": 0.17793, + "grad_norm": 1.0673077333577037, + "learning_rate": 0.003, + "loss": 4.0955, + "step": 17793 + }, + { + "epoch": 0.17794, + "grad_norm": 1.001312357548442, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 17794 + }, + { + "epoch": 0.17795, + "grad_norm": 0.9140041362328181, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 17795 + }, + { + "epoch": 0.17796, + "grad_norm": 0.882615935763939, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 17796 + }, + { + "epoch": 0.17797, + "grad_norm": 1.0612806870206761, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 17797 + }, + { + "epoch": 0.17798, + "grad_norm": 1.1152761812025842, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 17798 + }, + { + "epoch": 0.17799, + "grad_norm": 0.964577613039433, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 17799 + }, + { + "epoch": 0.178, + "grad_norm": 1.0799959902679477, + "learning_rate": 0.003, + "loss": 4.07, + "step": 17800 + }, + { + "epoch": 0.17801, + "grad_norm": 0.8976352144356766, + "learning_rate": 0.003, + "loss": 4.044, + "step": 17801 + }, + { + "epoch": 0.17802, + "grad_norm": 0.7843968269164642, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 17802 + }, + { + "epoch": 0.17803, + "grad_norm": 0.8210470078902311, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 17803 + }, + { + "epoch": 0.17804, + "grad_norm": 0.8192695195039678, + "learning_rate": 0.003, + "loss": 4.0819, + "step": 17804 + }, + { + "epoch": 0.17805, + "grad_norm": 0.8937685468818372, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 17805 + }, + { + "epoch": 0.17806, + "grad_norm": 1.004712423310537, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 17806 + }, + { + "epoch": 0.17807, + "grad_norm": 0.9739234285536098, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 17807 + }, + { + "epoch": 0.17808, + "grad_norm": 1.013316419788831, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 17808 + }, + { + "epoch": 0.17809, + "grad_norm": 1.3304290886820391, + "learning_rate": 0.003, + "loss": 4.09, + "step": 17809 + }, + { + "epoch": 0.1781, + "grad_norm": 0.9302629164377382, + "learning_rate": 0.003, + "loss": 4.054, + "step": 17810 + }, + { + "epoch": 0.17811, + "grad_norm": 0.9860362332645474, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 17811 + }, + { + "epoch": 0.17812, + "grad_norm": 1.3625527727187556, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 17812 + }, + { + "epoch": 0.17813, + "grad_norm": 0.885682354916934, + "learning_rate": 0.003, + "loss": 4.098, + "step": 17813 + }, + { + "epoch": 0.17814, + "grad_norm": 0.929712670200877, + "learning_rate": 0.003, + "loss": 4.082, + "step": 17814 + }, + { + "epoch": 0.17815, + "grad_norm": 1.0818968840157688, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 17815 + }, + { + "epoch": 0.17816, + "grad_norm": 1.0502349590262547, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 17816 + }, + { + "epoch": 0.17817, + "grad_norm": 1.0268988070562564, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 17817 + }, + { + "epoch": 0.17818, + "grad_norm": 1.0549192996280106, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 17818 + }, + { + "epoch": 0.17819, + "grad_norm": 1.099237327202239, + "learning_rate": 0.003, + "loss": 4.0871, + "step": 17819 + }, + { + "epoch": 0.1782, + "grad_norm": 1.0520405491770515, + "learning_rate": 0.003, + "loss": 4.089, + "step": 17820 + }, + { + "epoch": 0.17821, + "grad_norm": 1.2680224670507132, + "learning_rate": 0.003, + "loss": 4.0881, + "step": 17821 + }, + { + "epoch": 0.17822, + "grad_norm": 0.914965938656521, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 17822 + }, + { + "epoch": 0.17823, + "grad_norm": 0.9326951024131106, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 17823 + }, + { + "epoch": 0.17824, + "grad_norm": 1.1308858262090526, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 17824 + }, + { + "epoch": 0.17825, + "grad_norm": 0.9848591289600144, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 17825 + }, + { + "epoch": 0.17826, + "grad_norm": 0.988210425496162, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 17826 + }, + { + "epoch": 0.17827, + "grad_norm": 1.166381349388133, + "learning_rate": 0.003, + "loss": 4.08, + "step": 17827 + }, + { + "epoch": 0.17828, + "grad_norm": 0.8753340986392573, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 17828 + }, + { + "epoch": 0.17829, + "grad_norm": 0.9306863706953002, + "learning_rate": 0.003, + "loss": 4.07, + "step": 17829 + }, + { + "epoch": 0.1783, + "grad_norm": 1.0093616002461412, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 17830 + }, + { + "epoch": 0.17831, + "grad_norm": 1.1905015149492713, + "learning_rate": 0.003, + "loss": 4.0891, + "step": 17831 + }, + { + "epoch": 0.17832, + "grad_norm": 1.032751912045493, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 17832 + }, + { + "epoch": 0.17833, + "grad_norm": 1.1343903959242494, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 17833 + }, + { + "epoch": 0.17834, + "grad_norm": 1.133920787568314, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 17834 + }, + { + "epoch": 0.17835, + "grad_norm": 1.0165166819434492, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 17835 + }, + { + "epoch": 0.17836, + "grad_norm": 1.0100324537802907, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 17836 + }, + { + "epoch": 0.17837, + "grad_norm": 1.1961150247135557, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 17837 + }, + { + "epoch": 0.17838, + "grad_norm": 0.93693955479397, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 17838 + }, + { + "epoch": 0.17839, + "grad_norm": 0.9866716483122979, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 17839 + }, + { + "epoch": 0.1784, + "grad_norm": 0.9896915663179726, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 17840 + }, + { + "epoch": 0.17841, + "grad_norm": 1.121040186149045, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 17841 + }, + { + "epoch": 0.17842, + "grad_norm": 0.9370925146186201, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 17842 + }, + { + "epoch": 0.17843, + "grad_norm": 0.9921200585064218, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 17843 + }, + { + "epoch": 0.17844, + "grad_norm": 1.181858484123272, + "learning_rate": 0.003, + "loss": 4.11, + "step": 17844 + }, + { + "epoch": 0.17845, + "grad_norm": 1.0468827058793706, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 17845 + }, + { + "epoch": 0.17846, + "grad_norm": 1.0337474989183946, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 17846 + }, + { + "epoch": 0.17847, + "grad_norm": 1.1644704776407173, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 17847 + }, + { + "epoch": 0.17848, + "grad_norm": 0.9306252179444108, + "learning_rate": 0.003, + "loss": 4.07, + "step": 17848 + }, + { + "epoch": 0.17849, + "grad_norm": 1.1633250991739337, + "learning_rate": 0.003, + "loss": 4.1041, + "step": 17849 + }, + { + "epoch": 0.1785, + "grad_norm": 1.0460173701126316, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 17850 + }, + { + "epoch": 0.17851, + "grad_norm": 1.0496483191354153, + "learning_rate": 0.003, + "loss": 4.0991, + "step": 17851 + }, + { + "epoch": 0.17852, + "grad_norm": 1.0400900642142739, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 17852 + }, + { + "epoch": 0.17853, + "grad_norm": 1.0502822845930737, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 17853 + }, + { + "epoch": 0.17854, + "grad_norm": 1.075448091285323, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 17854 + }, + { + "epoch": 0.17855, + "grad_norm": 0.9246081782817402, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 17855 + }, + { + "epoch": 0.17856, + "grad_norm": 0.9690868441608264, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 17856 + }, + { + "epoch": 0.17857, + "grad_norm": 1.0572252873611134, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 17857 + }, + { + "epoch": 0.17858, + "grad_norm": 1.0010964713414423, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 17858 + }, + { + "epoch": 0.17859, + "grad_norm": 1.178562643416493, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 17859 + }, + { + "epoch": 0.1786, + "grad_norm": 1.068479989959543, + "learning_rate": 0.003, + "loss": 4.056, + "step": 17860 + }, + { + "epoch": 0.17861, + "grad_norm": 0.9608157722151996, + "learning_rate": 0.003, + "loss": 4.1012, + "step": 17861 + }, + { + "epoch": 0.17862, + "grad_norm": 1.148583636799406, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 17862 + }, + { + "epoch": 0.17863, + "grad_norm": 1.0129235571851296, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 17863 + }, + { + "epoch": 0.17864, + "grad_norm": 1.0284232793663282, + "learning_rate": 0.003, + "loss": 4.095, + "step": 17864 + }, + { + "epoch": 0.17865, + "grad_norm": 1.042156193429454, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 17865 + }, + { + "epoch": 0.17866, + "grad_norm": 1.129241467499866, + "learning_rate": 0.003, + "loss": 4.067, + "step": 17866 + }, + { + "epoch": 0.17867, + "grad_norm": 0.9071439789241961, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 17867 + }, + { + "epoch": 0.17868, + "grad_norm": 1.1075264496879136, + "learning_rate": 0.003, + "loss": 4.067, + "step": 17868 + }, + { + "epoch": 0.17869, + "grad_norm": 1.0485155959899803, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 17869 + }, + { + "epoch": 0.1787, + "grad_norm": 0.9131001851905814, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 17870 + }, + { + "epoch": 0.17871, + "grad_norm": 1.0102944575243078, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 17871 + }, + { + "epoch": 0.17872, + "grad_norm": 1.051501259695993, + "learning_rate": 0.003, + "loss": 4.054, + "step": 17872 + }, + { + "epoch": 0.17873, + "grad_norm": 1.0187508683721789, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 17873 + }, + { + "epoch": 0.17874, + "grad_norm": 1.0495253901799795, + "learning_rate": 0.003, + "loss": 4.063, + "step": 17874 + }, + { + "epoch": 0.17875, + "grad_norm": 0.9601716402492623, + "learning_rate": 0.003, + "loss": 4.096, + "step": 17875 + }, + { + "epoch": 0.17876, + "grad_norm": 1.0987957390488303, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 17876 + }, + { + "epoch": 0.17877, + "grad_norm": 0.9989590072292573, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 17877 + }, + { + "epoch": 0.17878, + "grad_norm": 0.972221642897546, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 17878 + }, + { + "epoch": 0.17879, + "grad_norm": 1.0252198800623218, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 17879 + }, + { + "epoch": 0.1788, + "grad_norm": 1.2502933673342704, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 17880 + }, + { + "epoch": 0.17881, + "grad_norm": 0.8941012966695218, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 17881 + }, + { + "epoch": 0.17882, + "grad_norm": 0.9562327060970383, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 17882 + }, + { + "epoch": 0.17883, + "grad_norm": 1.0719541652159068, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 17883 + }, + { + "epoch": 0.17884, + "grad_norm": 1.3487572905533791, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 17884 + }, + { + "epoch": 0.17885, + "grad_norm": 0.7280640453229087, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 17885 + }, + { + "epoch": 0.17886, + "grad_norm": 0.8278451643872338, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 17886 + }, + { + "epoch": 0.17887, + "grad_norm": 0.9677324824559796, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 17887 + }, + { + "epoch": 0.17888, + "grad_norm": 1.1078223997767713, + "learning_rate": 0.003, + "loss": 4.0876, + "step": 17888 + }, + { + "epoch": 0.17889, + "grad_norm": 0.8963349620016179, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 17889 + }, + { + "epoch": 0.1789, + "grad_norm": 0.9120392161019066, + "learning_rate": 0.003, + "loss": 4.0943, + "step": 17890 + }, + { + "epoch": 0.17891, + "grad_norm": 0.9740201800885981, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 17891 + }, + { + "epoch": 0.17892, + "grad_norm": 1.0361335385694996, + "learning_rate": 0.003, + "loss": 4.1047, + "step": 17892 + }, + { + "epoch": 0.17893, + "grad_norm": 1.0281711397463587, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 17893 + }, + { + "epoch": 0.17894, + "grad_norm": 0.9792771098048783, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 17894 + }, + { + "epoch": 0.17895, + "grad_norm": 0.9163501901276871, + "learning_rate": 0.003, + "loss": 4.0998, + "step": 17895 + }, + { + "epoch": 0.17896, + "grad_norm": 0.8328242194961628, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 17896 + }, + { + "epoch": 0.17897, + "grad_norm": 1.0043863387105492, + "learning_rate": 0.003, + "loss": 4.087, + "step": 17897 + }, + { + "epoch": 0.17898, + "grad_norm": 1.0760165838835616, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 17898 + }, + { + "epoch": 0.17899, + "grad_norm": 1.1550777021025278, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 17899 + }, + { + "epoch": 0.179, + "grad_norm": 1.1440704323917334, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 17900 + }, + { + "epoch": 0.17901, + "grad_norm": 0.9930517660076116, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 17901 + }, + { + "epoch": 0.17902, + "grad_norm": 0.967739396353277, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 17902 + }, + { + "epoch": 0.17903, + "grad_norm": 1.191521310377567, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 17903 + }, + { + "epoch": 0.17904, + "grad_norm": 0.9583245167237798, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 17904 + }, + { + "epoch": 0.17905, + "grad_norm": 1.0375511946659226, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 17905 + }, + { + "epoch": 0.17906, + "grad_norm": 1.113199476887078, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 17906 + }, + { + "epoch": 0.17907, + "grad_norm": 1.0730584488611583, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 17907 + }, + { + "epoch": 0.17908, + "grad_norm": 1.0657524221759531, + "learning_rate": 0.003, + "loss": 4.08, + "step": 17908 + }, + { + "epoch": 0.17909, + "grad_norm": 1.0320122990831833, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 17909 + }, + { + "epoch": 0.1791, + "grad_norm": 1.2985241765603914, + "learning_rate": 0.003, + "loss": 4.073, + "step": 17910 + }, + { + "epoch": 0.17911, + "grad_norm": 0.8047700486501976, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 17911 + }, + { + "epoch": 0.17912, + "grad_norm": 0.8016169956664643, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 17912 + }, + { + "epoch": 0.17913, + "grad_norm": 0.8575280590563721, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 17913 + }, + { + "epoch": 0.17914, + "grad_norm": 0.904253775349405, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 17914 + }, + { + "epoch": 0.17915, + "grad_norm": 1.070990251125325, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 17915 + }, + { + "epoch": 0.17916, + "grad_norm": 1.0651370421330626, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 17916 + }, + { + "epoch": 0.17917, + "grad_norm": 1.170271548767836, + "learning_rate": 0.003, + "loss": 4.0941, + "step": 17917 + }, + { + "epoch": 0.17918, + "grad_norm": 0.9351491712039274, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 17918 + }, + { + "epoch": 0.17919, + "grad_norm": 0.8466435900686171, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 17919 + }, + { + "epoch": 0.1792, + "grad_norm": 0.7977665430171612, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 17920 + }, + { + "epoch": 0.17921, + "grad_norm": 0.8793543918225755, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 17921 + }, + { + "epoch": 0.17922, + "grad_norm": 1.0124825993407154, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 17922 + }, + { + "epoch": 0.17923, + "grad_norm": 1.29066743389193, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 17923 + }, + { + "epoch": 0.17924, + "grad_norm": 0.703605725782825, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 17924 + }, + { + "epoch": 0.17925, + "grad_norm": 0.7313907853909201, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 17925 + }, + { + "epoch": 0.17926, + "grad_norm": 0.9546750192244967, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 17926 + }, + { + "epoch": 0.17927, + "grad_norm": 1.4515182558101798, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 17927 + }, + { + "epoch": 0.17928, + "grad_norm": 0.7440082530299986, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 17928 + }, + { + "epoch": 0.17929, + "grad_norm": 0.7108941415877245, + "learning_rate": 0.003, + "loss": 4.046, + "step": 17929 + }, + { + "epoch": 0.1793, + "grad_norm": 0.7754773873036243, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 17930 + }, + { + "epoch": 0.17931, + "grad_norm": 0.981267222422152, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 17931 + }, + { + "epoch": 0.17932, + "grad_norm": 1.0914852316707293, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 17932 + }, + { + "epoch": 0.17933, + "grad_norm": 1.1137314215385161, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 17933 + }, + { + "epoch": 0.17934, + "grad_norm": 1.1681662923654543, + "learning_rate": 0.003, + "loss": 4.0983, + "step": 17934 + }, + { + "epoch": 0.17935, + "grad_norm": 0.9166379121749517, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 17935 + }, + { + "epoch": 0.17936, + "grad_norm": 0.9264697383196641, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 17936 + }, + { + "epoch": 0.17937, + "grad_norm": 1.024736704784965, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 17937 + }, + { + "epoch": 0.17938, + "grad_norm": 1.2867197275577535, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 17938 + }, + { + "epoch": 0.17939, + "grad_norm": 0.7112418563271397, + "learning_rate": 0.003, + "loss": 4.078, + "step": 17939 + }, + { + "epoch": 0.1794, + "grad_norm": 0.593985809692867, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 17940 + }, + { + "epoch": 0.17941, + "grad_norm": 0.8738619259742181, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 17941 + }, + { + "epoch": 0.17942, + "grad_norm": 1.312958810847152, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 17942 + }, + { + "epoch": 0.17943, + "grad_norm": 0.8904618138261466, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 17943 + }, + { + "epoch": 0.17944, + "grad_norm": 0.8158876133510262, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 17944 + }, + { + "epoch": 0.17945, + "grad_norm": 0.9261976003832675, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 17945 + }, + { + "epoch": 0.17946, + "grad_norm": 0.9853658738796093, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 17946 + }, + { + "epoch": 0.17947, + "grad_norm": 0.9667794678059406, + "learning_rate": 0.003, + "loss": 4.056, + "step": 17947 + }, + { + "epoch": 0.17948, + "grad_norm": 1.0166160138076996, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 17948 + }, + { + "epoch": 0.17949, + "grad_norm": 1.4301913543246367, + "learning_rate": 0.003, + "loss": 4.0819, + "step": 17949 + }, + { + "epoch": 0.1795, + "grad_norm": 0.790546608653879, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 17950 + }, + { + "epoch": 0.17951, + "grad_norm": 0.8420410488044123, + "learning_rate": 0.003, + "loss": 4.054, + "step": 17951 + }, + { + "epoch": 0.17952, + "grad_norm": 1.005566128060179, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 17952 + }, + { + "epoch": 0.17953, + "grad_norm": 1.0889343252485184, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 17953 + }, + { + "epoch": 0.17954, + "grad_norm": 1.0430049362692269, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 17954 + }, + { + "epoch": 0.17955, + "grad_norm": 1.0269519007364034, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 17955 + }, + { + "epoch": 0.17956, + "grad_norm": 1.1043104579724807, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 17956 + }, + { + "epoch": 0.17957, + "grad_norm": 0.8890470147209761, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 17957 + }, + { + "epoch": 0.17958, + "grad_norm": 1.0273853553912045, + "learning_rate": 0.003, + "loss": 4.063, + "step": 17958 + }, + { + "epoch": 0.17959, + "grad_norm": 1.2856325777087365, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 17959 + }, + { + "epoch": 0.1796, + "grad_norm": 0.9905384534758195, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 17960 + }, + { + "epoch": 0.17961, + "grad_norm": 1.3273502206594483, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 17961 + }, + { + "epoch": 0.17962, + "grad_norm": 0.8851544671100873, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 17962 + }, + { + "epoch": 0.17963, + "grad_norm": 0.8797978196115954, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 17963 + }, + { + "epoch": 0.17964, + "grad_norm": 0.9791345709511954, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 17964 + }, + { + "epoch": 0.17965, + "grad_norm": 0.9831985393542777, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 17965 + }, + { + "epoch": 0.17966, + "grad_norm": 1.05395618338193, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 17966 + }, + { + "epoch": 0.17967, + "grad_norm": 1.2036230992464312, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 17967 + }, + { + "epoch": 0.17968, + "grad_norm": 0.9432805108109641, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 17968 + }, + { + "epoch": 0.17969, + "grad_norm": 0.9346933606588981, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 17969 + }, + { + "epoch": 0.1797, + "grad_norm": 0.936046581883216, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 17970 + }, + { + "epoch": 0.17971, + "grad_norm": 1.0748954074887904, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 17971 + }, + { + "epoch": 0.17972, + "grad_norm": 1.0776452139306145, + "learning_rate": 0.003, + "loss": 4.0944, + "step": 17972 + }, + { + "epoch": 0.17973, + "grad_norm": 1.0948718378054145, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 17973 + }, + { + "epoch": 0.17974, + "grad_norm": 1.2067397001143592, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 17974 + }, + { + "epoch": 0.17975, + "grad_norm": 0.8788918257809932, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 17975 + }, + { + "epoch": 0.17976, + "grad_norm": 0.7949383319057617, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 17976 + }, + { + "epoch": 0.17977, + "grad_norm": 0.9173349568424838, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 17977 + }, + { + "epoch": 0.17978, + "grad_norm": 0.9984270269933427, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 17978 + }, + { + "epoch": 0.17979, + "grad_norm": 1.0627665857863042, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 17979 + }, + { + "epoch": 0.1798, + "grad_norm": 1.0117210215886252, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 17980 + }, + { + "epoch": 0.17981, + "grad_norm": 0.947856707130065, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 17981 + }, + { + "epoch": 0.17982, + "grad_norm": 1.0994375557799472, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 17982 + }, + { + "epoch": 0.17983, + "grad_norm": 1.1313181065393143, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 17983 + }, + { + "epoch": 0.17984, + "grad_norm": 1.1647195164568591, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 17984 + }, + { + "epoch": 0.17985, + "grad_norm": 1.02767546751103, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 17985 + }, + { + "epoch": 0.17986, + "grad_norm": 1.0203663974503845, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 17986 + }, + { + "epoch": 0.17987, + "grad_norm": 1.1908831710803043, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 17987 + }, + { + "epoch": 0.17988, + "grad_norm": 0.7613958973838513, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 17988 + }, + { + "epoch": 0.17989, + "grad_norm": 0.7209069547995978, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 17989 + }, + { + "epoch": 0.1799, + "grad_norm": 0.7902514260521344, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 17990 + }, + { + "epoch": 0.17991, + "grad_norm": 0.874580386672636, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 17991 + }, + { + "epoch": 0.17992, + "grad_norm": 0.9943418221637598, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 17992 + }, + { + "epoch": 0.17993, + "grad_norm": 1.1361559930745488, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 17993 + }, + { + "epoch": 0.17994, + "grad_norm": 1.0993722043349115, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 17994 + }, + { + "epoch": 0.17995, + "grad_norm": 1.1286893367725095, + "learning_rate": 0.003, + "loss": 4.0975, + "step": 17995 + }, + { + "epoch": 0.17996, + "grad_norm": 1.0749468073244492, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 17996 + }, + { + "epoch": 0.17997, + "grad_norm": 1.012247526166339, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 17997 + }, + { + "epoch": 0.17998, + "grad_norm": 1.1561270538881523, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 17998 + }, + { + "epoch": 0.17999, + "grad_norm": 1.001500500816165, + "learning_rate": 0.003, + "loss": 4.081, + "step": 17999 + }, + { + "epoch": 0.18, + "grad_norm": 1.084169492035819, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 18000 + }, + { + "epoch": 0.18001, + "grad_norm": 1.157332802305126, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 18001 + }, + { + "epoch": 0.18002, + "grad_norm": 1.1542609429528623, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 18002 + }, + { + "epoch": 0.18003, + "grad_norm": 1.0720407096994307, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 18003 + }, + { + "epoch": 0.18004, + "grad_norm": 1.1013391965484167, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 18004 + }, + { + "epoch": 0.18005, + "grad_norm": 0.8607850992121242, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 18005 + }, + { + "epoch": 0.18006, + "grad_norm": 0.7056229946272707, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 18006 + }, + { + "epoch": 0.18007, + "grad_norm": 0.9372490390893289, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 18007 + }, + { + "epoch": 0.18008, + "grad_norm": 1.2036248224883062, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 18008 + }, + { + "epoch": 0.18009, + "grad_norm": 0.8493164470305848, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 18009 + }, + { + "epoch": 0.1801, + "grad_norm": 0.8931251893200244, + "learning_rate": 0.003, + "loss": 4.059, + "step": 18010 + }, + { + "epoch": 0.18011, + "grad_norm": 0.8562772347849715, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 18011 + }, + { + "epoch": 0.18012, + "grad_norm": 0.8921464327688934, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 18012 + }, + { + "epoch": 0.18013, + "grad_norm": 1.1498452459535395, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 18013 + }, + { + "epoch": 0.18014, + "grad_norm": 1.0311052683594082, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 18014 + }, + { + "epoch": 0.18015, + "grad_norm": 1.02203746517564, + "learning_rate": 0.003, + "loss": 4.1142, + "step": 18015 + }, + { + "epoch": 0.18016, + "grad_norm": 1.2226754107844533, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 18016 + }, + { + "epoch": 0.18017, + "grad_norm": 1.0153187447220673, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 18017 + }, + { + "epoch": 0.18018, + "grad_norm": 1.3194620562647383, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 18018 + }, + { + "epoch": 0.18019, + "grad_norm": 0.8629818705241493, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 18019 + }, + { + "epoch": 0.1802, + "grad_norm": 0.9647365916555045, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 18020 + }, + { + "epoch": 0.18021, + "grad_norm": 0.989820843405732, + "learning_rate": 0.003, + "loss": 4.0923, + "step": 18021 + }, + { + "epoch": 0.18022, + "grad_norm": 1.1671199229330438, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 18022 + }, + { + "epoch": 0.18023, + "grad_norm": 1.2234982426115364, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 18023 + }, + { + "epoch": 0.18024, + "grad_norm": 0.9621810113550203, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 18024 + }, + { + "epoch": 0.18025, + "grad_norm": 1.0283947560108164, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 18025 + }, + { + "epoch": 0.18026, + "grad_norm": 1.1257815534508089, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 18026 + }, + { + "epoch": 0.18027, + "grad_norm": 0.9599797148118149, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 18027 + }, + { + "epoch": 0.18028, + "grad_norm": 1.0865483155430586, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 18028 + }, + { + "epoch": 0.18029, + "grad_norm": 0.9826798638927311, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 18029 + }, + { + "epoch": 0.1803, + "grad_norm": 0.9788239017496804, + "learning_rate": 0.003, + "loss": 4.0938, + "step": 18030 + }, + { + "epoch": 0.18031, + "grad_norm": 1.0394624291263654, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 18031 + }, + { + "epoch": 0.18032, + "grad_norm": 1.1925609309496545, + "learning_rate": 0.003, + "loss": 4.1315, + "step": 18032 + }, + { + "epoch": 0.18033, + "grad_norm": 1.0079930289286658, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 18033 + }, + { + "epoch": 0.18034, + "grad_norm": 1.0684597914647542, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 18034 + }, + { + "epoch": 0.18035, + "grad_norm": 1.1239525955418117, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 18035 + }, + { + "epoch": 0.18036, + "grad_norm": 0.933789897761682, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 18036 + }, + { + "epoch": 0.18037, + "grad_norm": 1.0840242879739415, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 18037 + }, + { + "epoch": 0.18038, + "grad_norm": 1.055740522278367, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 18038 + }, + { + "epoch": 0.18039, + "grad_norm": 1.0171734030739439, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 18039 + }, + { + "epoch": 0.1804, + "grad_norm": 0.9588143173092722, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 18040 + }, + { + "epoch": 0.18041, + "grad_norm": 0.9663655816502914, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 18041 + }, + { + "epoch": 0.18042, + "grad_norm": 1.1906622123720898, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 18042 + }, + { + "epoch": 0.18043, + "grad_norm": 0.9982129588398057, + "learning_rate": 0.003, + "loss": 4.071, + "step": 18043 + }, + { + "epoch": 0.18044, + "grad_norm": 1.002452724977036, + "learning_rate": 0.003, + "loss": 4.1072, + "step": 18044 + }, + { + "epoch": 0.18045, + "grad_norm": 1.1744337502692606, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 18045 + }, + { + "epoch": 0.18046, + "grad_norm": 0.805281176561988, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 18046 + }, + { + "epoch": 0.18047, + "grad_norm": 0.8459115827457988, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 18047 + }, + { + "epoch": 0.18048, + "grad_norm": 0.8626919310661322, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 18048 + }, + { + "epoch": 0.18049, + "grad_norm": 0.8933032597805142, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 18049 + }, + { + "epoch": 0.1805, + "grad_norm": 1.0080683335188956, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 18050 + }, + { + "epoch": 0.18051, + "grad_norm": 0.8801479827633547, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 18051 + }, + { + "epoch": 0.18052, + "grad_norm": 1.0733027237379658, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 18052 + }, + { + "epoch": 0.18053, + "grad_norm": 1.299744541885555, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 18053 + }, + { + "epoch": 0.18054, + "grad_norm": 0.9926902867715208, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 18054 + }, + { + "epoch": 0.18055, + "grad_norm": 1.2528109816350064, + "learning_rate": 0.003, + "loss": 4.0878, + "step": 18055 + }, + { + "epoch": 0.18056, + "grad_norm": 1.1043374977284395, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 18056 + }, + { + "epoch": 0.18057, + "grad_norm": 1.1288434411539063, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 18057 + }, + { + "epoch": 0.18058, + "grad_norm": 0.884696668999549, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 18058 + }, + { + "epoch": 0.18059, + "grad_norm": 0.9852969513788737, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 18059 + }, + { + "epoch": 0.1806, + "grad_norm": 1.1509132920271576, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 18060 + }, + { + "epoch": 0.18061, + "grad_norm": 1.032391255358663, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 18061 + }, + { + "epoch": 0.18062, + "grad_norm": 1.275663592530344, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 18062 + }, + { + "epoch": 0.18063, + "grad_norm": 0.9039855148877469, + "learning_rate": 0.003, + "loss": 4.054, + "step": 18063 + }, + { + "epoch": 0.18064, + "grad_norm": 0.7723304713130585, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 18064 + }, + { + "epoch": 0.18065, + "grad_norm": 0.8434014840597956, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 18065 + }, + { + "epoch": 0.18066, + "grad_norm": 0.9874523162177533, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 18066 + }, + { + "epoch": 0.18067, + "grad_norm": 1.1256845113050378, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 18067 + }, + { + "epoch": 0.18068, + "grad_norm": 0.974090252760411, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 18068 + }, + { + "epoch": 0.18069, + "grad_norm": 1.0414792539580435, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 18069 + }, + { + "epoch": 0.1807, + "grad_norm": 1.1957832700044377, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 18070 + }, + { + "epoch": 0.18071, + "grad_norm": 1.1366416697936328, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 18071 + }, + { + "epoch": 0.18072, + "grad_norm": 0.8244779858521649, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 18072 + }, + { + "epoch": 0.18073, + "grad_norm": 0.736193972514852, + "learning_rate": 0.003, + "loss": 4.1001, + "step": 18073 + }, + { + "epoch": 0.18074, + "grad_norm": 0.8671350088767975, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 18074 + }, + { + "epoch": 0.18075, + "grad_norm": 0.9256618098869503, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 18075 + }, + { + "epoch": 0.18076, + "grad_norm": 0.930725678298883, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 18076 + }, + { + "epoch": 0.18077, + "grad_norm": 1.0558023434956116, + "learning_rate": 0.003, + "loss": 4.0858, + "step": 18077 + }, + { + "epoch": 0.18078, + "grad_norm": 1.1545075382746446, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 18078 + }, + { + "epoch": 0.18079, + "grad_norm": 1.0455709352888531, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 18079 + }, + { + "epoch": 0.1808, + "grad_norm": 1.2340408607780342, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 18080 + }, + { + "epoch": 0.18081, + "grad_norm": 0.9427188698628358, + "learning_rate": 0.003, + "loss": 4.0991, + "step": 18081 + }, + { + "epoch": 0.18082, + "grad_norm": 0.8074080765426477, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 18082 + }, + { + "epoch": 0.18083, + "grad_norm": 0.8964275495951045, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 18083 + }, + { + "epoch": 0.18084, + "grad_norm": 1.0353756940850554, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 18084 + }, + { + "epoch": 0.18085, + "grad_norm": 1.4005559671860468, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 18085 + }, + { + "epoch": 0.18086, + "grad_norm": 0.789213575013521, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 18086 + }, + { + "epoch": 0.18087, + "grad_norm": 0.7868082354374315, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 18087 + }, + { + "epoch": 0.18088, + "grad_norm": 0.9549445846340786, + "learning_rate": 0.003, + "loss": 4.072, + "step": 18088 + }, + { + "epoch": 0.18089, + "grad_norm": 1.1532409673422759, + "learning_rate": 0.003, + "loss": 4.043, + "step": 18089 + }, + { + "epoch": 0.1809, + "grad_norm": 1.1900216601377018, + "learning_rate": 0.003, + "loss": 4.09, + "step": 18090 + }, + { + "epoch": 0.18091, + "grad_norm": 0.9939755386023241, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 18091 + }, + { + "epoch": 0.18092, + "grad_norm": 1.0650430267892692, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 18092 + }, + { + "epoch": 0.18093, + "grad_norm": 0.9958006402117859, + "learning_rate": 0.003, + "loss": 4.0864, + "step": 18093 + }, + { + "epoch": 0.18094, + "grad_norm": 1.0613866978855453, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 18094 + }, + { + "epoch": 0.18095, + "grad_norm": 0.9060053979631656, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 18095 + }, + { + "epoch": 0.18096, + "grad_norm": 0.8730435485643874, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 18096 + }, + { + "epoch": 0.18097, + "grad_norm": 1.0511759231658397, + "learning_rate": 0.003, + "loss": 4.062, + "step": 18097 + }, + { + "epoch": 0.18098, + "grad_norm": 1.0326942678180235, + "learning_rate": 0.003, + "loss": 4.064, + "step": 18098 + }, + { + "epoch": 0.18099, + "grad_norm": 1.0621143822319237, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 18099 + }, + { + "epoch": 0.181, + "grad_norm": 1.0510654810834925, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 18100 + }, + { + "epoch": 0.18101, + "grad_norm": 1.1007285485055618, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 18101 + }, + { + "epoch": 0.18102, + "grad_norm": 0.9546278562170247, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 18102 + }, + { + "epoch": 0.18103, + "grad_norm": 1.0150766559112359, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 18103 + }, + { + "epoch": 0.18104, + "grad_norm": 1.1260284669500644, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 18104 + }, + { + "epoch": 0.18105, + "grad_norm": 0.920921370197899, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 18105 + }, + { + "epoch": 0.18106, + "grad_norm": 0.9752469759732258, + "learning_rate": 0.003, + "loss": 4.065, + "step": 18106 + }, + { + "epoch": 0.18107, + "grad_norm": 1.0248049081864017, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 18107 + }, + { + "epoch": 0.18108, + "grad_norm": 1.033747730737616, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 18108 + }, + { + "epoch": 0.18109, + "grad_norm": 1.041840023755259, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 18109 + }, + { + "epoch": 0.1811, + "grad_norm": 1.119016022543674, + "learning_rate": 0.003, + "loss": 4.1023, + "step": 18110 + }, + { + "epoch": 0.18111, + "grad_norm": 1.0484450695729495, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 18111 + }, + { + "epoch": 0.18112, + "grad_norm": 1.1873939308189874, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 18112 + }, + { + "epoch": 0.18113, + "grad_norm": 1.1310532319092064, + "learning_rate": 0.003, + "loss": 4.1029, + "step": 18113 + }, + { + "epoch": 0.18114, + "grad_norm": 1.149826983679774, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 18114 + }, + { + "epoch": 0.18115, + "grad_norm": 1.0880032392093062, + "learning_rate": 0.003, + "loss": 4.05, + "step": 18115 + }, + { + "epoch": 0.18116, + "grad_norm": 1.2764278766951527, + "learning_rate": 0.003, + "loss": 4.1062, + "step": 18116 + }, + { + "epoch": 0.18117, + "grad_norm": 0.8606357279788343, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 18117 + }, + { + "epoch": 0.18118, + "grad_norm": 0.9123554728486016, + "learning_rate": 0.003, + "loss": 4.064, + "step": 18118 + }, + { + "epoch": 0.18119, + "grad_norm": 1.0098279106776524, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 18119 + }, + { + "epoch": 0.1812, + "grad_norm": 1.3891339656647526, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 18120 + }, + { + "epoch": 0.18121, + "grad_norm": 0.9851769694719602, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 18121 + }, + { + "epoch": 0.18122, + "grad_norm": 0.926312559079871, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 18122 + }, + { + "epoch": 0.18123, + "grad_norm": 0.9482692134892446, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 18123 + }, + { + "epoch": 0.18124, + "grad_norm": 0.9621282586016507, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 18124 + }, + { + "epoch": 0.18125, + "grad_norm": 1.0859716303059248, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 18125 + }, + { + "epoch": 0.18126, + "grad_norm": 1.0346902814066081, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 18126 + }, + { + "epoch": 0.18127, + "grad_norm": 1.0069062509925983, + "learning_rate": 0.003, + "loss": 4.0956, + "step": 18127 + }, + { + "epoch": 0.18128, + "grad_norm": 1.0630084217721891, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 18128 + }, + { + "epoch": 0.18129, + "grad_norm": 1.1312489406846613, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 18129 + }, + { + "epoch": 0.1813, + "grad_norm": 0.9656396438926711, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 18130 + }, + { + "epoch": 0.18131, + "grad_norm": 0.9480415376515956, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 18131 + }, + { + "epoch": 0.18132, + "grad_norm": 0.8971282281093789, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 18132 + }, + { + "epoch": 0.18133, + "grad_norm": 0.9313407183316552, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 18133 + }, + { + "epoch": 0.18134, + "grad_norm": 1.162112901889005, + "learning_rate": 0.003, + "loss": 4.0934, + "step": 18134 + }, + { + "epoch": 0.18135, + "grad_norm": 1.1391437081838462, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 18135 + }, + { + "epoch": 0.18136, + "grad_norm": 0.9374879878029347, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 18136 + }, + { + "epoch": 0.18137, + "grad_norm": 0.9728122988703063, + "learning_rate": 0.003, + "loss": 4.0931, + "step": 18137 + }, + { + "epoch": 0.18138, + "grad_norm": 1.1645750749397854, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 18138 + }, + { + "epoch": 0.18139, + "grad_norm": 1.3524445777136922, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 18139 + }, + { + "epoch": 0.1814, + "grad_norm": 0.902403542667377, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 18140 + }, + { + "epoch": 0.18141, + "grad_norm": 0.8633392453307539, + "learning_rate": 0.003, + "loss": 4.055, + "step": 18141 + }, + { + "epoch": 0.18142, + "grad_norm": 0.8655899689459237, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 18142 + }, + { + "epoch": 0.18143, + "grad_norm": 0.9830098568671173, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 18143 + }, + { + "epoch": 0.18144, + "grad_norm": 1.2648797821843614, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 18144 + }, + { + "epoch": 0.18145, + "grad_norm": 0.8681932066359269, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 18145 + }, + { + "epoch": 0.18146, + "grad_norm": 0.972616769710075, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 18146 + }, + { + "epoch": 0.18147, + "grad_norm": 1.0174334411395036, + "learning_rate": 0.003, + "loss": 4.061, + "step": 18147 + }, + { + "epoch": 0.18148, + "grad_norm": 0.9548936971733968, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 18148 + }, + { + "epoch": 0.18149, + "grad_norm": 0.9784657612102272, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 18149 + }, + { + "epoch": 0.1815, + "grad_norm": 0.9706401311465016, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 18150 + }, + { + "epoch": 0.18151, + "grad_norm": 1.2084528559977497, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 18151 + }, + { + "epoch": 0.18152, + "grad_norm": 0.9193963147197685, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 18152 + }, + { + "epoch": 0.18153, + "grad_norm": 0.8213399007886859, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 18153 + }, + { + "epoch": 0.18154, + "grad_norm": 0.8778234253425335, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 18154 + }, + { + "epoch": 0.18155, + "grad_norm": 1.06124800530961, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 18155 + }, + { + "epoch": 0.18156, + "grad_norm": 1.1018982183563155, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 18156 + }, + { + "epoch": 0.18157, + "grad_norm": 1.0142820813782487, + "learning_rate": 0.003, + "loss": 4.073, + "step": 18157 + }, + { + "epoch": 0.18158, + "grad_norm": 1.0942231181022573, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 18158 + }, + { + "epoch": 0.18159, + "grad_norm": 0.8377943117103026, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 18159 + }, + { + "epoch": 0.1816, + "grad_norm": 0.784619028999677, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 18160 + }, + { + "epoch": 0.18161, + "grad_norm": 0.8101418991115407, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 18161 + }, + { + "epoch": 0.18162, + "grad_norm": 0.9314363867009436, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 18162 + }, + { + "epoch": 0.18163, + "grad_norm": 1.0220797087209432, + "learning_rate": 0.003, + "loss": 4.061, + "step": 18163 + }, + { + "epoch": 0.18164, + "grad_norm": 1.036868719634104, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 18164 + }, + { + "epoch": 0.18165, + "grad_norm": 1.1502905631633635, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 18165 + }, + { + "epoch": 0.18166, + "grad_norm": 0.8923177039712231, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 18166 + }, + { + "epoch": 0.18167, + "grad_norm": 1.0840389591360127, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 18167 + }, + { + "epoch": 0.18168, + "grad_norm": 1.2520165487484198, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 18168 + }, + { + "epoch": 0.18169, + "grad_norm": 1.0655041679117274, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 18169 + }, + { + "epoch": 0.1817, + "grad_norm": 1.109685866918144, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 18170 + }, + { + "epoch": 0.18171, + "grad_norm": 1.1634026535323296, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 18171 + }, + { + "epoch": 0.18172, + "grad_norm": 0.9309765333431995, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 18172 + }, + { + "epoch": 0.18173, + "grad_norm": 0.9101911908847053, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 18173 + }, + { + "epoch": 0.18174, + "grad_norm": 0.8884425822837404, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 18174 + }, + { + "epoch": 0.18175, + "grad_norm": 1.0988023418408224, + "learning_rate": 0.003, + "loss": 4.0903, + "step": 18175 + }, + { + "epoch": 0.18176, + "grad_norm": 1.213378518816908, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 18176 + }, + { + "epoch": 0.18177, + "grad_norm": 0.9356678790179462, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 18177 + }, + { + "epoch": 0.18178, + "grad_norm": 1.1289804773705714, + "learning_rate": 0.003, + "loss": 4.0906, + "step": 18178 + }, + { + "epoch": 0.18179, + "grad_norm": 1.1740311615200492, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 18179 + }, + { + "epoch": 0.1818, + "grad_norm": 1.0396255845733944, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 18180 + }, + { + "epoch": 0.18181, + "grad_norm": 1.1009256061767692, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 18181 + }, + { + "epoch": 0.18182, + "grad_norm": 1.0724226591293755, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 18182 + }, + { + "epoch": 0.18183, + "grad_norm": 1.1944156351902442, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 18183 + }, + { + "epoch": 0.18184, + "grad_norm": 0.9384874458953645, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 18184 + }, + { + "epoch": 0.18185, + "grad_norm": 1.0847958661736976, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 18185 + }, + { + "epoch": 0.18186, + "grad_norm": 1.0482725992207451, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 18186 + }, + { + "epoch": 0.18187, + "grad_norm": 1.0687295297641008, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 18187 + }, + { + "epoch": 0.18188, + "grad_norm": 0.8912928361240524, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 18188 + }, + { + "epoch": 0.18189, + "grad_norm": 0.9706848962604859, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 18189 + }, + { + "epoch": 0.1819, + "grad_norm": 1.0589632234545026, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 18190 + }, + { + "epoch": 0.18191, + "grad_norm": 1.0736309990208415, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 18191 + }, + { + "epoch": 0.18192, + "grad_norm": 0.9674265890471516, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 18192 + }, + { + "epoch": 0.18193, + "grad_norm": 1.1460608109056816, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 18193 + }, + { + "epoch": 0.18194, + "grad_norm": 0.9020589867884552, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 18194 + }, + { + "epoch": 0.18195, + "grad_norm": 0.8683288693754316, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 18195 + }, + { + "epoch": 0.18196, + "grad_norm": 0.8610165792973388, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 18196 + }, + { + "epoch": 0.18197, + "grad_norm": 0.9775492151926354, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 18197 + }, + { + "epoch": 0.18198, + "grad_norm": 1.2628170592491499, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 18198 + }, + { + "epoch": 0.18199, + "grad_norm": 0.9657950825623457, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 18199 + }, + { + "epoch": 0.182, + "grad_norm": 1.0414851471889592, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 18200 + }, + { + "epoch": 0.18201, + "grad_norm": 1.1852939151534179, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 18201 + }, + { + "epoch": 0.18202, + "grad_norm": 1.1081751111243061, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 18202 + }, + { + "epoch": 0.18203, + "grad_norm": 1.051438550226014, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 18203 + }, + { + "epoch": 0.18204, + "grad_norm": 1.1024983332434195, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 18204 + }, + { + "epoch": 0.18205, + "grad_norm": 1.0402212802671216, + "learning_rate": 0.003, + "loss": 4.033, + "step": 18205 + }, + { + "epoch": 0.18206, + "grad_norm": 0.9800800874216995, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 18206 + }, + { + "epoch": 0.18207, + "grad_norm": 0.9756825571245938, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 18207 + }, + { + "epoch": 0.18208, + "grad_norm": 1.1534596148925382, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 18208 + }, + { + "epoch": 0.18209, + "grad_norm": 0.9389736845121328, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 18209 + }, + { + "epoch": 0.1821, + "grad_norm": 1.0352101997646692, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 18210 + }, + { + "epoch": 0.18211, + "grad_norm": 1.03698143553655, + "learning_rate": 0.003, + "loss": 4.051, + "step": 18211 + }, + { + "epoch": 0.18212, + "grad_norm": 0.9762039176348934, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 18212 + }, + { + "epoch": 0.18213, + "grad_norm": 0.9857097746277337, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 18213 + }, + { + "epoch": 0.18214, + "grad_norm": 0.9896847291981593, + "learning_rate": 0.003, + "loss": 4.047, + "step": 18214 + }, + { + "epoch": 0.18215, + "grad_norm": 0.9711135638331198, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 18215 + }, + { + "epoch": 0.18216, + "grad_norm": 1.0157907964942883, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 18216 + }, + { + "epoch": 0.18217, + "grad_norm": 1.0126233589421034, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 18217 + }, + { + "epoch": 0.18218, + "grad_norm": 1.074366240842281, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 18218 + }, + { + "epoch": 0.18219, + "grad_norm": 1.068218630156494, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 18219 + }, + { + "epoch": 0.1822, + "grad_norm": 1.1904259104852517, + "learning_rate": 0.003, + "loss": 4.0873, + "step": 18220 + }, + { + "epoch": 0.18221, + "grad_norm": 0.9994295730138422, + "learning_rate": 0.003, + "loss": 4.0935, + "step": 18221 + }, + { + "epoch": 0.18222, + "grad_norm": 1.196224852626279, + "learning_rate": 0.003, + "loss": 4.0888, + "step": 18222 + }, + { + "epoch": 0.18223, + "grad_norm": 1.080684261934818, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 18223 + }, + { + "epoch": 0.18224, + "grad_norm": 1.025005656781295, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 18224 + }, + { + "epoch": 0.18225, + "grad_norm": 1.1000400815609939, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 18225 + }, + { + "epoch": 0.18226, + "grad_norm": 1.2693655042889926, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 18226 + }, + { + "epoch": 0.18227, + "grad_norm": 0.9395370866263622, + "learning_rate": 0.003, + "loss": 4.0912, + "step": 18227 + }, + { + "epoch": 0.18228, + "grad_norm": 0.8035165872998117, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 18228 + }, + { + "epoch": 0.18229, + "grad_norm": 0.7929080444312981, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 18229 + }, + { + "epoch": 0.1823, + "grad_norm": 0.926104780354193, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 18230 + }, + { + "epoch": 0.18231, + "grad_norm": 1.2972162167929158, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 18231 + }, + { + "epoch": 0.18232, + "grad_norm": 0.88266944138413, + "learning_rate": 0.003, + "loss": 4.068, + "step": 18232 + }, + { + "epoch": 0.18233, + "grad_norm": 0.9453537966469904, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 18233 + }, + { + "epoch": 0.18234, + "grad_norm": 1.007441976215157, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 18234 + }, + { + "epoch": 0.18235, + "grad_norm": 1.1205433138076473, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 18235 + }, + { + "epoch": 0.18236, + "grad_norm": 1.0116292686189725, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 18236 + }, + { + "epoch": 0.18237, + "grad_norm": 1.0445129270230764, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 18237 + }, + { + "epoch": 0.18238, + "grad_norm": 1.2926934943462762, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 18238 + }, + { + "epoch": 0.18239, + "grad_norm": 1.1441179958734502, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 18239 + }, + { + "epoch": 0.1824, + "grad_norm": 0.8515576955363109, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 18240 + }, + { + "epoch": 0.18241, + "grad_norm": 0.8092568137333931, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 18241 + }, + { + "epoch": 0.18242, + "grad_norm": 0.9688372169580559, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 18242 + }, + { + "epoch": 0.18243, + "grad_norm": 1.1721786658777427, + "learning_rate": 0.003, + "loss": 4.036, + "step": 18243 + }, + { + "epoch": 0.18244, + "grad_norm": 0.9030300222559061, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 18244 + }, + { + "epoch": 0.18245, + "grad_norm": 0.9001685786397269, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 18245 + }, + { + "epoch": 0.18246, + "grad_norm": 1.0503427581127769, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 18246 + }, + { + "epoch": 0.18247, + "grad_norm": 1.1077172921985285, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 18247 + }, + { + "epoch": 0.18248, + "grad_norm": 1.0240904519509122, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 18248 + }, + { + "epoch": 0.18249, + "grad_norm": 1.1152834925301305, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 18249 + }, + { + "epoch": 0.1825, + "grad_norm": 1.1928967327597875, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 18250 + }, + { + "epoch": 0.18251, + "grad_norm": 1.0138322530216233, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 18251 + }, + { + "epoch": 0.18252, + "grad_norm": 0.9176916637175803, + "learning_rate": 0.003, + "loss": 4.0968, + "step": 18252 + }, + { + "epoch": 0.18253, + "grad_norm": 0.8212330860467062, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 18253 + }, + { + "epoch": 0.18254, + "grad_norm": 0.7990286009235718, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 18254 + }, + { + "epoch": 0.18255, + "grad_norm": 0.9183064160300024, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 18255 + }, + { + "epoch": 0.18256, + "grad_norm": 0.867822812256223, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 18256 + }, + { + "epoch": 0.18257, + "grad_norm": 1.0462137730122982, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 18257 + }, + { + "epoch": 0.18258, + "grad_norm": 1.1768443764488872, + "learning_rate": 0.003, + "loss": 4.0864, + "step": 18258 + }, + { + "epoch": 0.18259, + "grad_norm": 0.8713564424635385, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 18259 + }, + { + "epoch": 0.1826, + "grad_norm": 0.870814764112766, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 18260 + }, + { + "epoch": 0.18261, + "grad_norm": 0.8967847345166235, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 18261 + }, + { + "epoch": 0.18262, + "grad_norm": 1.0505119511837913, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 18262 + }, + { + "epoch": 0.18263, + "grad_norm": 1.6873478177832761, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 18263 + }, + { + "epoch": 0.18264, + "grad_norm": 0.6892322903945147, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 18264 + }, + { + "epoch": 0.18265, + "grad_norm": 0.9343368053750315, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 18265 + }, + { + "epoch": 0.18266, + "grad_norm": 1.4920218210867104, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 18266 + }, + { + "epoch": 0.18267, + "grad_norm": 0.7643511362747499, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 18267 + }, + { + "epoch": 0.18268, + "grad_norm": 1.0272388631077953, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 18268 + }, + { + "epoch": 0.18269, + "grad_norm": 1.191528675149996, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 18269 + }, + { + "epoch": 0.1827, + "grad_norm": 1.0215737923728314, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 18270 + }, + { + "epoch": 0.18271, + "grad_norm": 0.9776904786948919, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 18271 + }, + { + "epoch": 0.18272, + "grad_norm": 0.8232590278722768, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 18272 + }, + { + "epoch": 0.18273, + "grad_norm": 0.9017083691880066, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 18273 + }, + { + "epoch": 0.18274, + "grad_norm": 1.246617309241457, + "learning_rate": 0.003, + "loss": 4.0919, + "step": 18274 + }, + { + "epoch": 0.18275, + "grad_norm": 0.9213658060103813, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 18275 + }, + { + "epoch": 0.18276, + "grad_norm": 1.0036485026088975, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 18276 + }, + { + "epoch": 0.18277, + "grad_norm": 1.3088781750707865, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 18277 + }, + { + "epoch": 0.18278, + "grad_norm": 0.979547569491654, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 18278 + }, + { + "epoch": 0.18279, + "grad_norm": 1.2007509825667384, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 18279 + }, + { + "epoch": 0.1828, + "grad_norm": 1.0200481314142245, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 18280 + }, + { + "epoch": 0.18281, + "grad_norm": 0.9854094696797915, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 18281 + }, + { + "epoch": 0.18282, + "grad_norm": 1.0100164441163, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 18282 + }, + { + "epoch": 0.18283, + "grad_norm": 1.1385991107442914, + "learning_rate": 0.003, + "loss": 4.0872, + "step": 18283 + }, + { + "epoch": 0.18284, + "grad_norm": 0.9267930475457726, + "learning_rate": 0.003, + "loss": 4.0925, + "step": 18284 + }, + { + "epoch": 0.18285, + "grad_norm": 1.1140373064449596, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 18285 + }, + { + "epoch": 0.18286, + "grad_norm": 1.0244058015654887, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 18286 + }, + { + "epoch": 0.18287, + "grad_norm": 1.1877878424992734, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 18287 + }, + { + "epoch": 0.18288, + "grad_norm": 0.8991253374572491, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 18288 + }, + { + "epoch": 0.18289, + "grad_norm": 1.028566242179825, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 18289 + }, + { + "epoch": 0.1829, + "grad_norm": 0.9909622599542173, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 18290 + }, + { + "epoch": 0.18291, + "grad_norm": 1.094698207105018, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 18291 + }, + { + "epoch": 0.18292, + "grad_norm": 0.9845366875228332, + "learning_rate": 0.003, + "loss": 4.0913, + "step": 18292 + }, + { + "epoch": 0.18293, + "grad_norm": 1.0819919364625812, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 18293 + }, + { + "epoch": 0.18294, + "grad_norm": 1.0913262838954416, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 18294 + }, + { + "epoch": 0.18295, + "grad_norm": 1.1126272973223525, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 18295 + }, + { + "epoch": 0.18296, + "grad_norm": 1.1617905025076987, + "learning_rate": 0.003, + "loss": 4.089, + "step": 18296 + }, + { + "epoch": 0.18297, + "grad_norm": 1.0511011492336304, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 18297 + }, + { + "epoch": 0.18298, + "grad_norm": 1.0376825728026353, + "learning_rate": 0.003, + "loss": 4.1084, + "step": 18298 + }, + { + "epoch": 0.18299, + "grad_norm": 1.1686920121454407, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 18299 + }, + { + "epoch": 0.183, + "grad_norm": 0.8958750332866732, + "learning_rate": 0.003, + "loss": 4.075, + "step": 18300 + }, + { + "epoch": 0.18301, + "grad_norm": 0.9304445035971647, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 18301 + }, + { + "epoch": 0.18302, + "grad_norm": 0.9559466565697079, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 18302 + }, + { + "epoch": 0.18303, + "grad_norm": 1.3589922421556402, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 18303 + }, + { + "epoch": 0.18304, + "grad_norm": 1.0261042865483352, + "learning_rate": 0.003, + "loss": 4.0917, + "step": 18304 + }, + { + "epoch": 0.18305, + "grad_norm": 1.2660312784526842, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 18305 + }, + { + "epoch": 0.18306, + "grad_norm": 0.9718659726420293, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 18306 + }, + { + "epoch": 0.18307, + "grad_norm": 0.9955859916385297, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 18307 + }, + { + "epoch": 0.18308, + "grad_norm": 0.9617256335387987, + "learning_rate": 0.003, + "loss": 4.052, + "step": 18308 + }, + { + "epoch": 0.18309, + "grad_norm": 1.0247017054494052, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 18309 + }, + { + "epoch": 0.1831, + "grad_norm": 0.9720373218632845, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 18310 + }, + { + "epoch": 0.18311, + "grad_norm": 1.1446854804082744, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 18311 + }, + { + "epoch": 0.18312, + "grad_norm": 0.9923514179152119, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 18312 + }, + { + "epoch": 0.18313, + "grad_norm": 1.1791866169074854, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 18313 + }, + { + "epoch": 0.18314, + "grad_norm": 0.9354298076487187, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 18314 + }, + { + "epoch": 0.18315, + "grad_norm": 0.7698359742957632, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 18315 + }, + { + "epoch": 0.18316, + "grad_norm": 0.8363463408655338, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 18316 + }, + { + "epoch": 0.18317, + "grad_norm": 1.0149336996553897, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 18317 + }, + { + "epoch": 0.18318, + "grad_norm": 1.0872761647967044, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 18318 + }, + { + "epoch": 0.18319, + "grad_norm": 0.9990339052661895, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 18319 + }, + { + "epoch": 0.1832, + "grad_norm": 1.1120922040811039, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 18320 + }, + { + "epoch": 0.18321, + "grad_norm": 1.0218645987515529, + "learning_rate": 0.003, + "loss": 4.1262, + "step": 18321 + }, + { + "epoch": 0.18322, + "grad_norm": 0.9987571898173779, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 18322 + }, + { + "epoch": 0.18323, + "grad_norm": 1.2765267261315303, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 18323 + }, + { + "epoch": 0.18324, + "grad_norm": 0.9464859504728964, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 18324 + }, + { + "epoch": 0.18325, + "grad_norm": 0.8663359124488759, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 18325 + }, + { + "epoch": 0.18326, + "grad_norm": 0.9980013708194013, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 18326 + }, + { + "epoch": 0.18327, + "grad_norm": 1.345685386793604, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 18327 + }, + { + "epoch": 0.18328, + "grad_norm": 0.9953471074231908, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 18328 + }, + { + "epoch": 0.18329, + "grad_norm": 0.9901312216825326, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 18329 + }, + { + "epoch": 0.1833, + "grad_norm": 1.2440007866831426, + "learning_rate": 0.003, + "loss": 4.1044, + "step": 18330 + }, + { + "epoch": 0.18331, + "grad_norm": 0.9089142387399571, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 18331 + }, + { + "epoch": 0.18332, + "grad_norm": 0.9185216683687568, + "learning_rate": 0.003, + "loss": 4.05, + "step": 18332 + }, + { + "epoch": 0.18333, + "grad_norm": 0.8849486487267425, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 18333 + }, + { + "epoch": 0.18334, + "grad_norm": 0.9432773704919585, + "learning_rate": 0.003, + "loss": 4.034, + "step": 18334 + }, + { + "epoch": 0.18335, + "grad_norm": 1.1369419789476078, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 18335 + }, + { + "epoch": 0.18336, + "grad_norm": 1.092176877848376, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 18336 + }, + { + "epoch": 0.18337, + "grad_norm": 0.8750582769255993, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 18337 + }, + { + "epoch": 0.18338, + "grad_norm": 0.9180409900596321, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 18338 + }, + { + "epoch": 0.18339, + "grad_norm": 1.0380409387277736, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 18339 + }, + { + "epoch": 0.1834, + "grad_norm": 1.2110837262791485, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 18340 + }, + { + "epoch": 0.18341, + "grad_norm": 0.9743049877406454, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 18341 + }, + { + "epoch": 0.18342, + "grad_norm": 1.008977256996627, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 18342 + }, + { + "epoch": 0.18343, + "grad_norm": 1.1818035205050363, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 18343 + }, + { + "epoch": 0.18344, + "grad_norm": 1.0849992801416393, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 18344 + }, + { + "epoch": 0.18345, + "grad_norm": 0.8381524938151212, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 18345 + }, + { + "epoch": 0.18346, + "grad_norm": 0.7416654482973836, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 18346 + }, + { + "epoch": 0.18347, + "grad_norm": 0.8680679902045735, + "learning_rate": 0.003, + "loss": 4.062, + "step": 18347 + }, + { + "epoch": 0.18348, + "grad_norm": 1.0142811249210657, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 18348 + }, + { + "epoch": 0.18349, + "grad_norm": 1.1581654765736493, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 18349 + }, + { + "epoch": 0.1835, + "grad_norm": 1.0262859862928602, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 18350 + }, + { + "epoch": 0.18351, + "grad_norm": 1.2239528588174924, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 18351 + }, + { + "epoch": 0.18352, + "grad_norm": 1.0375291896068208, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 18352 + }, + { + "epoch": 0.18353, + "grad_norm": 1.013195110500612, + "learning_rate": 0.003, + "loss": 4.064, + "step": 18353 + }, + { + "epoch": 0.18354, + "grad_norm": 1.0555206804767923, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 18354 + }, + { + "epoch": 0.18355, + "grad_norm": 1.0268780481187556, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 18355 + }, + { + "epoch": 0.18356, + "grad_norm": 1.2091053741928064, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 18356 + }, + { + "epoch": 0.18357, + "grad_norm": 0.8765279499843086, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 18357 + }, + { + "epoch": 0.18358, + "grad_norm": 0.8677092097177083, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 18358 + }, + { + "epoch": 0.18359, + "grad_norm": 0.8989748598410515, + "learning_rate": 0.003, + "loss": 4.0928, + "step": 18359 + }, + { + "epoch": 0.1836, + "grad_norm": 0.9902964299094082, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 18360 + }, + { + "epoch": 0.18361, + "grad_norm": 1.137006648234444, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 18361 + }, + { + "epoch": 0.18362, + "grad_norm": 0.8891331584289529, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 18362 + }, + { + "epoch": 0.18363, + "grad_norm": 0.9218315075877506, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 18363 + }, + { + "epoch": 0.18364, + "grad_norm": 0.9476907094856586, + "learning_rate": 0.003, + "loss": 4.103, + "step": 18364 + }, + { + "epoch": 0.18365, + "grad_norm": 1.1101123847563712, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 18365 + }, + { + "epoch": 0.18366, + "grad_norm": 0.9112333791219049, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 18366 + }, + { + "epoch": 0.18367, + "grad_norm": 1.090959603185802, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 18367 + }, + { + "epoch": 0.18368, + "grad_norm": 1.2988085502653668, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 18368 + }, + { + "epoch": 0.18369, + "grad_norm": 0.8937148133237743, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 18369 + }, + { + "epoch": 0.1837, + "grad_norm": 0.9758645363245967, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 18370 + }, + { + "epoch": 0.18371, + "grad_norm": 1.1376477180536697, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 18371 + }, + { + "epoch": 0.18372, + "grad_norm": 0.8732958806541976, + "learning_rate": 0.003, + "loss": 4.068, + "step": 18372 + }, + { + "epoch": 0.18373, + "grad_norm": 0.8230978519613023, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 18373 + }, + { + "epoch": 0.18374, + "grad_norm": 0.8677176011675312, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 18374 + }, + { + "epoch": 0.18375, + "grad_norm": 0.8975241396751247, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 18375 + }, + { + "epoch": 0.18376, + "grad_norm": 1.0338339295358738, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 18376 + }, + { + "epoch": 0.18377, + "grad_norm": 1.3130114559515753, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 18377 + }, + { + "epoch": 0.18378, + "grad_norm": 0.8203011185897067, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 18378 + }, + { + "epoch": 0.18379, + "grad_norm": 0.8810357094780679, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 18379 + }, + { + "epoch": 0.1838, + "grad_norm": 1.018207572861556, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 18380 + }, + { + "epoch": 0.18381, + "grad_norm": 1.307191831913616, + "learning_rate": 0.003, + "loss": 4.0918, + "step": 18381 + }, + { + "epoch": 0.18382, + "grad_norm": 0.8161908389796436, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 18382 + }, + { + "epoch": 0.18383, + "grad_norm": 0.6174531950688851, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 18383 + }, + { + "epoch": 0.18384, + "grad_norm": 0.6801494983125443, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 18384 + }, + { + "epoch": 0.18385, + "grad_norm": 0.7925620183096704, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 18385 + }, + { + "epoch": 0.18386, + "grad_norm": 1.03222703192264, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 18386 + }, + { + "epoch": 0.18387, + "grad_norm": 1.059279056724144, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 18387 + }, + { + "epoch": 0.18388, + "grad_norm": 1.0917341711305546, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 18388 + }, + { + "epoch": 0.18389, + "grad_norm": 1.217650319300019, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 18389 + }, + { + "epoch": 0.1839, + "grad_norm": 1.1892692943278462, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 18390 + }, + { + "epoch": 0.18391, + "grad_norm": 1.1085914205300513, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 18391 + }, + { + "epoch": 0.18392, + "grad_norm": 1.0555907668068372, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 18392 + }, + { + "epoch": 0.18393, + "grad_norm": 1.270322947011744, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 18393 + }, + { + "epoch": 0.18394, + "grad_norm": 1.017562527674312, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 18394 + }, + { + "epoch": 0.18395, + "grad_norm": 1.1419549253005403, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 18395 + }, + { + "epoch": 0.18396, + "grad_norm": 0.9540632525699252, + "learning_rate": 0.003, + "loss": 4.0932, + "step": 18396 + }, + { + "epoch": 0.18397, + "grad_norm": 1.1041581842958101, + "learning_rate": 0.003, + "loss": 4.0902, + "step": 18397 + }, + { + "epoch": 0.18398, + "grad_norm": 1.2217914711900832, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 18398 + }, + { + "epoch": 0.18399, + "grad_norm": 1.0109978544192313, + "learning_rate": 0.003, + "loss": 4.056, + "step": 18399 + }, + { + "epoch": 0.184, + "grad_norm": 1.0805682233474865, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 18400 + }, + { + "epoch": 0.18401, + "grad_norm": 1.1743094374761727, + "learning_rate": 0.003, + "loss": 4.072, + "step": 18401 + }, + { + "epoch": 0.18402, + "grad_norm": 0.9158559921541212, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 18402 + }, + { + "epoch": 0.18403, + "grad_norm": 0.9407847154964748, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 18403 + }, + { + "epoch": 0.18404, + "grad_norm": 1.0000026907013222, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 18404 + }, + { + "epoch": 0.18405, + "grad_norm": 0.8734803157141262, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 18405 + }, + { + "epoch": 0.18406, + "grad_norm": 0.788829844411896, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 18406 + }, + { + "epoch": 0.18407, + "grad_norm": 0.9621319984706638, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 18407 + }, + { + "epoch": 0.18408, + "grad_norm": 1.2940107279630644, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 18408 + }, + { + "epoch": 0.18409, + "grad_norm": 0.9820956238958781, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 18409 + }, + { + "epoch": 0.1841, + "grad_norm": 1.0445185062268691, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 18410 + }, + { + "epoch": 0.18411, + "grad_norm": 1.27855185451317, + "learning_rate": 0.003, + "loss": 4.0887, + "step": 18411 + }, + { + "epoch": 0.18412, + "grad_norm": 0.8639967971912106, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 18412 + }, + { + "epoch": 0.18413, + "grad_norm": 0.8997233676075509, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 18413 + }, + { + "epoch": 0.18414, + "grad_norm": 0.9273273906040044, + "learning_rate": 0.003, + "loss": 4.062, + "step": 18414 + }, + { + "epoch": 0.18415, + "grad_norm": 0.9579502253223268, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 18415 + }, + { + "epoch": 0.18416, + "grad_norm": 0.8517073735452731, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 18416 + }, + { + "epoch": 0.18417, + "grad_norm": 0.8773925511461428, + "learning_rate": 0.003, + "loss": 4.0931, + "step": 18417 + }, + { + "epoch": 0.18418, + "grad_norm": 0.9929906322983941, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 18418 + }, + { + "epoch": 0.18419, + "grad_norm": 1.2295719831612348, + "learning_rate": 0.003, + "loss": 4.049, + "step": 18419 + }, + { + "epoch": 0.1842, + "grad_norm": 1.0459956298217221, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 18420 + }, + { + "epoch": 0.18421, + "grad_norm": 1.1988689450677696, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 18421 + }, + { + "epoch": 0.18422, + "grad_norm": 0.9642460561423983, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 18422 + }, + { + "epoch": 0.18423, + "grad_norm": 0.8474762290044032, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 18423 + }, + { + "epoch": 0.18424, + "grad_norm": 0.8636845802446662, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 18424 + }, + { + "epoch": 0.18425, + "grad_norm": 0.9982215082404473, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 18425 + }, + { + "epoch": 0.18426, + "grad_norm": 1.0421903403897372, + "learning_rate": 0.003, + "loss": 4.065, + "step": 18426 + }, + { + "epoch": 0.18427, + "grad_norm": 1.03593566837555, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 18427 + }, + { + "epoch": 0.18428, + "grad_norm": 1.1797082293692724, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 18428 + }, + { + "epoch": 0.18429, + "grad_norm": 0.9417114017375848, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 18429 + }, + { + "epoch": 0.1843, + "grad_norm": 1.1456949136480872, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 18430 + }, + { + "epoch": 0.18431, + "grad_norm": 0.951799867111194, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 18431 + }, + { + "epoch": 0.18432, + "grad_norm": 1.0053357064305133, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 18432 + }, + { + "epoch": 0.18433, + "grad_norm": 1.0718748422679771, + "learning_rate": 0.003, + "loss": 4.047, + "step": 18433 + }, + { + "epoch": 0.18434, + "grad_norm": 1.0728986208289892, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 18434 + }, + { + "epoch": 0.18435, + "grad_norm": 1.1295080357990344, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 18435 + }, + { + "epoch": 0.18436, + "grad_norm": 1.0078854451836865, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 18436 + }, + { + "epoch": 0.18437, + "grad_norm": 1.0896271406240943, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 18437 + }, + { + "epoch": 0.18438, + "grad_norm": 0.911992936680652, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 18438 + }, + { + "epoch": 0.18439, + "grad_norm": 0.9820194215488122, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 18439 + }, + { + "epoch": 0.1844, + "grad_norm": 1.06921452899116, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 18440 + }, + { + "epoch": 0.18441, + "grad_norm": 1.0237784904450071, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 18441 + }, + { + "epoch": 0.18442, + "grad_norm": 1.146423205134172, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 18442 + }, + { + "epoch": 0.18443, + "grad_norm": 1.105152717943077, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 18443 + }, + { + "epoch": 0.18444, + "grad_norm": 0.9844056597067876, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 18444 + }, + { + "epoch": 0.18445, + "grad_norm": 0.9179369412959669, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 18445 + }, + { + "epoch": 0.18446, + "grad_norm": 0.956561926492257, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 18446 + }, + { + "epoch": 0.18447, + "grad_norm": 0.9627147792423177, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 18447 + }, + { + "epoch": 0.18448, + "grad_norm": 1.0511610110256289, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 18448 + }, + { + "epoch": 0.18449, + "grad_norm": 1.067537482826323, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 18449 + }, + { + "epoch": 0.1845, + "grad_norm": 1.1623575475125958, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 18450 + }, + { + "epoch": 0.18451, + "grad_norm": 0.9245988675120697, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 18451 + }, + { + "epoch": 0.18452, + "grad_norm": 1.32083607817048, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 18452 + }, + { + "epoch": 0.18453, + "grad_norm": 0.9446304331191059, + "learning_rate": 0.003, + "loss": 4.1083, + "step": 18453 + }, + { + "epoch": 0.18454, + "grad_norm": 0.8868010114532057, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 18454 + }, + { + "epoch": 0.18455, + "grad_norm": 0.8983656430088794, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 18455 + }, + { + "epoch": 0.18456, + "grad_norm": 0.9097532705463757, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 18456 + }, + { + "epoch": 0.18457, + "grad_norm": 1.2349288901695634, + "learning_rate": 0.003, + "loss": 4.0945, + "step": 18457 + }, + { + "epoch": 0.18458, + "grad_norm": 1.0568862192610602, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 18458 + }, + { + "epoch": 0.18459, + "grad_norm": 0.9643182296989079, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 18459 + }, + { + "epoch": 0.1846, + "grad_norm": 1.0288077281884063, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 18460 + }, + { + "epoch": 0.18461, + "grad_norm": 1.107015161372059, + "learning_rate": 0.003, + "loss": 4.067, + "step": 18461 + }, + { + "epoch": 0.18462, + "grad_norm": 1.128429844836738, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 18462 + }, + { + "epoch": 0.18463, + "grad_norm": 1.3061129606193311, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 18463 + }, + { + "epoch": 0.18464, + "grad_norm": 0.958575717136666, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 18464 + }, + { + "epoch": 0.18465, + "grad_norm": 1.08970698764656, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 18465 + }, + { + "epoch": 0.18466, + "grad_norm": 0.9847322047029925, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 18466 + }, + { + "epoch": 0.18467, + "grad_norm": 1.0233389137449376, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 18467 + }, + { + "epoch": 0.18468, + "grad_norm": 1.1481364533112246, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 18468 + }, + { + "epoch": 0.18469, + "grad_norm": 1.008469096224946, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 18469 + }, + { + "epoch": 0.1847, + "grad_norm": 1.1323517190826835, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 18470 + }, + { + "epoch": 0.18471, + "grad_norm": 0.9304222203464761, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 18471 + }, + { + "epoch": 0.18472, + "grad_norm": 1.0362745072179949, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 18472 + }, + { + "epoch": 0.18473, + "grad_norm": 1.093763913809579, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 18473 + }, + { + "epoch": 0.18474, + "grad_norm": 1.0398969376530685, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 18474 + }, + { + "epoch": 0.18475, + "grad_norm": 1.1120850283435315, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 18475 + }, + { + "epoch": 0.18476, + "grad_norm": 1.0642337602233973, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 18476 + }, + { + "epoch": 0.18477, + "grad_norm": 1.0365129586069282, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 18477 + }, + { + "epoch": 0.18478, + "grad_norm": 1.299003178653765, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 18478 + }, + { + "epoch": 0.18479, + "grad_norm": 1.2479022169270289, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 18479 + }, + { + "epoch": 0.1848, + "grad_norm": 0.9659538435352071, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 18480 + }, + { + "epoch": 0.18481, + "grad_norm": 1.0346500210676806, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 18481 + }, + { + "epoch": 0.18482, + "grad_norm": 0.9472493062478453, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 18482 + }, + { + "epoch": 0.18483, + "grad_norm": 0.9219886296427339, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 18483 + }, + { + "epoch": 0.18484, + "grad_norm": 0.9362506753474548, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 18484 + }, + { + "epoch": 0.18485, + "grad_norm": 1.0088138399607236, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 18485 + }, + { + "epoch": 0.18486, + "grad_norm": 1.1252293366711046, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 18486 + }, + { + "epoch": 0.18487, + "grad_norm": 1.0451430969371667, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 18487 + }, + { + "epoch": 0.18488, + "grad_norm": 0.9971953712760308, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 18488 + }, + { + "epoch": 0.18489, + "grad_norm": 1.115073592471031, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 18489 + }, + { + "epoch": 0.1849, + "grad_norm": 0.9649271173191476, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 18490 + }, + { + "epoch": 0.18491, + "grad_norm": 1.088425700548664, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 18491 + }, + { + "epoch": 0.18492, + "grad_norm": 0.837146605725308, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 18492 + }, + { + "epoch": 0.18493, + "grad_norm": 0.8624749898012253, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 18493 + }, + { + "epoch": 0.18494, + "grad_norm": 1.0164909766354353, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 18494 + }, + { + "epoch": 0.18495, + "grad_norm": 1.2180001480378022, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 18495 + }, + { + "epoch": 0.18496, + "grad_norm": 0.8718697397651684, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 18496 + }, + { + "epoch": 0.18497, + "grad_norm": 1.055432767034243, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 18497 + }, + { + "epoch": 0.18498, + "grad_norm": 1.0808698034518378, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 18498 + }, + { + "epoch": 0.18499, + "grad_norm": 1.0646855768636938, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 18499 + }, + { + "epoch": 0.185, + "grad_norm": 0.8855825507585664, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 18500 + }, + { + "epoch": 0.18501, + "grad_norm": 0.990235119578552, + "learning_rate": 0.003, + "loss": 4.036, + "step": 18501 + }, + { + "epoch": 0.18502, + "grad_norm": 0.9262326842948654, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 18502 + }, + { + "epoch": 0.18503, + "grad_norm": 0.8896525475792657, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 18503 + }, + { + "epoch": 0.18504, + "grad_norm": 0.9188725973254425, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 18504 + }, + { + "epoch": 0.18505, + "grad_norm": 0.9817176498171435, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 18505 + }, + { + "epoch": 0.18506, + "grad_norm": 1.442080319880222, + "learning_rate": 0.003, + "loss": 4.1006, + "step": 18506 + }, + { + "epoch": 0.18507, + "grad_norm": 1.2460039916619492, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 18507 + }, + { + "epoch": 0.18508, + "grad_norm": 1.0872978788349155, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 18508 + }, + { + "epoch": 0.18509, + "grad_norm": 1.0813005295611824, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 18509 + }, + { + "epoch": 0.1851, + "grad_norm": 0.9326726276274628, + "learning_rate": 0.003, + "loss": 4.0908, + "step": 18510 + }, + { + "epoch": 0.18511, + "grad_norm": 1.0819257887722924, + "learning_rate": 0.003, + "loss": 4.1018, + "step": 18511 + }, + { + "epoch": 0.18512, + "grad_norm": 1.100559218326041, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 18512 + }, + { + "epoch": 0.18513, + "grad_norm": 0.7935686926944789, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 18513 + }, + { + "epoch": 0.18514, + "grad_norm": 0.87111317068394, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 18514 + }, + { + "epoch": 0.18515, + "grad_norm": 0.9742556303388907, + "learning_rate": 0.003, + "loss": 4.021, + "step": 18515 + }, + { + "epoch": 0.18516, + "grad_norm": 1.1797120083572494, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 18516 + }, + { + "epoch": 0.18517, + "grad_norm": 1.0833231106009666, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 18517 + }, + { + "epoch": 0.18518, + "grad_norm": 1.0872452648168673, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 18518 + }, + { + "epoch": 0.18519, + "grad_norm": 0.9334536056839767, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 18519 + }, + { + "epoch": 0.1852, + "grad_norm": 0.983555471823677, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 18520 + }, + { + "epoch": 0.18521, + "grad_norm": 1.1280716007609901, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 18521 + }, + { + "epoch": 0.18522, + "grad_norm": 1.0842586056972137, + "learning_rate": 0.003, + "loss": 4.059, + "step": 18522 + }, + { + "epoch": 0.18523, + "grad_norm": 1.0222007448366213, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 18523 + }, + { + "epoch": 0.18524, + "grad_norm": 1.0475797750318427, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 18524 + }, + { + "epoch": 0.18525, + "grad_norm": 0.9313010818707739, + "learning_rate": 0.003, + "loss": 4.048, + "step": 18525 + }, + { + "epoch": 0.18526, + "grad_norm": 0.9722730056157963, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 18526 + }, + { + "epoch": 0.18527, + "grad_norm": 1.0107377377728928, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 18527 + }, + { + "epoch": 0.18528, + "grad_norm": 0.9679035463762834, + "learning_rate": 0.003, + "loss": 4.0845, + "step": 18528 + }, + { + "epoch": 0.18529, + "grad_norm": 1.1945495936972665, + "learning_rate": 0.003, + "loss": 4.049, + "step": 18529 + }, + { + "epoch": 0.1853, + "grad_norm": 1.1904945360950074, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 18530 + }, + { + "epoch": 0.18531, + "grad_norm": 1.2149470391078423, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 18531 + }, + { + "epoch": 0.18532, + "grad_norm": 0.9474623311673445, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 18532 + }, + { + "epoch": 0.18533, + "grad_norm": 0.9108239665409057, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 18533 + }, + { + "epoch": 0.18534, + "grad_norm": 1.0563180579026064, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 18534 + }, + { + "epoch": 0.18535, + "grad_norm": 1.2665730966629127, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 18535 + }, + { + "epoch": 0.18536, + "grad_norm": 0.8010821988892711, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 18536 + }, + { + "epoch": 0.18537, + "grad_norm": 0.8751683015517626, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 18537 + }, + { + "epoch": 0.18538, + "grad_norm": 1.0864198090956378, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 18538 + }, + { + "epoch": 0.18539, + "grad_norm": 1.240584218016658, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 18539 + }, + { + "epoch": 0.1854, + "grad_norm": 0.8896390387128419, + "learning_rate": 0.003, + "loss": 4.068, + "step": 18540 + }, + { + "epoch": 0.18541, + "grad_norm": 0.8024337578259342, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 18541 + }, + { + "epoch": 0.18542, + "grad_norm": 1.027767064946907, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 18542 + }, + { + "epoch": 0.18543, + "grad_norm": 1.5076433992059082, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 18543 + }, + { + "epoch": 0.18544, + "grad_norm": 0.6741152262499259, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 18544 + }, + { + "epoch": 0.18545, + "grad_norm": 0.8697041251558383, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 18545 + }, + { + "epoch": 0.18546, + "grad_norm": 1.228452532198602, + "learning_rate": 0.003, + "loss": 4.046, + "step": 18546 + }, + { + "epoch": 0.18547, + "grad_norm": 0.9429434820363315, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 18547 + }, + { + "epoch": 0.18548, + "grad_norm": 1.0565095810857705, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 18548 + }, + { + "epoch": 0.18549, + "grad_norm": 0.9057531546477974, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 18549 + }, + { + "epoch": 0.1855, + "grad_norm": 0.9507490218990253, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 18550 + }, + { + "epoch": 0.18551, + "grad_norm": 1.0664167570006196, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 18551 + }, + { + "epoch": 0.18552, + "grad_norm": 1.0690763175694167, + "learning_rate": 0.003, + "loss": 4.1003, + "step": 18552 + }, + { + "epoch": 0.18553, + "grad_norm": 1.0768131062970772, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 18553 + }, + { + "epoch": 0.18554, + "grad_norm": 0.9696580152927612, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 18554 + }, + { + "epoch": 0.18555, + "grad_norm": 1.0967400044448716, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 18555 + }, + { + "epoch": 0.18556, + "grad_norm": 1.1444487844798057, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 18556 + }, + { + "epoch": 0.18557, + "grad_norm": 1.1123978853308414, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 18557 + }, + { + "epoch": 0.18558, + "grad_norm": 0.9251052288899151, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 18558 + }, + { + "epoch": 0.18559, + "grad_norm": 0.9265911683114179, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 18559 + }, + { + "epoch": 0.1856, + "grad_norm": 0.9760288075037313, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 18560 + }, + { + "epoch": 0.18561, + "grad_norm": 0.9594363011111875, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 18561 + }, + { + "epoch": 0.18562, + "grad_norm": 0.9957342689373374, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 18562 + }, + { + "epoch": 0.18563, + "grad_norm": 1.0630219046163638, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 18563 + }, + { + "epoch": 0.18564, + "grad_norm": 0.8640397591473526, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 18564 + }, + { + "epoch": 0.18565, + "grad_norm": 0.9940656955708529, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 18565 + }, + { + "epoch": 0.18566, + "grad_norm": 1.0618983438265905, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 18566 + }, + { + "epoch": 0.18567, + "grad_norm": 0.9860046296775526, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 18567 + }, + { + "epoch": 0.18568, + "grad_norm": 1.16137170995902, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 18568 + }, + { + "epoch": 0.18569, + "grad_norm": 0.9202165079160673, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 18569 + }, + { + "epoch": 0.1857, + "grad_norm": 0.982053078931386, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 18570 + }, + { + "epoch": 0.18571, + "grad_norm": 1.1567836252268135, + "learning_rate": 0.003, + "loss": 4.0873, + "step": 18571 + }, + { + "epoch": 0.18572, + "grad_norm": 1.0347577566147594, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 18572 + }, + { + "epoch": 0.18573, + "grad_norm": 1.0469444777772607, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 18573 + }, + { + "epoch": 0.18574, + "grad_norm": 1.1838070862938586, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 18574 + }, + { + "epoch": 0.18575, + "grad_norm": 1.2431264317637685, + "learning_rate": 0.003, + "loss": 4.0946, + "step": 18575 + }, + { + "epoch": 0.18576, + "grad_norm": 1.0516178949522679, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 18576 + }, + { + "epoch": 0.18577, + "grad_norm": 1.067554280947935, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 18577 + }, + { + "epoch": 0.18578, + "grad_norm": 0.9812328257887662, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 18578 + }, + { + "epoch": 0.18579, + "grad_norm": 1.0345100123407154, + "learning_rate": 0.003, + "loss": 4.0901, + "step": 18579 + }, + { + "epoch": 0.1858, + "grad_norm": 1.161524135097281, + "learning_rate": 0.003, + "loss": 4.035, + "step": 18580 + }, + { + "epoch": 0.18581, + "grad_norm": 0.881767727621155, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 18581 + }, + { + "epoch": 0.18582, + "grad_norm": 0.9805368914783295, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 18582 + }, + { + "epoch": 0.18583, + "grad_norm": 1.0874464405115238, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 18583 + }, + { + "epoch": 0.18584, + "grad_norm": 0.9728185355689365, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 18584 + }, + { + "epoch": 0.18585, + "grad_norm": 1.03045827365126, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 18585 + }, + { + "epoch": 0.18586, + "grad_norm": 0.9935396921269278, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 18586 + }, + { + "epoch": 0.18587, + "grad_norm": 1.0121887587118545, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 18587 + }, + { + "epoch": 0.18588, + "grad_norm": 1.2656835242307474, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 18588 + }, + { + "epoch": 0.18589, + "grad_norm": 0.9757601165205843, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 18589 + }, + { + "epoch": 0.1859, + "grad_norm": 1.3382819846840293, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 18590 + }, + { + "epoch": 0.18591, + "grad_norm": 0.8255063864487283, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 18591 + }, + { + "epoch": 0.18592, + "grad_norm": 0.9239903871836477, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 18592 + }, + { + "epoch": 0.18593, + "grad_norm": 1.127306944296798, + "learning_rate": 0.003, + "loss": 4.1017, + "step": 18593 + }, + { + "epoch": 0.18594, + "grad_norm": 1.0291463213515077, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 18594 + }, + { + "epoch": 0.18595, + "grad_norm": 0.9002685265822157, + "learning_rate": 0.003, + "loss": 4.072, + "step": 18595 + }, + { + "epoch": 0.18596, + "grad_norm": 0.9116051812111878, + "learning_rate": 0.003, + "loss": 4.1018, + "step": 18596 + }, + { + "epoch": 0.18597, + "grad_norm": 1.1159920532053422, + "learning_rate": 0.003, + "loss": 4.0963, + "step": 18597 + }, + { + "epoch": 0.18598, + "grad_norm": 1.235682027758359, + "learning_rate": 0.003, + "loss": 4.1009, + "step": 18598 + }, + { + "epoch": 0.18599, + "grad_norm": 1.0435175141288648, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 18599 + }, + { + "epoch": 0.186, + "grad_norm": 1.14149032065731, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 18600 + }, + { + "epoch": 0.18601, + "grad_norm": 0.8843998718774485, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 18601 + }, + { + "epoch": 0.18602, + "grad_norm": 0.9388302167055703, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 18602 + }, + { + "epoch": 0.18603, + "grad_norm": 1.2049334686878859, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 18603 + }, + { + "epoch": 0.18604, + "grad_norm": 1.0737565094666142, + "learning_rate": 0.003, + "loss": 4.0926, + "step": 18604 + }, + { + "epoch": 0.18605, + "grad_norm": 1.005204142213794, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 18605 + }, + { + "epoch": 0.18606, + "grad_norm": 1.0638593305615718, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 18606 + }, + { + "epoch": 0.18607, + "grad_norm": 1.038551759573246, + "learning_rate": 0.003, + "loss": 4.0864, + "step": 18607 + }, + { + "epoch": 0.18608, + "grad_norm": 1.0385497463452955, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 18608 + }, + { + "epoch": 0.18609, + "grad_norm": 0.9401892885741805, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 18609 + }, + { + "epoch": 0.1861, + "grad_norm": 0.9222711076786131, + "learning_rate": 0.003, + "loss": 4.0871, + "step": 18610 + }, + { + "epoch": 0.18611, + "grad_norm": 1.0532714804275027, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 18611 + }, + { + "epoch": 0.18612, + "grad_norm": 1.0573097964473082, + "learning_rate": 0.003, + "loss": 4.075, + "step": 18612 + }, + { + "epoch": 0.18613, + "grad_norm": 1.1247639162598944, + "learning_rate": 0.003, + "loss": 4.1051, + "step": 18613 + }, + { + "epoch": 0.18614, + "grad_norm": 1.0129323977153715, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 18614 + }, + { + "epoch": 0.18615, + "grad_norm": 1.126440548607524, + "learning_rate": 0.003, + "loss": 4.0888, + "step": 18615 + }, + { + "epoch": 0.18616, + "grad_norm": 1.082772199450867, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 18616 + }, + { + "epoch": 0.18617, + "grad_norm": 1.0107000463127132, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 18617 + }, + { + "epoch": 0.18618, + "grad_norm": 1.1527452220596444, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 18618 + }, + { + "epoch": 0.18619, + "grad_norm": 0.9430014695999142, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 18619 + }, + { + "epoch": 0.1862, + "grad_norm": 1.0424110020025448, + "learning_rate": 0.003, + "loss": 4.0854, + "step": 18620 + }, + { + "epoch": 0.18621, + "grad_norm": 1.1077743069506298, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 18621 + }, + { + "epoch": 0.18622, + "grad_norm": 1.1604539252819168, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 18622 + }, + { + "epoch": 0.18623, + "grad_norm": 0.9487782545833757, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 18623 + }, + { + "epoch": 0.18624, + "grad_norm": 0.9345840585432955, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 18624 + }, + { + "epoch": 0.18625, + "grad_norm": 1.0300488890140642, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 18625 + }, + { + "epoch": 0.18626, + "grad_norm": 1.065051432007663, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 18626 + }, + { + "epoch": 0.18627, + "grad_norm": 0.9152565567482102, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 18627 + }, + { + "epoch": 0.18628, + "grad_norm": 0.8697835316664154, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 18628 + }, + { + "epoch": 0.18629, + "grad_norm": 0.8933052909257613, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 18629 + }, + { + "epoch": 0.1863, + "grad_norm": 0.9794791509808408, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 18630 + }, + { + "epoch": 0.18631, + "grad_norm": 1.0607676446042278, + "learning_rate": 0.003, + "loss": 4.0934, + "step": 18631 + }, + { + "epoch": 0.18632, + "grad_norm": 1.1903214950986176, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 18632 + }, + { + "epoch": 0.18633, + "grad_norm": 0.8456447816228931, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 18633 + }, + { + "epoch": 0.18634, + "grad_norm": 0.8156061877682977, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 18634 + }, + { + "epoch": 0.18635, + "grad_norm": 0.8081869135413564, + "learning_rate": 0.003, + "loss": 4.0873, + "step": 18635 + }, + { + "epoch": 0.18636, + "grad_norm": 0.928134029040471, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 18636 + }, + { + "epoch": 0.18637, + "grad_norm": 1.159402184754463, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 18637 + }, + { + "epoch": 0.18638, + "grad_norm": 1.0699783751016423, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 18638 + }, + { + "epoch": 0.18639, + "grad_norm": 1.1230604918664524, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 18639 + }, + { + "epoch": 0.1864, + "grad_norm": 1.11591507449526, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 18640 + }, + { + "epoch": 0.18641, + "grad_norm": 1.18915642447669, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 18641 + }, + { + "epoch": 0.18642, + "grad_norm": 1.0198360409438225, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 18642 + }, + { + "epoch": 0.18643, + "grad_norm": 1.1542057075571983, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 18643 + }, + { + "epoch": 0.18644, + "grad_norm": 0.9890882552836842, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 18644 + }, + { + "epoch": 0.18645, + "grad_norm": 1.0480087362316832, + "learning_rate": 0.003, + "loss": 4.051, + "step": 18645 + }, + { + "epoch": 0.18646, + "grad_norm": 1.0663102870339642, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 18646 + }, + { + "epoch": 0.18647, + "grad_norm": 0.9830140115246306, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 18647 + }, + { + "epoch": 0.18648, + "grad_norm": 0.9784059490414414, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 18648 + }, + { + "epoch": 0.18649, + "grad_norm": 1.1392625119267306, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 18649 + }, + { + "epoch": 0.1865, + "grad_norm": 1.065466915086816, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 18650 + }, + { + "epoch": 0.18651, + "grad_norm": 0.9021012733401452, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 18651 + }, + { + "epoch": 0.18652, + "grad_norm": 0.9809356517045076, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 18652 + }, + { + "epoch": 0.18653, + "grad_norm": 1.010882151932905, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 18653 + }, + { + "epoch": 0.18654, + "grad_norm": 1.1978204697732584, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 18654 + }, + { + "epoch": 0.18655, + "grad_norm": 1.0830463618135966, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 18655 + }, + { + "epoch": 0.18656, + "grad_norm": 0.9586133775847379, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 18656 + }, + { + "epoch": 0.18657, + "grad_norm": 0.9976819652374724, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 18657 + }, + { + "epoch": 0.18658, + "grad_norm": 1.0113824401293856, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 18658 + }, + { + "epoch": 0.18659, + "grad_norm": 0.9296132054742002, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 18659 + }, + { + "epoch": 0.1866, + "grad_norm": 0.8572582098584475, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 18660 + }, + { + "epoch": 0.18661, + "grad_norm": 0.9856786048249443, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 18661 + }, + { + "epoch": 0.18662, + "grad_norm": 1.2584624038589345, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 18662 + }, + { + "epoch": 0.18663, + "grad_norm": 0.9505816132684474, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 18663 + }, + { + "epoch": 0.18664, + "grad_norm": 0.8662513126293973, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 18664 + }, + { + "epoch": 0.18665, + "grad_norm": 0.9587126616062863, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 18665 + }, + { + "epoch": 0.18666, + "grad_norm": 1.3378015738606435, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 18666 + }, + { + "epoch": 0.18667, + "grad_norm": 0.7953459244757481, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 18667 + }, + { + "epoch": 0.18668, + "grad_norm": 0.8068212350456846, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 18668 + }, + { + "epoch": 0.18669, + "grad_norm": 0.9372854356408552, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 18669 + }, + { + "epoch": 0.1867, + "grad_norm": 1.2898928921752006, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 18670 + }, + { + "epoch": 0.18671, + "grad_norm": 0.9647587844139296, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 18671 + }, + { + "epoch": 0.18672, + "grad_norm": 1.029072362883631, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 18672 + }, + { + "epoch": 0.18673, + "grad_norm": 0.9209403596958489, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 18673 + }, + { + "epoch": 0.18674, + "grad_norm": 0.9373922495350042, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 18674 + }, + { + "epoch": 0.18675, + "grad_norm": 0.9967178059312145, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 18675 + }, + { + "epoch": 0.18676, + "grad_norm": 1.1200713431616491, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 18676 + }, + { + "epoch": 0.18677, + "grad_norm": 1.0689958678197289, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 18677 + }, + { + "epoch": 0.18678, + "grad_norm": 1.038058513345568, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 18678 + }, + { + "epoch": 0.18679, + "grad_norm": 1.0961003222875403, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 18679 + }, + { + "epoch": 0.1868, + "grad_norm": 1.1667367872846057, + "learning_rate": 0.003, + "loss": 4.062, + "step": 18680 + }, + { + "epoch": 0.18681, + "grad_norm": 0.9483313984134422, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 18681 + }, + { + "epoch": 0.18682, + "grad_norm": 1.08009886502674, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 18682 + }, + { + "epoch": 0.18683, + "grad_norm": 0.9956765090086841, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 18683 + }, + { + "epoch": 0.18684, + "grad_norm": 1.0014457627034954, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 18684 + }, + { + "epoch": 0.18685, + "grad_norm": 1.0632750321707838, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 18685 + }, + { + "epoch": 0.18686, + "grad_norm": 1.0931797365352698, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 18686 + }, + { + "epoch": 0.18687, + "grad_norm": 1.093174326274367, + "learning_rate": 0.003, + "loss": 4.0944, + "step": 18687 + }, + { + "epoch": 0.18688, + "grad_norm": 1.196419245272059, + "learning_rate": 0.003, + "loss": 4.07, + "step": 18688 + }, + { + "epoch": 0.18689, + "grad_norm": 0.9791243168315852, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 18689 + }, + { + "epoch": 0.1869, + "grad_norm": 1.1114547621140887, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 18690 + }, + { + "epoch": 0.18691, + "grad_norm": 1.1417105693737744, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 18691 + }, + { + "epoch": 0.18692, + "grad_norm": 1.0587969901522445, + "learning_rate": 0.003, + "loss": 4.047, + "step": 18692 + }, + { + "epoch": 0.18693, + "grad_norm": 1.052011888879862, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 18693 + }, + { + "epoch": 0.18694, + "grad_norm": 0.9368391393581812, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 18694 + }, + { + "epoch": 0.18695, + "grad_norm": 0.9251076736427094, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 18695 + }, + { + "epoch": 0.18696, + "grad_norm": 1.0549918981063626, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 18696 + }, + { + "epoch": 0.18697, + "grad_norm": 1.4960983903759504, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 18697 + }, + { + "epoch": 0.18698, + "grad_norm": 0.7859431911003104, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 18698 + }, + { + "epoch": 0.18699, + "grad_norm": 0.7721874452270914, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 18699 + }, + { + "epoch": 0.187, + "grad_norm": 0.7897013695767281, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 18700 + }, + { + "epoch": 0.18701, + "grad_norm": 0.8681199472048362, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 18701 + }, + { + "epoch": 0.18702, + "grad_norm": 0.9903299756570473, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 18702 + }, + { + "epoch": 0.18703, + "grad_norm": 1.0447928644698414, + "learning_rate": 0.003, + "loss": 4.069, + "step": 18703 + }, + { + "epoch": 0.18704, + "grad_norm": 0.9754515225033243, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 18704 + }, + { + "epoch": 0.18705, + "grad_norm": 1.0009876263167674, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 18705 + }, + { + "epoch": 0.18706, + "grad_norm": 1.112102263337297, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 18706 + }, + { + "epoch": 0.18707, + "grad_norm": 1.0173166497367596, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 18707 + }, + { + "epoch": 0.18708, + "grad_norm": 1.1777599753055243, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 18708 + }, + { + "epoch": 0.18709, + "grad_norm": 1.1025207688469114, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 18709 + }, + { + "epoch": 0.1871, + "grad_norm": 0.977481874572276, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 18710 + }, + { + "epoch": 0.18711, + "grad_norm": 1.0175263379979678, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 18711 + }, + { + "epoch": 0.18712, + "grad_norm": 1.2037012277838028, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 18712 + }, + { + "epoch": 0.18713, + "grad_norm": 0.8836625075061728, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 18713 + }, + { + "epoch": 0.18714, + "grad_norm": 1.1590077865291266, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 18714 + }, + { + "epoch": 0.18715, + "grad_norm": 1.0306482492433613, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 18715 + }, + { + "epoch": 0.18716, + "grad_norm": 1.1138564447880157, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 18716 + }, + { + "epoch": 0.18717, + "grad_norm": 0.9790259856641267, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 18717 + }, + { + "epoch": 0.18718, + "grad_norm": 1.0628155517562827, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 18718 + }, + { + "epoch": 0.18719, + "grad_norm": 1.0442341646338227, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 18719 + }, + { + "epoch": 0.1872, + "grad_norm": 1.0925768021843332, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 18720 + }, + { + "epoch": 0.18721, + "grad_norm": 1.0162838055295442, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 18721 + }, + { + "epoch": 0.18722, + "grad_norm": 0.9681197923909607, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 18722 + }, + { + "epoch": 0.18723, + "grad_norm": 0.8923649465985298, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 18723 + }, + { + "epoch": 0.18724, + "grad_norm": 0.8479351894471041, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 18724 + }, + { + "epoch": 0.18725, + "grad_norm": 0.9349314773474666, + "learning_rate": 0.003, + "loss": 4.051, + "step": 18725 + }, + { + "epoch": 0.18726, + "grad_norm": 1.290419220710696, + "learning_rate": 0.003, + "loss": 4.063, + "step": 18726 + }, + { + "epoch": 0.18727, + "grad_norm": 1.1317794836374053, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 18727 + }, + { + "epoch": 0.18728, + "grad_norm": 0.9738869164752744, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 18728 + }, + { + "epoch": 0.18729, + "grad_norm": 0.9902190012310191, + "learning_rate": 0.003, + "loss": 4.1011, + "step": 18729 + }, + { + "epoch": 0.1873, + "grad_norm": 1.140273704149311, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 18730 + }, + { + "epoch": 0.18731, + "grad_norm": 0.9864666450098403, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 18731 + }, + { + "epoch": 0.18732, + "grad_norm": 1.043359525407706, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 18732 + }, + { + "epoch": 0.18733, + "grad_norm": 1.1528608884186806, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 18733 + }, + { + "epoch": 0.18734, + "grad_norm": 0.9497518381317603, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 18734 + }, + { + "epoch": 0.18735, + "grad_norm": 0.8691265794835493, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 18735 + }, + { + "epoch": 0.18736, + "grad_norm": 0.9479297675635681, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 18736 + }, + { + "epoch": 0.18737, + "grad_norm": 1.1056554459844588, + "learning_rate": 0.003, + "loss": 4.09, + "step": 18737 + }, + { + "epoch": 0.18738, + "grad_norm": 1.1210343107004788, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 18738 + }, + { + "epoch": 0.18739, + "grad_norm": 1.1769054505105425, + "learning_rate": 0.003, + "loss": 4.0937, + "step": 18739 + }, + { + "epoch": 0.1874, + "grad_norm": 1.001599590423247, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 18740 + }, + { + "epoch": 0.18741, + "grad_norm": 1.1523143235962647, + "learning_rate": 0.003, + "loss": 4.067, + "step": 18741 + }, + { + "epoch": 0.18742, + "grad_norm": 0.9749281355250112, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 18742 + }, + { + "epoch": 0.18743, + "grad_norm": 1.1753410177369727, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 18743 + }, + { + "epoch": 0.18744, + "grad_norm": 1.0465119610108828, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 18744 + }, + { + "epoch": 0.18745, + "grad_norm": 0.9597280540602556, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 18745 + }, + { + "epoch": 0.18746, + "grad_norm": 1.047859341332082, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 18746 + }, + { + "epoch": 0.18747, + "grad_norm": 1.2463363569737471, + "learning_rate": 0.003, + "loss": 4.0947, + "step": 18747 + }, + { + "epoch": 0.18748, + "grad_norm": 1.0508070578647342, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 18748 + }, + { + "epoch": 0.18749, + "grad_norm": 0.9758553083748893, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 18749 + }, + { + "epoch": 0.1875, + "grad_norm": 0.9695786338444176, + "learning_rate": 0.003, + "loss": 4.0881, + "step": 18750 + }, + { + "epoch": 0.18751, + "grad_norm": 1.1111836442918612, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 18751 + }, + { + "epoch": 0.18752, + "grad_norm": 0.8298730589350851, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 18752 + }, + { + "epoch": 0.18753, + "grad_norm": 0.8316730642480865, + "learning_rate": 0.003, + "loss": 4.083, + "step": 18753 + }, + { + "epoch": 0.18754, + "grad_norm": 0.850415804381357, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 18754 + }, + { + "epoch": 0.18755, + "grad_norm": 0.8040728265455248, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 18755 + }, + { + "epoch": 0.18756, + "grad_norm": 0.8790358616866483, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 18756 + }, + { + "epoch": 0.18757, + "grad_norm": 0.8462322693695106, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 18757 + }, + { + "epoch": 0.18758, + "grad_norm": 0.9841939077161836, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 18758 + }, + { + "epoch": 0.18759, + "grad_norm": 1.1127067093303962, + "learning_rate": 0.003, + "loss": 4.0927, + "step": 18759 + }, + { + "epoch": 0.1876, + "grad_norm": 1.381828836151113, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 18760 + }, + { + "epoch": 0.18761, + "grad_norm": 1.0514246843002448, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 18761 + }, + { + "epoch": 0.18762, + "grad_norm": 1.0796669841396993, + "learning_rate": 0.003, + "loss": 4.0845, + "step": 18762 + }, + { + "epoch": 0.18763, + "grad_norm": 1.1160788626072016, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 18763 + }, + { + "epoch": 0.18764, + "grad_norm": 1.1291055432026835, + "learning_rate": 0.003, + "loss": 4.0893, + "step": 18764 + }, + { + "epoch": 0.18765, + "grad_norm": 0.8082501286050859, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 18765 + }, + { + "epoch": 0.18766, + "grad_norm": 0.8171993633039795, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 18766 + }, + { + "epoch": 0.18767, + "grad_norm": 0.9657869377286594, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 18767 + }, + { + "epoch": 0.18768, + "grad_norm": 0.9090022975078961, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 18768 + }, + { + "epoch": 0.18769, + "grad_norm": 0.9511019143162034, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 18769 + }, + { + "epoch": 0.1877, + "grad_norm": 1.321200505733317, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 18770 + }, + { + "epoch": 0.18771, + "grad_norm": 1.0086891038310786, + "learning_rate": 0.003, + "loss": 4.075, + "step": 18771 + }, + { + "epoch": 0.18772, + "grad_norm": 1.0292056359035422, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 18772 + }, + { + "epoch": 0.18773, + "grad_norm": 1.0027974901736976, + "learning_rate": 0.003, + "loss": 4.063, + "step": 18773 + }, + { + "epoch": 0.18774, + "grad_norm": 0.9186397891563183, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 18774 + }, + { + "epoch": 0.18775, + "grad_norm": 0.8641692292710332, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 18775 + }, + { + "epoch": 0.18776, + "grad_norm": 0.975433643335552, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 18776 + }, + { + "epoch": 0.18777, + "grad_norm": 1.2001710586008185, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 18777 + }, + { + "epoch": 0.18778, + "grad_norm": 1.0769574597087428, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 18778 + }, + { + "epoch": 0.18779, + "grad_norm": 1.2416383236749184, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 18779 + }, + { + "epoch": 0.1878, + "grad_norm": 0.9573292847462337, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 18780 + }, + { + "epoch": 0.18781, + "grad_norm": 0.8508329101007622, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 18781 + }, + { + "epoch": 0.18782, + "grad_norm": 0.912493073468906, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 18782 + }, + { + "epoch": 0.18783, + "grad_norm": 1.2003358511888564, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 18783 + }, + { + "epoch": 0.18784, + "grad_norm": 0.9945345559858317, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 18784 + }, + { + "epoch": 0.18785, + "grad_norm": 1.1800725879732161, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 18785 + }, + { + "epoch": 0.18786, + "grad_norm": 0.7792918851267874, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 18786 + }, + { + "epoch": 0.18787, + "grad_norm": 0.8432693376592952, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 18787 + }, + { + "epoch": 0.18788, + "grad_norm": 1.142513739958872, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 18788 + }, + { + "epoch": 0.18789, + "grad_norm": 1.1063784420657492, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 18789 + }, + { + "epoch": 0.1879, + "grad_norm": 0.9550275975432092, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 18790 + }, + { + "epoch": 0.18791, + "grad_norm": 1.0864369632453115, + "learning_rate": 0.003, + "loss": 4.04, + "step": 18791 + }, + { + "epoch": 0.18792, + "grad_norm": 1.22815369362515, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 18792 + }, + { + "epoch": 0.18793, + "grad_norm": 0.8465318088505633, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 18793 + }, + { + "epoch": 0.18794, + "grad_norm": 0.7398910891163203, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 18794 + }, + { + "epoch": 0.18795, + "grad_norm": 0.8340887687857778, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 18795 + }, + { + "epoch": 0.18796, + "grad_norm": 0.9527877588279003, + "learning_rate": 0.003, + "loss": 4.081, + "step": 18796 + }, + { + "epoch": 0.18797, + "grad_norm": 1.1886123129128123, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 18797 + }, + { + "epoch": 0.18798, + "grad_norm": 0.8504325147000925, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 18798 + }, + { + "epoch": 0.18799, + "grad_norm": 1.00917585982796, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 18799 + }, + { + "epoch": 0.188, + "grad_norm": 1.1161635793483478, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 18800 + }, + { + "epoch": 0.18801, + "grad_norm": 0.9889948902519162, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 18801 + }, + { + "epoch": 0.18802, + "grad_norm": 1.0755282781449103, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 18802 + }, + { + "epoch": 0.18803, + "grad_norm": 0.9862872626351783, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 18803 + }, + { + "epoch": 0.18804, + "grad_norm": 1.2148979879712134, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 18804 + }, + { + "epoch": 0.18805, + "grad_norm": 1.1194509217545878, + "learning_rate": 0.003, + "loss": 4.097, + "step": 18805 + }, + { + "epoch": 0.18806, + "grad_norm": 1.120377942582383, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 18806 + }, + { + "epoch": 0.18807, + "grad_norm": 1.0573766390691384, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 18807 + }, + { + "epoch": 0.18808, + "grad_norm": 0.9931334053811555, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 18808 + }, + { + "epoch": 0.18809, + "grad_norm": 1.1968490064244406, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 18809 + }, + { + "epoch": 0.1881, + "grad_norm": 1.0182778576170457, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 18810 + }, + { + "epoch": 0.18811, + "grad_norm": 1.0333875589138457, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 18811 + }, + { + "epoch": 0.18812, + "grad_norm": 1.0240451890967468, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 18812 + }, + { + "epoch": 0.18813, + "grad_norm": 1.009575679384057, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 18813 + }, + { + "epoch": 0.18814, + "grad_norm": 1.0139910496236546, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 18814 + }, + { + "epoch": 0.18815, + "grad_norm": 0.9871602638692379, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 18815 + }, + { + "epoch": 0.18816, + "grad_norm": 1.1384388556814873, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 18816 + }, + { + "epoch": 0.18817, + "grad_norm": 1.0148468948681189, + "learning_rate": 0.003, + "loss": 4.1037, + "step": 18817 + }, + { + "epoch": 0.18818, + "grad_norm": 0.9294043549453705, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 18818 + }, + { + "epoch": 0.18819, + "grad_norm": 1.0903641811815323, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 18819 + }, + { + "epoch": 0.1882, + "grad_norm": 1.1058699788578823, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 18820 + }, + { + "epoch": 0.18821, + "grad_norm": 1.081607161154618, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 18821 + }, + { + "epoch": 0.18822, + "grad_norm": 0.93881619407255, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 18822 + }, + { + "epoch": 0.18823, + "grad_norm": 1.0641220705992507, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 18823 + }, + { + "epoch": 0.18824, + "grad_norm": 1.3532425786003126, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 18824 + }, + { + "epoch": 0.18825, + "grad_norm": 0.9153291936239856, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 18825 + }, + { + "epoch": 0.18826, + "grad_norm": 0.9273686177780207, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 18826 + }, + { + "epoch": 0.18827, + "grad_norm": 0.8664271239728104, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 18827 + }, + { + "epoch": 0.18828, + "grad_norm": 0.92445496975903, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 18828 + }, + { + "epoch": 0.18829, + "grad_norm": 1.128197708424454, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 18829 + }, + { + "epoch": 0.1883, + "grad_norm": 1.0799969355579806, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 18830 + }, + { + "epoch": 0.18831, + "grad_norm": 0.8952002877998991, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 18831 + }, + { + "epoch": 0.18832, + "grad_norm": 0.846254803401692, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 18832 + }, + { + "epoch": 0.18833, + "grad_norm": 0.8032361387964057, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 18833 + }, + { + "epoch": 0.18834, + "grad_norm": 0.9193212581361108, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 18834 + }, + { + "epoch": 0.18835, + "grad_norm": 1.0058854650928517, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 18835 + }, + { + "epoch": 0.18836, + "grad_norm": 1.02962146554577, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 18836 + }, + { + "epoch": 0.18837, + "grad_norm": 1.0583849675533041, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 18837 + }, + { + "epoch": 0.18838, + "grad_norm": 1.1742953109096503, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 18838 + }, + { + "epoch": 0.18839, + "grad_norm": 1.110597136172856, + "learning_rate": 0.003, + "loss": 4.062, + "step": 18839 + }, + { + "epoch": 0.1884, + "grad_norm": 0.9163569837554064, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 18840 + }, + { + "epoch": 0.18841, + "grad_norm": 1.0453442700544062, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 18841 + }, + { + "epoch": 0.18842, + "grad_norm": 1.2767894174792445, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 18842 + }, + { + "epoch": 0.18843, + "grad_norm": 1.112235100708234, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 18843 + }, + { + "epoch": 0.18844, + "grad_norm": 1.054634806565749, + "learning_rate": 0.003, + "loss": 4.0881, + "step": 18844 + }, + { + "epoch": 0.18845, + "grad_norm": 1.0147620396692933, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 18845 + }, + { + "epoch": 0.18846, + "grad_norm": 1.153142132127355, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 18846 + }, + { + "epoch": 0.18847, + "grad_norm": 1.0263233543041375, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 18847 + }, + { + "epoch": 0.18848, + "grad_norm": 1.284479006973712, + "learning_rate": 0.003, + "loss": 4.054, + "step": 18848 + }, + { + "epoch": 0.18849, + "grad_norm": 1.0959921196111522, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 18849 + }, + { + "epoch": 0.1885, + "grad_norm": 0.8993691828315986, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 18850 + }, + { + "epoch": 0.18851, + "grad_norm": 0.8031780506480188, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 18851 + }, + { + "epoch": 0.18852, + "grad_norm": 0.8452634806646274, + "learning_rate": 0.003, + "loss": 4.0938, + "step": 18852 + }, + { + "epoch": 0.18853, + "grad_norm": 0.9929130155263682, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 18853 + }, + { + "epoch": 0.18854, + "grad_norm": 1.0574142856014204, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 18854 + }, + { + "epoch": 0.18855, + "grad_norm": 1.190017917262752, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 18855 + }, + { + "epoch": 0.18856, + "grad_norm": 0.962085304585998, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 18856 + }, + { + "epoch": 0.18857, + "grad_norm": 1.0298605778996732, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 18857 + }, + { + "epoch": 0.18858, + "grad_norm": 1.2640109967526214, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 18858 + }, + { + "epoch": 0.18859, + "grad_norm": 0.8442413920625037, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 18859 + }, + { + "epoch": 0.1886, + "grad_norm": 0.7906183221572707, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 18860 + }, + { + "epoch": 0.18861, + "grad_norm": 0.9043859993808829, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 18861 + }, + { + "epoch": 0.18862, + "grad_norm": 0.9171515502429097, + "learning_rate": 0.003, + "loss": 4.077, + "step": 18862 + }, + { + "epoch": 0.18863, + "grad_norm": 0.9968504000822103, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 18863 + }, + { + "epoch": 0.18864, + "grad_norm": 1.1641344180075532, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 18864 + }, + { + "epoch": 0.18865, + "grad_norm": 1.181599761919911, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 18865 + }, + { + "epoch": 0.18866, + "grad_norm": 0.9256633541758716, + "learning_rate": 0.003, + "loss": 4.062, + "step": 18866 + }, + { + "epoch": 0.18867, + "grad_norm": 1.0355856508568018, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 18867 + }, + { + "epoch": 0.18868, + "grad_norm": 0.9981036296860731, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 18868 + }, + { + "epoch": 0.18869, + "grad_norm": 1.1625692496707913, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 18869 + }, + { + "epoch": 0.1887, + "grad_norm": 0.9240750156025256, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 18870 + }, + { + "epoch": 0.18871, + "grad_norm": 0.9546138956401394, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 18871 + }, + { + "epoch": 0.18872, + "grad_norm": 1.0828242651262658, + "learning_rate": 0.003, + "loss": 4.061, + "step": 18872 + }, + { + "epoch": 0.18873, + "grad_norm": 1.173744342977259, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 18873 + }, + { + "epoch": 0.18874, + "grad_norm": 0.897546437301953, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 18874 + }, + { + "epoch": 0.18875, + "grad_norm": 1.0026551274514595, + "learning_rate": 0.003, + "loss": 4.1004, + "step": 18875 + }, + { + "epoch": 0.18876, + "grad_norm": 1.084610284756247, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 18876 + }, + { + "epoch": 0.18877, + "grad_norm": 1.0156080361892885, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 18877 + }, + { + "epoch": 0.18878, + "grad_norm": 1.2401442165599565, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 18878 + }, + { + "epoch": 0.18879, + "grad_norm": 0.9180298133820756, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 18879 + }, + { + "epoch": 0.1888, + "grad_norm": 1.0474143726177612, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 18880 + }, + { + "epoch": 0.18881, + "grad_norm": 1.0559364283762078, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 18881 + }, + { + "epoch": 0.18882, + "grad_norm": 1.0824880900742433, + "learning_rate": 0.003, + "loss": 4.0912, + "step": 18882 + }, + { + "epoch": 0.18883, + "grad_norm": 0.9172057227633789, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 18883 + }, + { + "epoch": 0.18884, + "grad_norm": 0.9082255414979549, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 18884 + }, + { + "epoch": 0.18885, + "grad_norm": 0.9827157189196616, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 18885 + }, + { + "epoch": 0.18886, + "grad_norm": 1.0973710200534368, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 18886 + }, + { + "epoch": 0.18887, + "grad_norm": 1.084210850939232, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 18887 + }, + { + "epoch": 0.18888, + "grad_norm": 1.1616612688063899, + "learning_rate": 0.003, + "loss": 4.0854, + "step": 18888 + }, + { + "epoch": 0.18889, + "grad_norm": 1.0903147926274903, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 18889 + }, + { + "epoch": 0.1889, + "grad_norm": 1.023947731885181, + "learning_rate": 0.003, + "loss": 4.078, + "step": 18890 + }, + { + "epoch": 0.18891, + "grad_norm": 0.9464050613396829, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 18891 + }, + { + "epoch": 0.18892, + "grad_norm": 1.0497239302247283, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 18892 + }, + { + "epoch": 0.18893, + "grad_norm": 1.0437064289007674, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 18893 + }, + { + "epoch": 0.18894, + "grad_norm": 1.040288243725371, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 18894 + }, + { + "epoch": 0.18895, + "grad_norm": 1.0690685843831285, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 18895 + }, + { + "epoch": 0.18896, + "grad_norm": 0.8724726224661488, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 18896 + }, + { + "epoch": 0.18897, + "grad_norm": 0.9403037027943237, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 18897 + }, + { + "epoch": 0.18898, + "grad_norm": 0.9665236863021748, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 18898 + }, + { + "epoch": 0.18899, + "grad_norm": 1.017725446150183, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 18899 + }, + { + "epoch": 0.189, + "grad_norm": 1.0403333166543673, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 18900 + }, + { + "epoch": 0.18901, + "grad_norm": 0.9553534066251744, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 18901 + }, + { + "epoch": 0.18902, + "grad_norm": 1.0834920297914512, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 18902 + }, + { + "epoch": 0.18903, + "grad_norm": 1.2632623666127873, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 18903 + }, + { + "epoch": 0.18904, + "grad_norm": 0.9636883823296541, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 18904 + }, + { + "epoch": 0.18905, + "grad_norm": 1.1532363014090505, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 18905 + }, + { + "epoch": 0.18906, + "grad_norm": 1.0565342368686448, + "learning_rate": 0.003, + "loss": 4.072, + "step": 18906 + }, + { + "epoch": 0.18907, + "grad_norm": 1.0277762748015322, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 18907 + }, + { + "epoch": 0.18908, + "grad_norm": 1.158928390824884, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 18908 + }, + { + "epoch": 0.18909, + "grad_norm": 0.9882038668297842, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 18909 + }, + { + "epoch": 0.1891, + "grad_norm": 1.0540293656555426, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 18910 + }, + { + "epoch": 0.18911, + "grad_norm": 1.0801368888382812, + "learning_rate": 0.003, + "loss": 4.1024, + "step": 18911 + }, + { + "epoch": 0.18912, + "grad_norm": 0.8565350344101561, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 18912 + }, + { + "epoch": 0.18913, + "grad_norm": 0.9190879026322392, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 18913 + }, + { + "epoch": 0.18914, + "grad_norm": 1.1281041159991403, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 18914 + }, + { + "epoch": 0.18915, + "grad_norm": 1.2021061275482066, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 18915 + }, + { + "epoch": 0.18916, + "grad_norm": 1.0806266707702947, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 18916 + }, + { + "epoch": 0.18917, + "grad_norm": 0.9980612748765478, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 18917 + }, + { + "epoch": 0.18918, + "grad_norm": 1.2194667403253487, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 18918 + }, + { + "epoch": 0.18919, + "grad_norm": 0.9492488198803907, + "learning_rate": 0.003, + "loss": 4.0881, + "step": 18919 + }, + { + "epoch": 0.1892, + "grad_norm": 0.9711293912660075, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 18920 + }, + { + "epoch": 0.18921, + "grad_norm": 1.2014580061278661, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 18921 + }, + { + "epoch": 0.18922, + "grad_norm": 0.9526721989964527, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 18922 + }, + { + "epoch": 0.18923, + "grad_norm": 0.9739524973613732, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 18923 + }, + { + "epoch": 0.18924, + "grad_norm": 1.1249562270638942, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 18924 + }, + { + "epoch": 0.18925, + "grad_norm": 1.0924465969570767, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 18925 + }, + { + "epoch": 0.18926, + "grad_norm": 0.8814652956043758, + "learning_rate": 0.003, + "loss": 4.0919, + "step": 18926 + }, + { + "epoch": 0.18927, + "grad_norm": 0.9181475729528666, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 18927 + }, + { + "epoch": 0.18928, + "grad_norm": 1.0268551565427015, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 18928 + }, + { + "epoch": 0.18929, + "grad_norm": 1.1886059448569883, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 18929 + }, + { + "epoch": 0.1893, + "grad_norm": 1.0214451785204388, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 18930 + }, + { + "epoch": 0.18931, + "grad_norm": 1.0511082685578708, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 18931 + }, + { + "epoch": 0.18932, + "grad_norm": 1.1020555256123672, + "learning_rate": 0.003, + "loss": 4.1079, + "step": 18932 + }, + { + "epoch": 0.18933, + "grad_norm": 0.9435295866352518, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 18933 + }, + { + "epoch": 0.18934, + "grad_norm": 1.0405539787752724, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 18934 + }, + { + "epoch": 0.18935, + "grad_norm": 1.0549119197848698, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 18935 + }, + { + "epoch": 0.18936, + "grad_norm": 1.1240798351647394, + "learning_rate": 0.003, + "loss": 4.0942, + "step": 18936 + }, + { + "epoch": 0.18937, + "grad_norm": 1.1228641710743612, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 18937 + }, + { + "epoch": 0.18938, + "grad_norm": 1.0920031030921848, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 18938 + }, + { + "epoch": 0.18939, + "grad_norm": 1.1138225212760167, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 18939 + }, + { + "epoch": 0.1894, + "grad_norm": 1.2005843094749384, + "learning_rate": 0.003, + "loss": 4.0959, + "step": 18940 + }, + { + "epoch": 0.18941, + "grad_norm": 0.9596756604223748, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 18941 + }, + { + "epoch": 0.18942, + "grad_norm": 0.9914967588367813, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 18942 + }, + { + "epoch": 0.18943, + "grad_norm": 1.0565295926238945, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 18943 + }, + { + "epoch": 0.18944, + "grad_norm": 0.8792268898496398, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 18944 + }, + { + "epoch": 0.18945, + "grad_norm": 0.8283893992576444, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 18945 + }, + { + "epoch": 0.18946, + "grad_norm": 0.904144338071763, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 18946 + }, + { + "epoch": 0.18947, + "grad_norm": 0.9402881675315479, + "learning_rate": 0.003, + "loss": 4.051, + "step": 18947 + }, + { + "epoch": 0.18948, + "grad_norm": 1.1983863037019837, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 18948 + }, + { + "epoch": 0.18949, + "grad_norm": 1.346192102118635, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 18949 + }, + { + "epoch": 0.1895, + "grad_norm": 0.9000420488924705, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 18950 + }, + { + "epoch": 0.18951, + "grad_norm": 0.9966938266799402, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 18951 + }, + { + "epoch": 0.18952, + "grad_norm": 0.9937385610371497, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 18952 + }, + { + "epoch": 0.18953, + "grad_norm": 1.1138394982604318, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 18953 + }, + { + "epoch": 0.18954, + "grad_norm": 1.314027643091253, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 18954 + }, + { + "epoch": 0.18955, + "grad_norm": 0.9191477466552569, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 18955 + }, + { + "epoch": 0.18956, + "grad_norm": 0.9524140525728139, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 18956 + }, + { + "epoch": 0.18957, + "grad_norm": 1.0962731083362354, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 18957 + }, + { + "epoch": 0.18958, + "grad_norm": 0.9150623579530411, + "learning_rate": 0.003, + "loss": 4.041, + "step": 18958 + }, + { + "epoch": 0.18959, + "grad_norm": 0.8936202381199325, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 18959 + }, + { + "epoch": 0.1896, + "grad_norm": 0.9385192908340084, + "learning_rate": 0.003, + "loss": 4.057, + "step": 18960 + }, + { + "epoch": 0.18961, + "grad_norm": 1.1083131065856682, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 18961 + }, + { + "epoch": 0.18962, + "grad_norm": 1.1153342782816245, + "learning_rate": 0.003, + "loss": 4.0954, + "step": 18962 + }, + { + "epoch": 0.18963, + "grad_norm": 0.8359261320585496, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 18963 + }, + { + "epoch": 0.18964, + "grad_norm": 0.7740038921319424, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 18964 + }, + { + "epoch": 0.18965, + "grad_norm": 0.8405141505105562, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 18965 + }, + { + "epoch": 0.18966, + "grad_norm": 0.837668343809781, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 18966 + }, + { + "epoch": 0.18967, + "grad_norm": 0.8975825902448171, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 18967 + }, + { + "epoch": 0.18968, + "grad_norm": 1.0632793750741536, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 18968 + }, + { + "epoch": 0.18969, + "grad_norm": 1.0602837007917518, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 18969 + }, + { + "epoch": 0.1897, + "grad_norm": 1.0671482386299562, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 18970 + }, + { + "epoch": 0.18971, + "grad_norm": 1.093459592323205, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 18971 + }, + { + "epoch": 0.18972, + "grad_norm": 1.019710872045341, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 18972 + }, + { + "epoch": 0.18973, + "grad_norm": 1.2594181729970262, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 18973 + }, + { + "epoch": 0.18974, + "grad_norm": 0.9477196193571867, + "learning_rate": 0.003, + "loss": 4.077, + "step": 18974 + }, + { + "epoch": 0.18975, + "grad_norm": 0.9048835433833766, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 18975 + }, + { + "epoch": 0.18976, + "grad_norm": 0.9574992752715522, + "learning_rate": 0.003, + "loss": 4.0871, + "step": 18976 + }, + { + "epoch": 0.18977, + "grad_norm": 1.1287328402894115, + "learning_rate": 0.003, + "loss": 4.0915, + "step": 18977 + }, + { + "epoch": 0.18978, + "grad_norm": 1.1140175461488955, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 18978 + }, + { + "epoch": 0.18979, + "grad_norm": 1.2672198152936929, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 18979 + }, + { + "epoch": 0.1898, + "grad_norm": 0.9863770448867728, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 18980 + }, + { + "epoch": 0.18981, + "grad_norm": 1.0406678539621108, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 18981 + }, + { + "epoch": 0.18982, + "grad_norm": 1.0691811171882206, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 18982 + }, + { + "epoch": 0.18983, + "grad_norm": 1.0617357861743053, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 18983 + }, + { + "epoch": 0.18984, + "grad_norm": 1.155728176780655, + "learning_rate": 0.003, + "loss": 4.066, + "step": 18984 + }, + { + "epoch": 0.18985, + "grad_norm": 0.9670353264178342, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 18985 + }, + { + "epoch": 0.18986, + "grad_norm": 1.1140816411343295, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 18986 + }, + { + "epoch": 0.18987, + "grad_norm": 1.0141494421772637, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 18987 + }, + { + "epoch": 0.18988, + "grad_norm": 1.1279295798634588, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 18988 + }, + { + "epoch": 0.18989, + "grad_norm": 0.9125721098507574, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 18989 + }, + { + "epoch": 0.1899, + "grad_norm": 0.8595128783315186, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 18990 + }, + { + "epoch": 0.18991, + "grad_norm": 0.8471799218150573, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 18991 + }, + { + "epoch": 0.18992, + "grad_norm": 0.7128631361368887, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 18992 + }, + { + "epoch": 0.18993, + "grad_norm": 0.8229666105089752, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 18993 + }, + { + "epoch": 0.18994, + "grad_norm": 1.021910438443167, + "learning_rate": 0.003, + "loss": 4.07, + "step": 18994 + }, + { + "epoch": 0.18995, + "grad_norm": 1.1358960090042656, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 18995 + }, + { + "epoch": 0.18996, + "grad_norm": 0.7770323963801032, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 18996 + }, + { + "epoch": 0.18997, + "grad_norm": 0.8300497909163747, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 18997 + }, + { + "epoch": 0.18998, + "grad_norm": 0.9419314417153006, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 18998 + }, + { + "epoch": 0.18999, + "grad_norm": 1.1917307851936518, + "learning_rate": 0.003, + "loss": 4.046, + "step": 18999 + }, + { + "epoch": 0.19, + "grad_norm": 1.0905596433086717, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 19000 + }, + { + "epoch": 0.19001, + "grad_norm": 1.2416221158700989, + "learning_rate": 0.003, + "loss": 4.075, + "step": 19001 + }, + { + "epoch": 0.19002, + "grad_norm": 0.986083995092308, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 19002 + }, + { + "epoch": 0.19003, + "grad_norm": 1.2678540579964068, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 19003 + }, + { + "epoch": 0.19004, + "grad_norm": 0.9301998791148526, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 19004 + }, + { + "epoch": 0.19005, + "grad_norm": 0.9763803987964047, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 19005 + }, + { + "epoch": 0.19006, + "grad_norm": 1.2670224401854455, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 19006 + }, + { + "epoch": 0.19007, + "grad_norm": 0.8409672726924401, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 19007 + }, + { + "epoch": 0.19008, + "grad_norm": 0.929883504267793, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 19008 + }, + { + "epoch": 0.19009, + "grad_norm": 1.1790005753868864, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 19009 + }, + { + "epoch": 0.1901, + "grad_norm": 1.0610809427204024, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 19010 + }, + { + "epoch": 0.19011, + "grad_norm": 1.1465936645329606, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 19011 + }, + { + "epoch": 0.19012, + "grad_norm": 1.0385829598829748, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 19012 + }, + { + "epoch": 0.19013, + "grad_norm": 0.937627494184569, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 19013 + }, + { + "epoch": 0.19014, + "grad_norm": 0.908300468330221, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 19014 + }, + { + "epoch": 0.19015, + "grad_norm": 1.052713138088683, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 19015 + }, + { + "epoch": 0.19016, + "grad_norm": 1.1400074347787965, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 19016 + }, + { + "epoch": 0.19017, + "grad_norm": 0.9516332670190728, + "learning_rate": 0.003, + "loss": 4.0854, + "step": 19017 + }, + { + "epoch": 0.19018, + "grad_norm": 0.9502763283876596, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 19018 + }, + { + "epoch": 0.19019, + "grad_norm": 1.0527567374173608, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 19019 + }, + { + "epoch": 0.1902, + "grad_norm": 0.9745268992907389, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 19020 + }, + { + "epoch": 0.19021, + "grad_norm": 1.0271682126686348, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 19021 + }, + { + "epoch": 0.19022, + "grad_norm": 1.06110691486001, + "learning_rate": 0.003, + "loss": 4.101, + "step": 19022 + }, + { + "epoch": 0.19023, + "grad_norm": 0.9845777157708752, + "learning_rate": 0.003, + "loss": 4.1165, + "step": 19023 + }, + { + "epoch": 0.19024, + "grad_norm": 1.2422070506385814, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 19024 + }, + { + "epoch": 0.19025, + "grad_norm": 0.8504529277255194, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 19025 + }, + { + "epoch": 0.19026, + "grad_norm": 0.9242771772180569, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 19026 + }, + { + "epoch": 0.19027, + "grad_norm": 1.086797195110357, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 19027 + }, + { + "epoch": 0.19028, + "grad_norm": 1.2480235375171684, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 19028 + }, + { + "epoch": 0.19029, + "grad_norm": 1.2653767337364723, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 19029 + }, + { + "epoch": 0.1903, + "grad_norm": 1.091663409167593, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 19030 + }, + { + "epoch": 0.19031, + "grad_norm": 1.0835714442360698, + "learning_rate": 0.003, + "loss": 4.058, + "step": 19031 + }, + { + "epoch": 0.19032, + "grad_norm": 1.0606914506358853, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 19032 + }, + { + "epoch": 0.19033, + "grad_norm": 1.0339514913909345, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 19033 + }, + { + "epoch": 0.19034, + "grad_norm": 1.0759193803218112, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 19034 + }, + { + "epoch": 0.19035, + "grad_norm": 0.9338906234448523, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 19035 + }, + { + "epoch": 0.19036, + "grad_norm": 1.0504410414140208, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 19036 + }, + { + "epoch": 0.19037, + "grad_norm": 1.1871855721014555, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 19037 + }, + { + "epoch": 0.19038, + "grad_norm": 1.0546673729412128, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 19038 + }, + { + "epoch": 0.19039, + "grad_norm": 1.123558473633067, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 19039 + }, + { + "epoch": 0.1904, + "grad_norm": 1.158495002040774, + "learning_rate": 0.003, + "loss": 4.045, + "step": 19040 + }, + { + "epoch": 0.19041, + "grad_norm": 0.8066000208240048, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 19041 + }, + { + "epoch": 0.19042, + "grad_norm": 0.866512871041666, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 19042 + }, + { + "epoch": 0.19043, + "grad_norm": 1.1189412151476774, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 19043 + }, + { + "epoch": 0.19044, + "grad_norm": 1.060439472404301, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 19044 + }, + { + "epoch": 0.19045, + "grad_norm": 1.08312614431398, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 19045 + }, + { + "epoch": 0.19046, + "grad_norm": 1.0007506956302583, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 19046 + }, + { + "epoch": 0.19047, + "grad_norm": 1.3402388408876749, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 19047 + }, + { + "epoch": 0.19048, + "grad_norm": 0.92561538322619, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 19048 + }, + { + "epoch": 0.19049, + "grad_norm": 1.3261229666445513, + "learning_rate": 0.003, + "loss": 4.0927, + "step": 19049 + }, + { + "epoch": 0.1905, + "grad_norm": 0.9765232901464445, + "learning_rate": 0.003, + "loss": 4.0907, + "step": 19050 + }, + { + "epoch": 0.19051, + "grad_norm": 0.9463462181482688, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 19051 + }, + { + "epoch": 0.19052, + "grad_norm": 1.0780802429940066, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 19052 + }, + { + "epoch": 0.19053, + "grad_norm": 0.9565413276495962, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 19053 + }, + { + "epoch": 0.19054, + "grad_norm": 1.0897093587605893, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 19054 + }, + { + "epoch": 0.19055, + "grad_norm": 1.1276835104661753, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 19055 + }, + { + "epoch": 0.19056, + "grad_norm": 1.0790519991790577, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 19056 + }, + { + "epoch": 0.19057, + "grad_norm": 1.1432327565811187, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 19057 + }, + { + "epoch": 0.19058, + "grad_norm": 1.2153861847307517, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 19058 + }, + { + "epoch": 0.19059, + "grad_norm": 1.1146138477698397, + "learning_rate": 0.003, + "loss": 4.073, + "step": 19059 + }, + { + "epoch": 0.1906, + "grad_norm": 0.8471988211516281, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 19060 + }, + { + "epoch": 0.19061, + "grad_norm": 0.8061507488961706, + "learning_rate": 0.003, + "loss": 4.062, + "step": 19061 + }, + { + "epoch": 0.19062, + "grad_norm": 0.8018481223790215, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 19062 + }, + { + "epoch": 0.19063, + "grad_norm": 0.7386996273419536, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 19063 + }, + { + "epoch": 0.19064, + "grad_norm": 0.888467760866657, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 19064 + }, + { + "epoch": 0.19065, + "grad_norm": 1.1245244432709685, + "learning_rate": 0.003, + "loss": 4.077, + "step": 19065 + }, + { + "epoch": 0.19066, + "grad_norm": 1.0569182427962698, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 19066 + }, + { + "epoch": 0.19067, + "grad_norm": 1.278330052360473, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 19067 + }, + { + "epoch": 0.19068, + "grad_norm": 1.0747837011163641, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 19068 + }, + { + "epoch": 0.19069, + "grad_norm": 1.1580580355470873, + "learning_rate": 0.003, + "loss": 4.074, + "step": 19069 + }, + { + "epoch": 0.1907, + "grad_norm": 1.1084424047547385, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 19070 + }, + { + "epoch": 0.19071, + "grad_norm": 1.0065788362299766, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 19071 + }, + { + "epoch": 0.19072, + "grad_norm": 0.9480110973160153, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 19072 + }, + { + "epoch": 0.19073, + "grad_norm": 0.9932287427517996, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 19073 + }, + { + "epoch": 0.19074, + "grad_norm": 1.1114540883953414, + "learning_rate": 0.003, + "loss": 4.0888, + "step": 19074 + }, + { + "epoch": 0.19075, + "grad_norm": 0.9752522227544241, + "learning_rate": 0.003, + "loss": 4.068, + "step": 19075 + }, + { + "epoch": 0.19076, + "grad_norm": 0.9640229231633557, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 19076 + }, + { + "epoch": 0.19077, + "grad_norm": 1.0978466662380697, + "learning_rate": 0.003, + "loss": 4.0988, + "step": 19077 + }, + { + "epoch": 0.19078, + "grad_norm": 1.1410938298632987, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 19078 + }, + { + "epoch": 0.19079, + "grad_norm": 1.1194694698748031, + "learning_rate": 0.003, + "loss": 4.0874, + "step": 19079 + }, + { + "epoch": 0.1908, + "grad_norm": 0.9633510401517809, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 19080 + }, + { + "epoch": 0.19081, + "grad_norm": 1.4950800256864205, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 19081 + }, + { + "epoch": 0.19082, + "grad_norm": 0.8936023918536756, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 19082 + }, + { + "epoch": 0.19083, + "grad_norm": 0.8377961747211358, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 19083 + }, + { + "epoch": 0.19084, + "grad_norm": 1.0753425477836072, + "learning_rate": 0.003, + "loss": 4.07, + "step": 19084 + }, + { + "epoch": 0.19085, + "grad_norm": 1.0278486432939995, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 19085 + }, + { + "epoch": 0.19086, + "grad_norm": 1.263245484316211, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 19086 + }, + { + "epoch": 0.19087, + "grad_norm": 1.3147023761755212, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 19087 + }, + { + "epoch": 0.19088, + "grad_norm": 0.9659983728421012, + "learning_rate": 0.003, + "loss": 4.0926, + "step": 19088 + }, + { + "epoch": 0.19089, + "grad_norm": 0.977057060267963, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 19089 + }, + { + "epoch": 0.1909, + "grad_norm": 1.0835561493800565, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 19090 + }, + { + "epoch": 0.19091, + "grad_norm": 1.188866903102991, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 19091 + }, + { + "epoch": 0.19092, + "grad_norm": 0.90879733260125, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 19092 + }, + { + "epoch": 0.19093, + "grad_norm": 0.900542400780217, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 19093 + }, + { + "epoch": 0.19094, + "grad_norm": 0.9332066239607978, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 19094 + }, + { + "epoch": 0.19095, + "grad_norm": 0.9868211514667526, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 19095 + }, + { + "epoch": 0.19096, + "grad_norm": 1.055585153331953, + "learning_rate": 0.003, + "loss": 4.073, + "step": 19096 + }, + { + "epoch": 0.19097, + "grad_norm": 1.1827968055355254, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 19097 + }, + { + "epoch": 0.19098, + "grad_norm": 0.8468615913094318, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 19098 + }, + { + "epoch": 0.19099, + "grad_norm": 0.8152298253102531, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 19099 + }, + { + "epoch": 0.191, + "grad_norm": 0.8519177835795184, + "learning_rate": 0.003, + "loss": 4.0856, + "step": 19100 + }, + { + "epoch": 0.19101, + "grad_norm": 0.9732750577477809, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 19101 + }, + { + "epoch": 0.19102, + "grad_norm": 1.1239912770009184, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 19102 + }, + { + "epoch": 0.19103, + "grad_norm": 1.130093434883689, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 19103 + }, + { + "epoch": 0.19104, + "grad_norm": 1.2196185496778085, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 19104 + }, + { + "epoch": 0.19105, + "grad_norm": 0.7301861965303266, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 19105 + }, + { + "epoch": 0.19106, + "grad_norm": 0.7022577676582528, + "learning_rate": 0.003, + "loss": 4.047, + "step": 19106 + }, + { + "epoch": 0.19107, + "grad_norm": 0.7588882843350907, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 19107 + }, + { + "epoch": 0.19108, + "grad_norm": 0.8702529271737249, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 19108 + }, + { + "epoch": 0.19109, + "grad_norm": 0.9376437204020522, + "learning_rate": 0.003, + "loss": 4.0868, + "step": 19109 + }, + { + "epoch": 0.1911, + "grad_norm": 1.0327110083015956, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 19110 + }, + { + "epoch": 0.19111, + "grad_norm": 1.2760750698492056, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 19111 + }, + { + "epoch": 0.19112, + "grad_norm": 0.8678167338033316, + "learning_rate": 0.003, + "loss": 4.077, + "step": 19112 + }, + { + "epoch": 0.19113, + "grad_norm": 0.9180457127341991, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 19113 + }, + { + "epoch": 0.19114, + "grad_norm": 0.8532000765717144, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 19114 + }, + { + "epoch": 0.19115, + "grad_norm": 0.9708799294073973, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 19115 + }, + { + "epoch": 0.19116, + "grad_norm": 1.2750468101651242, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 19116 + }, + { + "epoch": 0.19117, + "grad_norm": 1.0276620605492297, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 19117 + }, + { + "epoch": 0.19118, + "grad_norm": 1.2985656993192, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 19118 + }, + { + "epoch": 0.19119, + "grad_norm": 1.1288569290731891, + "learning_rate": 0.003, + "loss": 4.1076, + "step": 19119 + }, + { + "epoch": 0.1912, + "grad_norm": 0.9808016742182016, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 19120 + }, + { + "epoch": 0.19121, + "grad_norm": 1.1918640073129692, + "learning_rate": 0.003, + "loss": 4.08, + "step": 19121 + }, + { + "epoch": 0.19122, + "grad_norm": 0.9008532313977722, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 19122 + }, + { + "epoch": 0.19123, + "grad_norm": 0.9001514120043417, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 19123 + }, + { + "epoch": 0.19124, + "grad_norm": 0.9914688417208597, + "learning_rate": 0.003, + "loss": 4.0961, + "step": 19124 + }, + { + "epoch": 0.19125, + "grad_norm": 1.1093246126537564, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 19125 + }, + { + "epoch": 0.19126, + "grad_norm": 1.0086246874431481, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 19126 + }, + { + "epoch": 0.19127, + "grad_norm": 1.1546388904792078, + "learning_rate": 0.003, + "loss": 4.075, + "step": 19127 + }, + { + "epoch": 0.19128, + "grad_norm": 0.8911984569627887, + "learning_rate": 0.003, + "loss": 4.061, + "step": 19128 + }, + { + "epoch": 0.19129, + "grad_norm": 1.0409993604025332, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 19129 + }, + { + "epoch": 0.1913, + "grad_norm": 1.1780000006626623, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 19130 + }, + { + "epoch": 0.19131, + "grad_norm": 0.9688184465716454, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 19131 + }, + { + "epoch": 0.19132, + "grad_norm": 0.9495660199026361, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 19132 + }, + { + "epoch": 0.19133, + "grad_norm": 1.0614203556710307, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 19133 + }, + { + "epoch": 0.19134, + "grad_norm": 1.0810249925292732, + "learning_rate": 0.003, + "loss": 4.101, + "step": 19134 + }, + { + "epoch": 0.19135, + "grad_norm": 1.104885035146321, + "learning_rate": 0.003, + "loss": 4.0856, + "step": 19135 + }, + { + "epoch": 0.19136, + "grad_norm": 1.203923930897022, + "learning_rate": 0.003, + "loss": 4.066, + "step": 19136 + }, + { + "epoch": 0.19137, + "grad_norm": 1.2111552781116912, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 19137 + }, + { + "epoch": 0.19138, + "grad_norm": 1.0918391213762777, + "learning_rate": 0.003, + "loss": 4.072, + "step": 19138 + }, + { + "epoch": 0.19139, + "grad_norm": 1.0190820348762673, + "learning_rate": 0.003, + "loss": 4.0905, + "step": 19139 + }, + { + "epoch": 0.1914, + "grad_norm": 1.265414085438354, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 19140 + }, + { + "epoch": 0.19141, + "grad_norm": 0.8390892619529539, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 19141 + }, + { + "epoch": 0.19142, + "grad_norm": 0.760837142598633, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 19142 + }, + { + "epoch": 0.19143, + "grad_norm": 0.7549784415420118, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 19143 + }, + { + "epoch": 0.19144, + "grad_norm": 0.8136149524331792, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 19144 + }, + { + "epoch": 0.19145, + "grad_norm": 0.9921518642116638, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 19145 + }, + { + "epoch": 0.19146, + "grad_norm": 1.1051745726683495, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 19146 + }, + { + "epoch": 0.19147, + "grad_norm": 1.3258956218060172, + "learning_rate": 0.003, + "loss": 4.064, + "step": 19147 + }, + { + "epoch": 0.19148, + "grad_norm": 0.9532404704192361, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 19148 + }, + { + "epoch": 0.19149, + "grad_norm": 0.8757341012250994, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 19149 + }, + { + "epoch": 0.1915, + "grad_norm": 0.9160351742685405, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 19150 + }, + { + "epoch": 0.19151, + "grad_norm": 1.0427742324776192, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 19151 + }, + { + "epoch": 0.19152, + "grad_norm": 1.0514296787940591, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 19152 + }, + { + "epoch": 0.19153, + "grad_norm": 1.240808226920762, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 19153 + }, + { + "epoch": 0.19154, + "grad_norm": 1.0550087864518818, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 19154 + }, + { + "epoch": 0.19155, + "grad_norm": 0.9118191801432967, + "learning_rate": 0.003, + "loss": 4.029, + "step": 19155 + }, + { + "epoch": 0.19156, + "grad_norm": 0.7889774566100356, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 19156 + }, + { + "epoch": 0.19157, + "grad_norm": 0.9659906737124742, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 19157 + }, + { + "epoch": 0.19158, + "grad_norm": 1.322454114506548, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 19158 + }, + { + "epoch": 0.19159, + "grad_norm": 0.8517802877845722, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 19159 + }, + { + "epoch": 0.1916, + "grad_norm": 1.2261205719066928, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 19160 + }, + { + "epoch": 0.19161, + "grad_norm": 1.1684632963154713, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 19161 + }, + { + "epoch": 0.19162, + "grad_norm": 0.8416766997438629, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 19162 + }, + { + "epoch": 0.19163, + "grad_norm": 0.8717156238801331, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 19163 + }, + { + "epoch": 0.19164, + "grad_norm": 0.964842248080158, + "learning_rate": 0.003, + "loss": 4.1085, + "step": 19164 + }, + { + "epoch": 0.19165, + "grad_norm": 1.0145401677313641, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 19165 + }, + { + "epoch": 0.19166, + "grad_norm": 1.0682185790948437, + "learning_rate": 0.003, + "loss": 4.056, + "step": 19166 + }, + { + "epoch": 0.19167, + "grad_norm": 1.4658228711822456, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 19167 + }, + { + "epoch": 0.19168, + "grad_norm": 0.7310215409392945, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 19168 + }, + { + "epoch": 0.19169, + "grad_norm": 0.812194720729166, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 19169 + }, + { + "epoch": 0.1917, + "grad_norm": 0.9288841907657722, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 19170 + }, + { + "epoch": 0.19171, + "grad_norm": 1.0288639820513283, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 19171 + }, + { + "epoch": 0.19172, + "grad_norm": 1.4224585332801885, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 19172 + }, + { + "epoch": 0.19173, + "grad_norm": 0.8975095741964783, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 19173 + }, + { + "epoch": 0.19174, + "grad_norm": 0.91879401273857, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 19174 + }, + { + "epoch": 0.19175, + "grad_norm": 1.079460521565124, + "learning_rate": 0.003, + "loss": 4.1023, + "step": 19175 + }, + { + "epoch": 0.19176, + "grad_norm": 1.1543799176741782, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 19176 + }, + { + "epoch": 0.19177, + "grad_norm": 1.0409988231921665, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 19177 + }, + { + "epoch": 0.19178, + "grad_norm": 1.1337078981838211, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 19178 + }, + { + "epoch": 0.19179, + "grad_norm": 1.0761767352995197, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 19179 + }, + { + "epoch": 0.1918, + "grad_norm": 1.0117427280421019, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 19180 + }, + { + "epoch": 0.19181, + "grad_norm": 0.9594270414039656, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 19181 + }, + { + "epoch": 0.19182, + "grad_norm": 0.9280724532740865, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 19182 + }, + { + "epoch": 0.19183, + "grad_norm": 1.0548307715703285, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 19183 + }, + { + "epoch": 0.19184, + "grad_norm": 0.9935608828196411, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 19184 + }, + { + "epoch": 0.19185, + "grad_norm": 1.2873622378256648, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 19185 + }, + { + "epoch": 0.19186, + "grad_norm": 1.0543545074284166, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 19186 + }, + { + "epoch": 0.19187, + "grad_norm": 1.1006697886657537, + "learning_rate": 0.003, + "loss": 4.0891, + "step": 19187 + }, + { + "epoch": 0.19188, + "grad_norm": 1.2262701082877334, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 19188 + }, + { + "epoch": 0.19189, + "grad_norm": 0.8782386723807774, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 19189 + }, + { + "epoch": 0.1919, + "grad_norm": 1.1609374195854851, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 19190 + }, + { + "epoch": 0.19191, + "grad_norm": 0.8138831247913053, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 19191 + }, + { + "epoch": 0.19192, + "grad_norm": 0.9143657641877183, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 19192 + }, + { + "epoch": 0.19193, + "grad_norm": 1.3563433111999308, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 19193 + }, + { + "epoch": 0.19194, + "grad_norm": 1.12762635518305, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 19194 + }, + { + "epoch": 0.19195, + "grad_norm": 1.0034158836342875, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 19195 + }, + { + "epoch": 0.19196, + "grad_norm": 1.1566779995069918, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 19196 + }, + { + "epoch": 0.19197, + "grad_norm": 1.122368090543865, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 19197 + }, + { + "epoch": 0.19198, + "grad_norm": 0.861769648575424, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 19198 + }, + { + "epoch": 0.19199, + "grad_norm": 0.8372899854835236, + "learning_rate": 0.003, + "loss": 4.082, + "step": 19199 + }, + { + "epoch": 0.192, + "grad_norm": 0.8209186320173244, + "learning_rate": 0.003, + "loss": 4.0989, + "step": 19200 + }, + { + "epoch": 0.19201, + "grad_norm": 0.7176466184696128, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 19201 + }, + { + "epoch": 0.19202, + "grad_norm": 0.6498474185281552, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 19202 + }, + { + "epoch": 0.19203, + "grad_norm": 0.663922216953076, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 19203 + }, + { + "epoch": 0.19204, + "grad_norm": 0.8148757210839693, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 19204 + }, + { + "epoch": 0.19205, + "grad_norm": 1.076375186308871, + "learning_rate": 0.003, + "loss": 4.0919, + "step": 19205 + }, + { + "epoch": 0.19206, + "grad_norm": 1.0394353877185232, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 19206 + }, + { + "epoch": 0.19207, + "grad_norm": 1.0582204878830654, + "learning_rate": 0.003, + "loss": 4.079, + "step": 19207 + }, + { + "epoch": 0.19208, + "grad_norm": 1.0802773376433734, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 19208 + }, + { + "epoch": 0.19209, + "grad_norm": 1.0590259153004935, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 19209 + }, + { + "epoch": 0.1921, + "grad_norm": 1.211958870880975, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 19210 + }, + { + "epoch": 0.19211, + "grad_norm": 0.9916890774383754, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 19211 + }, + { + "epoch": 0.19212, + "grad_norm": 1.2128386145357724, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 19212 + }, + { + "epoch": 0.19213, + "grad_norm": 0.8752602975916279, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 19213 + }, + { + "epoch": 0.19214, + "grad_norm": 1.1292534891112187, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 19214 + }, + { + "epoch": 0.19215, + "grad_norm": 1.256744467815811, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 19215 + }, + { + "epoch": 0.19216, + "grad_norm": 0.9374740221303403, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 19216 + }, + { + "epoch": 0.19217, + "grad_norm": 0.9337064854957995, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 19217 + }, + { + "epoch": 0.19218, + "grad_norm": 1.077091360184908, + "learning_rate": 0.003, + "loss": 4.1058, + "step": 19218 + }, + { + "epoch": 0.19219, + "grad_norm": 1.026832485699907, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 19219 + }, + { + "epoch": 0.1922, + "grad_norm": 0.8401325658451464, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 19220 + }, + { + "epoch": 0.19221, + "grad_norm": 0.8169290475064839, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 19221 + }, + { + "epoch": 0.19222, + "grad_norm": 0.8741552454133128, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 19222 + }, + { + "epoch": 0.19223, + "grad_norm": 1.023186866376128, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 19223 + }, + { + "epoch": 0.19224, + "grad_norm": 1.0361683472113443, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 19224 + }, + { + "epoch": 0.19225, + "grad_norm": 1.2044072745561236, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 19225 + }, + { + "epoch": 0.19226, + "grad_norm": 1.0478834052455794, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 19226 + }, + { + "epoch": 0.19227, + "grad_norm": 1.3281732943474251, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 19227 + }, + { + "epoch": 0.19228, + "grad_norm": 0.9211000095988268, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 19228 + }, + { + "epoch": 0.19229, + "grad_norm": 0.9355703191329742, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 19229 + }, + { + "epoch": 0.1923, + "grad_norm": 1.1625516398521143, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 19230 + }, + { + "epoch": 0.19231, + "grad_norm": 1.1557699816998988, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 19231 + }, + { + "epoch": 0.19232, + "grad_norm": 0.862042291319803, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 19232 + }, + { + "epoch": 0.19233, + "grad_norm": 0.9134015113845708, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 19233 + }, + { + "epoch": 0.19234, + "grad_norm": 1.2023115901718613, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 19234 + }, + { + "epoch": 0.19235, + "grad_norm": 0.9991770565177834, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 19235 + }, + { + "epoch": 0.19236, + "grad_norm": 1.2058074759637103, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 19236 + }, + { + "epoch": 0.19237, + "grad_norm": 1.1926171040225677, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 19237 + }, + { + "epoch": 0.19238, + "grad_norm": 0.9349966536471271, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 19238 + }, + { + "epoch": 0.19239, + "grad_norm": 0.9660715778947635, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 19239 + }, + { + "epoch": 0.1924, + "grad_norm": 1.1902188591951817, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 19240 + }, + { + "epoch": 0.19241, + "grad_norm": 1.0644182826967346, + "learning_rate": 0.003, + "loss": 4.061, + "step": 19241 + }, + { + "epoch": 0.19242, + "grad_norm": 1.0481871468762456, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 19242 + }, + { + "epoch": 0.19243, + "grad_norm": 1.163243477371277, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 19243 + }, + { + "epoch": 0.19244, + "grad_norm": 1.0599252891113256, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 19244 + }, + { + "epoch": 0.19245, + "grad_norm": 0.9429980169922736, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 19245 + }, + { + "epoch": 0.19246, + "grad_norm": 0.9779985883682354, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 19246 + }, + { + "epoch": 0.19247, + "grad_norm": 1.1171114355015708, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 19247 + }, + { + "epoch": 0.19248, + "grad_norm": 1.0909667883154663, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 19248 + }, + { + "epoch": 0.19249, + "grad_norm": 1.0446101938057621, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 19249 + }, + { + "epoch": 0.1925, + "grad_norm": 1.0672846311911968, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 19250 + }, + { + "epoch": 0.19251, + "grad_norm": 1.1570569089923861, + "learning_rate": 0.003, + "loss": 4.0845, + "step": 19251 + }, + { + "epoch": 0.19252, + "grad_norm": 1.0984945072530528, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 19252 + }, + { + "epoch": 0.19253, + "grad_norm": 1.1692654713530661, + "learning_rate": 0.003, + "loss": 4.08, + "step": 19253 + }, + { + "epoch": 0.19254, + "grad_norm": 0.9110960640560931, + "learning_rate": 0.003, + "loss": 4.1069, + "step": 19254 + }, + { + "epoch": 0.19255, + "grad_norm": 0.9121788639379829, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 19255 + }, + { + "epoch": 0.19256, + "grad_norm": 1.1861097376227865, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 19256 + }, + { + "epoch": 0.19257, + "grad_norm": 1.025856411758768, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 19257 + }, + { + "epoch": 0.19258, + "grad_norm": 1.0974918047396145, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 19258 + }, + { + "epoch": 0.19259, + "grad_norm": 1.208740600530569, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 19259 + }, + { + "epoch": 0.1926, + "grad_norm": 1.0180907699060784, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 19260 + }, + { + "epoch": 0.19261, + "grad_norm": 1.0409949399056804, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 19261 + }, + { + "epoch": 0.19262, + "grad_norm": 1.1710731235046912, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 19262 + }, + { + "epoch": 0.19263, + "grad_norm": 1.037946366394642, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 19263 + }, + { + "epoch": 0.19264, + "grad_norm": 1.228019218249223, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 19264 + }, + { + "epoch": 0.19265, + "grad_norm": 0.9363897372917479, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 19265 + }, + { + "epoch": 0.19266, + "grad_norm": 0.9113268796884856, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 19266 + }, + { + "epoch": 0.19267, + "grad_norm": 0.9435356628608942, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 19267 + }, + { + "epoch": 0.19268, + "grad_norm": 1.082500756338579, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 19268 + }, + { + "epoch": 0.19269, + "grad_norm": 0.9130023433366301, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 19269 + }, + { + "epoch": 0.1927, + "grad_norm": 1.0310903173987505, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 19270 + }, + { + "epoch": 0.19271, + "grad_norm": 0.9974442093129793, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 19271 + }, + { + "epoch": 0.19272, + "grad_norm": 1.109513073024803, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 19272 + }, + { + "epoch": 0.19273, + "grad_norm": 1.0770862364125098, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 19273 + }, + { + "epoch": 0.19274, + "grad_norm": 1.0680570129035563, + "learning_rate": 0.003, + "loss": 4.0938, + "step": 19274 + }, + { + "epoch": 0.19275, + "grad_norm": 1.0760003158354785, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 19275 + }, + { + "epoch": 0.19276, + "grad_norm": 1.168771984278787, + "learning_rate": 0.003, + "loss": 4.066, + "step": 19276 + }, + { + "epoch": 0.19277, + "grad_norm": 1.0717970810978439, + "learning_rate": 0.003, + "loss": 4.1033, + "step": 19277 + }, + { + "epoch": 0.19278, + "grad_norm": 0.9892669296502719, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 19278 + }, + { + "epoch": 0.19279, + "grad_norm": 0.8727253844520091, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 19279 + }, + { + "epoch": 0.1928, + "grad_norm": 0.9082340706895862, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 19280 + }, + { + "epoch": 0.19281, + "grad_norm": 1.1510868176031246, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 19281 + }, + { + "epoch": 0.19282, + "grad_norm": 1.2084108888785712, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 19282 + }, + { + "epoch": 0.19283, + "grad_norm": 0.9201301551350098, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 19283 + }, + { + "epoch": 0.19284, + "grad_norm": 0.8701175421663908, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 19284 + }, + { + "epoch": 0.19285, + "grad_norm": 0.9491570633798165, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 19285 + }, + { + "epoch": 0.19286, + "grad_norm": 1.0489514826291693, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 19286 + }, + { + "epoch": 0.19287, + "grad_norm": 1.117859177958715, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 19287 + }, + { + "epoch": 0.19288, + "grad_norm": 1.1317507486665668, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 19288 + }, + { + "epoch": 0.19289, + "grad_norm": 0.9814617702003932, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 19289 + }, + { + "epoch": 0.1929, + "grad_norm": 0.9817332800757131, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 19290 + }, + { + "epoch": 0.19291, + "grad_norm": 1.1892560772944654, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 19291 + }, + { + "epoch": 0.19292, + "grad_norm": 0.9175493300912014, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 19292 + }, + { + "epoch": 0.19293, + "grad_norm": 1.0018774499999568, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 19293 + }, + { + "epoch": 0.19294, + "grad_norm": 1.296662426634145, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 19294 + }, + { + "epoch": 0.19295, + "grad_norm": 0.7665984444973306, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 19295 + }, + { + "epoch": 0.19296, + "grad_norm": 0.7924650507875266, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 19296 + }, + { + "epoch": 0.19297, + "grad_norm": 0.9955679986633871, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 19297 + }, + { + "epoch": 0.19298, + "grad_norm": 1.1310573256134389, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 19298 + }, + { + "epoch": 0.19299, + "grad_norm": 1.2582494770422097, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 19299 + }, + { + "epoch": 0.193, + "grad_norm": 1.4638016734428776, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 19300 + }, + { + "epoch": 0.19301, + "grad_norm": 0.805234632063388, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 19301 + }, + { + "epoch": 0.19302, + "grad_norm": 0.9418025883758193, + "learning_rate": 0.003, + "loss": 4.063, + "step": 19302 + }, + { + "epoch": 0.19303, + "grad_norm": 1.173349198965408, + "learning_rate": 0.003, + "loss": 4.0819, + "step": 19303 + }, + { + "epoch": 0.19304, + "grad_norm": 1.164102316755723, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 19304 + }, + { + "epoch": 0.19305, + "grad_norm": 1.0473109521179562, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 19305 + }, + { + "epoch": 0.19306, + "grad_norm": 1.2471282771268135, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 19306 + }, + { + "epoch": 0.19307, + "grad_norm": 0.9180035347055766, + "learning_rate": 0.003, + "loss": 4.073, + "step": 19307 + }, + { + "epoch": 0.19308, + "grad_norm": 1.0467980397627155, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 19308 + }, + { + "epoch": 0.19309, + "grad_norm": 1.1004386999395495, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 19309 + }, + { + "epoch": 0.1931, + "grad_norm": 0.9633141804583791, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 19310 + }, + { + "epoch": 0.19311, + "grad_norm": 1.0771059324669294, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 19311 + }, + { + "epoch": 0.19312, + "grad_norm": 1.0864903401608446, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 19312 + }, + { + "epoch": 0.19313, + "grad_norm": 1.1608262320979053, + "learning_rate": 0.003, + "loss": 4.1001, + "step": 19313 + }, + { + "epoch": 0.19314, + "grad_norm": 1.0098031546678037, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 19314 + }, + { + "epoch": 0.19315, + "grad_norm": 1.0539989424905194, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 19315 + }, + { + "epoch": 0.19316, + "grad_norm": 1.0390172650274812, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 19316 + }, + { + "epoch": 0.19317, + "grad_norm": 1.32647964590046, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 19317 + }, + { + "epoch": 0.19318, + "grad_norm": 0.8785936254392489, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 19318 + }, + { + "epoch": 0.19319, + "grad_norm": 0.8683495411909126, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 19319 + }, + { + "epoch": 0.1932, + "grad_norm": 1.0042166484230597, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 19320 + }, + { + "epoch": 0.19321, + "grad_norm": 1.1784752947786976, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 19321 + }, + { + "epoch": 0.19322, + "grad_norm": 1.197469831668031, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 19322 + }, + { + "epoch": 0.19323, + "grad_norm": 0.9222761320429943, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 19323 + }, + { + "epoch": 0.19324, + "grad_norm": 1.0417545947266198, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 19324 + }, + { + "epoch": 0.19325, + "grad_norm": 1.0819678167981466, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 19325 + }, + { + "epoch": 0.19326, + "grad_norm": 0.9707793026136153, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 19326 + }, + { + "epoch": 0.19327, + "grad_norm": 1.1217946456865704, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 19327 + }, + { + "epoch": 0.19328, + "grad_norm": 0.9841090713800797, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 19328 + }, + { + "epoch": 0.19329, + "grad_norm": 1.1690608083614729, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 19329 + }, + { + "epoch": 0.1933, + "grad_norm": 0.9088284296951473, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 19330 + }, + { + "epoch": 0.19331, + "grad_norm": 0.987859393884178, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 19331 + }, + { + "epoch": 0.19332, + "grad_norm": 1.203060766704908, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 19332 + }, + { + "epoch": 0.19333, + "grad_norm": 1.0484436885114863, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 19333 + }, + { + "epoch": 0.19334, + "grad_norm": 1.1121403584222151, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 19334 + }, + { + "epoch": 0.19335, + "grad_norm": 1.017585103370346, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 19335 + }, + { + "epoch": 0.19336, + "grad_norm": 1.116669890274972, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 19336 + }, + { + "epoch": 0.19337, + "grad_norm": 0.9650351336250899, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 19337 + }, + { + "epoch": 0.19338, + "grad_norm": 0.9542813979221435, + "learning_rate": 0.003, + "loss": 4.0958, + "step": 19338 + }, + { + "epoch": 0.19339, + "grad_norm": 1.0164090551463514, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 19339 + }, + { + "epoch": 0.1934, + "grad_norm": 1.2617196563915467, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 19340 + }, + { + "epoch": 0.19341, + "grad_norm": 0.9994122180691364, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 19341 + }, + { + "epoch": 0.19342, + "grad_norm": 1.0723678186050967, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 19342 + }, + { + "epoch": 0.19343, + "grad_norm": 1.158608504452686, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 19343 + }, + { + "epoch": 0.19344, + "grad_norm": 0.9289373278003584, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 19344 + }, + { + "epoch": 0.19345, + "grad_norm": 1.0902912781432672, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 19345 + }, + { + "epoch": 0.19346, + "grad_norm": 1.0473665911463008, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 19346 + }, + { + "epoch": 0.19347, + "grad_norm": 1.0901848514600527, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 19347 + }, + { + "epoch": 0.19348, + "grad_norm": 1.1976685709358577, + "learning_rate": 0.003, + "loss": 4.1146, + "step": 19348 + }, + { + "epoch": 0.19349, + "grad_norm": 0.9837028447902255, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 19349 + }, + { + "epoch": 0.1935, + "grad_norm": 1.017859366216322, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 19350 + }, + { + "epoch": 0.19351, + "grad_norm": 1.243626084811284, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 19351 + }, + { + "epoch": 0.19352, + "grad_norm": 1.1295253696148515, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 19352 + }, + { + "epoch": 0.19353, + "grad_norm": 1.0316071995495573, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 19353 + }, + { + "epoch": 0.19354, + "grad_norm": 1.1585415338406722, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 19354 + }, + { + "epoch": 0.19355, + "grad_norm": 0.8973139456384364, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 19355 + }, + { + "epoch": 0.19356, + "grad_norm": 0.7772111486893621, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 19356 + }, + { + "epoch": 0.19357, + "grad_norm": 0.7650238963435526, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 19357 + }, + { + "epoch": 0.19358, + "grad_norm": 0.8249716572301955, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 19358 + }, + { + "epoch": 0.19359, + "grad_norm": 1.0456647601566595, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 19359 + }, + { + "epoch": 0.1936, + "grad_norm": 1.208470664553363, + "learning_rate": 0.003, + "loss": 4.0917, + "step": 19360 + }, + { + "epoch": 0.19361, + "grad_norm": 0.9597149358237119, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 19361 + }, + { + "epoch": 0.19362, + "grad_norm": 0.9676112158187701, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 19362 + }, + { + "epoch": 0.19363, + "grad_norm": 1.0267558791421818, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 19363 + }, + { + "epoch": 0.19364, + "grad_norm": 1.409890970242339, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 19364 + }, + { + "epoch": 0.19365, + "grad_norm": 1.1465400266754044, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 19365 + }, + { + "epoch": 0.19366, + "grad_norm": 1.192702448660352, + "learning_rate": 0.003, + "loss": 4.077, + "step": 19366 + }, + { + "epoch": 0.19367, + "grad_norm": 0.9541312808820954, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 19367 + }, + { + "epoch": 0.19368, + "grad_norm": 0.9845725988962145, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 19368 + }, + { + "epoch": 0.19369, + "grad_norm": 1.3330858008574393, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 19369 + }, + { + "epoch": 0.1937, + "grad_norm": 0.7313327796100573, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 19370 + }, + { + "epoch": 0.19371, + "grad_norm": 0.7394308951177535, + "learning_rate": 0.003, + "loss": 4.056, + "step": 19371 + }, + { + "epoch": 0.19372, + "grad_norm": 0.9006152976443107, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 19372 + }, + { + "epoch": 0.19373, + "grad_norm": 1.1165844179946884, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 19373 + }, + { + "epoch": 0.19374, + "grad_norm": 1.1557835078179308, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 19374 + }, + { + "epoch": 0.19375, + "grad_norm": 1.070026170787068, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 19375 + }, + { + "epoch": 0.19376, + "grad_norm": 1.1621122953152492, + "learning_rate": 0.003, + "loss": 4.066, + "step": 19376 + }, + { + "epoch": 0.19377, + "grad_norm": 0.8673031890138061, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 19377 + }, + { + "epoch": 0.19378, + "grad_norm": 1.066320379227113, + "learning_rate": 0.003, + "loss": 4.0871, + "step": 19378 + }, + { + "epoch": 0.19379, + "grad_norm": 1.292914738580523, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 19379 + }, + { + "epoch": 0.1938, + "grad_norm": 0.7614257315008555, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 19380 + }, + { + "epoch": 0.19381, + "grad_norm": 0.7469458005120201, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 19381 + }, + { + "epoch": 0.19382, + "grad_norm": 0.8290465419922208, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 19382 + }, + { + "epoch": 0.19383, + "grad_norm": 0.9466559469084315, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 19383 + }, + { + "epoch": 0.19384, + "grad_norm": 1.2076908283475665, + "learning_rate": 0.003, + "loss": 4.1044, + "step": 19384 + }, + { + "epoch": 0.19385, + "grad_norm": 0.8904528852057874, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 19385 + }, + { + "epoch": 0.19386, + "grad_norm": 1.1424764428782548, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 19386 + }, + { + "epoch": 0.19387, + "grad_norm": 1.0119310456848885, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 19387 + }, + { + "epoch": 0.19388, + "grad_norm": 1.2604142529526263, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 19388 + }, + { + "epoch": 0.19389, + "grad_norm": 0.921967203735641, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 19389 + }, + { + "epoch": 0.1939, + "grad_norm": 0.8996300670983448, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 19390 + }, + { + "epoch": 0.19391, + "grad_norm": 1.0525491269946148, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 19391 + }, + { + "epoch": 0.19392, + "grad_norm": 1.0132115673034154, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 19392 + }, + { + "epoch": 0.19393, + "grad_norm": 0.9201811315466115, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 19393 + }, + { + "epoch": 0.19394, + "grad_norm": 1.0878780413729958, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 19394 + }, + { + "epoch": 0.19395, + "grad_norm": 1.126183542904692, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 19395 + }, + { + "epoch": 0.19396, + "grad_norm": 1.0714562545266466, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 19396 + }, + { + "epoch": 0.19397, + "grad_norm": 0.9824096893975103, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 19397 + }, + { + "epoch": 0.19398, + "grad_norm": 1.0452225193571192, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 19398 + }, + { + "epoch": 0.19399, + "grad_norm": 1.2289649890158996, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 19399 + }, + { + "epoch": 0.194, + "grad_norm": 1.1969932755949195, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 19400 + }, + { + "epoch": 0.19401, + "grad_norm": 1.1947489329858332, + "learning_rate": 0.003, + "loss": 4.058, + "step": 19401 + }, + { + "epoch": 0.19402, + "grad_norm": 0.8988953360105412, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 19402 + }, + { + "epoch": 0.19403, + "grad_norm": 1.0871505070337035, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 19403 + }, + { + "epoch": 0.19404, + "grad_norm": 1.1286113667952515, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 19404 + }, + { + "epoch": 0.19405, + "grad_norm": 0.969106561861846, + "learning_rate": 0.003, + "loss": 4.063, + "step": 19405 + }, + { + "epoch": 0.19406, + "grad_norm": 1.1636013184191678, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 19406 + }, + { + "epoch": 0.19407, + "grad_norm": 1.0191624910663375, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 19407 + }, + { + "epoch": 0.19408, + "grad_norm": 1.2153061701987697, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 19408 + }, + { + "epoch": 0.19409, + "grad_norm": 0.8410970019320878, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 19409 + }, + { + "epoch": 0.1941, + "grad_norm": 0.9362645133609997, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 19410 + }, + { + "epoch": 0.19411, + "grad_norm": 1.26480732751214, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 19411 + }, + { + "epoch": 0.19412, + "grad_norm": 1.27415110212814, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 19412 + }, + { + "epoch": 0.19413, + "grad_norm": 0.7058243424478735, + "learning_rate": 0.003, + "loss": 4.058, + "step": 19413 + }, + { + "epoch": 0.19414, + "grad_norm": 0.69557467000343, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 19414 + }, + { + "epoch": 0.19415, + "grad_norm": 0.8398239447466577, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 19415 + }, + { + "epoch": 0.19416, + "grad_norm": 0.8969463877744314, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 19416 + }, + { + "epoch": 0.19417, + "grad_norm": 1.0890506663099866, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 19417 + }, + { + "epoch": 0.19418, + "grad_norm": 1.2491655382848992, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 19418 + }, + { + "epoch": 0.19419, + "grad_norm": 0.9735096224997853, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 19419 + }, + { + "epoch": 0.1942, + "grad_norm": 1.1624933086771148, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 19420 + }, + { + "epoch": 0.19421, + "grad_norm": 1.0363929611258447, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 19421 + }, + { + "epoch": 0.19422, + "grad_norm": 1.1359661064794786, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 19422 + }, + { + "epoch": 0.19423, + "grad_norm": 1.0661651702219828, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 19423 + }, + { + "epoch": 0.19424, + "grad_norm": 0.9848198790617848, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 19424 + }, + { + "epoch": 0.19425, + "grad_norm": 1.2233143559862119, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 19425 + }, + { + "epoch": 0.19426, + "grad_norm": 0.9194329160027489, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 19426 + }, + { + "epoch": 0.19427, + "grad_norm": 0.851867183300549, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 19427 + }, + { + "epoch": 0.19428, + "grad_norm": 0.8149482008252208, + "learning_rate": 0.003, + "loss": 4.039, + "step": 19428 + }, + { + "epoch": 0.19429, + "grad_norm": 1.0276127984028876, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 19429 + }, + { + "epoch": 0.1943, + "grad_norm": 1.2644128095403526, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 19430 + }, + { + "epoch": 0.19431, + "grad_norm": 0.9538747787192434, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 19431 + }, + { + "epoch": 0.19432, + "grad_norm": 0.967313482901239, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 19432 + }, + { + "epoch": 0.19433, + "grad_norm": 1.1662466047124107, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 19433 + }, + { + "epoch": 0.19434, + "grad_norm": 1.031582523943195, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 19434 + }, + { + "epoch": 0.19435, + "grad_norm": 1.0437933426660613, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 19435 + }, + { + "epoch": 0.19436, + "grad_norm": 0.9304050357514089, + "learning_rate": 0.003, + "loss": 4.078, + "step": 19436 + }, + { + "epoch": 0.19437, + "grad_norm": 0.9608120433979089, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 19437 + }, + { + "epoch": 0.19438, + "grad_norm": 0.9723806643332686, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 19438 + }, + { + "epoch": 0.19439, + "grad_norm": 1.1190116455373578, + "learning_rate": 0.003, + "loss": 4.0873, + "step": 19439 + }, + { + "epoch": 0.1944, + "grad_norm": 1.0199657344432456, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 19440 + }, + { + "epoch": 0.19441, + "grad_norm": 1.1804818821126648, + "learning_rate": 0.003, + "loss": 4.0983, + "step": 19441 + }, + { + "epoch": 0.19442, + "grad_norm": 0.7955989971092, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 19442 + }, + { + "epoch": 0.19443, + "grad_norm": 1.0273821507921184, + "learning_rate": 0.003, + "loss": 4.0899, + "step": 19443 + }, + { + "epoch": 0.19444, + "grad_norm": 1.3222686108861788, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 19444 + }, + { + "epoch": 0.19445, + "grad_norm": 0.9512575013349508, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 19445 + }, + { + "epoch": 0.19446, + "grad_norm": 1.2859042453883895, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 19446 + }, + { + "epoch": 0.19447, + "grad_norm": 1.096194290637491, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 19447 + }, + { + "epoch": 0.19448, + "grad_norm": 1.1163965041149297, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 19448 + }, + { + "epoch": 0.19449, + "grad_norm": 0.9960028902766188, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 19449 + }, + { + "epoch": 0.1945, + "grad_norm": 1.2338934852758543, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 19450 + }, + { + "epoch": 0.19451, + "grad_norm": 0.8179977125079518, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 19451 + }, + { + "epoch": 0.19452, + "grad_norm": 0.8543269968795485, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 19452 + }, + { + "epoch": 0.19453, + "grad_norm": 1.0364160241167588, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 19453 + }, + { + "epoch": 0.19454, + "grad_norm": 1.273197543300573, + "learning_rate": 0.003, + "loss": 4.0913, + "step": 19454 + }, + { + "epoch": 0.19455, + "grad_norm": 0.8767271212309184, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 19455 + }, + { + "epoch": 0.19456, + "grad_norm": 0.9953134858800665, + "learning_rate": 0.003, + "loss": 4.093, + "step": 19456 + }, + { + "epoch": 0.19457, + "grad_norm": 1.177219374776552, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 19457 + }, + { + "epoch": 0.19458, + "grad_norm": 0.8624117850466086, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 19458 + }, + { + "epoch": 0.19459, + "grad_norm": 0.6912263685886758, + "learning_rate": 0.003, + "loss": 4.055, + "step": 19459 + }, + { + "epoch": 0.1946, + "grad_norm": 0.8882911415329025, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 19460 + }, + { + "epoch": 0.19461, + "grad_norm": 1.1936488549483462, + "learning_rate": 0.003, + "loss": 4.1098, + "step": 19461 + }, + { + "epoch": 0.19462, + "grad_norm": 1.031197079358379, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 19462 + }, + { + "epoch": 0.19463, + "grad_norm": 1.2706442778560993, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 19463 + }, + { + "epoch": 0.19464, + "grad_norm": 0.7948452857517353, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 19464 + }, + { + "epoch": 0.19465, + "grad_norm": 0.815103657961817, + "learning_rate": 0.003, + "loss": 4.053, + "step": 19465 + }, + { + "epoch": 0.19466, + "grad_norm": 0.9009981124547781, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 19466 + }, + { + "epoch": 0.19467, + "grad_norm": 1.1871821512802123, + "learning_rate": 0.003, + "loss": 4.079, + "step": 19467 + }, + { + "epoch": 0.19468, + "grad_norm": 1.0222587338093865, + "learning_rate": 0.003, + "loss": 4.047, + "step": 19468 + }, + { + "epoch": 0.19469, + "grad_norm": 1.0653658805542463, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 19469 + }, + { + "epoch": 0.1947, + "grad_norm": 1.0734183480522208, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 19470 + }, + { + "epoch": 0.19471, + "grad_norm": 1.0157503891987314, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 19471 + }, + { + "epoch": 0.19472, + "grad_norm": 1.085837988492662, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 19472 + }, + { + "epoch": 0.19473, + "grad_norm": 1.0752546145430986, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 19473 + }, + { + "epoch": 0.19474, + "grad_norm": 0.9651681292311738, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 19474 + }, + { + "epoch": 0.19475, + "grad_norm": 1.0614275247530274, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 19475 + }, + { + "epoch": 0.19476, + "grad_norm": 1.0732908336444935, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 19476 + }, + { + "epoch": 0.19477, + "grad_norm": 1.1820712348631295, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 19477 + }, + { + "epoch": 0.19478, + "grad_norm": 1.1955170346031967, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 19478 + }, + { + "epoch": 0.19479, + "grad_norm": 1.1809075698874647, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 19479 + }, + { + "epoch": 0.1948, + "grad_norm": 0.8726443353052283, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 19480 + }, + { + "epoch": 0.19481, + "grad_norm": 1.0629572891754375, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 19481 + }, + { + "epoch": 0.19482, + "grad_norm": 1.1844553094508796, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 19482 + }, + { + "epoch": 0.19483, + "grad_norm": 1.1684936825013028, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 19483 + }, + { + "epoch": 0.19484, + "grad_norm": 0.8856058524841478, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 19484 + }, + { + "epoch": 0.19485, + "grad_norm": 1.0064692391440815, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 19485 + }, + { + "epoch": 0.19486, + "grad_norm": 1.3219026671402243, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 19486 + }, + { + "epoch": 0.19487, + "grad_norm": 0.899639573565642, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 19487 + }, + { + "epoch": 0.19488, + "grad_norm": 0.9660928888496231, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 19488 + }, + { + "epoch": 0.19489, + "grad_norm": 0.9972243175529238, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 19489 + }, + { + "epoch": 0.1949, + "grad_norm": 1.263839648174591, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 19490 + }, + { + "epoch": 0.19491, + "grad_norm": 0.8975366963025061, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 19491 + }, + { + "epoch": 0.19492, + "grad_norm": 0.9775077265593591, + "learning_rate": 0.003, + "loss": 4.0916, + "step": 19492 + }, + { + "epoch": 0.19493, + "grad_norm": 1.1831183060020232, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 19493 + }, + { + "epoch": 0.19494, + "grad_norm": 1.0804497412333893, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 19494 + }, + { + "epoch": 0.19495, + "grad_norm": 1.2130930191586242, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 19495 + }, + { + "epoch": 0.19496, + "grad_norm": 1.0721086333141594, + "learning_rate": 0.003, + "loss": 4.052, + "step": 19496 + }, + { + "epoch": 0.19497, + "grad_norm": 0.95922054355939, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 19497 + }, + { + "epoch": 0.19498, + "grad_norm": 1.0776567685932843, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 19498 + }, + { + "epoch": 0.19499, + "grad_norm": 1.2332482486891303, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 19499 + }, + { + "epoch": 0.195, + "grad_norm": 0.9571014801195173, + "learning_rate": 0.003, + "loss": 4.066, + "step": 19500 + }, + { + "epoch": 0.19501, + "grad_norm": 1.023834763491347, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 19501 + }, + { + "epoch": 0.19502, + "grad_norm": 1.1413397272396921, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 19502 + }, + { + "epoch": 0.19503, + "grad_norm": 0.9291497918181236, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 19503 + }, + { + "epoch": 0.19504, + "grad_norm": 1.0821926834776672, + "learning_rate": 0.003, + "loss": 4.065, + "step": 19504 + }, + { + "epoch": 0.19505, + "grad_norm": 0.9945120763722275, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 19505 + }, + { + "epoch": 0.19506, + "grad_norm": 1.0526738326032545, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 19506 + }, + { + "epoch": 0.19507, + "grad_norm": 1.0247607689005436, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 19507 + }, + { + "epoch": 0.19508, + "grad_norm": 1.1568981621182008, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 19508 + }, + { + "epoch": 0.19509, + "grad_norm": 1.104031937957481, + "learning_rate": 0.003, + "loss": 4.0916, + "step": 19509 + }, + { + "epoch": 0.1951, + "grad_norm": 1.1239128834724506, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 19510 + }, + { + "epoch": 0.19511, + "grad_norm": 0.9748765352248675, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 19511 + }, + { + "epoch": 0.19512, + "grad_norm": 1.0777454631967849, + "learning_rate": 0.003, + "loss": 4.089, + "step": 19512 + }, + { + "epoch": 0.19513, + "grad_norm": 1.0002513844197876, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 19513 + }, + { + "epoch": 0.19514, + "grad_norm": 1.2226029590805514, + "learning_rate": 0.003, + "loss": 4.1104, + "step": 19514 + }, + { + "epoch": 0.19515, + "grad_norm": 1.0350587681455674, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 19515 + }, + { + "epoch": 0.19516, + "grad_norm": 1.082232742323968, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 19516 + }, + { + "epoch": 0.19517, + "grad_norm": 1.1979407277186234, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 19517 + }, + { + "epoch": 0.19518, + "grad_norm": 1.1188025517942874, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 19518 + }, + { + "epoch": 0.19519, + "grad_norm": 1.0678287676817793, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 19519 + }, + { + "epoch": 0.1952, + "grad_norm": 1.041867901717438, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 19520 + }, + { + "epoch": 0.19521, + "grad_norm": 1.1125296060024839, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 19521 + }, + { + "epoch": 0.19522, + "grad_norm": 1.2774639687474598, + "learning_rate": 0.003, + "loss": 4.0942, + "step": 19522 + }, + { + "epoch": 0.19523, + "grad_norm": 0.865775688994397, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 19523 + }, + { + "epoch": 0.19524, + "grad_norm": 0.8117819080799195, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 19524 + }, + { + "epoch": 0.19525, + "grad_norm": 0.7867619525588536, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 19525 + }, + { + "epoch": 0.19526, + "grad_norm": 0.9796429332908831, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 19526 + }, + { + "epoch": 0.19527, + "grad_norm": 1.1793803703118197, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 19527 + }, + { + "epoch": 0.19528, + "grad_norm": 1.0554429213522936, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 19528 + }, + { + "epoch": 0.19529, + "grad_norm": 1.3538769614330284, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 19529 + }, + { + "epoch": 0.1953, + "grad_norm": 1.0029753272838087, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 19530 + }, + { + "epoch": 0.19531, + "grad_norm": 1.0079266587925793, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 19531 + }, + { + "epoch": 0.19532, + "grad_norm": 0.8339986160911997, + "learning_rate": 0.003, + "loss": 4.0954, + "step": 19532 + }, + { + "epoch": 0.19533, + "grad_norm": 0.8308002072681483, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 19533 + }, + { + "epoch": 0.19534, + "grad_norm": 0.8628595627652501, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 19534 + }, + { + "epoch": 0.19535, + "grad_norm": 0.8584146901762636, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 19535 + }, + { + "epoch": 0.19536, + "grad_norm": 1.0172871906210925, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 19536 + }, + { + "epoch": 0.19537, + "grad_norm": 1.1546333558945094, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 19537 + }, + { + "epoch": 0.19538, + "grad_norm": 1.029606291174157, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 19538 + }, + { + "epoch": 0.19539, + "grad_norm": 1.0398258244735001, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 19539 + }, + { + "epoch": 0.1954, + "grad_norm": 1.060151386325284, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 19540 + }, + { + "epoch": 0.19541, + "grad_norm": 1.3577307855386096, + "learning_rate": 0.003, + "loss": 4.0864, + "step": 19541 + }, + { + "epoch": 0.19542, + "grad_norm": 0.8696290577722641, + "learning_rate": 0.003, + "loss": 4.0919, + "step": 19542 + }, + { + "epoch": 0.19543, + "grad_norm": 0.9165035378587159, + "learning_rate": 0.003, + "loss": 4.094, + "step": 19543 + }, + { + "epoch": 0.19544, + "grad_norm": 1.117653163789322, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 19544 + }, + { + "epoch": 0.19545, + "grad_norm": 1.2747189445666307, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 19545 + }, + { + "epoch": 0.19546, + "grad_norm": 0.9285618553134947, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 19546 + }, + { + "epoch": 0.19547, + "grad_norm": 1.0735925322992683, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 19547 + }, + { + "epoch": 0.19548, + "grad_norm": 1.3708154436601667, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 19548 + }, + { + "epoch": 0.19549, + "grad_norm": 1.0213839572975085, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 19549 + }, + { + "epoch": 0.1955, + "grad_norm": 0.9865489805430221, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 19550 + }, + { + "epoch": 0.19551, + "grad_norm": 1.1562811540819125, + "learning_rate": 0.003, + "loss": 4.0963, + "step": 19551 + }, + { + "epoch": 0.19552, + "grad_norm": 0.9870629764976921, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 19552 + }, + { + "epoch": 0.19553, + "grad_norm": 1.2332577500807513, + "learning_rate": 0.003, + "loss": 4.081, + "step": 19553 + }, + { + "epoch": 0.19554, + "grad_norm": 0.804700385570442, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 19554 + }, + { + "epoch": 0.19555, + "grad_norm": 0.8633221146973241, + "learning_rate": 0.003, + "loss": 4.06, + "step": 19555 + }, + { + "epoch": 0.19556, + "grad_norm": 1.03171031220499, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 19556 + }, + { + "epoch": 0.19557, + "grad_norm": 1.3983258166660677, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 19557 + }, + { + "epoch": 0.19558, + "grad_norm": 0.8429918976107377, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 19558 + }, + { + "epoch": 0.19559, + "grad_norm": 1.0008676717616385, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 19559 + }, + { + "epoch": 0.1956, + "grad_norm": 1.1854897973273935, + "learning_rate": 0.003, + "loss": 4.1003, + "step": 19560 + }, + { + "epoch": 0.19561, + "grad_norm": 1.2802044603615492, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 19561 + }, + { + "epoch": 0.19562, + "grad_norm": 0.7683115458137728, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 19562 + }, + { + "epoch": 0.19563, + "grad_norm": 0.6865842653444609, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 19563 + }, + { + "epoch": 0.19564, + "grad_norm": 0.7418166177040573, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 19564 + }, + { + "epoch": 0.19565, + "grad_norm": 0.8216851104877765, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 19565 + }, + { + "epoch": 0.19566, + "grad_norm": 1.0179842749720853, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 19566 + }, + { + "epoch": 0.19567, + "grad_norm": 1.3458362957735324, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 19567 + }, + { + "epoch": 0.19568, + "grad_norm": 0.9527846289579246, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 19568 + }, + { + "epoch": 0.19569, + "grad_norm": 0.874262248064775, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 19569 + }, + { + "epoch": 0.1957, + "grad_norm": 0.8106316584664504, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 19570 + }, + { + "epoch": 0.19571, + "grad_norm": 0.8479450323924783, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 19571 + }, + { + "epoch": 0.19572, + "grad_norm": 1.0088122131917534, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 19572 + }, + { + "epoch": 0.19573, + "grad_norm": 1.1633194090454788, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 19573 + }, + { + "epoch": 0.19574, + "grad_norm": 0.8960330601128592, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 19574 + }, + { + "epoch": 0.19575, + "grad_norm": 1.1253656505897451, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 19575 + }, + { + "epoch": 0.19576, + "grad_norm": 1.3446918563499104, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 19576 + }, + { + "epoch": 0.19577, + "grad_norm": 0.9417816698724147, + "learning_rate": 0.003, + "loss": 4.075, + "step": 19577 + }, + { + "epoch": 0.19578, + "grad_norm": 0.9563451726891925, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 19578 + }, + { + "epoch": 0.19579, + "grad_norm": 1.2714182267670642, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 19579 + }, + { + "epoch": 0.1958, + "grad_norm": 0.9981642532904149, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 19580 + }, + { + "epoch": 0.19581, + "grad_norm": 1.1636587163375327, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 19581 + }, + { + "epoch": 0.19582, + "grad_norm": 0.9119991738062687, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 19582 + }, + { + "epoch": 0.19583, + "grad_norm": 0.9243517100163055, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 19583 + }, + { + "epoch": 0.19584, + "grad_norm": 1.0197949650869493, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 19584 + }, + { + "epoch": 0.19585, + "grad_norm": 1.000627190880686, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 19585 + }, + { + "epoch": 0.19586, + "grad_norm": 1.0363408394918823, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 19586 + }, + { + "epoch": 0.19587, + "grad_norm": 0.9473734616758167, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 19587 + }, + { + "epoch": 0.19588, + "grad_norm": 0.958296908369999, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 19588 + }, + { + "epoch": 0.19589, + "grad_norm": 1.122980139154144, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 19589 + }, + { + "epoch": 0.1959, + "grad_norm": 1.0560370118778166, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 19590 + }, + { + "epoch": 0.19591, + "grad_norm": 1.2257818319613043, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 19591 + }, + { + "epoch": 0.19592, + "grad_norm": 1.0658277068282773, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 19592 + }, + { + "epoch": 0.19593, + "grad_norm": 1.0884496547978777, + "learning_rate": 0.003, + "loss": 4.063, + "step": 19593 + }, + { + "epoch": 0.19594, + "grad_norm": 0.8890222378692731, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 19594 + }, + { + "epoch": 0.19595, + "grad_norm": 1.0172064214448733, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 19595 + }, + { + "epoch": 0.19596, + "grad_norm": 1.251014932158139, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 19596 + }, + { + "epoch": 0.19597, + "grad_norm": 0.8679821676604227, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 19597 + }, + { + "epoch": 0.19598, + "grad_norm": 1.1374076129608401, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 19598 + }, + { + "epoch": 0.19599, + "grad_norm": 0.8967873114392836, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 19599 + }, + { + "epoch": 0.196, + "grad_norm": 1.1156739289657867, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 19600 + }, + { + "epoch": 0.19601, + "grad_norm": 1.3754344452073968, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 19601 + }, + { + "epoch": 0.19602, + "grad_norm": 1.013630981952612, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 19602 + }, + { + "epoch": 0.19603, + "grad_norm": 0.9943764749241032, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 19603 + }, + { + "epoch": 0.19604, + "grad_norm": 1.2078800414981548, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 19604 + }, + { + "epoch": 0.19605, + "grad_norm": 1.0011669179217135, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 19605 + }, + { + "epoch": 0.19606, + "grad_norm": 1.1282190039993816, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 19606 + }, + { + "epoch": 0.19607, + "grad_norm": 1.026367679068041, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 19607 + }, + { + "epoch": 0.19608, + "grad_norm": 1.0517422902374522, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 19608 + }, + { + "epoch": 0.19609, + "grad_norm": 1.1254082349914274, + "learning_rate": 0.003, + "loss": 4.073, + "step": 19609 + }, + { + "epoch": 0.1961, + "grad_norm": 1.2252819554226304, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 19610 + }, + { + "epoch": 0.19611, + "grad_norm": 0.8941700333860505, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 19611 + }, + { + "epoch": 0.19612, + "grad_norm": 0.9115427351156864, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 19612 + }, + { + "epoch": 0.19613, + "grad_norm": 0.9774129871751689, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 19613 + }, + { + "epoch": 0.19614, + "grad_norm": 1.204933351904217, + "learning_rate": 0.003, + "loss": 4.041, + "step": 19614 + }, + { + "epoch": 0.19615, + "grad_norm": 0.8162895983427717, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 19615 + }, + { + "epoch": 0.19616, + "grad_norm": 0.8783631371118935, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 19616 + }, + { + "epoch": 0.19617, + "grad_norm": 1.0088217546051081, + "learning_rate": 0.003, + "loss": 4.0944, + "step": 19617 + }, + { + "epoch": 0.19618, + "grad_norm": 1.102998605610352, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 19618 + }, + { + "epoch": 0.19619, + "grad_norm": 1.2961065017424278, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 19619 + }, + { + "epoch": 0.1962, + "grad_norm": 0.9538813126180983, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 19620 + }, + { + "epoch": 0.19621, + "grad_norm": 1.1764962127136702, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 19621 + }, + { + "epoch": 0.19622, + "grad_norm": 0.9647605258300946, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 19622 + }, + { + "epoch": 0.19623, + "grad_norm": 1.0138060008264889, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 19623 + }, + { + "epoch": 0.19624, + "grad_norm": 1.1032430342349775, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 19624 + }, + { + "epoch": 0.19625, + "grad_norm": 0.9170224689056333, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 19625 + }, + { + "epoch": 0.19626, + "grad_norm": 0.9432362198907326, + "learning_rate": 0.003, + "loss": 4.0819, + "step": 19626 + }, + { + "epoch": 0.19627, + "grad_norm": 1.0122426551707677, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 19627 + }, + { + "epoch": 0.19628, + "grad_norm": 1.1672551407266847, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 19628 + }, + { + "epoch": 0.19629, + "grad_norm": 1.1846719032941637, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 19629 + }, + { + "epoch": 0.1963, + "grad_norm": 1.249017972274037, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 19630 + }, + { + "epoch": 0.19631, + "grad_norm": 0.8735270143407934, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 19631 + }, + { + "epoch": 0.19632, + "grad_norm": 0.9675674653229092, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 19632 + }, + { + "epoch": 0.19633, + "grad_norm": 1.484701131541642, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 19633 + }, + { + "epoch": 0.19634, + "grad_norm": 0.8485241560036316, + "learning_rate": 0.003, + "loss": 4.0854, + "step": 19634 + }, + { + "epoch": 0.19635, + "grad_norm": 0.789719423299055, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 19635 + }, + { + "epoch": 0.19636, + "grad_norm": 0.9467278161364382, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 19636 + }, + { + "epoch": 0.19637, + "grad_norm": 0.9192933682836749, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 19637 + }, + { + "epoch": 0.19638, + "grad_norm": 0.8561919554801843, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 19638 + }, + { + "epoch": 0.19639, + "grad_norm": 1.1405848598367638, + "learning_rate": 0.003, + "loss": 4.097, + "step": 19639 + }, + { + "epoch": 0.1964, + "grad_norm": 1.2396779377395666, + "learning_rate": 0.003, + "loss": 4.045, + "step": 19640 + }, + { + "epoch": 0.19641, + "grad_norm": 0.9900675354721306, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 19641 + }, + { + "epoch": 0.19642, + "grad_norm": 1.2582754980439856, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 19642 + }, + { + "epoch": 0.19643, + "grad_norm": 0.898520388535626, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 19643 + }, + { + "epoch": 0.19644, + "grad_norm": 0.8245224220899114, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 19644 + }, + { + "epoch": 0.19645, + "grad_norm": 0.8521602801987644, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 19645 + }, + { + "epoch": 0.19646, + "grad_norm": 0.9126292775696665, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 19646 + }, + { + "epoch": 0.19647, + "grad_norm": 1.0055364234563897, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 19647 + }, + { + "epoch": 0.19648, + "grad_norm": 0.9947794940776054, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 19648 + }, + { + "epoch": 0.19649, + "grad_norm": 1.2496260378249877, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 19649 + }, + { + "epoch": 0.1965, + "grad_norm": 0.9778769467493066, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 19650 + }, + { + "epoch": 0.19651, + "grad_norm": 1.2424859200412788, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 19651 + }, + { + "epoch": 0.19652, + "grad_norm": 0.7918016065015122, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 19652 + }, + { + "epoch": 0.19653, + "grad_norm": 0.853859991192462, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 19653 + }, + { + "epoch": 0.19654, + "grad_norm": 1.1201368735125743, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 19654 + }, + { + "epoch": 0.19655, + "grad_norm": 1.2192227889468976, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 19655 + }, + { + "epoch": 0.19656, + "grad_norm": 1.0093522438825073, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 19656 + }, + { + "epoch": 0.19657, + "grad_norm": 1.1905818773459682, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 19657 + }, + { + "epoch": 0.19658, + "grad_norm": 1.0072468202998726, + "learning_rate": 0.003, + "loss": 4.0964, + "step": 19658 + }, + { + "epoch": 0.19659, + "grad_norm": 1.1052567000331015, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 19659 + }, + { + "epoch": 0.1966, + "grad_norm": 1.0766086495060634, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 19660 + }, + { + "epoch": 0.19661, + "grad_norm": 1.119781935797932, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 19661 + }, + { + "epoch": 0.19662, + "grad_norm": 1.0690402410656012, + "learning_rate": 0.003, + "loss": 4.085, + "step": 19662 + }, + { + "epoch": 0.19663, + "grad_norm": 1.268214664218795, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 19663 + }, + { + "epoch": 0.19664, + "grad_norm": 0.9263523578218694, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 19664 + }, + { + "epoch": 0.19665, + "grad_norm": 1.0024493123002174, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 19665 + }, + { + "epoch": 0.19666, + "grad_norm": 1.0737343091866198, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 19666 + }, + { + "epoch": 0.19667, + "grad_norm": 1.2007587963470145, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 19667 + }, + { + "epoch": 0.19668, + "grad_norm": 0.980510373049565, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 19668 + }, + { + "epoch": 0.19669, + "grad_norm": 0.9800534079984307, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 19669 + }, + { + "epoch": 0.1967, + "grad_norm": 1.001846683789748, + "learning_rate": 0.003, + "loss": 4.063, + "step": 19670 + }, + { + "epoch": 0.19671, + "grad_norm": 1.1025372359818557, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 19671 + }, + { + "epoch": 0.19672, + "grad_norm": 1.1321789022240898, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 19672 + }, + { + "epoch": 0.19673, + "grad_norm": 0.9286705323257387, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 19673 + }, + { + "epoch": 0.19674, + "grad_norm": 0.7814261203999033, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 19674 + }, + { + "epoch": 0.19675, + "grad_norm": 0.9085939561139585, + "learning_rate": 0.003, + "loss": 4.057, + "step": 19675 + }, + { + "epoch": 0.19676, + "grad_norm": 1.0952848623989881, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 19676 + }, + { + "epoch": 0.19677, + "grad_norm": 1.085538287838892, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 19677 + }, + { + "epoch": 0.19678, + "grad_norm": 1.073126154028048, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 19678 + }, + { + "epoch": 0.19679, + "grad_norm": 1.0288237877349964, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 19679 + }, + { + "epoch": 0.1968, + "grad_norm": 1.0984398827315445, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 19680 + }, + { + "epoch": 0.19681, + "grad_norm": 1.1216717416927258, + "learning_rate": 0.003, + "loss": 4.0954, + "step": 19681 + }, + { + "epoch": 0.19682, + "grad_norm": 1.0113222429692303, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 19682 + }, + { + "epoch": 0.19683, + "grad_norm": 1.0701776953943052, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 19683 + }, + { + "epoch": 0.19684, + "grad_norm": 1.3266592917286106, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 19684 + }, + { + "epoch": 0.19685, + "grad_norm": 0.9555187385056396, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 19685 + }, + { + "epoch": 0.19686, + "grad_norm": 0.9922552204043814, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 19686 + }, + { + "epoch": 0.19687, + "grad_norm": 1.2726826742279103, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 19687 + }, + { + "epoch": 0.19688, + "grad_norm": 1.1031500564903052, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 19688 + }, + { + "epoch": 0.19689, + "grad_norm": 0.9840065819635753, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 19689 + }, + { + "epoch": 0.1969, + "grad_norm": 1.0987685399486895, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 19690 + }, + { + "epoch": 0.19691, + "grad_norm": 1.22716031058259, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 19691 + }, + { + "epoch": 0.19692, + "grad_norm": 0.8838886265450338, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 19692 + }, + { + "epoch": 0.19693, + "grad_norm": 1.0809090121348586, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 19693 + }, + { + "epoch": 0.19694, + "grad_norm": 0.976333368660224, + "learning_rate": 0.003, + "loss": 4.049, + "step": 19694 + }, + { + "epoch": 0.19695, + "grad_norm": 1.1216348412438772, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 19695 + }, + { + "epoch": 0.19696, + "grad_norm": 1.0440020982964053, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 19696 + }, + { + "epoch": 0.19697, + "grad_norm": 1.0912295344507204, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 19697 + }, + { + "epoch": 0.19698, + "grad_norm": 1.2667665476784546, + "learning_rate": 0.003, + "loss": 4.1025, + "step": 19698 + }, + { + "epoch": 0.19699, + "grad_norm": 0.941386634333801, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 19699 + }, + { + "epoch": 0.197, + "grad_norm": 0.9435337903450359, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 19700 + }, + { + "epoch": 0.19701, + "grad_norm": 1.1094732331839676, + "learning_rate": 0.003, + "loss": 4.097, + "step": 19701 + }, + { + "epoch": 0.19702, + "grad_norm": 1.2451695029092207, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 19702 + }, + { + "epoch": 0.19703, + "grad_norm": 0.8391960597594307, + "learning_rate": 0.003, + "loss": 4.0876, + "step": 19703 + }, + { + "epoch": 0.19704, + "grad_norm": 0.91908667381626, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 19704 + }, + { + "epoch": 0.19705, + "grad_norm": 0.9938552412243009, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 19705 + }, + { + "epoch": 0.19706, + "grad_norm": 1.2185804745606796, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 19706 + }, + { + "epoch": 0.19707, + "grad_norm": 0.9290266984138505, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 19707 + }, + { + "epoch": 0.19708, + "grad_norm": 1.0802767310961074, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 19708 + }, + { + "epoch": 0.19709, + "grad_norm": 1.1447689726194905, + "learning_rate": 0.003, + "loss": 4.0985, + "step": 19709 + }, + { + "epoch": 0.1971, + "grad_norm": 1.1990613485856336, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 19710 + }, + { + "epoch": 0.19711, + "grad_norm": 0.851712018971272, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 19711 + }, + { + "epoch": 0.19712, + "grad_norm": 0.8964281204670373, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 19712 + }, + { + "epoch": 0.19713, + "grad_norm": 1.0721534503721215, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 19713 + }, + { + "epoch": 0.19714, + "grad_norm": 1.2296981867111196, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 19714 + }, + { + "epoch": 0.19715, + "grad_norm": 0.8770254677269856, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 19715 + }, + { + "epoch": 0.19716, + "grad_norm": 0.8834624903352598, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 19716 + }, + { + "epoch": 0.19717, + "grad_norm": 1.0866508829278358, + "learning_rate": 0.003, + "loss": 4.0893, + "step": 19717 + }, + { + "epoch": 0.19718, + "grad_norm": 1.0950545630197945, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 19718 + }, + { + "epoch": 0.19719, + "grad_norm": 0.9013082674068399, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 19719 + }, + { + "epoch": 0.1972, + "grad_norm": 0.952426986736761, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 19720 + }, + { + "epoch": 0.19721, + "grad_norm": 1.0827004470550563, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 19721 + }, + { + "epoch": 0.19722, + "grad_norm": 1.0211181743536386, + "learning_rate": 0.003, + "loss": 4.069, + "step": 19722 + }, + { + "epoch": 0.19723, + "grad_norm": 1.2943760295561586, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 19723 + }, + { + "epoch": 0.19724, + "grad_norm": 1.1142631710349287, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 19724 + }, + { + "epoch": 0.19725, + "grad_norm": 0.9684581870318089, + "learning_rate": 0.003, + "loss": 4.05, + "step": 19725 + }, + { + "epoch": 0.19726, + "grad_norm": 1.1245678011043916, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 19726 + }, + { + "epoch": 0.19727, + "grad_norm": 1.0530349252857751, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 19727 + }, + { + "epoch": 0.19728, + "grad_norm": 1.1582259949992828, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 19728 + }, + { + "epoch": 0.19729, + "grad_norm": 1.0118204807985547, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 19729 + }, + { + "epoch": 0.1973, + "grad_norm": 1.1069071954271341, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 19730 + }, + { + "epoch": 0.19731, + "grad_norm": 1.0591067840798596, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 19731 + }, + { + "epoch": 0.19732, + "grad_norm": 1.0122216335943788, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 19732 + }, + { + "epoch": 0.19733, + "grad_norm": 1.1049108169732538, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 19733 + }, + { + "epoch": 0.19734, + "grad_norm": 1.03068593102244, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 19734 + }, + { + "epoch": 0.19735, + "grad_norm": 0.9216751100790874, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 19735 + }, + { + "epoch": 0.19736, + "grad_norm": 1.02519585346266, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 19736 + }, + { + "epoch": 0.19737, + "grad_norm": 1.2591969385192279, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 19737 + }, + { + "epoch": 0.19738, + "grad_norm": 0.8769697387557531, + "learning_rate": 0.003, + "loss": 4.051, + "step": 19738 + }, + { + "epoch": 0.19739, + "grad_norm": 1.0008445731811877, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 19739 + }, + { + "epoch": 0.1974, + "grad_norm": 1.122411620893132, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 19740 + }, + { + "epoch": 0.19741, + "grad_norm": 0.9181875865989599, + "learning_rate": 0.003, + "loss": 4.064, + "step": 19741 + }, + { + "epoch": 0.19742, + "grad_norm": 1.1090964580430809, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 19742 + }, + { + "epoch": 0.19743, + "grad_norm": 1.18317729998744, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 19743 + }, + { + "epoch": 0.19744, + "grad_norm": 1.0080055573270916, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 19744 + }, + { + "epoch": 0.19745, + "grad_norm": 1.0816296163862171, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 19745 + }, + { + "epoch": 0.19746, + "grad_norm": 1.2039758356130552, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 19746 + }, + { + "epoch": 0.19747, + "grad_norm": 1.2775283898126437, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 19747 + }, + { + "epoch": 0.19748, + "grad_norm": 1.0757193447922522, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 19748 + }, + { + "epoch": 0.19749, + "grad_norm": 1.2513914428101447, + "learning_rate": 0.003, + "loss": 4.071, + "step": 19749 + }, + { + "epoch": 0.1975, + "grad_norm": 1.1401783656148807, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 19750 + }, + { + "epoch": 0.19751, + "grad_norm": 1.090882290227553, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 19751 + }, + { + "epoch": 0.19752, + "grad_norm": 1.127142580695729, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 19752 + }, + { + "epoch": 0.19753, + "grad_norm": 1.0067479600607931, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 19753 + }, + { + "epoch": 0.19754, + "grad_norm": 1.160743070522396, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 19754 + }, + { + "epoch": 0.19755, + "grad_norm": 0.9676086872924521, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 19755 + }, + { + "epoch": 0.19756, + "grad_norm": 1.1278542514663414, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 19756 + }, + { + "epoch": 0.19757, + "grad_norm": 0.9813440750902493, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 19757 + }, + { + "epoch": 0.19758, + "grad_norm": 0.9946776596427616, + "learning_rate": 0.003, + "loss": 4.0985, + "step": 19758 + }, + { + "epoch": 0.19759, + "grad_norm": 1.0277895191175592, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 19759 + }, + { + "epoch": 0.1976, + "grad_norm": 0.841022552838594, + "learning_rate": 0.003, + "loss": 4.0996, + "step": 19760 + }, + { + "epoch": 0.19761, + "grad_norm": 0.895190793552959, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 19761 + }, + { + "epoch": 0.19762, + "grad_norm": 0.9517698104673334, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 19762 + }, + { + "epoch": 0.19763, + "grad_norm": 1.1194277862995614, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 19763 + }, + { + "epoch": 0.19764, + "grad_norm": 0.9810270112936441, + "learning_rate": 0.003, + "loss": 4.061, + "step": 19764 + }, + { + "epoch": 0.19765, + "grad_norm": 0.9723715980972237, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 19765 + }, + { + "epoch": 0.19766, + "grad_norm": 1.1167243402018459, + "learning_rate": 0.003, + "loss": 4.0957, + "step": 19766 + }, + { + "epoch": 0.19767, + "grad_norm": 0.9992005340651623, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 19767 + }, + { + "epoch": 0.19768, + "grad_norm": 0.9553022892420305, + "learning_rate": 0.003, + "loss": 4.071, + "step": 19768 + }, + { + "epoch": 0.19769, + "grad_norm": 0.9930833650370362, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 19769 + }, + { + "epoch": 0.1977, + "grad_norm": 1.0467872587692324, + "learning_rate": 0.003, + "loss": 4.0913, + "step": 19770 + }, + { + "epoch": 0.19771, + "grad_norm": 1.1105399253310277, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 19771 + }, + { + "epoch": 0.19772, + "grad_norm": 0.907256044272153, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 19772 + }, + { + "epoch": 0.19773, + "grad_norm": 0.9913053504789294, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 19773 + }, + { + "epoch": 0.19774, + "grad_norm": 1.1477672848910199, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 19774 + }, + { + "epoch": 0.19775, + "grad_norm": 0.9811061508099871, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 19775 + }, + { + "epoch": 0.19776, + "grad_norm": 1.1586323672798708, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 19776 + }, + { + "epoch": 0.19777, + "grad_norm": 1.1554384283829118, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 19777 + }, + { + "epoch": 0.19778, + "grad_norm": 0.8947888499922675, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 19778 + }, + { + "epoch": 0.19779, + "grad_norm": 0.8355035314228024, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 19779 + }, + { + "epoch": 0.1978, + "grad_norm": 1.0072858017445776, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 19780 + }, + { + "epoch": 0.19781, + "grad_norm": 1.1769774778149924, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 19781 + }, + { + "epoch": 0.19782, + "grad_norm": 1.153525691546478, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 19782 + }, + { + "epoch": 0.19783, + "grad_norm": 1.110598618801948, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 19783 + }, + { + "epoch": 0.19784, + "grad_norm": 1.1772175930355984, + "learning_rate": 0.003, + "loss": 4.077, + "step": 19784 + }, + { + "epoch": 0.19785, + "grad_norm": 1.1677967995098004, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 19785 + }, + { + "epoch": 0.19786, + "grad_norm": 1.1894737517452079, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 19786 + }, + { + "epoch": 0.19787, + "grad_norm": 1.0765999043058847, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 19787 + }, + { + "epoch": 0.19788, + "grad_norm": 1.0501575192338788, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 19788 + }, + { + "epoch": 0.19789, + "grad_norm": 1.0127611873323827, + "learning_rate": 0.003, + "loss": 4.08, + "step": 19789 + }, + { + "epoch": 0.1979, + "grad_norm": 1.1262856356675508, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 19790 + }, + { + "epoch": 0.19791, + "grad_norm": 1.1830050172211235, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 19791 + }, + { + "epoch": 0.19792, + "grad_norm": 0.7332206783119853, + "learning_rate": 0.003, + "loss": 4.088, + "step": 19792 + }, + { + "epoch": 0.19793, + "grad_norm": 0.8945780239222727, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 19793 + }, + { + "epoch": 0.19794, + "grad_norm": 1.158612400409631, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 19794 + }, + { + "epoch": 0.19795, + "grad_norm": 1.1772032938925456, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 19795 + }, + { + "epoch": 0.19796, + "grad_norm": 1.1300287787951653, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 19796 + }, + { + "epoch": 0.19797, + "grad_norm": 1.0625062448432137, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 19797 + }, + { + "epoch": 0.19798, + "grad_norm": 0.8372880297743913, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 19798 + }, + { + "epoch": 0.19799, + "grad_norm": 0.9553083887436783, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 19799 + }, + { + "epoch": 0.198, + "grad_norm": 1.1966220830768395, + "learning_rate": 0.003, + "loss": 4.0998, + "step": 19800 + }, + { + "epoch": 0.19801, + "grad_norm": 0.9206156042146972, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 19801 + }, + { + "epoch": 0.19802, + "grad_norm": 1.160355196048666, + "learning_rate": 0.003, + "loss": 4.0954, + "step": 19802 + }, + { + "epoch": 0.19803, + "grad_norm": 0.9929214090450242, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 19803 + }, + { + "epoch": 0.19804, + "grad_norm": 0.9787583309152769, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 19804 + }, + { + "epoch": 0.19805, + "grad_norm": 1.113142801487708, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 19805 + }, + { + "epoch": 0.19806, + "grad_norm": 1.011813226074155, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 19806 + }, + { + "epoch": 0.19807, + "grad_norm": 1.2543631930940466, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 19807 + }, + { + "epoch": 0.19808, + "grad_norm": 1.2470147558741194, + "learning_rate": 0.003, + "loss": 4.1191, + "step": 19808 + }, + { + "epoch": 0.19809, + "grad_norm": 0.9943371952313821, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 19809 + }, + { + "epoch": 0.1981, + "grad_norm": 1.047437915441819, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 19810 + }, + { + "epoch": 0.19811, + "grad_norm": 0.9618221574390824, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 19811 + }, + { + "epoch": 0.19812, + "grad_norm": 1.0070748443170565, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 19812 + }, + { + "epoch": 0.19813, + "grad_norm": 0.9118101423530267, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 19813 + }, + { + "epoch": 0.19814, + "grad_norm": 0.8939868335112484, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 19814 + }, + { + "epoch": 0.19815, + "grad_norm": 0.9997997715845341, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 19815 + }, + { + "epoch": 0.19816, + "grad_norm": 1.0527358423263957, + "learning_rate": 0.003, + "loss": 4.061, + "step": 19816 + }, + { + "epoch": 0.19817, + "grad_norm": 1.1430989761865817, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 19817 + }, + { + "epoch": 0.19818, + "grad_norm": 0.9074851308200729, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 19818 + }, + { + "epoch": 0.19819, + "grad_norm": 1.002590328364556, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 19819 + }, + { + "epoch": 0.1982, + "grad_norm": 1.089035836088523, + "learning_rate": 0.003, + "loss": 4.072, + "step": 19820 + }, + { + "epoch": 0.19821, + "grad_norm": 0.9075691593876065, + "learning_rate": 0.003, + "loss": 4.074, + "step": 19821 + }, + { + "epoch": 0.19822, + "grad_norm": 1.1408859741416852, + "learning_rate": 0.003, + "loss": 4.047, + "step": 19822 + }, + { + "epoch": 0.19823, + "grad_norm": 1.5122168063110784, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 19823 + }, + { + "epoch": 0.19824, + "grad_norm": 0.8869113111716151, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 19824 + }, + { + "epoch": 0.19825, + "grad_norm": 0.9285987836878088, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 19825 + }, + { + "epoch": 0.19826, + "grad_norm": 1.14319630624869, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 19826 + }, + { + "epoch": 0.19827, + "grad_norm": 1.0390632770286254, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 19827 + }, + { + "epoch": 0.19828, + "grad_norm": 1.13055814753828, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 19828 + }, + { + "epoch": 0.19829, + "grad_norm": 1.1918464713497865, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 19829 + }, + { + "epoch": 0.1983, + "grad_norm": 0.9647515460212693, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 19830 + }, + { + "epoch": 0.19831, + "grad_norm": 1.029756376897784, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 19831 + }, + { + "epoch": 0.19832, + "grad_norm": 1.2215537374614216, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 19832 + }, + { + "epoch": 0.19833, + "grad_norm": 0.9768410602123577, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 19833 + }, + { + "epoch": 0.19834, + "grad_norm": 0.9776771477465113, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 19834 + }, + { + "epoch": 0.19835, + "grad_norm": 0.9530145804933323, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 19835 + }, + { + "epoch": 0.19836, + "grad_norm": 1.0159642617996845, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 19836 + }, + { + "epoch": 0.19837, + "grad_norm": 1.0172222755030425, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 19837 + }, + { + "epoch": 0.19838, + "grad_norm": 1.023649811289072, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 19838 + }, + { + "epoch": 0.19839, + "grad_norm": 1.0421901542199974, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 19839 + }, + { + "epoch": 0.1984, + "grad_norm": 1.0753993726726192, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 19840 + }, + { + "epoch": 0.19841, + "grad_norm": 1.1671221596909451, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 19841 + }, + { + "epoch": 0.19842, + "grad_norm": 1.1850039791581433, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 19842 + }, + { + "epoch": 0.19843, + "grad_norm": 1.0107616171566025, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 19843 + }, + { + "epoch": 0.19844, + "grad_norm": 1.0879338503411544, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 19844 + }, + { + "epoch": 0.19845, + "grad_norm": 1.0765351548999302, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 19845 + }, + { + "epoch": 0.19846, + "grad_norm": 1.0709463815017626, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 19846 + }, + { + "epoch": 0.19847, + "grad_norm": 1.0762241884781865, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 19847 + }, + { + "epoch": 0.19848, + "grad_norm": 1.0403514035314512, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 19848 + }, + { + "epoch": 0.19849, + "grad_norm": 1.0831150863390824, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 19849 + }, + { + "epoch": 0.1985, + "grad_norm": 1.090444956338347, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 19850 + }, + { + "epoch": 0.19851, + "grad_norm": 1.0978283849436352, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 19851 + }, + { + "epoch": 0.19852, + "grad_norm": 0.9174227455069103, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 19852 + }, + { + "epoch": 0.19853, + "grad_norm": 0.8662204416341773, + "learning_rate": 0.003, + "loss": 4.035, + "step": 19853 + }, + { + "epoch": 0.19854, + "grad_norm": 0.9499590489459081, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 19854 + }, + { + "epoch": 0.19855, + "grad_norm": 1.1439729674036212, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 19855 + }, + { + "epoch": 0.19856, + "grad_norm": 0.9825967229389033, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 19856 + }, + { + "epoch": 0.19857, + "grad_norm": 1.0586161238907035, + "learning_rate": 0.003, + "loss": 4.044, + "step": 19857 + }, + { + "epoch": 0.19858, + "grad_norm": 1.0431799883179453, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 19858 + }, + { + "epoch": 0.19859, + "grad_norm": 1.0710632590972815, + "learning_rate": 0.003, + "loss": 4.06, + "step": 19859 + }, + { + "epoch": 0.1986, + "grad_norm": 1.0114921164329083, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 19860 + }, + { + "epoch": 0.19861, + "grad_norm": 1.024169363103151, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 19861 + }, + { + "epoch": 0.19862, + "grad_norm": 1.1950130468125637, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 19862 + }, + { + "epoch": 0.19863, + "grad_norm": 0.8996316370994923, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 19863 + }, + { + "epoch": 0.19864, + "grad_norm": 0.9270993081313836, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 19864 + }, + { + "epoch": 0.19865, + "grad_norm": 0.948318478860946, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 19865 + }, + { + "epoch": 0.19866, + "grad_norm": 0.926939909487656, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 19866 + }, + { + "epoch": 0.19867, + "grad_norm": 1.0475921947252427, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 19867 + }, + { + "epoch": 0.19868, + "grad_norm": 1.2214946053434514, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 19868 + }, + { + "epoch": 0.19869, + "grad_norm": 1.0824062695664582, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 19869 + }, + { + "epoch": 0.1987, + "grad_norm": 1.0177976223676046, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 19870 + }, + { + "epoch": 0.19871, + "grad_norm": 1.100118373911228, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 19871 + }, + { + "epoch": 0.19872, + "grad_norm": 1.2066050789026999, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 19872 + }, + { + "epoch": 0.19873, + "grad_norm": 1.0422301053040883, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 19873 + }, + { + "epoch": 0.19874, + "grad_norm": 1.0371279432219476, + "learning_rate": 0.003, + "loss": 4.062, + "step": 19874 + }, + { + "epoch": 0.19875, + "grad_norm": 0.9690061553140517, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 19875 + }, + { + "epoch": 0.19876, + "grad_norm": 1.0756853410068368, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 19876 + }, + { + "epoch": 0.19877, + "grad_norm": 1.0683528594566978, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 19877 + }, + { + "epoch": 0.19878, + "grad_norm": 1.0672893494791524, + "learning_rate": 0.003, + "loss": 4.059, + "step": 19878 + }, + { + "epoch": 0.19879, + "grad_norm": 1.1632980150225545, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 19879 + }, + { + "epoch": 0.1988, + "grad_norm": 1.0526806445377341, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 19880 + }, + { + "epoch": 0.19881, + "grad_norm": 1.0488387092669185, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 19881 + }, + { + "epoch": 0.19882, + "grad_norm": 0.8264071127786685, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 19882 + }, + { + "epoch": 0.19883, + "grad_norm": 0.8348596972422054, + "learning_rate": 0.003, + "loss": 4.058, + "step": 19883 + }, + { + "epoch": 0.19884, + "grad_norm": 1.0813537492801162, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 19884 + }, + { + "epoch": 0.19885, + "grad_norm": 1.17649818625714, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 19885 + }, + { + "epoch": 0.19886, + "grad_norm": 0.9524816886360823, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 19886 + }, + { + "epoch": 0.19887, + "grad_norm": 1.0516485952776653, + "learning_rate": 0.003, + "loss": 4.0868, + "step": 19887 + }, + { + "epoch": 0.19888, + "grad_norm": 1.0469287909462217, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 19888 + }, + { + "epoch": 0.19889, + "grad_norm": 1.1586243054448133, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 19889 + }, + { + "epoch": 0.1989, + "grad_norm": 1.1485524287588211, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 19890 + }, + { + "epoch": 0.19891, + "grad_norm": 1.0471885482874481, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 19891 + }, + { + "epoch": 0.19892, + "grad_norm": 1.1435932604005468, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 19892 + }, + { + "epoch": 0.19893, + "grad_norm": 0.8536145987054171, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 19893 + }, + { + "epoch": 0.19894, + "grad_norm": 1.0370660863807306, + "learning_rate": 0.003, + "loss": 4.062, + "step": 19894 + }, + { + "epoch": 0.19895, + "grad_norm": 1.2931911316664961, + "learning_rate": 0.003, + "loss": 4.0975, + "step": 19895 + }, + { + "epoch": 0.19896, + "grad_norm": 0.8871615354336367, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 19896 + }, + { + "epoch": 0.19897, + "grad_norm": 0.9329635932026392, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 19897 + }, + { + "epoch": 0.19898, + "grad_norm": 1.0557013540320472, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 19898 + }, + { + "epoch": 0.19899, + "grad_norm": 1.2459863801665323, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 19899 + }, + { + "epoch": 0.199, + "grad_norm": 1.2375681483590995, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 19900 + }, + { + "epoch": 0.19901, + "grad_norm": 0.9145636588313064, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 19901 + }, + { + "epoch": 0.19902, + "grad_norm": 0.9906705081471063, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 19902 + }, + { + "epoch": 0.19903, + "grad_norm": 1.096865830508066, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 19903 + }, + { + "epoch": 0.19904, + "grad_norm": 0.9638739049554811, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 19904 + }, + { + "epoch": 0.19905, + "grad_norm": 1.1860943959504322, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 19905 + }, + { + "epoch": 0.19906, + "grad_norm": 0.8063448560983635, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 19906 + }, + { + "epoch": 0.19907, + "grad_norm": 0.8779602791911347, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 19907 + }, + { + "epoch": 0.19908, + "grad_norm": 0.9755334688330338, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 19908 + }, + { + "epoch": 0.19909, + "grad_norm": 1.37022253770862, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 19909 + }, + { + "epoch": 0.1991, + "grad_norm": 1.1478368434109705, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 19910 + }, + { + "epoch": 0.19911, + "grad_norm": 0.9244748838180736, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 19911 + }, + { + "epoch": 0.19912, + "grad_norm": 0.9297310058711756, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 19912 + }, + { + "epoch": 0.19913, + "grad_norm": 1.0311547097227036, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 19913 + }, + { + "epoch": 0.19914, + "grad_norm": 1.122780214960912, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 19914 + }, + { + "epoch": 0.19915, + "grad_norm": 0.9847123736505232, + "learning_rate": 0.003, + "loss": 4.059, + "step": 19915 + }, + { + "epoch": 0.19916, + "grad_norm": 1.2131676726819087, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 19916 + }, + { + "epoch": 0.19917, + "grad_norm": 0.9984762250229035, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 19917 + }, + { + "epoch": 0.19918, + "grad_norm": 1.0461596641756319, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 19918 + }, + { + "epoch": 0.19919, + "grad_norm": 0.9624503227693484, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 19919 + }, + { + "epoch": 0.1992, + "grad_norm": 0.9921368269293841, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 19920 + }, + { + "epoch": 0.19921, + "grad_norm": 1.0653588294481078, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 19921 + }, + { + "epoch": 0.19922, + "grad_norm": 0.9885061725273373, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 19922 + }, + { + "epoch": 0.19923, + "grad_norm": 1.0664394619722666, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 19923 + }, + { + "epoch": 0.19924, + "grad_norm": 0.9601336447017241, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 19924 + }, + { + "epoch": 0.19925, + "grad_norm": 1.0113656404423663, + "learning_rate": 0.003, + "loss": 4.08, + "step": 19925 + }, + { + "epoch": 0.19926, + "grad_norm": 1.35914640635023, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 19926 + }, + { + "epoch": 0.19927, + "grad_norm": 0.8349033704903348, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 19927 + }, + { + "epoch": 0.19928, + "grad_norm": 0.9624635014841716, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 19928 + }, + { + "epoch": 0.19929, + "grad_norm": 1.1291202857537734, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 19929 + }, + { + "epoch": 0.1993, + "grad_norm": 1.148429931039189, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 19930 + }, + { + "epoch": 0.19931, + "grad_norm": 1.0536227821389863, + "learning_rate": 0.003, + "loss": 4.0964, + "step": 19931 + }, + { + "epoch": 0.19932, + "grad_norm": 1.0869990762557453, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 19932 + }, + { + "epoch": 0.19933, + "grad_norm": 1.142320899194662, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 19933 + }, + { + "epoch": 0.19934, + "grad_norm": 1.1204984729473089, + "learning_rate": 0.003, + "loss": 4.078, + "step": 19934 + }, + { + "epoch": 0.19935, + "grad_norm": 0.9771835657829162, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 19935 + }, + { + "epoch": 0.19936, + "grad_norm": 1.0780948421610808, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 19936 + }, + { + "epoch": 0.19937, + "grad_norm": 1.148176903665943, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 19937 + }, + { + "epoch": 0.19938, + "grad_norm": 1.1746481380533733, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 19938 + }, + { + "epoch": 0.19939, + "grad_norm": 1.0331916294011603, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 19939 + }, + { + "epoch": 0.1994, + "grad_norm": 1.1865602871359768, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 19940 + }, + { + "epoch": 0.19941, + "grad_norm": 0.8736675592461197, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 19941 + }, + { + "epoch": 0.19942, + "grad_norm": 0.8900069007328221, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 19942 + }, + { + "epoch": 0.19943, + "grad_norm": 1.0470205368063563, + "learning_rate": 0.003, + "loss": 4.0856, + "step": 19943 + }, + { + "epoch": 0.19944, + "grad_norm": 1.3147287631187343, + "learning_rate": 0.003, + "loss": 4.0973, + "step": 19944 + }, + { + "epoch": 0.19945, + "grad_norm": 0.8127123262435298, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 19945 + }, + { + "epoch": 0.19946, + "grad_norm": 0.976419946094159, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 19946 + }, + { + "epoch": 0.19947, + "grad_norm": 1.1690681968277135, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 19947 + }, + { + "epoch": 0.19948, + "grad_norm": 1.2501708284039394, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 19948 + }, + { + "epoch": 0.19949, + "grad_norm": 0.9197444308137621, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 19949 + }, + { + "epoch": 0.1995, + "grad_norm": 1.08827695507248, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 19950 + }, + { + "epoch": 0.19951, + "grad_norm": 1.16951530494081, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 19951 + }, + { + "epoch": 0.19952, + "grad_norm": 0.9243863966579413, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 19952 + }, + { + "epoch": 0.19953, + "grad_norm": 0.8318287930724555, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 19953 + }, + { + "epoch": 0.19954, + "grad_norm": 0.8500937505346883, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 19954 + }, + { + "epoch": 0.19955, + "grad_norm": 1.0635892374387155, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 19955 + }, + { + "epoch": 0.19956, + "grad_norm": 1.1961423353253704, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 19956 + }, + { + "epoch": 0.19957, + "grad_norm": 0.8740377451982982, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 19957 + }, + { + "epoch": 0.19958, + "grad_norm": 0.9678239137116622, + "learning_rate": 0.003, + "loss": 4.0873, + "step": 19958 + }, + { + "epoch": 0.19959, + "grad_norm": 1.109898531517767, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 19959 + }, + { + "epoch": 0.1996, + "grad_norm": 1.3970406930764443, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 19960 + }, + { + "epoch": 0.19961, + "grad_norm": 0.8761720506069802, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 19961 + }, + { + "epoch": 0.19962, + "grad_norm": 1.0543140816716072, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 19962 + }, + { + "epoch": 0.19963, + "grad_norm": 1.129906074696742, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 19963 + }, + { + "epoch": 0.19964, + "grad_norm": 1.0462985602596482, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 19964 + }, + { + "epoch": 0.19965, + "grad_norm": 1.2044373702025555, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 19965 + }, + { + "epoch": 0.19966, + "grad_norm": 0.9135060875117148, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 19966 + }, + { + "epoch": 0.19967, + "grad_norm": 0.8493442262690899, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 19967 + }, + { + "epoch": 0.19968, + "grad_norm": 0.8567753566446166, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 19968 + }, + { + "epoch": 0.19969, + "grad_norm": 1.017916962704297, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 19969 + }, + { + "epoch": 0.1997, + "grad_norm": 1.2446746958852435, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 19970 + }, + { + "epoch": 0.19971, + "grad_norm": 0.9228320904253401, + "learning_rate": 0.003, + "loss": 4.021, + "step": 19971 + }, + { + "epoch": 0.19972, + "grad_norm": 0.9936986876262384, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 19972 + }, + { + "epoch": 0.19973, + "grad_norm": 1.3033261639564662, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 19973 + }, + { + "epoch": 0.19974, + "grad_norm": 1.006714505896375, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 19974 + }, + { + "epoch": 0.19975, + "grad_norm": 1.0755573695325717, + "learning_rate": 0.003, + "loss": 4.042, + "step": 19975 + }, + { + "epoch": 0.19976, + "grad_norm": 1.0502603027710726, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 19976 + }, + { + "epoch": 0.19977, + "grad_norm": 1.0981096961984473, + "learning_rate": 0.003, + "loss": 4.021, + "step": 19977 + }, + { + "epoch": 0.19978, + "grad_norm": 1.2007149673242994, + "learning_rate": 0.003, + "loss": 4.0878, + "step": 19978 + }, + { + "epoch": 0.19979, + "grad_norm": 0.8523753452060707, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 19979 + }, + { + "epoch": 0.1998, + "grad_norm": 0.91109591999719, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 19980 + }, + { + "epoch": 0.19981, + "grad_norm": 1.0374589118093063, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 19981 + }, + { + "epoch": 0.19982, + "grad_norm": 1.0643576850136525, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 19982 + }, + { + "epoch": 0.19983, + "grad_norm": 1.0605704724947707, + "learning_rate": 0.003, + "loss": 4.1016, + "step": 19983 + }, + { + "epoch": 0.19984, + "grad_norm": 1.1206092064975077, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 19984 + }, + { + "epoch": 0.19985, + "grad_norm": 1.266984022007813, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 19985 + }, + { + "epoch": 0.19986, + "grad_norm": 0.8494594711220513, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 19986 + }, + { + "epoch": 0.19987, + "grad_norm": 0.8089676136128172, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 19987 + }, + { + "epoch": 0.19988, + "grad_norm": 0.9757551708982394, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 19988 + }, + { + "epoch": 0.19989, + "grad_norm": 1.0806281044524597, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 19989 + }, + { + "epoch": 0.1999, + "grad_norm": 1.0348375175939237, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 19990 + }, + { + "epoch": 0.19991, + "grad_norm": 1.1955171191488299, + "learning_rate": 0.003, + "loss": 4.06, + "step": 19991 + }, + { + "epoch": 0.19992, + "grad_norm": 1.0860898035906374, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 19992 + }, + { + "epoch": 0.19993, + "grad_norm": 1.3056398609160775, + "learning_rate": 0.003, + "loss": 4.0864, + "step": 19993 + }, + { + "epoch": 0.19994, + "grad_norm": 0.8748535613835258, + "learning_rate": 0.003, + "loss": 4.079, + "step": 19994 + }, + { + "epoch": 0.19995, + "grad_norm": 1.0154761867192628, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 19995 + }, + { + "epoch": 0.19996, + "grad_norm": 1.2489501527644002, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 19996 + }, + { + "epoch": 0.19997, + "grad_norm": 1.0903096352851016, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 19997 + }, + { + "epoch": 0.19998, + "grad_norm": 1.0826318766689325, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 19998 + }, + { + "epoch": 0.19999, + "grad_norm": 1.0205576984196343, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 19999 + }, + { + "epoch": 0.2, + "grad_norm": 0.9770527725045092, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 20000 + }, + { + "epoch": 0.20001, + "grad_norm": 1.104819596070291, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 20001 + }, + { + "epoch": 0.20002, + "grad_norm": 1.1291194987788213, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 20002 + }, + { + "epoch": 0.20003, + "grad_norm": 1.0154383305762822, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 20003 + }, + { + "epoch": 0.20004, + "grad_norm": 1.0859456322922039, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 20004 + }, + { + "epoch": 0.20005, + "grad_norm": 0.969588575221007, + "learning_rate": 0.003, + "loss": 4.07, + "step": 20005 + }, + { + "epoch": 0.20006, + "grad_norm": 1.1105561407315958, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 20006 + }, + { + "epoch": 0.20007, + "grad_norm": 1.2096527210004855, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 20007 + }, + { + "epoch": 0.20008, + "grad_norm": 0.9912917196054614, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 20008 + }, + { + "epoch": 0.20009, + "grad_norm": 0.8596378978044155, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 20009 + }, + { + "epoch": 0.2001, + "grad_norm": 0.9495329373021918, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 20010 + }, + { + "epoch": 0.20011, + "grad_norm": 1.2031384698599754, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 20011 + }, + { + "epoch": 0.20012, + "grad_norm": 0.945528553698067, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 20012 + }, + { + "epoch": 0.20013, + "grad_norm": 1.0037702927871375, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 20013 + }, + { + "epoch": 0.20014, + "grad_norm": 1.031972922178036, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 20014 + }, + { + "epoch": 0.20015, + "grad_norm": 1.1544379293294438, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 20015 + }, + { + "epoch": 0.20016, + "grad_norm": 0.9801690642594753, + "learning_rate": 0.003, + "loss": 4.071, + "step": 20016 + }, + { + "epoch": 0.20017, + "grad_norm": 1.0743948701174013, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 20017 + }, + { + "epoch": 0.20018, + "grad_norm": 1.0985449497095807, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 20018 + }, + { + "epoch": 0.20019, + "grad_norm": 1.152044107879043, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 20019 + }, + { + "epoch": 0.2002, + "grad_norm": 0.9333031308140537, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 20020 + }, + { + "epoch": 0.20021, + "grad_norm": 0.9873854080702816, + "learning_rate": 0.003, + "loss": 4.09, + "step": 20021 + }, + { + "epoch": 0.20022, + "grad_norm": 1.0547783292486428, + "learning_rate": 0.003, + "loss": 4.067, + "step": 20022 + }, + { + "epoch": 0.20023, + "grad_norm": 1.2405387683253337, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 20023 + }, + { + "epoch": 0.20024, + "grad_norm": 1.141685646430772, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 20024 + }, + { + "epoch": 0.20025, + "grad_norm": 0.959286860327659, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 20025 + }, + { + "epoch": 0.20026, + "grad_norm": 1.0926780826115616, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 20026 + }, + { + "epoch": 0.20027, + "grad_norm": 1.0773914409153584, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 20027 + }, + { + "epoch": 0.20028, + "grad_norm": 1.1154042185053785, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 20028 + }, + { + "epoch": 0.20029, + "grad_norm": 1.1766612908604677, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 20029 + }, + { + "epoch": 0.2003, + "grad_norm": 1.0912795646541482, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 20030 + }, + { + "epoch": 0.20031, + "grad_norm": 1.136046733951346, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 20031 + }, + { + "epoch": 0.20032, + "grad_norm": 0.9503960052653992, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 20032 + }, + { + "epoch": 0.20033, + "grad_norm": 1.008040427084458, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 20033 + }, + { + "epoch": 0.20034, + "grad_norm": 1.2034342054056162, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 20034 + }, + { + "epoch": 0.20035, + "grad_norm": 0.8954946034540286, + "learning_rate": 0.003, + "loss": 4.056, + "step": 20035 + }, + { + "epoch": 0.20036, + "grad_norm": 1.0711132087812196, + "learning_rate": 0.003, + "loss": 4.097, + "step": 20036 + }, + { + "epoch": 0.20037, + "grad_norm": 1.0774797501119175, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 20037 + }, + { + "epoch": 0.20038, + "grad_norm": 1.0461157601611193, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 20038 + }, + { + "epoch": 0.20039, + "grad_norm": 0.9523653930703428, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 20039 + }, + { + "epoch": 0.2004, + "grad_norm": 1.0551069788003995, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 20040 + }, + { + "epoch": 0.20041, + "grad_norm": 1.011407537085384, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 20041 + }, + { + "epoch": 0.20042, + "grad_norm": 0.9859012874995439, + "learning_rate": 0.003, + "loss": 4.1066, + "step": 20042 + }, + { + "epoch": 0.20043, + "grad_norm": 1.1151827513891501, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 20043 + }, + { + "epoch": 0.20044, + "grad_norm": 1.0347602346780056, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 20044 + }, + { + "epoch": 0.20045, + "grad_norm": 1.2050285835562153, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 20045 + }, + { + "epoch": 0.20046, + "grad_norm": 1.1697498506788333, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 20046 + }, + { + "epoch": 0.20047, + "grad_norm": 1.0690843691888614, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 20047 + }, + { + "epoch": 0.20048, + "grad_norm": 1.2427151905941718, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 20048 + }, + { + "epoch": 0.20049, + "grad_norm": 0.8877648458467586, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 20049 + }, + { + "epoch": 0.2005, + "grad_norm": 0.7900410029124588, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 20050 + }, + { + "epoch": 0.20051, + "grad_norm": 0.7992694360293541, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 20051 + }, + { + "epoch": 0.20052, + "grad_norm": 1.0272990686428278, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 20052 + }, + { + "epoch": 0.20053, + "grad_norm": 1.097053994801997, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 20053 + }, + { + "epoch": 0.20054, + "grad_norm": 1.241372644842951, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 20054 + }, + { + "epoch": 0.20055, + "grad_norm": 0.99035047524419, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 20055 + }, + { + "epoch": 0.20056, + "grad_norm": 1.042799541033347, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 20056 + }, + { + "epoch": 0.20057, + "grad_norm": 1.024121064813439, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 20057 + }, + { + "epoch": 0.20058, + "grad_norm": 1.1717790996035293, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 20058 + }, + { + "epoch": 0.20059, + "grad_norm": 0.8327867590851366, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 20059 + }, + { + "epoch": 0.2006, + "grad_norm": 0.8111869940767044, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 20060 + }, + { + "epoch": 0.20061, + "grad_norm": 0.8525444405932977, + "learning_rate": 0.003, + "loss": 4.069, + "step": 20061 + }, + { + "epoch": 0.20062, + "grad_norm": 0.8468689092854959, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 20062 + }, + { + "epoch": 0.20063, + "grad_norm": 0.8449637708303165, + "learning_rate": 0.003, + "loss": 4.0931, + "step": 20063 + }, + { + "epoch": 0.20064, + "grad_norm": 1.10929136426329, + "learning_rate": 0.003, + "loss": 4.0954, + "step": 20064 + }, + { + "epoch": 0.20065, + "grad_norm": 1.0539647232597982, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 20065 + }, + { + "epoch": 0.20066, + "grad_norm": 0.9872043825507649, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 20066 + }, + { + "epoch": 0.20067, + "grad_norm": 1.1372473318770386, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 20067 + }, + { + "epoch": 0.20068, + "grad_norm": 0.990611886602288, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 20068 + }, + { + "epoch": 0.20069, + "grad_norm": 1.0545452126347132, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 20069 + }, + { + "epoch": 0.2007, + "grad_norm": 1.3221551365370856, + "learning_rate": 0.003, + "loss": 4.0848, + "step": 20070 + }, + { + "epoch": 0.20071, + "grad_norm": 0.9337111536769449, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 20071 + }, + { + "epoch": 0.20072, + "grad_norm": 0.9298550525577872, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 20072 + }, + { + "epoch": 0.20073, + "grad_norm": 1.0680095501009508, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 20073 + }, + { + "epoch": 0.20074, + "grad_norm": 1.3035003257483728, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 20074 + }, + { + "epoch": 0.20075, + "grad_norm": 0.8937405431133216, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 20075 + }, + { + "epoch": 0.20076, + "grad_norm": 0.9784292826595019, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 20076 + }, + { + "epoch": 0.20077, + "grad_norm": 1.308372572430098, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 20077 + }, + { + "epoch": 0.20078, + "grad_norm": 0.8802207665924311, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 20078 + }, + { + "epoch": 0.20079, + "grad_norm": 0.9091535757071906, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 20079 + }, + { + "epoch": 0.2008, + "grad_norm": 1.0410863420169651, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 20080 + }, + { + "epoch": 0.20081, + "grad_norm": 1.1451744862214102, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 20081 + }, + { + "epoch": 0.20082, + "grad_norm": 0.9984541079934195, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 20082 + }, + { + "epoch": 0.20083, + "grad_norm": 1.2688095266329746, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 20083 + }, + { + "epoch": 0.20084, + "grad_norm": 0.9249505820171302, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 20084 + }, + { + "epoch": 0.20085, + "grad_norm": 1.177788131350806, + "learning_rate": 0.003, + "loss": 4.0876, + "step": 20085 + }, + { + "epoch": 0.20086, + "grad_norm": 0.9553215039221508, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 20086 + }, + { + "epoch": 0.20087, + "grad_norm": 1.0117029413784422, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 20087 + }, + { + "epoch": 0.20088, + "grad_norm": 1.1552495515139698, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 20088 + }, + { + "epoch": 0.20089, + "grad_norm": 1.349278930346668, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 20089 + }, + { + "epoch": 0.2009, + "grad_norm": 0.7324801185183755, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 20090 + }, + { + "epoch": 0.20091, + "grad_norm": 0.7941018837768566, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 20091 + }, + { + "epoch": 0.20092, + "grad_norm": 0.9158233280665841, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 20092 + }, + { + "epoch": 0.20093, + "grad_norm": 1.1341842280314918, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 20093 + }, + { + "epoch": 0.20094, + "grad_norm": 1.0964931050624174, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 20094 + }, + { + "epoch": 0.20095, + "grad_norm": 1.2012093823001364, + "learning_rate": 0.003, + "loss": 4.0928, + "step": 20095 + }, + { + "epoch": 0.20096, + "grad_norm": 1.2131101372629949, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 20096 + }, + { + "epoch": 0.20097, + "grad_norm": 1.0884479700002327, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 20097 + }, + { + "epoch": 0.20098, + "grad_norm": 1.1750722298080802, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 20098 + }, + { + "epoch": 0.20099, + "grad_norm": 1.0420135857356139, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 20099 + }, + { + "epoch": 0.201, + "grad_norm": 1.1007774340539218, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 20100 + }, + { + "epoch": 0.20101, + "grad_norm": 0.9268279552261454, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 20101 + }, + { + "epoch": 0.20102, + "grad_norm": 1.0929260324451326, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 20102 + }, + { + "epoch": 0.20103, + "grad_norm": 1.2662511350683263, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 20103 + }, + { + "epoch": 0.20104, + "grad_norm": 0.9969779357340661, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 20104 + }, + { + "epoch": 0.20105, + "grad_norm": 1.0480755103312087, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 20105 + }, + { + "epoch": 0.20106, + "grad_norm": 0.9930424641235542, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 20106 + }, + { + "epoch": 0.20107, + "grad_norm": 1.0872275449364202, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 20107 + }, + { + "epoch": 0.20108, + "grad_norm": 1.0225176849181072, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 20108 + }, + { + "epoch": 0.20109, + "grad_norm": 1.0860868216078132, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 20109 + }, + { + "epoch": 0.2011, + "grad_norm": 0.9898436993716675, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 20110 + }, + { + "epoch": 0.20111, + "grad_norm": 1.070132113265644, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 20111 + }, + { + "epoch": 0.20112, + "grad_norm": 1.0583348349105566, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 20112 + }, + { + "epoch": 0.20113, + "grad_norm": 1.2169301538315058, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 20113 + }, + { + "epoch": 0.20114, + "grad_norm": 1.1704484939314983, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 20114 + }, + { + "epoch": 0.20115, + "grad_norm": 0.9305512781587795, + "learning_rate": 0.003, + "loss": 4.0919, + "step": 20115 + }, + { + "epoch": 0.20116, + "grad_norm": 1.0302006387655913, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 20116 + }, + { + "epoch": 0.20117, + "grad_norm": 1.0779116841911758, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 20117 + }, + { + "epoch": 0.20118, + "grad_norm": 1.1002541010698357, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 20118 + }, + { + "epoch": 0.20119, + "grad_norm": 0.9448146920043455, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 20119 + }, + { + "epoch": 0.2012, + "grad_norm": 1.0459318092282186, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 20120 + }, + { + "epoch": 0.20121, + "grad_norm": 1.0029655150663725, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 20121 + }, + { + "epoch": 0.20122, + "grad_norm": 1.0381547208831676, + "learning_rate": 0.003, + "loss": 4.0933, + "step": 20122 + }, + { + "epoch": 0.20123, + "grad_norm": 1.0689085524853492, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 20123 + }, + { + "epoch": 0.20124, + "grad_norm": 0.9812168531419083, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 20124 + }, + { + "epoch": 0.20125, + "grad_norm": 1.2094429737799957, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 20125 + }, + { + "epoch": 0.20126, + "grad_norm": 1.1898120360421671, + "learning_rate": 0.003, + "loss": 4.043, + "step": 20126 + }, + { + "epoch": 0.20127, + "grad_norm": 1.0002876459389063, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 20127 + }, + { + "epoch": 0.20128, + "grad_norm": 0.9288857594366416, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 20128 + }, + { + "epoch": 0.20129, + "grad_norm": 0.9617040679007447, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 20129 + }, + { + "epoch": 0.2013, + "grad_norm": 1.230146488825064, + "learning_rate": 0.003, + "loss": 4.0874, + "step": 20130 + }, + { + "epoch": 0.20131, + "grad_norm": 1.0147758974948389, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 20131 + }, + { + "epoch": 0.20132, + "grad_norm": 1.0036338758257264, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 20132 + }, + { + "epoch": 0.20133, + "grad_norm": 1.1850724180325538, + "learning_rate": 0.003, + "loss": 4.065, + "step": 20133 + }, + { + "epoch": 0.20134, + "grad_norm": 0.9253895275391559, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 20134 + }, + { + "epoch": 0.20135, + "grad_norm": 1.126589551673606, + "learning_rate": 0.003, + "loss": 4.068, + "step": 20135 + }, + { + "epoch": 0.20136, + "grad_norm": 1.4138326536937396, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 20136 + }, + { + "epoch": 0.20137, + "grad_norm": 0.7155629488793994, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 20137 + }, + { + "epoch": 0.20138, + "grad_norm": 0.7916312926015955, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 20138 + }, + { + "epoch": 0.20139, + "grad_norm": 0.9056996454744319, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 20139 + }, + { + "epoch": 0.2014, + "grad_norm": 1.002850418700553, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 20140 + }, + { + "epoch": 0.20141, + "grad_norm": 1.0558304175005546, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 20141 + }, + { + "epoch": 0.20142, + "grad_norm": 0.8877249607384835, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 20142 + }, + { + "epoch": 0.20143, + "grad_norm": 1.0683020610905936, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 20143 + }, + { + "epoch": 0.20144, + "grad_norm": 1.3340741130800287, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 20144 + }, + { + "epoch": 0.20145, + "grad_norm": 0.9406433859597987, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 20145 + }, + { + "epoch": 0.20146, + "grad_norm": 1.1458049160017492, + "learning_rate": 0.003, + "loss": 4.0857, + "step": 20146 + }, + { + "epoch": 0.20147, + "grad_norm": 1.225214048742926, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 20147 + }, + { + "epoch": 0.20148, + "grad_norm": 1.0772187861257152, + "learning_rate": 0.003, + "loss": 4.067, + "step": 20148 + }, + { + "epoch": 0.20149, + "grad_norm": 1.0757031585498547, + "learning_rate": 0.003, + "loss": 4.053, + "step": 20149 + }, + { + "epoch": 0.2015, + "grad_norm": 1.0028875941096915, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 20150 + }, + { + "epoch": 0.20151, + "grad_norm": 1.1790640036193782, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 20151 + }, + { + "epoch": 0.20152, + "grad_norm": 0.9174324096678705, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 20152 + }, + { + "epoch": 0.20153, + "grad_norm": 1.0415560776537363, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 20153 + }, + { + "epoch": 0.20154, + "grad_norm": 0.9917630667683197, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 20154 + }, + { + "epoch": 0.20155, + "grad_norm": 1.125270358269216, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 20155 + }, + { + "epoch": 0.20156, + "grad_norm": 0.9608908355017673, + "learning_rate": 0.003, + "loss": 4.065, + "step": 20156 + }, + { + "epoch": 0.20157, + "grad_norm": 1.101442670424311, + "learning_rate": 0.003, + "loss": 4.042, + "step": 20157 + }, + { + "epoch": 0.20158, + "grad_norm": 1.3140214331172575, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 20158 + }, + { + "epoch": 0.20159, + "grad_norm": 1.0470803564991324, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 20159 + }, + { + "epoch": 0.2016, + "grad_norm": 1.1486794100731799, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 20160 + }, + { + "epoch": 0.20161, + "grad_norm": 1.0868142956950861, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 20161 + }, + { + "epoch": 0.20162, + "grad_norm": 1.1513222955050189, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 20162 + }, + { + "epoch": 0.20163, + "grad_norm": 1.0114539887003384, + "learning_rate": 0.003, + "loss": 4.072, + "step": 20163 + }, + { + "epoch": 0.20164, + "grad_norm": 1.1892111512077475, + "learning_rate": 0.003, + "loss": 4.0972, + "step": 20164 + }, + { + "epoch": 0.20165, + "grad_norm": 1.1564408987650778, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 20165 + }, + { + "epoch": 0.20166, + "grad_norm": 1.0123972026945887, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 20166 + }, + { + "epoch": 0.20167, + "grad_norm": 1.1828758818437044, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 20167 + }, + { + "epoch": 0.20168, + "grad_norm": 0.9930207212342277, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 20168 + }, + { + "epoch": 0.20169, + "grad_norm": 1.0945872167791806, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 20169 + }, + { + "epoch": 0.2017, + "grad_norm": 0.9885331086596368, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 20170 + }, + { + "epoch": 0.20171, + "grad_norm": 0.9235160087429681, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 20171 + }, + { + "epoch": 0.20172, + "grad_norm": 1.0189892262236893, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 20172 + }, + { + "epoch": 0.20173, + "grad_norm": 1.1555895917102725, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 20173 + }, + { + "epoch": 0.20174, + "grad_norm": 0.9497367350214524, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 20174 + }, + { + "epoch": 0.20175, + "grad_norm": 1.1045217005081487, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 20175 + }, + { + "epoch": 0.20176, + "grad_norm": 1.074818106163867, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 20176 + }, + { + "epoch": 0.20177, + "grad_norm": 0.8841230077435683, + "learning_rate": 0.003, + "loss": 4.065, + "step": 20177 + }, + { + "epoch": 0.20178, + "grad_norm": 0.8395926621628859, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 20178 + }, + { + "epoch": 0.20179, + "grad_norm": 1.0716909676690753, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 20179 + }, + { + "epoch": 0.2018, + "grad_norm": 1.0881773599669438, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 20180 + }, + { + "epoch": 0.20181, + "grad_norm": 1.1172400885009979, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 20181 + }, + { + "epoch": 0.20182, + "grad_norm": 1.1137852551091212, + "learning_rate": 0.003, + "loss": 4.069, + "step": 20182 + }, + { + "epoch": 0.20183, + "grad_norm": 1.0687341878027536, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 20183 + }, + { + "epoch": 0.20184, + "grad_norm": 1.0781173500427328, + "learning_rate": 0.003, + "loss": 4.1013, + "step": 20184 + }, + { + "epoch": 0.20185, + "grad_norm": 1.1000172540743685, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 20185 + }, + { + "epoch": 0.20186, + "grad_norm": 1.0602143130722947, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 20186 + }, + { + "epoch": 0.20187, + "grad_norm": 1.2189877498747546, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 20187 + }, + { + "epoch": 0.20188, + "grad_norm": 0.8208298080464387, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 20188 + }, + { + "epoch": 0.20189, + "grad_norm": 0.8903098120836465, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 20189 + }, + { + "epoch": 0.2019, + "grad_norm": 0.9170882026135706, + "learning_rate": 0.003, + "loss": 4.066, + "step": 20190 + }, + { + "epoch": 0.20191, + "grad_norm": 0.9744210879454257, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 20191 + }, + { + "epoch": 0.20192, + "grad_norm": 1.3244950281479533, + "learning_rate": 0.003, + "loss": 4.072, + "step": 20192 + }, + { + "epoch": 0.20193, + "grad_norm": 1.059466424831226, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 20193 + }, + { + "epoch": 0.20194, + "grad_norm": 1.1227881368449835, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 20194 + }, + { + "epoch": 0.20195, + "grad_norm": 0.8964923480547893, + "learning_rate": 0.003, + "loss": 4.11, + "step": 20195 + }, + { + "epoch": 0.20196, + "grad_norm": 0.8808052665093309, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 20196 + }, + { + "epoch": 0.20197, + "grad_norm": 1.012819300751165, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 20197 + }, + { + "epoch": 0.20198, + "grad_norm": 1.1756749441609071, + "learning_rate": 0.003, + "loss": 4.04, + "step": 20198 + }, + { + "epoch": 0.20199, + "grad_norm": 0.9588678516669384, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 20199 + }, + { + "epoch": 0.202, + "grad_norm": 1.2068256528734576, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 20200 + }, + { + "epoch": 0.20201, + "grad_norm": 1.106936914847982, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 20201 + }, + { + "epoch": 0.20202, + "grad_norm": 1.1733306201001892, + "learning_rate": 0.003, + "loss": 4.1066, + "step": 20202 + }, + { + "epoch": 0.20203, + "grad_norm": 1.0747506452538336, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 20203 + }, + { + "epoch": 0.20204, + "grad_norm": 1.1742226598984955, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 20204 + }, + { + "epoch": 0.20205, + "grad_norm": 1.0580734797744125, + "learning_rate": 0.003, + "loss": 4.074, + "step": 20205 + }, + { + "epoch": 0.20206, + "grad_norm": 1.0488014788817372, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 20206 + }, + { + "epoch": 0.20207, + "grad_norm": 1.047286334456402, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 20207 + }, + { + "epoch": 0.20208, + "grad_norm": 1.2440867360777985, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 20208 + }, + { + "epoch": 0.20209, + "grad_norm": 1.1602244835018176, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 20209 + }, + { + "epoch": 0.2021, + "grad_norm": 0.9165956703538142, + "learning_rate": 0.003, + "loss": 4.057, + "step": 20210 + }, + { + "epoch": 0.20211, + "grad_norm": 0.8136076966903752, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 20211 + }, + { + "epoch": 0.20212, + "grad_norm": 0.9015133132782237, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 20212 + }, + { + "epoch": 0.20213, + "grad_norm": 0.8920500908422263, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 20213 + }, + { + "epoch": 0.20214, + "grad_norm": 0.9369172045179152, + "learning_rate": 0.003, + "loss": 4.019, + "step": 20214 + }, + { + "epoch": 0.20215, + "grad_norm": 1.0995632235646036, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 20215 + }, + { + "epoch": 0.20216, + "grad_norm": 0.9553266997951407, + "learning_rate": 0.003, + "loss": 4.0937, + "step": 20216 + }, + { + "epoch": 0.20217, + "grad_norm": 1.1756549040888895, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 20217 + }, + { + "epoch": 0.20218, + "grad_norm": 1.251952337457048, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 20218 + }, + { + "epoch": 0.20219, + "grad_norm": 1.0048003967428603, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 20219 + }, + { + "epoch": 0.2022, + "grad_norm": 1.133589681815616, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 20220 + }, + { + "epoch": 0.20221, + "grad_norm": 1.0893221105292012, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 20221 + }, + { + "epoch": 0.20222, + "grad_norm": 1.106023311537448, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 20222 + }, + { + "epoch": 0.20223, + "grad_norm": 0.9261104591683657, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 20223 + }, + { + "epoch": 0.20224, + "grad_norm": 1.1097276587069098, + "learning_rate": 0.003, + "loss": 4.0913, + "step": 20224 + }, + { + "epoch": 0.20225, + "grad_norm": 1.0638230534596647, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 20225 + }, + { + "epoch": 0.20226, + "grad_norm": 1.028596752703774, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 20226 + }, + { + "epoch": 0.20227, + "grad_norm": 1.008829623846466, + "learning_rate": 0.003, + "loss": 4.038, + "step": 20227 + }, + { + "epoch": 0.20228, + "grad_norm": 1.0328628829530948, + "learning_rate": 0.003, + "loss": 4.033, + "step": 20228 + }, + { + "epoch": 0.20229, + "grad_norm": 1.011256724598991, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 20229 + }, + { + "epoch": 0.2023, + "grad_norm": 1.0072197418661677, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 20230 + }, + { + "epoch": 0.20231, + "grad_norm": 1.1643173761988441, + "learning_rate": 0.003, + "loss": 4.078, + "step": 20231 + }, + { + "epoch": 0.20232, + "grad_norm": 0.767790200817118, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 20232 + }, + { + "epoch": 0.20233, + "grad_norm": 0.8723225750675377, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 20233 + }, + { + "epoch": 0.20234, + "grad_norm": 1.0003320193108192, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 20234 + }, + { + "epoch": 0.20235, + "grad_norm": 1.2125566555463978, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 20235 + }, + { + "epoch": 0.20236, + "grad_norm": 0.9397580203532276, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 20236 + }, + { + "epoch": 0.20237, + "grad_norm": 0.9402575790453416, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 20237 + }, + { + "epoch": 0.20238, + "grad_norm": 1.0638283658419836, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 20238 + }, + { + "epoch": 0.20239, + "grad_norm": 1.0857376608490426, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 20239 + }, + { + "epoch": 0.2024, + "grad_norm": 1.1890718383103422, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 20240 + }, + { + "epoch": 0.20241, + "grad_norm": 1.055998062886969, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 20241 + }, + { + "epoch": 0.20242, + "grad_norm": 1.1273917624566197, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 20242 + }, + { + "epoch": 0.20243, + "grad_norm": 1.262613076956378, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 20243 + }, + { + "epoch": 0.20244, + "grad_norm": 1.3942107151504828, + "learning_rate": 0.003, + "loss": 4.0912, + "step": 20244 + }, + { + "epoch": 0.20245, + "grad_norm": 0.9201464968727514, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 20245 + }, + { + "epoch": 0.20246, + "grad_norm": 0.9636337408925404, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 20246 + }, + { + "epoch": 0.20247, + "grad_norm": 0.9523496776257118, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 20247 + }, + { + "epoch": 0.20248, + "grad_norm": 1.0953526107417304, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 20248 + }, + { + "epoch": 0.20249, + "grad_norm": 1.1159268985297133, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 20249 + }, + { + "epoch": 0.2025, + "grad_norm": 0.9396667515011566, + "learning_rate": 0.003, + "loss": 4.055, + "step": 20250 + }, + { + "epoch": 0.20251, + "grad_norm": 1.15733558826939, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 20251 + }, + { + "epoch": 0.20252, + "grad_norm": 1.03754333135072, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 20252 + }, + { + "epoch": 0.20253, + "grad_norm": 1.1502786017260327, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 20253 + }, + { + "epoch": 0.20254, + "grad_norm": 0.8447197694645501, + "learning_rate": 0.003, + "loss": 4.1001, + "step": 20254 + }, + { + "epoch": 0.20255, + "grad_norm": 0.8651490852140094, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 20255 + }, + { + "epoch": 0.20256, + "grad_norm": 1.0947173244407642, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 20256 + }, + { + "epoch": 0.20257, + "grad_norm": 0.9306859843440506, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 20257 + }, + { + "epoch": 0.20258, + "grad_norm": 0.8328752297060158, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 20258 + }, + { + "epoch": 0.20259, + "grad_norm": 1.0412792086740676, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 20259 + }, + { + "epoch": 0.2026, + "grad_norm": 1.292793576986174, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 20260 + }, + { + "epoch": 0.20261, + "grad_norm": 1.057609034361059, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 20261 + }, + { + "epoch": 0.20262, + "grad_norm": 1.2681415158758542, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 20262 + }, + { + "epoch": 0.20263, + "grad_norm": 0.7946262682702172, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 20263 + }, + { + "epoch": 0.20264, + "grad_norm": 0.8272498431845223, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 20264 + }, + { + "epoch": 0.20265, + "grad_norm": 1.0457795804742915, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 20265 + }, + { + "epoch": 0.20266, + "grad_norm": 1.269320725838732, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 20266 + }, + { + "epoch": 0.20267, + "grad_norm": 1.0755031065985503, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 20267 + }, + { + "epoch": 0.20268, + "grad_norm": 1.0035815089462619, + "learning_rate": 0.003, + "loss": 4.072, + "step": 20268 + }, + { + "epoch": 0.20269, + "grad_norm": 1.0204514292292992, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 20269 + }, + { + "epoch": 0.2027, + "grad_norm": 1.0232549361480245, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 20270 + }, + { + "epoch": 0.20271, + "grad_norm": 1.1374232753052398, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 20271 + }, + { + "epoch": 0.20272, + "grad_norm": 0.9188737976550341, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 20272 + }, + { + "epoch": 0.20273, + "grad_norm": 1.0089889730370019, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 20273 + }, + { + "epoch": 0.20274, + "grad_norm": 1.2935559737481868, + "learning_rate": 0.003, + "loss": 4.0905, + "step": 20274 + }, + { + "epoch": 0.20275, + "grad_norm": 1.0357376468389752, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 20275 + }, + { + "epoch": 0.20276, + "grad_norm": 1.3613024682018602, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 20276 + }, + { + "epoch": 0.20277, + "grad_norm": 1.0991853565096068, + "learning_rate": 0.003, + "loss": 4.0884, + "step": 20277 + }, + { + "epoch": 0.20278, + "grad_norm": 1.0348305401819256, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 20278 + }, + { + "epoch": 0.20279, + "grad_norm": 1.0911615088758024, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 20279 + }, + { + "epoch": 0.2028, + "grad_norm": 1.1325942830066007, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 20280 + }, + { + "epoch": 0.20281, + "grad_norm": 1.0123851830845978, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 20281 + }, + { + "epoch": 0.20282, + "grad_norm": 1.0923042726820358, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 20282 + }, + { + "epoch": 0.20283, + "grad_norm": 0.9658994455476915, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 20283 + }, + { + "epoch": 0.20284, + "grad_norm": 0.9102219249549409, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 20284 + }, + { + "epoch": 0.20285, + "grad_norm": 0.9228670451117957, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 20285 + }, + { + "epoch": 0.20286, + "grad_norm": 0.9824486381795884, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 20286 + }, + { + "epoch": 0.20287, + "grad_norm": 1.212435782598217, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 20287 + }, + { + "epoch": 0.20288, + "grad_norm": 0.9071292748486482, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 20288 + }, + { + "epoch": 0.20289, + "grad_norm": 1.2129968082045786, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 20289 + }, + { + "epoch": 0.2029, + "grad_norm": 1.3479534796025638, + "learning_rate": 0.003, + "loss": 4.082, + "step": 20290 + }, + { + "epoch": 0.20291, + "grad_norm": 0.7509246358952387, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 20291 + }, + { + "epoch": 0.20292, + "grad_norm": 0.6996775641513888, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 20292 + }, + { + "epoch": 0.20293, + "grad_norm": 0.7941231309897544, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 20293 + }, + { + "epoch": 0.20294, + "grad_norm": 1.0029755388367063, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 20294 + }, + { + "epoch": 0.20295, + "grad_norm": 1.1209857163082242, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 20295 + }, + { + "epoch": 0.20296, + "grad_norm": 0.8883299370113532, + "learning_rate": 0.003, + "loss": 4.039, + "step": 20296 + }, + { + "epoch": 0.20297, + "grad_norm": 0.9359765257747988, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 20297 + }, + { + "epoch": 0.20298, + "grad_norm": 1.1143836324179799, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 20298 + }, + { + "epoch": 0.20299, + "grad_norm": 0.9486260856275727, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 20299 + }, + { + "epoch": 0.203, + "grad_norm": 1.0544649921340832, + "learning_rate": 0.003, + "loss": 4.0819, + "step": 20300 + }, + { + "epoch": 0.20301, + "grad_norm": 1.130496361246096, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 20301 + }, + { + "epoch": 0.20302, + "grad_norm": 1.0424236415678805, + "learning_rate": 0.003, + "loss": 4.077, + "step": 20302 + }, + { + "epoch": 0.20303, + "grad_norm": 1.2283920234558934, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 20303 + }, + { + "epoch": 0.20304, + "grad_norm": 1.0411782341209161, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 20304 + }, + { + "epoch": 0.20305, + "grad_norm": 1.2035430751475944, + "learning_rate": 0.003, + "loss": 4.083, + "step": 20305 + }, + { + "epoch": 0.20306, + "grad_norm": 1.0653964108031593, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 20306 + }, + { + "epoch": 0.20307, + "grad_norm": 1.1778191464524344, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 20307 + }, + { + "epoch": 0.20308, + "grad_norm": 1.2292087170983588, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 20308 + }, + { + "epoch": 0.20309, + "grad_norm": 1.206608797680219, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 20309 + }, + { + "epoch": 0.2031, + "grad_norm": 0.9341674512761203, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 20310 + }, + { + "epoch": 0.20311, + "grad_norm": 0.9236553384814545, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 20311 + }, + { + "epoch": 0.20312, + "grad_norm": 0.9466216760250811, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 20312 + }, + { + "epoch": 0.20313, + "grad_norm": 0.9936103849820657, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 20313 + }, + { + "epoch": 0.20314, + "grad_norm": 1.0256017860157565, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 20314 + }, + { + "epoch": 0.20315, + "grad_norm": 1.0529409603127173, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 20315 + }, + { + "epoch": 0.20316, + "grad_norm": 1.1006368286160666, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 20316 + }, + { + "epoch": 0.20317, + "grad_norm": 1.0861371125312453, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 20317 + }, + { + "epoch": 0.20318, + "grad_norm": 1.0733141876102592, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 20318 + }, + { + "epoch": 0.20319, + "grad_norm": 1.0810815751686011, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 20319 + }, + { + "epoch": 0.2032, + "grad_norm": 1.0373565419119792, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 20320 + }, + { + "epoch": 0.20321, + "grad_norm": 1.0677290075762826, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 20321 + }, + { + "epoch": 0.20322, + "grad_norm": 1.149668847460244, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 20322 + }, + { + "epoch": 0.20323, + "grad_norm": 0.8979819909567669, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 20323 + }, + { + "epoch": 0.20324, + "grad_norm": 0.9877737926050604, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 20324 + }, + { + "epoch": 0.20325, + "grad_norm": 1.1616341930927203, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 20325 + }, + { + "epoch": 0.20326, + "grad_norm": 1.1605391919805796, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 20326 + }, + { + "epoch": 0.20327, + "grad_norm": 0.9812484340563865, + "learning_rate": 0.003, + "loss": 4.06, + "step": 20327 + }, + { + "epoch": 0.20328, + "grad_norm": 0.9722216579746589, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 20328 + }, + { + "epoch": 0.20329, + "grad_norm": 1.1386935461463064, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 20329 + }, + { + "epoch": 0.2033, + "grad_norm": 1.056052701995016, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 20330 + }, + { + "epoch": 0.20331, + "grad_norm": 1.0056807924151139, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 20331 + }, + { + "epoch": 0.20332, + "grad_norm": 1.1827779129570648, + "learning_rate": 0.003, + "loss": 4.048, + "step": 20332 + }, + { + "epoch": 0.20333, + "grad_norm": 0.9068937728003621, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 20333 + }, + { + "epoch": 0.20334, + "grad_norm": 0.9750639783836725, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 20334 + }, + { + "epoch": 0.20335, + "grad_norm": 1.118756436563454, + "learning_rate": 0.003, + "loss": 4.031, + "step": 20335 + }, + { + "epoch": 0.20336, + "grad_norm": 1.0801039147725, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 20336 + }, + { + "epoch": 0.20337, + "grad_norm": 1.0489924411646294, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 20337 + }, + { + "epoch": 0.20338, + "grad_norm": 1.1688067204945856, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 20338 + }, + { + "epoch": 0.20339, + "grad_norm": 1.055490165451117, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 20339 + }, + { + "epoch": 0.2034, + "grad_norm": 1.2353779069155888, + "learning_rate": 0.003, + "loss": 4.0886, + "step": 20340 + }, + { + "epoch": 0.20341, + "grad_norm": 0.9857478280806118, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 20341 + }, + { + "epoch": 0.20342, + "grad_norm": 0.9980001967339112, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 20342 + }, + { + "epoch": 0.20343, + "grad_norm": 1.105429601179383, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 20343 + }, + { + "epoch": 0.20344, + "grad_norm": 0.8814173134184274, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 20344 + }, + { + "epoch": 0.20345, + "grad_norm": 0.9142590174120341, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 20345 + }, + { + "epoch": 0.20346, + "grad_norm": 0.9245820511312445, + "learning_rate": 0.003, + "loss": 4.023, + "step": 20346 + }, + { + "epoch": 0.20347, + "grad_norm": 1.0366952625173518, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 20347 + }, + { + "epoch": 0.20348, + "grad_norm": 1.0805789556178416, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 20348 + }, + { + "epoch": 0.20349, + "grad_norm": 1.1192169908872946, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 20349 + }, + { + "epoch": 0.2035, + "grad_norm": 0.9046129314561049, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 20350 + }, + { + "epoch": 0.20351, + "grad_norm": 1.0752809198149396, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 20351 + }, + { + "epoch": 0.20352, + "grad_norm": 0.9497476260950626, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 20352 + }, + { + "epoch": 0.20353, + "grad_norm": 1.0145092670724505, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 20353 + }, + { + "epoch": 0.20354, + "grad_norm": 1.1823863845629197, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 20354 + }, + { + "epoch": 0.20355, + "grad_norm": 1.1646275144247464, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 20355 + }, + { + "epoch": 0.20356, + "grad_norm": 1.12034985268271, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 20356 + }, + { + "epoch": 0.20357, + "grad_norm": 1.124969966942962, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 20357 + }, + { + "epoch": 0.20358, + "grad_norm": 1.035532428916575, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 20358 + }, + { + "epoch": 0.20359, + "grad_norm": 1.014249033854247, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 20359 + }, + { + "epoch": 0.2036, + "grad_norm": 1.038576449978239, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 20360 + }, + { + "epoch": 0.20361, + "grad_norm": 1.0233229617805308, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 20361 + }, + { + "epoch": 0.20362, + "grad_norm": 1.1350664653382718, + "learning_rate": 0.003, + "loss": 4.052, + "step": 20362 + }, + { + "epoch": 0.20363, + "grad_norm": 1.0676652667191704, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 20363 + }, + { + "epoch": 0.20364, + "grad_norm": 0.981725716608127, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 20364 + }, + { + "epoch": 0.20365, + "grad_norm": 1.0419508294170274, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 20365 + }, + { + "epoch": 0.20366, + "grad_norm": 1.1440024933738542, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 20366 + }, + { + "epoch": 0.20367, + "grad_norm": 1.082510430070646, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 20367 + }, + { + "epoch": 0.20368, + "grad_norm": 1.0509117989728258, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 20368 + }, + { + "epoch": 0.20369, + "grad_norm": 1.0208555790880747, + "learning_rate": 0.003, + "loss": 4.0819, + "step": 20369 + }, + { + "epoch": 0.2037, + "grad_norm": 1.0803306206446845, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 20370 + }, + { + "epoch": 0.20371, + "grad_norm": 1.1340664883250027, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 20371 + }, + { + "epoch": 0.20372, + "grad_norm": 0.9453621094893181, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 20372 + }, + { + "epoch": 0.20373, + "grad_norm": 1.2372478986981708, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 20373 + }, + { + "epoch": 0.20374, + "grad_norm": 1.1982784023854653, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 20374 + }, + { + "epoch": 0.20375, + "grad_norm": 1.218262309503704, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 20375 + }, + { + "epoch": 0.20376, + "grad_norm": 1.111008000591327, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 20376 + }, + { + "epoch": 0.20377, + "grad_norm": 0.9093508678910939, + "learning_rate": 0.003, + "loss": 4.077, + "step": 20377 + }, + { + "epoch": 0.20378, + "grad_norm": 1.1206151212097026, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 20378 + }, + { + "epoch": 0.20379, + "grad_norm": 0.9324756073946757, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 20379 + }, + { + "epoch": 0.2038, + "grad_norm": 1.182259095151916, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 20380 + }, + { + "epoch": 0.20381, + "grad_norm": 1.089646557562275, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 20381 + }, + { + "epoch": 0.20382, + "grad_norm": 0.9682840924809557, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 20382 + }, + { + "epoch": 0.20383, + "grad_norm": 1.0302645324763875, + "learning_rate": 0.003, + "loss": 4.075, + "step": 20383 + }, + { + "epoch": 0.20384, + "grad_norm": 1.2318952065222333, + "learning_rate": 0.003, + "loss": 4.0933, + "step": 20384 + }, + { + "epoch": 0.20385, + "grad_norm": 0.7923564505047604, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 20385 + }, + { + "epoch": 0.20386, + "grad_norm": 0.7792002214825937, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 20386 + }, + { + "epoch": 0.20387, + "grad_norm": 0.9187868655813682, + "learning_rate": 0.003, + "loss": 4.066, + "step": 20387 + }, + { + "epoch": 0.20388, + "grad_norm": 1.0587817252866132, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 20388 + }, + { + "epoch": 0.20389, + "grad_norm": 1.1786439258092174, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 20389 + }, + { + "epoch": 0.2039, + "grad_norm": 0.9711945135997272, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 20390 + }, + { + "epoch": 0.20391, + "grad_norm": 1.1030889828599049, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 20391 + }, + { + "epoch": 0.20392, + "grad_norm": 1.3314982474462895, + "learning_rate": 0.003, + "loss": 4.0991, + "step": 20392 + }, + { + "epoch": 0.20393, + "grad_norm": 1.0258103885094643, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 20393 + }, + { + "epoch": 0.20394, + "grad_norm": 1.0650273103540555, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 20394 + }, + { + "epoch": 0.20395, + "grad_norm": 1.0604349930252408, + "learning_rate": 0.003, + "loss": 4.0848, + "step": 20395 + }, + { + "epoch": 0.20396, + "grad_norm": 1.0259895391533753, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 20396 + }, + { + "epoch": 0.20397, + "grad_norm": 1.1170823160204773, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 20397 + }, + { + "epoch": 0.20398, + "grad_norm": 1.1467649821710673, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 20398 + }, + { + "epoch": 0.20399, + "grad_norm": 1.137855330586396, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 20399 + }, + { + "epoch": 0.204, + "grad_norm": 0.8077475089549623, + "learning_rate": 0.003, + "loss": 4.0884, + "step": 20400 + }, + { + "epoch": 0.20401, + "grad_norm": 0.7855646679398862, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 20401 + }, + { + "epoch": 0.20402, + "grad_norm": 0.9178093035387119, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 20402 + }, + { + "epoch": 0.20403, + "grad_norm": 1.2475596079616316, + "learning_rate": 0.003, + "loss": 4.052, + "step": 20403 + }, + { + "epoch": 0.20404, + "grad_norm": 1.0984179357240316, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 20404 + }, + { + "epoch": 0.20405, + "grad_norm": 1.1714864920764012, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 20405 + }, + { + "epoch": 0.20406, + "grad_norm": 1.1180575004351172, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 20406 + }, + { + "epoch": 0.20407, + "grad_norm": 0.985955738904536, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 20407 + }, + { + "epoch": 0.20408, + "grad_norm": 1.0644769937375158, + "learning_rate": 0.003, + "loss": 4.065, + "step": 20408 + }, + { + "epoch": 0.20409, + "grad_norm": 1.064960727954865, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 20409 + }, + { + "epoch": 0.2041, + "grad_norm": 1.1972086565987772, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 20410 + }, + { + "epoch": 0.20411, + "grad_norm": 1.052531726671447, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 20411 + }, + { + "epoch": 0.20412, + "grad_norm": 1.0471877347353198, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 20412 + }, + { + "epoch": 0.20413, + "grad_norm": 1.215294351440432, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 20413 + }, + { + "epoch": 0.20414, + "grad_norm": 0.9548994361355647, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 20414 + }, + { + "epoch": 0.20415, + "grad_norm": 0.8542376885470483, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 20415 + }, + { + "epoch": 0.20416, + "grad_norm": 0.962127170027431, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 20416 + }, + { + "epoch": 0.20417, + "grad_norm": 1.0670317575482393, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 20417 + }, + { + "epoch": 0.20418, + "grad_norm": 1.1635460702878937, + "learning_rate": 0.003, + "loss": 4.051, + "step": 20418 + }, + { + "epoch": 0.20419, + "grad_norm": 0.9199136381346804, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 20419 + }, + { + "epoch": 0.2042, + "grad_norm": 0.9547927375251636, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 20420 + }, + { + "epoch": 0.20421, + "grad_norm": 0.9660170141099914, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 20421 + }, + { + "epoch": 0.20422, + "grad_norm": 1.4171885967984, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 20422 + }, + { + "epoch": 0.20423, + "grad_norm": 0.9390863933150786, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 20423 + }, + { + "epoch": 0.20424, + "grad_norm": 0.9698916970071754, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 20424 + }, + { + "epoch": 0.20425, + "grad_norm": 1.2267882575275422, + "learning_rate": 0.003, + "loss": 4.065, + "step": 20425 + }, + { + "epoch": 0.20426, + "grad_norm": 0.9798368096312204, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 20426 + }, + { + "epoch": 0.20427, + "grad_norm": 1.0640744721766973, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 20427 + }, + { + "epoch": 0.20428, + "grad_norm": 1.1381108802965312, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 20428 + }, + { + "epoch": 0.20429, + "grad_norm": 1.002232343286511, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 20429 + }, + { + "epoch": 0.2043, + "grad_norm": 1.0406377296057876, + "learning_rate": 0.003, + "loss": 4.0878, + "step": 20430 + }, + { + "epoch": 0.20431, + "grad_norm": 1.0446398814940294, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 20431 + }, + { + "epoch": 0.20432, + "grad_norm": 1.1451070811266124, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 20432 + }, + { + "epoch": 0.20433, + "grad_norm": 1.0059838066103037, + "learning_rate": 0.003, + "loss": 4.054, + "step": 20433 + }, + { + "epoch": 0.20434, + "grad_norm": 1.0964234703357283, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 20434 + }, + { + "epoch": 0.20435, + "grad_norm": 0.9816518615523494, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 20435 + }, + { + "epoch": 0.20436, + "grad_norm": 1.0709926349279455, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 20436 + }, + { + "epoch": 0.20437, + "grad_norm": 1.0879043812610842, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 20437 + }, + { + "epoch": 0.20438, + "grad_norm": 0.945308628862119, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 20438 + }, + { + "epoch": 0.20439, + "grad_norm": 1.0698629109815088, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 20439 + }, + { + "epoch": 0.2044, + "grad_norm": 1.23783389376985, + "learning_rate": 0.003, + "loss": 4.063, + "step": 20440 + }, + { + "epoch": 0.20441, + "grad_norm": 1.2598569706255083, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 20441 + }, + { + "epoch": 0.20442, + "grad_norm": 1.0320493290871886, + "learning_rate": 0.003, + "loss": 4.061, + "step": 20442 + }, + { + "epoch": 0.20443, + "grad_norm": 1.1576723769838313, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 20443 + }, + { + "epoch": 0.20444, + "grad_norm": 1.0044484508433689, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 20444 + }, + { + "epoch": 0.20445, + "grad_norm": 1.0782668338295993, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 20445 + }, + { + "epoch": 0.20446, + "grad_norm": 0.945432675756064, + "learning_rate": 0.003, + "loss": 4.064, + "step": 20446 + }, + { + "epoch": 0.20447, + "grad_norm": 0.9477154424280836, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 20447 + }, + { + "epoch": 0.20448, + "grad_norm": 1.0759059509692537, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 20448 + }, + { + "epoch": 0.20449, + "grad_norm": 1.1298996199514046, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 20449 + }, + { + "epoch": 0.2045, + "grad_norm": 1.0237715627856718, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 20450 + }, + { + "epoch": 0.20451, + "grad_norm": 1.1721146880333415, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 20451 + }, + { + "epoch": 0.20452, + "grad_norm": 1.0583564154275742, + "learning_rate": 0.003, + "loss": 4.045, + "step": 20452 + }, + { + "epoch": 0.20453, + "grad_norm": 1.173819572640895, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 20453 + }, + { + "epoch": 0.20454, + "grad_norm": 1.1883257566544159, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 20454 + }, + { + "epoch": 0.20455, + "grad_norm": 0.9328589408806967, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 20455 + }, + { + "epoch": 0.20456, + "grad_norm": 0.9617793615147908, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 20456 + }, + { + "epoch": 0.20457, + "grad_norm": 1.1791390615254953, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 20457 + }, + { + "epoch": 0.20458, + "grad_norm": 1.071627735044496, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 20458 + }, + { + "epoch": 0.20459, + "grad_norm": 1.079291622012755, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 20459 + }, + { + "epoch": 0.2046, + "grad_norm": 0.9839497542256438, + "learning_rate": 0.003, + "loss": 4.052, + "step": 20460 + }, + { + "epoch": 0.20461, + "grad_norm": 0.955425504804964, + "learning_rate": 0.003, + "loss": 4.063, + "step": 20461 + }, + { + "epoch": 0.20462, + "grad_norm": 0.9390055912307708, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 20462 + }, + { + "epoch": 0.20463, + "grad_norm": 0.9875929378562353, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 20463 + }, + { + "epoch": 0.20464, + "grad_norm": 1.1796055193483816, + "learning_rate": 0.003, + "loss": 4.0952, + "step": 20464 + }, + { + "epoch": 0.20465, + "grad_norm": 1.0221628985725175, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 20465 + }, + { + "epoch": 0.20466, + "grad_norm": 1.133141371058565, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 20466 + }, + { + "epoch": 0.20467, + "grad_norm": 1.0648465760066406, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 20467 + }, + { + "epoch": 0.20468, + "grad_norm": 1.0591576363894424, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 20468 + }, + { + "epoch": 0.20469, + "grad_norm": 1.125221270822581, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 20469 + }, + { + "epoch": 0.2047, + "grad_norm": 1.0602995857764592, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 20470 + }, + { + "epoch": 0.20471, + "grad_norm": 1.0550731169838814, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 20471 + }, + { + "epoch": 0.20472, + "grad_norm": 1.2590096146798935, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 20472 + }, + { + "epoch": 0.20473, + "grad_norm": 0.9110621412819447, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 20473 + }, + { + "epoch": 0.20474, + "grad_norm": 0.9931426028807101, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 20474 + }, + { + "epoch": 0.20475, + "grad_norm": 1.2340544897801664, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 20475 + }, + { + "epoch": 0.20476, + "grad_norm": 0.8475701780679876, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 20476 + }, + { + "epoch": 0.20477, + "grad_norm": 0.8554975373875899, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 20477 + }, + { + "epoch": 0.20478, + "grad_norm": 0.8284540639677718, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 20478 + }, + { + "epoch": 0.20479, + "grad_norm": 0.9988679443547973, + "learning_rate": 0.003, + "loss": 4.0888, + "step": 20479 + }, + { + "epoch": 0.2048, + "grad_norm": 1.1026289133227725, + "learning_rate": 0.003, + "loss": 4.072, + "step": 20480 + }, + { + "epoch": 0.20481, + "grad_norm": 1.0138613392162443, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 20481 + }, + { + "epoch": 0.20482, + "grad_norm": 1.1556578735719398, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 20482 + }, + { + "epoch": 0.20483, + "grad_norm": 1.0818788547728153, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 20483 + }, + { + "epoch": 0.20484, + "grad_norm": 1.1897951228329762, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 20484 + }, + { + "epoch": 0.20485, + "grad_norm": 1.0070456340780194, + "learning_rate": 0.003, + "loss": 4.0912, + "step": 20485 + }, + { + "epoch": 0.20486, + "grad_norm": 1.0706144508825242, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 20486 + }, + { + "epoch": 0.20487, + "grad_norm": 1.310107310358264, + "learning_rate": 0.003, + "loss": 4.082, + "step": 20487 + }, + { + "epoch": 0.20488, + "grad_norm": 1.0191068418068485, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 20488 + }, + { + "epoch": 0.20489, + "grad_norm": 1.2166226857919342, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 20489 + }, + { + "epoch": 0.2049, + "grad_norm": 0.9240886833542857, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 20490 + }, + { + "epoch": 0.20491, + "grad_norm": 1.0488970291252975, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 20491 + }, + { + "epoch": 0.20492, + "grad_norm": 1.1635085332214876, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 20492 + }, + { + "epoch": 0.20493, + "grad_norm": 1.078578800500168, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 20493 + }, + { + "epoch": 0.20494, + "grad_norm": 1.1807858927578427, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 20494 + }, + { + "epoch": 0.20495, + "grad_norm": 0.9976728644420545, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 20495 + }, + { + "epoch": 0.20496, + "grad_norm": 1.077379372521244, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 20496 + }, + { + "epoch": 0.20497, + "grad_norm": 1.1373834774733125, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 20497 + }, + { + "epoch": 0.20498, + "grad_norm": 0.9250759640975156, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 20498 + }, + { + "epoch": 0.20499, + "grad_norm": 0.8795581671114251, + "learning_rate": 0.003, + "loss": 4.0857, + "step": 20499 + }, + { + "epoch": 0.205, + "grad_norm": 0.938854306760688, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 20500 + }, + { + "epoch": 0.20501, + "grad_norm": 1.1833892339910494, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 20501 + }, + { + "epoch": 0.20502, + "grad_norm": 0.9187567681267947, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 20502 + }, + { + "epoch": 0.20503, + "grad_norm": 0.8932484574305813, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 20503 + }, + { + "epoch": 0.20504, + "grad_norm": 0.8634375872456692, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 20504 + }, + { + "epoch": 0.20505, + "grad_norm": 1.0111671811038372, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 20505 + }, + { + "epoch": 0.20506, + "grad_norm": 1.1266771866697345, + "learning_rate": 0.003, + "loss": 4.1212, + "step": 20506 + }, + { + "epoch": 0.20507, + "grad_norm": 0.9338390904526495, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 20507 + }, + { + "epoch": 0.20508, + "grad_norm": 1.0472147331663144, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 20508 + }, + { + "epoch": 0.20509, + "grad_norm": 1.159978132605313, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 20509 + }, + { + "epoch": 0.2051, + "grad_norm": 1.1133184075244211, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 20510 + }, + { + "epoch": 0.20511, + "grad_norm": 1.0018463461503606, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 20511 + }, + { + "epoch": 0.20512, + "grad_norm": 1.385974756181029, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 20512 + }, + { + "epoch": 0.20513, + "grad_norm": 0.8479723382333061, + "learning_rate": 0.003, + "loss": 4.077, + "step": 20513 + }, + { + "epoch": 0.20514, + "grad_norm": 0.7814281083203123, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 20514 + }, + { + "epoch": 0.20515, + "grad_norm": 0.9755607026462095, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 20515 + }, + { + "epoch": 0.20516, + "grad_norm": 1.1538315748223849, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 20516 + }, + { + "epoch": 0.20517, + "grad_norm": 1.0920045331763581, + "learning_rate": 0.003, + "loss": 4.072, + "step": 20517 + }, + { + "epoch": 0.20518, + "grad_norm": 1.1786030228325326, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 20518 + }, + { + "epoch": 0.20519, + "grad_norm": 1.158214121220045, + "learning_rate": 0.003, + "loss": 4.069, + "step": 20519 + }, + { + "epoch": 0.2052, + "grad_norm": 1.0084195178265112, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 20520 + }, + { + "epoch": 0.20521, + "grad_norm": 0.9696390483411483, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 20521 + }, + { + "epoch": 0.20522, + "grad_norm": 1.0175816054491447, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 20522 + }, + { + "epoch": 0.20523, + "grad_norm": 1.1965245103523052, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 20523 + }, + { + "epoch": 0.20524, + "grad_norm": 0.9823113910962292, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 20524 + }, + { + "epoch": 0.20525, + "grad_norm": 1.1570339468537238, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 20525 + }, + { + "epoch": 0.20526, + "grad_norm": 1.1820165969208218, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 20526 + }, + { + "epoch": 0.20527, + "grad_norm": 0.979501774035726, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 20527 + }, + { + "epoch": 0.20528, + "grad_norm": 1.1421075654091617, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 20528 + }, + { + "epoch": 0.20529, + "grad_norm": 1.044496215403326, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 20529 + }, + { + "epoch": 0.2053, + "grad_norm": 1.0835765610165682, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 20530 + }, + { + "epoch": 0.20531, + "grad_norm": 1.0257150180954069, + "learning_rate": 0.003, + "loss": 4.076, + "step": 20531 + }, + { + "epoch": 0.20532, + "grad_norm": 1.122645249150263, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 20532 + }, + { + "epoch": 0.20533, + "grad_norm": 1.0506798410819864, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 20533 + }, + { + "epoch": 0.20534, + "grad_norm": 1.1887307672162466, + "learning_rate": 0.003, + "loss": 4.072, + "step": 20534 + }, + { + "epoch": 0.20535, + "grad_norm": 1.0146953201105537, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 20535 + }, + { + "epoch": 0.20536, + "grad_norm": 1.1708270860848262, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 20536 + }, + { + "epoch": 0.20537, + "grad_norm": 1.1173222444374487, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 20537 + }, + { + "epoch": 0.20538, + "grad_norm": 1.127657749272563, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 20538 + }, + { + "epoch": 0.20539, + "grad_norm": 1.3087378398621563, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 20539 + }, + { + "epoch": 0.2054, + "grad_norm": 0.9801206272562986, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 20540 + }, + { + "epoch": 0.20541, + "grad_norm": 0.9473758280119332, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 20541 + }, + { + "epoch": 0.20542, + "grad_norm": 0.9105488079927226, + "learning_rate": 0.003, + "loss": 4.111, + "step": 20542 + }, + { + "epoch": 0.20543, + "grad_norm": 0.9237172499239538, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 20543 + }, + { + "epoch": 0.20544, + "grad_norm": 1.1665833067101368, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 20544 + }, + { + "epoch": 0.20545, + "grad_norm": 1.1482379294568845, + "learning_rate": 0.003, + "loss": 4.052, + "step": 20545 + }, + { + "epoch": 0.20546, + "grad_norm": 1.0203929558119926, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 20546 + }, + { + "epoch": 0.20547, + "grad_norm": 1.173802260005915, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 20547 + }, + { + "epoch": 0.20548, + "grad_norm": 1.1013117267706523, + "learning_rate": 0.003, + "loss": 4.0933, + "step": 20548 + }, + { + "epoch": 0.20549, + "grad_norm": 0.9850835789035832, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 20549 + }, + { + "epoch": 0.2055, + "grad_norm": 1.0226482108719344, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 20550 + }, + { + "epoch": 0.20551, + "grad_norm": 1.078975474706088, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 20551 + }, + { + "epoch": 0.20552, + "grad_norm": 1.0976958870534375, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 20552 + }, + { + "epoch": 0.20553, + "grad_norm": 1.178753857235989, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 20553 + }, + { + "epoch": 0.20554, + "grad_norm": 1.2004298546084728, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 20554 + }, + { + "epoch": 0.20555, + "grad_norm": 1.0616842031611884, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 20555 + }, + { + "epoch": 0.20556, + "grad_norm": 0.9403225573386969, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 20556 + }, + { + "epoch": 0.20557, + "grad_norm": 1.0075883985057532, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 20557 + }, + { + "epoch": 0.20558, + "grad_norm": 1.1680775669364434, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 20558 + }, + { + "epoch": 0.20559, + "grad_norm": 0.9425945597891857, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 20559 + }, + { + "epoch": 0.2056, + "grad_norm": 1.0196760982515571, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 20560 + }, + { + "epoch": 0.20561, + "grad_norm": 1.0975233103898299, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 20561 + }, + { + "epoch": 0.20562, + "grad_norm": 1.0312204191387822, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 20562 + }, + { + "epoch": 0.20563, + "grad_norm": 0.9185752254300784, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 20563 + }, + { + "epoch": 0.20564, + "grad_norm": 0.8625221268652373, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 20564 + }, + { + "epoch": 0.20565, + "grad_norm": 0.9363844632112057, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 20565 + }, + { + "epoch": 0.20566, + "grad_norm": 1.118573777063897, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 20566 + }, + { + "epoch": 0.20567, + "grad_norm": 1.1143071482940945, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 20567 + }, + { + "epoch": 0.20568, + "grad_norm": 0.855376954462423, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 20568 + }, + { + "epoch": 0.20569, + "grad_norm": 0.8206618826367794, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 20569 + }, + { + "epoch": 0.2057, + "grad_norm": 0.9771511188066676, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 20570 + }, + { + "epoch": 0.20571, + "grad_norm": 1.5738936853354584, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 20571 + }, + { + "epoch": 0.20572, + "grad_norm": 0.6924679432304148, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 20572 + }, + { + "epoch": 0.20573, + "grad_norm": 0.833465177967506, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 20573 + }, + { + "epoch": 0.20574, + "grad_norm": 0.9574485185578163, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 20574 + }, + { + "epoch": 0.20575, + "grad_norm": 1.200948415551815, + "learning_rate": 0.003, + "loss": 4.0871, + "step": 20575 + }, + { + "epoch": 0.20576, + "grad_norm": 1.0399926072367027, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 20576 + }, + { + "epoch": 0.20577, + "grad_norm": 1.1511694950256275, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 20577 + }, + { + "epoch": 0.20578, + "grad_norm": 0.994317677026091, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 20578 + }, + { + "epoch": 0.20579, + "grad_norm": 1.2433558399240716, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 20579 + }, + { + "epoch": 0.2058, + "grad_norm": 1.2138290627518498, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 20580 + }, + { + "epoch": 0.20581, + "grad_norm": 1.0665179267353084, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 20581 + }, + { + "epoch": 0.20582, + "grad_norm": 1.6454898432107934, + "learning_rate": 0.003, + "loss": 4.0845, + "step": 20582 + }, + { + "epoch": 0.20583, + "grad_norm": 1.0566868709855683, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 20583 + }, + { + "epoch": 0.20584, + "grad_norm": 1.159021256726994, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 20584 + }, + { + "epoch": 0.20585, + "grad_norm": 1.0221186153791622, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 20585 + }, + { + "epoch": 0.20586, + "grad_norm": 1.1423009205870016, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 20586 + }, + { + "epoch": 0.20587, + "grad_norm": 1.014975572214994, + "learning_rate": 0.003, + "loss": 4.079, + "step": 20587 + }, + { + "epoch": 0.20588, + "grad_norm": 1.016854182644124, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 20588 + }, + { + "epoch": 0.20589, + "grad_norm": 0.9772283172566587, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 20589 + }, + { + "epoch": 0.2059, + "grad_norm": 0.9954541701887865, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 20590 + }, + { + "epoch": 0.20591, + "grad_norm": 1.2468510195236293, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 20591 + }, + { + "epoch": 0.20592, + "grad_norm": 1.0176109600484098, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 20592 + }, + { + "epoch": 0.20593, + "grad_norm": 1.2474828112357066, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 20593 + }, + { + "epoch": 0.20594, + "grad_norm": 1.1022873894998697, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 20594 + }, + { + "epoch": 0.20595, + "grad_norm": 1.0495822473442384, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 20595 + }, + { + "epoch": 0.20596, + "grad_norm": 1.2086387053120595, + "learning_rate": 0.003, + "loss": 4.098, + "step": 20596 + }, + { + "epoch": 0.20597, + "grad_norm": 0.9863873287479966, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 20597 + }, + { + "epoch": 0.20598, + "grad_norm": 0.9924504079242689, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 20598 + }, + { + "epoch": 0.20599, + "grad_norm": 0.9161886240131877, + "learning_rate": 0.003, + "loss": 4.1153, + "step": 20599 + }, + { + "epoch": 0.206, + "grad_norm": 0.8586873024241293, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 20600 + }, + { + "epoch": 0.20601, + "grad_norm": 0.8627505993482487, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 20601 + }, + { + "epoch": 0.20602, + "grad_norm": 1.0357442526393967, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 20602 + }, + { + "epoch": 0.20603, + "grad_norm": 1.3094103399660142, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 20603 + }, + { + "epoch": 0.20604, + "grad_norm": 0.9129292905503593, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 20604 + }, + { + "epoch": 0.20605, + "grad_norm": 0.9813090338979268, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 20605 + }, + { + "epoch": 0.20606, + "grad_norm": 1.3331248786348315, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 20606 + }, + { + "epoch": 0.20607, + "grad_norm": 1.077540093500715, + "learning_rate": 0.003, + "loss": 4.0941, + "step": 20607 + }, + { + "epoch": 0.20608, + "grad_norm": 0.9434550628488495, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 20608 + }, + { + "epoch": 0.20609, + "grad_norm": 1.0948536051717321, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 20609 + }, + { + "epoch": 0.2061, + "grad_norm": 1.2331287352699838, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 20610 + }, + { + "epoch": 0.20611, + "grad_norm": 0.8367121558464378, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 20611 + }, + { + "epoch": 0.20612, + "grad_norm": 0.8478281522570646, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 20612 + }, + { + "epoch": 0.20613, + "grad_norm": 0.9157915601409786, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 20613 + }, + { + "epoch": 0.20614, + "grad_norm": 1.0394839142542278, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 20614 + }, + { + "epoch": 0.20615, + "grad_norm": 1.0125879470555106, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 20615 + }, + { + "epoch": 0.20616, + "grad_norm": 1.038583305476503, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 20616 + }, + { + "epoch": 0.20617, + "grad_norm": 0.9706200728177345, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 20617 + }, + { + "epoch": 0.20618, + "grad_norm": 1.3481259287099192, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 20618 + }, + { + "epoch": 0.20619, + "grad_norm": 0.9168913209444879, + "learning_rate": 0.003, + "loss": 4.0957, + "step": 20619 + }, + { + "epoch": 0.2062, + "grad_norm": 0.9161192410759126, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 20620 + }, + { + "epoch": 0.20621, + "grad_norm": 1.0237413338117487, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 20621 + }, + { + "epoch": 0.20622, + "grad_norm": 1.2462714820180236, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 20622 + }, + { + "epoch": 0.20623, + "grad_norm": 0.989067452349252, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 20623 + }, + { + "epoch": 0.20624, + "grad_norm": 1.093665805075143, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 20624 + }, + { + "epoch": 0.20625, + "grad_norm": 1.1493565929314613, + "learning_rate": 0.003, + "loss": 4.047, + "step": 20625 + }, + { + "epoch": 0.20626, + "grad_norm": 0.9726679431842697, + "learning_rate": 0.003, + "loss": 4.0898, + "step": 20626 + }, + { + "epoch": 0.20627, + "grad_norm": 0.9851272383993226, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 20627 + }, + { + "epoch": 0.20628, + "grad_norm": 1.0346740754679782, + "learning_rate": 0.003, + "loss": 4.0898, + "step": 20628 + }, + { + "epoch": 0.20629, + "grad_norm": 1.0943162653310972, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 20629 + }, + { + "epoch": 0.2063, + "grad_norm": 1.122841102405217, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 20630 + }, + { + "epoch": 0.20631, + "grad_norm": 0.8851768262182617, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 20631 + }, + { + "epoch": 0.20632, + "grad_norm": 0.9347815416170635, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 20632 + }, + { + "epoch": 0.20633, + "grad_norm": 1.0959302409267342, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 20633 + }, + { + "epoch": 0.20634, + "grad_norm": 0.9762459656516308, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 20634 + }, + { + "epoch": 0.20635, + "grad_norm": 0.964939806092724, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 20635 + }, + { + "epoch": 0.20636, + "grad_norm": 1.1006657529571797, + "learning_rate": 0.003, + "loss": 4.092, + "step": 20636 + }, + { + "epoch": 0.20637, + "grad_norm": 1.0473415986910473, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 20637 + }, + { + "epoch": 0.20638, + "grad_norm": 1.0853163915728636, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 20638 + }, + { + "epoch": 0.20639, + "grad_norm": 1.0476139959686217, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 20639 + }, + { + "epoch": 0.2064, + "grad_norm": 1.111698324476178, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 20640 + }, + { + "epoch": 0.20641, + "grad_norm": 1.1202457082207664, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 20641 + }, + { + "epoch": 0.20642, + "grad_norm": 0.9546565147380373, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 20642 + }, + { + "epoch": 0.20643, + "grad_norm": 0.965618637187201, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 20643 + }, + { + "epoch": 0.20644, + "grad_norm": 1.1462173477306608, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 20644 + }, + { + "epoch": 0.20645, + "grad_norm": 1.076834435107136, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 20645 + }, + { + "epoch": 0.20646, + "grad_norm": 1.121135484433155, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 20646 + }, + { + "epoch": 0.20647, + "grad_norm": 1.1958467206651566, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 20647 + }, + { + "epoch": 0.20648, + "grad_norm": 0.9353687925019336, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 20648 + }, + { + "epoch": 0.20649, + "grad_norm": 1.027072360189581, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 20649 + }, + { + "epoch": 0.2065, + "grad_norm": 1.0200070711660998, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 20650 + }, + { + "epoch": 0.20651, + "grad_norm": 1.2144182774343661, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 20651 + }, + { + "epoch": 0.20652, + "grad_norm": 0.9490302465644939, + "learning_rate": 0.003, + "loss": 4.005, + "step": 20652 + }, + { + "epoch": 0.20653, + "grad_norm": 1.0877165576466268, + "learning_rate": 0.003, + "loss": 4.062, + "step": 20653 + }, + { + "epoch": 0.20654, + "grad_norm": 1.0024782760329543, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 20654 + }, + { + "epoch": 0.20655, + "grad_norm": 1.1626383714977664, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 20655 + }, + { + "epoch": 0.20656, + "grad_norm": 0.9609125415237895, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 20656 + }, + { + "epoch": 0.20657, + "grad_norm": 0.9929042131090778, + "learning_rate": 0.003, + "loss": 4.0948, + "step": 20657 + }, + { + "epoch": 0.20658, + "grad_norm": 1.0332100656146044, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 20658 + }, + { + "epoch": 0.20659, + "grad_norm": 1.0156274946260577, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 20659 + }, + { + "epoch": 0.2066, + "grad_norm": 1.5352321934271573, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 20660 + }, + { + "epoch": 0.20661, + "grad_norm": 1.0097316481818481, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 20661 + }, + { + "epoch": 0.20662, + "grad_norm": 1.1026120278199323, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 20662 + }, + { + "epoch": 0.20663, + "grad_norm": 1.0372254042932956, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 20663 + }, + { + "epoch": 0.20664, + "grad_norm": 0.9453533838702906, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 20664 + }, + { + "epoch": 0.20665, + "grad_norm": 0.9368207377747209, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 20665 + }, + { + "epoch": 0.20666, + "grad_norm": 1.1818585127878891, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 20666 + }, + { + "epoch": 0.20667, + "grad_norm": 0.9488431486843422, + "learning_rate": 0.003, + "loss": 4.082, + "step": 20667 + }, + { + "epoch": 0.20668, + "grad_norm": 1.009109108764311, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 20668 + }, + { + "epoch": 0.20669, + "grad_norm": 1.1741715342233507, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 20669 + }, + { + "epoch": 0.2067, + "grad_norm": 0.8478551222876168, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 20670 + }, + { + "epoch": 0.20671, + "grad_norm": 0.8641147130807628, + "learning_rate": 0.003, + "loss": 4.051, + "step": 20671 + }, + { + "epoch": 0.20672, + "grad_norm": 0.8431096639348223, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 20672 + }, + { + "epoch": 0.20673, + "grad_norm": 0.9375291182058684, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 20673 + }, + { + "epoch": 0.20674, + "grad_norm": 1.1727708824704923, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 20674 + }, + { + "epoch": 0.20675, + "grad_norm": 1.1930098832426104, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 20675 + }, + { + "epoch": 0.20676, + "grad_norm": 0.9498229892891717, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 20676 + }, + { + "epoch": 0.20677, + "grad_norm": 1.2044396720791133, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 20677 + }, + { + "epoch": 0.20678, + "grad_norm": 1.0093217296969081, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 20678 + }, + { + "epoch": 0.20679, + "grad_norm": 1.2763190648210971, + "learning_rate": 0.003, + "loss": 4.0981, + "step": 20679 + }, + { + "epoch": 0.2068, + "grad_norm": 0.9419293259692796, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 20680 + }, + { + "epoch": 0.20681, + "grad_norm": 1.148088750102926, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 20681 + }, + { + "epoch": 0.20682, + "grad_norm": 1.1967780692636842, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 20682 + }, + { + "epoch": 0.20683, + "grad_norm": 0.8661317986700543, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 20683 + }, + { + "epoch": 0.20684, + "grad_norm": 0.786137183323681, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 20684 + }, + { + "epoch": 0.20685, + "grad_norm": 0.7921826152982543, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 20685 + }, + { + "epoch": 0.20686, + "grad_norm": 0.8449272858952941, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 20686 + }, + { + "epoch": 0.20687, + "grad_norm": 1.0780879717133192, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 20687 + }, + { + "epoch": 0.20688, + "grad_norm": 1.3225068076916102, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 20688 + }, + { + "epoch": 0.20689, + "grad_norm": 0.6686893915650635, + "learning_rate": 0.003, + "loss": 4.04, + "step": 20689 + }, + { + "epoch": 0.2069, + "grad_norm": 0.8228045833792297, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 20690 + }, + { + "epoch": 0.20691, + "grad_norm": 1.1225723450252534, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 20691 + }, + { + "epoch": 0.20692, + "grad_norm": 1.1388513869925954, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 20692 + }, + { + "epoch": 0.20693, + "grad_norm": 0.9541572057046324, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 20693 + }, + { + "epoch": 0.20694, + "grad_norm": 0.8624706797331997, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 20694 + }, + { + "epoch": 0.20695, + "grad_norm": 1.085398483244086, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 20695 + }, + { + "epoch": 0.20696, + "grad_norm": 1.2369892359086265, + "learning_rate": 0.003, + "loss": 4.06, + "step": 20696 + }, + { + "epoch": 0.20697, + "grad_norm": 0.9901844683249218, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 20697 + }, + { + "epoch": 0.20698, + "grad_norm": 1.2519813379810645, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 20698 + }, + { + "epoch": 0.20699, + "grad_norm": 1.0294651396532886, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 20699 + }, + { + "epoch": 0.207, + "grad_norm": 1.1043487154032898, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 20700 + }, + { + "epoch": 0.20701, + "grad_norm": 0.958147523835993, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 20701 + }, + { + "epoch": 0.20702, + "grad_norm": 0.8856346895248662, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 20702 + }, + { + "epoch": 0.20703, + "grad_norm": 0.9932155536991769, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 20703 + }, + { + "epoch": 0.20704, + "grad_norm": 1.0605053662792405, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 20704 + }, + { + "epoch": 0.20705, + "grad_norm": 0.9558034687116492, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 20705 + }, + { + "epoch": 0.20706, + "grad_norm": 0.9451653045251248, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 20706 + }, + { + "epoch": 0.20707, + "grad_norm": 0.9412584927673379, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 20707 + }, + { + "epoch": 0.20708, + "grad_norm": 1.2434307018151256, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 20708 + }, + { + "epoch": 0.20709, + "grad_norm": 1.1553415197667454, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 20709 + }, + { + "epoch": 0.2071, + "grad_norm": 0.9274753519464971, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 20710 + }, + { + "epoch": 0.20711, + "grad_norm": 0.9507935287664252, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 20711 + }, + { + "epoch": 0.20712, + "grad_norm": 1.1750773137703157, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 20712 + }, + { + "epoch": 0.20713, + "grad_norm": 1.225751144340155, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 20713 + }, + { + "epoch": 0.20714, + "grad_norm": 0.9471821496864262, + "learning_rate": 0.003, + "loss": 4.051, + "step": 20714 + }, + { + "epoch": 0.20715, + "grad_norm": 1.1348108064665425, + "learning_rate": 0.003, + "loss": 4.054, + "step": 20715 + }, + { + "epoch": 0.20716, + "grad_norm": 0.9728747556350429, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 20716 + }, + { + "epoch": 0.20717, + "grad_norm": 0.9364931706019148, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 20717 + }, + { + "epoch": 0.20718, + "grad_norm": 0.8975771117520668, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 20718 + }, + { + "epoch": 0.20719, + "grad_norm": 1.0118958757854923, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 20719 + }, + { + "epoch": 0.2072, + "grad_norm": 1.1452094594846352, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 20720 + }, + { + "epoch": 0.20721, + "grad_norm": 0.9917884396600504, + "learning_rate": 0.003, + "loss": 4.0949, + "step": 20721 + }, + { + "epoch": 0.20722, + "grad_norm": 1.248966684827715, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 20722 + }, + { + "epoch": 0.20723, + "grad_norm": 0.8814664786073738, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 20723 + }, + { + "epoch": 0.20724, + "grad_norm": 0.9877947145219791, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 20724 + }, + { + "epoch": 0.20725, + "grad_norm": 1.0537724239833088, + "learning_rate": 0.003, + "loss": 4.0946, + "step": 20725 + }, + { + "epoch": 0.20726, + "grad_norm": 1.2489465386400784, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 20726 + }, + { + "epoch": 0.20727, + "grad_norm": 1.1600892153042306, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 20727 + }, + { + "epoch": 0.20728, + "grad_norm": 1.085491183264916, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 20728 + }, + { + "epoch": 0.20729, + "grad_norm": 1.278057063116854, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 20729 + }, + { + "epoch": 0.2073, + "grad_norm": 1.0869900775420653, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 20730 + }, + { + "epoch": 0.20731, + "grad_norm": 1.044631897933949, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 20731 + }, + { + "epoch": 0.20732, + "grad_norm": 1.078650313956936, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 20732 + }, + { + "epoch": 0.20733, + "grad_norm": 1.1065531411971083, + "learning_rate": 0.003, + "loss": 4.048, + "step": 20733 + }, + { + "epoch": 0.20734, + "grad_norm": 0.9465542233547177, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 20734 + }, + { + "epoch": 0.20735, + "grad_norm": 1.175771317991426, + "learning_rate": 0.003, + "loss": 4.0932, + "step": 20735 + }, + { + "epoch": 0.20736, + "grad_norm": 1.1066814121297668, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 20736 + }, + { + "epoch": 0.20737, + "grad_norm": 0.9577523637977857, + "learning_rate": 0.003, + "loss": 4.06, + "step": 20737 + }, + { + "epoch": 0.20738, + "grad_norm": 1.0805101494292937, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 20738 + }, + { + "epoch": 0.20739, + "grad_norm": 1.2086256098599457, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 20739 + }, + { + "epoch": 0.2074, + "grad_norm": 1.0473277908235488, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 20740 + }, + { + "epoch": 0.20741, + "grad_norm": 1.3309295659587799, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 20741 + }, + { + "epoch": 0.20742, + "grad_norm": 0.8698572443332067, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 20742 + }, + { + "epoch": 0.20743, + "grad_norm": 0.8038604823843878, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 20743 + }, + { + "epoch": 0.20744, + "grad_norm": 0.8096737733303885, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 20744 + }, + { + "epoch": 0.20745, + "grad_norm": 0.9623035086154108, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 20745 + }, + { + "epoch": 0.20746, + "grad_norm": 1.0984014521192602, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 20746 + }, + { + "epoch": 0.20747, + "grad_norm": 0.91321220460166, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 20747 + }, + { + "epoch": 0.20748, + "grad_norm": 1.0599779309119177, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 20748 + }, + { + "epoch": 0.20749, + "grad_norm": 1.0861518058072823, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 20749 + }, + { + "epoch": 0.2075, + "grad_norm": 1.241538892132348, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 20750 + }, + { + "epoch": 0.20751, + "grad_norm": 1.0259868636356215, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 20751 + }, + { + "epoch": 0.20752, + "grad_norm": 0.9951206205002214, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 20752 + }, + { + "epoch": 0.20753, + "grad_norm": 1.0615104553874197, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 20753 + }, + { + "epoch": 0.20754, + "grad_norm": 1.0328772558612058, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 20754 + }, + { + "epoch": 0.20755, + "grad_norm": 1.0736136848733902, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 20755 + }, + { + "epoch": 0.20756, + "grad_norm": 1.0118881183831738, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 20756 + }, + { + "epoch": 0.20757, + "grad_norm": 0.9546087893605105, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 20757 + }, + { + "epoch": 0.20758, + "grad_norm": 1.1116364210739331, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 20758 + }, + { + "epoch": 0.20759, + "grad_norm": 1.0373788784801725, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 20759 + }, + { + "epoch": 0.2076, + "grad_norm": 1.0597017973638307, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 20760 + }, + { + "epoch": 0.20761, + "grad_norm": 0.9811530008506719, + "learning_rate": 0.003, + "loss": 4.051, + "step": 20761 + }, + { + "epoch": 0.20762, + "grad_norm": 0.9851499420930591, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 20762 + }, + { + "epoch": 0.20763, + "grad_norm": 1.1645407623785977, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 20763 + }, + { + "epoch": 0.20764, + "grad_norm": 1.1811638854146216, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 20764 + }, + { + "epoch": 0.20765, + "grad_norm": 1.1441888552978754, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 20765 + }, + { + "epoch": 0.20766, + "grad_norm": 1.064862184928858, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 20766 + }, + { + "epoch": 0.20767, + "grad_norm": 1.1151839314768486, + "learning_rate": 0.003, + "loss": 4.078, + "step": 20767 + }, + { + "epoch": 0.20768, + "grad_norm": 1.0160544876735944, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 20768 + }, + { + "epoch": 0.20769, + "grad_norm": 1.0486543444393077, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 20769 + }, + { + "epoch": 0.2077, + "grad_norm": 0.9533763350036811, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 20770 + }, + { + "epoch": 0.20771, + "grad_norm": 0.8936365499655853, + "learning_rate": 0.003, + "loss": 4.015, + "step": 20771 + }, + { + "epoch": 0.20772, + "grad_norm": 0.9042372363273592, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 20772 + }, + { + "epoch": 0.20773, + "grad_norm": 1.1655327154158655, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 20773 + }, + { + "epoch": 0.20774, + "grad_norm": 1.0757014259415456, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 20774 + }, + { + "epoch": 0.20775, + "grad_norm": 1.0125211249412955, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 20775 + }, + { + "epoch": 0.20776, + "grad_norm": 1.2410754800567012, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 20776 + }, + { + "epoch": 0.20777, + "grad_norm": 0.8479148161991825, + "learning_rate": 0.003, + "loss": 4.083, + "step": 20777 + }, + { + "epoch": 0.20778, + "grad_norm": 0.840499280288675, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 20778 + }, + { + "epoch": 0.20779, + "grad_norm": 0.9164997842397298, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 20779 + }, + { + "epoch": 0.2078, + "grad_norm": 1.1060649156004752, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 20780 + }, + { + "epoch": 0.20781, + "grad_norm": 1.0972152341876487, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 20781 + }, + { + "epoch": 0.20782, + "grad_norm": 1.0196796719724093, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 20782 + }, + { + "epoch": 0.20783, + "grad_norm": 1.2063006911711558, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 20783 + }, + { + "epoch": 0.20784, + "grad_norm": 1.1897385460796979, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 20784 + }, + { + "epoch": 0.20785, + "grad_norm": 1.1494312145828385, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 20785 + }, + { + "epoch": 0.20786, + "grad_norm": 0.9176114029470254, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 20786 + }, + { + "epoch": 0.20787, + "grad_norm": 0.9263850494149862, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 20787 + }, + { + "epoch": 0.20788, + "grad_norm": 1.0471384349035155, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 20788 + }, + { + "epoch": 0.20789, + "grad_norm": 0.9180398657993246, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 20789 + }, + { + "epoch": 0.2079, + "grad_norm": 1.2272830823486112, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 20790 + }, + { + "epoch": 0.20791, + "grad_norm": 1.0565278974581844, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 20791 + }, + { + "epoch": 0.20792, + "grad_norm": 0.9332198202052218, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 20792 + }, + { + "epoch": 0.20793, + "grad_norm": 1.0268292367877714, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 20793 + }, + { + "epoch": 0.20794, + "grad_norm": 1.1865354058971989, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 20794 + }, + { + "epoch": 0.20795, + "grad_norm": 1.0062082140222934, + "learning_rate": 0.003, + "loss": 4.05, + "step": 20795 + }, + { + "epoch": 0.20796, + "grad_norm": 1.164337336695232, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 20796 + }, + { + "epoch": 0.20797, + "grad_norm": 1.0004854387046063, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 20797 + }, + { + "epoch": 0.20798, + "grad_norm": 1.0308668167046988, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 20798 + }, + { + "epoch": 0.20799, + "grad_norm": 1.1312967366839155, + "learning_rate": 0.003, + "loss": 4.062, + "step": 20799 + }, + { + "epoch": 0.208, + "grad_norm": 1.063253905185582, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 20800 + }, + { + "epoch": 0.20801, + "grad_norm": 1.3274768471569989, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 20801 + }, + { + "epoch": 0.20802, + "grad_norm": 0.84896722609775, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 20802 + }, + { + "epoch": 0.20803, + "grad_norm": 0.9523331928486481, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 20803 + }, + { + "epoch": 0.20804, + "grad_norm": 1.1635007974893898, + "learning_rate": 0.003, + "loss": 4.0981, + "step": 20804 + }, + { + "epoch": 0.20805, + "grad_norm": 1.0045896265557979, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 20805 + }, + { + "epoch": 0.20806, + "grad_norm": 1.1757151402281818, + "learning_rate": 0.003, + "loss": 4.094, + "step": 20806 + }, + { + "epoch": 0.20807, + "grad_norm": 0.9047949491140973, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 20807 + }, + { + "epoch": 0.20808, + "grad_norm": 1.099910903883058, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 20808 + }, + { + "epoch": 0.20809, + "grad_norm": 1.2571554320607468, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 20809 + }, + { + "epoch": 0.2081, + "grad_norm": 1.0123505796541634, + "learning_rate": 0.003, + "loss": 4.066, + "step": 20810 + }, + { + "epoch": 0.20811, + "grad_norm": 1.4518436019535483, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 20811 + }, + { + "epoch": 0.20812, + "grad_norm": 0.6794970402660218, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 20812 + }, + { + "epoch": 0.20813, + "grad_norm": 0.66884767367758, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 20813 + }, + { + "epoch": 0.20814, + "grad_norm": 0.8831443769805793, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 20814 + }, + { + "epoch": 0.20815, + "grad_norm": 1.142522189663606, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 20815 + }, + { + "epoch": 0.20816, + "grad_norm": 1.3414804412692956, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 20816 + }, + { + "epoch": 0.20817, + "grad_norm": 0.6391003800819705, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 20817 + }, + { + "epoch": 0.20818, + "grad_norm": 0.7622539602062272, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 20818 + }, + { + "epoch": 0.20819, + "grad_norm": 1.0525653831877808, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 20819 + }, + { + "epoch": 0.2082, + "grad_norm": 1.1310714357547171, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 20820 + }, + { + "epoch": 0.20821, + "grad_norm": 1.0474154047488926, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 20821 + }, + { + "epoch": 0.20822, + "grad_norm": 1.2118323189164297, + "learning_rate": 0.003, + "loss": 4.0845, + "step": 20822 + }, + { + "epoch": 0.20823, + "grad_norm": 0.9900252561573274, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 20823 + }, + { + "epoch": 0.20824, + "grad_norm": 1.1164116282537144, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 20824 + }, + { + "epoch": 0.20825, + "grad_norm": 1.042094343690858, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 20825 + }, + { + "epoch": 0.20826, + "grad_norm": 1.1296108717716264, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 20826 + }, + { + "epoch": 0.20827, + "grad_norm": 0.9329538880980894, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 20827 + }, + { + "epoch": 0.20828, + "grad_norm": 1.0250871194052538, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 20828 + }, + { + "epoch": 0.20829, + "grad_norm": 1.0997611769940838, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 20829 + }, + { + "epoch": 0.2083, + "grad_norm": 1.1023188344997072, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 20830 + }, + { + "epoch": 0.20831, + "grad_norm": 1.1640356564145946, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 20831 + }, + { + "epoch": 0.20832, + "grad_norm": 0.9559470785715548, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 20832 + }, + { + "epoch": 0.20833, + "grad_norm": 1.1265014194109597, + "learning_rate": 0.003, + "loss": 4.0936, + "step": 20833 + }, + { + "epoch": 0.20834, + "grad_norm": 1.2203891788823338, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 20834 + }, + { + "epoch": 0.20835, + "grad_norm": 0.9279618562935751, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 20835 + }, + { + "epoch": 0.20836, + "grad_norm": 1.029611737576409, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 20836 + }, + { + "epoch": 0.20837, + "grad_norm": 1.1473787261857578, + "learning_rate": 0.003, + "loss": 4.0973, + "step": 20837 + }, + { + "epoch": 0.20838, + "grad_norm": 1.214673567025561, + "learning_rate": 0.003, + "loss": 4.038, + "step": 20838 + }, + { + "epoch": 0.20839, + "grad_norm": 1.2867336847027315, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 20839 + }, + { + "epoch": 0.2084, + "grad_norm": 0.9574257434294623, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 20840 + }, + { + "epoch": 0.20841, + "grad_norm": 0.9658781970501911, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 20841 + }, + { + "epoch": 0.20842, + "grad_norm": 1.104574032974007, + "learning_rate": 0.003, + "loss": 4.0871, + "step": 20842 + }, + { + "epoch": 0.20843, + "grad_norm": 0.9896003045826756, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 20843 + }, + { + "epoch": 0.20844, + "grad_norm": 1.082960616962342, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 20844 + }, + { + "epoch": 0.20845, + "grad_norm": 1.0581241151406446, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 20845 + }, + { + "epoch": 0.20846, + "grad_norm": 1.0682886150276885, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 20846 + }, + { + "epoch": 0.20847, + "grad_norm": 1.2253053617480538, + "learning_rate": 0.003, + "loss": 4.0946, + "step": 20847 + }, + { + "epoch": 0.20848, + "grad_norm": 1.0968662353246514, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 20848 + }, + { + "epoch": 0.20849, + "grad_norm": 1.0104681033608185, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 20849 + }, + { + "epoch": 0.2085, + "grad_norm": 1.3129749620904831, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 20850 + }, + { + "epoch": 0.20851, + "grad_norm": 0.9938699027836155, + "learning_rate": 0.003, + "loss": 4.044, + "step": 20851 + }, + { + "epoch": 0.20852, + "grad_norm": 1.2852944119353713, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 20852 + }, + { + "epoch": 0.20853, + "grad_norm": 1.012492290616181, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 20853 + }, + { + "epoch": 0.20854, + "grad_norm": 1.2162078102173304, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 20854 + }, + { + "epoch": 0.20855, + "grad_norm": 0.8729367843873709, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 20855 + }, + { + "epoch": 0.20856, + "grad_norm": 0.8627826124774218, + "learning_rate": 0.003, + "loss": 4.049, + "step": 20856 + }, + { + "epoch": 0.20857, + "grad_norm": 0.9870728471001707, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 20857 + }, + { + "epoch": 0.20858, + "grad_norm": 1.3558878859115444, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 20858 + }, + { + "epoch": 0.20859, + "grad_norm": 0.9479633626347346, + "learning_rate": 0.003, + "loss": 4.0934, + "step": 20859 + }, + { + "epoch": 0.2086, + "grad_norm": 0.9911778952013034, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 20860 + }, + { + "epoch": 0.20861, + "grad_norm": 1.0516493391206059, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 20861 + }, + { + "epoch": 0.20862, + "grad_norm": 0.8885805229930269, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 20862 + }, + { + "epoch": 0.20863, + "grad_norm": 0.9497236943318167, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 20863 + }, + { + "epoch": 0.20864, + "grad_norm": 0.9049695770469504, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 20864 + }, + { + "epoch": 0.20865, + "grad_norm": 1.1160644105466235, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 20865 + }, + { + "epoch": 0.20866, + "grad_norm": 1.3205716209311305, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 20866 + }, + { + "epoch": 0.20867, + "grad_norm": 0.8307913825383603, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 20867 + }, + { + "epoch": 0.20868, + "grad_norm": 0.960789443333116, + "learning_rate": 0.003, + "loss": 4.051, + "step": 20868 + }, + { + "epoch": 0.20869, + "grad_norm": 1.2686978851139956, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 20869 + }, + { + "epoch": 0.2087, + "grad_norm": 1.0317783326928518, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 20870 + }, + { + "epoch": 0.20871, + "grad_norm": 1.0695951839064524, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 20871 + }, + { + "epoch": 0.20872, + "grad_norm": 0.952916589350123, + "learning_rate": 0.003, + "loss": 4.0944, + "step": 20872 + }, + { + "epoch": 0.20873, + "grad_norm": 0.9796525257097245, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 20873 + }, + { + "epoch": 0.20874, + "grad_norm": 1.1015945327571157, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 20874 + }, + { + "epoch": 0.20875, + "grad_norm": 1.120588474692363, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 20875 + }, + { + "epoch": 0.20876, + "grad_norm": 0.9742284329594857, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 20876 + }, + { + "epoch": 0.20877, + "grad_norm": 1.0078990963842376, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 20877 + }, + { + "epoch": 0.20878, + "grad_norm": 1.1143872145216511, + "learning_rate": 0.003, + "loss": 4.071, + "step": 20878 + }, + { + "epoch": 0.20879, + "grad_norm": 1.042321421160097, + "learning_rate": 0.003, + "loss": 4.055, + "step": 20879 + }, + { + "epoch": 0.2088, + "grad_norm": 0.9632887565611077, + "learning_rate": 0.003, + "loss": 4.06, + "step": 20880 + }, + { + "epoch": 0.20881, + "grad_norm": 0.9957379160952496, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 20881 + }, + { + "epoch": 0.20882, + "grad_norm": 1.0747968890723556, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 20882 + }, + { + "epoch": 0.20883, + "grad_norm": 1.1905672782561554, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 20883 + }, + { + "epoch": 0.20884, + "grad_norm": 0.9137229310393785, + "learning_rate": 0.003, + "loss": 4.0888, + "step": 20884 + }, + { + "epoch": 0.20885, + "grad_norm": 0.9865038279558007, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 20885 + }, + { + "epoch": 0.20886, + "grad_norm": 1.222379905735362, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 20886 + }, + { + "epoch": 0.20887, + "grad_norm": 0.9360543015756947, + "learning_rate": 0.003, + "loss": 4.099, + "step": 20887 + }, + { + "epoch": 0.20888, + "grad_norm": 1.102348159052255, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 20888 + }, + { + "epoch": 0.20889, + "grad_norm": 1.0169724376076015, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 20889 + }, + { + "epoch": 0.2089, + "grad_norm": 1.151649383182817, + "learning_rate": 0.003, + "loss": 4.036, + "step": 20890 + }, + { + "epoch": 0.20891, + "grad_norm": 1.2036909171150674, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 20891 + }, + { + "epoch": 0.20892, + "grad_norm": 1.1498314445336297, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 20892 + }, + { + "epoch": 0.20893, + "grad_norm": 0.8950217037038283, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 20893 + }, + { + "epoch": 0.20894, + "grad_norm": 0.8883674753671914, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 20894 + }, + { + "epoch": 0.20895, + "grad_norm": 0.9326690919151186, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 20895 + }, + { + "epoch": 0.20896, + "grad_norm": 0.9424985463219923, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 20896 + }, + { + "epoch": 0.20897, + "grad_norm": 0.9973338351027307, + "learning_rate": 0.003, + "loss": 4.049, + "step": 20897 + }, + { + "epoch": 0.20898, + "grad_norm": 1.0945235351385671, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 20898 + }, + { + "epoch": 0.20899, + "grad_norm": 1.0388283607174973, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 20899 + }, + { + "epoch": 0.209, + "grad_norm": 1.1240785793508847, + "learning_rate": 0.003, + "loss": 4.049, + "step": 20900 + }, + { + "epoch": 0.20901, + "grad_norm": 0.99760744927545, + "learning_rate": 0.003, + "loss": 4.0884, + "step": 20901 + }, + { + "epoch": 0.20902, + "grad_norm": 1.0679497501862973, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 20902 + }, + { + "epoch": 0.20903, + "grad_norm": 1.1797052259468936, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 20903 + }, + { + "epoch": 0.20904, + "grad_norm": 0.9958329905279298, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 20904 + }, + { + "epoch": 0.20905, + "grad_norm": 1.1883633041748016, + "learning_rate": 0.003, + "loss": 4.1014, + "step": 20905 + }, + { + "epoch": 0.20906, + "grad_norm": 1.0221716175876139, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 20906 + }, + { + "epoch": 0.20907, + "grad_norm": 1.1166307140703549, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 20907 + }, + { + "epoch": 0.20908, + "grad_norm": 1.070409649502791, + "learning_rate": 0.003, + "loss": 4.039, + "step": 20908 + }, + { + "epoch": 0.20909, + "grad_norm": 1.0915998573922376, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 20909 + }, + { + "epoch": 0.2091, + "grad_norm": 1.0071018919022448, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 20910 + }, + { + "epoch": 0.20911, + "grad_norm": 1.2656152568059242, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 20911 + }, + { + "epoch": 0.20912, + "grad_norm": 0.9943635004170801, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 20912 + }, + { + "epoch": 0.20913, + "grad_norm": 1.056609693542207, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 20913 + }, + { + "epoch": 0.20914, + "grad_norm": 1.0786964667407428, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 20914 + }, + { + "epoch": 0.20915, + "grad_norm": 0.8314956791387484, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 20915 + }, + { + "epoch": 0.20916, + "grad_norm": 0.9685247110905923, + "learning_rate": 0.003, + "loss": 4.0893, + "step": 20916 + }, + { + "epoch": 0.20917, + "grad_norm": 1.2118654335862826, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 20917 + }, + { + "epoch": 0.20918, + "grad_norm": 1.0312127272956535, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 20918 + }, + { + "epoch": 0.20919, + "grad_norm": 1.2615255078675625, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 20919 + }, + { + "epoch": 0.2092, + "grad_norm": 0.9964236114538393, + "learning_rate": 0.003, + "loss": 4.07, + "step": 20920 + }, + { + "epoch": 0.20921, + "grad_norm": 1.292628435269601, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 20921 + }, + { + "epoch": 0.20922, + "grad_norm": 0.8983527082940497, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 20922 + }, + { + "epoch": 0.20923, + "grad_norm": 1.002776971500041, + "learning_rate": 0.003, + "loss": 4.0878, + "step": 20923 + }, + { + "epoch": 0.20924, + "grad_norm": 0.9518499213608802, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 20924 + }, + { + "epoch": 0.20925, + "grad_norm": 1.0010550816157102, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 20925 + }, + { + "epoch": 0.20926, + "grad_norm": 1.3326726907949216, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 20926 + }, + { + "epoch": 0.20927, + "grad_norm": 0.8327472894994723, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 20927 + }, + { + "epoch": 0.20928, + "grad_norm": 0.8801252985982598, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 20928 + }, + { + "epoch": 0.20929, + "grad_norm": 1.1392574161474622, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 20929 + }, + { + "epoch": 0.2093, + "grad_norm": 0.9967296444395508, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 20930 + }, + { + "epoch": 0.20931, + "grad_norm": 0.9764938153869399, + "learning_rate": 0.003, + "loss": 4.076, + "step": 20931 + }, + { + "epoch": 0.20932, + "grad_norm": 1.2825649218177002, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 20932 + }, + { + "epoch": 0.20933, + "grad_norm": 1.0032726821715023, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 20933 + }, + { + "epoch": 0.20934, + "grad_norm": 1.102744014750965, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 20934 + }, + { + "epoch": 0.20935, + "grad_norm": 0.9377335273868062, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 20935 + }, + { + "epoch": 0.20936, + "grad_norm": 1.1034559379464997, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 20936 + }, + { + "epoch": 0.20937, + "grad_norm": 1.4791163100325675, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 20937 + }, + { + "epoch": 0.20938, + "grad_norm": 0.8689356414037108, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 20938 + }, + { + "epoch": 0.20939, + "grad_norm": 0.8427071401070674, + "learning_rate": 0.003, + "loss": 4.063, + "step": 20939 + }, + { + "epoch": 0.2094, + "grad_norm": 0.9265815826054483, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 20940 + }, + { + "epoch": 0.20941, + "grad_norm": 1.0099437891981076, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 20941 + }, + { + "epoch": 0.20942, + "grad_norm": 1.203262229847494, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 20942 + }, + { + "epoch": 0.20943, + "grad_norm": 0.8348180876511309, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 20943 + }, + { + "epoch": 0.20944, + "grad_norm": 0.8904297413839721, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 20944 + }, + { + "epoch": 0.20945, + "grad_norm": 1.0041248531154272, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 20945 + }, + { + "epoch": 0.20946, + "grad_norm": 1.0916479263931063, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 20946 + }, + { + "epoch": 0.20947, + "grad_norm": 1.1951028674907425, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 20947 + }, + { + "epoch": 0.20948, + "grad_norm": 1.081571999505606, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 20948 + }, + { + "epoch": 0.20949, + "grad_norm": 1.199490662596358, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 20949 + }, + { + "epoch": 0.2095, + "grad_norm": 0.9305400894836376, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 20950 + }, + { + "epoch": 0.20951, + "grad_norm": 1.0618272552214656, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 20951 + }, + { + "epoch": 0.20952, + "grad_norm": 0.9608030117120973, + "learning_rate": 0.003, + "loss": 4.0941, + "step": 20952 + }, + { + "epoch": 0.20953, + "grad_norm": 1.139156789673186, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 20953 + }, + { + "epoch": 0.20954, + "grad_norm": 0.8782823478777583, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 20954 + }, + { + "epoch": 0.20955, + "grad_norm": 0.7772783185286388, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 20955 + }, + { + "epoch": 0.20956, + "grad_norm": 0.9129380619038314, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 20956 + }, + { + "epoch": 0.20957, + "grad_norm": 1.0872611180083562, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 20957 + }, + { + "epoch": 0.20958, + "grad_norm": 1.2178611699364201, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 20958 + }, + { + "epoch": 0.20959, + "grad_norm": 1.1993922633002072, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 20959 + }, + { + "epoch": 0.2096, + "grad_norm": 0.922061806343558, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 20960 + }, + { + "epoch": 0.20961, + "grad_norm": 0.9799893420060268, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 20961 + }, + { + "epoch": 0.20962, + "grad_norm": 1.2921558841566678, + "learning_rate": 0.003, + "loss": 4.069, + "step": 20962 + }, + { + "epoch": 0.20963, + "grad_norm": 1.1268863148822506, + "learning_rate": 0.003, + "loss": 4.043, + "step": 20963 + }, + { + "epoch": 0.20964, + "grad_norm": 1.1532252987516078, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 20964 + }, + { + "epoch": 0.20965, + "grad_norm": 0.9043483174045643, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 20965 + }, + { + "epoch": 0.20966, + "grad_norm": 1.0271697225499592, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 20966 + }, + { + "epoch": 0.20967, + "grad_norm": 1.213138303150783, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 20967 + }, + { + "epoch": 0.20968, + "grad_norm": 0.8610200438705731, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 20968 + }, + { + "epoch": 0.20969, + "grad_norm": 0.8288653996696551, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 20969 + }, + { + "epoch": 0.2097, + "grad_norm": 0.8666926404800426, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 20970 + }, + { + "epoch": 0.20971, + "grad_norm": 0.9515725871412093, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 20971 + }, + { + "epoch": 0.20972, + "grad_norm": 1.140846365626235, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 20972 + }, + { + "epoch": 0.20973, + "grad_norm": 0.8851341959683362, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 20973 + }, + { + "epoch": 0.20974, + "grad_norm": 0.9693628592223124, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 20974 + }, + { + "epoch": 0.20975, + "grad_norm": 1.0443438376481937, + "learning_rate": 0.003, + "loss": 4.0873, + "step": 20975 + }, + { + "epoch": 0.20976, + "grad_norm": 1.047085087588925, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 20976 + }, + { + "epoch": 0.20977, + "grad_norm": 0.9292370776347733, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 20977 + }, + { + "epoch": 0.20978, + "grad_norm": 1.0241510190596212, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 20978 + }, + { + "epoch": 0.20979, + "grad_norm": 1.398230360330995, + "learning_rate": 0.003, + "loss": 4.1076, + "step": 20979 + }, + { + "epoch": 0.2098, + "grad_norm": 0.9607854600432582, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 20980 + }, + { + "epoch": 0.20981, + "grad_norm": 1.0105909476062682, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 20981 + }, + { + "epoch": 0.20982, + "grad_norm": 1.338923953903629, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 20982 + }, + { + "epoch": 0.20983, + "grad_norm": 1.067515178619015, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 20983 + }, + { + "epoch": 0.20984, + "grad_norm": 1.0277166346833038, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 20984 + }, + { + "epoch": 0.20985, + "grad_norm": 1.3025771763802771, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 20985 + }, + { + "epoch": 0.20986, + "grad_norm": 0.9479824920826533, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 20986 + }, + { + "epoch": 0.20987, + "grad_norm": 0.9581118739627525, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 20987 + }, + { + "epoch": 0.20988, + "grad_norm": 1.4270429274421466, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 20988 + }, + { + "epoch": 0.20989, + "grad_norm": 0.8231542439549997, + "learning_rate": 0.003, + "loss": 4.045, + "step": 20989 + }, + { + "epoch": 0.2099, + "grad_norm": 0.8498482904732212, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 20990 + }, + { + "epoch": 0.20991, + "grad_norm": 1.178381890049208, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 20991 + }, + { + "epoch": 0.20992, + "grad_norm": 1.1847373521840892, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 20992 + }, + { + "epoch": 0.20993, + "grad_norm": 0.7404675534544347, + "learning_rate": 0.003, + "loss": 4.1105, + "step": 20993 + }, + { + "epoch": 0.20994, + "grad_norm": 0.7034640100530521, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 20994 + }, + { + "epoch": 0.20995, + "grad_norm": 0.7968891352310264, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 20995 + }, + { + "epoch": 0.20996, + "grad_norm": 0.8924647823177969, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 20996 + }, + { + "epoch": 0.20997, + "grad_norm": 1.027322505467095, + "learning_rate": 0.003, + "loss": 4.041, + "step": 20997 + }, + { + "epoch": 0.20998, + "grad_norm": 1.0788060235596744, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 20998 + }, + { + "epoch": 0.20999, + "grad_norm": 1.0483352365942207, + "learning_rate": 0.003, + "loss": 4.062, + "step": 20999 + }, + { + "epoch": 0.21, + "grad_norm": 1.1613640300326153, + "learning_rate": 0.003, + "loss": 4.0925, + "step": 21000 + }, + { + "epoch": 0.21001, + "grad_norm": 1.1170056977478917, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 21001 + }, + { + "epoch": 0.21002, + "grad_norm": 0.9902692015212363, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 21002 + }, + { + "epoch": 0.21003, + "grad_norm": 1.191784942713295, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 21003 + }, + { + "epoch": 0.21004, + "grad_norm": 1.0278858191273157, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 21004 + }, + { + "epoch": 0.21005, + "grad_norm": 1.1342117591967515, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 21005 + }, + { + "epoch": 0.21006, + "grad_norm": 1.008502068231316, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 21006 + }, + { + "epoch": 0.21007, + "grad_norm": 1.0775225195642366, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 21007 + }, + { + "epoch": 0.21008, + "grad_norm": 1.0931518850229958, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 21008 + }, + { + "epoch": 0.21009, + "grad_norm": 1.1360310846999258, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 21009 + }, + { + "epoch": 0.2101, + "grad_norm": 1.1189053562425344, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 21010 + }, + { + "epoch": 0.21011, + "grad_norm": 1.1844672140087518, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 21011 + }, + { + "epoch": 0.21012, + "grad_norm": 1.0237576111364075, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 21012 + }, + { + "epoch": 0.21013, + "grad_norm": 1.1618257948162096, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 21013 + }, + { + "epoch": 0.21014, + "grad_norm": 1.049426391170783, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 21014 + }, + { + "epoch": 0.21015, + "grad_norm": 1.1499152007126543, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 21015 + }, + { + "epoch": 0.21016, + "grad_norm": 0.9892802007463918, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 21016 + }, + { + "epoch": 0.21017, + "grad_norm": 1.284459078788656, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 21017 + }, + { + "epoch": 0.21018, + "grad_norm": 0.8443909703049393, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 21018 + }, + { + "epoch": 0.21019, + "grad_norm": 0.9002606413114878, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 21019 + }, + { + "epoch": 0.2102, + "grad_norm": 1.0011664691136566, + "learning_rate": 0.003, + "loss": 4.055, + "step": 21020 + }, + { + "epoch": 0.21021, + "grad_norm": 1.256452744696525, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 21021 + }, + { + "epoch": 0.21022, + "grad_norm": 1.0655016338519239, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 21022 + }, + { + "epoch": 0.21023, + "grad_norm": 1.1152538507623835, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 21023 + }, + { + "epoch": 0.21024, + "grad_norm": 0.997278305436515, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 21024 + }, + { + "epoch": 0.21025, + "grad_norm": 0.9822731235443738, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 21025 + }, + { + "epoch": 0.21026, + "grad_norm": 1.1321904943188026, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 21026 + }, + { + "epoch": 0.21027, + "grad_norm": 1.105307477088627, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 21027 + }, + { + "epoch": 0.21028, + "grad_norm": 1.1725000570313406, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 21028 + }, + { + "epoch": 0.21029, + "grad_norm": 1.0825563441124535, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 21029 + }, + { + "epoch": 0.2103, + "grad_norm": 1.056382876166678, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 21030 + }, + { + "epoch": 0.21031, + "grad_norm": 0.9972971714065579, + "learning_rate": 0.003, + "loss": 4.051, + "step": 21031 + }, + { + "epoch": 0.21032, + "grad_norm": 0.9131290759296145, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 21032 + }, + { + "epoch": 0.21033, + "grad_norm": 1.01618682643476, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 21033 + }, + { + "epoch": 0.21034, + "grad_norm": 1.1240245192044853, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 21034 + }, + { + "epoch": 0.21035, + "grad_norm": 1.2151600267975669, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 21035 + }, + { + "epoch": 0.21036, + "grad_norm": 0.9307170684564446, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 21036 + }, + { + "epoch": 0.21037, + "grad_norm": 0.9926716208371283, + "learning_rate": 0.003, + "loss": 4.055, + "step": 21037 + }, + { + "epoch": 0.21038, + "grad_norm": 1.1452056695109978, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 21038 + }, + { + "epoch": 0.21039, + "grad_norm": 1.0842388293432828, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 21039 + }, + { + "epoch": 0.2104, + "grad_norm": 0.9705041419665481, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 21040 + }, + { + "epoch": 0.21041, + "grad_norm": 1.127993431926338, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 21041 + }, + { + "epoch": 0.21042, + "grad_norm": 1.001136653289757, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 21042 + }, + { + "epoch": 0.21043, + "grad_norm": 1.3471637209211096, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 21043 + }, + { + "epoch": 0.21044, + "grad_norm": 1.1306140792421662, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 21044 + }, + { + "epoch": 0.21045, + "grad_norm": 1.055579713387114, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 21045 + }, + { + "epoch": 0.21046, + "grad_norm": 1.2275402222363736, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 21046 + }, + { + "epoch": 0.21047, + "grad_norm": 1.0999710490705354, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 21047 + }, + { + "epoch": 0.21048, + "grad_norm": 1.0231109307520014, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 21048 + }, + { + "epoch": 0.21049, + "grad_norm": 1.1379161670911586, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 21049 + }, + { + "epoch": 0.2105, + "grad_norm": 0.9266030440661334, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 21050 + }, + { + "epoch": 0.21051, + "grad_norm": 0.874309124436618, + "learning_rate": 0.003, + "loss": 4.0901, + "step": 21051 + }, + { + "epoch": 0.21052, + "grad_norm": 0.9143231323277458, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 21052 + }, + { + "epoch": 0.21053, + "grad_norm": 1.0042323211846782, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 21053 + }, + { + "epoch": 0.21054, + "grad_norm": 1.1524900800753866, + "learning_rate": 0.003, + "loss": 4.0884, + "step": 21054 + }, + { + "epoch": 0.21055, + "grad_norm": 1.1083882941643974, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 21055 + }, + { + "epoch": 0.21056, + "grad_norm": 0.9274259210602794, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 21056 + }, + { + "epoch": 0.21057, + "grad_norm": 1.1489559489106986, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 21057 + }, + { + "epoch": 0.21058, + "grad_norm": 1.2408275405306983, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 21058 + }, + { + "epoch": 0.21059, + "grad_norm": 0.8486623643715242, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 21059 + }, + { + "epoch": 0.2106, + "grad_norm": 1.0471705734007617, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 21060 + }, + { + "epoch": 0.21061, + "grad_norm": 1.348373720690715, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 21061 + }, + { + "epoch": 0.21062, + "grad_norm": 0.8049920440899453, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 21062 + }, + { + "epoch": 0.21063, + "grad_norm": 0.7563208292149448, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 21063 + }, + { + "epoch": 0.21064, + "grad_norm": 0.7280160088049493, + "learning_rate": 0.003, + "loss": 4.066, + "step": 21064 + }, + { + "epoch": 0.21065, + "grad_norm": 0.8287869092148357, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 21065 + }, + { + "epoch": 0.21066, + "grad_norm": 0.9468256966142923, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 21066 + }, + { + "epoch": 0.21067, + "grad_norm": 1.1034057761393183, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 21067 + }, + { + "epoch": 0.21068, + "grad_norm": 0.996549224299107, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 21068 + }, + { + "epoch": 0.21069, + "grad_norm": 1.1776743684597348, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 21069 + }, + { + "epoch": 0.2107, + "grad_norm": 1.0017674157748013, + "learning_rate": 0.003, + "loss": 4.065, + "step": 21070 + }, + { + "epoch": 0.21071, + "grad_norm": 1.072558149944195, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 21071 + }, + { + "epoch": 0.21072, + "grad_norm": 0.9320075751191523, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 21072 + }, + { + "epoch": 0.21073, + "grad_norm": 1.034043400376183, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 21073 + }, + { + "epoch": 0.21074, + "grad_norm": 1.2606855148913554, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 21074 + }, + { + "epoch": 0.21075, + "grad_norm": 1.0414473557787123, + "learning_rate": 0.003, + "loss": 4.075, + "step": 21075 + }, + { + "epoch": 0.21076, + "grad_norm": 1.4975124470544352, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 21076 + }, + { + "epoch": 0.21077, + "grad_norm": 0.7833852296351864, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 21077 + }, + { + "epoch": 0.21078, + "grad_norm": 0.8508238135751062, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 21078 + }, + { + "epoch": 0.21079, + "grad_norm": 1.0369104167385037, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 21079 + }, + { + "epoch": 0.2108, + "grad_norm": 1.1640235151345146, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 21080 + }, + { + "epoch": 0.21081, + "grad_norm": 1.0658886064463113, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 21081 + }, + { + "epoch": 0.21082, + "grad_norm": 1.3707504280746186, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 21082 + }, + { + "epoch": 0.21083, + "grad_norm": 0.7963079017934281, + "learning_rate": 0.003, + "loss": 4.07, + "step": 21083 + }, + { + "epoch": 0.21084, + "grad_norm": 0.8494277386438776, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 21084 + }, + { + "epoch": 0.21085, + "grad_norm": 1.1012351993429084, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 21085 + }, + { + "epoch": 0.21086, + "grad_norm": 1.5152904485714744, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 21086 + }, + { + "epoch": 0.21087, + "grad_norm": 0.8273732826748955, + "learning_rate": 0.003, + "loss": 4.072, + "step": 21087 + }, + { + "epoch": 0.21088, + "grad_norm": 0.8837087645282795, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 21088 + }, + { + "epoch": 0.21089, + "grad_norm": 1.4172643500170043, + "learning_rate": 0.003, + "loss": 4.0922, + "step": 21089 + }, + { + "epoch": 0.2109, + "grad_norm": 1.0951719973155558, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 21090 + }, + { + "epoch": 0.21091, + "grad_norm": 1.215087718837464, + "learning_rate": 0.003, + "loss": 4.044, + "step": 21091 + }, + { + "epoch": 0.21092, + "grad_norm": 1.0507656265642218, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 21092 + }, + { + "epoch": 0.21093, + "grad_norm": 1.1422060353476664, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 21093 + }, + { + "epoch": 0.21094, + "grad_norm": 0.9467077063982493, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 21094 + }, + { + "epoch": 0.21095, + "grad_norm": 1.0474445672617263, + "learning_rate": 0.003, + "loss": 4.074, + "step": 21095 + }, + { + "epoch": 0.21096, + "grad_norm": 1.0214835847862145, + "learning_rate": 0.003, + "loss": 4.048, + "step": 21096 + }, + { + "epoch": 0.21097, + "grad_norm": 1.01803712130632, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 21097 + }, + { + "epoch": 0.21098, + "grad_norm": 1.1624159370104825, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 21098 + }, + { + "epoch": 0.21099, + "grad_norm": 0.9264206442840789, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 21099 + }, + { + "epoch": 0.211, + "grad_norm": 1.0682842185800343, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 21100 + }, + { + "epoch": 0.21101, + "grad_norm": 0.9198583645221149, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 21101 + }, + { + "epoch": 0.21102, + "grad_norm": 1.0321338192935856, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 21102 + }, + { + "epoch": 0.21103, + "grad_norm": 1.1016361635912424, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 21103 + }, + { + "epoch": 0.21104, + "grad_norm": 1.0747401197448212, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 21104 + }, + { + "epoch": 0.21105, + "grad_norm": 0.9674497934465337, + "learning_rate": 0.003, + "loss": 4.08, + "step": 21105 + }, + { + "epoch": 0.21106, + "grad_norm": 1.2429970636767618, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 21106 + }, + { + "epoch": 0.21107, + "grad_norm": 1.0781659140202244, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 21107 + }, + { + "epoch": 0.21108, + "grad_norm": 1.4371258815712247, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 21108 + }, + { + "epoch": 0.21109, + "grad_norm": 1.170785255926806, + "learning_rate": 0.003, + "loss": 4.075, + "step": 21109 + }, + { + "epoch": 0.2111, + "grad_norm": 0.9414136694730864, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 21110 + }, + { + "epoch": 0.21111, + "grad_norm": 0.9867083515451793, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 21111 + }, + { + "epoch": 0.21112, + "grad_norm": 1.1334283308256854, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 21112 + }, + { + "epoch": 0.21113, + "grad_norm": 1.0107037985502143, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 21113 + }, + { + "epoch": 0.21114, + "grad_norm": 0.9742291224143317, + "learning_rate": 0.003, + "loss": 4.0854, + "step": 21114 + }, + { + "epoch": 0.21115, + "grad_norm": 1.14361417207491, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 21115 + }, + { + "epoch": 0.21116, + "grad_norm": 1.2765508814910087, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 21116 + }, + { + "epoch": 0.21117, + "grad_norm": 1.413909482716925, + "learning_rate": 0.003, + "loss": 4.067, + "step": 21117 + }, + { + "epoch": 0.21118, + "grad_norm": 0.7872668712865801, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 21118 + }, + { + "epoch": 0.21119, + "grad_norm": 0.8591173506829681, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 21119 + }, + { + "epoch": 0.2112, + "grad_norm": 0.9359465830906611, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 21120 + }, + { + "epoch": 0.21121, + "grad_norm": 1.2607249762786281, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 21121 + }, + { + "epoch": 0.21122, + "grad_norm": 0.8581067595241416, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 21122 + }, + { + "epoch": 0.21123, + "grad_norm": 0.7614011496776761, + "learning_rate": 0.003, + "loss": 4.057, + "step": 21123 + }, + { + "epoch": 0.21124, + "grad_norm": 0.9513677602788703, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 21124 + }, + { + "epoch": 0.21125, + "grad_norm": 1.4190716565433221, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 21125 + }, + { + "epoch": 0.21126, + "grad_norm": 0.8282320037977055, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 21126 + }, + { + "epoch": 0.21127, + "grad_norm": 0.7743475011895226, + "learning_rate": 0.003, + "loss": 4.072, + "step": 21127 + }, + { + "epoch": 0.21128, + "grad_norm": 0.926317892028432, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 21128 + }, + { + "epoch": 0.21129, + "grad_norm": 1.1769196368023551, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 21129 + }, + { + "epoch": 0.2113, + "grad_norm": 1.308737597315486, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 21130 + }, + { + "epoch": 0.21131, + "grad_norm": 1.0532483222023556, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 21131 + }, + { + "epoch": 0.21132, + "grad_norm": 1.1197918170488919, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 21132 + }, + { + "epoch": 0.21133, + "grad_norm": 1.0345567289242423, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 21133 + }, + { + "epoch": 0.21134, + "grad_norm": 1.1324222281388348, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 21134 + }, + { + "epoch": 0.21135, + "grad_norm": 0.9282220041228886, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 21135 + }, + { + "epoch": 0.21136, + "grad_norm": 0.9674914561444075, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 21136 + }, + { + "epoch": 0.21137, + "grad_norm": 1.2687235563379828, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 21137 + }, + { + "epoch": 0.21138, + "grad_norm": 0.8657081976237382, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 21138 + }, + { + "epoch": 0.21139, + "grad_norm": 0.9294609560823931, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 21139 + }, + { + "epoch": 0.2114, + "grad_norm": 1.0415528456132592, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 21140 + }, + { + "epoch": 0.21141, + "grad_norm": 1.1501707969877513, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 21141 + }, + { + "epoch": 0.21142, + "grad_norm": 0.9872340329683469, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 21142 + }, + { + "epoch": 0.21143, + "grad_norm": 1.2582514687207667, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 21143 + }, + { + "epoch": 0.21144, + "grad_norm": 0.7353421234580598, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 21144 + }, + { + "epoch": 0.21145, + "grad_norm": 0.7080670287406137, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 21145 + }, + { + "epoch": 0.21146, + "grad_norm": 0.8781066024870269, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 21146 + }, + { + "epoch": 0.21147, + "grad_norm": 0.9861328464682059, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 21147 + }, + { + "epoch": 0.21148, + "grad_norm": 1.0108476784716045, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 21148 + }, + { + "epoch": 0.21149, + "grad_norm": 0.9543800427374821, + "learning_rate": 0.003, + "loss": 4.0956, + "step": 21149 + }, + { + "epoch": 0.2115, + "grad_norm": 1.186819166620684, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 21150 + }, + { + "epoch": 0.21151, + "grad_norm": 1.019207007857277, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 21151 + }, + { + "epoch": 0.21152, + "grad_norm": 1.1277092052136801, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 21152 + }, + { + "epoch": 0.21153, + "grad_norm": 1.283313494757284, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 21153 + }, + { + "epoch": 0.21154, + "grad_norm": 1.1119586335157217, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 21154 + }, + { + "epoch": 0.21155, + "grad_norm": 0.8642657181956677, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 21155 + }, + { + "epoch": 0.21156, + "grad_norm": 0.9024274031243885, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 21156 + }, + { + "epoch": 0.21157, + "grad_norm": 0.8606706114729045, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 21157 + }, + { + "epoch": 0.21158, + "grad_norm": 0.8695576359471315, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 21158 + }, + { + "epoch": 0.21159, + "grad_norm": 1.131432950949515, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 21159 + }, + { + "epoch": 0.2116, + "grad_norm": 1.1062739911215254, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 21160 + }, + { + "epoch": 0.21161, + "grad_norm": 1.0665855260766952, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 21161 + }, + { + "epoch": 0.21162, + "grad_norm": 1.3457845559537047, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 21162 + }, + { + "epoch": 0.21163, + "grad_norm": 0.9916486151551507, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 21163 + }, + { + "epoch": 0.21164, + "grad_norm": 0.9817638750079845, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 21164 + }, + { + "epoch": 0.21165, + "grad_norm": 1.1402245141552292, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 21165 + }, + { + "epoch": 0.21166, + "grad_norm": 1.0786175953047288, + "learning_rate": 0.003, + "loss": 4.039, + "step": 21166 + }, + { + "epoch": 0.21167, + "grad_norm": 1.2184072912401867, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 21167 + }, + { + "epoch": 0.21168, + "grad_norm": 1.018868015249159, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 21168 + }, + { + "epoch": 0.21169, + "grad_norm": 0.9654407659403722, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 21169 + }, + { + "epoch": 0.2117, + "grad_norm": 1.0737816433866847, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 21170 + }, + { + "epoch": 0.21171, + "grad_norm": 1.0465884249291126, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 21171 + }, + { + "epoch": 0.21172, + "grad_norm": 1.0388966739631085, + "learning_rate": 0.003, + "loss": 4.079, + "step": 21172 + }, + { + "epoch": 0.21173, + "grad_norm": 1.0525191680142962, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 21173 + }, + { + "epoch": 0.21174, + "grad_norm": 1.1224102671301817, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 21174 + }, + { + "epoch": 0.21175, + "grad_norm": 1.1506454197815508, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 21175 + }, + { + "epoch": 0.21176, + "grad_norm": 1.0286531766901235, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 21176 + }, + { + "epoch": 0.21177, + "grad_norm": 1.1031336538398755, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 21177 + }, + { + "epoch": 0.21178, + "grad_norm": 1.103615048860668, + "learning_rate": 0.003, + "loss": 4.028, + "step": 21178 + }, + { + "epoch": 0.21179, + "grad_norm": 1.1784003376170336, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 21179 + }, + { + "epoch": 0.2118, + "grad_norm": 1.3064519980528813, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 21180 + }, + { + "epoch": 0.21181, + "grad_norm": 0.9257082365139948, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 21181 + }, + { + "epoch": 0.21182, + "grad_norm": 0.9709863194533105, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 21182 + }, + { + "epoch": 0.21183, + "grad_norm": 1.277651064275812, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 21183 + }, + { + "epoch": 0.21184, + "grad_norm": 1.0005916767091565, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 21184 + }, + { + "epoch": 0.21185, + "grad_norm": 1.1744765236090258, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 21185 + }, + { + "epoch": 0.21186, + "grad_norm": 1.0415434611549872, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 21186 + }, + { + "epoch": 0.21187, + "grad_norm": 1.023434740173438, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 21187 + }, + { + "epoch": 0.21188, + "grad_norm": 1.0660054597268966, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 21188 + }, + { + "epoch": 0.21189, + "grad_norm": 1.0542262256544255, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 21189 + }, + { + "epoch": 0.2119, + "grad_norm": 1.1227100858135617, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 21190 + }, + { + "epoch": 0.21191, + "grad_norm": 1.0388698321085599, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 21191 + }, + { + "epoch": 0.21192, + "grad_norm": 1.026668394460733, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 21192 + }, + { + "epoch": 0.21193, + "grad_norm": 1.0308557252111394, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 21193 + }, + { + "epoch": 0.21194, + "grad_norm": 1.2289457075132089, + "learning_rate": 0.003, + "loss": 4.0958, + "step": 21194 + }, + { + "epoch": 0.21195, + "grad_norm": 0.8351735417352819, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 21195 + }, + { + "epoch": 0.21196, + "grad_norm": 1.0419982588804575, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 21196 + }, + { + "epoch": 0.21197, + "grad_norm": 1.0865624460745271, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 21197 + }, + { + "epoch": 0.21198, + "grad_norm": 1.1634838100463858, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 21198 + }, + { + "epoch": 0.21199, + "grad_norm": 1.2707059103733171, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 21199 + }, + { + "epoch": 0.212, + "grad_norm": 1.1129507046361597, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 21200 + }, + { + "epoch": 0.21201, + "grad_norm": 1.050552859867441, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 21201 + }, + { + "epoch": 0.21202, + "grad_norm": 1.1203542396625805, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 21202 + }, + { + "epoch": 0.21203, + "grad_norm": 1.165992701411825, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 21203 + }, + { + "epoch": 0.21204, + "grad_norm": 1.1381979150787762, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 21204 + }, + { + "epoch": 0.21205, + "grad_norm": 1.0437299447903416, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 21205 + }, + { + "epoch": 0.21206, + "grad_norm": 0.8994809711819914, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 21206 + }, + { + "epoch": 0.21207, + "grad_norm": 1.0038062888266812, + "learning_rate": 0.003, + "loss": 4.0953, + "step": 21207 + }, + { + "epoch": 0.21208, + "grad_norm": 1.2222132616294945, + "learning_rate": 0.003, + "loss": 4.0929, + "step": 21208 + }, + { + "epoch": 0.21209, + "grad_norm": 0.9991496603948783, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 21209 + }, + { + "epoch": 0.2121, + "grad_norm": 1.1517568372092961, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 21210 + }, + { + "epoch": 0.21211, + "grad_norm": 0.8843049272468909, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 21211 + }, + { + "epoch": 0.21212, + "grad_norm": 0.9203772081322751, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 21212 + }, + { + "epoch": 0.21213, + "grad_norm": 1.0413206849623011, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 21213 + }, + { + "epoch": 0.21214, + "grad_norm": 1.099332279688137, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 21214 + }, + { + "epoch": 0.21215, + "grad_norm": 1.0413957953024968, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 21215 + }, + { + "epoch": 0.21216, + "grad_norm": 1.2138781057201848, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 21216 + }, + { + "epoch": 0.21217, + "grad_norm": 1.0301348425777108, + "learning_rate": 0.003, + "loss": 4.038, + "step": 21217 + }, + { + "epoch": 0.21218, + "grad_norm": 1.1814174073169428, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 21218 + }, + { + "epoch": 0.21219, + "grad_norm": 1.0352412328683613, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 21219 + }, + { + "epoch": 0.2122, + "grad_norm": 1.132434823616984, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 21220 + }, + { + "epoch": 0.21221, + "grad_norm": 1.0523503147964524, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 21221 + }, + { + "epoch": 0.21222, + "grad_norm": 1.2317523630586569, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 21222 + }, + { + "epoch": 0.21223, + "grad_norm": 1.0881509659213509, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 21223 + }, + { + "epoch": 0.21224, + "grad_norm": 0.9658866950861134, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 21224 + }, + { + "epoch": 0.21225, + "grad_norm": 0.8779914655710239, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 21225 + }, + { + "epoch": 0.21226, + "grad_norm": 0.9376666135639579, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 21226 + }, + { + "epoch": 0.21227, + "grad_norm": 1.0409608232736678, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 21227 + }, + { + "epoch": 0.21228, + "grad_norm": 1.4252209016108264, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 21228 + }, + { + "epoch": 0.21229, + "grad_norm": 0.9696898705982024, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 21229 + }, + { + "epoch": 0.2123, + "grad_norm": 1.238001309370028, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 21230 + }, + { + "epoch": 0.21231, + "grad_norm": 0.8632046406159173, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 21231 + }, + { + "epoch": 0.21232, + "grad_norm": 0.8453743941040195, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 21232 + }, + { + "epoch": 0.21233, + "grad_norm": 0.980621385896423, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 21233 + }, + { + "epoch": 0.21234, + "grad_norm": 1.0532937429684637, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 21234 + }, + { + "epoch": 0.21235, + "grad_norm": 1.1084266812303676, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 21235 + }, + { + "epoch": 0.21236, + "grad_norm": 1.2757058555295784, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 21236 + }, + { + "epoch": 0.21237, + "grad_norm": 1.0512956839573728, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 21237 + }, + { + "epoch": 0.21238, + "grad_norm": 1.3708828198449563, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 21238 + }, + { + "epoch": 0.21239, + "grad_norm": 0.8695917620397281, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 21239 + }, + { + "epoch": 0.2124, + "grad_norm": 0.9164838288244584, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 21240 + }, + { + "epoch": 0.21241, + "grad_norm": 0.9888145503249258, + "learning_rate": 0.003, + "loss": 4.074, + "step": 21241 + }, + { + "epoch": 0.21242, + "grad_norm": 1.4175814041586778, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 21242 + }, + { + "epoch": 0.21243, + "grad_norm": 0.9135014392827527, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 21243 + }, + { + "epoch": 0.21244, + "grad_norm": 0.8407080432996875, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 21244 + }, + { + "epoch": 0.21245, + "grad_norm": 0.9520872940598882, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 21245 + }, + { + "epoch": 0.21246, + "grad_norm": 1.1088222376126498, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 21246 + }, + { + "epoch": 0.21247, + "grad_norm": 1.2415378069350207, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 21247 + }, + { + "epoch": 0.21248, + "grad_norm": 0.9858931667070718, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 21248 + }, + { + "epoch": 0.21249, + "grad_norm": 1.0976044230518636, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 21249 + }, + { + "epoch": 0.2125, + "grad_norm": 1.1866902027272417, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 21250 + }, + { + "epoch": 0.21251, + "grad_norm": 0.9228628192162529, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 21251 + }, + { + "epoch": 0.21252, + "grad_norm": 0.9164005337547431, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 21252 + }, + { + "epoch": 0.21253, + "grad_norm": 0.9951260023279028, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 21253 + }, + { + "epoch": 0.21254, + "grad_norm": 1.0385497806728212, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 21254 + }, + { + "epoch": 0.21255, + "grad_norm": 0.9492064615153181, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 21255 + }, + { + "epoch": 0.21256, + "grad_norm": 1.0520546754572844, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 21256 + }, + { + "epoch": 0.21257, + "grad_norm": 1.0301308501873816, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 21257 + }, + { + "epoch": 0.21258, + "grad_norm": 0.9647738161928094, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 21258 + }, + { + "epoch": 0.21259, + "grad_norm": 1.0847308688182467, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 21259 + }, + { + "epoch": 0.2126, + "grad_norm": 1.0469093041945001, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 21260 + }, + { + "epoch": 0.21261, + "grad_norm": 1.1371552023445968, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 21261 + }, + { + "epoch": 0.21262, + "grad_norm": 1.0215211671101787, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 21262 + }, + { + "epoch": 0.21263, + "grad_norm": 1.0709946656880625, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 21263 + }, + { + "epoch": 0.21264, + "grad_norm": 1.1467731449192007, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 21264 + }, + { + "epoch": 0.21265, + "grad_norm": 1.3695516574227042, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 21265 + }, + { + "epoch": 0.21266, + "grad_norm": 1.0568074289625726, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 21266 + }, + { + "epoch": 0.21267, + "grad_norm": 0.9803316838462555, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 21267 + }, + { + "epoch": 0.21268, + "grad_norm": 1.0196405142167484, + "learning_rate": 0.003, + "loss": 4.061, + "step": 21268 + }, + { + "epoch": 0.21269, + "grad_norm": 1.130077977744749, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 21269 + }, + { + "epoch": 0.2127, + "grad_norm": 1.0663996725092588, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 21270 + }, + { + "epoch": 0.21271, + "grad_norm": 1.0762335888635823, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 21271 + }, + { + "epoch": 0.21272, + "grad_norm": 1.1918260747561114, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 21272 + }, + { + "epoch": 0.21273, + "grad_norm": 1.1207452612588593, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 21273 + }, + { + "epoch": 0.21274, + "grad_norm": 1.091533064304827, + "learning_rate": 0.003, + "loss": 4.0907, + "step": 21274 + }, + { + "epoch": 0.21275, + "grad_norm": 0.9244898722569342, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 21275 + }, + { + "epoch": 0.21276, + "grad_norm": 0.9182839316088798, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 21276 + }, + { + "epoch": 0.21277, + "grad_norm": 1.0195067176539017, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 21277 + }, + { + "epoch": 0.21278, + "grad_norm": 1.041109226606181, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 21278 + }, + { + "epoch": 0.21279, + "grad_norm": 1.351688053129703, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 21279 + }, + { + "epoch": 0.2128, + "grad_norm": 1.0616044428595397, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 21280 + }, + { + "epoch": 0.21281, + "grad_norm": 1.0598757788183744, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 21281 + }, + { + "epoch": 0.21282, + "grad_norm": 1.3224803467946695, + "learning_rate": 0.003, + "loss": 4.077, + "step": 21282 + }, + { + "epoch": 0.21283, + "grad_norm": 1.0029534797331092, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 21283 + }, + { + "epoch": 0.21284, + "grad_norm": 1.0543117805306779, + "learning_rate": 0.003, + "loss": 4.053, + "step": 21284 + }, + { + "epoch": 0.21285, + "grad_norm": 1.1315251981885368, + "learning_rate": 0.003, + "loss": 4.0925, + "step": 21285 + }, + { + "epoch": 0.21286, + "grad_norm": 0.9815035748139855, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 21286 + }, + { + "epoch": 0.21287, + "grad_norm": 0.9610969968538601, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 21287 + }, + { + "epoch": 0.21288, + "grad_norm": 1.2393680432268255, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 21288 + }, + { + "epoch": 0.21289, + "grad_norm": 1.0177564444396159, + "learning_rate": 0.003, + "loss": 4.054, + "step": 21289 + }, + { + "epoch": 0.2129, + "grad_norm": 1.1042183660720273, + "learning_rate": 0.003, + "loss": 4.032, + "step": 21290 + }, + { + "epoch": 0.21291, + "grad_norm": 1.0305703129610535, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 21291 + }, + { + "epoch": 0.21292, + "grad_norm": 1.057102497673516, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 21292 + }, + { + "epoch": 0.21293, + "grad_norm": 1.041018912930225, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 21293 + }, + { + "epoch": 0.21294, + "grad_norm": 1.1361529513831168, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 21294 + }, + { + "epoch": 0.21295, + "grad_norm": 1.1055784884657307, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 21295 + }, + { + "epoch": 0.21296, + "grad_norm": 1.0333546892353782, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 21296 + }, + { + "epoch": 0.21297, + "grad_norm": 1.0736958524876654, + "learning_rate": 0.003, + "loss": 4.06, + "step": 21297 + }, + { + "epoch": 0.21298, + "grad_norm": 0.803830091263215, + "learning_rate": 0.003, + "loss": 4.0933, + "step": 21298 + }, + { + "epoch": 0.21299, + "grad_norm": 0.9722511176626223, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 21299 + }, + { + "epoch": 0.213, + "grad_norm": 1.1310426282317743, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 21300 + }, + { + "epoch": 0.21301, + "grad_norm": 1.0488193728986075, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 21301 + }, + { + "epoch": 0.21302, + "grad_norm": 1.1645479539545371, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 21302 + }, + { + "epoch": 0.21303, + "grad_norm": 1.0062554455709156, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 21303 + }, + { + "epoch": 0.21304, + "grad_norm": 1.299915059950392, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 21304 + }, + { + "epoch": 0.21305, + "grad_norm": 1.093559053457306, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 21305 + }, + { + "epoch": 0.21306, + "grad_norm": 1.0883361025501108, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 21306 + }, + { + "epoch": 0.21307, + "grad_norm": 1.0988379839171039, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 21307 + }, + { + "epoch": 0.21308, + "grad_norm": 1.1640255284919887, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 21308 + }, + { + "epoch": 0.21309, + "grad_norm": 1.0870001345706013, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 21309 + }, + { + "epoch": 0.2131, + "grad_norm": 1.1227746622733463, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 21310 + }, + { + "epoch": 0.21311, + "grad_norm": 1.0266613388728314, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 21311 + }, + { + "epoch": 0.21312, + "grad_norm": 0.9614590017188844, + "learning_rate": 0.003, + "loss": 4.053, + "step": 21312 + }, + { + "epoch": 0.21313, + "grad_norm": 1.2825618639604133, + "learning_rate": 0.003, + "loss": 4.066, + "step": 21313 + }, + { + "epoch": 0.21314, + "grad_norm": 1.1902740746719067, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 21314 + }, + { + "epoch": 0.21315, + "grad_norm": 0.9447775224537129, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 21315 + }, + { + "epoch": 0.21316, + "grad_norm": 1.083973548388085, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 21316 + }, + { + "epoch": 0.21317, + "grad_norm": 0.8572082997722463, + "learning_rate": 0.003, + "loss": 4.0819, + "step": 21317 + }, + { + "epoch": 0.21318, + "grad_norm": 0.9977982290509118, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 21318 + }, + { + "epoch": 0.21319, + "grad_norm": 1.3891924593718181, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 21319 + }, + { + "epoch": 0.2132, + "grad_norm": 1.0874543582870284, + "learning_rate": 0.003, + "loss": 4.072, + "step": 21320 + }, + { + "epoch": 0.21321, + "grad_norm": 1.2019573166948891, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 21321 + }, + { + "epoch": 0.21322, + "grad_norm": 1.0347282435098937, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 21322 + }, + { + "epoch": 0.21323, + "grad_norm": 1.2034436167110603, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 21323 + }, + { + "epoch": 0.21324, + "grad_norm": 0.885922906751551, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 21324 + }, + { + "epoch": 0.21325, + "grad_norm": 0.9590146991565432, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 21325 + }, + { + "epoch": 0.21326, + "grad_norm": 1.1963103107356456, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 21326 + }, + { + "epoch": 0.21327, + "grad_norm": 0.9932700832504807, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 21327 + }, + { + "epoch": 0.21328, + "grad_norm": 1.1525767014190553, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 21328 + }, + { + "epoch": 0.21329, + "grad_norm": 1.001112945836027, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 21329 + }, + { + "epoch": 0.2133, + "grad_norm": 1.0988764028280216, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 21330 + }, + { + "epoch": 0.21331, + "grad_norm": 0.9602538726630094, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 21331 + }, + { + "epoch": 0.21332, + "grad_norm": 0.9458915613773757, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 21332 + }, + { + "epoch": 0.21333, + "grad_norm": 0.9888058027936597, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 21333 + }, + { + "epoch": 0.21334, + "grad_norm": 1.078888100779112, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 21334 + }, + { + "epoch": 0.21335, + "grad_norm": 0.9458970572838213, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 21335 + }, + { + "epoch": 0.21336, + "grad_norm": 0.9555174257855069, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 21336 + }, + { + "epoch": 0.21337, + "grad_norm": 1.1488752914702607, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 21337 + }, + { + "epoch": 0.21338, + "grad_norm": 0.9715884092280883, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 21338 + }, + { + "epoch": 0.21339, + "grad_norm": 1.1178694051511993, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 21339 + }, + { + "epoch": 0.2134, + "grad_norm": 1.1154442748964397, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 21340 + }, + { + "epoch": 0.21341, + "grad_norm": 0.9687621949623174, + "learning_rate": 0.003, + "loss": 4.075, + "step": 21341 + }, + { + "epoch": 0.21342, + "grad_norm": 1.3715170705662738, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 21342 + }, + { + "epoch": 0.21343, + "grad_norm": 1.077102320338362, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 21343 + }, + { + "epoch": 0.21344, + "grad_norm": 1.0023038801670139, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 21344 + }, + { + "epoch": 0.21345, + "grad_norm": 1.111371779252382, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 21345 + }, + { + "epoch": 0.21346, + "grad_norm": 1.0100206801172757, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 21346 + }, + { + "epoch": 0.21347, + "grad_norm": 0.9441230084689467, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 21347 + }, + { + "epoch": 0.21348, + "grad_norm": 1.0471060151731317, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 21348 + }, + { + "epoch": 0.21349, + "grad_norm": 0.9968844990410214, + "learning_rate": 0.003, + "loss": 4.054, + "step": 21349 + }, + { + "epoch": 0.2135, + "grad_norm": 0.9489091776710803, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 21350 + }, + { + "epoch": 0.21351, + "grad_norm": 1.268211185345142, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 21351 + }, + { + "epoch": 0.21352, + "grad_norm": 1.0385361086720344, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 21352 + }, + { + "epoch": 0.21353, + "grad_norm": 1.08091372616286, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 21353 + }, + { + "epoch": 0.21354, + "grad_norm": 1.1974717344739148, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 21354 + }, + { + "epoch": 0.21355, + "grad_norm": 0.9718376529169811, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 21355 + }, + { + "epoch": 0.21356, + "grad_norm": 1.2234818716253428, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 21356 + }, + { + "epoch": 0.21357, + "grad_norm": 1.0372558712271882, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 21357 + }, + { + "epoch": 0.21358, + "grad_norm": 1.242518520130103, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 21358 + }, + { + "epoch": 0.21359, + "grad_norm": 1.0190713660754291, + "learning_rate": 0.003, + "loss": 4.053, + "step": 21359 + }, + { + "epoch": 0.2136, + "grad_norm": 1.0112384234303786, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 21360 + }, + { + "epoch": 0.21361, + "grad_norm": 1.34437856963742, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 21361 + }, + { + "epoch": 0.21362, + "grad_norm": 0.9699481316784562, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 21362 + }, + { + "epoch": 0.21363, + "grad_norm": 1.1351045936307729, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 21363 + }, + { + "epoch": 0.21364, + "grad_norm": 1.1630293508264975, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 21364 + }, + { + "epoch": 0.21365, + "grad_norm": 1.2309563316406178, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 21365 + }, + { + "epoch": 0.21366, + "grad_norm": 0.8938243412157417, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 21366 + }, + { + "epoch": 0.21367, + "grad_norm": 0.8828367699548293, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 21367 + }, + { + "epoch": 0.21368, + "grad_norm": 0.8885672503302571, + "learning_rate": 0.003, + "loss": 4.067, + "step": 21368 + }, + { + "epoch": 0.21369, + "grad_norm": 0.9410422050327574, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 21369 + }, + { + "epoch": 0.2137, + "grad_norm": 1.0100640350043806, + "learning_rate": 0.003, + "loss": 4.0972, + "step": 21370 + }, + { + "epoch": 0.21371, + "grad_norm": 1.2798428426249322, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 21371 + }, + { + "epoch": 0.21372, + "grad_norm": 1.2173514756390007, + "learning_rate": 0.003, + "loss": 4.0919, + "step": 21372 + }, + { + "epoch": 0.21373, + "grad_norm": 0.9751227843939065, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 21373 + }, + { + "epoch": 0.21374, + "grad_norm": 1.018469391934704, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 21374 + }, + { + "epoch": 0.21375, + "grad_norm": 1.130508419645794, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 21375 + }, + { + "epoch": 0.21376, + "grad_norm": 0.9538295137789555, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 21376 + }, + { + "epoch": 0.21377, + "grad_norm": 0.9709002258576594, + "learning_rate": 0.003, + "loss": 4.077, + "step": 21377 + }, + { + "epoch": 0.21378, + "grad_norm": 0.9439936063373492, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 21378 + }, + { + "epoch": 0.21379, + "grad_norm": 1.2041302721869762, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 21379 + }, + { + "epoch": 0.2138, + "grad_norm": 1.0215467802313887, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 21380 + }, + { + "epoch": 0.21381, + "grad_norm": 0.9181711217225806, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 21381 + }, + { + "epoch": 0.21382, + "grad_norm": 1.0347427063982764, + "learning_rate": 0.003, + "loss": 4.026, + "step": 21382 + }, + { + "epoch": 0.21383, + "grad_norm": 1.1648096000090369, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 21383 + }, + { + "epoch": 0.21384, + "grad_norm": 1.0705423403232506, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 21384 + }, + { + "epoch": 0.21385, + "grad_norm": 1.024950445907933, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 21385 + }, + { + "epoch": 0.21386, + "grad_norm": 1.0311862625136272, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 21386 + }, + { + "epoch": 0.21387, + "grad_norm": 1.0794602235387516, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 21387 + }, + { + "epoch": 0.21388, + "grad_norm": 1.1718506455920172, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 21388 + }, + { + "epoch": 0.21389, + "grad_norm": 1.1726654696155594, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 21389 + }, + { + "epoch": 0.2139, + "grad_norm": 1.0582346899808128, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 21390 + }, + { + "epoch": 0.21391, + "grad_norm": 1.0043596217618964, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 21391 + }, + { + "epoch": 0.21392, + "grad_norm": 1.0775654233919447, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 21392 + }, + { + "epoch": 0.21393, + "grad_norm": 0.9677637280638006, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 21393 + }, + { + "epoch": 0.21394, + "grad_norm": 1.1289953215627928, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 21394 + }, + { + "epoch": 0.21395, + "grad_norm": 1.228688124343099, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 21395 + }, + { + "epoch": 0.21396, + "grad_norm": 0.8113490298378131, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 21396 + }, + { + "epoch": 0.21397, + "grad_norm": 0.7629936109605011, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 21397 + }, + { + "epoch": 0.21398, + "grad_norm": 0.7278206521248173, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 21398 + }, + { + "epoch": 0.21399, + "grad_norm": 0.7493933615497212, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 21399 + }, + { + "epoch": 0.214, + "grad_norm": 1.0004180821306414, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 21400 + }, + { + "epoch": 0.21401, + "grad_norm": 1.3199113292963491, + "learning_rate": 0.003, + "loss": 4.067, + "step": 21401 + }, + { + "epoch": 0.21402, + "grad_norm": 0.8880254412598829, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 21402 + }, + { + "epoch": 0.21403, + "grad_norm": 1.0139000804985943, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 21403 + }, + { + "epoch": 0.21404, + "grad_norm": 1.3567449995277974, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 21404 + }, + { + "epoch": 0.21405, + "grad_norm": 0.9834485530293444, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 21405 + }, + { + "epoch": 0.21406, + "grad_norm": 1.1204094250738625, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 21406 + }, + { + "epoch": 0.21407, + "grad_norm": 0.9378443807622204, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 21407 + }, + { + "epoch": 0.21408, + "grad_norm": 1.0627947395071697, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 21408 + }, + { + "epoch": 0.21409, + "grad_norm": 1.1524686781228457, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 21409 + }, + { + "epoch": 0.2141, + "grad_norm": 1.0322951553442128, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 21410 + }, + { + "epoch": 0.21411, + "grad_norm": 1.0734813706891986, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 21411 + }, + { + "epoch": 0.21412, + "grad_norm": 0.968756227796403, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 21412 + }, + { + "epoch": 0.21413, + "grad_norm": 1.0228286881238144, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 21413 + }, + { + "epoch": 0.21414, + "grad_norm": 1.0936880654223944, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 21414 + }, + { + "epoch": 0.21415, + "grad_norm": 1.17901186518045, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 21415 + }, + { + "epoch": 0.21416, + "grad_norm": 1.0412350796161367, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 21416 + }, + { + "epoch": 0.21417, + "grad_norm": 0.9822807477773742, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 21417 + }, + { + "epoch": 0.21418, + "grad_norm": 1.1333934745724676, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 21418 + }, + { + "epoch": 0.21419, + "grad_norm": 0.8748994680298862, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 21419 + }, + { + "epoch": 0.2142, + "grad_norm": 0.8155989235483306, + "learning_rate": 0.003, + "loss": 4.048, + "step": 21420 + }, + { + "epoch": 0.21421, + "grad_norm": 0.9071467486588297, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 21421 + }, + { + "epoch": 0.21422, + "grad_norm": 1.2216199894966777, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 21422 + }, + { + "epoch": 0.21423, + "grad_norm": 1.1198206464754956, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 21423 + }, + { + "epoch": 0.21424, + "grad_norm": 1.0434044892047645, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 21424 + }, + { + "epoch": 0.21425, + "grad_norm": 1.08231273112757, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 21425 + }, + { + "epoch": 0.21426, + "grad_norm": 1.1423211937866262, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 21426 + }, + { + "epoch": 0.21427, + "grad_norm": 0.9329592327925849, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 21427 + }, + { + "epoch": 0.21428, + "grad_norm": 1.0535562754462775, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 21428 + }, + { + "epoch": 0.21429, + "grad_norm": 1.182105989060146, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 21429 + }, + { + "epoch": 0.2143, + "grad_norm": 1.0010078598365137, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 21430 + }, + { + "epoch": 0.21431, + "grad_norm": 1.0006613899674832, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 21431 + }, + { + "epoch": 0.21432, + "grad_norm": 1.212287758666535, + "learning_rate": 0.003, + "loss": 4.0874, + "step": 21432 + }, + { + "epoch": 0.21433, + "grad_norm": 1.002641684582012, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 21433 + }, + { + "epoch": 0.21434, + "grad_norm": 1.2007037879352644, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 21434 + }, + { + "epoch": 0.21435, + "grad_norm": 1.2116934137161046, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 21435 + }, + { + "epoch": 0.21436, + "grad_norm": 1.1995763337506056, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 21436 + }, + { + "epoch": 0.21437, + "grad_norm": 1.225059225649724, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 21437 + }, + { + "epoch": 0.21438, + "grad_norm": 0.9853024509809132, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 21438 + }, + { + "epoch": 0.21439, + "grad_norm": 1.097340706268492, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 21439 + }, + { + "epoch": 0.2144, + "grad_norm": 1.0870458082698866, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 21440 + }, + { + "epoch": 0.21441, + "grad_norm": 1.0601003571615435, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 21441 + }, + { + "epoch": 0.21442, + "grad_norm": 1.086480290555551, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 21442 + }, + { + "epoch": 0.21443, + "grad_norm": 1.0243149262090492, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 21443 + }, + { + "epoch": 0.21444, + "grad_norm": 1.0737713450196862, + "learning_rate": 0.003, + "loss": 4.0929, + "step": 21444 + }, + { + "epoch": 0.21445, + "grad_norm": 1.0073952721527655, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 21445 + }, + { + "epoch": 0.21446, + "grad_norm": 0.9492008459168798, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 21446 + }, + { + "epoch": 0.21447, + "grad_norm": 0.8799527733694723, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 21447 + }, + { + "epoch": 0.21448, + "grad_norm": 0.932864852359693, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 21448 + }, + { + "epoch": 0.21449, + "grad_norm": 1.1970463467991124, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 21449 + }, + { + "epoch": 0.2145, + "grad_norm": 1.1171908212677994, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 21450 + }, + { + "epoch": 0.21451, + "grad_norm": 1.1179187054677384, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 21451 + }, + { + "epoch": 0.21452, + "grad_norm": 1.0086477912515213, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 21452 + }, + { + "epoch": 0.21453, + "grad_norm": 1.3443500123593468, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 21453 + }, + { + "epoch": 0.21454, + "grad_norm": 0.9684426488105321, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 21454 + }, + { + "epoch": 0.21455, + "grad_norm": 1.0529668716187763, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 21455 + }, + { + "epoch": 0.21456, + "grad_norm": 1.1058121385362811, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 21456 + }, + { + "epoch": 0.21457, + "grad_norm": 1.0430591256303883, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 21457 + }, + { + "epoch": 0.21458, + "grad_norm": 1.032253679987056, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 21458 + }, + { + "epoch": 0.21459, + "grad_norm": 1.1409950026340119, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 21459 + }, + { + "epoch": 0.2146, + "grad_norm": 1.0820431361526979, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 21460 + }, + { + "epoch": 0.21461, + "grad_norm": 0.9301495059152519, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 21461 + }, + { + "epoch": 0.21462, + "grad_norm": 0.9186144859160273, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 21462 + }, + { + "epoch": 0.21463, + "grad_norm": 0.9421215332657836, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 21463 + }, + { + "epoch": 0.21464, + "grad_norm": 1.085448527982102, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 21464 + }, + { + "epoch": 0.21465, + "grad_norm": 1.1243946926905768, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 21465 + }, + { + "epoch": 0.21466, + "grad_norm": 1.3673501548518185, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 21466 + }, + { + "epoch": 0.21467, + "grad_norm": 0.9680478023136071, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 21467 + }, + { + "epoch": 0.21468, + "grad_norm": 1.0061415185045852, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 21468 + }, + { + "epoch": 0.21469, + "grad_norm": 1.3832152996009708, + "learning_rate": 0.003, + "loss": 4.0992, + "step": 21469 + }, + { + "epoch": 0.2147, + "grad_norm": 0.9833466742170001, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 21470 + }, + { + "epoch": 0.21471, + "grad_norm": 1.1059631757863784, + "learning_rate": 0.003, + "loss": 4.05, + "step": 21471 + }, + { + "epoch": 0.21472, + "grad_norm": 1.143677730458373, + "learning_rate": 0.003, + "loss": 4.081, + "step": 21472 + }, + { + "epoch": 0.21473, + "grad_norm": 1.0988709850464011, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 21473 + }, + { + "epoch": 0.21474, + "grad_norm": 1.1652929929355134, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 21474 + }, + { + "epoch": 0.21475, + "grad_norm": 0.9178046154440261, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 21475 + }, + { + "epoch": 0.21476, + "grad_norm": 0.97457680862927, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 21476 + }, + { + "epoch": 0.21477, + "grad_norm": 1.0894444159203291, + "learning_rate": 0.003, + "loss": 4.083, + "step": 21477 + }, + { + "epoch": 0.21478, + "grad_norm": 1.151979702409531, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 21478 + }, + { + "epoch": 0.21479, + "grad_norm": 1.074566819769084, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 21479 + }, + { + "epoch": 0.2148, + "grad_norm": 1.1525483021484935, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 21480 + }, + { + "epoch": 0.21481, + "grad_norm": 1.0255079857392, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 21481 + }, + { + "epoch": 0.21482, + "grad_norm": 1.1193984559257095, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 21482 + }, + { + "epoch": 0.21483, + "grad_norm": 1.1918950813013083, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 21483 + }, + { + "epoch": 0.21484, + "grad_norm": 1.1100355231620087, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 21484 + }, + { + "epoch": 0.21485, + "grad_norm": 1.0110859782848514, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 21485 + }, + { + "epoch": 0.21486, + "grad_norm": 1.001132844809797, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 21486 + }, + { + "epoch": 0.21487, + "grad_norm": 1.0800240558800722, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 21487 + }, + { + "epoch": 0.21488, + "grad_norm": 0.9262323191653086, + "learning_rate": 0.003, + "loss": 4.0925, + "step": 21488 + }, + { + "epoch": 0.21489, + "grad_norm": 0.9905654177980457, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 21489 + }, + { + "epoch": 0.2149, + "grad_norm": 1.034511270160601, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 21490 + }, + { + "epoch": 0.21491, + "grad_norm": 1.0307990103798377, + "learning_rate": 0.003, + "loss": 4.1015, + "step": 21491 + }, + { + "epoch": 0.21492, + "grad_norm": 1.1026562330744605, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 21492 + }, + { + "epoch": 0.21493, + "grad_norm": 1.179476360999142, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 21493 + }, + { + "epoch": 0.21494, + "grad_norm": 1.1466075962449616, + "learning_rate": 0.003, + "loss": 4.1016, + "step": 21494 + }, + { + "epoch": 0.21495, + "grad_norm": 1.052061380055686, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 21495 + }, + { + "epoch": 0.21496, + "grad_norm": 1.0654138293308901, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 21496 + }, + { + "epoch": 0.21497, + "grad_norm": 1.046008325116196, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 21497 + }, + { + "epoch": 0.21498, + "grad_norm": 0.8620468076074921, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 21498 + }, + { + "epoch": 0.21499, + "grad_norm": 1.0140445614204305, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 21499 + }, + { + "epoch": 0.215, + "grad_norm": 0.9698435032958104, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 21500 + }, + { + "epoch": 0.21501, + "grad_norm": 1.037292650328754, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 21501 + }, + { + "epoch": 0.21502, + "grad_norm": 1.242057196699381, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 21502 + }, + { + "epoch": 0.21503, + "grad_norm": 1.0458972064414578, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 21503 + }, + { + "epoch": 0.21504, + "grad_norm": 1.1125838783984106, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 21504 + }, + { + "epoch": 0.21505, + "grad_norm": 1.160146307146324, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 21505 + }, + { + "epoch": 0.21506, + "grad_norm": 1.0226546209587806, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 21506 + }, + { + "epoch": 0.21507, + "grad_norm": 1.1705374063349419, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 21507 + }, + { + "epoch": 0.21508, + "grad_norm": 1.0245365937485644, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 21508 + }, + { + "epoch": 0.21509, + "grad_norm": 1.0323978164519798, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 21509 + }, + { + "epoch": 0.2151, + "grad_norm": 0.9904211399440084, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 21510 + }, + { + "epoch": 0.21511, + "grad_norm": 1.1340780747972494, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 21511 + }, + { + "epoch": 0.21512, + "grad_norm": 1.077652382315818, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 21512 + }, + { + "epoch": 0.21513, + "grad_norm": 1.3119707168057697, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 21513 + }, + { + "epoch": 0.21514, + "grad_norm": 1.0630002081882177, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 21514 + }, + { + "epoch": 0.21515, + "grad_norm": 1.0499508424611121, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 21515 + }, + { + "epoch": 0.21516, + "grad_norm": 0.9856889766748522, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 21516 + }, + { + "epoch": 0.21517, + "grad_norm": 1.020488469157323, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 21517 + }, + { + "epoch": 0.21518, + "grad_norm": 1.1351038602578452, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 21518 + }, + { + "epoch": 0.21519, + "grad_norm": 0.944197598459694, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 21519 + }, + { + "epoch": 0.2152, + "grad_norm": 1.0486599721618586, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 21520 + }, + { + "epoch": 0.21521, + "grad_norm": 1.2364691000270138, + "learning_rate": 0.003, + "loss": 4.065, + "step": 21521 + }, + { + "epoch": 0.21522, + "grad_norm": 0.9736407135475406, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 21522 + }, + { + "epoch": 0.21523, + "grad_norm": 1.094409044543281, + "learning_rate": 0.003, + "loss": 4.054, + "step": 21523 + }, + { + "epoch": 0.21524, + "grad_norm": 1.204764243253608, + "learning_rate": 0.003, + "loss": 4.0854, + "step": 21524 + }, + { + "epoch": 0.21525, + "grad_norm": 0.9036680801847775, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 21525 + }, + { + "epoch": 0.21526, + "grad_norm": 1.1046584650432338, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 21526 + }, + { + "epoch": 0.21527, + "grad_norm": 1.3610997995709242, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 21527 + }, + { + "epoch": 0.21528, + "grad_norm": 1.1910687272273388, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 21528 + }, + { + "epoch": 0.21529, + "grad_norm": 1.0185632502925615, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 21529 + }, + { + "epoch": 0.2153, + "grad_norm": 1.0925719523265098, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 21530 + }, + { + "epoch": 0.21531, + "grad_norm": 1.1217712720233952, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 21531 + }, + { + "epoch": 0.21532, + "grad_norm": 1.0868449959467181, + "learning_rate": 0.003, + "loss": 4.068, + "step": 21532 + }, + { + "epoch": 0.21533, + "grad_norm": 1.1109213621016323, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 21533 + }, + { + "epoch": 0.21534, + "grad_norm": 1.3936280902079439, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 21534 + }, + { + "epoch": 0.21535, + "grad_norm": 1.0684037777113122, + "learning_rate": 0.003, + "loss": 4.0884, + "step": 21535 + }, + { + "epoch": 0.21536, + "grad_norm": 1.01897972237249, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 21536 + }, + { + "epoch": 0.21537, + "grad_norm": 1.0157261831083233, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 21537 + }, + { + "epoch": 0.21538, + "grad_norm": 1.2389400541979256, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 21538 + }, + { + "epoch": 0.21539, + "grad_norm": 1.2215916905808901, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 21539 + }, + { + "epoch": 0.2154, + "grad_norm": 0.9500499371511821, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 21540 + }, + { + "epoch": 0.21541, + "grad_norm": 1.0079057117778578, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 21541 + }, + { + "epoch": 0.21542, + "grad_norm": 1.0084755561417007, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 21542 + }, + { + "epoch": 0.21543, + "grad_norm": 0.9933382894616929, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 21543 + }, + { + "epoch": 0.21544, + "grad_norm": 1.0917971235965132, + "learning_rate": 0.003, + "loss": 4.055, + "step": 21544 + }, + { + "epoch": 0.21545, + "grad_norm": 0.9895021667529471, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 21545 + }, + { + "epoch": 0.21546, + "grad_norm": 1.2213439401819528, + "learning_rate": 0.003, + "loss": 4.054, + "step": 21546 + }, + { + "epoch": 0.21547, + "grad_norm": 1.0530090941183796, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 21547 + }, + { + "epoch": 0.21548, + "grad_norm": 1.0516129304049857, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 21548 + }, + { + "epoch": 0.21549, + "grad_norm": 1.1654976777794916, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 21549 + }, + { + "epoch": 0.2155, + "grad_norm": 0.9401186743037688, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 21550 + }, + { + "epoch": 0.21551, + "grad_norm": 0.8185550013538742, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 21551 + }, + { + "epoch": 0.21552, + "grad_norm": 0.8587299186207644, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 21552 + }, + { + "epoch": 0.21553, + "grad_norm": 1.116414997418698, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 21553 + }, + { + "epoch": 0.21554, + "grad_norm": 1.187919606502965, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 21554 + }, + { + "epoch": 0.21555, + "grad_norm": 0.9298291666881922, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 21555 + }, + { + "epoch": 0.21556, + "grad_norm": 1.0915999762600372, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 21556 + }, + { + "epoch": 0.21557, + "grad_norm": 1.1047564844049087, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 21557 + }, + { + "epoch": 0.21558, + "grad_norm": 1.0512254765593416, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 21558 + }, + { + "epoch": 0.21559, + "grad_norm": 1.08489460376171, + "learning_rate": 0.003, + "loss": 4.094, + "step": 21559 + }, + { + "epoch": 0.2156, + "grad_norm": 1.125996200595082, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 21560 + }, + { + "epoch": 0.21561, + "grad_norm": 0.9620208437827598, + "learning_rate": 0.003, + "loss": 4.0934, + "step": 21561 + }, + { + "epoch": 0.21562, + "grad_norm": 1.2009947089726176, + "learning_rate": 0.003, + "loss": 4.0922, + "step": 21562 + }, + { + "epoch": 0.21563, + "grad_norm": 1.120535573745468, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 21563 + }, + { + "epoch": 0.21564, + "grad_norm": 1.0961239065920505, + "learning_rate": 0.003, + "loss": 4.036, + "step": 21564 + }, + { + "epoch": 0.21565, + "grad_norm": 0.9972427359732287, + "learning_rate": 0.003, + "loss": 4.0942, + "step": 21565 + }, + { + "epoch": 0.21566, + "grad_norm": 1.2627685648927507, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 21566 + }, + { + "epoch": 0.21567, + "grad_norm": 1.2466614169489239, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 21567 + }, + { + "epoch": 0.21568, + "grad_norm": 0.949120186205377, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 21568 + }, + { + "epoch": 0.21569, + "grad_norm": 0.9149349451036899, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 21569 + }, + { + "epoch": 0.2157, + "grad_norm": 1.0657052625505827, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 21570 + }, + { + "epoch": 0.21571, + "grad_norm": 1.0613119771963557, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 21571 + }, + { + "epoch": 0.21572, + "grad_norm": 0.956317688187834, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 21572 + }, + { + "epoch": 0.21573, + "grad_norm": 1.0394576173684436, + "learning_rate": 0.003, + "loss": 4.05, + "step": 21573 + }, + { + "epoch": 0.21574, + "grad_norm": 1.119924607809392, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 21574 + }, + { + "epoch": 0.21575, + "grad_norm": 0.9596816120893071, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 21575 + }, + { + "epoch": 0.21576, + "grad_norm": 1.1107971619571202, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 21576 + }, + { + "epoch": 0.21577, + "grad_norm": 1.256311701587101, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 21577 + }, + { + "epoch": 0.21578, + "grad_norm": 1.05604087591441, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 21578 + }, + { + "epoch": 0.21579, + "grad_norm": 1.2523234453295735, + "learning_rate": 0.003, + "loss": 4.074, + "step": 21579 + }, + { + "epoch": 0.2158, + "grad_norm": 0.988895341901757, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 21580 + }, + { + "epoch": 0.21581, + "grad_norm": 0.955021278945525, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 21581 + }, + { + "epoch": 0.21582, + "grad_norm": 1.0235067560762974, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 21582 + }, + { + "epoch": 0.21583, + "grad_norm": 1.1243545104433899, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 21583 + }, + { + "epoch": 0.21584, + "grad_norm": 1.1417195502895636, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 21584 + }, + { + "epoch": 0.21585, + "grad_norm": 0.8923262728478317, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 21585 + }, + { + "epoch": 0.21586, + "grad_norm": 0.9290711978380102, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 21586 + }, + { + "epoch": 0.21587, + "grad_norm": 0.9136984137604982, + "learning_rate": 0.003, + "loss": 4.0918, + "step": 21587 + }, + { + "epoch": 0.21588, + "grad_norm": 0.8854598801547823, + "learning_rate": 0.003, + "loss": 4.048, + "step": 21588 + }, + { + "epoch": 0.21589, + "grad_norm": 0.8789871007451678, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 21589 + }, + { + "epoch": 0.2159, + "grad_norm": 1.0322688582856423, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 21590 + }, + { + "epoch": 0.21591, + "grad_norm": 1.3784608764824404, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 21591 + }, + { + "epoch": 0.21592, + "grad_norm": 0.8374654878088557, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 21592 + }, + { + "epoch": 0.21593, + "grad_norm": 0.9636594277255688, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 21593 + }, + { + "epoch": 0.21594, + "grad_norm": 1.3144439004705126, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 21594 + }, + { + "epoch": 0.21595, + "grad_norm": 1.009494035681793, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 21595 + }, + { + "epoch": 0.21596, + "grad_norm": 1.094377880206675, + "learning_rate": 0.003, + "loss": 4.042, + "step": 21596 + }, + { + "epoch": 0.21597, + "grad_norm": 1.1606336419248025, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 21597 + }, + { + "epoch": 0.21598, + "grad_norm": 0.8957440528966354, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 21598 + }, + { + "epoch": 0.21599, + "grad_norm": 0.830804160145562, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 21599 + }, + { + "epoch": 0.216, + "grad_norm": 0.9366034180261275, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 21600 + }, + { + "epoch": 0.21601, + "grad_norm": 1.1801575198083991, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 21601 + }, + { + "epoch": 0.21602, + "grad_norm": 0.8696499151918002, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 21602 + }, + { + "epoch": 0.21603, + "grad_norm": 1.1273238099588543, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 21603 + }, + { + "epoch": 0.21604, + "grad_norm": 1.3279995136758755, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 21604 + }, + { + "epoch": 0.21605, + "grad_norm": 0.9514415194720047, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 21605 + }, + { + "epoch": 0.21606, + "grad_norm": 0.98260037045213, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 21606 + }, + { + "epoch": 0.21607, + "grad_norm": 1.0513849474443124, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 21607 + }, + { + "epoch": 0.21608, + "grad_norm": 1.0781945909701984, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 21608 + }, + { + "epoch": 0.21609, + "grad_norm": 0.9887748215717975, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 21609 + }, + { + "epoch": 0.2161, + "grad_norm": 1.0032940292079808, + "learning_rate": 0.003, + "loss": 4.038, + "step": 21610 + }, + { + "epoch": 0.21611, + "grad_norm": 1.209697490600328, + "learning_rate": 0.003, + "loss": 4.061, + "step": 21611 + }, + { + "epoch": 0.21612, + "grad_norm": 1.1080189858453595, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 21612 + }, + { + "epoch": 0.21613, + "grad_norm": 1.069745829054759, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 21613 + }, + { + "epoch": 0.21614, + "grad_norm": 1.133027223058248, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 21614 + }, + { + "epoch": 0.21615, + "grad_norm": 0.8893743165204914, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 21615 + }, + { + "epoch": 0.21616, + "grad_norm": 1.0438289287743263, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 21616 + }, + { + "epoch": 0.21617, + "grad_norm": 1.0605673580752484, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 21617 + }, + { + "epoch": 0.21618, + "grad_norm": 1.203898239498179, + "learning_rate": 0.003, + "loss": 4.0905, + "step": 21618 + }, + { + "epoch": 0.21619, + "grad_norm": 1.029089388572224, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 21619 + }, + { + "epoch": 0.2162, + "grad_norm": 1.1895386987973444, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 21620 + }, + { + "epoch": 0.21621, + "grad_norm": 1.0348035326246454, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 21621 + }, + { + "epoch": 0.21622, + "grad_norm": 1.18543458305186, + "learning_rate": 0.003, + "loss": 4.0931, + "step": 21622 + }, + { + "epoch": 0.21623, + "grad_norm": 1.206155251262532, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 21623 + }, + { + "epoch": 0.21624, + "grad_norm": 1.023388289923947, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 21624 + }, + { + "epoch": 0.21625, + "grad_norm": 0.9924819885420434, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 21625 + }, + { + "epoch": 0.21626, + "grad_norm": 1.1287309523211717, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 21626 + }, + { + "epoch": 0.21627, + "grad_norm": 1.0578059148351417, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 21627 + }, + { + "epoch": 0.21628, + "grad_norm": 1.0471991345076561, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 21628 + }, + { + "epoch": 0.21629, + "grad_norm": 1.2106860822718304, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 21629 + }, + { + "epoch": 0.2163, + "grad_norm": 0.9418886217383721, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 21630 + }, + { + "epoch": 0.21631, + "grad_norm": 1.0185276535430683, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 21631 + }, + { + "epoch": 0.21632, + "grad_norm": 0.8795498248765667, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 21632 + }, + { + "epoch": 0.21633, + "grad_norm": 0.8167051916373704, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 21633 + }, + { + "epoch": 0.21634, + "grad_norm": 1.0908389385304265, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 21634 + }, + { + "epoch": 0.21635, + "grad_norm": 1.0491741597383908, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 21635 + }, + { + "epoch": 0.21636, + "grad_norm": 1.0648304631594245, + "learning_rate": 0.003, + "loss": 4.067, + "step": 21636 + }, + { + "epoch": 0.21637, + "grad_norm": 1.282777897965275, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 21637 + }, + { + "epoch": 0.21638, + "grad_norm": 0.9534034599062005, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 21638 + }, + { + "epoch": 0.21639, + "grad_norm": 1.092985257004578, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 21639 + }, + { + "epoch": 0.2164, + "grad_norm": 1.2243194703624636, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 21640 + }, + { + "epoch": 0.21641, + "grad_norm": 1.1830254609972009, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 21641 + }, + { + "epoch": 0.21642, + "grad_norm": 1.0200016953999798, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 21642 + }, + { + "epoch": 0.21643, + "grad_norm": 1.0346518180239368, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 21643 + }, + { + "epoch": 0.21644, + "grad_norm": 0.9517013538409715, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 21644 + }, + { + "epoch": 0.21645, + "grad_norm": 0.9701109098183953, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 21645 + }, + { + "epoch": 0.21646, + "grad_norm": 0.990618392986658, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 21646 + }, + { + "epoch": 0.21647, + "grad_norm": 1.0160498541341705, + "learning_rate": 0.003, + "loss": 4.04, + "step": 21647 + }, + { + "epoch": 0.21648, + "grad_norm": 1.0364672061683544, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 21648 + }, + { + "epoch": 0.21649, + "grad_norm": 1.155877905685031, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 21649 + }, + { + "epoch": 0.2165, + "grad_norm": 1.0709786731240922, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 21650 + }, + { + "epoch": 0.21651, + "grad_norm": 1.0082681486089675, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 21651 + }, + { + "epoch": 0.21652, + "grad_norm": 1.0493318215232867, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 21652 + }, + { + "epoch": 0.21653, + "grad_norm": 1.1562688410270574, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 21653 + }, + { + "epoch": 0.21654, + "grad_norm": 0.9400444885047932, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 21654 + }, + { + "epoch": 0.21655, + "grad_norm": 1.18456347305482, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 21655 + }, + { + "epoch": 0.21656, + "grad_norm": 1.0209804048914426, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 21656 + }, + { + "epoch": 0.21657, + "grad_norm": 1.2558382159263803, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 21657 + }, + { + "epoch": 0.21658, + "grad_norm": 0.9312478159122216, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 21658 + }, + { + "epoch": 0.21659, + "grad_norm": 1.227822954377595, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 21659 + }, + { + "epoch": 0.2166, + "grad_norm": 0.9126866955406738, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 21660 + }, + { + "epoch": 0.21661, + "grad_norm": 1.019334707523089, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 21661 + }, + { + "epoch": 0.21662, + "grad_norm": 1.067831645835217, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 21662 + }, + { + "epoch": 0.21663, + "grad_norm": 1.0680907472527872, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 21663 + }, + { + "epoch": 0.21664, + "grad_norm": 1.1342416677231122, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 21664 + }, + { + "epoch": 0.21665, + "grad_norm": 0.9661093284119754, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 21665 + }, + { + "epoch": 0.21666, + "grad_norm": 1.311476391658844, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 21666 + }, + { + "epoch": 0.21667, + "grad_norm": 0.9441414871389969, + "learning_rate": 0.003, + "loss": 4.051, + "step": 21667 + }, + { + "epoch": 0.21668, + "grad_norm": 1.1030550013831546, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 21668 + }, + { + "epoch": 0.21669, + "grad_norm": 1.2777600028607583, + "learning_rate": 0.003, + "loss": 4.04, + "step": 21669 + }, + { + "epoch": 0.2167, + "grad_norm": 1.0390908196970854, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 21670 + }, + { + "epoch": 0.21671, + "grad_norm": 1.1913379320907644, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 21671 + }, + { + "epoch": 0.21672, + "grad_norm": 0.9393098298739957, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 21672 + }, + { + "epoch": 0.21673, + "grad_norm": 1.0670305614094764, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 21673 + }, + { + "epoch": 0.21674, + "grad_norm": 1.0263762774091783, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 21674 + }, + { + "epoch": 0.21675, + "grad_norm": 0.944924845245466, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 21675 + }, + { + "epoch": 0.21676, + "grad_norm": 0.9024801549098858, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 21676 + }, + { + "epoch": 0.21677, + "grad_norm": 0.9968989893386268, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 21677 + }, + { + "epoch": 0.21678, + "grad_norm": 1.390557597155966, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 21678 + }, + { + "epoch": 0.21679, + "grad_norm": 1.0051455057431322, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 21679 + }, + { + "epoch": 0.2168, + "grad_norm": 1.2631678714398271, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 21680 + }, + { + "epoch": 0.21681, + "grad_norm": 0.9406672697962443, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 21681 + }, + { + "epoch": 0.21682, + "grad_norm": 0.909916695839513, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 21682 + }, + { + "epoch": 0.21683, + "grad_norm": 1.0658157286784247, + "learning_rate": 0.003, + "loss": 4.0871, + "step": 21683 + }, + { + "epoch": 0.21684, + "grad_norm": 1.1911704952664846, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 21684 + }, + { + "epoch": 0.21685, + "grad_norm": 1.0063029457641142, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 21685 + }, + { + "epoch": 0.21686, + "grad_norm": 1.134439923995109, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 21686 + }, + { + "epoch": 0.21687, + "grad_norm": 1.073348447622698, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 21687 + }, + { + "epoch": 0.21688, + "grad_norm": 1.087188845614753, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 21688 + }, + { + "epoch": 0.21689, + "grad_norm": 1.123049985482784, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 21689 + }, + { + "epoch": 0.2169, + "grad_norm": 0.8699528029141139, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 21690 + }, + { + "epoch": 0.21691, + "grad_norm": 0.9096379285030786, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 21691 + }, + { + "epoch": 0.21692, + "grad_norm": 1.0764346538230987, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 21692 + }, + { + "epoch": 0.21693, + "grad_norm": 1.0820955426428855, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 21693 + }, + { + "epoch": 0.21694, + "grad_norm": 1.1463747605657717, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 21694 + }, + { + "epoch": 0.21695, + "grad_norm": 0.9727994072509841, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 21695 + }, + { + "epoch": 0.21696, + "grad_norm": 0.9956025016808704, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 21696 + }, + { + "epoch": 0.21697, + "grad_norm": 1.3692930437906932, + "learning_rate": 0.003, + "loss": 4.064, + "step": 21697 + }, + { + "epoch": 0.21698, + "grad_norm": 0.8801660811938229, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 21698 + }, + { + "epoch": 0.21699, + "grad_norm": 0.8933224001199339, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 21699 + }, + { + "epoch": 0.217, + "grad_norm": 0.907786405247936, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 21700 + }, + { + "epoch": 0.21701, + "grad_norm": 1.126130657378152, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 21701 + }, + { + "epoch": 0.21702, + "grad_norm": 1.0592292117696351, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 21702 + }, + { + "epoch": 0.21703, + "grad_norm": 1.0708545506297031, + "learning_rate": 0.003, + "loss": 4.043, + "step": 21703 + }, + { + "epoch": 0.21704, + "grad_norm": 1.057463962841132, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 21704 + }, + { + "epoch": 0.21705, + "grad_norm": 0.8935918211873309, + "learning_rate": 0.003, + "loss": 4.033, + "step": 21705 + }, + { + "epoch": 0.21706, + "grad_norm": 1.0730556318947846, + "learning_rate": 0.003, + "loss": 4.048, + "step": 21706 + }, + { + "epoch": 0.21707, + "grad_norm": 1.2578951444442374, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 21707 + }, + { + "epoch": 0.21708, + "grad_norm": 1.0530993230540482, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 21708 + }, + { + "epoch": 0.21709, + "grad_norm": 1.1990128861281393, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 21709 + }, + { + "epoch": 0.2171, + "grad_norm": 1.0001748315621062, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 21710 + }, + { + "epoch": 0.21711, + "grad_norm": 1.416497083482433, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 21711 + }, + { + "epoch": 0.21712, + "grad_norm": 1.0939395786773787, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 21712 + }, + { + "epoch": 0.21713, + "grad_norm": 1.0124115974172727, + "learning_rate": 0.003, + "loss": 4.082, + "step": 21713 + }, + { + "epoch": 0.21714, + "grad_norm": 1.0346636179223017, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 21714 + }, + { + "epoch": 0.21715, + "grad_norm": 1.147594281262174, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 21715 + }, + { + "epoch": 0.21716, + "grad_norm": 0.8454227710223359, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 21716 + }, + { + "epoch": 0.21717, + "grad_norm": 0.9837982402144017, + "learning_rate": 0.003, + "loss": 4.051, + "step": 21717 + }, + { + "epoch": 0.21718, + "grad_norm": 1.1565109400247808, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 21718 + }, + { + "epoch": 0.21719, + "grad_norm": 1.0824388166225245, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 21719 + }, + { + "epoch": 0.2172, + "grad_norm": 1.2181128379430055, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 21720 + }, + { + "epoch": 0.21721, + "grad_norm": 0.9599378844633476, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 21721 + }, + { + "epoch": 0.21722, + "grad_norm": 0.9114999569314388, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 21722 + }, + { + "epoch": 0.21723, + "grad_norm": 0.9835861930749623, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 21723 + }, + { + "epoch": 0.21724, + "grad_norm": 1.2756099577861488, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 21724 + }, + { + "epoch": 0.21725, + "grad_norm": 1.0499715332923916, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 21725 + }, + { + "epoch": 0.21726, + "grad_norm": 1.1734172301646209, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 21726 + }, + { + "epoch": 0.21727, + "grad_norm": 0.8826410618392563, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 21727 + }, + { + "epoch": 0.21728, + "grad_norm": 0.8956504728047184, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 21728 + }, + { + "epoch": 0.21729, + "grad_norm": 1.0822699303951953, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 21729 + }, + { + "epoch": 0.2173, + "grad_norm": 1.0996474480923613, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 21730 + }, + { + "epoch": 0.21731, + "grad_norm": 0.9924643299354381, + "learning_rate": 0.003, + "loss": 4.1011, + "step": 21731 + }, + { + "epoch": 0.21732, + "grad_norm": 1.015547211468348, + "learning_rate": 0.003, + "loss": 4.076, + "step": 21732 + }, + { + "epoch": 0.21733, + "grad_norm": 1.0299637965494899, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 21733 + }, + { + "epoch": 0.21734, + "grad_norm": 1.1679542448870912, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 21734 + }, + { + "epoch": 0.21735, + "grad_norm": 1.2370716351489661, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 21735 + }, + { + "epoch": 0.21736, + "grad_norm": 1.0441962582638746, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 21736 + }, + { + "epoch": 0.21737, + "grad_norm": 1.2665459581226008, + "learning_rate": 0.003, + "loss": 4.044, + "step": 21737 + }, + { + "epoch": 0.21738, + "grad_norm": 1.0041843008176363, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 21738 + }, + { + "epoch": 0.21739, + "grad_norm": 1.0088386112757253, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 21739 + }, + { + "epoch": 0.2174, + "grad_norm": 1.0360429383017693, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 21740 + }, + { + "epoch": 0.21741, + "grad_norm": 1.0155065579289504, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 21741 + }, + { + "epoch": 0.21742, + "grad_norm": 1.1239763306471062, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 21742 + }, + { + "epoch": 0.21743, + "grad_norm": 1.0797825963810062, + "learning_rate": 0.003, + "loss": 4.0907, + "step": 21743 + }, + { + "epoch": 0.21744, + "grad_norm": 1.0484754290540608, + "learning_rate": 0.003, + "loss": 4.071, + "step": 21744 + }, + { + "epoch": 0.21745, + "grad_norm": 1.0571381591032232, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 21745 + }, + { + "epoch": 0.21746, + "grad_norm": 1.0601439507684296, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 21746 + }, + { + "epoch": 0.21747, + "grad_norm": 1.1115418201487968, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 21747 + }, + { + "epoch": 0.21748, + "grad_norm": 1.0043688847385357, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 21748 + }, + { + "epoch": 0.21749, + "grad_norm": 0.9420365376330115, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 21749 + }, + { + "epoch": 0.2175, + "grad_norm": 1.0856953823912208, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 21750 + }, + { + "epoch": 0.21751, + "grad_norm": 1.252716455497068, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 21751 + }, + { + "epoch": 0.21752, + "grad_norm": 1.1519951798530181, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 21752 + }, + { + "epoch": 0.21753, + "grad_norm": 0.8606953747458421, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 21753 + }, + { + "epoch": 0.21754, + "grad_norm": 0.8686151890633858, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 21754 + }, + { + "epoch": 0.21755, + "grad_norm": 0.9727797012264362, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 21755 + }, + { + "epoch": 0.21756, + "grad_norm": 1.4292010430720254, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 21756 + }, + { + "epoch": 0.21757, + "grad_norm": 0.899769554891279, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 21757 + }, + { + "epoch": 0.21758, + "grad_norm": 0.9529757323971476, + "learning_rate": 0.003, + "loss": 4.064, + "step": 21758 + }, + { + "epoch": 0.21759, + "grad_norm": 0.9847812642472042, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 21759 + }, + { + "epoch": 0.2176, + "grad_norm": 1.1110315894871545, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 21760 + }, + { + "epoch": 0.21761, + "grad_norm": 0.8753746901201969, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 21761 + }, + { + "epoch": 0.21762, + "grad_norm": 1.0254855603828748, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 21762 + }, + { + "epoch": 0.21763, + "grad_norm": 1.2433844113312553, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 21763 + }, + { + "epoch": 0.21764, + "grad_norm": 0.8522878466068323, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 21764 + }, + { + "epoch": 0.21765, + "grad_norm": 0.9876485683469493, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 21765 + }, + { + "epoch": 0.21766, + "grad_norm": 1.2711575889237838, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 21766 + }, + { + "epoch": 0.21767, + "grad_norm": 0.8005620212517366, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 21767 + }, + { + "epoch": 0.21768, + "grad_norm": 0.9108670808408771, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 21768 + }, + { + "epoch": 0.21769, + "grad_norm": 1.149361614519133, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 21769 + }, + { + "epoch": 0.2177, + "grad_norm": 1.1243365313741949, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 21770 + }, + { + "epoch": 0.21771, + "grad_norm": 1.00932831116816, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 21771 + }, + { + "epoch": 0.21772, + "grad_norm": 1.1692001832235916, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 21772 + }, + { + "epoch": 0.21773, + "grad_norm": 0.9964948394499694, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 21773 + }, + { + "epoch": 0.21774, + "grad_norm": 1.0686058603648438, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 21774 + }, + { + "epoch": 0.21775, + "grad_norm": 1.1009170276952898, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 21775 + }, + { + "epoch": 0.21776, + "grad_norm": 0.9871967899211062, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 21776 + }, + { + "epoch": 0.21777, + "grad_norm": 1.1595699941913098, + "learning_rate": 0.003, + "loss": 4.064, + "step": 21777 + }, + { + "epoch": 0.21778, + "grad_norm": 0.9002972170588126, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 21778 + }, + { + "epoch": 0.21779, + "grad_norm": 0.9940573290646327, + "learning_rate": 0.003, + "loss": 4.035, + "step": 21779 + }, + { + "epoch": 0.2178, + "grad_norm": 1.2090027711629767, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 21780 + }, + { + "epoch": 0.21781, + "grad_norm": 1.0263848800683288, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 21781 + }, + { + "epoch": 0.21782, + "grad_norm": 1.3013206956355017, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 21782 + }, + { + "epoch": 0.21783, + "grad_norm": 1.0137361557875562, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 21783 + }, + { + "epoch": 0.21784, + "grad_norm": 1.1905340905017938, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 21784 + }, + { + "epoch": 0.21785, + "grad_norm": 0.9562384331876861, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 21785 + }, + { + "epoch": 0.21786, + "grad_norm": 0.9836198243622734, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 21786 + }, + { + "epoch": 0.21787, + "grad_norm": 1.0791083461134667, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 21787 + }, + { + "epoch": 0.21788, + "grad_norm": 0.9280518823051347, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 21788 + }, + { + "epoch": 0.21789, + "grad_norm": 0.9389722363183011, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 21789 + }, + { + "epoch": 0.2179, + "grad_norm": 1.1758027190865634, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 21790 + }, + { + "epoch": 0.21791, + "grad_norm": 1.1977566618742412, + "learning_rate": 0.003, + "loss": 4.06, + "step": 21791 + }, + { + "epoch": 0.21792, + "grad_norm": 0.9388120347533434, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 21792 + }, + { + "epoch": 0.21793, + "grad_norm": 0.9877643500577494, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 21793 + }, + { + "epoch": 0.21794, + "grad_norm": 1.2808636400112037, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 21794 + }, + { + "epoch": 0.21795, + "grad_norm": 0.8406508297255363, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 21795 + }, + { + "epoch": 0.21796, + "grad_norm": 0.8954666901145352, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 21796 + }, + { + "epoch": 0.21797, + "grad_norm": 1.1377009624852723, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 21797 + }, + { + "epoch": 0.21798, + "grad_norm": 1.0429225023720603, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 21798 + }, + { + "epoch": 0.21799, + "grad_norm": 1.0816558026905745, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 21799 + }, + { + "epoch": 0.218, + "grad_norm": 1.098307550201317, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 21800 + }, + { + "epoch": 0.21801, + "grad_norm": 1.1469685319586898, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 21801 + }, + { + "epoch": 0.21802, + "grad_norm": 0.9053303424003006, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 21802 + }, + { + "epoch": 0.21803, + "grad_norm": 1.0390771986461802, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 21803 + }, + { + "epoch": 0.21804, + "grad_norm": 1.4116675557850629, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 21804 + }, + { + "epoch": 0.21805, + "grad_norm": 1.1703731728278717, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 21805 + }, + { + "epoch": 0.21806, + "grad_norm": 0.9486357423021219, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 21806 + }, + { + "epoch": 0.21807, + "grad_norm": 1.0073928878597103, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 21807 + }, + { + "epoch": 0.21808, + "grad_norm": 1.0937543963464336, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 21808 + }, + { + "epoch": 0.21809, + "grad_norm": 0.9732932211151658, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 21809 + }, + { + "epoch": 0.2181, + "grad_norm": 1.1198876593185725, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 21810 + }, + { + "epoch": 0.21811, + "grad_norm": 1.0304696938518834, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 21811 + }, + { + "epoch": 0.21812, + "grad_norm": 1.0452556758592917, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 21812 + }, + { + "epoch": 0.21813, + "grad_norm": 1.2006946741732127, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 21813 + }, + { + "epoch": 0.21814, + "grad_norm": 1.1782791625057099, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 21814 + }, + { + "epoch": 0.21815, + "grad_norm": 0.9811177395723667, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 21815 + }, + { + "epoch": 0.21816, + "grad_norm": 1.219324291094324, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 21816 + }, + { + "epoch": 0.21817, + "grad_norm": 0.9201818652894872, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 21817 + }, + { + "epoch": 0.21818, + "grad_norm": 0.9791614104490955, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 21818 + }, + { + "epoch": 0.21819, + "grad_norm": 1.2030028652610103, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 21819 + }, + { + "epoch": 0.2182, + "grad_norm": 0.9819947443032402, + "learning_rate": 0.003, + "loss": 4.054, + "step": 21820 + }, + { + "epoch": 0.21821, + "grad_norm": 1.1881825676829347, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 21821 + }, + { + "epoch": 0.21822, + "grad_norm": 0.9866910765706978, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 21822 + }, + { + "epoch": 0.21823, + "grad_norm": 1.1042249627499925, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 21823 + }, + { + "epoch": 0.21824, + "grad_norm": 1.095270378577289, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 21824 + }, + { + "epoch": 0.21825, + "grad_norm": 1.1184579210578858, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 21825 + }, + { + "epoch": 0.21826, + "grad_norm": 1.4015999232059693, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 21826 + }, + { + "epoch": 0.21827, + "grad_norm": 0.8720645247192133, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 21827 + }, + { + "epoch": 0.21828, + "grad_norm": 0.8332065831433242, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 21828 + }, + { + "epoch": 0.21829, + "grad_norm": 0.9329611279055753, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 21829 + }, + { + "epoch": 0.2183, + "grad_norm": 0.9588285545832566, + "learning_rate": 0.003, + "loss": 4.061, + "step": 21830 + }, + { + "epoch": 0.21831, + "grad_norm": 1.3064167995965914, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 21831 + }, + { + "epoch": 0.21832, + "grad_norm": 0.9359949293489347, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 21832 + }, + { + "epoch": 0.21833, + "grad_norm": 1.0700894916076886, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 21833 + }, + { + "epoch": 0.21834, + "grad_norm": 1.1866254506250609, + "learning_rate": 0.003, + "loss": 4.074, + "step": 21834 + }, + { + "epoch": 0.21835, + "grad_norm": 1.1163459551907222, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 21835 + }, + { + "epoch": 0.21836, + "grad_norm": 0.8988583840757879, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 21836 + }, + { + "epoch": 0.21837, + "grad_norm": 0.94386468463435, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 21837 + }, + { + "epoch": 0.21838, + "grad_norm": 1.1821374735324925, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 21838 + }, + { + "epoch": 0.21839, + "grad_norm": 1.0952448498712766, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 21839 + }, + { + "epoch": 0.2184, + "grad_norm": 1.0879702691209918, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 21840 + }, + { + "epoch": 0.21841, + "grad_norm": 1.0821978394490959, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 21841 + }, + { + "epoch": 0.21842, + "grad_norm": 0.9124459135058113, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 21842 + }, + { + "epoch": 0.21843, + "grad_norm": 0.8403319414289636, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 21843 + }, + { + "epoch": 0.21844, + "grad_norm": 0.9634016956149898, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 21844 + }, + { + "epoch": 0.21845, + "grad_norm": 1.2318622119171379, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 21845 + }, + { + "epoch": 0.21846, + "grad_norm": 1.081906465733042, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 21846 + }, + { + "epoch": 0.21847, + "grad_norm": 0.9901170616070224, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 21847 + }, + { + "epoch": 0.21848, + "grad_norm": 1.0586231605646887, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 21848 + }, + { + "epoch": 0.21849, + "grad_norm": 0.9893857365144102, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 21849 + }, + { + "epoch": 0.2185, + "grad_norm": 1.0990766794671925, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 21850 + }, + { + "epoch": 0.21851, + "grad_norm": 1.1240325543022698, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 21851 + }, + { + "epoch": 0.21852, + "grad_norm": 0.7834189101232806, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 21852 + }, + { + "epoch": 0.21853, + "grad_norm": 0.9727643038092221, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 21853 + }, + { + "epoch": 0.21854, + "grad_norm": 1.1328850047323018, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 21854 + }, + { + "epoch": 0.21855, + "grad_norm": 1.1717335647230558, + "learning_rate": 0.003, + "loss": 4.068, + "step": 21855 + }, + { + "epoch": 0.21856, + "grad_norm": 0.9900892012403634, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 21856 + }, + { + "epoch": 0.21857, + "grad_norm": 1.037216483699412, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 21857 + }, + { + "epoch": 0.21858, + "grad_norm": 1.1159289937307781, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 21858 + }, + { + "epoch": 0.21859, + "grad_norm": 1.2080141684974375, + "learning_rate": 0.003, + "loss": 4.056, + "step": 21859 + }, + { + "epoch": 0.2186, + "grad_norm": 1.025357295483933, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 21860 + }, + { + "epoch": 0.21861, + "grad_norm": 1.2284600727428254, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 21861 + }, + { + "epoch": 0.21862, + "grad_norm": 0.8479547457153721, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 21862 + }, + { + "epoch": 0.21863, + "grad_norm": 0.7995594247542647, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 21863 + }, + { + "epoch": 0.21864, + "grad_norm": 0.8797310619297531, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 21864 + }, + { + "epoch": 0.21865, + "grad_norm": 1.0835624483361808, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 21865 + }, + { + "epoch": 0.21866, + "grad_norm": 1.191645019856226, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 21866 + }, + { + "epoch": 0.21867, + "grad_norm": 1.262882763842499, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 21867 + }, + { + "epoch": 0.21868, + "grad_norm": 1.1387139373471502, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 21868 + }, + { + "epoch": 0.21869, + "grad_norm": 1.046097897161438, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 21869 + }, + { + "epoch": 0.2187, + "grad_norm": 1.1332567192753684, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 21870 + }, + { + "epoch": 0.21871, + "grad_norm": 0.9359379868390781, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 21871 + }, + { + "epoch": 0.21872, + "grad_norm": 0.9519205667873816, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 21872 + }, + { + "epoch": 0.21873, + "grad_norm": 0.9957608355917833, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 21873 + }, + { + "epoch": 0.21874, + "grad_norm": 1.2047717429665121, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 21874 + }, + { + "epoch": 0.21875, + "grad_norm": 1.072638128750961, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 21875 + }, + { + "epoch": 0.21876, + "grad_norm": 1.0575192965996736, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 21876 + }, + { + "epoch": 0.21877, + "grad_norm": 1.2527685139645577, + "learning_rate": 0.003, + "loss": 4.0868, + "step": 21877 + }, + { + "epoch": 0.21878, + "grad_norm": 1.0553827236219049, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 21878 + }, + { + "epoch": 0.21879, + "grad_norm": 1.0398360642102262, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 21879 + }, + { + "epoch": 0.2188, + "grad_norm": 1.033751679437056, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 21880 + }, + { + "epoch": 0.21881, + "grad_norm": 1.211391858985043, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 21881 + }, + { + "epoch": 0.21882, + "grad_norm": 1.0700948245263877, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 21882 + }, + { + "epoch": 0.21883, + "grad_norm": 1.070910445838634, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 21883 + }, + { + "epoch": 0.21884, + "grad_norm": 1.1231575095656516, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 21884 + }, + { + "epoch": 0.21885, + "grad_norm": 1.0796922225601255, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 21885 + }, + { + "epoch": 0.21886, + "grad_norm": 0.932258366301731, + "learning_rate": 0.003, + "loss": 4.1114, + "step": 21886 + }, + { + "epoch": 0.21887, + "grad_norm": 0.8971397584113258, + "learning_rate": 0.003, + "loss": 4.033, + "step": 21887 + }, + { + "epoch": 0.21888, + "grad_norm": 1.073128825409268, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 21888 + }, + { + "epoch": 0.21889, + "grad_norm": 1.026889694896479, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 21889 + }, + { + "epoch": 0.2189, + "grad_norm": 1.0356128184842517, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 21890 + }, + { + "epoch": 0.21891, + "grad_norm": 1.1799709579371358, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 21891 + }, + { + "epoch": 0.21892, + "grad_norm": 1.1444331860703008, + "learning_rate": 0.003, + "loss": 4.0887, + "step": 21892 + }, + { + "epoch": 0.21893, + "grad_norm": 1.0971273030193682, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 21893 + }, + { + "epoch": 0.21894, + "grad_norm": 1.2768457695987183, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 21894 + }, + { + "epoch": 0.21895, + "grad_norm": 0.9336398015111492, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 21895 + }, + { + "epoch": 0.21896, + "grad_norm": 0.8966030724832944, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 21896 + }, + { + "epoch": 0.21897, + "grad_norm": 0.9567604686131651, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 21897 + }, + { + "epoch": 0.21898, + "grad_norm": 1.0880398902088928, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 21898 + }, + { + "epoch": 0.21899, + "grad_norm": 1.0899696534048156, + "learning_rate": 0.003, + "loss": 4.053, + "step": 21899 + }, + { + "epoch": 0.219, + "grad_norm": 0.9867342159508794, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 21900 + }, + { + "epoch": 0.21901, + "grad_norm": 1.07462308445344, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 21901 + }, + { + "epoch": 0.21902, + "grad_norm": 1.2316995956328967, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 21902 + }, + { + "epoch": 0.21903, + "grad_norm": 1.0683787548166255, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 21903 + }, + { + "epoch": 0.21904, + "grad_norm": 0.9952593489962709, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 21904 + }, + { + "epoch": 0.21905, + "grad_norm": 1.0662716244457868, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 21905 + }, + { + "epoch": 0.21906, + "grad_norm": 1.1144114145893405, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 21906 + }, + { + "epoch": 0.21907, + "grad_norm": 1.0217005152701855, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 21907 + }, + { + "epoch": 0.21908, + "grad_norm": 1.3233050960425763, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 21908 + }, + { + "epoch": 0.21909, + "grad_norm": 1.0893850605376774, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 21909 + }, + { + "epoch": 0.2191, + "grad_norm": 0.9159319041648941, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 21910 + }, + { + "epoch": 0.21911, + "grad_norm": 0.9318743809582366, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 21911 + }, + { + "epoch": 0.21912, + "grad_norm": 1.270328701323107, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 21912 + }, + { + "epoch": 0.21913, + "grad_norm": 0.841030188529176, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 21913 + }, + { + "epoch": 0.21914, + "grad_norm": 0.8738673446564376, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 21914 + }, + { + "epoch": 0.21915, + "grad_norm": 1.0436274107151402, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 21915 + }, + { + "epoch": 0.21916, + "grad_norm": 1.2206547990020633, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 21916 + }, + { + "epoch": 0.21917, + "grad_norm": 0.9530228337367064, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 21917 + }, + { + "epoch": 0.21918, + "grad_norm": 0.8058791168378249, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 21918 + }, + { + "epoch": 0.21919, + "grad_norm": 0.9404533257131441, + "learning_rate": 0.003, + "loss": 4.081, + "step": 21919 + }, + { + "epoch": 0.2192, + "grad_norm": 1.1416176076343725, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 21920 + }, + { + "epoch": 0.21921, + "grad_norm": 1.0000932711188928, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 21921 + }, + { + "epoch": 0.21922, + "grad_norm": 0.9884219277399243, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 21922 + }, + { + "epoch": 0.21923, + "grad_norm": 1.0019106430896854, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 21923 + }, + { + "epoch": 0.21924, + "grad_norm": 1.1779470954228424, + "learning_rate": 0.003, + "loss": 4.0948, + "step": 21924 + }, + { + "epoch": 0.21925, + "grad_norm": 0.9484038810332698, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 21925 + }, + { + "epoch": 0.21926, + "grad_norm": 1.245161319043485, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 21926 + }, + { + "epoch": 0.21927, + "grad_norm": 1.021604189150329, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 21927 + }, + { + "epoch": 0.21928, + "grad_norm": 1.076509974664768, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 21928 + }, + { + "epoch": 0.21929, + "grad_norm": 1.091488257309248, + "learning_rate": 0.003, + "loss": 4.087, + "step": 21929 + }, + { + "epoch": 0.2193, + "grad_norm": 0.8807705978304886, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 21930 + }, + { + "epoch": 0.21931, + "grad_norm": 0.9341518058192211, + "learning_rate": 0.003, + "loss": 4.062, + "step": 21931 + }, + { + "epoch": 0.21932, + "grad_norm": 0.9010381355284397, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 21932 + }, + { + "epoch": 0.21933, + "grad_norm": 0.9692510706118285, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 21933 + }, + { + "epoch": 0.21934, + "grad_norm": 1.2351411675773536, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 21934 + }, + { + "epoch": 0.21935, + "grad_norm": 1.043415650558395, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 21935 + }, + { + "epoch": 0.21936, + "grad_norm": 1.2176520030096332, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 21936 + }, + { + "epoch": 0.21937, + "grad_norm": 1.0028357600006315, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 21937 + }, + { + "epoch": 0.21938, + "grad_norm": 1.2816731261739953, + "learning_rate": 0.003, + "loss": 4.1075, + "step": 21938 + }, + { + "epoch": 0.21939, + "grad_norm": 0.9046620945754946, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 21939 + }, + { + "epoch": 0.2194, + "grad_norm": 1.102778859662781, + "learning_rate": 0.003, + "loss": 4.071, + "step": 21940 + }, + { + "epoch": 0.21941, + "grad_norm": 1.1462819016387535, + "learning_rate": 0.003, + "loss": 4.1036, + "step": 21941 + }, + { + "epoch": 0.21942, + "grad_norm": 1.2411195541512452, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 21942 + }, + { + "epoch": 0.21943, + "grad_norm": 1.1345942284858206, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 21943 + }, + { + "epoch": 0.21944, + "grad_norm": 1.204951810633587, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 21944 + }, + { + "epoch": 0.21945, + "grad_norm": 1.010178487940068, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 21945 + }, + { + "epoch": 0.21946, + "grad_norm": 1.0508680298965007, + "learning_rate": 0.003, + "loss": 4.047, + "step": 21946 + }, + { + "epoch": 0.21947, + "grad_norm": 0.9521758475266423, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 21947 + }, + { + "epoch": 0.21948, + "grad_norm": 1.1176565842033925, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 21948 + }, + { + "epoch": 0.21949, + "grad_norm": 1.0190680720696423, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 21949 + }, + { + "epoch": 0.2195, + "grad_norm": 1.0984134032214858, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 21950 + }, + { + "epoch": 0.21951, + "grad_norm": 0.9979701374897622, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 21951 + }, + { + "epoch": 0.21952, + "grad_norm": 1.1977665887738582, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 21952 + }, + { + "epoch": 0.21953, + "grad_norm": 1.1071435941928778, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 21953 + }, + { + "epoch": 0.21954, + "grad_norm": 1.3614050092598842, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 21954 + }, + { + "epoch": 0.21955, + "grad_norm": 0.850510648426961, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 21955 + }, + { + "epoch": 0.21956, + "grad_norm": 1.0031907473899282, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 21956 + }, + { + "epoch": 0.21957, + "grad_norm": 1.3319128978015278, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 21957 + }, + { + "epoch": 0.21958, + "grad_norm": 0.9660190362531299, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 21958 + }, + { + "epoch": 0.21959, + "grad_norm": 0.9508574073555329, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 21959 + }, + { + "epoch": 0.2196, + "grad_norm": 1.1086189463228375, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 21960 + }, + { + "epoch": 0.21961, + "grad_norm": 0.9769180184251998, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 21961 + }, + { + "epoch": 0.21962, + "grad_norm": 1.0607616702014842, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 21962 + }, + { + "epoch": 0.21963, + "grad_norm": 0.9362030192506503, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 21963 + }, + { + "epoch": 0.21964, + "grad_norm": 0.9799303119822954, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 21964 + }, + { + "epoch": 0.21965, + "grad_norm": 1.2288423866023974, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 21965 + }, + { + "epoch": 0.21966, + "grad_norm": 0.934793448219046, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 21966 + }, + { + "epoch": 0.21967, + "grad_norm": 1.0067134585299555, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 21967 + }, + { + "epoch": 0.21968, + "grad_norm": 1.009410847476633, + "learning_rate": 0.003, + "loss": 4.042, + "step": 21968 + }, + { + "epoch": 0.21969, + "grad_norm": 1.1812845219372061, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 21969 + }, + { + "epoch": 0.2197, + "grad_norm": 0.9414780331347194, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 21970 + }, + { + "epoch": 0.21971, + "grad_norm": 0.9362032534450997, + "learning_rate": 0.003, + "loss": 4.059, + "step": 21971 + }, + { + "epoch": 0.21972, + "grad_norm": 1.1257118513004098, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 21972 + }, + { + "epoch": 0.21973, + "grad_norm": 1.1244191409740334, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 21973 + }, + { + "epoch": 0.21974, + "grad_norm": 1.188052325411122, + "learning_rate": 0.003, + "loss": 4.0997, + "step": 21974 + }, + { + "epoch": 0.21975, + "grad_norm": 0.9559467710113063, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 21975 + }, + { + "epoch": 0.21976, + "grad_norm": 0.9453077824015989, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 21976 + }, + { + "epoch": 0.21977, + "grad_norm": 1.1005409211573578, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 21977 + }, + { + "epoch": 0.21978, + "grad_norm": 1.0664288599909775, + "learning_rate": 0.003, + "loss": 4.0942, + "step": 21978 + }, + { + "epoch": 0.21979, + "grad_norm": 1.2982316675694832, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 21979 + }, + { + "epoch": 0.2198, + "grad_norm": 0.9755869008909936, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 21980 + }, + { + "epoch": 0.21981, + "grad_norm": 1.129345102739605, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 21981 + }, + { + "epoch": 0.21982, + "grad_norm": 0.9104254283929122, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 21982 + }, + { + "epoch": 0.21983, + "grad_norm": 1.0912337132808585, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 21983 + }, + { + "epoch": 0.21984, + "grad_norm": 1.1281763534152205, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 21984 + }, + { + "epoch": 0.21985, + "grad_norm": 0.9013075838936454, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 21985 + }, + { + "epoch": 0.21986, + "grad_norm": 1.185858959113043, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 21986 + }, + { + "epoch": 0.21987, + "grad_norm": 1.4231211787508016, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 21987 + }, + { + "epoch": 0.21988, + "grad_norm": 0.9150059638472299, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 21988 + }, + { + "epoch": 0.21989, + "grad_norm": 0.9010900289857722, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 21989 + }, + { + "epoch": 0.2199, + "grad_norm": 0.9767844263290876, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 21990 + }, + { + "epoch": 0.21991, + "grad_norm": 1.2705762039255306, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 21991 + }, + { + "epoch": 0.21992, + "grad_norm": 1.0297154985831858, + "learning_rate": 0.003, + "loss": 4.044, + "step": 21992 + }, + { + "epoch": 0.21993, + "grad_norm": 1.1503954503709115, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 21993 + }, + { + "epoch": 0.21994, + "grad_norm": 1.1025741802312625, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 21994 + }, + { + "epoch": 0.21995, + "grad_norm": 1.0484167355275913, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 21995 + }, + { + "epoch": 0.21996, + "grad_norm": 0.9498797202069863, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 21996 + }, + { + "epoch": 0.21997, + "grad_norm": 1.1377179562176651, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 21997 + }, + { + "epoch": 0.21998, + "grad_norm": 0.9740361014226149, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 21998 + }, + { + "epoch": 0.21999, + "grad_norm": 1.0888309851110627, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 21999 + }, + { + "epoch": 0.22, + "grad_norm": 1.2301028946185568, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 22000 + }, + { + "epoch": 0.22001, + "grad_norm": 0.8799155276212549, + "learning_rate": 0.003, + "loss": 4.056, + "step": 22001 + }, + { + "epoch": 0.22002, + "grad_norm": 0.7536568082737142, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 22002 + }, + { + "epoch": 0.22003, + "grad_norm": 0.841781976690513, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 22003 + }, + { + "epoch": 0.22004, + "grad_norm": 0.8636001273108609, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 22004 + }, + { + "epoch": 0.22005, + "grad_norm": 0.8919707730300951, + "learning_rate": 0.003, + "loss": 4.064, + "step": 22005 + }, + { + "epoch": 0.22006, + "grad_norm": 1.2209789268563258, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 22006 + }, + { + "epoch": 0.22007, + "grad_norm": 0.9676408914557496, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 22007 + }, + { + "epoch": 0.22008, + "grad_norm": 1.1138212017099323, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 22008 + }, + { + "epoch": 0.22009, + "grad_norm": 1.282271991669187, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 22009 + }, + { + "epoch": 0.2201, + "grad_norm": 0.9979071274808283, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 22010 + }, + { + "epoch": 0.22011, + "grad_norm": 0.9300007565963104, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 22011 + }, + { + "epoch": 0.22012, + "grad_norm": 0.8476015526076758, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 22012 + }, + { + "epoch": 0.22013, + "grad_norm": 0.8784403928589986, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 22013 + }, + { + "epoch": 0.22014, + "grad_norm": 0.9036578507805185, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 22014 + }, + { + "epoch": 0.22015, + "grad_norm": 1.055852152749332, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 22015 + }, + { + "epoch": 0.22016, + "grad_norm": 1.0790358016941337, + "learning_rate": 0.003, + "loss": 4.068, + "step": 22016 + }, + { + "epoch": 0.22017, + "grad_norm": 1.33260034638369, + "learning_rate": 0.003, + "loss": 4.064, + "step": 22017 + }, + { + "epoch": 0.22018, + "grad_norm": 0.9291850881983121, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 22018 + }, + { + "epoch": 0.22019, + "grad_norm": 0.91359832751277, + "learning_rate": 0.003, + "loss": 4.044, + "step": 22019 + }, + { + "epoch": 0.2202, + "grad_norm": 1.1837805395593957, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 22020 + }, + { + "epoch": 0.22021, + "grad_norm": 1.0242600779885618, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 22021 + }, + { + "epoch": 0.22022, + "grad_norm": 1.065351159628495, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 22022 + }, + { + "epoch": 0.22023, + "grad_norm": 1.171636259450973, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 22023 + }, + { + "epoch": 0.22024, + "grad_norm": 1.0460373718969016, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 22024 + }, + { + "epoch": 0.22025, + "grad_norm": 1.2114706322052353, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 22025 + }, + { + "epoch": 0.22026, + "grad_norm": 1.1388287249194442, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 22026 + }, + { + "epoch": 0.22027, + "grad_norm": 1.1609272270032163, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 22027 + }, + { + "epoch": 0.22028, + "grad_norm": 1.2840317656816451, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 22028 + }, + { + "epoch": 0.22029, + "grad_norm": 0.9734592517757922, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 22029 + }, + { + "epoch": 0.2203, + "grad_norm": 1.1144137332746902, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 22030 + }, + { + "epoch": 0.22031, + "grad_norm": 1.374195411026943, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 22031 + }, + { + "epoch": 0.22032, + "grad_norm": 1.1383458637295605, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 22032 + }, + { + "epoch": 0.22033, + "grad_norm": 0.9429598884147194, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 22033 + }, + { + "epoch": 0.22034, + "grad_norm": 0.989385253905555, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 22034 + }, + { + "epoch": 0.22035, + "grad_norm": 1.2136774995035935, + "learning_rate": 0.003, + "loss": 4.078, + "step": 22035 + }, + { + "epoch": 0.22036, + "grad_norm": 0.9773462832019811, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 22036 + }, + { + "epoch": 0.22037, + "grad_norm": 1.2113145133796195, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 22037 + }, + { + "epoch": 0.22038, + "grad_norm": 1.2654501878230824, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 22038 + }, + { + "epoch": 0.22039, + "grad_norm": 0.7913835431537962, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 22039 + }, + { + "epoch": 0.2204, + "grad_norm": 0.7607424465447554, + "learning_rate": 0.003, + "loss": 4.065, + "step": 22040 + }, + { + "epoch": 0.22041, + "grad_norm": 0.749496327382705, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 22041 + }, + { + "epoch": 0.22042, + "grad_norm": 0.9451128832281389, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 22042 + }, + { + "epoch": 0.22043, + "grad_norm": 1.2543569233620557, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 22043 + }, + { + "epoch": 0.22044, + "grad_norm": 0.9651959776062369, + "learning_rate": 0.003, + "loss": 4.053, + "step": 22044 + }, + { + "epoch": 0.22045, + "grad_norm": 1.1205428873858805, + "learning_rate": 0.003, + "loss": 4.069, + "step": 22045 + }, + { + "epoch": 0.22046, + "grad_norm": 0.91608415763565, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 22046 + }, + { + "epoch": 0.22047, + "grad_norm": 0.9283991486435444, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 22047 + }, + { + "epoch": 0.22048, + "grad_norm": 1.1125520136210267, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 22048 + }, + { + "epoch": 0.22049, + "grad_norm": 1.0523580570135502, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 22049 + }, + { + "epoch": 0.2205, + "grad_norm": 1.0623282185308198, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 22050 + }, + { + "epoch": 0.22051, + "grad_norm": 1.017634091946446, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 22051 + }, + { + "epoch": 0.22052, + "grad_norm": 1.3714484834691925, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 22052 + }, + { + "epoch": 0.22053, + "grad_norm": 0.9468871897030142, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 22053 + }, + { + "epoch": 0.22054, + "grad_norm": 0.8894667078161215, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 22054 + }, + { + "epoch": 0.22055, + "grad_norm": 0.9538163847813176, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 22055 + }, + { + "epoch": 0.22056, + "grad_norm": 1.0353390367082436, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 22056 + }, + { + "epoch": 0.22057, + "grad_norm": 1.1036838703570486, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 22057 + }, + { + "epoch": 0.22058, + "grad_norm": 1.277674403540984, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 22058 + }, + { + "epoch": 0.22059, + "grad_norm": 0.9854092153645654, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 22059 + }, + { + "epoch": 0.2206, + "grad_norm": 1.1004045386916625, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 22060 + }, + { + "epoch": 0.22061, + "grad_norm": 1.0291920282220117, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 22061 + }, + { + "epoch": 0.22062, + "grad_norm": 1.214074662207699, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 22062 + }, + { + "epoch": 0.22063, + "grad_norm": 0.8852685767847551, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 22063 + }, + { + "epoch": 0.22064, + "grad_norm": 0.9802746247281702, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 22064 + }, + { + "epoch": 0.22065, + "grad_norm": 0.9934708265727853, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 22065 + }, + { + "epoch": 0.22066, + "grad_norm": 1.2323966704941063, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 22066 + }, + { + "epoch": 0.22067, + "grad_norm": 1.0300687053312463, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 22067 + }, + { + "epoch": 0.22068, + "grad_norm": 0.9144849864751317, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 22068 + }, + { + "epoch": 0.22069, + "grad_norm": 1.2037138052878606, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 22069 + }, + { + "epoch": 0.2207, + "grad_norm": 1.2970163374889876, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 22070 + }, + { + "epoch": 0.22071, + "grad_norm": 1.1930113787598, + "learning_rate": 0.003, + "loss": 4.08, + "step": 22071 + }, + { + "epoch": 0.22072, + "grad_norm": 1.0324809136029667, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 22072 + }, + { + "epoch": 0.22073, + "grad_norm": 0.9618610198569468, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 22073 + }, + { + "epoch": 0.22074, + "grad_norm": 1.1117763440631725, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 22074 + }, + { + "epoch": 0.22075, + "grad_norm": 0.9534078961059544, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 22075 + }, + { + "epoch": 0.22076, + "grad_norm": 1.074647776629819, + "learning_rate": 0.003, + "loss": 4.094, + "step": 22076 + }, + { + "epoch": 0.22077, + "grad_norm": 1.2105334444328661, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 22077 + }, + { + "epoch": 0.22078, + "grad_norm": 1.0812382141502532, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 22078 + }, + { + "epoch": 0.22079, + "grad_norm": 1.207690563712945, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 22079 + }, + { + "epoch": 0.2208, + "grad_norm": 0.8965593104761895, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 22080 + }, + { + "epoch": 0.22081, + "grad_norm": 1.215903667442995, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 22081 + }, + { + "epoch": 0.22082, + "grad_norm": 1.193401299376855, + "learning_rate": 0.003, + "loss": 4.068, + "step": 22082 + }, + { + "epoch": 0.22083, + "grad_norm": 0.9513605021626332, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 22083 + }, + { + "epoch": 0.22084, + "grad_norm": 0.9648708327678751, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 22084 + }, + { + "epoch": 0.22085, + "grad_norm": 1.0755706954615447, + "learning_rate": 0.003, + "loss": 4.063, + "step": 22085 + }, + { + "epoch": 0.22086, + "grad_norm": 1.0533901142482172, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 22086 + }, + { + "epoch": 0.22087, + "grad_norm": 0.8739382724093631, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 22087 + }, + { + "epoch": 0.22088, + "grad_norm": 0.9719033823098067, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 22088 + }, + { + "epoch": 0.22089, + "grad_norm": 1.4176208398032975, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 22089 + }, + { + "epoch": 0.2209, + "grad_norm": 1.1370916350107647, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 22090 + }, + { + "epoch": 0.22091, + "grad_norm": 1.018806059327387, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 22091 + }, + { + "epoch": 0.22092, + "grad_norm": 1.1319337515761905, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 22092 + }, + { + "epoch": 0.22093, + "grad_norm": 0.9072209885155598, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 22093 + }, + { + "epoch": 0.22094, + "grad_norm": 0.9618062804516794, + "learning_rate": 0.003, + "loss": 4.046, + "step": 22094 + }, + { + "epoch": 0.22095, + "grad_norm": 1.0014023951291084, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 22095 + }, + { + "epoch": 0.22096, + "grad_norm": 1.093008799080337, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 22096 + }, + { + "epoch": 0.22097, + "grad_norm": 1.2142325458275705, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 22097 + }, + { + "epoch": 0.22098, + "grad_norm": 0.9164875653960181, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 22098 + }, + { + "epoch": 0.22099, + "grad_norm": 0.9929284392731871, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 22099 + }, + { + "epoch": 0.221, + "grad_norm": 1.2515226215425053, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 22100 + }, + { + "epoch": 0.22101, + "grad_norm": 0.9474753532694903, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 22101 + }, + { + "epoch": 0.22102, + "grad_norm": 1.1723768185795456, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 22102 + }, + { + "epoch": 0.22103, + "grad_norm": 1.0252875131272121, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 22103 + }, + { + "epoch": 0.22104, + "grad_norm": 1.2660031060462824, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 22104 + }, + { + "epoch": 0.22105, + "grad_norm": 0.870103545303669, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 22105 + }, + { + "epoch": 0.22106, + "grad_norm": 1.0716096270885103, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 22106 + }, + { + "epoch": 0.22107, + "grad_norm": 1.2709848565058195, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 22107 + }, + { + "epoch": 0.22108, + "grad_norm": 0.9917211318511614, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 22108 + }, + { + "epoch": 0.22109, + "grad_norm": 1.1559317738982766, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 22109 + }, + { + "epoch": 0.2211, + "grad_norm": 1.0391426850612342, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 22110 + }, + { + "epoch": 0.22111, + "grad_norm": 1.101169253343214, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 22111 + }, + { + "epoch": 0.22112, + "grad_norm": 1.1054563931415329, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 22112 + }, + { + "epoch": 0.22113, + "grad_norm": 1.1714510457196528, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 22113 + }, + { + "epoch": 0.22114, + "grad_norm": 1.0841264988932064, + "learning_rate": 0.003, + "loss": 4.051, + "step": 22114 + }, + { + "epoch": 0.22115, + "grad_norm": 1.1678816358557755, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 22115 + }, + { + "epoch": 0.22116, + "grad_norm": 1.065119219155598, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 22116 + }, + { + "epoch": 0.22117, + "grad_norm": 0.8720654285597286, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 22117 + }, + { + "epoch": 0.22118, + "grad_norm": 0.8574511637676461, + "learning_rate": 0.003, + "loss": 4.0899, + "step": 22118 + }, + { + "epoch": 0.22119, + "grad_norm": 0.8868760394299372, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 22119 + }, + { + "epoch": 0.2212, + "grad_norm": 1.1076779455034, + "learning_rate": 0.003, + "loss": 4.077, + "step": 22120 + }, + { + "epoch": 0.22121, + "grad_norm": 1.0373981425938772, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 22121 + }, + { + "epoch": 0.22122, + "grad_norm": 0.9778246745290635, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 22122 + }, + { + "epoch": 0.22123, + "grad_norm": 1.23260040256254, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 22123 + }, + { + "epoch": 0.22124, + "grad_norm": 1.2034149429618017, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 22124 + }, + { + "epoch": 0.22125, + "grad_norm": 0.9779936712133855, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 22125 + }, + { + "epoch": 0.22126, + "grad_norm": 1.1185993434114172, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 22126 + }, + { + "epoch": 0.22127, + "grad_norm": 1.2085949359982222, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 22127 + }, + { + "epoch": 0.22128, + "grad_norm": 1.0272110794263585, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 22128 + }, + { + "epoch": 0.22129, + "grad_norm": 1.1034149653405565, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 22129 + }, + { + "epoch": 0.2213, + "grad_norm": 1.0831822905163915, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 22130 + }, + { + "epoch": 0.22131, + "grad_norm": 1.0478732402940512, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 22131 + }, + { + "epoch": 0.22132, + "grad_norm": 0.9739849574769375, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 22132 + }, + { + "epoch": 0.22133, + "grad_norm": 1.1866951608439855, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 22133 + }, + { + "epoch": 0.22134, + "grad_norm": 0.9441966258269561, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 22134 + }, + { + "epoch": 0.22135, + "grad_norm": 1.0899329424259943, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 22135 + }, + { + "epoch": 0.22136, + "grad_norm": 1.118187767806871, + "learning_rate": 0.003, + "loss": 4.066, + "step": 22136 + }, + { + "epoch": 0.22137, + "grad_norm": 1.0687823958746339, + "learning_rate": 0.003, + "loss": 4.058, + "step": 22137 + }, + { + "epoch": 0.22138, + "grad_norm": 0.9709383538556655, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 22138 + }, + { + "epoch": 0.22139, + "grad_norm": 1.0886919275445215, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 22139 + }, + { + "epoch": 0.2214, + "grad_norm": 0.9084822887283244, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 22140 + }, + { + "epoch": 0.22141, + "grad_norm": 0.8855526022681398, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 22141 + }, + { + "epoch": 0.22142, + "grad_norm": 1.1679419478119977, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 22142 + }, + { + "epoch": 0.22143, + "grad_norm": 1.0464690487687005, + "learning_rate": 0.003, + "loss": 4.071, + "step": 22143 + }, + { + "epoch": 0.22144, + "grad_norm": 1.2378629301849458, + "learning_rate": 0.003, + "loss": 4.058, + "step": 22144 + }, + { + "epoch": 0.22145, + "grad_norm": 1.1193998692303972, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 22145 + }, + { + "epoch": 0.22146, + "grad_norm": 0.9543361067364833, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 22146 + }, + { + "epoch": 0.22147, + "grad_norm": 1.0526398276608397, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 22147 + }, + { + "epoch": 0.22148, + "grad_norm": 1.0521255374886898, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 22148 + }, + { + "epoch": 0.22149, + "grad_norm": 1.2282808699043781, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 22149 + }, + { + "epoch": 0.2215, + "grad_norm": 1.0528009657825599, + "learning_rate": 0.003, + "loss": 4.021, + "step": 22150 + }, + { + "epoch": 0.22151, + "grad_norm": 1.2021158535757976, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 22151 + }, + { + "epoch": 0.22152, + "grad_norm": 0.9878931132245843, + "learning_rate": 0.003, + "loss": 4.0927, + "step": 22152 + }, + { + "epoch": 0.22153, + "grad_norm": 1.0954229815060532, + "learning_rate": 0.003, + "loss": 4.05, + "step": 22153 + }, + { + "epoch": 0.22154, + "grad_norm": 1.2465107826698634, + "learning_rate": 0.003, + "loss": 4.061, + "step": 22154 + }, + { + "epoch": 0.22155, + "grad_norm": 0.9879143708250671, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 22155 + }, + { + "epoch": 0.22156, + "grad_norm": 1.2570372661673526, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 22156 + }, + { + "epoch": 0.22157, + "grad_norm": 1.1306391380901502, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 22157 + }, + { + "epoch": 0.22158, + "grad_norm": 1.1271171509079259, + "learning_rate": 0.003, + "loss": 4.0893, + "step": 22158 + }, + { + "epoch": 0.22159, + "grad_norm": 1.220559181378633, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 22159 + }, + { + "epoch": 0.2216, + "grad_norm": 0.92067687821397, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 22160 + }, + { + "epoch": 0.22161, + "grad_norm": 0.9026544881233186, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 22161 + }, + { + "epoch": 0.22162, + "grad_norm": 1.1296221792021035, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 22162 + }, + { + "epoch": 0.22163, + "grad_norm": 1.3522348404098625, + "learning_rate": 0.003, + "loss": 4.088, + "step": 22163 + }, + { + "epoch": 0.22164, + "grad_norm": 0.8873495878638926, + "learning_rate": 0.003, + "loss": 4.059, + "step": 22164 + }, + { + "epoch": 0.22165, + "grad_norm": 0.9750667702086907, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 22165 + }, + { + "epoch": 0.22166, + "grad_norm": 1.0099678512291974, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 22166 + }, + { + "epoch": 0.22167, + "grad_norm": 1.015849175037569, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 22167 + }, + { + "epoch": 0.22168, + "grad_norm": 1.349760808589888, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 22168 + }, + { + "epoch": 0.22169, + "grad_norm": 0.928959179729019, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 22169 + }, + { + "epoch": 0.2217, + "grad_norm": 1.0753165587349351, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 22170 + }, + { + "epoch": 0.22171, + "grad_norm": 1.2318499720224279, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 22171 + }, + { + "epoch": 0.22172, + "grad_norm": 0.9841668160259697, + "learning_rate": 0.003, + "loss": 4.066, + "step": 22172 + }, + { + "epoch": 0.22173, + "grad_norm": 0.9207807122194273, + "learning_rate": 0.003, + "loss": 4.047, + "step": 22173 + }, + { + "epoch": 0.22174, + "grad_norm": 0.8745437163656563, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 22174 + }, + { + "epoch": 0.22175, + "grad_norm": 0.9715681627316696, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 22175 + }, + { + "epoch": 0.22176, + "grad_norm": 1.2810837206270642, + "learning_rate": 0.003, + "loss": 4.0983, + "step": 22176 + }, + { + "epoch": 0.22177, + "grad_norm": 0.9741092495938862, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 22177 + }, + { + "epoch": 0.22178, + "grad_norm": 1.000083159670246, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 22178 + }, + { + "epoch": 0.22179, + "grad_norm": 1.088885611303467, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 22179 + }, + { + "epoch": 0.2218, + "grad_norm": 1.2111988441798596, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 22180 + }, + { + "epoch": 0.22181, + "grad_norm": 1.0788850241368066, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 22181 + }, + { + "epoch": 0.22182, + "grad_norm": 0.9718929925158949, + "learning_rate": 0.003, + "loss": 4.052, + "step": 22182 + }, + { + "epoch": 0.22183, + "grad_norm": 0.9806067141863312, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 22183 + }, + { + "epoch": 0.22184, + "grad_norm": 1.3852412433343637, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 22184 + }, + { + "epoch": 0.22185, + "grad_norm": 0.9482231102204444, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 22185 + }, + { + "epoch": 0.22186, + "grad_norm": 1.1426081635612542, + "learning_rate": 0.003, + "loss": 4.073, + "step": 22186 + }, + { + "epoch": 0.22187, + "grad_norm": 0.9876760318472187, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 22187 + }, + { + "epoch": 0.22188, + "grad_norm": 1.0469305998802174, + "learning_rate": 0.003, + "loss": 4.0942, + "step": 22188 + }, + { + "epoch": 0.22189, + "grad_norm": 0.9843666034141816, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 22189 + }, + { + "epoch": 0.2219, + "grad_norm": 0.9185483788703165, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 22190 + }, + { + "epoch": 0.22191, + "grad_norm": 1.0714665284585694, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 22191 + }, + { + "epoch": 0.22192, + "grad_norm": 1.1700629033916932, + "learning_rate": 0.003, + "loss": 4.052, + "step": 22192 + }, + { + "epoch": 0.22193, + "grad_norm": 1.2604926970679071, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 22193 + }, + { + "epoch": 0.22194, + "grad_norm": 0.9783307228390339, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 22194 + }, + { + "epoch": 0.22195, + "grad_norm": 1.0398001120082274, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 22195 + }, + { + "epoch": 0.22196, + "grad_norm": 1.1106954852611237, + "learning_rate": 0.003, + "loss": 4.081, + "step": 22196 + }, + { + "epoch": 0.22197, + "grad_norm": 0.9903478136796539, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 22197 + }, + { + "epoch": 0.22198, + "grad_norm": 1.0529405848432538, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 22198 + }, + { + "epoch": 0.22199, + "grad_norm": 1.14778677133365, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 22199 + }, + { + "epoch": 0.222, + "grad_norm": 0.9270817393412104, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 22200 + }, + { + "epoch": 0.22201, + "grad_norm": 1.0746954255107017, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 22201 + }, + { + "epoch": 0.22202, + "grad_norm": 1.1562773284163492, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 22202 + }, + { + "epoch": 0.22203, + "grad_norm": 0.9448377773408985, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 22203 + }, + { + "epoch": 0.22204, + "grad_norm": 1.0497852571998043, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 22204 + }, + { + "epoch": 0.22205, + "grad_norm": 1.0327735568660665, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 22205 + }, + { + "epoch": 0.22206, + "grad_norm": 1.169133875469224, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 22206 + }, + { + "epoch": 0.22207, + "grad_norm": 0.9696334574500572, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 22207 + }, + { + "epoch": 0.22208, + "grad_norm": 1.0637937465433778, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 22208 + }, + { + "epoch": 0.22209, + "grad_norm": 1.1412367172357876, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 22209 + }, + { + "epoch": 0.2221, + "grad_norm": 0.8700889565652512, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 22210 + }, + { + "epoch": 0.22211, + "grad_norm": 0.8373796950086766, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 22211 + }, + { + "epoch": 0.22212, + "grad_norm": 0.8295257155305376, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 22212 + }, + { + "epoch": 0.22213, + "grad_norm": 0.9803956096553748, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 22213 + }, + { + "epoch": 0.22214, + "grad_norm": 1.522171043812501, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 22214 + }, + { + "epoch": 0.22215, + "grad_norm": 0.6856217982491738, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 22215 + }, + { + "epoch": 0.22216, + "grad_norm": 0.9347371465208557, + "learning_rate": 0.003, + "loss": 4.063, + "step": 22216 + }, + { + "epoch": 0.22217, + "grad_norm": 1.4206568110411346, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 22217 + }, + { + "epoch": 0.22218, + "grad_norm": 0.7526040613236351, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 22218 + }, + { + "epoch": 0.22219, + "grad_norm": 0.7739579709255766, + "learning_rate": 0.003, + "loss": 4.03, + "step": 22219 + }, + { + "epoch": 0.2222, + "grad_norm": 1.04848508661123, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 22220 + }, + { + "epoch": 0.22221, + "grad_norm": 1.3844048808563134, + "learning_rate": 0.003, + "loss": 4.085, + "step": 22221 + }, + { + "epoch": 0.22222, + "grad_norm": 1.0919482963174638, + "learning_rate": 0.003, + "loss": 4.0819, + "step": 22222 + }, + { + "epoch": 0.22223, + "grad_norm": 1.3298762462623102, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 22223 + }, + { + "epoch": 0.22224, + "grad_norm": 1.0577158930357622, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 22224 + }, + { + "epoch": 0.22225, + "grad_norm": 0.973627136925897, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 22225 + }, + { + "epoch": 0.22226, + "grad_norm": 1.2068539151590618, + "learning_rate": 0.003, + "loss": 4.05, + "step": 22226 + }, + { + "epoch": 0.22227, + "grad_norm": 1.1041444557014708, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 22227 + }, + { + "epoch": 0.22228, + "grad_norm": 1.1192921252923778, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 22228 + }, + { + "epoch": 0.22229, + "grad_norm": 0.8886626914983616, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 22229 + }, + { + "epoch": 0.2223, + "grad_norm": 0.9956554355722442, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 22230 + }, + { + "epoch": 0.22231, + "grad_norm": 1.217867687762391, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 22231 + }, + { + "epoch": 0.22232, + "grad_norm": 1.1099984915228862, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 22232 + }, + { + "epoch": 0.22233, + "grad_norm": 1.1036584829699105, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 22233 + }, + { + "epoch": 0.22234, + "grad_norm": 1.0771420175924245, + "learning_rate": 0.003, + "loss": 4.068, + "step": 22234 + }, + { + "epoch": 0.22235, + "grad_norm": 1.146500475569834, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 22235 + }, + { + "epoch": 0.22236, + "grad_norm": 1.140022365078969, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 22236 + }, + { + "epoch": 0.22237, + "grad_norm": 1.1714603888028337, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 22237 + }, + { + "epoch": 0.22238, + "grad_norm": 1.0647876960178806, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 22238 + }, + { + "epoch": 0.22239, + "grad_norm": 1.1327015801396911, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 22239 + }, + { + "epoch": 0.2224, + "grad_norm": 1.1129439571422857, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 22240 + }, + { + "epoch": 0.22241, + "grad_norm": 0.9670846330070334, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 22241 + }, + { + "epoch": 0.22242, + "grad_norm": 1.0549919348434558, + "learning_rate": 0.003, + "loss": 4.036, + "step": 22242 + }, + { + "epoch": 0.22243, + "grad_norm": 1.0255196479253745, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 22243 + }, + { + "epoch": 0.22244, + "grad_norm": 1.0796516939290839, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 22244 + }, + { + "epoch": 0.22245, + "grad_norm": 0.9224334755756555, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 22245 + }, + { + "epoch": 0.22246, + "grad_norm": 0.9368787107832641, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 22246 + }, + { + "epoch": 0.22247, + "grad_norm": 1.0089142754546816, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 22247 + }, + { + "epoch": 0.22248, + "grad_norm": 1.2179783874190009, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 22248 + }, + { + "epoch": 0.22249, + "grad_norm": 1.0181180032131043, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 22249 + }, + { + "epoch": 0.2225, + "grad_norm": 1.1939081056570935, + "learning_rate": 0.003, + "loss": 4.044, + "step": 22250 + }, + { + "epoch": 0.22251, + "grad_norm": 0.9339562029276065, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 22251 + }, + { + "epoch": 0.22252, + "grad_norm": 1.035871234019731, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 22252 + }, + { + "epoch": 0.22253, + "grad_norm": 1.4761426152942385, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 22253 + }, + { + "epoch": 0.22254, + "grad_norm": 1.0737039095820475, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 22254 + }, + { + "epoch": 0.22255, + "grad_norm": 0.9851708218452359, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 22255 + }, + { + "epoch": 0.22256, + "grad_norm": 1.0251058339406423, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 22256 + }, + { + "epoch": 0.22257, + "grad_norm": 1.1767205351238077, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 22257 + }, + { + "epoch": 0.22258, + "grad_norm": 0.896738964461162, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 22258 + }, + { + "epoch": 0.22259, + "grad_norm": 1.1022929945027633, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 22259 + }, + { + "epoch": 0.2226, + "grad_norm": 0.9844543134852343, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 22260 + }, + { + "epoch": 0.22261, + "grad_norm": 1.0518559418681108, + "learning_rate": 0.003, + "loss": 4.07, + "step": 22261 + }, + { + "epoch": 0.22262, + "grad_norm": 1.0164051541138535, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 22262 + }, + { + "epoch": 0.22263, + "grad_norm": 1.1576992167140285, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 22263 + }, + { + "epoch": 0.22264, + "grad_norm": 0.9495067271264949, + "learning_rate": 0.003, + "loss": 4.069, + "step": 22264 + }, + { + "epoch": 0.22265, + "grad_norm": 1.0999364514019112, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 22265 + }, + { + "epoch": 0.22266, + "grad_norm": 0.8851471431032033, + "learning_rate": 0.003, + "loss": 4.052, + "step": 22266 + }, + { + "epoch": 0.22267, + "grad_norm": 0.8431342047627507, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 22267 + }, + { + "epoch": 0.22268, + "grad_norm": 0.8837420983986594, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 22268 + }, + { + "epoch": 0.22269, + "grad_norm": 1.0880931693592621, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 22269 + }, + { + "epoch": 0.2227, + "grad_norm": 1.2080536062950191, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 22270 + }, + { + "epoch": 0.22271, + "grad_norm": 0.9474317664570654, + "learning_rate": 0.003, + "loss": 4.062, + "step": 22271 + }, + { + "epoch": 0.22272, + "grad_norm": 1.0231568615825237, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 22272 + }, + { + "epoch": 0.22273, + "grad_norm": 1.1190786499611443, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 22273 + }, + { + "epoch": 0.22274, + "grad_norm": 1.113788713885249, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 22274 + }, + { + "epoch": 0.22275, + "grad_norm": 1.2485713434627368, + "learning_rate": 0.003, + "loss": 4.069, + "step": 22275 + }, + { + "epoch": 0.22276, + "grad_norm": 1.2172895981413447, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 22276 + }, + { + "epoch": 0.22277, + "grad_norm": 0.9123608144218428, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 22277 + }, + { + "epoch": 0.22278, + "grad_norm": 1.2627123129939064, + "learning_rate": 0.003, + "loss": 4.051, + "step": 22278 + }, + { + "epoch": 0.22279, + "grad_norm": 0.967261451421243, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 22279 + }, + { + "epoch": 0.2228, + "grad_norm": 1.0286880413273014, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 22280 + }, + { + "epoch": 0.22281, + "grad_norm": 1.2052750447712857, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 22281 + }, + { + "epoch": 0.22282, + "grad_norm": 1.0127724089247927, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 22282 + }, + { + "epoch": 0.22283, + "grad_norm": 1.269549190982028, + "learning_rate": 0.003, + "loss": 4.051, + "step": 22283 + }, + { + "epoch": 0.22284, + "grad_norm": 0.8467489001524422, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 22284 + }, + { + "epoch": 0.22285, + "grad_norm": 1.0575268301789376, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 22285 + }, + { + "epoch": 0.22286, + "grad_norm": 1.1068526589285197, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 22286 + }, + { + "epoch": 0.22287, + "grad_norm": 1.0292846998497767, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 22287 + }, + { + "epoch": 0.22288, + "grad_norm": 1.054598006397313, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 22288 + }, + { + "epoch": 0.22289, + "grad_norm": 1.1683219928852846, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 22289 + }, + { + "epoch": 0.2229, + "grad_norm": 1.0381387236434874, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 22290 + }, + { + "epoch": 0.22291, + "grad_norm": 1.2713566071876572, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 22291 + }, + { + "epoch": 0.22292, + "grad_norm": 1.4136651628956474, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 22292 + }, + { + "epoch": 0.22293, + "grad_norm": 0.7081920557061824, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 22293 + }, + { + "epoch": 0.22294, + "grad_norm": 0.7870112843126854, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 22294 + }, + { + "epoch": 0.22295, + "grad_norm": 1.4884724730105408, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 22295 + }, + { + "epoch": 0.22296, + "grad_norm": 1.0502629052793597, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 22296 + }, + { + "epoch": 0.22297, + "grad_norm": 1.124289839949527, + "learning_rate": 0.003, + "loss": 4.038, + "step": 22297 + }, + { + "epoch": 0.22298, + "grad_norm": 0.8289209464916832, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 22298 + }, + { + "epoch": 0.22299, + "grad_norm": 0.7909209849246809, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 22299 + }, + { + "epoch": 0.223, + "grad_norm": 0.8041186417409784, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 22300 + }, + { + "epoch": 0.22301, + "grad_norm": 0.8434753288496988, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 22301 + }, + { + "epoch": 0.22302, + "grad_norm": 1.062527990436084, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 22302 + }, + { + "epoch": 0.22303, + "grad_norm": 1.1246496361455733, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 22303 + }, + { + "epoch": 0.22304, + "grad_norm": 0.9662876988563978, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 22304 + }, + { + "epoch": 0.22305, + "grad_norm": 1.1609771814745893, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 22305 + }, + { + "epoch": 0.22306, + "grad_norm": 1.0449666559075692, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 22306 + }, + { + "epoch": 0.22307, + "grad_norm": 1.1417211585096338, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 22307 + }, + { + "epoch": 0.22308, + "grad_norm": 1.0103857106942562, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 22308 + }, + { + "epoch": 0.22309, + "grad_norm": 1.1649804032747604, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 22309 + }, + { + "epoch": 0.2231, + "grad_norm": 1.0695847686929982, + "learning_rate": 0.003, + "loss": 4.06, + "step": 22310 + }, + { + "epoch": 0.22311, + "grad_norm": 1.0966823990601045, + "learning_rate": 0.003, + "loss": 4.083, + "step": 22311 + }, + { + "epoch": 0.22312, + "grad_norm": 1.0156643752812893, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 22312 + }, + { + "epoch": 0.22313, + "grad_norm": 1.1264417058878147, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 22313 + }, + { + "epoch": 0.22314, + "grad_norm": 1.1419263072847998, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 22314 + }, + { + "epoch": 0.22315, + "grad_norm": 1.0836516641508607, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 22315 + }, + { + "epoch": 0.22316, + "grad_norm": 1.0629557236322618, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 22316 + }, + { + "epoch": 0.22317, + "grad_norm": 1.0192957688359923, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 22317 + }, + { + "epoch": 0.22318, + "grad_norm": 1.1641672168728556, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 22318 + }, + { + "epoch": 0.22319, + "grad_norm": 1.0730176563582312, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 22319 + }, + { + "epoch": 0.2232, + "grad_norm": 1.1456276841068775, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 22320 + }, + { + "epoch": 0.22321, + "grad_norm": 0.9911699433003762, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 22321 + }, + { + "epoch": 0.22322, + "grad_norm": 1.3645844462210186, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 22322 + }, + { + "epoch": 0.22323, + "grad_norm": 0.8884706228943804, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 22323 + }, + { + "epoch": 0.22324, + "grad_norm": 0.9140335078751279, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 22324 + }, + { + "epoch": 0.22325, + "grad_norm": 1.1149216349369933, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 22325 + }, + { + "epoch": 0.22326, + "grad_norm": 1.107966107657045, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 22326 + }, + { + "epoch": 0.22327, + "grad_norm": 1.2664483943982636, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 22327 + }, + { + "epoch": 0.22328, + "grad_norm": 1.1048913907149147, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 22328 + }, + { + "epoch": 0.22329, + "grad_norm": 0.9379343353571749, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 22329 + }, + { + "epoch": 0.2233, + "grad_norm": 0.8426721322169656, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 22330 + }, + { + "epoch": 0.22331, + "grad_norm": 0.9756569951531048, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 22331 + }, + { + "epoch": 0.22332, + "grad_norm": 1.2504126802927444, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 22332 + }, + { + "epoch": 0.22333, + "grad_norm": 0.815253028666604, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 22333 + }, + { + "epoch": 0.22334, + "grad_norm": 0.760110983337267, + "learning_rate": 0.003, + "loss": 4.0912, + "step": 22334 + }, + { + "epoch": 0.22335, + "grad_norm": 0.8980258378845584, + "learning_rate": 0.003, + "loss": 4.053, + "step": 22335 + }, + { + "epoch": 0.22336, + "grad_norm": 0.9719714756266843, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 22336 + }, + { + "epoch": 0.22337, + "grad_norm": 1.1775990461179406, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 22337 + }, + { + "epoch": 0.22338, + "grad_norm": 0.9996456249202219, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 22338 + }, + { + "epoch": 0.22339, + "grad_norm": 1.0261903324145105, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 22339 + }, + { + "epoch": 0.2234, + "grad_norm": 1.1872445086807997, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 22340 + }, + { + "epoch": 0.22341, + "grad_norm": 1.1803846687378126, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 22341 + }, + { + "epoch": 0.22342, + "grad_norm": 1.363894748280635, + "learning_rate": 0.003, + "loss": 4.039, + "step": 22342 + }, + { + "epoch": 0.22343, + "grad_norm": 0.8605749812535854, + "learning_rate": 0.003, + "loss": 4.067, + "step": 22343 + }, + { + "epoch": 0.22344, + "grad_norm": 0.8649724396346711, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 22344 + }, + { + "epoch": 0.22345, + "grad_norm": 0.8409942401497909, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 22345 + }, + { + "epoch": 0.22346, + "grad_norm": 0.9625204080670493, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 22346 + }, + { + "epoch": 0.22347, + "grad_norm": 1.2345681750245177, + "learning_rate": 0.003, + "loss": 4.046, + "step": 22347 + }, + { + "epoch": 0.22348, + "grad_norm": 0.9574771136823234, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 22348 + }, + { + "epoch": 0.22349, + "grad_norm": 1.1574218155445837, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 22349 + }, + { + "epoch": 0.2235, + "grad_norm": 1.0604384863676906, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 22350 + }, + { + "epoch": 0.22351, + "grad_norm": 1.2903704790464843, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 22351 + }, + { + "epoch": 0.22352, + "grad_norm": 0.9312923997095506, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 22352 + }, + { + "epoch": 0.22353, + "grad_norm": 1.1219332489419693, + "learning_rate": 0.003, + "loss": 4.051, + "step": 22353 + }, + { + "epoch": 0.22354, + "grad_norm": 1.1513047452229437, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 22354 + }, + { + "epoch": 0.22355, + "grad_norm": 0.9026450391404515, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 22355 + }, + { + "epoch": 0.22356, + "grad_norm": 0.9858402288403314, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 22356 + }, + { + "epoch": 0.22357, + "grad_norm": 1.3203179296898109, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 22357 + }, + { + "epoch": 0.22358, + "grad_norm": 1.1500540062723554, + "learning_rate": 0.003, + "loss": 4.0938, + "step": 22358 + }, + { + "epoch": 0.22359, + "grad_norm": 1.0749890573467222, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 22359 + }, + { + "epoch": 0.2236, + "grad_norm": 1.0338195337426912, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 22360 + }, + { + "epoch": 0.22361, + "grad_norm": 1.0578181054805744, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 22361 + }, + { + "epoch": 0.22362, + "grad_norm": 0.95995701411904, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 22362 + }, + { + "epoch": 0.22363, + "grad_norm": 1.0242912916311282, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 22363 + }, + { + "epoch": 0.22364, + "grad_norm": 1.0311171853723922, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 22364 + }, + { + "epoch": 0.22365, + "grad_norm": 0.9630545218655423, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 22365 + }, + { + "epoch": 0.22366, + "grad_norm": 1.2074370412226065, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 22366 + }, + { + "epoch": 0.22367, + "grad_norm": 1.0407705127892635, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 22367 + }, + { + "epoch": 0.22368, + "grad_norm": 1.284986655774507, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 22368 + }, + { + "epoch": 0.22369, + "grad_norm": 1.2016365698528375, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 22369 + }, + { + "epoch": 0.2237, + "grad_norm": 0.890688254568509, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 22370 + }, + { + "epoch": 0.22371, + "grad_norm": 0.7750132139438226, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 22371 + }, + { + "epoch": 0.22372, + "grad_norm": 0.8606812050309082, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 22372 + }, + { + "epoch": 0.22373, + "grad_norm": 1.1093696572681067, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 22373 + }, + { + "epoch": 0.22374, + "grad_norm": 1.082751568254946, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 22374 + }, + { + "epoch": 0.22375, + "grad_norm": 0.9922431158388926, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 22375 + }, + { + "epoch": 0.22376, + "grad_norm": 1.0412632038523095, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 22376 + }, + { + "epoch": 0.22377, + "grad_norm": 1.2350557623316951, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 22377 + }, + { + "epoch": 0.22378, + "grad_norm": 1.0522811284874776, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 22378 + }, + { + "epoch": 0.22379, + "grad_norm": 0.931441749624425, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 22379 + }, + { + "epoch": 0.2238, + "grad_norm": 1.1511161789195818, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 22380 + }, + { + "epoch": 0.22381, + "grad_norm": 1.2072087935010476, + "learning_rate": 0.003, + "loss": 4.023, + "step": 22381 + }, + { + "epoch": 0.22382, + "grad_norm": 0.7908197429176597, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 22382 + }, + { + "epoch": 0.22383, + "grad_norm": 0.8161567424496725, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 22383 + }, + { + "epoch": 0.22384, + "grad_norm": 0.9050118614277113, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 22384 + }, + { + "epoch": 0.22385, + "grad_norm": 1.0653597500816099, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 22385 + }, + { + "epoch": 0.22386, + "grad_norm": 1.2323665394333672, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 22386 + }, + { + "epoch": 0.22387, + "grad_norm": 0.9212457866526074, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 22387 + }, + { + "epoch": 0.22388, + "grad_norm": 0.9512145617495454, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 22388 + }, + { + "epoch": 0.22389, + "grad_norm": 1.081577454517777, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 22389 + }, + { + "epoch": 0.2239, + "grad_norm": 1.04543313099483, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 22390 + }, + { + "epoch": 0.22391, + "grad_norm": 1.14027403031076, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 22391 + }, + { + "epoch": 0.22392, + "grad_norm": 1.2607546422870277, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 22392 + }, + { + "epoch": 0.22393, + "grad_norm": 0.9763862474889891, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 22393 + }, + { + "epoch": 0.22394, + "grad_norm": 1.1340796721773192, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 22394 + }, + { + "epoch": 0.22395, + "grad_norm": 0.9994378419991179, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 22395 + }, + { + "epoch": 0.22396, + "grad_norm": 1.1420976323539886, + "learning_rate": 0.003, + "loss": 4.071, + "step": 22396 + }, + { + "epoch": 0.22397, + "grad_norm": 0.9337958137292957, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 22397 + }, + { + "epoch": 0.22398, + "grad_norm": 0.9422830440870098, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 22398 + }, + { + "epoch": 0.22399, + "grad_norm": 1.122146846452604, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 22399 + }, + { + "epoch": 0.224, + "grad_norm": 0.9856013436130137, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 22400 + }, + { + "epoch": 0.22401, + "grad_norm": 1.464681807150187, + "learning_rate": 0.003, + "loss": 4.1065, + "step": 22401 + }, + { + "epoch": 0.22402, + "grad_norm": 1.166844063109591, + "learning_rate": 0.003, + "loss": 4.056, + "step": 22402 + }, + { + "epoch": 0.22403, + "grad_norm": 0.9647551491373044, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 22403 + }, + { + "epoch": 0.22404, + "grad_norm": 1.3238062624249805, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 22404 + }, + { + "epoch": 0.22405, + "grad_norm": 0.9377254954376487, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 22405 + }, + { + "epoch": 0.22406, + "grad_norm": 1.1066266408180574, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 22406 + }, + { + "epoch": 0.22407, + "grad_norm": 1.1270257811973734, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 22407 + }, + { + "epoch": 0.22408, + "grad_norm": 0.9951630792310795, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 22408 + }, + { + "epoch": 0.22409, + "grad_norm": 1.0505536617191236, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 22409 + }, + { + "epoch": 0.2241, + "grad_norm": 1.0638123236668728, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 22410 + }, + { + "epoch": 0.22411, + "grad_norm": 1.3262434922212083, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 22411 + }, + { + "epoch": 0.22412, + "grad_norm": 1.0346536895195944, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 22412 + }, + { + "epoch": 0.22413, + "grad_norm": 1.1465942939118468, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 22413 + }, + { + "epoch": 0.22414, + "grad_norm": 0.9452643607528397, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 22414 + }, + { + "epoch": 0.22415, + "grad_norm": 0.8670075102874131, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 22415 + }, + { + "epoch": 0.22416, + "grad_norm": 1.0062895629041335, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 22416 + }, + { + "epoch": 0.22417, + "grad_norm": 1.3410391043624368, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 22417 + }, + { + "epoch": 0.22418, + "grad_norm": 0.9402666175344355, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 22418 + }, + { + "epoch": 0.22419, + "grad_norm": 0.9401614299685327, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 22419 + }, + { + "epoch": 0.2242, + "grad_norm": 0.9956359950590042, + "learning_rate": 0.003, + "loss": 4.075, + "step": 22420 + }, + { + "epoch": 0.22421, + "grad_norm": 1.198120550278088, + "learning_rate": 0.003, + "loss": 4.042, + "step": 22421 + }, + { + "epoch": 0.22422, + "grad_norm": 0.8758664882478415, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 22422 + }, + { + "epoch": 0.22423, + "grad_norm": 0.8291055337040151, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 22423 + }, + { + "epoch": 0.22424, + "grad_norm": 0.8337878478091165, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 22424 + }, + { + "epoch": 0.22425, + "grad_norm": 0.9923569583274027, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 22425 + }, + { + "epoch": 0.22426, + "grad_norm": 1.299299554197855, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 22426 + }, + { + "epoch": 0.22427, + "grad_norm": 0.9086022736754195, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 22427 + }, + { + "epoch": 0.22428, + "grad_norm": 1.2052287615388082, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 22428 + }, + { + "epoch": 0.22429, + "grad_norm": 1.1526802522403703, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 22429 + }, + { + "epoch": 0.2243, + "grad_norm": 1.218163705242154, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 22430 + }, + { + "epoch": 0.22431, + "grad_norm": 1.0169369393606453, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 22431 + }, + { + "epoch": 0.22432, + "grad_norm": 1.2499713537133363, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 22432 + }, + { + "epoch": 0.22433, + "grad_norm": 0.9620738540510864, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 22433 + }, + { + "epoch": 0.22434, + "grad_norm": 1.211918017935265, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 22434 + }, + { + "epoch": 0.22435, + "grad_norm": 1.217969732375626, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 22435 + }, + { + "epoch": 0.22436, + "grad_norm": 1.000089047335068, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 22436 + }, + { + "epoch": 0.22437, + "grad_norm": 1.0579807395821288, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 22437 + }, + { + "epoch": 0.22438, + "grad_norm": 1.0896149436407443, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 22438 + }, + { + "epoch": 0.22439, + "grad_norm": 1.0527934028901602, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 22439 + }, + { + "epoch": 0.2244, + "grad_norm": 1.1719320434333342, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 22440 + }, + { + "epoch": 0.22441, + "grad_norm": 0.9970470925963152, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 22441 + }, + { + "epoch": 0.22442, + "grad_norm": 1.067484613345627, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 22442 + }, + { + "epoch": 0.22443, + "grad_norm": 0.9396095075188677, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 22443 + }, + { + "epoch": 0.22444, + "grad_norm": 1.202751742890538, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 22444 + }, + { + "epoch": 0.22445, + "grad_norm": 0.8743578008597525, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 22445 + }, + { + "epoch": 0.22446, + "grad_norm": 0.9509808680456987, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 22446 + }, + { + "epoch": 0.22447, + "grad_norm": 1.2210233036887506, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 22447 + }, + { + "epoch": 0.22448, + "grad_norm": 1.1675214944700019, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 22448 + }, + { + "epoch": 0.22449, + "grad_norm": 0.8809677959311882, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 22449 + }, + { + "epoch": 0.2245, + "grad_norm": 0.9366451118415756, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 22450 + }, + { + "epoch": 0.22451, + "grad_norm": 1.037789863402292, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 22451 + }, + { + "epoch": 0.22452, + "grad_norm": 1.1095798834015436, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 22452 + }, + { + "epoch": 0.22453, + "grad_norm": 1.0641087467608297, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 22453 + }, + { + "epoch": 0.22454, + "grad_norm": 1.0998289448934275, + "learning_rate": 0.003, + "loss": 4.061, + "step": 22454 + }, + { + "epoch": 0.22455, + "grad_norm": 0.9848683579449208, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 22455 + }, + { + "epoch": 0.22456, + "grad_norm": 1.0477287538528925, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 22456 + }, + { + "epoch": 0.22457, + "grad_norm": 1.0800295987385093, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 22457 + }, + { + "epoch": 0.22458, + "grad_norm": 0.9069061971664648, + "learning_rate": 0.003, + "loss": 4.061, + "step": 22458 + }, + { + "epoch": 0.22459, + "grad_norm": 0.992274402568411, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 22459 + }, + { + "epoch": 0.2246, + "grad_norm": 1.489990556256964, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 22460 + }, + { + "epoch": 0.22461, + "grad_norm": 0.9592683019485003, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 22461 + }, + { + "epoch": 0.22462, + "grad_norm": 1.018171436210344, + "learning_rate": 0.003, + "loss": 4.057, + "step": 22462 + }, + { + "epoch": 0.22463, + "grad_norm": 1.2063105704166224, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 22463 + }, + { + "epoch": 0.22464, + "grad_norm": 1.2352108047958466, + "learning_rate": 0.003, + "loss": 4.0868, + "step": 22464 + }, + { + "epoch": 0.22465, + "grad_norm": 0.9940271085978912, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 22465 + }, + { + "epoch": 0.22466, + "grad_norm": 1.0266965345085155, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 22466 + }, + { + "epoch": 0.22467, + "grad_norm": 1.0219860697814622, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 22467 + }, + { + "epoch": 0.22468, + "grad_norm": 1.3824058576377525, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 22468 + }, + { + "epoch": 0.22469, + "grad_norm": 1.0637943265507082, + "learning_rate": 0.003, + "loss": 4.0902, + "step": 22469 + }, + { + "epoch": 0.2247, + "grad_norm": 1.1992157503159058, + "learning_rate": 0.003, + "loss": 4.064, + "step": 22470 + }, + { + "epoch": 0.22471, + "grad_norm": 0.9708445738302144, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 22471 + }, + { + "epoch": 0.22472, + "grad_norm": 1.2628429901184606, + "learning_rate": 0.003, + "loss": 4.0938, + "step": 22472 + }, + { + "epoch": 0.22473, + "grad_norm": 0.9934260489045026, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 22473 + }, + { + "epoch": 0.22474, + "grad_norm": 1.1148497213234265, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 22474 + }, + { + "epoch": 0.22475, + "grad_norm": 1.1958913598031464, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 22475 + }, + { + "epoch": 0.22476, + "grad_norm": 1.0305830149225952, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 22476 + }, + { + "epoch": 0.22477, + "grad_norm": 0.8928436337803399, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 22477 + }, + { + "epoch": 0.22478, + "grad_norm": 0.9179774133809768, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 22478 + }, + { + "epoch": 0.22479, + "grad_norm": 1.020851331223456, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 22479 + }, + { + "epoch": 0.2248, + "grad_norm": 1.19966082138903, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 22480 + }, + { + "epoch": 0.22481, + "grad_norm": 0.8885730782900061, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 22481 + }, + { + "epoch": 0.22482, + "grad_norm": 1.0463477026138692, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 22482 + }, + { + "epoch": 0.22483, + "grad_norm": 1.0216758498859966, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 22483 + }, + { + "epoch": 0.22484, + "grad_norm": 1.0691074718426543, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 22484 + }, + { + "epoch": 0.22485, + "grad_norm": 1.1426421372302191, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 22485 + }, + { + "epoch": 0.22486, + "grad_norm": 1.152159934253922, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 22486 + }, + { + "epoch": 0.22487, + "grad_norm": 1.373567029456339, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 22487 + }, + { + "epoch": 0.22488, + "grad_norm": 0.846375852693257, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 22488 + }, + { + "epoch": 0.22489, + "grad_norm": 0.8666567488221609, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 22489 + }, + { + "epoch": 0.2249, + "grad_norm": 0.8869781789922504, + "learning_rate": 0.003, + "loss": 4.059, + "step": 22490 + }, + { + "epoch": 0.22491, + "grad_norm": 1.0275102526620044, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 22491 + }, + { + "epoch": 0.22492, + "grad_norm": 0.9914543013202267, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 22492 + }, + { + "epoch": 0.22493, + "grad_norm": 1.219099209913054, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 22493 + }, + { + "epoch": 0.22494, + "grad_norm": 1.3039657650067895, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 22494 + }, + { + "epoch": 0.22495, + "grad_norm": 0.8904152823985702, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 22495 + }, + { + "epoch": 0.22496, + "grad_norm": 0.8918816477717991, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 22496 + }, + { + "epoch": 0.22497, + "grad_norm": 0.9081766363056867, + "learning_rate": 0.003, + "loss": 4.057, + "step": 22497 + }, + { + "epoch": 0.22498, + "grad_norm": 1.0390265363976225, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 22498 + }, + { + "epoch": 0.22499, + "grad_norm": 1.248297813781798, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 22499 + }, + { + "epoch": 0.225, + "grad_norm": 0.9619697584341302, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 22500 + }, + { + "epoch": 0.22501, + "grad_norm": 1.116176116041455, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 22501 + }, + { + "epoch": 0.22502, + "grad_norm": 1.1141485501074664, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 22502 + }, + { + "epoch": 0.22503, + "grad_norm": 0.9907648030594403, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 22503 + }, + { + "epoch": 0.22504, + "grad_norm": 1.0857498541061046, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 22504 + }, + { + "epoch": 0.22505, + "grad_norm": 0.9625932127002438, + "learning_rate": 0.003, + "loss": 4.068, + "step": 22505 + }, + { + "epoch": 0.22506, + "grad_norm": 1.1407774471744334, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 22506 + }, + { + "epoch": 0.22507, + "grad_norm": 1.2747392023177753, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 22507 + }, + { + "epoch": 0.22508, + "grad_norm": 1.192286518069241, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 22508 + }, + { + "epoch": 0.22509, + "grad_norm": 1.1539759071341258, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 22509 + }, + { + "epoch": 0.2251, + "grad_norm": 0.9417076585933534, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 22510 + }, + { + "epoch": 0.22511, + "grad_norm": 0.9721884021336454, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 22511 + }, + { + "epoch": 0.22512, + "grad_norm": 1.1257824100748415, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 22512 + }, + { + "epoch": 0.22513, + "grad_norm": 1.2825916329982192, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 22513 + }, + { + "epoch": 0.22514, + "grad_norm": 0.970275019234373, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 22514 + }, + { + "epoch": 0.22515, + "grad_norm": 1.0277952848804648, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 22515 + }, + { + "epoch": 0.22516, + "grad_norm": 1.175346972088747, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 22516 + }, + { + "epoch": 0.22517, + "grad_norm": 0.9365503320010827, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 22517 + }, + { + "epoch": 0.22518, + "grad_norm": 0.7715348743940424, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 22518 + }, + { + "epoch": 0.22519, + "grad_norm": 0.7926918509399645, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 22519 + }, + { + "epoch": 0.2252, + "grad_norm": 0.9108619226710537, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 22520 + }, + { + "epoch": 0.22521, + "grad_norm": 0.9972824725503685, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 22521 + }, + { + "epoch": 0.22522, + "grad_norm": 1.0917275122167198, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 22522 + }, + { + "epoch": 0.22523, + "grad_norm": 1.151270310744497, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 22523 + }, + { + "epoch": 0.22524, + "grad_norm": 1.1161333446664703, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 22524 + }, + { + "epoch": 0.22525, + "grad_norm": 1.0099718446008723, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 22525 + }, + { + "epoch": 0.22526, + "grad_norm": 1.2088921613163528, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 22526 + }, + { + "epoch": 0.22527, + "grad_norm": 0.9937614833347866, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 22527 + }, + { + "epoch": 0.22528, + "grad_norm": 1.0991812647483523, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 22528 + }, + { + "epoch": 0.22529, + "grad_norm": 1.0676478189137333, + "learning_rate": 0.003, + "loss": 4.051, + "step": 22529 + }, + { + "epoch": 0.2253, + "grad_norm": 1.2076072599434338, + "learning_rate": 0.003, + "loss": 4.047, + "step": 22530 + }, + { + "epoch": 0.22531, + "grad_norm": 1.605930759643696, + "learning_rate": 0.003, + "loss": 4.084, + "step": 22531 + }, + { + "epoch": 0.22532, + "grad_norm": 0.7845284355237903, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 22532 + }, + { + "epoch": 0.22533, + "grad_norm": 0.7409177384915987, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 22533 + }, + { + "epoch": 0.22534, + "grad_norm": 0.8163684820543265, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 22534 + }, + { + "epoch": 0.22535, + "grad_norm": 1.0136439823861143, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 22535 + }, + { + "epoch": 0.22536, + "grad_norm": 1.2681219318900987, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 22536 + }, + { + "epoch": 0.22537, + "grad_norm": 0.9628036182335371, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 22537 + }, + { + "epoch": 0.22538, + "grad_norm": 1.0866049445477368, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 22538 + }, + { + "epoch": 0.22539, + "grad_norm": 1.1690819872125466, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 22539 + }, + { + "epoch": 0.2254, + "grad_norm": 1.2240443629725801, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 22540 + }, + { + "epoch": 0.22541, + "grad_norm": 1.1040830976317288, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 22541 + }, + { + "epoch": 0.22542, + "grad_norm": 1.2478642950414598, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 22542 + }, + { + "epoch": 0.22543, + "grad_norm": 0.9297330347461877, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 22543 + }, + { + "epoch": 0.22544, + "grad_norm": 1.0372142281741699, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 22544 + }, + { + "epoch": 0.22545, + "grad_norm": 1.2339331490334051, + "learning_rate": 0.003, + "loss": 4.068, + "step": 22545 + }, + { + "epoch": 0.22546, + "grad_norm": 1.0286625428822491, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 22546 + }, + { + "epoch": 0.22547, + "grad_norm": 1.2361473555261813, + "learning_rate": 0.003, + "loss": 4.053, + "step": 22547 + }, + { + "epoch": 0.22548, + "grad_norm": 0.9976035735946247, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 22548 + }, + { + "epoch": 0.22549, + "grad_norm": 1.3305338698439537, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 22549 + }, + { + "epoch": 0.2255, + "grad_norm": 0.971512787788029, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 22550 + }, + { + "epoch": 0.22551, + "grad_norm": 1.0564859725051992, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 22551 + }, + { + "epoch": 0.22552, + "grad_norm": 1.1651723173845807, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 22552 + }, + { + "epoch": 0.22553, + "grad_norm": 0.9674531761832571, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 22553 + }, + { + "epoch": 0.22554, + "grad_norm": 1.179146380350289, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 22554 + }, + { + "epoch": 0.22555, + "grad_norm": 1.0450341295723669, + "learning_rate": 0.003, + "loss": 4.07, + "step": 22555 + }, + { + "epoch": 0.22556, + "grad_norm": 1.1384844723572152, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 22556 + }, + { + "epoch": 0.22557, + "grad_norm": 0.9892363146469764, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 22557 + }, + { + "epoch": 0.22558, + "grad_norm": 1.0427137717830648, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 22558 + }, + { + "epoch": 0.22559, + "grad_norm": 1.1588897786132946, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 22559 + }, + { + "epoch": 0.2256, + "grad_norm": 1.2783967101538578, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 22560 + }, + { + "epoch": 0.22561, + "grad_norm": 0.8699119858186579, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 22561 + }, + { + "epoch": 0.22562, + "grad_norm": 1.025007892426665, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 22562 + }, + { + "epoch": 0.22563, + "grad_norm": 1.060391600307385, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 22563 + }, + { + "epoch": 0.22564, + "grad_norm": 1.3040983564903983, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 22564 + }, + { + "epoch": 0.22565, + "grad_norm": 1.1628309820101068, + "learning_rate": 0.003, + "loss": 4.084, + "step": 22565 + }, + { + "epoch": 0.22566, + "grad_norm": 0.9612582687284942, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 22566 + }, + { + "epoch": 0.22567, + "grad_norm": 0.9990077134497042, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 22567 + }, + { + "epoch": 0.22568, + "grad_norm": 1.0716375350304883, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 22568 + }, + { + "epoch": 0.22569, + "grad_norm": 0.9568970403178704, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 22569 + }, + { + "epoch": 0.2257, + "grad_norm": 0.8993114722119465, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 22570 + }, + { + "epoch": 0.22571, + "grad_norm": 0.9220546225161004, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 22571 + }, + { + "epoch": 0.22572, + "grad_norm": 1.0715928298468227, + "learning_rate": 0.003, + "loss": 4.076, + "step": 22572 + }, + { + "epoch": 0.22573, + "grad_norm": 1.0843929121978422, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 22573 + }, + { + "epoch": 0.22574, + "grad_norm": 1.1056951948295102, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 22574 + }, + { + "epoch": 0.22575, + "grad_norm": 1.0065518468082317, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 22575 + }, + { + "epoch": 0.22576, + "grad_norm": 1.222049836427641, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 22576 + }, + { + "epoch": 0.22577, + "grad_norm": 0.9657958957211283, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 22577 + }, + { + "epoch": 0.22578, + "grad_norm": 1.1751564285865475, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 22578 + }, + { + "epoch": 0.22579, + "grad_norm": 0.8947230123104734, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 22579 + }, + { + "epoch": 0.2258, + "grad_norm": 0.885642356315052, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 22580 + }, + { + "epoch": 0.22581, + "grad_norm": 1.0897548650886653, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 22581 + }, + { + "epoch": 0.22582, + "grad_norm": 1.2209224117084332, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 22582 + }, + { + "epoch": 0.22583, + "grad_norm": 1.424179768816507, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 22583 + }, + { + "epoch": 0.22584, + "grad_norm": 1.0919015013107554, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 22584 + }, + { + "epoch": 0.22585, + "grad_norm": 1.093540550330662, + "learning_rate": 0.003, + "loss": 4.087, + "step": 22585 + }, + { + "epoch": 0.22586, + "grad_norm": 1.1357768115199334, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 22586 + }, + { + "epoch": 0.22587, + "grad_norm": 1.0906378093256033, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 22587 + }, + { + "epoch": 0.22588, + "grad_norm": 1.1430325609998344, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 22588 + }, + { + "epoch": 0.22589, + "grad_norm": 1.137948919161406, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 22589 + }, + { + "epoch": 0.2259, + "grad_norm": 1.2133906797005338, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 22590 + }, + { + "epoch": 0.22591, + "grad_norm": 0.9909429343652391, + "learning_rate": 0.003, + "loss": 4.06, + "step": 22591 + }, + { + "epoch": 0.22592, + "grad_norm": 1.0515151735497181, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 22592 + }, + { + "epoch": 0.22593, + "grad_norm": 1.2442206427572016, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 22593 + }, + { + "epoch": 0.22594, + "grad_norm": 0.8126405329698843, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 22594 + }, + { + "epoch": 0.22595, + "grad_norm": 0.8320018571412467, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 22595 + }, + { + "epoch": 0.22596, + "grad_norm": 0.9858003554158359, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 22596 + }, + { + "epoch": 0.22597, + "grad_norm": 1.0340049273803233, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 22597 + }, + { + "epoch": 0.22598, + "grad_norm": 0.9777775575609512, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 22598 + }, + { + "epoch": 0.22599, + "grad_norm": 1.2739817921277785, + "learning_rate": 0.003, + "loss": 4.049, + "step": 22599 + }, + { + "epoch": 0.226, + "grad_norm": 0.9638460487949367, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 22600 + }, + { + "epoch": 0.22601, + "grad_norm": 1.1086840721177997, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 22601 + }, + { + "epoch": 0.22602, + "grad_norm": 1.258125649367284, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 22602 + }, + { + "epoch": 0.22603, + "grad_norm": 1.0982101986436015, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 22603 + }, + { + "epoch": 0.22604, + "grad_norm": 1.0439336249713866, + "learning_rate": 0.003, + "loss": 4.058, + "step": 22604 + }, + { + "epoch": 0.22605, + "grad_norm": 1.00496145418761, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 22605 + }, + { + "epoch": 0.22606, + "grad_norm": 0.9996766090855498, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 22606 + }, + { + "epoch": 0.22607, + "grad_norm": 1.1246127985359498, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 22607 + }, + { + "epoch": 0.22608, + "grad_norm": 0.9041507061112296, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 22608 + }, + { + "epoch": 0.22609, + "grad_norm": 1.0264778503163519, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 22609 + }, + { + "epoch": 0.2261, + "grad_norm": 1.2304341942798516, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 22610 + }, + { + "epoch": 0.22611, + "grad_norm": 1.0111904110414276, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 22611 + }, + { + "epoch": 0.22612, + "grad_norm": 1.1254076898399807, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 22612 + }, + { + "epoch": 0.22613, + "grad_norm": 1.0892756178675247, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 22613 + }, + { + "epoch": 0.22614, + "grad_norm": 0.9962081455469234, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 22614 + }, + { + "epoch": 0.22615, + "grad_norm": 1.0936502318169192, + "learning_rate": 0.003, + "loss": 4.05, + "step": 22615 + }, + { + "epoch": 0.22616, + "grad_norm": 1.088623962496488, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 22616 + }, + { + "epoch": 0.22617, + "grad_norm": 1.0764705165709685, + "learning_rate": 0.003, + "loss": 4.044, + "step": 22617 + }, + { + "epoch": 0.22618, + "grad_norm": 1.2837152810577468, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 22618 + }, + { + "epoch": 0.22619, + "grad_norm": 0.7287123489079286, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 22619 + }, + { + "epoch": 0.2262, + "grad_norm": 0.8436457474250535, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 22620 + }, + { + "epoch": 0.22621, + "grad_norm": 0.9839484903684502, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 22621 + }, + { + "epoch": 0.22622, + "grad_norm": 1.2097007955948635, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 22622 + }, + { + "epoch": 0.22623, + "grad_norm": 0.7944323254006829, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 22623 + }, + { + "epoch": 0.22624, + "grad_norm": 0.816218620767578, + "learning_rate": 0.003, + "loss": 4.048, + "step": 22624 + }, + { + "epoch": 0.22625, + "grad_norm": 0.8386337043830674, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 22625 + }, + { + "epoch": 0.22626, + "grad_norm": 0.9234617679831854, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 22626 + }, + { + "epoch": 0.22627, + "grad_norm": 1.3803364419775297, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 22627 + }, + { + "epoch": 0.22628, + "grad_norm": 1.057156778682424, + "learning_rate": 0.003, + "loss": 4.084, + "step": 22628 + }, + { + "epoch": 0.22629, + "grad_norm": 1.2810117451951746, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 22629 + }, + { + "epoch": 0.2263, + "grad_norm": 1.03524726709079, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 22630 + }, + { + "epoch": 0.22631, + "grad_norm": 1.2787159747337848, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 22631 + }, + { + "epoch": 0.22632, + "grad_norm": 1.0363851497583005, + "learning_rate": 0.003, + "loss": 4.05, + "step": 22632 + }, + { + "epoch": 0.22633, + "grad_norm": 1.0985776497552882, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 22633 + }, + { + "epoch": 0.22634, + "grad_norm": 1.201430871781674, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 22634 + }, + { + "epoch": 0.22635, + "grad_norm": 1.1548619501128752, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 22635 + }, + { + "epoch": 0.22636, + "grad_norm": 1.1624177353847014, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 22636 + }, + { + "epoch": 0.22637, + "grad_norm": 1.1008968986576237, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 22637 + }, + { + "epoch": 0.22638, + "grad_norm": 1.058566109100919, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 22638 + }, + { + "epoch": 0.22639, + "grad_norm": 0.8366433430323444, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 22639 + }, + { + "epoch": 0.2264, + "grad_norm": 0.7168204310008179, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 22640 + }, + { + "epoch": 0.22641, + "grad_norm": 0.7175345054002166, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 22641 + }, + { + "epoch": 0.22642, + "grad_norm": 0.7056817292501352, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 22642 + }, + { + "epoch": 0.22643, + "grad_norm": 0.9362064150014895, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 22643 + }, + { + "epoch": 0.22644, + "grad_norm": 1.2676159853043472, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 22644 + }, + { + "epoch": 0.22645, + "grad_norm": 0.9771353762653823, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 22645 + }, + { + "epoch": 0.22646, + "grad_norm": 1.138771067596515, + "learning_rate": 0.003, + "loss": 4.0857, + "step": 22646 + }, + { + "epoch": 0.22647, + "grad_norm": 1.0526609011035137, + "learning_rate": 0.003, + "loss": 4.011, + "step": 22647 + }, + { + "epoch": 0.22648, + "grad_norm": 1.0648542441403084, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 22648 + }, + { + "epoch": 0.22649, + "grad_norm": 1.0989158674329171, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 22649 + }, + { + "epoch": 0.2265, + "grad_norm": 1.1019154711375403, + "learning_rate": 0.003, + "loss": 4.021, + "step": 22650 + }, + { + "epoch": 0.22651, + "grad_norm": 1.1301329476796718, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 22651 + }, + { + "epoch": 0.22652, + "grad_norm": 1.0249888631957, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 22652 + }, + { + "epoch": 0.22653, + "grad_norm": 1.2520159026559352, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 22653 + }, + { + "epoch": 0.22654, + "grad_norm": 0.9668496231526368, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 22654 + }, + { + "epoch": 0.22655, + "grad_norm": 1.0587015289610346, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 22655 + }, + { + "epoch": 0.22656, + "grad_norm": 1.0380669683156214, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 22656 + }, + { + "epoch": 0.22657, + "grad_norm": 1.19180406547561, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 22657 + }, + { + "epoch": 0.22658, + "grad_norm": 1.1236748424885916, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 22658 + }, + { + "epoch": 0.22659, + "grad_norm": 1.2188816146620425, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 22659 + }, + { + "epoch": 0.2266, + "grad_norm": 1.0350292915465633, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 22660 + }, + { + "epoch": 0.22661, + "grad_norm": 1.1408440427158653, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 22661 + }, + { + "epoch": 0.22662, + "grad_norm": 1.108248540629782, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 22662 + }, + { + "epoch": 0.22663, + "grad_norm": 0.8346215299535009, + "learning_rate": 0.003, + "loss": 4.036, + "step": 22663 + }, + { + "epoch": 0.22664, + "grad_norm": 0.9245651774533653, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 22664 + }, + { + "epoch": 0.22665, + "grad_norm": 1.11578333649667, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 22665 + }, + { + "epoch": 0.22666, + "grad_norm": 1.026922038202442, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 22666 + }, + { + "epoch": 0.22667, + "grad_norm": 1.2608870656607427, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 22667 + }, + { + "epoch": 0.22668, + "grad_norm": 0.974483746483182, + "learning_rate": 0.003, + "loss": 4.033, + "step": 22668 + }, + { + "epoch": 0.22669, + "grad_norm": 0.9549363202525608, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 22669 + }, + { + "epoch": 0.2267, + "grad_norm": 1.1034487043500556, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 22670 + }, + { + "epoch": 0.22671, + "grad_norm": 1.3011647436060687, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 22671 + }, + { + "epoch": 0.22672, + "grad_norm": 1.2316702636875723, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 22672 + }, + { + "epoch": 0.22673, + "grad_norm": 1.016666476299037, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 22673 + }, + { + "epoch": 0.22674, + "grad_norm": 1.1213261042203733, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 22674 + }, + { + "epoch": 0.22675, + "grad_norm": 1.0664058831215755, + "learning_rate": 0.003, + "loss": 4.086, + "step": 22675 + }, + { + "epoch": 0.22676, + "grad_norm": 1.184898645611014, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 22676 + }, + { + "epoch": 0.22677, + "grad_norm": 0.9693146668756276, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 22677 + }, + { + "epoch": 0.22678, + "grad_norm": 0.939935303368772, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 22678 + }, + { + "epoch": 0.22679, + "grad_norm": 0.9236611139385635, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 22679 + }, + { + "epoch": 0.2268, + "grad_norm": 0.9498501084638807, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 22680 + }, + { + "epoch": 0.22681, + "grad_norm": 1.1499360779323902, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 22681 + }, + { + "epoch": 0.22682, + "grad_norm": 0.9581411630812054, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 22682 + }, + { + "epoch": 0.22683, + "grad_norm": 1.0767946181917907, + "learning_rate": 0.003, + "loss": 4.057, + "step": 22683 + }, + { + "epoch": 0.22684, + "grad_norm": 1.2363029210148875, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 22684 + }, + { + "epoch": 0.22685, + "grad_norm": 1.203371398069033, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 22685 + }, + { + "epoch": 0.22686, + "grad_norm": 1.1242599966649176, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 22686 + }, + { + "epoch": 0.22687, + "grad_norm": 1.136977697725721, + "learning_rate": 0.003, + "loss": 4.071, + "step": 22687 + }, + { + "epoch": 0.22688, + "grad_norm": 1.13299072066787, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 22688 + }, + { + "epoch": 0.22689, + "grad_norm": 1.0514642693898273, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 22689 + }, + { + "epoch": 0.2269, + "grad_norm": 1.034432683673166, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 22690 + }, + { + "epoch": 0.22691, + "grad_norm": 1.263968803760662, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 22691 + }, + { + "epoch": 0.22692, + "grad_norm": 1.0221288871512209, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 22692 + }, + { + "epoch": 0.22693, + "grad_norm": 1.0093924871350197, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 22693 + }, + { + "epoch": 0.22694, + "grad_norm": 1.1230190349831581, + "learning_rate": 0.003, + "loss": 4.06, + "step": 22694 + }, + { + "epoch": 0.22695, + "grad_norm": 0.9691961946294024, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 22695 + }, + { + "epoch": 0.22696, + "grad_norm": 1.179312845377929, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 22696 + }, + { + "epoch": 0.22697, + "grad_norm": 0.9681160251969262, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 22697 + }, + { + "epoch": 0.22698, + "grad_norm": 0.9644485226381941, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 22698 + }, + { + "epoch": 0.22699, + "grad_norm": 0.9382515170163946, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 22699 + }, + { + "epoch": 0.227, + "grad_norm": 1.0973750126560924, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 22700 + }, + { + "epoch": 0.22701, + "grad_norm": 1.1936771488026676, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 22701 + }, + { + "epoch": 0.22702, + "grad_norm": 1.1870841627576656, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 22702 + }, + { + "epoch": 0.22703, + "grad_norm": 0.987723050407163, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 22703 + }, + { + "epoch": 0.22704, + "grad_norm": 1.080743665829685, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 22704 + }, + { + "epoch": 0.22705, + "grad_norm": 1.0442605914374299, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 22705 + }, + { + "epoch": 0.22706, + "grad_norm": 1.2578470208687231, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 22706 + }, + { + "epoch": 0.22707, + "grad_norm": 1.0141949463446303, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 22707 + }, + { + "epoch": 0.22708, + "grad_norm": 1.1757375731690285, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 22708 + }, + { + "epoch": 0.22709, + "grad_norm": 0.9184804681302048, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 22709 + }, + { + "epoch": 0.2271, + "grad_norm": 1.1453415138128997, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 22710 + }, + { + "epoch": 0.22711, + "grad_norm": 1.2125852394516135, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 22711 + }, + { + "epoch": 0.22712, + "grad_norm": 0.8974900363646532, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 22712 + }, + { + "epoch": 0.22713, + "grad_norm": 0.9150739061657054, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 22713 + }, + { + "epoch": 0.22714, + "grad_norm": 1.026141883398064, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 22714 + }, + { + "epoch": 0.22715, + "grad_norm": 1.2768131983184139, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 22715 + }, + { + "epoch": 0.22716, + "grad_norm": 0.9980571561628434, + "learning_rate": 0.003, + "loss": 4.046, + "step": 22716 + }, + { + "epoch": 0.22717, + "grad_norm": 1.1876702628818767, + "learning_rate": 0.003, + "loss": 4.0854, + "step": 22717 + }, + { + "epoch": 0.22718, + "grad_norm": 1.0122391932911974, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 22718 + }, + { + "epoch": 0.22719, + "grad_norm": 1.1116318447799785, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 22719 + }, + { + "epoch": 0.2272, + "grad_norm": 0.9496550837609811, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 22720 + }, + { + "epoch": 0.22721, + "grad_norm": 1.0035078956876617, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 22721 + }, + { + "epoch": 0.22722, + "grad_norm": 1.260076626869755, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 22722 + }, + { + "epoch": 0.22723, + "grad_norm": 1.0329112326582155, + "learning_rate": 0.003, + "loss": 4.06, + "step": 22723 + }, + { + "epoch": 0.22724, + "grad_norm": 0.9981329655814446, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 22724 + }, + { + "epoch": 0.22725, + "grad_norm": 1.1570967429837145, + "learning_rate": 0.003, + "loss": 4.0848, + "step": 22725 + }, + { + "epoch": 0.22726, + "grad_norm": 1.2857995484678038, + "learning_rate": 0.003, + "loss": 4.066, + "step": 22726 + }, + { + "epoch": 0.22727, + "grad_norm": 1.0824255553972295, + "learning_rate": 0.003, + "loss": 4.079, + "step": 22727 + }, + { + "epoch": 0.22728, + "grad_norm": 0.8693234521599711, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 22728 + }, + { + "epoch": 0.22729, + "grad_norm": 0.9190207455022776, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 22729 + }, + { + "epoch": 0.2273, + "grad_norm": 1.2214552256393842, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 22730 + }, + { + "epoch": 0.22731, + "grad_norm": 1.1424052228645174, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 22731 + }, + { + "epoch": 0.22732, + "grad_norm": 0.891506080611325, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 22732 + }, + { + "epoch": 0.22733, + "grad_norm": 1.0207824003995911, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 22733 + }, + { + "epoch": 0.22734, + "grad_norm": 1.2152045260151811, + "learning_rate": 0.003, + "loss": 4.078, + "step": 22734 + }, + { + "epoch": 0.22735, + "grad_norm": 0.905374222280776, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 22735 + }, + { + "epoch": 0.22736, + "grad_norm": 0.9253632383552521, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 22736 + }, + { + "epoch": 0.22737, + "grad_norm": 0.9421334442940558, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 22737 + }, + { + "epoch": 0.22738, + "grad_norm": 0.8096811473200695, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 22738 + }, + { + "epoch": 0.22739, + "grad_norm": 0.8581294950726412, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 22739 + }, + { + "epoch": 0.2274, + "grad_norm": 0.9081598394279394, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 22740 + }, + { + "epoch": 0.22741, + "grad_norm": 0.9719596088360146, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 22741 + }, + { + "epoch": 0.22742, + "grad_norm": 1.1903533929582912, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 22742 + }, + { + "epoch": 0.22743, + "grad_norm": 1.2052493083169507, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 22743 + }, + { + "epoch": 0.22744, + "grad_norm": 1.1629365521808726, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 22744 + }, + { + "epoch": 0.22745, + "grad_norm": 1.005482686052524, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 22745 + }, + { + "epoch": 0.22746, + "grad_norm": 1.082103654574431, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 22746 + }, + { + "epoch": 0.22747, + "grad_norm": 1.1369615113324556, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 22747 + }, + { + "epoch": 0.22748, + "grad_norm": 1.3171934352025798, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 22748 + }, + { + "epoch": 0.22749, + "grad_norm": 0.8016163223810633, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 22749 + }, + { + "epoch": 0.2275, + "grad_norm": 0.8142453055282701, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 22750 + }, + { + "epoch": 0.22751, + "grad_norm": 1.147338967788604, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 22751 + }, + { + "epoch": 0.22752, + "grad_norm": 1.2026415958060126, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 22752 + }, + { + "epoch": 0.22753, + "grad_norm": 0.9903184399749815, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 22753 + }, + { + "epoch": 0.22754, + "grad_norm": 1.1588169123853003, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 22754 + }, + { + "epoch": 0.22755, + "grad_norm": 0.9652576447105401, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 22755 + }, + { + "epoch": 0.22756, + "grad_norm": 1.0077786346017974, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 22756 + }, + { + "epoch": 0.22757, + "grad_norm": 1.0944189422154287, + "learning_rate": 0.003, + "loss": 4.05, + "step": 22757 + }, + { + "epoch": 0.22758, + "grad_norm": 1.0542426057962648, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 22758 + }, + { + "epoch": 0.22759, + "grad_norm": 1.2005773003128315, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 22759 + }, + { + "epoch": 0.2276, + "grad_norm": 0.9940384831744792, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 22760 + }, + { + "epoch": 0.22761, + "grad_norm": 1.0410035551855064, + "learning_rate": 0.003, + "loss": 4.065, + "step": 22761 + }, + { + "epoch": 0.22762, + "grad_norm": 0.9023209169034415, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 22762 + }, + { + "epoch": 0.22763, + "grad_norm": 1.0314246411640609, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 22763 + }, + { + "epoch": 0.22764, + "grad_norm": 1.0485417061010875, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 22764 + }, + { + "epoch": 0.22765, + "grad_norm": 1.2608281380877528, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 22765 + }, + { + "epoch": 0.22766, + "grad_norm": 0.9474321630209444, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 22766 + }, + { + "epoch": 0.22767, + "grad_norm": 1.4337894837192782, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 22767 + }, + { + "epoch": 0.22768, + "grad_norm": 1.0004910699405996, + "learning_rate": 0.003, + "loss": 4.0893, + "step": 22768 + }, + { + "epoch": 0.22769, + "grad_norm": 1.113319531900149, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 22769 + }, + { + "epoch": 0.2277, + "grad_norm": 1.2778478011091141, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 22770 + }, + { + "epoch": 0.22771, + "grad_norm": 1.196854012180637, + "learning_rate": 0.003, + "loss": 4.065, + "step": 22771 + }, + { + "epoch": 0.22772, + "grad_norm": 1.06424146231236, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 22772 + }, + { + "epoch": 0.22773, + "grad_norm": 1.1098352905315998, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 22773 + }, + { + "epoch": 0.22774, + "grad_norm": 0.991549638836504, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 22774 + }, + { + "epoch": 0.22775, + "grad_norm": 1.2252666273647663, + "learning_rate": 0.003, + "loss": 4.055, + "step": 22775 + }, + { + "epoch": 0.22776, + "grad_norm": 1.0691133690730525, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 22776 + }, + { + "epoch": 0.22777, + "grad_norm": 1.2964127589273702, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 22777 + }, + { + "epoch": 0.22778, + "grad_norm": 0.9434142828315909, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 22778 + }, + { + "epoch": 0.22779, + "grad_norm": 1.1154346936487058, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 22779 + }, + { + "epoch": 0.2278, + "grad_norm": 1.2814448214959362, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 22780 + }, + { + "epoch": 0.22781, + "grad_norm": 0.9867564092741077, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 22781 + }, + { + "epoch": 0.22782, + "grad_norm": 1.0369846215599425, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 22782 + }, + { + "epoch": 0.22783, + "grad_norm": 1.0091287542084428, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 22783 + }, + { + "epoch": 0.22784, + "grad_norm": 1.1653032443029752, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 22784 + }, + { + "epoch": 0.22785, + "grad_norm": 1.0627753732124292, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 22785 + }, + { + "epoch": 0.22786, + "grad_norm": 1.0670837288669779, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 22786 + }, + { + "epoch": 0.22787, + "grad_norm": 0.9892611328274344, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 22787 + }, + { + "epoch": 0.22788, + "grad_norm": 1.1540710300987744, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 22788 + }, + { + "epoch": 0.22789, + "grad_norm": 1.0139711026990825, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 22789 + }, + { + "epoch": 0.2279, + "grad_norm": 1.093753324698676, + "learning_rate": 0.003, + "loss": 4.0819, + "step": 22790 + }, + { + "epoch": 0.22791, + "grad_norm": 1.0948234880202896, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 22791 + }, + { + "epoch": 0.22792, + "grad_norm": 1.1125449286745646, + "learning_rate": 0.003, + "loss": 4.043, + "step": 22792 + }, + { + "epoch": 0.22793, + "grad_norm": 1.162695987030045, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 22793 + }, + { + "epoch": 0.22794, + "grad_norm": 0.8231343333735222, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 22794 + }, + { + "epoch": 0.22795, + "grad_norm": 1.0314564147316923, + "learning_rate": 0.003, + "loss": 4.1077, + "step": 22795 + }, + { + "epoch": 0.22796, + "grad_norm": 1.00998476714245, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 22796 + }, + { + "epoch": 0.22797, + "grad_norm": 1.499244642221076, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 22797 + }, + { + "epoch": 0.22798, + "grad_norm": 1.0599084745791225, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 22798 + }, + { + "epoch": 0.22799, + "grad_norm": 1.1017192468706576, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 22799 + }, + { + "epoch": 0.228, + "grad_norm": 1.1215818457961728, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 22800 + }, + { + "epoch": 0.22801, + "grad_norm": 1.1453621400112506, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 22801 + }, + { + "epoch": 0.22802, + "grad_norm": 0.861777631823021, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 22802 + }, + { + "epoch": 0.22803, + "grad_norm": 0.9576839366795911, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 22803 + }, + { + "epoch": 0.22804, + "grad_norm": 1.068133338653689, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 22804 + }, + { + "epoch": 0.22805, + "grad_norm": 0.9213959494677512, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 22805 + }, + { + "epoch": 0.22806, + "grad_norm": 1.1453234272052268, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 22806 + }, + { + "epoch": 0.22807, + "grad_norm": 0.9466814693193292, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 22807 + }, + { + "epoch": 0.22808, + "grad_norm": 1.086220822904849, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 22808 + }, + { + "epoch": 0.22809, + "grad_norm": 1.0095032778629776, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 22809 + }, + { + "epoch": 0.2281, + "grad_norm": 1.1862254296106665, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 22810 + }, + { + "epoch": 0.22811, + "grad_norm": 1.2317937476387095, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 22811 + }, + { + "epoch": 0.22812, + "grad_norm": 1.0198483525180178, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 22812 + }, + { + "epoch": 0.22813, + "grad_norm": 1.2203452988785826, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 22813 + }, + { + "epoch": 0.22814, + "grad_norm": 1.136524800859732, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 22814 + }, + { + "epoch": 0.22815, + "grad_norm": 0.963138419924167, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 22815 + }, + { + "epoch": 0.22816, + "grad_norm": 0.9472192183313011, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 22816 + }, + { + "epoch": 0.22817, + "grad_norm": 1.009345708856213, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 22817 + }, + { + "epoch": 0.22818, + "grad_norm": 1.1868763728990446, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 22818 + }, + { + "epoch": 0.22819, + "grad_norm": 1.1393409284743279, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 22819 + }, + { + "epoch": 0.2282, + "grad_norm": 1.0017417928561267, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 22820 + }, + { + "epoch": 0.22821, + "grad_norm": 0.9371237205566023, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 22821 + }, + { + "epoch": 0.22822, + "grad_norm": 1.0235088416844893, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 22822 + }, + { + "epoch": 0.22823, + "grad_norm": 1.1999229571439278, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 22823 + }, + { + "epoch": 0.22824, + "grad_norm": 0.9271238489425409, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 22824 + }, + { + "epoch": 0.22825, + "grad_norm": 0.9844131171577343, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 22825 + }, + { + "epoch": 0.22826, + "grad_norm": 1.0040622481942323, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 22826 + }, + { + "epoch": 0.22827, + "grad_norm": 1.1252785091414204, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 22827 + }, + { + "epoch": 0.22828, + "grad_norm": 0.906829969837117, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 22828 + }, + { + "epoch": 0.22829, + "grad_norm": 0.9804371009472621, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 22829 + }, + { + "epoch": 0.2283, + "grad_norm": 1.1270445577661332, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 22830 + }, + { + "epoch": 0.22831, + "grad_norm": 1.0077331433052432, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 22831 + }, + { + "epoch": 0.22832, + "grad_norm": 1.46190062472228, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 22832 + }, + { + "epoch": 0.22833, + "grad_norm": 0.8878557206196854, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 22833 + }, + { + "epoch": 0.22834, + "grad_norm": 0.9165465097874225, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 22834 + }, + { + "epoch": 0.22835, + "grad_norm": 1.0992057760991696, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 22835 + }, + { + "epoch": 0.22836, + "grad_norm": 1.058914182331367, + "learning_rate": 0.003, + "loss": 4.095, + "step": 22836 + }, + { + "epoch": 0.22837, + "grad_norm": 1.2357784476796534, + "learning_rate": 0.003, + "loss": 4.076, + "step": 22837 + }, + { + "epoch": 0.22838, + "grad_norm": 1.0728650251757574, + "learning_rate": 0.003, + "loss": 4.1071, + "step": 22838 + }, + { + "epoch": 0.22839, + "grad_norm": 1.1050052463100206, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 22839 + }, + { + "epoch": 0.2284, + "grad_norm": 1.0422826578124216, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 22840 + }, + { + "epoch": 0.22841, + "grad_norm": 1.1110325102474816, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 22841 + }, + { + "epoch": 0.22842, + "grad_norm": 0.9048058905218529, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 22842 + }, + { + "epoch": 0.22843, + "grad_norm": 0.9503307661070878, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 22843 + }, + { + "epoch": 0.22844, + "grad_norm": 1.3559390426123594, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 22844 + }, + { + "epoch": 0.22845, + "grad_norm": 0.8457605191369061, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 22845 + }, + { + "epoch": 0.22846, + "grad_norm": 1.0294168653989435, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 22846 + }, + { + "epoch": 0.22847, + "grad_norm": 1.0375261633154422, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 22847 + }, + { + "epoch": 0.22848, + "grad_norm": 1.239290644514706, + "learning_rate": 0.003, + "loss": 4.0939, + "step": 22848 + }, + { + "epoch": 0.22849, + "grad_norm": 1.3173747687902935, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 22849 + }, + { + "epoch": 0.2285, + "grad_norm": 1.1224710282344228, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 22850 + }, + { + "epoch": 0.22851, + "grad_norm": 1.1880691597444575, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 22851 + }, + { + "epoch": 0.22852, + "grad_norm": 1.0490178895135605, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 22852 + }, + { + "epoch": 0.22853, + "grad_norm": 0.9356348468343376, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 22853 + }, + { + "epoch": 0.22854, + "grad_norm": 1.0016183627293493, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 22854 + }, + { + "epoch": 0.22855, + "grad_norm": 1.3923452926293296, + "learning_rate": 0.003, + "loss": 4.085, + "step": 22855 + }, + { + "epoch": 0.22856, + "grad_norm": 0.658112468077093, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 22856 + }, + { + "epoch": 0.22857, + "grad_norm": 0.7076553086938133, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 22857 + }, + { + "epoch": 0.22858, + "grad_norm": 0.7864709896328681, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 22858 + }, + { + "epoch": 0.22859, + "grad_norm": 0.90116916012843, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 22859 + }, + { + "epoch": 0.2286, + "grad_norm": 1.0345566509398967, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 22860 + }, + { + "epoch": 0.22861, + "grad_norm": 1.160555244399162, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 22861 + }, + { + "epoch": 0.22862, + "grad_norm": 0.8581843568615816, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 22862 + }, + { + "epoch": 0.22863, + "grad_norm": 0.9581025739242088, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 22863 + }, + { + "epoch": 0.22864, + "grad_norm": 1.0804972007338332, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 22864 + }, + { + "epoch": 0.22865, + "grad_norm": 1.2486324494807717, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 22865 + }, + { + "epoch": 0.22866, + "grad_norm": 1.0475401494103287, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 22866 + }, + { + "epoch": 0.22867, + "grad_norm": 1.1534436706639162, + "learning_rate": 0.003, + "loss": 4.044, + "step": 22867 + }, + { + "epoch": 0.22868, + "grad_norm": 1.0993735599350887, + "learning_rate": 0.003, + "loss": 4.1011, + "step": 22868 + }, + { + "epoch": 0.22869, + "grad_norm": 0.8900643456381807, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 22869 + }, + { + "epoch": 0.2287, + "grad_norm": 1.107160654645508, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 22870 + }, + { + "epoch": 0.22871, + "grad_norm": 1.087591730293768, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 22871 + }, + { + "epoch": 0.22872, + "grad_norm": 1.015509092047475, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 22872 + }, + { + "epoch": 0.22873, + "grad_norm": 1.0116604881134166, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 22873 + }, + { + "epoch": 0.22874, + "grad_norm": 1.401326741071446, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 22874 + }, + { + "epoch": 0.22875, + "grad_norm": 1.1931126848346199, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 22875 + }, + { + "epoch": 0.22876, + "grad_norm": 1.1921987293563259, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 22876 + }, + { + "epoch": 0.22877, + "grad_norm": 1.1973203029717259, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 22877 + }, + { + "epoch": 0.22878, + "grad_norm": 1.0292158876236621, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 22878 + }, + { + "epoch": 0.22879, + "grad_norm": 1.0984645212379915, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 22879 + }, + { + "epoch": 0.2288, + "grad_norm": 0.9314027507327496, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 22880 + }, + { + "epoch": 0.22881, + "grad_norm": 1.0173967783592437, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 22881 + }, + { + "epoch": 0.22882, + "grad_norm": 1.1058881770499598, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 22882 + }, + { + "epoch": 0.22883, + "grad_norm": 1.1093448784704423, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 22883 + }, + { + "epoch": 0.22884, + "grad_norm": 1.028338632084557, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 22884 + }, + { + "epoch": 0.22885, + "grad_norm": 1.2735618687010544, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 22885 + }, + { + "epoch": 0.22886, + "grad_norm": 1.2234772517772161, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 22886 + }, + { + "epoch": 0.22887, + "grad_norm": 0.9798743444837071, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 22887 + }, + { + "epoch": 0.22888, + "grad_norm": 0.9883135264632797, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 22888 + }, + { + "epoch": 0.22889, + "grad_norm": 1.1546438255140392, + "learning_rate": 0.003, + "loss": 4.067, + "step": 22889 + }, + { + "epoch": 0.2289, + "grad_norm": 0.9982256005150282, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 22890 + }, + { + "epoch": 0.22891, + "grad_norm": 1.0688781164534575, + "learning_rate": 0.003, + "loss": 4.059, + "step": 22891 + }, + { + "epoch": 0.22892, + "grad_norm": 0.9120808033929035, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 22892 + }, + { + "epoch": 0.22893, + "grad_norm": 0.9665753067883914, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 22893 + }, + { + "epoch": 0.22894, + "grad_norm": 1.1440954106697183, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 22894 + }, + { + "epoch": 0.22895, + "grad_norm": 1.2741841718006288, + "learning_rate": 0.003, + "loss": 4.052, + "step": 22895 + }, + { + "epoch": 0.22896, + "grad_norm": 0.9033007397263914, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 22896 + }, + { + "epoch": 0.22897, + "grad_norm": 0.991824575799767, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 22897 + }, + { + "epoch": 0.22898, + "grad_norm": 1.0912235856614454, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 22898 + }, + { + "epoch": 0.22899, + "grad_norm": 1.128080606935541, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 22899 + }, + { + "epoch": 0.229, + "grad_norm": 1.0027871749954884, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 22900 + }, + { + "epoch": 0.22901, + "grad_norm": 1.1292498956303025, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 22901 + }, + { + "epoch": 0.22902, + "grad_norm": 1.0111873327823366, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 22902 + }, + { + "epoch": 0.22903, + "grad_norm": 1.1946515915579896, + "learning_rate": 0.003, + "loss": 4.0886, + "step": 22903 + }, + { + "epoch": 0.22904, + "grad_norm": 0.9450145791949425, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 22904 + }, + { + "epoch": 0.22905, + "grad_norm": 1.0857070605863846, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 22905 + }, + { + "epoch": 0.22906, + "grad_norm": 1.0581673210097393, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 22906 + }, + { + "epoch": 0.22907, + "grad_norm": 1.0011351437353648, + "learning_rate": 0.003, + "loss": 4.067, + "step": 22907 + }, + { + "epoch": 0.22908, + "grad_norm": 1.0836296355118598, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 22908 + }, + { + "epoch": 0.22909, + "grad_norm": 1.1800982430582134, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 22909 + }, + { + "epoch": 0.2291, + "grad_norm": 1.0415760404521235, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 22910 + }, + { + "epoch": 0.22911, + "grad_norm": 1.2450330587407126, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 22911 + }, + { + "epoch": 0.22912, + "grad_norm": 1.2813423356942366, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 22912 + }, + { + "epoch": 0.22913, + "grad_norm": 1.0498160513286101, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 22913 + }, + { + "epoch": 0.22914, + "grad_norm": 0.9733605629580131, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 22914 + }, + { + "epoch": 0.22915, + "grad_norm": 1.0897550329129226, + "learning_rate": 0.003, + "loss": 4.065, + "step": 22915 + }, + { + "epoch": 0.22916, + "grad_norm": 0.8719964317271184, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 22916 + }, + { + "epoch": 0.22917, + "grad_norm": 0.9988909468958036, + "learning_rate": 0.003, + "loss": 4.041, + "step": 22917 + }, + { + "epoch": 0.22918, + "grad_norm": 1.0847212112006923, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 22918 + }, + { + "epoch": 0.22919, + "grad_norm": 1.0672284611685507, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 22919 + }, + { + "epoch": 0.2292, + "grad_norm": 0.9794401093574652, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 22920 + }, + { + "epoch": 0.22921, + "grad_norm": 1.327433123217694, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 22921 + }, + { + "epoch": 0.22922, + "grad_norm": 1.1583355464957559, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 22922 + }, + { + "epoch": 0.22923, + "grad_norm": 0.7978418434371057, + "learning_rate": 0.003, + "loss": 4.05, + "step": 22923 + }, + { + "epoch": 0.22924, + "grad_norm": 0.8087723084380822, + "learning_rate": 0.003, + "loss": 4.037, + "step": 22924 + }, + { + "epoch": 0.22925, + "grad_norm": 0.9761931578126225, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 22925 + }, + { + "epoch": 0.22926, + "grad_norm": 1.048620718784688, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 22926 + }, + { + "epoch": 0.22927, + "grad_norm": 1.1952304604470327, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 22927 + }, + { + "epoch": 0.22928, + "grad_norm": 1.0523645893340254, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 22928 + }, + { + "epoch": 0.22929, + "grad_norm": 1.1494222921610575, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 22929 + }, + { + "epoch": 0.2293, + "grad_norm": 1.0401928927318627, + "learning_rate": 0.003, + "loss": 4.025, + "step": 22930 + }, + { + "epoch": 0.22931, + "grad_norm": 1.0020682013966122, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 22931 + }, + { + "epoch": 0.22932, + "grad_norm": 0.9723330361614368, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 22932 + }, + { + "epoch": 0.22933, + "grad_norm": 0.9261326187723209, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 22933 + }, + { + "epoch": 0.22934, + "grad_norm": 0.9138221828184387, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 22934 + }, + { + "epoch": 0.22935, + "grad_norm": 1.0144587781740186, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 22935 + }, + { + "epoch": 0.22936, + "grad_norm": 0.9869385577954342, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 22936 + }, + { + "epoch": 0.22937, + "grad_norm": 1.0871914349013165, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 22937 + }, + { + "epoch": 0.22938, + "grad_norm": 1.1756722776065447, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 22938 + }, + { + "epoch": 0.22939, + "grad_norm": 1.152208478418597, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 22939 + }, + { + "epoch": 0.2294, + "grad_norm": 1.1594316064441164, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 22940 + }, + { + "epoch": 0.22941, + "grad_norm": 0.9250483881861781, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 22941 + }, + { + "epoch": 0.22942, + "grad_norm": 1.141217882003522, + "learning_rate": 0.003, + "loss": 4.019, + "step": 22942 + }, + { + "epoch": 0.22943, + "grad_norm": 1.0722765966943424, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 22943 + }, + { + "epoch": 0.22944, + "grad_norm": 1.0302497075369774, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 22944 + }, + { + "epoch": 0.22945, + "grad_norm": 1.1871681277156412, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 22945 + }, + { + "epoch": 0.22946, + "grad_norm": 0.9999705329306845, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 22946 + }, + { + "epoch": 0.22947, + "grad_norm": 1.2080601724168047, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 22947 + }, + { + "epoch": 0.22948, + "grad_norm": 1.0627626121047542, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 22948 + }, + { + "epoch": 0.22949, + "grad_norm": 1.044490136804917, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 22949 + }, + { + "epoch": 0.2295, + "grad_norm": 1.0942332257711702, + "learning_rate": 0.003, + "loss": 4.059, + "step": 22950 + }, + { + "epoch": 0.22951, + "grad_norm": 0.9803495895896501, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 22951 + }, + { + "epoch": 0.22952, + "grad_norm": 1.0723626578554561, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 22952 + }, + { + "epoch": 0.22953, + "grad_norm": 1.2356457891702575, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 22953 + }, + { + "epoch": 0.22954, + "grad_norm": 0.9797751685553377, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 22954 + }, + { + "epoch": 0.22955, + "grad_norm": 1.0150428034166952, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 22955 + }, + { + "epoch": 0.22956, + "grad_norm": 1.0744151255738221, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 22956 + }, + { + "epoch": 0.22957, + "grad_norm": 1.130233926867794, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 22957 + }, + { + "epoch": 0.22958, + "grad_norm": 1.01266280912793, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 22958 + }, + { + "epoch": 0.22959, + "grad_norm": 1.2027916588073841, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 22959 + }, + { + "epoch": 0.2296, + "grad_norm": 0.9938669192369084, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 22960 + }, + { + "epoch": 0.22961, + "grad_norm": 1.1138109915240904, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 22961 + }, + { + "epoch": 0.22962, + "grad_norm": 0.9114678652075974, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 22962 + }, + { + "epoch": 0.22963, + "grad_norm": 0.9692734144938564, + "learning_rate": 0.003, + "loss": 4.068, + "step": 22963 + }, + { + "epoch": 0.22964, + "grad_norm": 0.8411493147358887, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 22964 + }, + { + "epoch": 0.22965, + "grad_norm": 0.8618934653385096, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 22965 + }, + { + "epoch": 0.22966, + "grad_norm": 1.0738218980576655, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 22966 + }, + { + "epoch": 0.22967, + "grad_norm": 1.325000690018282, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 22967 + }, + { + "epoch": 0.22968, + "grad_norm": 1.008287654228617, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 22968 + }, + { + "epoch": 0.22969, + "grad_norm": 1.2048482278870043, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 22969 + }, + { + "epoch": 0.2297, + "grad_norm": 0.9321159893766899, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 22970 + }, + { + "epoch": 0.22971, + "grad_norm": 1.0463823002753065, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 22971 + }, + { + "epoch": 0.22972, + "grad_norm": 1.0913302869182766, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 22972 + }, + { + "epoch": 0.22973, + "grad_norm": 0.9808116633085656, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 22973 + }, + { + "epoch": 0.22974, + "grad_norm": 1.1584515108665903, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 22974 + }, + { + "epoch": 0.22975, + "grad_norm": 1.0149544296533424, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 22975 + }, + { + "epoch": 0.22976, + "grad_norm": 1.0538204597992338, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 22976 + }, + { + "epoch": 0.22977, + "grad_norm": 1.158772823840195, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 22977 + }, + { + "epoch": 0.22978, + "grad_norm": 0.957857788507489, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 22978 + }, + { + "epoch": 0.22979, + "grad_norm": 1.0652585065294828, + "learning_rate": 0.003, + "loss": 4.052, + "step": 22979 + }, + { + "epoch": 0.2298, + "grad_norm": 1.145901799525801, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 22980 + }, + { + "epoch": 0.22981, + "grad_norm": 1.0082984908284687, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 22981 + }, + { + "epoch": 0.22982, + "grad_norm": 0.9968504423650877, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 22982 + }, + { + "epoch": 0.22983, + "grad_norm": 0.9677776670695288, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 22983 + }, + { + "epoch": 0.22984, + "grad_norm": 1.1578533450802326, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 22984 + }, + { + "epoch": 0.22985, + "grad_norm": 1.011302012732689, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 22985 + }, + { + "epoch": 0.22986, + "grad_norm": 1.2686640556142987, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 22986 + }, + { + "epoch": 0.22987, + "grad_norm": 1.0863057420979207, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 22987 + }, + { + "epoch": 0.22988, + "grad_norm": 1.253742290594758, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 22988 + }, + { + "epoch": 0.22989, + "grad_norm": 0.9439190962780126, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 22989 + }, + { + "epoch": 0.2299, + "grad_norm": 1.1484684034187351, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 22990 + }, + { + "epoch": 0.22991, + "grad_norm": 1.0712764694942105, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 22991 + }, + { + "epoch": 0.22992, + "grad_norm": 1.1288558463442677, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 22992 + }, + { + "epoch": 0.22993, + "grad_norm": 1.253975296657525, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 22993 + }, + { + "epoch": 0.22994, + "grad_norm": 0.9703634967012335, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 22994 + }, + { + "epoch": 0.22995, + "grad_norm": 0.9871063950501945, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 22995 + }, + { + "epoch": 0.22996, + "grad_norm": 0.9502168159598257, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 22996 + }, + { + "epoch": 0.22997, + "grad_norm": 1.1043291905414765, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 22997 + }, + { + "epoch": 0.22998, + "grad_norm": 1.1086740388989749, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 22998 + }, + { + "epoch": 0.22999, + "grad_norm": 1.178540645252304, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 22999 + }, + { + "epoch": 0.23, + "grad_norm": 0.9722375821922435, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 23000 + }, + { + "epoch": 0.23001, + "grad_norm": 1.025149612740098, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 23001 + }, + { + "epoch": 0.23002, + "grad_norm": 1.0370248871569465, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 23002 + }, + { + "epoch": 0.23003, + "grad_norm": 0.9653547173875021, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 23003 + }, + { + "epoch": 0.23004, + "grad_norm": 1.3382142900283882, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 23004 + }, + { + "epoch": 0.23005, + "grad_norm": 0.7562719447102688, + "learning_rate": 0.003, + "loss": 4.071, + "step": 23005 + }, + { + "epoch": 0.23006, + "grad_norm": 0.7793552038138384, + "learning_rate": 0.003, + "loss": 4.063, + "step": 23006 + }, + { + "epoch": 0.23007, + "grad_norm": 0.8667555013067949, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 23007 + }, + { + "epoch": 0.23008, + "grad_norm": 1.0050313167590563, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 23008 + }, + { + "epoch": 0.23009, + "grad_norm": 1.1415997301818643, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 23009 + }, + { + "epoch": 0.2301, + "grad_norm": 0.9403140435913515, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 23010 + }, + { + "epoch": 0.23011, + "grad_norm": 1.1764838646658005, + "learning_rate": 0.003, + "loss": 4.062, + "step": 23011 + }, + { + "epoch": 0.23012, + "grad_norm": 1.1072868837856285, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 23012 + }, + { + "epoch": 0.23013, + "grad_norm": 1.165097664905161, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 23013 + }, + { + "epoch": 0.23014, + "grad_norm": 1.2805111807475102, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 23014 + }, + { + "epoch": 0.23015, + "grad_norm": 0.9573433491629244, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 23015 + }, + { + "epoch": 0.23016, + "grad_norm": 0.953054487496113, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 23016 + }, + { + "epoch": 0.23017, + "grad_norm": 1.0147237297113296, + "learning_rate": 0.003, + "loss": 4.062, + "step": 23017 + }, + { + "epoch": 0.23018, + "grad_norm": 1.2298567653823833, + "learning_rate": 0.003, + "loss": 4.029, + "step": 23018 + }, + { + "epoch": 0.23019, + "grad_norm": 1.1126334881362752, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 23019 + }, + { + "epoch": 0.2302, + "grad_norm": 1.3092882234339658, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 23020 + }, + { + "epoch": 0.23021, + "grad_norm": 1.0093755392673764, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 23021 + }, + { + "epoch": 0.23022, + "grad_norm": 0.9209914307640483, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 23022 + }, + { + "epoch": 0.23023, + "grad_norm": 0.9441815648354497, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 23023 + }, + { + "epoch": 0.23024, + "grad_norm": 0.9409456744567426, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 23024 + }, + { + "epoch": 0.23025, + "grad_norm": 1.0857767602649258, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 23025 + }, + { + "epoch": 0.23026, + "grad_norm": 1.1819022032126731, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 23026 + }, + { + "epoch": 0.23027, + "grad_norm": 1.1012679934837564, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 23027 + }, + { + "epoch": 0.23028, + "grad_norm": 1.0461079325176135, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 23028 + }, + { + "epoch": 0.23029, + "grad_norm": 1.1661978572458316, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 23029 + }, + { + "epoch": 0.2303, + "grad_norm": 0.9910961255779784, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 23030 + }, + { + "epoch": 0.23031, + "grad_norm": 1.3717486867168285, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 23031 + }, + { + "epoch": 0.23032, + "grad_norm": 0.8882222511863238, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 23032 + }, + { + "epoch": 0.23033, + "grad_norm": 1.0646988931280739, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 23033 + }, + { + "epoch": 0.23034, + "grad_norm": 1.1693404159594334, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 23034 + }, + { + "epoch": 0.23035, + "grad_norm": 1.2059938514161526, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 23035 + }, + { + "epoch": 0.23036, + "grad_norm": 1.1227960505666439, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 23036 + }, + { + "epoch": 0.23037, + "grad_norm": 0.9548337109857752, + "learning_rate": 0.003, + "loss": 4.079, + "step": 23037 + }, + { + "epoch": 0.23038, + "grad_norm": 0.893144541770601, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 23038 + }, + { + "epoch": 0.23039, + "grad_norm": 0.8868787618653565, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 23039 + }, + { + "epoch": 0.2304, + "grad_norm": 1.0986046953040953, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 23040 + }, + { + "epoch": 0.23041, + "grad_norm": 1.0858957181386737, + "learning_rate": 0.003, + "loss": 4.078, + "step": 23041 + }, + { + "epoch": 0.23042, + "grad_norm": 1.541023736317091, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 23042 + }, + { + "epoch": 0.23043, + "grad_norm": 0.8088445742761614, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 23043 + }, + { + "epoch": 0.23044, + "grad_norm": 0.9471589112936857, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 23044 + }, + { + "epoch": 0.23045, + "grad_norm": 1.167798126596269, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 23045 + }, + { + "epoch": 0.23046, + "grad_norm": 1.2454501852511157, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 23046 + }, + { + "epoch": 0.23047, + "grad_norm": 1.0527893415275735, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 23047 + }, + { + "epoch": 0.23048, + "grad_norm": 1.0642213885679732, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 23048 + }, + { + "epoch": 0.23049, + "grad_norm": 1.3106625485006722, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 23049 + }, + { + "epoch": 0.2305, + "grad_norm": 0.8610063332597675, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 23050 + }, + { + "epoch": 0.23051, + "grad_norm": 0.8854738471433293, + "learning_rate": 0.003, + "loss": 4.047, + "step": 23051 + }, + { + "epoch": 0.23052, + "grad_norm": 1.0933486734422937, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 23052 + }, + { + "epoch": 0.23053, + "grad_norm": 1.0559888733323155, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 23053 + }, + { + "epoch": 0.23054, + "grad_norm": 0.9511879972915817, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 23054 + }, + { + "epoch": 0.23055, + "grad_norm": 1.1061979110114208, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 23055 + }, + { + "epoch": 0.23056, + "grad_norm": 1.1947748746205804, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 23056 + }, + { + "epoch": 0.23057, + "grad_norm": 0.9824748870624569, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 23057 + }, + { + "epoch": 0.23058, + "grad_norm": 1.0428649431436274, + "learning_rate": 0.003, + "loss": 4.021, + "step": 23058 + }, + { + "epoch": 0.23059, + "grad_norm": 1.0346883824944166, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 23059 + }, + { + "epoch": 0.2306, + "grad_norm": 1.1028573032748423, + "learning_rate": 0.003, + "loss": 4.056, + "step": 23060 + }, + { + "epoch": 0.23061, + "grad_norm": 0.8562061898007992, + "learning_rate": 0.003, + "loss": 4.085, + "step": 23061 + }, + { + "epoch": 0.23062, + "grad_norm": 0.9788076670690179, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 23062 + }, + { + "epoch": 0.23063, + "grad_norm": 1.0858311836717391, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 23063 + }, + { + "epoch": 0.23064, + "grad_norm": 0.9698190491799236, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 23064 + }, + { + "epoch": 0.23065, + "grad_norm": 1.1239095787812046, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 23065 + }, + { + "epoch": 0.23066, + "grad_norm": 1.0666429038451464, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 23066 + }, + { + "epoch": 0.23067, + "grad_norm": 1.3650842984339948, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 23067 + }, + { + "epoch": 0.23068, + "grad_norm": 0.9336108690855124, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 23068 + }, + { + "epoch": 0.23069, + "grad_norm": 1.0324792128519071, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 23069 + }, + { + "epoch": 0.2307, + "grad_norm": 1.1133718289790655, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 23070 + }, + { + "epoch": 0.23071, + "grad_norm": 1.1647308848113653, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 23071 + }, + { + "epoch": 0.23072, + "grad_norm": 1.190644003269401, + "learning_rate": 0.003, + "loss": 4.045, + "step": 23072 + }, + { + "epoch": 0.23073, + "grad_norm": 1.1244886786363948, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 23073 + }, + { + "epoch": 0.23074, + "grad_norm": 1.1138308808968203, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 23074 + }, + { + "epoch": 0.23075, + "grad_norm": 1.240621844505005, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 23075 + }, + { + "epoch": 0.23076, + "grad_norm": 1.011293306870554, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 23076 + }, + { + "epoch": 0.23077, + "grad_norm": 1.0927475570938843, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 23077 + }, + { + "epoch": 0.23078, + "grad_norm": 0.9319585749779603, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 23078 + }, + { + "epoch": 0.23079, + "grad_norm": 0.9362618983447523, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 23079 + }, + { + "epoch": 0.2308, + "grad_norm": 1.052455858546931, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 23080 + }, + { + "epoch": 0.23081, + "grad_norm": 1.0625089605527087, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 23081 + }, + { + "epoch": 0.23082, + "grad_norm": 0.9355554615797105, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 23082 + }, + { + "epoch": 0.23083, + "grad_norm": 1.0605792332465904, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 23083 + }, + { + "epoch": 0.23084, + "grad_norm": 1.140656892499274, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 23084 + }, + { + "epoch": 0.23085, + "grad_norm": 0.9992260827394697, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 23085 + }, + { + "epoch": 0.23086, + "grad_norm": 1.2030509990182239, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 23086 + }, + { + "epoch": 0.23087, + "grad_norm": 1.186029273268042, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 23087 + }, + { + "epoch": 0.23088, + "grad_norm": 1.0053687322253577, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 23088 + }, + { + "epoch": 0.23089, + "grad_norm": 0.9789416996724749, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 23089 + }, + { + "epoch": 0.2309, + "grad_norm": 1.1607287972406788, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 23090 + }, + { + "epoch": 0.23091, + "grad_norm": 1.007846782376759, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 23091 + }, + { + "epoch": 0.23092, + "grad_norm": 1.0546221805556975, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 23092 + }, + { + "epoch": 0.23093, + "grad_norm": 1.3011984365712288, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 23093 + }, + { + "epoch": 0.23094, + "grad_norm": 1.1233484389324009, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 23094 + }, + { + "epoch": 0.23095, + "grad_norm": 1.0754994106523186, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 23095 + }, + { + "epoch": 0.23096, + "grad_norm": 1.0124212368852106, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 23096 + }, + { + "epoch": 0.23097, + "grad_norm": 1.1634696980775712, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 23097 + }, + { + "epoch": 0.23098, + "grad_norm": 1.0275161791409257, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 23098 + }, + { + "epoch": 0.23099, + "grad_norm": 1.1408075859944709, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 23099 + }, + { + "epoch": 0.231, + "grad_norm": 1.0695988447780593, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 23100 + }, + { + "epoch": 0.23101, + "grad_norm": 1.0357184509943844, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 23101 + }, + { + "epoch": 0.23102, + "grad_norm": 0.923711218439492, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 23102 + }, + { + "epoch": 0.23103, + "grad_norm": 1.0938946467880468, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 23103 + }, + { + "epoch": 0.23104, + "grad_norm": 1.1710636079380274, + "learning_rate": 0.003, + "loss": 4.06, + "step": 23104 + }, + { + "epoch": 0.23105, + "grad_norm": 1.1682217979125153, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 23105 + }, + { + "epoch": 0.23106, + "grad_norm": 0.9390867266926818, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 23106 + }, + { + "epoch": 0.23107, + "grad_norm": 0.9310995779261454, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 23107 + }, + { + "epoch": 0.23108, + "grad_norm": 1.1331529080803593, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 23108 + }, + { + "epoch": 0.23109, + "grad_norm": 1.0392659047354005, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 23109 + }, + { + "epoch": 0.2311, + "grad_norm": 1.1643878061867747, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 23110 + }, + { + "epoch": 0.23111, + "grad_norm": 0.9819810724760509, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 23111 + }, + { + "epoch": 0.23112, + "grad_norm": 1.024087505876934, + "learning_rate": 0.003, + "loss": 4.0932, + "step": 23112 + }, + { + "epoch": 0.23113, + "grad_norm": 1.0842346412467232, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 23113 + }, + { + "epoch": 0.23114, + "grad_norm": 1.0945513392562516, + "learning_rate": 0.003, + "loss": 4.029, + "step": 23114 + }, + { + "epoch": 0.23115, + "grad_norm": 1.0963240066790572, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 23115 + }, + { + "epoch": 0.23116, + "grad_norm": 1.065811350918353, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 23116 + }, + { + "epoch": 0.23117, + "grad_norm": 1.1162386016196757, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 23117 + }, + { + "epoch": 0.23118, + "grad_norm": 0.935582932335054, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 23118 + }, + { + "epoch": 0.23119, + "grad_norm": 1.0250760448266285, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 23119 + }, + { + "epoch": 0.2312, + "grad_norm": 1.3654057406977596, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 23120 + }, + { + "epoch": 0.23121, + "grad_norm": 0.9090867240307798, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 23121 + }, + { + "epoch": 0.23122, + "grad_norm": 0.9420987720824722, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 23122 + }, + { + "epoch": 0.23123, + "grad_norm": 0.9595977282441917, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 23123 + }, + { + "epoch": 0.23124, + "grad_norm": 1.0381982343307155, + "learning_rate": 0.003, + "loss": 4.058, + "step": 23124 + }, + { + "epoch": 0.23125, + "grad_norm": 1.119262939713756, + "learning_rate": 0.003, + "loss": 4.076, + "step": 23125 + }, + { + "epoch": 0.23126, + "grad_norm": 0.9791990712928906, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 23126 + }, + { + "epoch": 0.23127, + "grad_norm": 1.4166068129447913, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 23127 + }, + { + "epoch": 0.23128, + "grad_norm": 0.9172738155977773, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 23128 + }, + { + "epoch": 0.23129, + "grad_norm": 1.0306941846316313, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 23129 + }, + { + "epoch": 0.2313, + "grad_norm": 1.0516552402865527, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 23130 + }, + { + "epoch": 0.23131, + "grad_norm": 1.1305859046033226, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 23131 + }, + { + "epoch": 0.23132, + "grad_norm": 1.313588600227736, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 23132 + }, + { + "epoch": 0.23133, + "grad_norm": 0.979527166250691, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 23133 + }, + { + "epoch": 0.23134, + "grad_norm": 1.2134919792874443, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 23134 + }, + { + "epoch": 0.23135, + "grad_norm": 1.066100744959211, + "learning_rate": 0.003, + "loss": 4.073, + "step": 23135 + }, + { + "epoch": 0.23136, + "grad_norm": 1.1133356082102115, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 23136 + }, + { + "epoch": 0.23137, + "grad_norm": 0.9552527118393349, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 23137 + }, + { + "epoch": 0.23138, + "grad_norm": 1.0195078850369272, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 23138 + }, + { + "epoch": 0.23139, + "grad_norm": 1.0973046197197294, + "learning_rate": 0.003, + "loss": 4.0901, + "step": 23139 + }, + { + "epoch": 0.2314, + "grad_norm": 1.0005397045736615, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 23140 + }, + { + "epoch": 0.23141, + "grad_norm": 1.070528930755608, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 23141 + }, + { + "epoch": 0.23142, + "grad_norm": 1.0040044824246792, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 23142 + }, + { + "epoch": 0.23143, + "grad_norm": 1.2050599178579984, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 23143 + }, + { + "epoch": 0.23144, + "grad_norm": 1.017216325257727, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 23144 + }, + { + "epoch": 0.23145, + "grad_norm": 1.2460945603477995, + "learning_rate": 0.003, + "loss": 4.04, + "step": 23145 + }, + { + "epoch": 0.23146, + "grad_norm": 1.1032988497983622, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 23146 + }, + { + "epoch": 0.23147, + "grad_norm": 1.036315233953266, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 23147 + }, + { + "epoch": 0.23148, + "grad_norm": 0.9918414958866598, + "learning_rate": 0.003, + "loss": 4.0955, + "step": 23148 + }, + { + "epoch": 0.23149, + "grad_norm": 1.1335979970007297, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 23149 + }, + { + "epoch": 0.2315, + "grad_norm": 1.103915237930881, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 23150 + }, + { + "epoch": 0.23151, + "grad_norm": 1.0935036110350032, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 23151 + }, + { + "epoch": 0.23152, + "grad_norm": 1.0630190178816423, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 23152 + }, + { + "epoch": 0.23153, + "grad_norm": 1.2111409208679378, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 23153 + }, + { + "epoch": 0.23154, + "grad_norm": 1.1598867537868738, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 23154 + }, + { + "epoch": 0.23155, + "grad_norm": 1.0216901323722298, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 23155 + }, + { + "epoch": 0.23156, + "grad_norm": 1.1133716450871882, + "learning_rate": 0.003, + "loss": 4.032, + "step": 23156 + }, + { + "epoch": 0.23157, + "grad_norm": 1.0394621051656576, + "learning_rate": 0.003, + "loss": 4.063, + "step": 23157 + }, + { + "epoch": 0.23158, + "grad_norm": 1.1109251112982637, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 23158 + }, + { + "epoch": 0.23159, + "grad_norm": 0.9831724007601131, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 23159 + }, + { + "epoch": 0.2316, + "grad_norm": 1.1340853330721734, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 23160 + }, + { + "epoch": 0.23161, + "grad_norm": 0.8532803967592357, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 23161 + }, + { + "epoch": 0.23162, + "grad_norm": 0.8073921759553571, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 23162 + }, + { + "epoch": 0.23163, + "grad_norm": 0.8651027685985229, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 23163 + }, + { + "epoch": 0.23164, + "grad_norm": 1.0658368093915904, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 23164 + }, + { + "epoch": 0.23165, + "grad_norm": 0.9990906424083029, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 23165 + }, + { + "epoch": 0.23166, + "grad_norm": 1.0012622059186043, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 23166 + }, + { + "epoch": 0.23167, + "grad_norm": 1.2111031133878345, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 23167 + }, + { + "epoch": 0.23168, + "grad_norm": 0.9467391123793605, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 23168 + }, + { + "epoch": 0.23169, + "grad_norm": 1.1173183009314431, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 23169 + }, + { + "epoch": 0.2317, + "grad_norm": 1.1986633698234581, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 23170 + }, + { + "epoch": 0.23171, + "grad_norm": 1.2689140465157385, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 23171 + }, + { + "epoch": 0.23172, + "grad_norm": 0.9076264180512771, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 23172 + }, + { + "epoch": 0.23173, + "grad_norm": 0.9223468537242725, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 23173 + }, + { + "epoch": 0.23174, + "grad_norm": 1.1782531754348893, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 23174 + }, + { + "epoch": 0.23175, + "grad_norm": 0.991086808718113, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 23175 + }, + { + "epoch": 0.23176, + "grad_norm": 1.08905064053076, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 23176 + }, + { + "epoch": 0.23177, + "grad_norm": 1.1744403996200075, + "learning_rate": 0.003, + "loss": 4.055, + "step": 23177 + }, + { + "epoch": 0.23178, + "grad_norm": 0.8914134690538925, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 23178 + }, + { + "epoch": 0.23179, + "grad_norm": 0.8398350488703997, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 23179 + }, + { + "epoch": 0.2318, + "grad_norm": 0.9276067940757818, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 23180 + }, + { + "epoch": 0.23181, + "grad_norm": 0.9982570690006467, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 23181 + }, + { + "epoch": 0.23182, + "grad_norm": 1.2642145125323294, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 23182 + }, + { + "epoch": 0.23183, + "grad_norm": 0.9840202312136026, + "learning_rate": 0.003, + "loss": 4.07, + "step": 23183 + }, + { + "epoch": 0.23184, + "grad_norm": 1.3032819721439526, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 23184 + }, + { + "epoch": 0.23185, + "grad_norm": 0.9084997303772201, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 23185 + }, + { + "epoch": 0.23186, + "grad_norm": 1.1646832815835588, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 23186 + }, + { + "epoch": 0.23187, + "grad_norm": 1.2641934772116006, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 23187 + }, + { + "epoch": 0.23188, + "grad_norm": 0.9438063244440258, + "learning_rate": 0.003, + "loss": 4.099, + "step": 23188 + }, + { + "epoch": 0.23189, + "grad_norm": 1.0585439539107455, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 23189 + }, + { + "epoch": 0.2319, + "grad_norm": 1.1467578313549762, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 23190 + }, + { + "epoch": 0.23191, + "grad_norm": 1.0219039959579839, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 23191 + }, + { + "epoch": 0.23192, + "grad_norm": 1.230510816394557, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 23192 + }, + { + "epoch": 0.23193, + "grad_norm": 1.0255773128717187, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 23193 + }, + { + "epoch": 0.23194, + "grad_norm": 1.2706686978554007, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 23194 + }, + { + "epoch": 0.23195, + "grad_norm": 1.1081796886608786, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 23195 + }, + { + "epoch": 0.23196, + "grad_norm": 1.0756761982545358, + "learning_rate": 0.003, + "loss": 4.0819, + "step": 23196 + }, + { + "epoch": 0.23197, + "grad_norm": 1.1114354332457268, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 23197 + }, + { + "epoch": 0.23198, + "grad_norm": 0.9891400921180366, + "learning_rate": 0.003, + "loss": 4.04, + "step": 23198 + }, + { + "epoch": 0.23199, + "grad_norm": 1.0046175062085005, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 23199 + }, + { + "epoch": 0.232, + "grad_norm": 1.0141490802382194, + "learning_rate": 0.003, + "loss": 4.063, + "step": 23200 + }, + { + "epoch": 0.23201, + "grad_norm": 1.112343076863886, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 23201 + }, + { + "epoch": 0.23202, + "grad_norm": 0.9087359185055534, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 23202 + }, + { + "epoch": 0.23203, + "grad_norm": 1.213286967228212, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 23203 + }, + { + "epoch": 0.23204, + "grad_norm": 0.9440474291020265, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 23204 + }, + { + "epoch": 0.23205, + "grad_norm": 1.037633608699151, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 23205 + }, + { + "epoch": 0.23206, + "grad_norm": 1.311936405731073, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 23206 + }, + { + "epoch": 0.23207, + "grad_norm": 1.0395463991186182, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 23207 + }, + { + "epoch": 0.23208, + "grad_norm": 1.0875011319954218, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 23208 + }, + { + "epoch": 0.23209, + "grad_norm": 1.1673222056137604, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 23209 + }, + { + "epoch": 0.2321, + "grad_norm": 1.1414941864674921, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 23210 + }, + { + "epoch": 0.23211, + "grad_norm": 1.1547712156666075, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 23211 + }, + { + "epoch": 0.23212, + "grad_norm": 1.0099484091245763, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 23212 + }, + { + "epoch": 0.23213, + "grad_norm": 1.0027394482476562, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 23213 + }, + { + "epoch": 0.23214, + "grad_norm": 1.2339998841715643, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 23214 + }, + { + "epoch": 0.23215, + "grad_norm": 1.0961584799938844, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 23215 + }, + { + "epoch": 0.23216, + "grad_norm": 1.0396600506412232, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 23216 + }, + { + "epoch": 0.23217, + "grad_norm": 1.083867758641769, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 23217 + }, + { + "epoch": 0.23218, + "grad_norm": 1.0837467004466728, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 23218 + }, + { + "epoch": 0.23219, + "grad_norm": 1.044045188929117, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 23219 + }, + { + "epoch": 0.2322, + "grad_norm": 1.0290550756245642, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 23220 + }, + { + "epoch": 0.23221, + "grad_norm": 1.1083037656864458, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 23221 + }, + { + "epoch": 0.23222, + "grad_norm": 1.0274508148356079, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 23222 + }, + { + "epoch": 0.23223, + "grad_norm": 0.9937698175218881, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 23223 + }, + { + "epoch": 0.23224, + "grad_norm": 1.1797329556233802, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 23224 + }, + { + "epoch": 0.23225, + "grad_norm": 0.9806488894457404, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 23225 + }, + { + "epoch": 0.23226, + "grad_norm": 1.0656182195714081, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 23226 + }, + { + "epoch": 0.23227, + "grad_norm": 1.0661328774801475, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 23227 + }, + { + "epoch": 0.23228, + "grad_norm": 1.1514825300473297, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 23228 + }, + { + "epoch": 0.23229, + "grad_norm": 0.9772388576187193, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 23229 + }, + { + "epoch": 0.2323, + "grad_norm": 1.0716738241118662, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 23230 + }, + { + "epoch": 0.23231, + "grad_norm": 1.057956658275955, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 23231 + }, + { + "epoch": 0.23232, + "grad_norm": 1.1366086628409933, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 23232 + }, + { + "epoch": 0.23233, + "grad_norm": 0.8905385618193005, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 23233 + }, + { + "epoch": 0.23234, + "grad_norm": 0.8067496584539019, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 23234 + }, + { + "epoch": 0.23235, + "grad_norm": 0.8357406999574927, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 23235 + }, + { + "epoch": 0.23236, + "grad_norm": 1.0239028778803494, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 23236 + }, + { + "epoch": 0.23237, + "grad_norm": 1.1877579461617664, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 23237 + }, + { + "epoch": 0.23238, + "grad_norm": 0.9871155790247844, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 23238 + }, + { + "epoch": 0.23239, + "grad_norm": 1.0861262670744174, + "learning_rate": 0.003, + "loss": 4.0876, + "step": 23239 + }, + { + "epoch": 0.2324, + "grad_norm": 1.0030085029220785, + "learning_rate": 0.003, + "loss": 4.055, + "step": 23240 + }, + { + "epoch": 0.23241, + "grad_norm": 1.0572880111107825, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 23241 + }, + { + "epoch": 0.23242, + "grad_norm": 1.1261183688141614, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 23242 + }, + { + "epoch": 0.23243, + "grad_norm": 1.1683964289056492, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 23243 + }, + { + "epoch": 0.23244, + "grad_norm": 1.0899398907254365, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 23244 + }, + { + "epoch": 0.23245, + "grad_norm": 1.066094360832435, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 23245 + }, + { + "epoch": 0.23246, + "grad_norm": 1.1989757127310716, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 23246 + }, + { + "epoch": 0.23247, + "grad_norm": 0.9167186022901703, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 23247 + }, + { + "epoch": 0.23248, + "grad_norm": 1.1207672893192362, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 23248 + }, + { + "epoch": 0.23249, + "grad_norm": 1.294393129548955, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 23249 + }, + { + "epoch": 0.2325, + "grad_norm": 0.9011521617625725, + "learning_rate": 0.003, + "loss": 4.0873, + "step": 23250 + }, + { + "epoch": 0.23251, + "grad_norm": 0.9511370308544078, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 23251 + }, + { + "epoch": 0.23252, + "grad_norm": 1.1757987712928446, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 23252 + }, + { + "epoch": 0.23253, + "grad_norm": 0.9906856405469748, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 23253 + }, + { + "epoch": 0.23254, + "grad_norm": 1.225769124322137, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 23254 + }, + { + "epoch": 0.23255, + "grad_norm": 0.9934974725677224, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 23255 + }, + { + "epoch": 0.23256, + "grad_norm": 1.2557406274463185, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 23256 + }, + { + "epoch": 0.23257, + "grad_norm": 0.8175904155907018, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 23257 + }, + { + "epoch": 0.23258, + "grad_norm": 0.9371339076654251, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 23258 + }, + { + "epoch": 0.23259, + "grad_norm": 1.2043224705425535, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 23259 + }, + { + "epoch": 0.2326, + "grad_norm": 0.9681495079958585, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 23260 + }, + { + "epoch": 0.23261, + "grad_norm": 1.0057533130665934, + "learning_rate": 0.003, + "loss": 4.049, + "step": 23261 + }, + { + "epoch": 0.23262, + "grad_norm": 1.0839614970462763, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 23262 + }, + { + "epoch": 0.23263, + "grad_norm": 1.097992690426, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 23263 + }, + { + "epoch": 0.23264, + "grad_norm": 1.1179041379201653, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 23264 + }, + { + "epoch": 0.23265, + "grad_norm": 0.8661487144888289, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 23265 + }, + { + "epoch": 0.23266, + "grad_norm": 0.9157691175967809, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 23266 + }, + { + "epoch": 0.23267, + "grad_norm": 0.9751858489149507, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 23267 + }, + { + "epoch": 0.23268, + "grad_norm": 1.2903485617475061, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 23268 + }, + { + "epoch": 0.23269, + "grad_norm": 1.3494634699541161, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 23269 + }, + { + "epoch": 0.2327, + "grad_norm": 0.98379440588935, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 23270 + }, + { + "epoch": 0.23271, + "grad_norm": 0.9976072213671036, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 23271 + }, + { + "epoch": 0.23272, + "grad_norm": 1.1756361904204502, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 23272 + }, + { + "epoch": 0.23273, + "grad_norm": 1.1397951857688118, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 23273 + }, + { + "epoch": 0.23274, + "grad_norm": 1.1880480746591116, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 23274 + }, + { + "epoch": 0.23275, + "grad_norm": 1.2148820853330589, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 23275 + }, + { + "epoch": 0.23276, + "grad_norm": 0.9361919344787414, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 23276 + }, + { + "epoch": 0.23277, + "grad_norm": 1.127663220593326, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 23277 + }, + { + "epoch": 0.23278, + "grad_norm": 1.1167286826996226, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 23278 + }, + { + "epoch": 0.23279, + "grad_norm": 1.026396024734567, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 23279 + }, + { + "epoch": 0.2328, + "grad_norm": 1.1701143742170492, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 23280 + }, + { + "epoch": 0.23281, + "grad_norm": 0.967541192486795, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 23281 + }, + { + "epoch": 0.23282, + "grad_norm": 0.9164460366260724, + "learning_rate": 0.003, + "loss": 4.043, + "step": 23282 + }, + { + "epoch": 0.23283, + "grad_norm": 0.9387104839480317, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 23283 + }, + { + "epoch": 0.23284, + "grad_norm": 1.164991750225874, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 23284 + }, + { + "epoch": 0.23285, + "grad_norm": 1.0884489412826945, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 23285 + }, + { + "epoch": 0.23286, + "grad_norm": 0.9532368675645632, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 23286 + }, + { + "epoch": 0.23287, + "grad_norm": 1.1014684468503042, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 23287 + }, + { + "epoch": 0.23288, + "grad_norm": 1.06035998251999, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 23288 + }, + { + "epoch": 0.23289, + "grad_norm": 1.0089324518641782, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 23289 + }, + { + "epoch": 0.2329, + "grad_norm": 1.2359538585519076, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 23290 + }, + { + "epoch": 0.23291, + "grad_norm": 0.8474387499236767, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 23291 + }, + { + "epoch": 0.23292, + "grad_norm": 0.8092569490192474, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 23292 + }, + { + "epoch": 0.23293, + "grad_norm": 0.8636980160954677, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 23293 + }, + { + "epoch": 0.23294, + "grad_norm": 0.9822980214742939, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 23294 + }, + { + "epoch": 0.23295, + "grad_norm": 1.0485859814663188, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 23295 + }, + { + "epoch": 0.23296, + "grad_norm": 1.302153430904368, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 23296 + }, + { + "epoch": 0.23297, + "grad_norm": 1.1897591138725538, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 23297 + }, + { + "epoch": 0.23298, + "grad_norm": 0.9007319235618058, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 23298 + }, + { + "epoch": 0.23299, + "grad_norm": 0.9605275469959805, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 23299 + }, + { + "epoch": 0.233, + "grad_norm": 1.6335364354875108, + "learning_rate": 0.003, + "loss": 4.1144, + "step": 23300 + }, + { + "epoch": 0.23301, + "grad_norm": 0.8434647478115622, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 23301 + }, + { + "epoch": 0.23302, + "grad_norm": 1.0731560170090264, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 23302 + }, + { + "epoch": 0.23303, + "grad_norm": 1.2947515961996936, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 23303 + }, + { + "epoch": 0.23304, + "grad_norm": 0.9700780317570465, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 23304 + }, + { + "epoch": 0.23305, + "grad_norm": 1.1385996080815721, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 23305 + }, + { + "epoch": 0.23306, + "grad_norm": 1.0014561958997144, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 23306 + }, + { + "epoch": 0.23307, + "grad_norm": 1.3414588396163432, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 23307 + }, + { + "epoch": 0.23308, + "grad_norm": 0.896520273215258, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 23308 + }, + { + "epoch": 0.23309, + "grad_norm": 1.0142093136528723, + "learning_rate": 0.003, + "loss": 4.075, + "step": 23309 + }, + { + "epoch": 0.2331, + "grad_norm": 1.2909386767787028, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 23310 + }, + { + "epoch": 0.23311, + "grad_norm": 1.0481776227365667, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 23311 + }, + { + "epoch": 0.23312, + "grad_norm": 1.4086026399696814, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 23312 + }, + { + "epoch": 0.23313, + "grad_norm": 0.9824354969005721, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 23313 + }, + { + "epoch": 0.23314, + "grad_norm": 0.9722438282185253, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 23314 + }, + { + "epoch": 0.23315, + "grad_norm": 1.2645789303878971, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 23315 + }, + { + "epoch": 0.23316, + "grad_norm": 1.2533103894973028, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 23316 + }, + { + "epoch": 0.23317, + "grad_norm": 1.047356493965884, + "learning_rate": 0.003, + "loss": 4.0945, + "step": 23317 + }, + { + "epoch": 0.23318, + "grad_norm": 0.9210488372697351, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 23318 + }, + { + "epoch": 0.23319, + "grad_norm": 1.0858012126984034, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 23319 + }, + { + "epoch": 0.2332, + "grad_norm": 1.0652818527648464, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 23320 + }, + { + "epoch": 0.23321, + "grad_norm": 0.931646812829688, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 23321 + }, + { + "epoch": 0.23322, + "grad_norm": 1.0053273765336548, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 23322 + }, + { + "epoch": 0.23323, + "grad_norm": 1.0308882411565996, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 23323 + }, + { + "epoch": 0.23324, + "grad_norm": 1.3513530622983758, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 23324 + }, + { + "epoch": 0.23325, + "grad_norm": 0.9764121604099635, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 23325 + }, + { + "epoch": 0.23326, + "grad_norm": 1.043784483722594, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 23326 + }, + { + "epoch": 0.23327, + "grad_norm": 1.071810296025223, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 23327 + }, + { + "epoch": 0.23328, + "grad_norm": 1.1234462707277517, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 23328 + }, + { + "epoch": 0.23329, + "grad_norm": 1.1000895525535423, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 23329 + }, + { + "epoch": 0.2333, + "grad_norm": 1.1002067143371825, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 23330 + }, + { + "epoch": 0.23331, + "grad_norm": 0.9210630550030137, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 23331 + }, + { + "epoch": 0.23332, + "grad_norm": 0.9526892088727477, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 23332 + }, + { + "epoch": 0.23333, + "grad_norm": 1.157224688783885, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 23333 + }, + { + "epoch": 0.23334, + "grad_norm": 1.2724222448198343, + "learning_rate": 0.003, + "loss": 4.041, + "step": 23334 + }, + { + "epoch": 0.23335, + "grad_norm": 0.824522810171713, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 23335 + }, + { + "epoch": 0.23336, + "grad_norm": 0.9652988478580697, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 23336 + }, + { + "epoch": 0.23337, + "grad_norm": 1.202878880012682, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 23337 + }, + { + "epoch": 0.23338, + "grad_norm": 1.2653114920711364, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 23338 + }, + { + "epoch": 0.23339, + "grad_norm": 0.9830570593818382, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 23339 + }, + { + "epoch": 0.2334, + "grad_norm": 0.9856728813578017, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 23340 + }, + { + "epoch": 0.23341, + "grad_norm": 1.0596701865158977, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 23341 + }, + { + "epoch": 0.23342, + "grad_norm": 1.1219601611246444, + "learning_rate": 0.003, + "loss": 4.036, + "step": 23342 + }, + { + "epoch": 0.23343, + "grad_norm": 1.0148114704451738, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 23343 + }, + { + "epoch": 0.23344, + "grad_norm": 1.0492161013290748, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 23344 + }, + { + "epoch": 0.23345, + "grad_norm": 1.02530636463363, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 23345 + }, + { + "epoch": 0.23346, + "grad_norm": 1.2628046873995327, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 23346 + }, + { + "epoch": 0.23347, + "grad_norm": 1.0761806350707914, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 23347 + }, + { + "epoch": 0.23348, + "grad_norm": 1.3509689018886173, + "learning_rate": 0.003, + "loss": 4.077, + "step": 23348 + }, + { + "epoch": 0.23349, + "grad_norm": 0.8282639961393438, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 23349 + }, + { + "epoch": 0.2335, + "grad_norm": 0.8075912789843954, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 23350 + }, + { + "epoch": 0.23351, + "grad_norm": 0.8567297855003785, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 23351 + }, + { + "epoch": 0.23352, + "grad_norm": 1.0630871942978355, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 23352 + }, + { + "epoch": 0.23353, + "grad_norm": 1.3265048077060029, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 23353 + }, + { + "epoch": 0.23354, + "grad_norm": 0.9636561808574756, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 23354 + }, + { + "epoch": 0.23355, + "grad_norm": 1.0837289332981568, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 23355 + }, + { + "epoch": 0.23356, + "grad_norm": 1.383860808949056, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 23356 + }, + { + "epoch": 0.23357, + "grad_norm": 0.9552128039584551, + "learning_rate": 0.003, + "loss": 4.075, + "step": 23357 + }, + { + "epoch": 0.23358, + "grad_norm": 0.9301966703436234, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 23358 + }, + { + "epoch": 0.23359, + "grad_norm": 0.988759815802235, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 23359 + }, + { + "epoch": 0.2336, + "grad_norm": 1.0843317274737967, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 23360 + }, + { + "epoch": 0.23361, + "grad_norm": 1.032180221213396, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 23361 + }, + { + "epoch": 0.23362, + "grad_norm": 1.1480939188043875, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 23362 + }, + { + "epoch": 0.23363, + "grad_norm": 1.0212389490428089, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 23363 + }, + { + "epoch": 0.23364, + "grad_norm": 1.1790180827143864, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 23364 + }, + { + "epoch": 0.23365, + "grad_norm": 1.156253031023978, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 23365 + }, + { + "epoch": 0.23366, + "grad_norm": 1.1862177790527397, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 23366 + }, + { + "epoch": 0.23367, + "grad_norm": 1.0143890656938566, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 23367 + }, + { + "epoch": 0.23368, + "grad_norm": 1.1821741804888835, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 23368 + }, + { + "epoch": 0.23369, + "grad_norm": 0.9671886074514608, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 23369 + }, + { + "epoch": 0.2337, + "grad_norm": 1.2447298221171221, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 23370 + }, + { + "epoch": 0.23371, + "grad_norm": 1.088767165246868, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 23371 + }, + { + "epoch": 0.23372, + "grad_norm": 1.1010418368260242, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 23372 + }, + { + "epoch": 0.23373, + "grad_norm": 1.029346658396718, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 23373 + }, + { + "epoch": 0.23374, + "grad_norm": 1.0879602527429764, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 23374 + }, + { + "epoch": 0.23375, + "grad_norm": 0.9579486056437087, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 23375 + }, + { + "epoch": 0.23376, + "grad_norm": 1.1825711319467291, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 23376 + }, + { + "epoch": 0.23377, + "grad_norm": 0.8367333265204594, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 23377 + }, + { + "epoch": 0.23378, + "grad_norm": 0.8599179183506531, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 23378 + }, + { + "epoch": 0.23379, + "grad_norm": 0.9196973829265874, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 23379 + }, + { + "epoch": 0.2338, + "grad_norm": 1.533727451943196, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 23380 + }, + { + "epoch": 0.23381, + "grad_norm": 1.0631251129030261, + "learning_rate": 0.003, + "loss": 4.068, + "step": 23381 + }, + { + "epoch": 0.23382, + "grad_norm": 0.9773900558348693, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 23382 + }, + { + "epoch": 0.23383, + "grad_norm": 1.1360861951920698, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 23383 + }, + { + "epoch": 0.23384, + "grad_norm": 1.0957299607396642, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 23384 + }, + { + "epoch": 0.23385, + "grad_norm": 0.9544578190840922, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 23385 + }, + { + "epoch": 0.23386, + "grad_norm": 0.9644297725129956, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 23386 + }, + { + "epoch": 0.23387, + "grad_norm": 1.1139800299797182, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 23387 + }, + { + "epoch": 0.23388, + "grad_norm": 1.1139204057611027, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 23388 + }, + { + "epoch": 0.23389, + "grad_norm": 0.9148773462458125, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 23389 + }, + { + "epoch": 0.2339, + "grad_norm": 0.9842732716117093, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 23390 + }, + { + "epoch": 0.23391, + "grad_norm": 1.1720537994980547, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 23391 + }, + { + "epoch": 0.23392, + "grad_norm": 0.8301287722775733, + "learning_rate": 0.003, + "loss": 4.0855, + "step": 23392 + }, + { + "epoch": 0.23393, + "grad_norm": 1.1353856963755358, + "learning_rate": 0.003, + "loss": 4.0946, + "step": 23393 + }, + { + "epoch": 0.23394, + "grad_norm": 1.0377872887302493, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 23394 + }, + { + "epoch": 0.23395, + "grad_norm": 1.1200727305519933, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 23395 + }, + { + "epoch": 0.23396, + "grad_norm": 1.1338443222431087, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 23396 + }, + { + "epoch": 0.23397, + "grad_norm": 1.093273702631995, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 23397 + }, + { + "epoch": 0.23398, + "grad_norm": 1.0546992791538066, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 23398 + }, + { + "epoch": 0.23399, + "grad_norm": 1.1867918080730908, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 23399 + }, + { + "epoch": 0.234, + "grad_norm": 1.093985764014062, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 23400 + }, + { + "epoch": 0.23401, + "grad_norm": 1.1298911504478302, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 23401 + }, + { + "epoch": 0.23402, + "grad_norm": 0.8867635695311883, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 23402 + }, + { + "epoch": 0.23403, + "grad_norm": 0.8998422973288971, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 23403 + }, + { + "epoch": 0.23404, + "grad_norm": 0.9942810304138244, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 23404 + }, + { + "epoch": 0.23405, + "grad_norm": 1.3269050861586684, + "learning_rate": 0.003, + "loss": 4.055, + "step": 23405 + }, + { + "epoch": 0.23406, + "grad_norm": 0.8233991559481522, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 23406 + }, + { + "epoch": 0.23407, + "grad_norm": 0.8734288282120704, + "learning_rate": 0.003, + "loss": 4.055, + "step": 23407 + }, + { + "epoch": 0.23408, + "grad_norm": 1.1156491935267336, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 23408 + }, + { + "epoch": 0.23409, + "grad_norm": 1.0221309607262126, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 23409 + }, + { + "epoch": 0.2341, + "grad_norm": 1.1330728751348726, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 23410 + }, + { + "epoch": 0.23411, + "grad_norm": 1.2142994313901114, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 23411 + }, + { + "epoch": 0.23412, + "grad_norm": 0.9367588104802832, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 23412 + }, + { + "epoch": 0.23413, + "grad_norm": 0.9265582643001952, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 23413 + }, + { + "epoch": 0.23414, + "grad_norm": 1.1197201417421865, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 23414 + }, + { + "epoch": 0.23415, + "grad_norm": 1.104591831417941, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 23415 + }, + { + "epoch": 0.23416, + "grad_norm": 1.2467452323584045, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 23416 + }, + { + "epoch": 0.23417, + "grad_norm": 0.9968604779209335, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 23417 + }, + { + "epoch": 0.23418, + "grad_norm": 1.509781130402687, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 23418 + }, + { + "epoch": 0.23419, + "grad_norm": 0.8729392391137132, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 23419 + }, + { + "epoch": 0.2342, + "grad_norm": 0.9225408597240045, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 23420 + }, + { + "epoch": 0.23421, + "grad_norm": 1.1740979327146226, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 23421 + }, + { + "epoch": 0.23422, + "grad_norm": 1.1234082015680658, + "learning_rate": 0.003, + "loss": 4.045, + "step": 23422 + }, + { + "epoch": 0.23423, + "grad_norm": 1.1233391999579245, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 23423 + }, + { + "epoch": 0.23424, + "grad_norm": 1.000110760603854, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 23424 + }, + { + "epoch": 0.23425, + "grad_norm": 1.0984021341540569, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 23425 + }, + { + "epoch": 0.23426, + "grad_norm": 1.047913363399878, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 23426 + }, + { + "epoch": 0.23427, + "grad_norm": 1.0740433493781008, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 23427 + }, + { + "epoch": 0.23428, + "grad_norm": 1.0876155078184555, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 23428 + }, + { + "epoch": 0.23429, + "grad_norm": 1.0078202331214268, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 23429 + }, + { + "epoch": 0.2343, + "grad_norm": 1.106019557784853, + "learning_rate": 0.003, + "loss": 4.041, + "step": 23430 + }, + { + "epoch": 0.23431, + "grad_norm": 1.1078376649735358, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 23431 + }, + { + "epoch": 0.23432, + "grad_norm": 1.1890742896390034, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 23432 + }, + { + "epoch": 0.23433, + "grad_norm": 0.9411558827751707, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 23433 + }, + { + "epoch": 0.23434, + "grad_norm": 1.0461176681857616, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 23434 + }, + { + "epoch": 0.23435, + "grad_norm": 1.1746932055634043, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 23435 + }, + { + "epoch": 0.23436, + "grad_norm": 0.9723144572336818, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 23436 + }, + { + "epoch": 0.23437, + "grad_norm": 1.3137584220550291, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 23437 + }, + { + "epoch": 0.23438, + "grad_norm": 0.9516584798584137, + "learning_rate": 0.003, + "loss": 4.1081, + "step": 23438 + }, + { + "epoch": 0.23439, + "grad_norm": 0.9378252357820119, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 23439 + }, + { + "epoch": 0.2344, + "grad_norm": 1.2457251640033264, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 23440 + }, + { + "epoch": 0.23441, + "grad_norm": 0.9709732476014801, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 23441 + }, + { + "epoch": 0.23442, + "grad_norm": 0.9222264737649798, + "learning_rate": 0.003, + "loss": 4.08, + "step": 23442 + }, + { + "epoch": 0.23443, + "grad_norm": 1.056164811914167, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 23443 + }, + { + "epoch": 0.23444, + "grad_norm": 1.1665956263383699, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 23444 + }, + { + "epoch": 0.23445, + "grad_norm": 1.0008981484035766, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 23445 + }, + { + "epoch": 0.23446, + "grad_norm": 1.0678470210815496, + "learning_rate": 0.003, + "loss": 4.0858, + "step": 23446 + }, + { + "epoch": 0.23447, + "grad_norm": 1.0643404412152064, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 23447 + }, + { + "epoch": 0.23448, + "grad_norm": 1.1032524645655748, + "learning_rate": 0.003, + "loss": 4.031, + "step": 23448 + }, + { + "epoch": 0.23449, + "grad_norm": 0.9061698903509238, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 23449 + }, + { + "epoch": 0.2345, + "grad_norm": 1.1607818694322503, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 23450 + }, + { + "epoch": 0.23451, + "grad_norm": 1.214922144566054, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 23451 + }, + { + "epoch": 0.23452, + "grad_norm": 1.0181117597664378, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 23452 + }, + { + "epoch": 0.23453, + "grad_norm": 1.1819719515630704, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 23453 + }, + { + "epoch": 0.23454, + "grad_norm": 0.9776654412837813, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 23454 + }, + { + "epoch": 0.23455, + "grad_norm": 1.3052415834050175, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 23455 + }, + { + "epoch": 0.23456, + "grad_norm": 0.9274490653813234, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 23456 + }, + { + "epoch": 0.23457, + "grad_norm": 0.9404165301367674, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 23457 + }, + { + "epoch": 0.23458, + "grad_norm": 1.2091565742513213, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 23458 + }, + { + "epoch": 0.23459, + "grad_norm": 1.0276916449577682, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 23459 + }, + { + "epoch": 0.2346, + "grad_norm": 1.041179797642017, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 23460 + }, + { + "epoch": 0.23461, + "grad_norm": 1.0460449775708476, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 23461 + }, + { + "epoch": 0.23462, + "grad_norm": 1.3302836428754803, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 23462 + }, + { + "epoch": 0.23463, + "grad_norm": 0.8967052051221618, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 23463 + }, + { + "epoch": 0.23464, + "grad_norm": 1.0219210271397918, + "learning_rate": 0.003, + "loss": 4.062, + "step": 23464 + }, + { + "epoch": 0.23465, + "grad_norm": 1.0564265249252731, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 23465 + }, + { + "epoch": 0.23466, + "grad_norm": 1.1748340337380585, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 23466 + }, + { + "epoch": 0.23467, + "grad_norm": 1.0173538529940016, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 23467 + }, + { + "epoch": 0.23468, + "grad_norm": 1.0985414226135624, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 23468 + }, + { + "epoch": 0.23469, + "grad_norm": 1.060657173180982, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 23469 + }, + { + "epoch": 0.2347, + "grad_norm": 1.0836146037261283, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 23470 + }, + { + "epoch": 0.23471, + "grad_norm": 1.1716373514071767, + "learning_rate": 0.003, + "loss": 4.072, + "step": 23471 + }, + { + "epoch": 0.23472, + "grad_norm": 0.853075996471349, + "learning_rate": 0.003, + "loss": 4.044, + "step": 23472 + }, + { + "epoch": 0.23473, + "grad_norm": 0.8890788828107103, + "learning_rate": 0.003, + "loss": 4.048, + "step": 23473 + }, + { + "epoch": 0.23474, + "grad_norm": 1.0106190290347543, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 23474 + }, + { + "epoch": 0.23475, + "grad_norm": 1.1176315763021085, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 23475 + }, + { + "epoch": 0.23476, + "grad_norm": 1.001507257864232, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 23476 + }, + { + "epoch": 0.23477, + "grad_norm": 1.1554920467330316, + "learning_rate": 0.003, + "loss": 4.065, + "step": 23477 + }, + { + "epoch": 0.23478, + "grad_norm": 1.0726274844268984, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 23478 + }, + { + "epoch": 0.23479, + "grad_norm": 1.0093146228042342, + "learning_rate": 0.003, + "loss": 4.062, + "step": 23479 + }, + { + "epoch": 0.2348, + "grad_norm": 1.073764214543599, + "learning_rate": 0.003, + "loss": 4.1114, + "step": 23480 + }, + { + "epoch": 0.23481, + "grad_norm": 1.00565088233453, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 23481 + }, + { + "epoch": 0.23482, + "grad_norm": 1.030496033704009, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 23482 + }, + { + "epoch": 0.23483, + "grad_norm": 0.9742380239494429, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 23483 + }, + { + "epoch": 0.23484, + "grad_norm": 1.3334096627652519, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 23484 + }, + { + "epoch": 0.23485, + "grad_norm": 0.8042936708920002, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 23485 + }, + { + "epoch": 0.23486, + "grad_norm": 1.1019251274460784, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 23486 + }, + { + "epoch": 0.23487, + "grad_norm": 1.4083826556078827, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 23487 + }, + { + "epoch": 0.23488, + "grad_norm": 1.0295030108574317, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 23488 + }, + { + "epoch": 0.23489, + "grad_norm": 1.1872651632356204, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 23489 + }, + { + "epoch": 0.2349, + "grad_norm": 1.2160309230989739, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 23490 + }, + { + "epoch": 0.23491, + "grad_norm": 1.0981438392515879, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 23491 + }, + { + "epoch": 0.23492, + "grad_norm": 1.0490755365232907, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 23492 + }, + { + "epoch": 0.23493, + "grad_norm": 1.0469655307799723, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 23493 + }, + { + "epoch": 0.23494, + "grad_norm": 1.146374869303461, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 23494 + }, + { + "epoch": 0.23495, + "grad_norm": 0.9907171345519402, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 23495 + }, + { + "epoch": 0.23496, + "grad_norm": 0.992010292800779, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 23496 + }, + { + "epoch": 0.23497, + "grad_norm": 1.2962539119186627, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 23497 + }, + { + "epoch": 0.23498, + "grad_norm": 0.8269237341321786, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 23498 + }, + { + "epoch": 0.23499, + "grad_norm": 0.9569894753453498, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 23499 + }, + { + "epoch": 0.235, + "grad_norm": 0.8623599075284318, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 23500 + }, + { + "epoch": 0.23501, + "grad_norm": 1.10329407173948, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 23501 + }, + { + "epoch": 0.23502, + "grad_norm": 1.428182344856015, + "learning_rate": 0.003, + "loss": 4.072, + "step": 23502 + }, + { + "epoch": 0.23503, + "grad_norm": 0.7757731078710093, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 23503 + }, + { + "epoch": 0.23504, + "grad_norm": 0.821728670554105, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 23504 + }, + { + "epoch": 0.23505, + "grad_norm": 0.950719408978859, + "learning_rate": 0.003, + "loss": 4.0955, + "step": 23505 + }, + { + "epoch": 0.23506, + "grad_norm": 1.2356618347981159, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 23506 + }, + { + "epoch": 0.23507, + "grad_norm": 0.8489218826633983, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 23507 + }, + { + "epoch": 0.23508, + "grad_norm": 1.1228972697264972, + "learning_rate": 0.003, + "loss": 4.0898, + "step": 23508 + }, + { + "epoch": 0.23509, + "grad_norm": 1.1870075341037234, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 23509 + }, + { + "epoch": 0.2351, + "grad_norm": 0.8398455970548591, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 23510 + }, + { + "epoch": 0.23511, + "grad_norm": 0.8242852487376205, + "learning_rate": 0.003, + "loss": 4.053, + "step": 23511 + }, + { + "epoch": 0.23512, + "grad_norm": 0.9615322075008323, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 23512 + }, + { + "epoch": 0.23513, + "grad_norm": 1.2430427614890378, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 23513 + }, + { + "epoch": 0.23514, + "grad_norm": 0.9792920114192762, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 23514 + }, + { + "epoch": 0.23515, + "grad_norm": 0.9613949238865114, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 23515 + }, + { + "epoch": 0.23516, + "grad_norm": 1.1882329952226038, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 23516 + }, + { + "epoch": 0.23517, + "grad_norm": 1.0659443750906297, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 23517 + }, + { + "epoch": 0.23518, + "grad_norm": 1.0513209345248276, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 23518 + }, + { + "epoch": 0.23519, + "grad_norm": 1.211528622867414, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 23519 + }, + { + "epoch": 0.2352, + "grad_norm": 1.3267832915017088, + "learning_rate": 0.003, + "loss": 4.1004, + "step": 23520 + }, + { + "epoch": 0.23521, + "grad_norm": 0.9034913182852142, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 23521 + }, + { + "epoch": 0.23522, + "grad_norm": 0.8656686682002407, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 23522 + }, + { + "epoch": 0.23523, + "grad_norm": 0.9976644360747943, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 23523 + }, + { + "epoch": 0.23524, + "grad_norm": 1.5678814754121182, + "learning_rate": 0.003, + "loss": 4.1161, + "step": 23524 + }, + { + "epoch": 0.23525, + "grad_norm": 0.7430911932072025, + "learning_rate": 0.003, + "loss": 4.062, + "step": 23525 + }, + { + "epoch": 0.23526, + "grad_norm": 0.8742679319084571, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 23526 + }, + { + "epoch": 0.23527, + "grad_norm": 0.9969362169376504, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 23527 + }, + { + "epoch": 0.23528, + "grad_norm": 1.1821063567802572, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 23528 + }, + { + "epoch": 0.23529, + "grad_norm": 0.9754004095250101, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 23529 + }, + { + "epoch": 0.2353, + "grad_norm": 1.1202349001819414, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 23530 + }, + { + "epoch": 0.23531, + "grad_norm": 1.0712233123648724, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 23531 + }, + { + "epoch": 0.23532, + "grad_norm": 1.1639769674358054, + "learning_rate": 0.003, + "loss": 4.0916, + "step": 23532 + }, + { + "epoch": 0.23533, + "grad_norm": 1.117354533507665, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 23533 + }, + { + "epoch": 0.23534, + "grad_norm": 1.0458394818172345, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 23534 + }, + { + "epoch": 0.23535, + "grad_norm": 1.1690569109975768, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 23535 + }, + { + "epoch": 0.23536, + "grad_norm": 1.0140305969452317, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 23536 + }, + { + "epoch": 0.23537, + "grad_norm": 1.3056746689879422, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 23537 + }, + { + "epoch": 0.23538, + "grad_norm": 1.0544458981748615, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 23538 + }, + { + "epoch": 0.23539, + "grad_norm": 1.06048605886552, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 23539 + }, + { + "epoch": 0.2354, + "grad_norm": 1.0220231950297236, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 23540 + }, + { + "epoch": 0.23541, + "grad_norm": 1.0414942824869606, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 23541 + }, + { + "epoch": 0.23542, + "grad_norm": 0.9582303363679554, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 23542 + }, + { + "epoch": 0.23543, + "grad_norm": 1.0609730017811372, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 23543 + }, + { + "epoch": 0.23544, + "grad_norm": 1.1082741230936484, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 23544 + }, + { + "epoch": 0.23545, + "grad_norm": 1.023413101162999, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 23545 + }, + { + "epoch": 0.23546, + "grad_norm": 1.0408997736026981, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 23546 + }, + { + "epoch": 0.23547, + "grad_norm": 1.249931445330235, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 23547 + }, + { + "epoch": 0.23548, + "grad_norm": 1.0698298876321286, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 23548 + }, + { + "epoch": 0.23549, + "grad_norm": 1.0327841173729044, + "learning_rate": 0.003, + "loss": 4.067, + "step": 23549 + }, + { + "epoch": 0.2355, + "grad_norm": 0.9045755706587648, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 23550 + }, + { + "epoch": 0.23551, + "grad_norm": 1.0201380520932704, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 23551 + }, + { + "epoch": 0.23552, + "grad_norm": 1.3597669906647236, + "learning_rate": 0.003, + "loss": 4.056, + "step": 23552 + }, + { + "epoch": 0.23553, + "grad_norm": 0.9928226862024238, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 23553 + }, + { + "epoch": 0.23554, + "grad_norm": 1.1541720342745825, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 23554 + }, + { + "epoch": 0.23555, + "grad_norm": 1.0578778535609665, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 23555 + }, + { + "epoch": 0.23556, + "grad_norm": 1.021938868215982, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 23556 + }, + { + "epoch": 0.23557, + "grad_norm": 1.0542771145357601, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 23557 + }, + { + "epoch": 0.23558, + "grad_norm": 1.0925001112046735, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 23558 + }, + { + "epoch": 0.23559, + "grad_norm": 0.9447575889837548, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 23559 + }, + { + "epoch": 0.2356, + "grad_norm": 0.9335602156379257, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 23560 + }, + { + "epoch": 0.23561, + "grad_norm": 0.9734282773632583, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 23561 + }, + { + "epoch": 0.23562, + "grad_norm": 1.151729347531325, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 23562 + }, + { + "epoch": 0.23563, + "grad_norm": 1.1868573804103117, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 23563 + }, + { + "epoch": 0.23564, + "grad_norm": 0.9834442583572628, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 23564 + }, + { + "epoch": 0.23565, + "grad_norm": 1.2231321506840225, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 23565 + }, + { + "epoch": 0.23566, + "grad_norm": 1.038193855300545, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 23566 + }, + { + "epoch": 0.23567, + "grad_norm": 1.0376300205990103, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 23567 + }, + { + "epoch": 0.23568, + "grad_norm": 1.0642254280411814, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 23568 + }, + { + "epoch": 0.23569, + "grad_norm": 1.0020688518571121, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 23569 + }, + { + "epoch": 0.2357, + "grad_norm": 1.1081802158864986, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 23570 + }, + { + "epoch": 0.23571, + "grad_norm": 1.1280132019522795, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 23571 + }, + { + "epoch": 0.23572, + "grad_norm": 1.1503011556899279, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 23572 + }, + { + "epoch": 0.23573, + "grad_norm": 1.0229577152861995, + "learning_rate": 0.003, + "loss": 4.054, + "step": 23573 + }, + { + "epoch": 0.23574, + "grad_norm": 1.2856315949673378, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 23574 + }, + { + "epoch": 0.23575, + "grad_norm": 0.8758465544881264, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 23575 + }, + { + "epoch": 0.23576, + "grad_norm": 0.9295089762885905, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 23576 + }, + { + "epoch": 0.23577, + "grad_norm": 1.091954940040125, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 23577 + }, + { + "epoch": 0.23578, + "grad_norm": 0.9980322695551633, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 23578 + }, + { + "epoch": 0.23579, + "grad_norm": 1.0739817142445998, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 23579 + }, + { + "epoch": 0.2358, + "grad_norm": 1.1936572203694018, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 23580 + }, + { + "epoch": 0.23581, + "grad_norm": 1.1166090754382842, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 23581 + }, + { + "epoch": 0.23582, + "grad_norm": 1.2670988065781532, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 23582 + }, + { + "epoch": 0.23583, + "grad_norm": 0.8590821838457953, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 23583 + }, + { + "epoch": 0.23584, + "grad_norm": 0.7728114114842981, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 23584 + }, + { + "epoch": 0.23585, + "grad_norm": 0.8270221424711749, + "learning_rate": 0.003, + "loss": 4.033, + "step": 23585 + }, + { + "epoch": 0.23586, + "grad_norm": 0.8649648665268009, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 23586 + }, + { + "epoch": 0.23587, + "grad_norm": 0.8067669439468499, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 23587 + }, + { + "epoch": 0.23588, + "grad_norm": 0.7146830563975983, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 23588 + }, + { + "epoch": 0.23589, + "grad_norm": 0.7884912874193454, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 23589 + }, + { + "epoch": 0.2359, + "grad_norm": 0.936344108651867, + "learning_rate": 0.003, + "loss": 4.045, + "step": 23590 + }, + { + "epoch": 0.23591, + "grad_norm": 1.5582550604892536, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 23591 + }, + { + "epoch": 0.23592, + "grad_norm": 0.8166613468587481, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 23592 + }, + { + "epoch": 0.23593, + "grad_norm": 0.7881663279072546, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 23593 + }, + { + "epoch": 0.23594, + "grad_norm": 0.9765705999295781, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 23594 + }, + { + "epoch": 0.23595, + "grad_norm": 1.4191470272071691, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 23595 + }, + { + "epoch": 0.23596, + "grad_norm": 0.8075246426080014, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 23596 + }, + { + "epoch": 0.23597, + "grad_norm": 0.9015227209602722, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 23597 + }, + { + "epoch": 0.23598, + "grad_norm": 1.1375545589949696, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 23598 + }, + { + "epoch": 0.23599, + "grad_norm": 1.2816027500370863, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 23599 + }, + { + "epoch": 0.236, + "grad_norm": 0.959901885993596, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 23600 + }, + { + "epoch": 0.23601, + "grad_norm": 1.1623973153876366, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 23601 + }, + { + "epoch": 0.23602, + "grad_norm": 1.169802564217323, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 23602 + }, + { + "epoch": 0.23603, + "grad_norm": 0.9932708349230118, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 23603 + }, + { + "epoch": 0.23604, + "grad_norm": 1.0298252545340216, + "learning_rate": 0.003, + "loss": 4.051, + "step": 23604 + }, + { + "epoch": 0.23605, + "grad_norm": 1.1616884708676094, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 23605 + }, + { + "epoch": 0.23606, + "grad_norm": 1.160083185111852, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 23606 + }, + { + "epoch": 0.23607, + "grad_norm": 1.4193814204632347, + "learning_rate": 0.003, + "loss": 4.072, + "step": 23607 + }, + { + "epoch": 0.23608, + "grad_norm": 0.7206553212913136, + "learning_rate": 0.003, + "loss": 4.024, + "step": 23608 + }, + { + "epoch": 0.23609, + "grad_norm": 0.8929016825640773, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 23609 + }, + { + "epoch": 0.2361, + "grad_norm": 1.3182246469440908, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 23610 + }, + { + "epoch": 0.23611, + "grad_norm": 1.0863763114865783, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 23611 + }, + { + "epoch": 0.23612, + "grad_norm": 0.8948591670290258, + "learning_rate": 0.003, + "loss": 4.048, + "step": 23612 + }, + { + "epoch": 0.23613, + "grad_norm": 1.1094977527771286, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 23613 + }, + { + "epoch": 0.23614, + "grad_norm": 1.1174211503682565, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 23614 + }, + { + "epoch": 0.23615, + "grad_norm": 1.0789192697753553, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 23615 + }, + { + "epoch": 0.23616, + "grad_norm": 0.9644907238739461, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 23616 + }, + { + "epoch": 0.23617, + "grad_norm": 0.9723917960526969, + "learning_rate": 0.003, + "loss": 4.065, + "step": 23617 + }, + { + "epoch": 0.23618, + "grad_norm": 1.0996412645235902, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 23618 + }, + { + "epoch": 0.23619, + "grad_norm": 1.2063418250967195, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 23619 + }, + { + "epoch": 0.2362, + "grad_norm": 1.0254813877468685, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 23620 + }, + { + "epoch": 0.23621, + "grad_norm": 1.1103277947797734, + "learning_rate": 0.003, + "loss": 4.051, + "step": 23621 + }, + { + "epoch": 0.23622, + "grad_norm": 1.11405328149298, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 23622 + }, + { + "epoch": 0.23623, + "grad_norm": 0.9733671440944722, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 23623 + }, + { + "epoch": 0.23624, + "grad_norm": 1.0512473070495054, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 23624 + }, + { + "epoch": 0.23625, + "grad_norm": 1.0275136678595305, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 23625 + }, + { + "epoch": 0.23626, + "grad_norm": 1.3728507335215976, + "learning_rate": 0.003, + "loss": 4.092, + "step": 23626 + }, + { + "epoch": 0.23627, + "grad_norm": 0.921117496580854, + "learning_rate": 0.003, + "loss": 4.046, + "step": 23627 + }, + { + "epoch": 0.23628, + "grad_norm": 1.0891333030866115, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 23628 + }, + { + "epoch": 0.23629, + "grad_norm": 1.0836400483597988, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 23629 + }, + { + "epoch": 0.2363, + "grad_norm": 1.3589097020535168, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 23630 + }, + { + "epoch": 0.23631, + "grad_norm": 1.0171299781068013, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 23631 + }, + { + "epoch": 0.23632, + "grad_norm": 1.367222496871783, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 23632 + }, + { + "epoch": 0.23633, + "grad_norm": 1.225575368564778, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 23633 + }, + { + "epoch": 0.23634, + "grad_norm": 0.9685658294564806, + "learning_rate": 0.003, + "loss": 4.0905, + "step": 23634 + }, + { + "epoch": 0.23635, + "grad_norm": 1.0897922177140058, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 23635 + }, + { + "epoch": 0.23636, + "grad_norm": 1.2495356492793985, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 23636 + }, + { + "epoch": 0.23637, + "grad_norm": 1.1431117874914118, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 23637 + }, + { + "epoch": 0.23638, + "grad_norm": 0.8248356053185932, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 23638 + }, + { + "epoch": 0.23639, + "grad_norm": 0.8591472654521626, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 23639 + }, + { + "epoch": 0.2364, + "grad_norm": 1.0238998912251185, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 23640 + }, + { + "epoch": 0.23641, + "grad_norm": 0.9809559643290829, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 23641 + }, + { + "epoch": 0.23642, + "grad_norm": 1.111102959112007, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 23642 + }, + { + "epoch": 0.23643, + "grad_norm": 1.1646776300912602, + "learning_rate": 0.003, + "loss": 4.0923, + "step": 23643 + }, + { + "epoch": 0.23644, + "grad_norm": 1.0665046985288036, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 23644 + }, + { + "epoch": 0.23645, + "grad_norm": 1.2760260866783752, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 23645 + }, + { + "epoch": 0.23646, + "grad_norm": 1.0229511274333023, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 23646 + }, + { + "epoch": 0.23647, + "grad_norm": 1.2193664595704328, + "learning_rate": 0.003, + "loss": 4.051, + "step": 23647 + }, + { + "epoch": 0.23648, + "grad_norm": 0.8786564349504338, + "learning_rate": 0.003, + "loss": 4.054, + "step": 23648 + }, + { + "epoch": 0.23649, + "grad_norm": 0.9802042357905627, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 23649 + }, + { + "epoch": 0.2365, + "grad_norm": 1.2348166032590733, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 23650 + }, + { + "epoch": 0.23651, + "grad_norm": 1.079098037056643, + "learning_rate": 0.003, + "loss": 4.064, + "step": 23651 + }, + { + "epoch": 0.23652, + "grad_norm": 1.3729278645221161, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 23652 + }, + { + "epoch": 0.23653, + "grad_norm": 1.1210114137552265, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 23653 + }, + { + "epoch": 0.23654, + "grad_norm": 1.1383310093022951, + "learning_rate": 0.003, + "loss": 4.029, + "step": 23654 + }, + { + "epoch": 0.23655, + "grad_norm": 1.0661985760555224, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 23655 + }, + { + "epoch": 0.23656, + "grad_norm": 1.036443051961566, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 23656 + }, + { + "epoch": 0.23657, + "grad_norm": 1.1528230451020876, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 23657 + }, + { + "epoch": 0.23658, + "grad_norm": 1.0526461595942163, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 23658 + }, + { + "epoch": 0.23659, + "grad_norm": 1.1166673807000544, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 23659 + }, + { + "epoch": 0.2366, + "grad_norm": 1.2169001121904273, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 23660 + }, + { + "epoch": 0.23661, + "grad_norm": 1.0292575320497783, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 23661 + }, + { + "epoch": 0.23662, + "grad_norm": 1.1131179886299214, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 23662 + }, + { + "epoch": 0.23663, + "grad_norm": 1.055101539432986, + "learning_rate": 0.003, + "loss": 4.052, + "step": 23663 + }, + { + "epoch": 0.23664, + "grad_norm": 1.0559363684229275, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 23664 + }, + { + "epoch": 0.23665, + "grad_norm": 1.0920347552032552, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 23665 + }, + { + "epoch": 0.23666, + "grad_norm": 1.1371920798357849, + "learning_rate": 0.003, + "loss": 4.1105, + "step": 23666 + }, + { + "epoch": 0.23667, + "grad_norm": 1.0469875966840019, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 23667 + }, + { + "epoch": 0.23668, + "grad_norm": 1.177636298911308, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 23668 + }, + { + "epoch": 0.23669, + "grad_norm": 0.966649860710648, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 23669 + }, + { + "epoch": 0.2367, + "grad_norm": 1.1180213619579933, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 23670 + }, + { + "epoch": 0.23671, + "grad_norm": 1.017403945652298, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 23671 + }, + { + "epoch": 0.23672, + "grad_norm": 1.1710128400804352, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 23672 + }, + { + "epoch": 0.23673, + "grad_norm": 1.068501374595374, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 23673 + }, + { + "epoch": 0.23674, + "grad_norm": 0.9611940924140384, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 23674 + }, + { + "epoch": 0.23675, + "grad_norm": 1.0520761898714208, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 23675 + }, + { + "epoch": 0.23676, + "grad_norm": 1.1356125320016206, + "learning_rate": 0.003, + "loss": 4.0949, + "step": 23676 + }, + { + "epoch": 0.23677, + "grad_norm": 0.9692083553135872, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 23677 + }, + { + "epoch": 0.23678, + "grad_norm": 1.0870789656746258, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 23678 + }, + { + "epoch": 0.23679, + "grad_norm": 1.063053076910004, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 23679 + }, + { + "epoch": 0.2368, + "grad_norm": 1.1942942047916494, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 23680 + }, + { + "epoch": 0.23681, + "grad_norm": 1.2447884813507135, + "learning_rate": 0.003, + "loss": 4.037, + "step": 23681 + }, + { + "epoch": 0.23682, + "grad_norm": 0.8853577540883502, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 23682 + }, + { + "epoch": 0.23683, + "grad_norm": 0.8237422960892312, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 23683 + }, + { + "epoch": 0.23684, + "grad_norm": 0.9469489593343755, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 23684 + }, + { + "epoch": 0.23685, + "grad_norm": 1.073668390976808, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 23685 + }, + { + "epoch": 0.23686, + "grad_norm": 0.9582037203141565, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 23686 + }, + { + "epoch": 0.23687, + "grad_norm": 1.1527368896680183, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 23687 + }, + { + "epoch": 0.23688, + "grad_norm": 1.038243333536318, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 23688 + }, + { + "epoch": 0.23689, + "grad_norm": 1.0916161536336308, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 23689 + }, + { + "epoch": 0.2369, + "grad_norm": 1.104747725270206, + "learning_rate": 0.003, + "loss": 4.077, + "step": 23690 + }, + { + "epoch": 0.23691, + "grad_norm": 1.2941283655908429, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 23691 + }, + { + "epoch": 0.23692, + "grad_norm": 0.9392258644323469, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 23692 + }, + { + "epoch": 0.23693, + "grad_norm": 1.0714250347700582, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 23693 + }, + { + "epoch": 0.23694, + "grad_norm": 1.178210227520043, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 23694 + }, + { + "epoch": 0.23695, + "grad_norm": 1.0639946038670638, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 23695 + }, + { + "epoch": 0.23696, + "grad_norm": 1.2666849893313061, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 23696 + }, + { + "epoch": 0.23697, + "grad_norm": 0.9378951232205407, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 23697 + }, + { + "epoch": 0.23698, + "grad_norm": 1.0261822657180493, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 23698 + }, + { + "epoch": 0.23699, + "grad_norm": 1.1239070144068233, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 23699 + }, + { + "epoch": 0.237, + "grad_norm": 1.1139034371085395, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 23700 + }, + { + "epoch": 0.23701, + "grad_norm": 1.1165238338931263, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 23701 + }, + { + "epoch": 0.23702, + "grad_norm": 1.0423469933319764, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 23702 + }, + { + "epoch": 0.23703, + "grad_norm": 1.1094047196161323, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 23703 + }, + { + "epoch": 0.23704, + "grad_norm": 0.9259113606592585, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 23704 + }, + { + "epoch": 0.23705, + "grad_norm": 0.9179486026252701, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 23705 + }, + { + "epoch": 0.23706, + "grad_norm": 1.0653429311625073, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 23706 + }, + { + "epoch": 0.23707, + "grad_norm": 1.036076782336441, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 23707 + }, + { + "epoch": 0.23708, + "grad_norm": 1.2137238845626308, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 23708 + }, + { + "epoch": 0.23709, + "grad_norm": 1.0815204115133397, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 23709 + }, + { + "epoch": 0.2371, + "grad_norm": 1.0788050711530803, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 23710 + }, + { + "epoch": 0.23711, + "grad_norm": 0.9230220345670156, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 23711 + }, + { + "epoch": 0.23712, + "grad_norm": 1.1856144764014744, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 23712 + }, + { + "epoch": 0.23713, + "grad_norm": 1.1215840249716014, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 23713 + }, + { + "epoch": 0.23714, + "grad_norm": 0.8493909579995295, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 23714 + }, + { + "epoch": 0.23715, + "grad_norm": 0.8604670357356798, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 23715 + }, + { + "epoch": 0.23716, + "grad_norm": 0.7492179425633426, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 23716 + }, + { + "epoch": 0.23717, + "grad_norm": 0.8238963472720289, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 23717 + }, + { + "epoch": 0.23718, + "grad_norm": 0.9798572327034545, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 23718 + }, + { + "epoch": 0.23719, + "grad_norm": 1.2249864116638436, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 23719 + }, + { + "epoch": 0.2372, + "grad_norm": 1.0580301436784667, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 23720 + }, + { + "epoch": 0.23721, + "grad_norm": 1.1945363894614676, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 23721 + }, + { + "epoch": 0.23722, + "grad_norm": 1.1282403289708818, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 23722 + }, + { + "epoch": 0.23723, + "grad_norm": 1.1143080883272796, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 23723 + }, + { + "epoch": 0.23724, + "grad_norm": 1.245968227708912, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 23724 + }, + { + "epoch": 0.23725, + "grad_norm": 0.9281550711123931, + "learning_rate": 0.003, + "loss": 4.055, + "step": 23725 + }, + { + "epoch": 0.23726, + "grad_norm": 1.0634568283516386, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 23726 + }, + { + "epoch": 0.23727, + "grad_norm": 1.124376987937097, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 23727 + }, + { + "epoch": 0.23728, + "grad_norm": 1.2520793307691531, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 23728 + }, + { + "epoch": 0.23729, + "grad_norm": 0.9850174983746043, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 23729 + }, + { + "epoch": 0.2373, + "grad_norm": 1.2022853563424263, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 23730 + }, + { + "epoch": 0.23731, + "grad_norm": 0.8648805130322849, + "learning_rate": 0.003, + "loss": 4.047, + "step": 23731 + }, + { + "epoch": 0.23732, + "grad_norm": 0.8923228245742846, + "learning_rate": 0.003, + "loss": 4.054, + "step": 23732 + }, + { + "epoch": 0.23733, + "grad_norm": 1.1310872614457812, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 23733 + }, + { + "epoch": 0.23734, + "grad_norm": 0.9418295508432908, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 23734 + }, + { + "epoch": 0.23735, + "grad_norm": 0.9344608137338329, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 23735 + }, + { + "epoch": 0.23736, + "grad_norm": 1.1360638983286573, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 23736 + }, + { + "epoch": 0.23737, + "grad_norm": 1.120158600679455, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 23737 + }, + { + "epoch": 0.23738, + "grad_norm": 1.0056875254833064, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 23738 + }, + { + "epoch": 0.23739, + "grad_norm": 1.035319345313085, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 23739 + }, + { + "epoch": 0.2374, + "grad_norm": 1.0483499652294834, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 23740 + }, + { + "epoch": 0.23741, + "grad_norm": 1.5602290776644656, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 23741 + }, + { + "epoch": 0.23742, + "grad_norm": 0.7204464033424287, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 23742 + }, + { + "epoch": 0.23743, + "grad_norm": 0.7905891343739269, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 23743 + }, + { + "epoch": 0.23744, + "grad_norm": 0.8374221865167389, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 23744 + }, + { + "epoch": 0.23745, + "grad_norm": 1.0774882545057223, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 23745 + }, + { + "epoch": 0.23746, + "grad_norm": 1.227228621217819, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 23746 + }, + { + "epoch": 0.23747, + "grad_norm": 1.0855360949829131, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 23747 + }, + { + "epoch": 0.23748, + "grad_norm": 1.3035428544010856, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 23748 + }, + { + "epoch": 0.23749, + "grad_norm": 0.8697134581259565, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 23749 + }, + { + "epoch": 0.2375, + "grad_norm": 0.917653416715854, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 23750 + }, + { + "epoch": 0.23751, + "grad_norm": 1.1119193945706618, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 23751 + }, + { + "epoch": 0.23752, + "grad_norm": 1.3775739256604083, + "learning_rate": 0.003, + "loss": 4.053, + "step": 23752 + }, + { + "epoch": 0.23753, + "grad_norm": 1.1662698083883343, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 23753 + }, + { + "epoch": 0.23754, + "grad_norm": 1.3099198205848868, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 23754 + }, + { + "epoch": 0.23755, + "grad_norm": 0.882872443844075, + "learning_rate": 0.003, + "loss": 4.045, + "step": 23755 + }, + { + "epoch": 0.23756, + "grad_norm": 0.7989929876481974, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 23756 + }, + { + "epoch": 0.23757, + "grad_norm": 1.0169718814265936, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 23757 + }, + { + "epoch": 0.23758, + "grad_norm": 1.1995961827454678, + "learning_rate": 0.003, + "loss": 4.073, + "step": 23758 + }, + { + "epoch": 0.23759, + "grad_norm": 1.0300998891356041, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 23759 + }, + { + "epoch": 0.2376, + "grad_norm": 1.0204842497304152, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 23760 + }, + { + "epoch": 0.23761, + "grad_norm": 1.148952865521914, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 23761 + }, + { + "epoch": 0.23762, + "grad_norm": 0.9934661590627203, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 23762 + }, + { + "epoch": 0.23763, + "grad_norm": 0.9925910370809509, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 23763 + }, + { + "epoch": 0.23764, + "grad_norm": 0.9434707047994529, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 23764 + }, + { + "epoch": 0.23765, + "grad_norm": 1.0662178479675934, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 23765 + }, + { + "epoch": 0.23766, + "grad_norm": 1.0040106598271696, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 23766 + }, + { + "epoch": 0.23767, + "grad_norm": 1.0578445978838449, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 23767 + }, + { + "epoch": 0.23768, + "grad_norm": 1.4705319781118704, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 23768 + }, + { + "epoch": 0.23769, + "grad_norm": 1.1061762635257946, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 23769 + }, + { + "epoch": 0.2377, + "grad_norm": 0.9980925069538616, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 23770 + }, + { + "epoch": 0.23771, + "grad_norm": 1.1334355129078424, + "learning_rate": 0.003, + "loss": 4.039, + "step": 23771 + }, + { + "epoch": 0.23772, + "grad_norm": 1.2124954322857469, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 23772 + }, + { + "epoch": 0.23773, + "grad_norm": 1.0990255549273866, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 23773 + }, + { + "epoch": 0.23774, + "grad_norm": 1.1691494394248776, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 23774 + }, + { + "epoch": 0.23775, + "grad_norm": 1.0878880133065036, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 23775 + }, + { + "epoch": 0.23776, + "grad_norm": 1.0039317219797612, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 23776 + }, + { + "epoch": 0.23777, + "grad_norm": 1.0698438144382, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 23777 + }, + { + "epoch": 0.23778, + "grad_norm": 1.1740933701020895, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 23778 + }, + { + "epoch": 0.23779, + "grad_norm": 1.0469185903163827, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 23779 + }, + { + "epoch": 0.2378, + "grad_norm": 1.3310251434755593, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 23780 + }, + { + "epoch": 0.23781, + "grad_norm": 0.8174473518488273, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 23781 + }, + { + "epoch": 0.23782, + "grad_norm": 0.7009976372341422, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 23782 + }, + { + "epoch": 0.23783, + "grad_norm": 0.7172973666510695, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 23783 + }, + { + "epoch": 0.23784, + "grad_norm": 0.7373644672107555, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 23784 + }, + { + "epoch": 0.23785, + "grad_norm": 0.7824029521775288, + "learning_rate": 0.003, + "loss": 4.023, + "step": 23785 + }, + { + "epoch": 0.23786, + "grad_norm": 0.9472151389070049, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 23786 + }, + { + "epoch": 0.23787, + "grad_norm": 1.4405235580744562, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 23787 + }, + { + "epoch": 0.23788, + "grad_norm": 1.008822502308013, + "learning_rate": 0.003, + "loss": 4.058, + "step": 23788 + }, + { + "epoch": 0.23789, + "grad_norm": 1.0691408247426528, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 23789 + }, + { + "epoch": 0.2379, + "grad_norm": 1.0550135223833992, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 23790 + }, + { + "epoch": 0.23791, + "grad_norm": 1.3836328665994548, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 23791 + }, + { + "epoch": 0.23792, + "grad_norm": 1.013464023063325, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 23792 + }, + { + "epoch": 0.23793, + "grad_norm": 1.1166482674703455, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 23793 + }, + { + "epoch": 0.23794, + "grad_norm": 1.348518446701531, + "learning_rate": 0.003, + "loss": 4.0855, + "step": 23794 + }, + { + "epoch": 0.23795, + "grad_norm": 1.1432554958674896, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 23795 + }, + { + "epoch": 0.23796, + "grad_norm": 0.9751798031996699, + "learning_rate": 0.003, + "loss": 4.086, + "step": 23796 + }, + { + "epoch": 0.23797, + "grad_norm": 1.3179821527346258, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 23797 + }, + { + "epoch": 0.23798, + "grad_norm": 1.0753880038735073, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 23798 + }, + { + "epoch": 0.23799, + "grad_norm": 1.0746997075338092, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 23799 + }, + { + "epoch": 0.238, + "grad_norm": 1.292638103761781, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 23800 + }, + { + "epoch": 0.23801, + "grad_norm": 0.8805578459761717, + "learning_rate": 0.003, + "loss": 4.0965, + "step": 23801 + }, + { + "epoch": 0.23802, + "grad_norm": 0.9674260246096528, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 23802 + }, + { + "epoch": 0.23803, + "grad_norm": 1.093431782353197, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 23803 + }, + { + "epoch": 0.23804, + "grad_norm": 0.9675116294389068, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 23804 + }, + { + "epoch": 0.23805, + "grad_norm": 1.0912956123916449, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 23805 + }, + { + "epoch": 0.23806, + "grad_norm": 1.1753508173003067, + "learning_rate": 0.003, + "loss": 4.053, + "step": 23806 + }, + { + "epoch": 0.23807, + "grad_norm": 1.0397221561092398, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 23807 + }, + { + "epoch": 0.23808, + "grad_norm": 1.0075554735918983, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 23808 + }, + { + "epoch": 0.23809, + "grad_norm": 1.1538430487848166, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 23809 + }, + { + "epoch": 0.2381, + "grad_norm": 1.0786692628375345, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 23810 + }, + { + "epoch": 0.23811, + "grad_norm": 1.0687245541056585, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 23811 + }, + { + "epoch": 0.23812, + "grad_norm": 1.0266875187144224, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 23812 + }, + { + "epoch": 0.23813, + "grad_norm": 1.1348177755877902, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 23813 + }, + { + "epoch": 0.23814, + "grad_norm": 0.9251776648361479, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 23814 + }, + { + "epoch": 0.23815, + "grad_norm": 1.1196475337825191, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 23815 + }, + { + "epoch": 0.23816, + "grad_norm": 1.2097896005150115, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 23816 + }, + { + "epoch": 0.23817, + "grad_norm": 0.9700615210089315, + "learning_rate": 0.003, + "loss": 4.0923, + "step": 23817 + }, + { + "epoch": 0.23818, + "grad_norm": 1.005368970763991, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 23818 + }, + { + "epoch": 0.23819, + "grad_norm": 1.2123446749237288, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 23819 + }, + { + "epoch": 0.2382, + "grad_norm": 1.0236524380477225, + "learning_rate": 0.003, + "loss": 4.051, + "step": 23820 + }, + { + "epoch": 0.23821, + "grad_norm": 1.0543212688576906, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 23821 + }, + { + "epoch": 0.23822, + "grad_norm": 1.2787878177094183, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 23822 + }, + { + "epoch": 0.23823, + "grad_norm": 0.9124305788332148, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 23823 + }, + { + "epoch": 0.23824, + "grad_norm": 1.030346822609352, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 23824 + }, + { + "epoch": 0.23825, + "grad_norm": 1.2122540570027307, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 23825 + }, + { + "epoch": 0.23826, + "grad_norm": 1.0603403342216815, + "learning_rate": 0.003, + "loss": 4.028, + "step": 23826 + }, + { + "epoch": 0.23827, + "grad_norm": 1.209867262503206, + "learning_rate": 0.003, + "loss": 4.038, + "step": 23827 + }, + { + "epoch": 0.23828, + "grad_norm": 1.0084937819365778, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 23828 + }, + { + "epoch": 0.23829, + "grad_norm": 1.1914443889834128, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 23829 + }, + { + "epoch": 0.2383, + "grad_norm": 1.0864368079911577, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 23830 + }, + { + "epoch": 0.23831, + "grad_norm": 1.0497425588099127, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 23831 + }, + { + "epoch": 0.23832, + "grad_norm": 0.9847912712690293, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 23832 + }, + { + "epoch": 0.23833, + "grad_norm": 1.3193212222741582, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 23833 + }, + { + "epoch": 0.23834, + "grad_norm": 1.0650705558605496, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 23834 + }, + { + "epoch": 0.23835, + "grad_norm": 1.1286202692629714, + "learning_rate": 0.003, + "loss": 4.023, + "step": 23835 + }, + { + "epoch": 0.23836, + "grad_norm": 0.9699471793244705, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 23836 + }, + { + "epoch": 0.23837, + "grad_norm": 1.163004997461789, + "learning_rate": 0.003, + "loss": 4.0942, + "step": 23837 + }, + { + "epoch": 0.23838, + "grad_norm": 1.014867956985326, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 23838 + }, + { + "epoch": 0.23839, + "grad_norm": 0.963949839009725, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 23839 + }, + { + "epoch": 0.2384, + "grad_norm": 1.169655366799792, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 23840 + }, + { + "epoch": 0.23841, + "grad_norm": 0.9938811298791316, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 23841 + }, + { + "epoch": 0.23842, + "grad_norm": 1.0056575223155713, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 23842 + }, + { + "epoch": 0.23843, + "grad_norm": 1.0078363936153272, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 23843 + }, + { + "epoch": 0.23844, + "grad_norm": 1.008838199852202, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 23844 + }, + { + "epoch": 0.23845, + "grad_norm": 0.9578606990775936, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 23845 + }, + { + "epoch": 0.23846, + "grad_norm": 1.0902177689213621, + "learning_rate": 0.003, + "loss": 4.028, + "step": 23846 + }, + { + "epoch": 0.23847, + "grad_norm": 1.1453656927995977, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 23847 + }, + { + "epoch": 0.23848, + "grad_norm": 0.9829265912319546, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 23848 + }, + { + "epoch": 0.23849, + "grad_norm": 0.9635718061604223, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 23849 + }, + { + "epoch": 0.2385, + "grad_norm": 1.0522262052169493, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 23850 + }, + { + "epoch": 0.23851, + "grad_norm": 1.1803356994226453, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 23851 + }, + { + "epoch": 0.23852, + "grad_norm": 0.9012917979887443, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 23852 + }, + { + "epoch": 0.23853, + "grad_norm": 0.9588956998175372, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 23853 + }, + { + "epoch": 0.23854, + "grad_norm": 1.0846997198688972, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 23854 + }, + { + "epoch": 0.23855, + "grad_norm": 0.9177515318954331, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 23855 + }, + { + "epoch": 0.23856, + "grad_norm": 0.9728662974293291, + "learning_rate": 0.003, + "loss": 4.033, + "step": 23856 + }, + { + "epoch": 0.23857, + "grad_norm": 1.3682722187176248, + "learning_rate": 0.003, + "loss": 4.0967, + "step": 23857 + }, + { + "epoch": 0.23858, + "grad_norm": 1.1572304783453824, + "learning_rate": 0.003, + "loss": 4.1065, + "step": 23858 + }, + { + "epoch": 0.23859, + "grad_norm": 1.190031981499841, + "learning_rate": 0.003, + "loss": 4.05, + "step": 23859 + }, + { + "epoch": 0.2386, + "grad_norm": 1.1804720517730813, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 23860 + }, + { + "epoch": 0.23861, + "grad_norm": 1.0842915223698015, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 23861 + }, + { + "epoch": 0.23862, + "grad_norm": 1.2522398944348216, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 23862 + }, + { + "epoch": 0.23863, + "grad_norm": 1.0030258724852101, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 23863 + }, + { + "epoch": 0.23864, + "grad_norm": 1.1741997895633685, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 23864 + }, + { + "epoch": 0.23865, + "grad_norm": 0.9629932330722331, + "learning_rate": 0.003, + "loss": 4.048, + "step": 23865 + }, + { + "epoch": 0.23866, + "grad_norm": 0.956398627072388, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 23866 + }, + { + "epoch": 0.23867, + "grad_norm": 1.0256419991483237, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 23867 + }, + { + "epoch": 0.23868, + "grad_norm": 1.297024721004735, + "learning_rate": 0.003, + "loss": 4.054, + "step": 23868 + }, + { + "epoch": 0.23869, + "grad_norm": 0.9925445169656774, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 23869 + }, + { + "epoch": 0.2387, + "grad_norm": 1.5205825167855842, + "learning_rate": 0.003, + "loss": 4.099, + "step": 23870 + }, + { + "epoch": 0.23871, + "grad_norm": 0.7489085371703331, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 23871 + }, + { + "epoch": 0.23872, + "grad_norm": 0.8418156000946146, + "learning_rate": 0.003, + "loss": 4.058, + "step": 23872 + }, + { + "epoch": 0.23873, + "grad_norm": 0.8341774932590992, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 23873 + }, + { + "epoch": 0.23874, + "grad_norm": 1.0444422320622988, + "learning_rate": 0.003, + "loss": 4.048, + "step": 23874 + }, + { + "epoch": 0.23875, + "grad_norm": 1.04502161786984, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 23875 + }, + { + "epoch": 0.23876, + "grad_norm": 1.0588848402427813, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 23876 + }, + { + "epoch": 0.23877, + "grad_norm": 0.9158269507612589, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 23877 + }, + { + "epoch": 0.23878, + "grad_norm": 1.1416508153502618, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 23878 + }, + { + "epoch": 0.23879, + "grad_norm": 1.0375538873085566, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 23879 + }, + { + "epoch": 0.2388, + "grad_norm": 1.3487874469135388, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 23880 + }, + { + "epoch": 0.23881, + "grad_norm": 0.9576030087860853, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 23881 + }, + { + "epoch": 0.23882, + "grad_norm": 1.1022304558783569, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 23882 + }, + { + "epoch": 0.23883, + "grad_norm": 1.225227543707656, + "learning_rate": 0.003, + "loss": 4.084, + "step": 23883 + }, + { + "epoch": 0.23884, + "grad_norm": 0.928586030945311, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 23884 + }, + { + "epoch": 0.23885, + "grad_norm": 1.0645954793766284, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 23885 + }, + { + "epoch": 0.23886, + "grad_norm": 1.0037796510042707, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 23886 + }, + { + "epoch": 0.23887, + "grad_norm": 1.0035009503541328, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 23887 + }, + { + "epoch": 0.23888, + "grad_norm": 1.2187677018364267, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 23888 + }, + { + "epoch": 0.23889, + "grad_norm": 0.9821911941555987, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 23889 + }, + { + "epoch": 0.2389, + "grad_norm": 1.1817124411127036, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 23890 + }, + { + "epoch": 0.23891, + "grad_norm": 0.8219277470074154, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 23891 + }, + { + "epoch": 0.23892, + "grad_norm": 0.999005700411657, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 23892 + }, + { + "epoch": 0.23893, + "grad_norm": 1.097406463672634, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 23893 + }, + { + "epoch": 0.23894, + "grad_norm": 0.9383456785186689, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 23894 + }, + { + "epoch": 0.23895, + "grad_norm": 0.9922254023813701, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 23895 + }, + { + "epoch": 0.23896, + "grad_norm": 1.4279967833042906, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 23896 + }, + { + "epoch": 0.23897, + "grad_norm": 1.0430446859195102, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 23897 + }, + { + "epoch": 0.23898, + "grad_norm": 1.1685623683923159, + "learning_rate": 0.003, + "loss": 4.072, + "step": 23898 + }, + { + "epoch": 0.23899, + "grad_norm": 1.0206589601096223, + "learning_rate": 0.003, + "loss": 4.029, + "step": 23899 + }, + { + "epoch": 0.239, + "grad_norm": 1.2379217417186377, + "learning_rate": 0.003, + "loss": 4.042, + "step": 23900 + }, + { + "epoch": 0.23901, + "grad_norm": 1.0165844259353574, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 23901 + }, + { + "epoch": 0.23902, + "grad_norm": 1.295110387260934, + "learning_rate": 0.003, + "loss": 4.058, + "step": 23902 + }, + { + "epoch": 0.23903, + "grad_norm": 0.847171098594224, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 23903 + }, + { + "epoch": 0.23904, + "grad_norm": 0.7338390106070883, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 23904 + }, + { + "epoch": 0.23905, + "grad_norm": 0.7843112930634589, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 23905 + }, + { + "epoch": 0.23906, + "grad_norm": 0.8107291700953856, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 23906 + }, + { + "epoch": 0.23907, + "grad_norm": 0.9756949349149188, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 23907 + }, + { + "epoch": 0.23908, + "grad_norm": 1.1171232096739234, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 23908 + }, + { + "epoch": 0.23909, + "grad_norm": 1.0355456424739131, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 23909 + }, + { + "epoch": 0.2391, + "grad_norm": 1.2916279975800702, + "learning_rate": 0.003, + "loss": 4.075, + "step": 23910 + }, + { + "epoch": 0.23911, + "grad_norm": 1.02718864201699, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 23911 + }, + { + "epoch": 0.23912, + "grad_norm": 1.0661139272360525, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 23912 + }, + { + "epoch": 0.23913, + "grad_norm": 1.0411211038559842, + "learning_rate": 0.003, + "loss": 4.047, + "step": 23913 + }, + { + "epoch": 0.23914, + "grad_norm": 1.1235752987712526, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 23914 + }, + { + "epoch": 0.23915, + "grad_norm": 0.9171862339488368, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 23915 + }, + { + "epoch": 0.23916, + "grad_norm": 1.1726807221961495, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 23916 + }, + { + "epoch": 0.23917, + "grad_norm": 1.3322219236880184, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 23917 + }, + { + "epoch": 0.23918, + "grad_norm": 0.9855559502948431, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 23918 + }, + { + "epoch": 0.23919, + "grad_norm": 1.4274438218110415, + "learning_rate": 0.003, + "loss": 4.1, + "step": 23919 + }, + { + "epoch": 0.2392, + "grad_norm": 0.9884559092798703, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 23920 + }, + { + "epoch": 0.23921, + "grad_norm": 1.1368051177969087, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 23921 + }, + { + "epoch": 0.23922, + "grad_norm": 0.9660438343088658, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 23922 + }, + { + "epoch": 0.23923, + "grad_norm": 1.0721746172762174, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 23923 + }, + { + "epoch": 0.23924, + "grad_norm": 1.177113851471357, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 23924 + }, + { + "epoch": 0.23925, + "grad_norm": 1.2072268907337627, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 23925 + }, + { + "epoch": 0.23926, + "grad_norm": 0.8340317935751297, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 23926 + }, + { + "epoch": 0.23927, + "grad_norm": 0.8235705382670261, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 23927 + }, + { + "epoch": 0.23928, + "grad_norm": 0.9448779674667251, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 23928 + }, + { + "epoch": 0.23929, + "grad_norm": 1.2422491112563736, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 23929 + }, + { + "epoch": 0.2393, + "grad_norm": 0.9267470880897487, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 23930 + }, + { + "epoch": 0.23931, + "grad_norm": 0.9339691169629697, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 23931 + }, + { + "epoch": 0.23932, + "grad_norm": 1.1409772848751243, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 23932 + }, + { + "epoch": 0.23933, + "grad_norm": 1.1265021305624607, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 23933 + }, + { + "epoch": 0.23934, + "grad_norm": 0.9290079470408341, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 23934 + }, + { + "epoch": 0.23935, + "grad_norm": 0.9930219555224072, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 23935 + }, + { + "epoch": 0.23936, + "grad_norm": 1.0289235034334716, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 23936 + }, + { + "epoch": 0.23937, + "grad_norm": 0.9509961942911722, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 23937 + }, + { + "epoch": 0.23938, + "grad_norm": 1.055523936167296, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 23938 + }, + { + "epoch": 0.23939, + "grad_norm": 1.1196866177545763, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 23939 + }, + { + "epoch": 0.2394, + "grad_norm": 1.2178141518481278, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 23940 + }, + { + "epoch": 0.23941, + "grad_norm": 1.1289001399707395, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 23941 + }, + { + "epoch": 0.23942, + "grad_norm": 1.1133696036411562, + "learning_rate": 0.003, + "loss": 4.056, + "step": 23942 + }, + { + "epoch": 0.23943, + "grad_norm": 1.0672336815889725, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 23943 + }, + { + "epoch": 0.23944, + "grad_norm": 1.1753144183699902, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 23944 + }, + { + "epoch": 0.23945, + "grad_norm": 0.8637120476357367, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 23945 + }, + { + "epoch": 0.23946, + "grad_norm": 0.8419396337175905, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 23946 + }, + { + "epoch": 0.23947, + "grad_norm": 0.9639569565726857, + "learning_rate": 0.003, + "loss": 4.058, + "step": 23947 + }, + { + "epoch": 0.23948, + "grad_norm": 1.164804967053413, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 23948 + }, + { + "epoch": 0.23949, + "grad_norm": 0.9231094014751717, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 23949 + }, + { + "epoch": 0.2395, + "grad_norm": 1.0516122232044138, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 23950 + }, + { + "epoch": 0.23951, + "grad_norm": 1.3316496297614904, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 23951 + }, + { + "epoch": 0.23952, + "grad_norm": 1.089365466989078, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 23952 + }, + { + "epoch": 0.23953, + "grad_norm": 1.1027529083963692, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 23953 + }, + { + "epoch": 0.23954, + "grad_norm": 1.1161694102669157, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 23954 + }, + { + "epoch": 0.23955, + "grad_norm": 1.0120900102407515, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 23955 + }, + { + "epoch": 0.23956, + "grad_norm": 1.2476219005429474, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 23956 + }, + { + "epoch": 0.23957, + "grad_norm": 0.9190375497199925, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 23957 + }, + { + "epoch": 0.23958, + "grad_norm": 0.9813181068712381, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 23958 + }, + { + "epoch": 0.23959, + "grad_norm": 1.2916429106474054, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 23959 + }, + { + "epoch": 0.2396, + "grad_norm": 0.8878862674448511, + "learning_rate": 0.003, + "loss": 4.047, + "step": 23960 + }, + { + "epoch": 0.23961, + "grad_norm": 1.1685544400991232, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 23961 + }, + { + "epoch": 0.23962, + "grad_norm": 1.2179723996671548, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 23962 + }, + { + "epoch": 0.23963, + "grad_norm": 0.8777269555071104, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 23963 + }, + { + "epoch": 0.23964, + "grad_norm": 0.9036269552942755, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 23964 + }, + { + "epoch": 0.23965, + "grad_norm": 1.1931149352217016, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 23965 + }, + { + "epoch": 0.23966, + "grad_norm": 1.1833253120771872, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 23966 + }, + { + "epoch": 0.23967, + "grad_norm": 0.9207754794533318, + "learning_rate": 0.003, + "loss": 4.062, + "step": 23967 + }, + { + "epoch": 0.23968, + "grad_norm": 1.0729953577736249, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 23968 + }, + { + "epoch": 0.23969, + "grad_norm": 1.1520901854346355, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 23969 + }, + { + "epoch": 0.2397, + "grad_norm": 0.9889493755553943, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 23970 + }, + { + "epoch": 0.23971, + "grad_norm": 1.088103653701929, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 23971 + }, + { + "epoch": 0.23972, + "grad_norm": 1.0407527148234603, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 23972 + }, + { + "epoch": 0.23973, + "grad_norm": 1.3168303257993927, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 23973 + }, + { + "epoch": 0.23974, + "grad_norm": 0.8966716592766779, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 23974 + }, + { + "epoch": 0.23975, + "grad_norm": 0.8725938034676675, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 23975 + }, + { + "epoch": 0.23976, + "grad_norm": 0.9159904161762514, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 23976 + }, + { + "epoch": 0.23977, + "grad_norm": 0.9568708249056218, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 23977 + }, + { + "epoch": 0.23978, + "grad_norm": 1.1629632513113666, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 23978 + }, + { + "epoch": 0.23979, + "grad_norm": 0.9928690993931124, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 23979 + }, + { + "epoch": 0.2398, + "grad_norm": 1.0047071413954176, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 23980 + }, + { + "epoch": 0.23981, + "grad_norm": 1.2161209348980706, + "learning_rate": 0.003, + "loss": 4.0916, + "step": 23981 + }, + { + "epoch": 0.23982, + "grad_norm": 0.7700864935088126, + "learning_rate": 0.003, + "loss": 4.046, + "step": 23982 + }, + { + "epoch": 0.23983, + "grad_norm": 0.8857330464977635, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 23983 + }, + { + "epoch": 0.23984, + "grad_norm": 1.0201836377652849, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 23984 + }, + { + "epoch": 0.23985, + "grad_norm": 1.2767542318289817, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 23985 + }, + { + "epoch": 0.23986, + "grad_norm": 1.1024013058306985, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 23986 + }, + { + "epoch": 0.23987, + "grad_norm": 1.3373578070825578, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 23987 + }, + { + "epoch": 0.23988, + "grad_norm": 1.0861394284941046, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 23988 + }, + { + "epoch": 0.23989, + "grad_norm": 1.1099751015137602, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 23989 + }, + { + "epoch": 0.2399, + "grad_norm": 1.0795310842026296, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 23990 + }, + { + "epoch": 0.23991, + "grad_norm": 1.0800380260268918, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 23991 + }, + { + "epoch": 0.23992, + "grad_norm": 1.018488368390995, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 23992 + }, + { + "epoch": 0.23993, + "grad_norm": 0.951096404910504, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 23993 + }, + { + "epoch": 0.23994, + "grad_norm": 1.0720231233908768, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 23994 + }, + { + "epoch": 0.23995, + "grad_norm": 1.1676569491212088, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 23995 + }, + { + "epoch": 0.23996, + "grad_norm": 1.0660518448900063, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 23996 + }, + { + "epoch": 0.23997, + "grad_norm": 1.266271928393999, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 23997 + }, + { + "epoch": 0.23998, + "grad_norm": 0.9815678708265406, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 23998 + }, + { + "epoch": 0.23999, + "grad_norm": 1.0665189431291664, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 23999 + }, + { + "epoch": 0.24, + "grad_norm": 1.2323467217466333, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 24000 + }, + { + "epoch": 0.24001, + "grad_norm": 0.8922656349884103, + "learning_rate": 0.003, + "loss": 4.035, + "step": 24001 + }, + { + "epoch": 0.24002, + "grad_norm": 1.013794577051766, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 24002 + }, + { + "epoch": 0.24003, + "grad_norm": 1.0635798991978533, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 24003 + }, + { + "epoch": 0.24004, + "grad_norm": 1.1232025650469686, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 24004 + }, + { + "epoch": 0.24005, + "grad_norm": 0.9613986857021359, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 24005 + }, + { + "epoch": 0.24006, + "grad_norm": 1.0230032488545135, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 24006 + }, + { + "epoch": 0.24007, + "grad_norm": 1.3189404454617228, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 24007 + }, + { + "epoch": 0.24008, + "grad_norm": 0.8285200982792265, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 24008 + }, + { + "epoch": 0.24009, + "grad_norm": 0.8787875570739027, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 24009 + }, + { + "epoch": 0.2401, + "grad_norm": 1.0586052299901565, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 24010 + }, + { + "epoch": 0.24011, + "grad_norm": 1.1672312421779802, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 24011 + }, + { + "epoch": 0.24012, + "grad_norm": 0.9972448683897215, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 24012 + }, + { + "epoch": 0.24013, + "grad_norm": 1.0492047808656748, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 24013 + }, + { + "epoch": 0.24014, + "grad_norm": 0.9220104910252845, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 24014 + }, + { + "epoch": 0.24015, + "grad_norm": 1.0235857057861502, + "learning_rate": 0.003, + "loss": 4.033, + "step": 24015 + }, + { + "epoch": 0.24016, + "grad_norm": 1.2848050260778248, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 24016 + }, + { + "epoch": 0.24017, + "grad_norm": 1.3595364785177046, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 24017 + }, + { + "epoch": 0.24018, + "grad_norm": 1.1420362863604538, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 24018 + }, + { + "epoch": 0.24019, + "grad_norm": 0.9778091138451199, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 24019 + }, + { + "epoch": 0.2402, + "grad_norm": 1.190778647111199, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 24020 + }, + { + "epoch": 0.24021, + "grad_norm": 0.9742810923112724, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 24021 + }, + { + "epoch": 0.24022, + "grad_norm": 1.2327247562992212, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 24022 + }, + { + "epoch": 0.24023, + "grad_norm": 0.9748065361892074, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 24023 + }, + { + "epoch": 0.24024, + "grad_norm": 1.2977654804559182, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 24024 + }, + { + "epoch": 0.24025, + "grad_norm": 0.972050060274605, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 24025 + }, + { + "epoch": 0.24026, + "grad_norm": 0.9182204515961072, + "learning_rate": 0.003, + "loss": 4.049, + "step": 24026 + }, + { + "epoch": 0.24027, + "grad_norm": 0.991243397627054, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 24027 + }, + { + "epoch": 0.24028, + "grad_norm": 1.0371449652234355, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 24028 + }, + { + "epoch": 0.24029, + "grad_norm": 0.8628224187582935, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 24029 + }, + { + "epoch": 0.2403, + "grad_norm": 0.8900191973316712, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 24030 + }, + { + "epoch": 0.24031, + "grad_norm": 0.9797666164978069, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 24031 + }, + { + "epoch": 0.24032, + "grad_norm": 1.099643428198633, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 24032 + }, + { + "epoch": 0.24033, + "grad_norm": 0.9946618301046031, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 24033 + }, + { + "epoch": 0.24034, + "grad_norm": 1.2176453806532388, + "learning_rate": 0.003, + "loss": 4.0856, + "step": 24034 + }, + { + "epoch": 0.24035, + "grad_norm": 1.147898423420466, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 24035 + }, + { + "epoch": 0.24036, + "grad_norm": 1.2943197855900728, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 24036 + }, + { + "epoch": 0.24037, + "grad_norm": 0.9965328234931012, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 24037 + }, + { + "epoch": 0.24038, + "grad_norm": 1.2279150025571122, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 24038 + }, + { + "epoch": 0.24039, + "grad_norm": 1.1599151708586077, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 24039 + }, + { + "epoch": 0.2404, + "grad_norm": 1.0679685261822085, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 24040 + }, + { + "epoch": 0.24041, + "grad_norm": 1.0412818064586875, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 24041 + }, + { + "epoch": 0.24042, + "grad_norm": 1.1277065871231575, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 24042 + }, + { + "epoch": 0.24043, + "grad_norm": 1.0870010609320075, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 24043 + }, + { + "epoch": 0.24044, + "grad_norm": 1.00820392312493, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 24044 + }, + { + "epoch": 0.24045, + "grad_norm": 1.0472760054881896, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 24045 + }, + { + "epoch": 0.24046, + "grad_norm": 0.9444706977976434, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 24046 + }, + { + "epoch": 0.24047, + "grad_norm": 0.9374099567259032, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 24047 + }, + { + "epoch": 0.24048, + "grad_norm": 1.014158752103189, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 24048 + }, + { + "epoch": 0.24049, + "grad_norm": 1.1691802441584573, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 24049 + }, + { + "epoch": 0.2405, + "grad_norm": 0.9006890609093515, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 24050 + }, + { + "epoch": 0.24051, + "grad_norm": 0.9315882701366784, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 24051 + }, + { + "epoch": 0.24052, + "grad_norm": 1.2398321983460856, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 24052 + }, + { + "epoch": 0.24053, + "grad_norm": 1.235548557715879, + "learning_rate": 0.003, + "loss": 4.056, + "step": 24053 + }, + { + "epoch": 0.24054, + "grad_norm": 0.9606671991178264, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 24054 + }, + { + "epoch": 0.24055, + "grad_norm": 1.0805635306112542, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 24055 + }, + { + "epoch": 0.24056, + "grad_norm": 1.2771992721500693, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 24056 + }, + { + "epoch": 0.24057, + "grad_norm": 0.8440457435875898, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 24057 + }, + { + "epoch": 0.24058, + "grad_norm": 0.8719248069864602, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 24058 + }, + { + "epoch": 0.24059, + "grad_norm": 1.1337309302615044, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 24059 + }, + { + "epoch": 0.2406, + "grad_norm": 1.2974719998325344, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 24060 + }, + { + "epoch": 0.24061, + "grad_norm": 0.9017808208024526, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 24061 + }, + { + "epoch": 0.24062, + "grad_norm": 0.8455730851011245, + "learning_rate": 0.003, + "loss": 4.039, + "step": 24062 + }, + { + "epoch": 0.24063, + "grad_norm": 0.8021571415114502, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 24063 + }, + { + "epoch": 0.24064, + "grad_norm": 0.852386976914747, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 24064 + }, + { + "epoch": 0.24065, + "grad_norm": 1.1028215101406798, + "learning_rate": 0.003, + "loss": 4.0938, + "step": 24065 + }, + { + "epoch": 0.24066, + "grad_norm": 1.1759942411143685, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 24066 + }, + { + "epoch": 0.24067, + "grad_norm": 0.9726639052275374, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 24067 + }, + { + "epoch": 0.24068, + "grad_norm": 0.9025777853177296, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 24068 + }, + { + "epoch": 0.24069, + "grad_norm": 0.8584729301288968, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 24069 + }, + { + "epoch": 0.2407, + "grad_norm": 1.0907675257832585, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 24070 + }, + { + "epoch": 0.24071, + "grad_norm": 1.3252726506590422, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 24071 + }, + { + "epoch": 0.24072, + "grad_norm": 0.9458459276233341, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 24072 + }, + { + "epoch": 0.24073, + "grad_norm": 1.1314362202522183, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 24073 + }, + { + "epoch": 0.24074, + "grad_norm": 1.2364184908684712, + "learning_rate": 0.003, + "loss": 4.0949, + "step": 24074 + }, + { + "epoch": 0.24075, + "grad_norm": 0.8668163221624492, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 24075 + }, + { + "epoch": 0.24076, + "grad_norm": 0.9112848724505388, + "learning_rate": 0.003, + "loss": 4.0993, + "step": 24076 + }, + { + "epoch": 0.24077, + "grad_norm": 1.1041455333904564, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 24077 + }, + { + "epoch": 0.24078, + "grad_norm": 1.0176919415938155, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 24078 + }, + { + "epoch": 0.24079, + "grad_norm": 1.0083775238464336, + "learning_rate": 0.003, + "loss": 4.066, + "step": 24079 + }, + { + "epoch": 0.2408, + "grad_norm": 1.132020435525448, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 24080 + }, + { + "epoch": 0.24081, + "grad_norm": 0.8445470658147082, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 24081 + }, + { + "epoch": 0.24082, + "grad_norm": 1.1115658690328167, + "learning_rate": 0.003, + "loss": 4.047, + "step": 24082 + }, + { + "epoch": 0.24083, + "grad_norm": 1.0736596661647817, + "learning_rate": 0.003, + "loss": 4.062, + "step": 24083 + }, + { + "epoch": 0.24084, + "grad_norm": 1.1325531125962927, + "learning_rate": 0.003, + "loss": 4.0991, + "step": 24084 + }, + { + "epoch": 0.24085, + "grad_norm": 0.8191562538832515, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 24085 + }, + { + "epoch": 0.24086, + "grad_norm": 0.9119082660069995, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 24086 + }, + { + "epoch": 0.24087, + "grad_norm": 1.2524938047811875, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 24087 + }, + { + "epoch": 0.24088, + "grad_norm": 1.3007358936816038, + "learning_rate": 0.003, + "loss": 4.07, + "step": 24088 + }, + { + "epoch": 0.24089, + "grad_norm": 0.8853275121673594, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 24089 + }, + { + "epoch": 0.2409, + "grad_norm": 1.0291357896247795, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 24090 + }, + { + "epoch": 0.24091, + "grad_norm": 1.1097191127364319, + "learning_rate": 0.003, + "loss": 4.058, + "step": 24091 + }, + { + "epoch": 0.24092, + "grad_norm": 1.1791032437529645, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 24092 + }, + { + "epoch": 0.24093, + "grad_norm": 1.2330266872071856, + "learning_rate": 0.003, + "loss": 4.011, + "step": 24093 + }, + { + "epoch": 0.24094, + "grad_norm": 1.0145479438372447, + "learning_rate": 0.003, + "loss": 4.026, + "step": 24094 + }, + { + "epoch": 0.24095, + "grad_norm": 1.1654600214956885, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 24095 + }, + { + "epoch": 0.24096, + "grad_norm": 1.14000943190656, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 24096 + }, + { + "epoch": 0.24097, + "grad_norm": 1.0877849418295429, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 24097 + }, + { + "epoch": 0.24098, + "grad_norm": 1.0645230270859263, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 24098 + }, + { + "epoch": 0.24099, + "grad_norm": 1.1179851628225876, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 24099 + }, + { + "epoch": 0.241, + "grad_norm": 0.8102347152184571, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 24100 + }, + { + "epoch": 0.24101, + "grad_norm": 0.9166559907789378, + "learning_rate": 0.003, + "loss": 4.038, + "step": 24101 + }, + { + "epoch": 0.24102, + "grad_norm": 1.2395464818707116, + "learning_rate": 0.003, + "loss": 4.048, + "step": 24102 + }, + { + "epoch": 0.24103, + "grad_norm": 1.0771962233405679, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 24103 + }, + { + "epoch": 0.24104, + "grad_norm": 1.2573065271709876, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 24104 + }, + { + "epoch": 0.24105, + "grad_norm": 0.9146224211892161, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 24105 + }, + { + "epoch": 0.24106, + "grad_norm": 1.1153897739287957, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 24106 + }, + { + "epoch": 0.24107, + "grad_norm": 1.3594748705752069, + "learning_rate": 0.003, + "loss": 4.037, + "step": 24107 + }, + { + "epoch": 0.24108, + "grad_norm": 0.7079155539063188, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 24108 + }, + { + "epoch": 0.24109, + "grad_norm": 0.7617767903278616, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 24109 + }, + { + "epoch": 0.2411, + "grad_norm": 0.9505362332316223, + "learning_rate": 0.003, + "loss": 4.064, + "step": 24110 + }, + { + "epoch": 0.24111, + "grad_norm": 1.1220253532731097, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 24111 + }, + { + "epoch": 0.24112, + "grad_norm": 0.9986520109483957, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 24112 + }, + { + "epoch": 0.24113, + "grad_norm": 1.5117584555352745, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 24113 + }, + { + "epoch": 0.24114, + "grad_norm": 0.8844652614590492, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 24114 + }, + { + "epoch": 0.24115, + "grad_norm": 0.9176602371411989, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 24115 + }, + { + "epoch": 0.24116, + "grad_norm": 1.0119625655899793, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 24116 + }, + { + "epoch": 0.24117, + "grad_norm": 1.2290080478875722, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 24117 + }, + { + "epoch": 0.24118, + "grad_norm": 0.996948144501788, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 24118 + }, + { + "epoch": 0.24119, + "grad_norm": 1.0471276075806757, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 24119 + }, + { + "epoch": 0.2412, + "grad_norm": 1.1598938807688015, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 24120 + }, + { + "epoch": 0.24121, + "grad_norm": 1.0012341391249389, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 24121 + }, + { + "epoch": 0.24122, + "grad_norm": 1.0284702756890542, + "learning_rate": 0.003, + "loss": 4.013, + "step": 24122 + }, + { + "epoch": 0.24123, + "grad_norm": 1.09825148229697, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 24123 + }, + { + "epoch": 0.24124, + "grad_norm": 1.2969506828540118, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 24124 + }, + { + "epoch": 0.24125, + "grad_norm": 0.9847219376277163, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 24125 + }, + { + "epoch": 0.24126, + "grad_norm": 1.0920870701896896, + "learning_rate": 0.003, + "loss": 4.053, + "step": 24126 + }, + { + "epoch": 0.24127, + "grad_norm": 0.967886048781256, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 24127 + }, + { + "epoch": 0.24128, + "grad_norm": 1.1657958128371344, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 24128 + }, + { + "epoch": 0.24129, + "grad_norm": 1.1305745764678012, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 24129 + }, + { + "epoch": 0.2413, + "grad_norm": 1.0092076314021217, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 24130 + }, + { + "epoch": 0.24131, + "grad_norm": 1.219810453413088, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 24131 + }, + { + "epoch": 0.24132, + "grad_norm": 1.0531758946825154, + "learning_rate": 0.003, + "loss": 4.042, + "step": 24132 + }, + { + "epoch": 0.24133, + "grad_norm": 1.1111557082866665, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 24133 + }, + { + "epoch": 0.24134, + "grad_norm": 1.000623815684043, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 24134 + }, + { + "epoch": 0.24135, + "grad_norm": 1.3438277302156427, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 24135 + }, + { + "epoch": 0.24136, + "grad_norm": 1.323634774182362, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 24136 + }, + { + "epoch": 0.24137, + "grad_norm": 0.9956283896923329, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 24137 + }, + { + "epoch": 0.24138, + "grad_norm": 1.1601830105986761, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 24138 + }, + { + "epoch": 0.24139, + "grad_norm": 1.1884183757846036, + "learning_rate": 0.003, + "loss": 4.052, + "step": 24139 + }, + { + "epoch": 0.2414, + "grad_norm": 0.9917738974524729, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 24140 + }, + { + "epoch": 0.24141, + "grad_norm": 1.0963891822097478, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 24141 + }, + { + "epoch": 0.24142, + "grad_norm": 1.0645601162009928, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 24142 + }, + { + "epoch": 0.24143, + "grad_norm": 1.0728525165351057, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 24143 + }, + { + "epoch": 0.24144, + "grad_norm": 1.0798894435564188, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 24144 + }, + { + "epoch": 0.24145, + "grad_norm": 1.034801274757726, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 24145 + }, + { + "epoch": 0.24146, + "grad_norm": 1.084128040742064, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 24146 + }, + { + "epoch": 0.24147, + "grad_norm": 1.1171568145889692, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 24147 + }, + { + "epoch": 0.24148, + "grad_norm": 1.0452720363777188, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 24148 + }, + { + "epoch": 0.24149, + "grad_norm": 1.2840072481112235, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 24149 + }, + { + "epoch": 0.2415, + "grad_norm": 0.960334428498968, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 24150 + }, + { + "epoch": 0.24151, + "grad_norm": 1.0948420231285805, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 24151 + }, + { + "epoch": 0.24152, + "grad_norm": 1.3548224307571008, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 24152 + }, + { + "epoch": 0.24153, + "grad_norm": 0.7744292398646301, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 24153 + }, + { + "epoch": 0.24154, + "grad_norm": 0.7240703434580535, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 24154 + }, + { + "epoch": 0.24155, + "grad_norm": 0.9224852535086641, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 24155 + }, + { + "epoch": 0.24156, + "grad_norm": 1.1195049800396883, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 24156 + }, + { + "epoch": 0.24157, + "grad_norm": 0.92654721298054, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 24157 + }, + { + "epoch": 0.24158, + "grad_norm": 1.1682499803430817, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 24158 + }, + { + "epoch": 0.24159, + "grad_norm": 0.9920776946975021, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 24159 + }, + { + "epoch": 0.2416, + "grad_norm": 1.2799746122396916, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 24160 + }, + { + "epoch": 0.24161, + "grad_norm": 0.9864698190625147, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 24161 + }, + { + "epoch": 0.24162, + "grad_norm": 1.164176257157679, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 24162 + }, + { + "epoch": 0.24163, + "grad_norm": 0.9097546445679869, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 24163 + }, + { + "epoch": 0.24164, + "grad_norm": 0.9320887569876325, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 24164 + }, + { + "epoch": 0.24165, + "grad_norm": 1.0674942877575373, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 24165 + }, + { + "epoch": 0.24166, + "grad_norm": 1.074230325492355, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 24166 + }, + { + "epoch": 0.24167, + "grad_norm": 1.043741384932895, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 24167 + }, + { + "epoch": 0.24168, + "grad_norm": 1.077210864029984, + "learning_rate": 0.003, + "loss": 4.0968, + "step": 24168 + }, + { + "epoch": 0.24169, + "grad_norm": 1.0389740837547172, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 24169 + }, + { + "epoch": 0.2417, + "grad_norm": 1.0554762011426801, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 24170 + }, + { + "epoch": 0.24171, + "grad_norm": 1.0010706095668231, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 24171 + }, + { + "epoch": 0.24172, + "grad_norm": 1.0170776202491962, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 24172 + }, + { + "epoch": 0.24173, + "grad_norm": 1.0348718538434551, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 24173 + }, + { + "epoch": 0.24174, + "grad_norm": 1.2072533514124322, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 24174 + }, + { + "epoch": 0.24175, + "grad_norm": 1.0532538284726365, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 24175 + }, + { + "epoch": 0.24176, + "grad_norm": 1.2147432502647997, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 24176 + }, + { + "epoch": 0.24177, + "grad_norm": 1.1000932744950456, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 24177 + }, + { + "epoch": 0.24178, + "grad_norm": 1.2670972446128115, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 24178 + }, + { + "epoch": 0.24179, + "grad_norm": 0.8012685527368546, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 24179 + }, + { + "epoch": 0.2418, + "grad_norm": 0.8209371635174778, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 24180 + }, + { + "epoch": 0.24181, + "grad_norm": 0.9482428244147603, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 24181 + }, + { + "epoch": 0.24182, + "grad_norm": 1.0355140098073154, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 24182 + }, + { + "epoch": 0.24183, + "grad_norm": 1.1655864837712815, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 24183 + }, + { + "epoch": 0.24184, + "grad_norm": 1.214543702146661, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 24184 + }, + { + "epoch": 0.24185, + "grad_norm": 0.9557962286158198, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 24185 + }, + { + "epoch": 0.24186, + "grad_norm": 1.0647010113496371, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 24186 + }, + { + "epoch": 0.24187, + "grad_norm": 1.1476059518636283, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 24187 + }, + { + "epoch": 0.24188, + "grad_norm": 1.0141042206668103, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 24188 + }, + { + "epoch": 0.24189, + "grad_norm": 1.033258908880472, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 24189 + }, + { + "epoch": 0.2419, + "grad_norm": 1.2020923491253248, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 24190 + }, + { + "epoch": 0.24191, + "grad_norm": 0.9078934959944964, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 24191 + }, + { + "epoch": 0.24192, + "grad_norm": 0.9621663258875008, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 24192 + }, + { + "epoch": 0.24193, + "grad_norm": 1.0272138852876802, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 24193 + }, + { + "epoch": 0.24194, + "grad_norm": 1.2039558337204377, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 24194 + }, + { + "epoch": 0.24195, + "grad_norm": 1.0616682721745643, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 24195 + }, + { + "epoch": 0.24196, + "grad_norm": 1.257985624342846, + "learning_rate": 0.003, + "loss": 4.101, + "step": 24196 + }, + { + "epoch": 0.24197, + "grad_norm": 0.998338300240225, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 24197 + }, + { + "epoch": 0.24198, + "grad_norm": 0.9561200679358208, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 24198 + }, + { + "epoch": 0.24199, + "grad_norm": 1.0768053980003776, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 24199 + }, + { + "epoch": 0.242, + "grad_norm": 1.124738558376181, + "learning_rate": 0.003, + "loss": 4.077, + "step": 24200 + }, + { + "epoch": 0.24201, + "grad_norm": 1.184338797300607, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 24201 + }, + { + "epoch": 0.24202, + "grad_norm": 1.2388633633773751, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 24202 + }, + { + "epoch": 0.24203, + "grad_norm": 0.9155328724787437, + "learning_rate": 0.003, + "loss": 4.043, + "step": 24203 + }, + { + "epoch": 0.24204, + "grad_norm": 1.0036501373886864, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 24204 + }, + { + "epoch": 0.24205, + "grad_norm": 1.3831832175051775, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 24205 + }, + { + "epoch": 0.24206, + "grad_norm": 1.0026759848198694, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 24206 + }, + { + "epoch": 0.24207, + "grad_norm": 1.377202628194043, + "learning_rate": 0.003, + "loss": 4.0848, + "step": 24207 + }, + { + "epoch": 0.24208, + "grad_norm": 1.0853332434190097, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 24208 + }, + { + "epoch": 0.24209, + "grad_norm": 1.0643822319217369, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 24209 + }, + { + "epoch": 0.2421, + "grad_norm": 1.1533356255292895, + "learning_rate": 0.003, + "loss": 4.073, + "step": 24210 + }, + { + "epoch": 0.24211, + "grad_norm": 1.0975930038610404, + "learning_rate": 0.003, + "loss": 4.086, + "step": 24211 + }, + { + "epoch": 0.24212, + "grad_norm": 0.9028215001477858, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 24212 + }, + { + "epoch": 0.24213, + "grad_norm": 0.9993411757732606, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 24213 + }, + { + "epoch": 0.24214, + "grad_norm": 1.1898075459425677, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 24214 + }, + { + "epoch": 0.24215, + "grad_norm": 1.0289523009723405, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 24215 + }, + { + "epoch": 0.24216, + "grad_norm": 1.0596647730187214, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 24216 + }, + { + "epoch": 0.24217, + "grad_norm": 0.9428329380464351, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 24217 + }, + { + "epoch": 0.24218, + "grad_norm": 0.9623135487217724, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 24218 + }, + { + "epoch": 0.24219, + "grad_norm": 1.1055214965714408, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 24219 + }, + { + "epoch": 0.2422, + "grad_norm": 0.8633183165981227, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 24220 + }, + { + "epoch": 0.24221, + "grad_norm": 0.8932232120377513, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 24221 + }, + { + "epoch": 0.24222, + "grad_norm": 1.072352649868967, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 24222 + }, + { + "epoch": 0.24223, + "grad_norm": 1.4808311135766419, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 24223 + }, + { + "epoch": 0.24224, + "grad_norm": 0.9818979601631758, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 24224 + }, + { + "epoch": 0.24225, + "grad_norm": 1.291937442833253, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 24225 + }, + { + "epoch": 0.24226, + "grad_norm": 1.0365915070527956, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 24226 + }, + { + "epoch": 0.24227, + "grad_norm": 1.0972734793856858, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 24227 + }, + { + "epoch": 0.24228, + "grad_norm": 1.1094238773766028, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 24228 + }, + { + "epoch": 0.24229, + "grad_norm": 0.9131644565907042, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 24229 + }, + { + "epoch": 0.2423, + "grad_norm": 1.0103776480500528, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 24230 + }, + { + "epoch": 0.24231, + "grad_norm": 1.0291213217300565, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 24231 + }, + { + "epoch": 0.24232, + "grad_norm": 1.014410337229161, + "learning_rate": 0.003, + "loss": 4.024, + "step": 24232 + }, + { + "epoch": 0.24233, + "grad_norm": 1.1551046688253361, + "learning_rate": 0.003, + "loss": 4.063, + "step": 24233 + }, + { + "epoch": 0.24234, + "grad_norm": 1.0577993948416384, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 24234 + }, + { + "epoch": 0.24235, + "grad_norm": 1.0084616901132737, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 24235 + }, + { + "epoch": 0.24236, + "grad_norm": 1.012172871298913, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 24236 + }, + { + "epoch": 0.24237, + "grad_norm": 1.0028146583516784, + "learning_rate": 0.003, + "loss": 4.049, + "step": 24237 + }, + { + "epoch": 0.24238, + "grad_norm": 1.0004847823104852, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 24238 + }, + { + "epoch": 0.24239, + "grad_norm": 1.0817492116999445, + "learning_rate": 0.003, + "loss": 4.07, + "step": 24239 + }, + { + "epoch": 0.2424, + "grad_norm": 1.1109115836843229, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 24240 + }, + { + "epoch": 0.24241, + "grad_norm": 0.9749936375692445, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 24241 + }, + { + "epoch": 0.24242, + "grad_norm": 1.067156235978742, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 24242 + }, + { + "epoch": 0.24243, + "grad_norm": 0.8890299121419734, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 24243 + }, + { + "epoch": 0.24244, + "grad_norm": 1.001161043476274, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 24244 + }, + { + "epoch": 0.24245, + "grad_norm": 1.3111114328214892, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 24245 + }, + { + "epoch": 0.24246, + "grad_norm": 1.2625759423838512, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 24246 + }, + { + "epoch": 0.24247, + "grad_norm": 1.1484524842331731, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 24247 + }, + { + "epoch": 0.24248, + "grad_norm": 1.240178067320192, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 24248 + }, + { + "epoch": 0.24249, + "grad_norm": 1.1868110526457374, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 24249 + }, + { + "epoch": 0.2425, + "grad_norm": 0.8903760795383244, + "learning_rate": 0.003, + "loss": 4.04, + "step": 24250 + }, + { + "epoch": 0.24251, + "grad_norm": 0.9268451319936355, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 24251 + }, + { + "epoch": 0.24252, + "grad_norm": 1.0247329118442787, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 24252 + }, + { + "epoch": 0.24253, + "grad_norm": 1.4666349187239602, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 24253 + }, + { + "epoch": 0.24254, + "grad_norm": 1.0249002629428363, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 24254 + }, + { + "epoch": 0.24255, + "grad_norm": 0.9419202258160086, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 24255 + }, + { + "epoch": 0.24256, + "grad_norm": 1.1687258536725087, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 24256 + }, + { + "epoch": 0.24257, + "grad_norm": 1.146177595283099, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 24257 + }, + { + "epoch": 0.24258, + "grad_norm": 1.1025361157464986, + "learning_rate": 0.003, + "loss": 4.045, + "step": 24258 + }, + { + "epoch": 0.24259, + "grad_norm": 1.0322540198139036, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 24259 + }, + { + "epoch": 0.2426, + "grad_norm": 1.038061161725506, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 24260 + }, + { + "epoch": 0.24261, + "grad_norm": 0.9596744463356618, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 24261 + }, + { + "epoch": 0.24262, + "grad_norm": 1.0052588868354069, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 24262 + }, + { + "epoch": 0.24263, + "grad_norm": 1.1787962224844712, + "learning_rate": 0.003, + "loss": 4.067, + "step": 24263 + }, + { + "epoch": 0.24264, + "grad_norm": 1.054724354832868, + "learning_rate": 0.003, + "loss": 4.089, + "step": 24264 + }, + { + "epoch": 0.24265, + "grad_norm": 1.0998644407020164, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 24265 + }, + { + "epoch": 0.24266, + "grad_norm": 0.9553362812822045, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 24266 + }, + { + "epoch": 0.24267, + "grad_norm": 1.157859018376865, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 24267 + }, + { + "epoch": 0.24268, + "grad_norm": 1.1140924908406056, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 24268 + }, + { + "epoch": 0.24269, + "grad_norm": 1.040105726090819, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 24269 + }, + { + "epoch": 0.2427, + "grad_norm": 1.1622001929443904, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 24270 + }, + { + "epoch": 0.24271, + "grad_norm": 1.0998975255608927, + "learning_rate": 0.003, + "loss": 4.031, + "step": 24271 + }, + { + "epoch": 0.24272, + "grad_norm": 1.0015716618279493, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 24272 + }, + { + "epoch": 0.24273, + "grad_norm": 0.9843430201747475, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 24273 + }, + { + "epoch": 0.24274, + "grad_norm": 1.1561092981585157, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 24274 + }, + { + "epoch": 0.24275, + "grad_norm": 1.18336859842934, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 24275 + }, + { + "epoch": 0.24276, + "grad_norm": 1.2193841259025893, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 24276 + }, + { + "epoch": 0.24277, + "grad_norm": 1.0068624734046727, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 24277 + }, + { + "epoch": 0.24278, + "grad_norm": 0.9027220593045131, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 24278 + }, + { + "epoch": 0.24279, + "grad_norm": 0.8792435975504496, + "learning_rate": 0.003, + "loss": 4.05, + "step": 24279 + }, + { + "epoch": 0.2428, + "grad_norm": 0.9611822576476848, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 24280 + }, + { + "epoch": 0.24281, + "grad_norm": 1.0731414376676602, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 24281 + }, + { + "epoch": 0.24282, + "grad_norm": 1.0963294587756387, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 24282 + }, + { + "epoch": 0.24283, + "grad_norm": 1.0715451257560378, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 24283 + }, + { + "epoch": 0.24284, + "grad_norm": 1.0405270364075438, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 24284 + }, + { + "epoch": 0.24285, + "grad_norm": 0.9845205775309261, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 24285 + }, + { + "epoch": 0.24286, + "grad_norm": 1.2565452532537065, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 24286 + }, + { + "epoch": 0.24287, + "grad_norm": 1.176939796355312, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 24287 + }, + { + "epoch": 0.24288, + "grad_norm": 1.1975554157036998, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 24288 + }, + { + "epoch": 0.24289, + "grad_norm": 1.1144226066298462, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 24289 + }, + { + "epoch": 0.2429, + "grad_norm": 1.1425712609325953, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 24290 + }, + { + "epoch": 0.24291, + "grad_norm": 0.9053777238675509, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 24291 + }, + { + "epoch": 0.24292, + "grad_norm": 0.9810922913058597, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 24292 + }, + { + "epoch": 0.24293, + "grad_norm": 1.269001641031243, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 24293 + }, + { + "epoch": 0.24294, + "grad_norm": 0.913044867480457, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 24294 + }, + { + "epoch": 0.24295, + "grad_norm": 0.7787197450826933, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 24295 + }, + { + "epoch": 0.24296, + "grad_norm": 0.9464150197153832, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 24296 + }, + { + "epoch": 0.24297, + "grad_norm": 1.20467946969929, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 24297 + }, + { + "epoch": 0.24298, + "grad_norm": 0.8877428540087946, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 24298 + }, + { + "epoch": 0.24299, + "grad_norm": 1.0881210445702143, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 24299 + }, + { + "epoch": 0.243, + "grad_norm": 1.1397515535948115, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 24300 + }, + { + "epoch": 0.24301, + "grad_norm": 0.9206550635803028, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 24301 + }, + { + "epoch": 0.24302, + "grad_norm": 1.0223494914105815, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 24302 + }, + { + "epoch": 0.24303, + "grad_norm": 1.1455789605637539, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 24303 + }, + { + "epoch": 0.24304, + "grad_norm": 1.223734434499614, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 24304 + }, + { + "epoch": 0.24305, + "grad_norm": 1.1627831972084723, + "learning_rate": 0.003, + "loss": 4.062, + "step": 24305 + }, + { + "epoch": 0.24306, + "grad_norm": 0.9099201528384143, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 24306 + }, + { + "epoch": 0.24307, + "grad_norm": 0.7522109086159879, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 24307 + }, + { + "epoch": 0.24308, + "grad_norm": 0.7684704335211698, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 24308 + }, + { + "epoch": 0.24309, + "grad_norm": 0.8966031962351703, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 24309 + }, + { + "epoch": 0.2431, + "grad_norm": 1.049217899591865, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 24310 + }, + { + "epoch": 0.24311, + "grad_norm": 0.9839194049451325, + "learning_rate": 0.003, + "loss": 4.045, + "step": 24311 + }, + { + "epoch": 0.24312, + "grad_norm": 1.138659209593223, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 24312 + }, + { + "epoch": 0.24313, + "grad_norm": 1.3264510393177016, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 24313 + }, + { + "epoch": 0.24314, + "grad_norm": 0.9250826137923849, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 24314 + }, + { + "epoch": 0.24315, + "grad_norm": 1.0274282507269878, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 24315 + }, + { + "epoch": 0.24316, + "grad_norm": 1.1614021682393783, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 24316 + }, + { + "epoch": 0.24317, + "grad_norm": 1.1687420212890054, + "learning_rate": 0.003, + "loss": 4.0908, + "step": 24317 + }, + { + "epoch": 0.24318, + "grad_norm": 0.9776072993497097, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 24318 + }, + { + "epoch": 0.24319, + "grad_norm": 1.0392598925236938, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 24319 + }, + { + "epoch": 0.2432, + "grad_norm": 1.128692679642286, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 24320 + }, + { + "epoch": 0.24321, + "grad_norm": 1.0137188465486153, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 24321 + }, + { + "epoch": 0.24322, + "grad_norm": 1.1640022237236531, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 24322 + }, + { + "epoch": 0.24323, + "grad_norm": 0.9313293476160628, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 24323 + }, + { + "epoch": 0.24324, + "grad_norm": 1.195368554254754, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 24324 + }, + { + "epoch": 0.24325, + "grad_norm": 1.0989404075522071, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 24325 + }, + { + "epoch": 0.24326, + "grad_norm": 1.249022875909844, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 24326 + }, + { + "epoch": 0.24327, + "grad_norm": 1.0252351827478892, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 24327 + }, + { + "epoch": 0.24328, + "grad_norm": 1.16877078963475, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 24328 + }, + { + "epoch": 0.24329, + "grad_norm": 0.9845075160190336, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 24329 + }, + { + "epoch": 0.2433, + "grad_norm": 1.332261774424252, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 24330 + }, + { + "epoch": 0.24331, + "grad_norm": 0.8527331270905638, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 24331 + }, + { + "epoch": 0.24332, + "grad_norm": 0.9543887456655874, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 24332 + }, + { + "epoch": 0.24333, + "grad_norm": 1.1106325541621593, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 24333 + }, + { + "epoch": 0.24334, + "grad_norm": 1.1552948656074595, + "learning_rate": 0.003, + "loss": 4.054, + "step": 24334 + }, + { + "epoch": 0.24335, + "grad_norm": 1.0345942867424502, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 24335 + }, + { + "epoch": 0.24336, + "grad_norm": 1.1867691856048583, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 24336 + }, + { + "epoch": 0.24337, + "grad_norm": 1.0766351085146495, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 24337 + }, + { + "epoch": 0.24338, + "grad_norm": 0.9666703015815187, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 24338 + }, + { + "epoch": 0.24339, + "grad_norm": 1.171836877541629, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 24339 + }, + { + "epoch": 0.2434, + "grad_norm": 0.9902835648673942, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 24340 + }, + { + "epoch": 0.24341, + "grad_norm": 1.007872550518532, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 24341 + }, + { + "epoch": 0.24342, + "grad_norm": 1.1240054347543786, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 24342 + }, + { + "epoch": 0.24343, + "grad_norm": 0.9986393026543436, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 24343 + }, + { + "epoch": 0.24344, + "grad_norm": 1.065373216883283, + "learning_rate": 0.003, + "loss": 4.068, + "step": 24344 + }, + { + "epoch": 0.24345, + "grad_norm": 1.0841546751720286, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 24345 + }, + { + "epoch": 0.24346, + "grad_norm": 0.9435403884105734, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 24346 + }, + { + "epoch": 0.24347, + "grad_norm": 0.9176540333598013, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 24347 + }, + { + "epoch": 0.24348, + "grad_norm": 0.9889264676438655, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 24348 + }, + { + "epoch": 0.24349, + "grad_norm": 1.095786520049029, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 24349 + }, + { + "epoch": 0.2435, + "grad_norm": 0.9987498187204157, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 24350 + }, + { + "epoch": 0.24351, + "grad_norm": 1.1350696193890728, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 24351 + }, + { + "epoch": 0.24352, + "grad_norm": 1.0053485454389648, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 24352 + }, + { + "epoch": 0.24353, + "grad_norm": 1.148533418193212, + "learning_rate": 0.003, + "loss": 4.053, + "step": 24353 + }, + { + "epoch": 0.24354, + "grad_norm": 1.1846381179540963, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 24354 + }, + { + "epoch": 0.24355, + "grad_norm": 1.1325920940053127, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 24355 + }, + { + "epoch": 0.24356, + "grad_norm": 1.0119599681615965, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 24356 + }, + { + "epoch": 0.24357, + "grad_norm": 0.9231271482429225, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 24357 + }, + { + "epoch": 0.24358, + "grad_norm": 1.1570320282933761, + "learning_rate": 0.003, + "loss": 4.048, + "step": 24358 + }, + { + "epoch": 0.24359, + "grad_norm": 1.2393230855231199, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 24359 + }, + { + "epoch": 0.2436, + "grad_norm": 0.9315407652484744, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 24360 + }, + { + "epoch": 0.24361, + "grad_norm": 1.1165883964777987, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 24361 + }, + { + "epoch": 0.24362, + "grad_norm": 1.2828512848546922, + "learning_rate": 0.003, + "loss": 4.05, + "step": 24362 + }, + { + "epoch": 0.24363, + "grad_norm": 0.8986061575926603, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 24363 + }, + { + "epoch": 0.24364, + "grad_norm": 0.8642852849425261, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 24364 + }, + { + "epoch": 0.24365, + "grad_norm": 0.9487762363908786, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 24365 + }, + { + "epoch": 0.24366, + "grad_norm": 1.0450952262297653, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 24366 + }, + { + "epoch": 0.24367, + "grad_norm": 1.0954364154263105, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 24367 + }, + { + "epoch": 0.24368, + "grad_norm": 1.171434883869945, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 24368 + }, + { + "epoch": 0.24369, + "grad_norm": 1.0528685898507744, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 24369 + }, + { + "epoch": 0.2437, + "grad_norm": 1.167225099613884, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 24370 + }, + { + "epoch": 0.24371, + "grad_norm": 0.8960110331436049, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 24371 + }, + { + "epoch": 0.24372, + "grad_norm": 1.1440934513440069, + "learning_rate": 0.003, + "loss": 4.063, + "step": 24372 + }, + { + "epoch": 0.24373, + "grad_norm": 1.2979194523877857, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 24373 + }, + { + "epoch": 0.24374, + "grad_norm": 0.853683656380877, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 24374 + }, + { + "epoch": 0.24375, + "grad_norm": 0.9737576478144697, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 24375 + }, + { + "epoch": 0.24376, + "grad_norm": 1.2654304854661242, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 24376 + }, + { + "epoch": 0.24377, + "grad_norm": 0.8457245748691913, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 24377 + }, + { + "epoch": 0.24378, + "grad_norm": 1.08100429768971, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 24378 + }, + { + "epoch": 0.24379, + "grad_norm": 1.187572470716943, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 24379 + }, + { + "epoch": 0.2438, + "grad_norm": 1.1196013800447853, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 24380 + }, + { + "epoch": 0.24381, + "grad_norm": 1.1195527012124065, + "learning_rate": 0.003, + "loss": 4.059, + "step": 24381 + }, + { + "epoch": 0.24382, + "grad_norm": 1.0263094118329879, + "learning_rate": 0.003, + "loss": 4.054, + "step": 24382 + }, + { + "epoch": 0.24383, + "grad_norm": 1.0459352097741361, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 24383 + }, + { + "epoch": 0.24384, + "grad_norm": 1.0599498957520028, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 24384 + }, + { + "epoch": 0.24385, + "grad_norm": 0.974211872606265, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 24385 + }, + { + "epoch": 0.24386, + "grad_norm": 1.1181430389876574, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 24386 + }, + { + "epoch": 0.24387, + "grad_norm": 1.162791798997232, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 24387 + }, + { + "epoch": 0.24388, + "grad_norm": 1.0305446895064758, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 24388 + }, + { + "epoch": 0.24389, + "grad_norm": 0.8901643916078669, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 24389 + }, + { + "epoch": 0.2439, + "grad_norm": 1.0530306870800499, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 24390 + }, + { + "epoch": 0.24391, + "grad_norm": 1.0051611689743172, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 24391 + }, + { + "epoch": 0.24392, + "grad_norm": 1.0071380467303397, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 24392 + }, + { + "epoch": 0.24393, + "grad_norm": 1.00368719707391, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 24393 + }, + { + "epoch": 0.24394, + "grad_norm": 1.0824991081660387, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 24394 + }, + { + "epoch": 0.24395, + "grad_norm": 1.0888025771436318, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 24395 + }, + { + "epoch": 0.24396, + "grad_norm": 1.2986297876157964, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 24396 + }, + { + "epoch": 0.24397, + "grad_norm": 1.1962890394272405, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 24397 + }, + { + "epoch": 0.24398, + "grad_norm": 0.9347925372662131, + "learning_rate": 0.003, + "loss": 4.035, + "step": 24398 + }, + { + "epoch": 0.24399, + "grad_norm": 0.9845994783162298, + "learning_rate": 0.003, + "loss": 4.065, + "step": 24399 + }, + { + "epoch": 0.244, + "grad_norm": 1.2382403882763744, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 24400 + }, + { + "epoch": 0.24401, + "grad_norm": 1.0533591302893304, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 24401 + }, + { + "epoch": 0.24402, + "grad_norm": 1.145938733489789, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 24402 + }, + { + "epoch": 0.24403, + "grad_norm": 1.267062886643476, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 24403 + }, + { + "epoch": 0.24404, + "grad_norm": 1.0168186974552855, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 24404 + }, + { + "epoch": 0.24405, + "grad_norm": 1.2089954806755698, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 24405 + }, + { + "epoch": 0.24406, + "grad_norm": 0.8308035835102188, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 24406 + }, + { + "epoch": 0.24407, + "grad_norm": 0.9659052569697022, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 24407 + }, + { + "epoch": 0.24408, + "grad_norm": 1.3003882215461582, + "learning_rate": 0.003, + "loss": 4.057, + "step": 24408 + }, + { + "epoch": 0.24409, + "grad_norm": 0.9421419655496542, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 24409 + }, + { + "epoch": 0.2441, + "grad_norm": 0.9772508944410727, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 24410 + }, + { + "epoch": 0.24411, + "grad_norm": 1.0955072966611856, + "learning_rate": 0.003, + "loss": 4.039, + "step": 24411 + }, + { + "epoch": 0.24412, + "grad_norm": 1.19781925818305, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 24412 + }, + { + "epoch": 0.24413, + "grad_norm": 1.017288128860357, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 24413 + }, + { + "epoch": 0.24414, + "grad_norm": 1.1583813464651636, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 24414 + }, + { + "epoch": 0.24415, + "grad_norm": 1.074358783009818, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 24415 + }, + { + "epoch": 0.24416, + "grad_norm": 0.9747509933429614, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 24416 + }, + { + "epoch": 0.24417, + "grad_norm": 0.9732619608337425, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 24417 + }, + { + "epoch": 0.24418, + "grad_norm": 1.1599147190291224, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 24418 + }, + { + "epoch": 0.24419, + "grad_norm": 1.0434547954631437, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 24419 + }, + { + "epoch": 0.2442, + "grad_norm": 1.103987607147282, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 24420 + }, + { + "epoch": 0.24421, + "grad_norm": 1.0648577224727558, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 24421 + }, + { + "epoch": 0.24422, + "grad_norm": 1.0715671851431805, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 24422 + }, + { + "epoch": 0.24423, + "grad_norm": 1.1189930198582487, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 24423 + }, + { + "epoch": 0.24424, + "grad_norm": 1.230392282430919, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 24424 + }, + { + "epoch": 0.24425, + "grad_norm": 1.084617807323761, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 24425 + }, + { + "epoch": 0.24426, + "grad_norm": 1.044915820715204, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 24426 + }, + { + "epoch": 0.24427, + "grad_norm": 1.284051199141401, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 24427 + }, + { + "epoch": 0.24428, + "grad_norm": 0.877915031842583, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 24428 + }, + { + "epoch": 0.24429, + "grad_norm": 1.018671881472148, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 24429 + }, + { + "epoch": 0.2443, + "grad_norm": 1.1061560284151357, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 24430 + }, + { + "epoch": 0.24431, + "grad_norm": 0.9203434689392098, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 24431 + }, + { + "epoch": 0.24432, + "grad_norm": 1.1172916391566607, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 24432 + }, + { + "epoch": 0.24433, + "grad_norm": 1.2095793490342932, + "learning_rate": 0.003, + "loss": 4.073, + "step": 24433 + }, + { + "epoch": 0.24434, + "grad_norm": 1.2096113620841014, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 24434 + }, + { + "epoch": 0.24435, + "grad_norm": 0.9501522698528554, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 24435 + }, + { + "epoch": 0.24436, + "grad_norm": 1.049287879224242, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 24436 + }, + { + "epoch": 0.24437, + "grad_norm": 1.1850767069791839, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 24437 + }, + { + "epoch": 0.24438, + "grad_norm": 1.0989466610375618, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 24438 + }, + { + "epoch": 0.24439, + "grad_norm": 1.0914023905592214, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 24439 + }, + { + "epoch": 0.2444, + "grad_norm": 1.042961923628348, + "learning_rate": 0.003, + "loss": 4.053, + "step": 24440 + }, + { + "epoch": 0.24441, + "grad_norm": 1.1507555131877132, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 24441 + }, + { + "epoch": 0.24442, + "grad_norm": 1.1567555956598687, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 24442 + }, + { + "epoch": 0.24443, + "grad_norm": 1.0161140790471654, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 24443 + }, + { + "epoch": 0.24444, + "grad_norm": 1.0542846212996573, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 24444 + }, + { + "epoch": 0.24445, + "grad_norm": 1.21240322567265, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 24445 + }, + { + "epoch": 0.24446, + "grad_norm": 0.9816498227597553, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 24446 + }, + { + "epoch": 0.24447, + "grad_norm": 1.0592434610918047, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 24447 + }, + { + "epoch": 0.24448, + "grad_norm": 1.048122768172596, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 24448 + }, + { + "epoch": 0.24449, + "grad_norm": 1.0749916678564537, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 24449 + }, + { + "epoch": 0.2445, + "grad_norm": 0.9680544664485675, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 24450 + }, + { + "epoch": 0.24451, + "grad_norm": 1.105973153475953, + "learning_rate": 0.003, + "loss": 4.059, + "step": 24451 + }, + { + "epoch": 0.24452, + "grad_norm": 1.1785202840098692, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 24452 + }, + { + "epoch": 0.24453, + "grad_norm": 0.9851046887969028, + "learning_rate": 0.003, + "loss": 4.068, + "step": 24453 + }, + { + "epoch": 0.24454, + "grad_norm": 1.0843960799918466, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 24454 + }, + { + "epoch": 0.24455, + "grad_norm": 1.078294963408901, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 24455 + }, + { + "epoch": 0.24456, + "grad_norm": 1.154329749432934, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 24456 + }, + { + "epoch": 0.24457, + "grad_norm": 0.9464613654617325, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 24457 + }, + { + "epoch": 0.24458, + "grad_norm": 0.9645166008219593, + "learning_rate": 0.003, + "loss": 4.069, + "step": 24458 + }, + { + "epoch": 0.24459, + "grad_norm": 1.1133181445954663, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 24459 + }, + { + "epoch": 0.2446, + "grad_norm": 1.003838995653022, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 24460 + }, + { + "epoch": 0.24461, + "grad_norm": 1.099904822645435, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 24461 + }, + { + "epoch": 0.24462, + "grad_norm": 1.1263366281538953, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 24462 + }, + { + "epoch": 0.24463, + "grad_norm": 1.1708798047457256, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 24463 + }, + { + "epoch": 0.24464, + "grad_norm": 0.8890112378583809, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 24464 + }, + { + "epoch": 0.24465, + "grad_norm": 0.9571128245653899, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 24465 + }, + { + "epoch": 0.24466, + "grad_norm": 1.1717241395200504, + "learning_rate": 0.003, + "loss": 4.08, + "step": 24466 + }, + { + "epoch": 0.24467, + "grad_norm": 1.139960030909704, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 24467 + }, + { + "epoch": 0.24468, + "grad_norm": 1.004612514866232, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 24468 + }, + { + "epoch": 0.24469, + "grad_norm": 1.1972300002153207, + "learning_rate": 0.003, + "loss": 4.1036, + "step": 24469 + }, + { + "epoch": 0.2447, + "grad_norm": 0.9794256820912776, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 24470 + }, + { + "epoch": 0.24471, + "grad_norm": 1.0474305483234372, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 24471 + }, + { + "epoch": 0.24472, + "grad_norm": 0.9633885145664801, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 24472 + }, + { + "epoch": 0.24473, + "grad_norm": 1.2859579898368854, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 24473 + }, + { + "epoch": 0.24474, + "grad_norm": 0.9830890451844011, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 24474 + }, + { + "epoch": 0.24475, + "grad_norm": 0.931304401950387, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 24475 + }, + { + "epoch": 0.24476, + "grad_norm": 0.9477029286091504, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 24476 + }, + { + "epoch": 0.24477, + "grad_norm": 1.054924714483454, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 24477 + }, + { + "epoch": 0.24478, + "grad_norm": 1.0995671530576965, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 24478 + }, + { + "epoch": 0.24479, + "grad_norm": 1.03621998574258, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 24479 + }, + { + "epoch": 0.2448, + "grad_norm": 0.9997218864536697, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 24480 + }, + { + "epoch": 0.24481, + "grad_norm": 1.334168576199029, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 24481 + }, + { + "epoch": 0.24482, + "grad_norm": 0.9663266568501995, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 24482 + }, + { + "epoch": 0.24483, + "grad_norm": 1.2731057640356582, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 24483 + }, + { + "epoch": 0.24484, + "grad_norm": 1.0655539121116473, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 24484 + }, + { + "epoch": 0.24485, + "grad_norm": 1.3786617832844588, + "learning_rate": 0.003, + "loss": 4.049, + "step": 24485 + }, + { + "epoch": 0.24486, + "grad_norm": 0.9096812253772166, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 24486 + }, + { + "epoch": 0.24487, + "grad_norm": 1.0462759547212908, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 24487 + }, + { + "epoch": 0.24488, + "grad_norm": 1.2297644033193498, + "learning_rate": 0.003, + "loss": 4.089, + "step": 24488 + }, + { + "epoch": 0.24489, + "grad_norm": 1.0309210201818442, + "learning_rate": 0.003, + "loss": 4.079, + "step": 24489 + }, + { + "epoch": 0.2449, + "grad_norm": 1.107839565666321, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 24490 + }, + { + "epoch": 0.24491, + "grad_norm": 1.2688508067316655, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 24491 + }, + { + "epoch": 0.24492, + "grad_norm": 0.8195801702189403, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 24492 + }, + { + "epoch": 0.24493, + "grad_norm": 0.8892097882372949, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 24493 + }, + { + "epoch": 0.24494, + "grad_norm": 1.0847117119377716, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 24494 + }, + { + "epoch": 0.24495, + "grad_norm": 0.9876723689645918, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 24495 + }, + { + "epoch": 0.24496, + "grad_norm": 1.2442670988212836, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 24496 + }, + { + "epoch": 0.24497, + "grad_norm": 1.025294873574308, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 24497 + }, + { + "epoch": 0.24498, + "grad_norm": 1.292705845264272, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 24498 + }, + { + "epoch": 0.24499, + "grad_norm": 0.9802374483993138, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 24499 + }, + { + "epoch": 0.245, + "grad_norm": 0.9776039096955471, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 24500 + }, + { + "epoch": 0.24501, + "grad_norm": 1.1591987222955755, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 24501 + }, + { + "epoch": 0.24502, + "grad_norm": 1.0230609647415772, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 24502 + }, + { + "epoch": 0.24503, + "grad_norm": 0.9685661983926828, + "learning_rate": 0.003, + "loss": 4.035, + "step": 24503 + }, + { + "epoch": 0.24504, + "grad_norm": 0.9950429169356849, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 24504 + }, + { + "epoch": 0.24505, + "grad_norm": 1.020209180047942, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 24505 + }, + { + "epoch": 0.24506, + "grad_norm": 1.0858922906014479, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 24506 + }, + { + "epoch": 0.24507, + "grad_norm": 1.0127047559911593, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 24507 + }, + { + "epoch": 0.24508, + "grad_norm": 0.9625933294670748, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 24508 + }, + { + "epoch": 0.24509, + "grad_norm": 1.0957118928935998, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 24509 + }, + { + "epoch": 0.2451, + "grad_norm": 0.957476730529024, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 24510 + }, + { + "epoch": 0.24511, + "grad_norm": 0.9229956501339138, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 24511 + }, + { + "epoch": 0.24512, + "grad_norm": 1.1673810095114048, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 24512 + }, + { + "epoch": 0.24513, + "grad_norm": 1.1804067174352204, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 24513 + }, + { + "epoch": 0.24514, + "grad_norm": 1.0852277592163264, + "learning_rate": 0.003, + "loss": 4.0923, + "step": 24514 + }, + { + "epoch": 0.24515, + "grad_norm": 1.1031380012765153, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 24515 + }, + { + "epoch": 0.24516, + "grad_norm": 1.1129747518759212, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 24516 + }, + { + "epoch": 0.24517, + "grad_norm": 1.2422181758099604, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 24517 + }, + { + "epoch": 0.24518, + "grad_norm": 1.1737298469010269, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 24518 + }, + { + "epoch": 0.24519, + "grad_norm": 1.0971201567271232, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 24519 + }, + { + "epoch": 0.2452, + "grad_norm": 1.0424543618965005, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 24520 + }, + { + "epoch": 0.24521, + "grad_norm": 1.1941279377256537, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 24521 + }, + { + "epoch": 0.24522, + "grad_norm": 0.929898161975911, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 24522 + }, + { + "epoch": 0.24523, + "grad_norm": 1.0341633265874755, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 24523 + }, + { + "epoch": 0.24524, + "grad_norm": 1.051720208817227, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 24524 + }, + { + "epoch": 0.24525, + "grad_norm": 1.1018619859443741, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 24525 + }, + { + "epoch": 0.24526, + "grad_norm": 1.2761655710546569, + "learning_rate": 0.003, + "loss": 4.043, + "step": 24526 + }, + { + "epoch": 0.24527, + "grad_norm": 0.9814363455838192, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 24527 + }, + { + "epoch": 0.24528, + "grad_norm": 1.0662633163473234, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 24528 + }, + { + "epoch": 0.24529, + "grad_norm": 1.0051029859563558, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 24529 + }, + { + "epoch": 0.2453, + "grad_norm": 1.2041560842640628, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 24530 + }, + { + "epoch": 0.24531, + "grad_norm": 0.8204586043329545, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 24531 + }, + { + "epoch": 0.24532, + "grad_norm": 0.8176921530178726, + "learning_rate": 0.003, + "loss": 4.05, + "step": 24532 + }, + { + "epoch": 0.24533, + "grad_norm": 0.8789597515684178, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 24533 + }, + { + "epoch": 0.24534, + "grad_norm": 1.0683707200303285, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 24534 + }, + { + "epoch": 0.24535, + "grad_norm": 1.105660637482608, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 24535 + }, + { + "epoch": 0.24536, + "grad_norm": 1.0523905370346969, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 24536 + }, + { + "epoch": 0.24537, + "grad_norm": 1.1498541073507964, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 24537 + }, + { + "epoch": 0.24538, + "grad_norm": 0.9115404183457538, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 24538 + }, + { + "epoch": 0.24539, + "grad_norm": 1.1203994720964912, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 24539 + }, + { + "epoch": 0.2454, + "grad_norm": 1.080937355786388, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 24540 + }, + { + "epoch": 0.24541, + "grad_norm": 1.0190835578169006, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 24541 + }, + { + "epoch": 0.24542, + "grad_norm": 1.066145739471434, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 24542 + }, + { + "epoch": 0.24543, + "grad_norm": 1.1703553782425595, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 24543 + }, + { + "epoch": 0.24544, + "grad_norm": 0.9996081757987486, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 24544 + }, + { + "epoch": 0.24545, + "grad_norm": 1.2522037647345854, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 24545 + }, + { + "epoch": 0.24546, + "grad_norm": 0.8925015850156302, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 24546 + }, + { + "epoch": 0.24547, + "grad_norm": 1.1163501777126204, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 24547 + }, + { + "epoch": 0.24548, + "grad_norm": 1.2296744356557265, + "learning_rate": 0.003, + "loss": 4.066, + "step": 24548 + }, + { + "epoch": 0.24549, + "grad_norm": 1.0941373963633223, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 24549 + }, + { + "epoch": 0.2455, + "grad_norm": 1.2600300331684005, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 24550 + }, + { + "epoch": 0.24551, + "grad_norm": 0.974344247301205, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 24551 + }, + { + "epoch": 0.24552, + "grad_norm": 1.0821670433854287, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 24552 + }, + { + "epoch": 0.24553, + "grad_norm": 1.2339438152271822, + "learning_rate": 0.003, + "loss": 4.068, + "step": 24553 + }, + { + "epoch": 0.24554, + "grad_norm": 1.0532354875367689, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 24554 + }, + { + "epoch": 0.24555, + "grad_norm": 1.1385891564029733, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 24555 + }, + { + "epoch": 0.24556, + "grad_norm": 0.7574755745545025, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 24556 + }, + { + "epoch": 0.24557, + "grad_norm": 0.7922976347052285, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 24557 + }, + { + "epoch": 0.24558, + "grad_norm": 0.9797282165010098, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 24558 + }, + { + "epoch": 0.24559, + "grad_norm": 1.0782823922208948, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 24559 + }, + { + "epoch": 0.2456, + "grad_norm": 1.0678072467300708, + "learning_rate": 0.003, + "loss": 4.0915, + "step": 24560 + }, + { + "epoch": 0.24561, + "grad_norm": 1.0512388409887676, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 24561 + }, + { + "epoch": 0.24562, + "grad_norm": 0.9441438617576656, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 24562 + }, + { + "epoch": 0.24563, + "grad_norm": 1.2572379365836577, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 24563 + }, + { + "epoch": 0.24564, + "grad_norm": 0.9825370286769212, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 24564 + }, + { + "epoch": 0.24565, + "grad_norm": 1.195483735800537, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 24565 + }, + { + "epoch": 0.24566, + "grad_norm": 0.7708110116228793, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 24566 + }, + { + "epoch": 0.24567, + "grad_norm": 0.8083896515884819, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 24567 + }, + { + "epoch": 0.24568, + "grad_norm": 0.9349205176463263, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 24568 + }, + { + "epoch": 0.24569, + "grad_norm": 1.395219696038983, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 24569 + }, + { + "epoch": 0.2457, + "grad_norm": 0.866807041560469, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 24570 + }, + { + "epoch": 0.24571, + "grad_norm": 1.012772305505143, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 24571 + }, + { + "epoch": 0.24572, + "grad_norm": 1.1976076979688128, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 24572 + }, + { + "epoch": 0.24573, + "grad_norm": 0.9208882708598122, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 24573 + }, + { + "epoch": 0.24574, + "grad_norm": 0.9268744293584502, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 24574 + }, + { + "epoch": 0.24575, + "grad_norm": 1.0306100643609395, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 24575 + }, + { + "epoch": 0.24576, + "grad_norm": 0.9755637336171271, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 24576 + }, + { + "epoch": 0.24577, + "grad_norm": 0.9795431038269763, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 24577 + }, + { + "epoch": 0.24578, + "grad_norm": 1.0586831095229023, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 24578 + }, + { + "epoch": 0.24579, + "grad_norm": 0.9880484491467221, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 24579 + }, + { + "epoch": 0.2458, + "grad_norm": 1.4104441594332504, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 24580 + }, + { + "epoch": 0.24581, + "grad_norm": 0.7477646703305221, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 24581 + }, + { + "epoch": 0.24582, + "grad_norm": 0.9085867365045369, + "learning_rate": 0.003, + "loss": 4.1019, + "step": 24582 + }, + { + "epoch": 0.24583, + "grad_norm": 0.9420063782933656, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 24583 + }, + { + "epoch": 0.24584, + "grad_norm": 0.9459340574423004, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 24584 + }, + { + "epoch": 0.24585, + "grad_norm": 1.2894771011733297, + "learning_rate": 0.003, + "loss": 4.1015, + "step": 24585 + }, + { + "epoch": 0.24586, + "grad_norm": 1.029870789553319, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 24586 + }, + { + "epoch": 0.24587, + "grad_norm": 1.1097054237890456, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 24587 + }, + { + "epoch": 0.24588, + "grad_norm": 0.9542712068651599, + "learning_rate": 0.003, + "loss": 4.042, + "step": 24588 + }, + { + "epoch": 0.24589, + "grad_norm": 1.268335245481498, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 24589 + }, + { + "epoch": 0.2459, + "grad_norm": 1.1724322031276377, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 24590 + }, + { + "epoch": 0.24591, + "grad_norm": 1.3480297868968887, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 24591 + }, + { + "epoch": 0.24592, + "grad_norm": 0.9839394111024741, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 24592 + }, + { + "epoch": 0.24593, + "grad_norm": 1.2189058486885178, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 24593 + }, + { + "epoch": 0.24594, + "grad_norm": 1.287416752857632, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 24594 + }, + { + "epoch": 0.24595, + "grad_norm": 1.0331020451080666, + "learning_rate": 0.003, + "loss": 4.048, + "step": 24595 + }, + { + "epoch": 0.24596, + "grad_norm": 1.1880565912802026, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 24596 + }, + { + "epoch": 0.24597, + "grad_norm": 0.9171880247491911, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 24597 + }, + { + "epoch": 0.24598, + "grad_norm": 1.0888074325230517, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 24598 + }, + { + "epoch": 0.24599, + "grad_norm": 1.0865663766077078, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 24599 + }, + { + "epoch": 0.246, + "grad_norm": 1.2678963364412588, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 24600 + }, + { + "epoch": 0.24601, + "grad_norm": 0.8679679711309632, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 24601 + }, + { + "epoch": 0.24602, + "grad_norm": 0.90586873423968, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 24602 + }, + { + "epoch": 0.24603, + "grad_norm": 1.1579345067493658, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 24603 + }, + { + "epoch": 0.24604, + "grad_norm": 1.023985909974375, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 24604 + }, + { + "epoch": 0.24605, + "grad_norm": 1.068631236495759, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 24605 + }, + { + "epoch": 0.24606, + "grad_norm": 0.9777394599771564, + "learning_rate": 0.003, + "loss": 4.033, + "step": 24606 + }, + { + "epoch": 0.24607, + "grad_norm": 1.0126370979656854, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 24607 + }, + { + "epoch": 0.24608, + "grad_norm": 1.3760268515264735, + "learning_rate": 0.003, + "loss": 4.072, + "step": 24608 + }, + { + "epoch": 0.24609, + "grad_norm": 1.1129526058966732, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 24609 + }, + { + "epoch": 0.2461, + "grad_norm": 1.0122081261594276, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 24610 + }, + { + "epoch": 0.24611, + "grad_norm": 1.2736676400850426, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 24611 + }, + { + "epoch": 0.24612, + "grad_norm": 0.9799148507842179, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 24612 + }, + { + "epoch": 0.24613, + "grad_norm": 1.0582295365959657, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 24613 + }, + { + "epoch": 0.24614, + "grad_norm": 1.1148315759545722, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 24614 + }, + { + "epoch": 0.24615, + "grad_norm": 0.9651090289392745, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 24615 + }, + { + "epoch": 0.24616, + "grad_norm": 1.0844737440753454, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 24616 + }, + { + "epoch": 0.24617, + "grad_norm": 0.9068386466315139, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 24617 + }, + { + "epoch": 0.24618, + "grad_norm": 1.0352781870941024, + "learning_rate": 0.003, + "loss": 4.059, + "step": 24618 + }, + { + "epoch": 0.24619, + "grad_norm": 1.2317420249113433, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 24619 + }, + { + "epoch": 0.2462, + "grad_norm": 0.9816508286008248, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 24620 + }, + { + "epoch": 0.24621, + "grad_norm": 1.081753666011136, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 24621 + }, + { + "epoch": 0.24622, + "grad_norm": 1.1889033401043563, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 24622 + }, + { + "epoch": 0.24623, + "grad_norm": 1.1164112910275001, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 24623 + }, + { + "epoch": 0.24624, + "grad_norm": 1.1315287692995322, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 24624 + }, + { + "epoch": 0.24625, + "grad_norm": 0.9895492204887854, + "learning_rate": 0.003, + "loss": 4.057, + "step": 24625 + }, + { + "epoch": 0.24626, + "grad_norm": 1.0121504243959005, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 24626 + }, + { + "epoch": 0.24627, + "grad_norm": 1.1046131117842612, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 24627 + }, + { + "epoch": 0.24628, + "grad_norm": 0.9735874902535265, + "learning_rate": 0.003, + "loss": 4.059, + "step": 24628 + }, + { + "epoch": 0.24629, + "grad_norm": 1.0674178064612259, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 24629 + }, + { + "epoch": 0.2463, + "grad_norm": 0.9545794998547547, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 24630 + }, + { + "epoch": 0.24631, + "grad_norm": 1.1290941625331625, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 24631 + }, + { + "epoch": 0.24632, + "grad_norm": 1.0739010674877743, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 24632 + }, + { + "epoch": 0.24633, + "grad_norm": 0.9372640461189755, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 24633 + }, + { + "epoch": 0.24634, + "grad_norm": 1.275738946897645, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 24634 + }, + { + "epoch": 0.24635, + "grad_norm": 1.1819955848041108, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 24635 + }, + { + "epoch": 0.24636, + "grad_norm": 1.1251933181499558, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 24636 + }, + { + "epoch": 0.24637, + "grad_norm": 0.9882850740637155, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 24637 + }, + { + "epoch": 0.24638, + "grad_norm": 1.2461754811760133, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 24638 + }, + { + "epoch": 0.24639, + "grad_norm": 0.9419937725258088, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 24639 + }, + { + "epoch": 0.2464, + "grad_norm": 1.1242112877292134, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 24640 + }, + { + "epoch": 0.24641, + "grad_norm": 1.102949354893846, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 24641 + }, + { + "epoch": 0.24642, + "grad_norm": 1.0428665506861903, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 24642 + }, + { + "epoch": 0.24643, + "grad_norm": 1.2199657751560864, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 24643 + }, + { + "epoch": 0.24644, + "grad_norm": 0.9457519197874904, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 24644 + }, + { + "epoch": 0.24645, + "grad_norm": 0.8270665325796519, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 24645 + }, + { + "epoch": 0.24646, + "grad_norm": 0.8787015968346407, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 24646 + }, + { + "epoch": 0.24647, + "grad_norm": 1.0652943970165598, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 24647 + }, + { + "epoch": 0.24648, + "grad_norm": 0.9859673635189228, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 24648 + }, + { + "epoch": 0.24649, + "grad_norm": 0.9626445335979179, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 24649 + }, + { + "epoch": 0.2465, + "grad_norm": 1.0309394944637005, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 24650 + }, + { + "epoch": 0.24651, + "grad_norm": 1.1913254865539624, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 24651 + }, + { + "epoch": 0.24652, + "grad_norm": 1.0886319793858816, + "learning_rate": 0.003, + "loss": 4.035, + "step": 24652 + }, + { + "epoch": 0.24653, + "grad_norm": 1.122188542611205, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 24653 + }, + { + "epoch": 0.24654, + "grad_norm": 0.9680917494258479, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 24654 + }, + { + "epoch": 0.24655, + "grad_norm": 0.9063213324171143, + "learning_rate": 0.003, + "loss": 4.0938, + "step": 24655 + }, + { + "epoch": 0.24656, + "grad_norm": 1.138624711641011, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 24656 + }, + { + "epoch": 0.24657, + "grad_norm": 1.2850059026710892, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 24657 + }, + { + "epoch": 0.24658, + "grad_norm": 1.0449640588400524, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 24658 + }, + { + "epoch": 0.24659, + "grad_norm": 1.0829111318076052, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 24659 + }, + { + "epoch": 0.2466, + "grad_norm": 1.1924375214540703, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 24660 + }, + { + "epoch": 0.24661, + "grad_norm": 0.9631581156702572, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 24661 + }, + { + "epoch": 0.24662, + "grad_norm": 0.9628292792211719, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 24662 + }, + { + "epoch": 0.24663, + "grad_norm": 1.0969427774628147, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 24663 + }, + { + "epoch": 0.24664, + "grad_norm": 0.9142557235714011, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 24664 + }, + { + "epoch": 0.24665, + "grad_norm": 1.1589246408504001, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 24665 + }, + { + "epoch": 0.24666, + "grad_norm": 0.9601551485986024, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 24666 + }, + { + "epoch": 0.24667, + "grad_norm": 1.2696245306977803, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 24667 + }, + { + "epoch": 0.24668, + "grad_norm": 1.0861904661713524, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 24668 + }, + { + "epoch": 0.24669, + "grad_norm": 0.9606563221282256, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 24669 + }, + { + "epoch": 0.2467, + "grad_norm": 0.903189706998714, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 24670 + }, + { + "epoch": 0.24671, + "grad_norm": 0.9524934969920437, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 24671 + }, + { + "epoch": 0.24672, + "grad_norm": 1.2346257567859866, + "learning_rate": 0.003, + "loss": 4.0967, + "step": 24672 + }, + { + "epoch": 0.24673, + "grad_norm": 0.8618374425524838, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 24673 + }, + { + "epoch": 0.24674, + "grad_norm": 1.0380771496986911, + "learning_rate": 0.003, + "loss": 4.057, + "step": 24674 + }, + { + "epoch": 0.24675, + "grad_norm": 1.3201672256490344, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 24675 + }, + { + "epoch": 0.24676, + "grad_norm": 1.1599049691212444, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 24676 + }, + { + "epoch": 0.24677, + "grad_norm": 1.0644257042390872, + "learning_rate": 0.003, + "loss": 4.071, + "step": 24677 + }, + { + "epoch": 0.24678, + "grad_norm": 1.0011729397352491, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 24678 + }, + { + "epoch": 0.24679, + "grad_norm": 1.3184625732011106, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 24679 + }, + { + "epoch": 0.2468, + "grad_norm": 0.972753823663196, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 24680 + }, + { + "epoch": 0.24681, + "grad_norm": 1.0631977085989497, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 24681 + }, + { + "epoch": 0.24682, + "grad_norm": 0.9090813311022216, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 24682 + }, + { + "epoch": 0.24683, + "grad_norm": 0.9261940604648667, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 24683 + }, + { + "epoch": 0.24684, + "grad_norm": 1.0062404006393881, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 24684 + }, + { + "epoch": 0.24685, + "grad_norm": 1.2894968645887905, + "learning_rate": 0.003, + "loss": 4.03, + "step": 24685 + }, + { + "epoch": 0.24686, + "grad_norm": 0.8513812817091038, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 24686 + }, + { + "epoch": 0.24687, + "grad_norm": 0.9750961881587251, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 24687 + }, + { + "epoch": 0.24688, + "grad_norm": 1.1562796673053788, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 24688 + }, + { + "epoch": 0.24689, + "grad_norm": 0.938513280115268, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 24689 + }, + { + "epoch": 0.2469, + "grad_norm": 1.1004722248174656, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 24690 + }, + { + "epoch": 0.24691, + "grad_norm": 1.325707361941631, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 24691 + }, + { + "epoch": 0.24692, + "grad_norm": 1.0744256167843902, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 24692 + }, + { + "epoch": 0.24693, + "grad_norm": 1.0044853853882822, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 24693 + }, + { + "epoch": 0.24694, + "grad_norm": 1.2064458424084925, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 24694 + }, + { + "epoch": 0.24695, + "grad_norm": 0.8495179185231407, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 24695 + }, + { + "epoch": 0.24696, + "grad_norm": 0.8067193156650647, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 24696 + }, + { + "epoch": 0.24697, + "grad_norm": 0.927529083655388, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 24697 + }, + { + "epoch": 0.24698, + "grad_norm": 1.0853247158307318, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 24698 + }, + { + "epoch": 0.24699, + "grad_norm": 0.9906051210618717, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 24699 + }, + { + "epoch": 0.247, + "grad_norm": 1.0094961141484666, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 24700 + }, + { + "epoch": 0.24701, + "grad_norm": 1.2714840335973399, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 24701 + }, + { + "epoch": 0.24702, + "grad_norm": 1.2181138171170283, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 24702 + }, + { + "epoch": 0.24703, + "grad_norm": 0.9552688499721164, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 24703 + }, + { + "epoch": 0.24704, + "grad_norm": 0.8840797884971562, + "learning_rate": 0.003, + "loss": 4.065, + "step": 24704 + }, + { + "epoch": 0.24705, + "grad_norm": 1.011438837918084, + "learning_rate": 0.003, + "loss": 4.065, + "step": 24705 + }, + { + "epoch": 0.24706, + "grad_norm": 1.2336886069044553, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 24706 + }, + { + "epoch": 0.24707, + "grad_norm": 0.9917943307363697, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 24707 + }, + { + "epoch": 0.24708, + "grad_norm": 1.1769915922368026, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 24708 + }, + { + "epoch": 0.24709, + "grad_norm": 0.8437144126876495, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 24709 + }, + { + "epoch": 0.2471, + "grad_norm": 1.1087054927557878, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 24710 + }, + { + "epoch": 0.24711, + "grad_norm": 1.2040942913039943, + "learning_rate": 0.003, + "loss": 4.05, + "step": 24711 + }, + { + "epoch": 0.24712, + "grad_norm": 0.9900888574243972, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 24712 + }, + { + "epoch": 0.24713, + "grad_norm": 1.2301278090727008, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 24713 + }, + { + "epoch": 0.24714, + "grad_norm": 1.0170243437345805, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 24714 + }, + { + "epoch": 0.24715, + "grad_norm": 1.1922051525733572, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 24715 + }, + { + "epoch": 0.24716, + "grad_norm": 1.1690875834725394, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 24716 + }, + { + "epoch": 0.24717, + "grad_norm": 1.1830840801787743, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 24717 + }, + { + "epoch": 0.24718, + "grad_norm": 1.0736680669791836, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 24718 + }, + { + "epoch": 0.24719, + "grad_norm": 1.226598371320371, + "learning_rate": 0.003, + "loss": 4.049, + "step": 24719 + }, + { + "epoch": 0.2472, + "grad_norm": 1.0353988543270114, + "learning_rate": 0.003, + "loss": 4.048, + "step": 24720 + }, + { + "epoch": 0.24721, + "grad_norm": 1.1374863817569243, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 24721 + }, + { + "epoch": 0.24722, + "grad_norm": 1.0387157044555735, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 24722 + }, + { + "epoch": 0.24723, + "grad_norm": 1.0748797394695377, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 24723 + }, + { + "epoch": 0.24724, + "grad_norm": 0.941054651802274, + "learning_rate": 0.003, + "loss": 4.064, + "step": 24724 + }, + { + "epoch": 0.24725, + "grad_norm": 0.9818369073644678, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 24725 + }, + { + "epoch": 0.24726, + "grad_norm": 1.0802605258962947, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 24726 + }, + { + "epoch": 0.24727, + "grad_norm": 1.0405045735711804, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 24727 + }, + { + "epoch": 0.24728, + "grad_norm": 1.0152697955935541, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 24728 + }, + { + "epoch": 0.24729, + "grad_norm": 1.179633303695429, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 24729 + }, + { + "epoch": 0.2473, + "grad_norm": 1.0384860231006403, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 24730 + }, + { + "epoch": 0.24731, + "grad_norm": 0.9705909795051036, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 24731 + }, + { + "epoch": 0.24732, + "grad_norm": 1.1242225535889303, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 24732 + }, + { + "epoch": 0.24733, + "grad_norm": 1.186225636820815, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 24733 + }, + { + "epoch": 0.24734, + "grad_norm": 1.1198902641712936, + "learning_rate": 0.003, + "loss": 4.018, + "step": 24734 + }, + { + "epoch": 0.24735, + "grad_norm": 1.1747043568493887, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 24735 + }, + { + "epoch": 0.24736, + "grad_norm": 1.132694515750506, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 24736 + }, + { + "epoch": 0.24737, + "grad_norm": 0.9525065047406372, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 24737 + }, + { + "epoch": 0.24738, + "grad_norm": 1.1500165123220936, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 24738 + }, + { + "epoch": 0.24739, + "grad_norm": 0.8732708054891144, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 24739 + }, + { + "epoch": 0.2474, + "grad_norm": 0.9375918960374737, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 24740 + }, + { + "epoch": 0.24741, + "grad_norm": 1.140689485628179, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 24741 + }, + { + "epoch": 0.24742, + "grad_norm": 1.0518270192333863, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 24742 + }, + { + "epoch": 0.24743, + "grad_norm": 0.9900462726590269, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 24743 + }, + { + "epoch": 0.24744, + "grad_norm": 1.2864058202133488, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 24744 + }, + { + "epoch": 0.24745, + "grad_norm": 0.9504085115847941, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 24745 + }, + { + "epoch": 0.24746, + "grad_norm": 1.0160944353505394, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 24746 + }, + { + "epoch": 0.24747, + "grad_norm": 1.4239290333808914, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 24747 + }, + { + "epoch": 0.24748, + "grad_norm": 0.7884948217300976, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 24748 + }, + { + "epoch": 0.24749, + "grad_norm": 0.7445876622538873, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 24749 + }, + { + "epoch": 0.2475, + "grad_norm": 0.8743593127485874, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 24750 + }, + { + "epoch": 0.24751, + "grad_norm": 0.9942577860146912, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 24751 + }, + { + "epoch": 0.24752, + "grad_norm": 1.086846215409882, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 24752 + }, + { + "epoch": 0.24753, + "grad_norm": 1.1907725338213866, + "learning_rate": 0.003, + "loss": 4.054, + "step": 24753 + }, + { + "epoch": 0.24754, + "grad_norm": 1.0688038139142573, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 24754 + }, + { + "epoch": 0.24755, + "grad_norm": 1.1142046514186008, + "learning_rate": 0.003, + "loss": 4.011, + "step": 24755 + }, + { + "epoch": 0.24756, + "grad_norm": 1.0866499331033443, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 24756 + }, + { + "epoch": 0.24757, + "grad_norm": 1.0944474695824318, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 24757 + }, + { + "epoch": 0.24758, + "grad_norm": 1.109911042119174, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 24758 + }, + { + "epoch": 0.24759, + "grad_norm": 1.0961452217204894, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 24759 + }, + { + "epoch": 0.2476, + "grad_norm": 1.0037060140158256, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 24760 + }, + { + "epoch": 0.24761, + "grad_norm": 1.1421555019850844, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 24761 + }, + { + "epoch": 0.24762, + "grad_norm": 1.0725437592937563, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 24762 + }, + { + "epoch": 0.24763, + "grad_norm": 1.1117354805127408, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 24763 + }, + { + "epoch": 0.24764, + "grad_norm": 1.0137812034449336, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 24764 + }, + { + "epoch": 0.24765, + "grad_norm": 1.2148714246280825, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 24765 + }, + { + "epoch": 0.24766, + "grad_norm": 1.097720804377704, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 24766 + }, + { + "epoch": 0.24767, + "grad_norm": 1.1905623233411347, + "learning_rate": 0.003, + "loss": 4.0913, + "step": 24767 + }, + { + "epoch": 0.24768, + "grad_norm": 1.2324221018751076, + "learning_rate": 0.003, + "loss": 4.058, + "step": 24768 + }, + { + "epoch": 0.24769, + "grad_norm": 0.8177297220107068, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 24769 + }, + { + "epoch": 0.2477, + "grad_norm": 0.8313083258670777, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 24770 + }, + { + "epoch": 0.24771, + "grad_norm": 1.0098552101732607, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 24771 + }, + { + "epoch": 0.24772, + "grad_norm": 1.2519711811965568, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 24772 + }, + { + "epoch": 0.24773, + "grad_norm": 0.9481976685918833, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 24773 + }, + { + "epoch": 0.24774, + "grad_norm": 1.2200377281779555, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 24774 + }, + { + "epoch": 0.24775, + "grad_norm": 1.075428310762756, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 24775 + }, + { + "epoch": 0.24776, + "grad_norm": 1.233502503287914, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 24776 + }, + { + "epoch": 0.24777, + "grad_norm": 1.0660513017357165, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 24777 + }, + { + "epoch": 0.24778, + "grad_norm": 1.2896667377914846, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 24778 + }, + { + "epoch": 0.24779, + "grad_norm": 0.9609389660697545, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 24779 + }, + { + "epoch": 0.2478, + "grad_norm": 0.9253553974541081, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 24780 + }, + { + "epoch": 0.24781, + "grad_norm": 0.8559734630112821, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 24781 + }, + { + "epoch": 0.24782, + "grad_norm": 0.9645879116043375, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 24782 + }, + { + "epoch": 0.24783, + "grad_norm": 1.2882345281969723, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 24783 + }, + { + "epoch": 0.24784, + "grad_norm": 0.876162436139419, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 24784 + }, + { + "epoch": 0.24785, + "grad_norm": 0.8665955674973349, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 24785 + }, + { + "epoch": 0.24786, + "grad_norm": 0.9773249631153904, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 24786 + }, + { + "epoch": 0.24787, + "grad_norm": 1.2580531334760643, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 24787 + }, + { + "epoch": 0.24788, + "grad_norm": 0.9888973190496023, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 24788 + }, + { + "epoch": 0.24789, + "grad_norm": 0.9363933849822386, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 24789 + }, + { + "epoch": 0.2479, + "grad_norm": 0.9424170305935237, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 24790 + }, + { + "epoch": 0.24791, + "grad_norm": 1.2681468809368612, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 24791 + }, + { + "epoch": 0.24792, + "grad_norm": 1.044080664279814, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 24792 + }, + { + "epoch": 0.24793, + "grad_norm": 1.3018984754486889, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 24793 + }, + { + "epoch": 0.24794, + "grad_norm": 0.9843436726715499, + "learning_rate": 0.003, + "loss": 4.054, + "step": 24794 + }, + { + "epoch": 0.24795, + "grad_norm": 1.303557357562636, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 24795 + }, + { + "epoch": 0.24796, + "grad_norm": 1.1760974020112973, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 24796 + }, + { + "epoch": 0.24797, + "grad_norm": 1.107843136801216, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 24797 + }, + { + "epoch": 0.24798, + "grad_norm": 0.9681455179893603, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 24798 + }, + { + "epoch": 0.24799, + "grad_norm": 1.0083723162932803, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 24799 + }, + { + "epoch": 0.248, + "grad_norm": 1.0500739679173716, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 24800 + }, + { + "epoch": 0.24801, + "grad_norm": 1.12816148739073, + "learning_rate": 0.003, + "loss": 4.085, + "step": 24801 + }, + { + "epoch": 0.24802, + "grad_norm": 0.895276505463727, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 24802 + }, + { + "epoch": 0.24803, + "grad_norm": 1.0580229603811733, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 24803 + }, + { + "epoch": 0.24804, + "grad_norm": 1.265807100669761, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 24804 + }, + { + "epoch": 0.24805, + "grad_norm": 1.1276407517159412, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 24805 + }, + { + "epoch": 0.24806, + "grad_norm": 1.0822040517507203, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 24806 + }, + { + "epoch": 0.24807, + "grad_norm": 1.1022435809202336, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 24807 + }, + { + "epoch": 0.24808, + "grad_norm": 1.0323619064382232, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 24808 + }, + { + "epoch": 0.24809, + "grad_norm": 1.0889079993310413, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 24809 + }, + { + "epoch": 0.2481, + "grad_norm": 1.0204812711357896, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 24810 + }, + { + "epoch": 0.24811, + "grad_norm": 1.1044017807923157, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 24811 + }, + { + "epoch": 0.24812, + "grad_norm": 1.227664509275974, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 24812 + }, + { + "epoch": 0.24813, + "grad_norm": 0.9557321887232326, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 24813 + }, + { + "epoch": 0.24814, + "grad_norm": 0.9993117921031706, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 24814 + }, + { + "epoch": 0.24815, + "grad_norm": 1.1387375652287113, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 24815 + }, + { + "epoch": 0.24816, + "grad_norm": 1.0520042838267838, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 24816 + }, + { + "epoch": 0.24817, + "grad_norm": 1.1036540214941648, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 24817 + }, + { + "epoch": 0.24818, + "grad_norm": 1.0834836010244604, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 24818 + }, + { + "epoch": 0.24819, + "grad_norm": 0.8838445538796877, + "learning_rate": 0.003, + "loss": 4.027, + "step": 24819 + }, + { + "epoch": 0.2482, + "grad_norm": 0.8552523576404096, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 24820 + }, + { + "epoch": 0.24821, + "grad_norm": 0.9679878224310879, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 24821 + }, + { + "epoch": 0.24822, + "grad_norm": 1.2813256910534838, + "learning_rate": 0.003, + "loss": 4.063, + "step": 24822 + }, + { + "epoch": 0.24823, + "grad_norm": 1.029936074222357, + "learning_rate": 0.003, + "loss": 4.058, + "step": 24823 + }, + { + "epoch": 0.24824, + "grad_norm": 1.2448401359596364, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 24824 + }, + { + "epoch": 0.24825, + "grad_norm": 1.0443579598850283, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 24825 + }, + { + "epoch": 0.24826, + "grad_norm": 1.217212209650394, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 24826 + }, + { + "epoch": 0.24827, + "grad_norm": 0.988159330205415, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 24827 + }, + { + "epoch": 0.24828, + "grad_norm": 1.14193594453651, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 24828 + }, + { + "epoch": 0.24829, + "grad_norm": 0.851278776280311, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 24829 + }, + { + "epoch": 0.2483, + "grad_norm": 0.992773376282867, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 24830 + }, + { + "epoch": 0.24831, + "grad_norm": 1.0593539806917875, + "learning_rate": 0.003, + "loss": 4.073, + "step": 24831 + }, + { + "epoch": 0.24832, + "grad_norm": 0.8832510473895722, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 24832 + }, + { + "epoch": 0.24833, + "grad_norm": 1.135776825879357, + "learning_rate": 0.003, + "loss": 4.049, + "step": 24833 + }, + { + "epoch": 0.24834, + "grad_norm": 1.1438600908271406, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 24834 + }, + { + "epoch": 0.24835, + "grad_norm": 1.084154702865052, + "learning_rate": 0.003, + "loss": 4.0997, + "step": 24835 + }, + { + "epoch": 0.24836, + "grad_norm": 1.294111204266152, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 24836 + }, + { + "epoch": 0.24837, + "grad_norm": 1.132577244689692, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 24837 + }, + { + "epoch": 0.24838, + "grad_norm": 1.1654449891515632, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 24838 + }, + { + "epoch": 0.24839, + "grad_norm": 0.9614132591527729, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 24839 + }, + { + "epoch": 0.2484, + "grad_norm": 1.0909915997557376, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 24840 + }, + { + "epoch": 0.24841, + "grad_norm": 1.2984085756713848, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 24841 + }, + { + "epoch": 0.24842, + "grad_norm": 0.8519083475815673, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 24842 + }, + { + "epoch": 0.24843, + "grad_norm": 0.82945793582198, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 24843 + }, + { + "epoch": 0.24844, + "grad_norm": 0.9264190682490512, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 24844 + }, + { + "epoch": 0.24845, + "grad_norm": 1.2213701382698796, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 24845 + }, + { + "epoch": 0.24846, + "grad_norm": 1.08920296435768, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 24846 + }, + { + "epoch": 0.24847, + "grad_norm": 0.963045260265732, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 24847 + }, + { + "epoch": 0.24848, + "grad_norm": 1.154563230200878, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 24848 + }, + { + "epoch": 0.24849, + "grad_norm": 1.1394012972900927, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 24849 + }, + { + "epoch": 0.2485, + "grad_norm": 0.8048756822312236, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 24850 + }, + { + "epoch": 0.24851, + "grad_norm": 0.8248866506612944, + "learning_rate": 0.003, + "loss": 4.035, + "step": 24851 + }, + { + "epoch": 0.24852, + "grad_norm": 1.0519998758650209, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 24852 + }, + { + "epoch": 0.24853, + "grad_norm": 1.2643392247089633, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 24853 + }, + { + "epoch": 0.24854, + "grad_norm": 0.9194679316050718, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 24854 + }, + { + "epoch": 0.24855, + "grad_norm": 0.9502520443490123, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 24855 + }, + { + "epoch": 0.24856, + "grad_norm": 0.9050536931782389, + "learning_rate": 0.003, + "loss": 4.063, + "step": 24856 + }, + { + "epoch": 0.24857, + "grad_norm": 0.9653759686036988, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 24857 + }, + { + "epoch": 0.24858, + "grad_norm": 1.2100431594448355, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 24858 + }, + { + "epoch": 0.24859, + "grad_norm": 0.7872995760049382, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 24859 + }, + { + "epoch": 0.2486, + "grad_norm": 0.8874522313548019, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 24860 + }, + { + "epoch": 0.24861, + "grad_norm": 1.0828749503870954, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 24861 + }, + { + "epoch": 0.24862, + "grad_norm": 1.088736868222223, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 24862 + }, + { + "epoch": 0.24863, + "grad_norm": 1.0820645136486238, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 24863 + }, + { + "epoch": 0.24864, + "grad_norm": 0.9136378095708828, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 24864 + }, + { + "epoch": 0.24865, + "grad_norm": 0.8819608011534515, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 24865 + }, + { + "epoch": 0.24866, + "grad_norm": 1.0347730067789773, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 24866 + }, + { + "epoch": 0.24867, + "grad_norm": 1.6241600546194297, + "learning_rate": 0.003, + "loss": 4.046, + "step": 24867 + }, + { + "epoch": 0.24868, + "grad_norm": 1.0203559032734775, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 24868 + }, + { + "epoch": 0.24869, + "grad_norm": 1.2091148983418556, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 24869 + }, + { + "epoch": 0.2487, + "grad_norm": 1.0873235656584144, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 24870 + }, + { + "epoch": 0.24871, + "grad_norm": 1.1656806820912908, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 24871 + }, + { + "epoch": 0.24872, + "grad_norm": 0.8474682328940225, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 24872 + }, + { + "epoch": 0.24873, + "grad_norm": 1.0472852850224623, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 24873 + }, + { + "epoch": 0.24874, + "grad_norm": 1.24217880896953, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 24874 + }, + { + "epoch": 0.24875, + "grad_norm": 0.696679793983281, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 24875 + }, + { + "epoch": 0.24876, + "grad_norm": 0.7407744194293339, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 24876 + }, + { + "epoch": 0.24877, + "grad_norm": 0.922528455322865, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 24877 + }, + { + "epoch": 0.24878, + "grad_norm": 1.1394367999742392, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 24878 + }, + { + "epoch": 0.24879, + "grad_norm": 1.2167072318518601, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 24879 + }, + { + "epoch": 0.2488, + "grad_norm": 1.0770058072072792, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 24880 + }, + { + "epoch": 0.24881, + "grad_norm": 0.9678445371664982, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 24881 + }, + { + "epoch": 0.24882, + "grad_norm": 0.9537172049363611, + "learning_rate": 0.003, + "loss": 4.029, + "step": 24882 + }, + { + "epoch": 0.24883, + "grad_norm": 1.1483118358727828, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 24883 + }, + { + "epoch": 0.24884, + "grad_norm": 1.0981589715736404, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 24884 + }, + { + "epoch": 0.24885, + "grad_norm": 1.0630276477199767, + "learning_rate": 0.003, + "loss": 4.072, + "step": 24885 + }, + { + "epoch": 0.24886, + "grad_norm": 0.9311137386994214, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 24886 + }, + { + "epoch": 0.24887, + "grad_norm": 1.115802083827793, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 24887 + }, + { + "epoch": 0.24888, + "grad_norm": 0.9606212237551618, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 24888 + }, + { + "epoch": 0.24889, + "grad_norm": 1.1748361266734033, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 24889 + }, + { + "epoch": 0.2489, + "grad_norm": 1.219687541816208, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 24890 + }, + { + "epoch": 0.24891, + "grad_norm": 1.2939899572025186, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 24891 + }, + { + "epoch": 0.24892, + "grad_norm": 0.9771818886735693, + "learning_rate": 0.003, + "loss": 4.0964, + "step": 24892 + }, + { + "epoch": 0.24893, + "grad_norm": 1.0613524268159757, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 24893 + }, + { + "epoch": 0.24894, + "grad_norm": 1.0700850904241723, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 24894 + }, + { + "epoch": 0.24895, + "grad_norm": 1.0718442320815917, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 24895 + }, + { + "epoch": 0.24896, + "grad_norm": 1.3316698730999508, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 24896 + }, + { + "epoch": 0.24897, + "grad_norm": 0.9961487129832411, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 24897 + }, + { + "epoch": 0.24898, + "grad_norm": 0.9811108253576529, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 24898 + }, + { + "epoch": 0.24899, + "grad_norm": 1.13364100435044, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 24899 + }, + { + "epoch": 0.249, + "grad_norm": 1.0401056970181863, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 24900 + }, + { + "epoch": 0.24901, + "grad_norm": 1.0957215128426971, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 24901 + }, + { + "epoch": 0.24902, + "grad_norm": 0.9647718448917143, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 24902 + }, + { + "epoch": 0.24903, + "grad_norm": 1.136089972085343, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 24903 + }, + { + "epoch": 0.24904, + "grad_norm": 1.1310112829019894, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 24904 + }, + { + "epoch": 0.24905, + "grad_norm": 0.9838031822190026, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 24905 + }, + { + "epoch": 0.24906, + "grad_norm": 1.3377100797335977, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 24906 + }, + { + "epoch": 0.24907, + "grad_norm": 0.931644323017362, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 24907 + }, + { + "epoch": 0.24908, + "grad_norm": 1.0282015981620234, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 24908 + }, + { + "epoch": 0.24909, + "grad_norm": 1.412134740344243, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 24909 + }, + { + "epoch": 0.2491, + "grad_norm": 0.9477505802332908, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 24910 + }, + { + "epoch": 0.24911, + "grad_norm": 1.1823878739392266, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 24911 + }, + { + "epoch": 0.24912, + "grad_norm": 1.0136295466036043, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 24912 + }, + { + "epoch": 0.24913, + "grad_norm": 1.3105573544397495, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 24913 + }, + { + "epoch": 0.24914, + "grad_norm": 0.8975524742662339, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 24914 + }, + { + "epoch": 0.24915, + "grad_norm": 0.9099622003257015, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 24915 + }, + { + "epoch": 0.24916, + "grad_norm": 1.1105908131099282, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 24916 + }, + { + "epoch": 0.24917, + "grad_norm": 1.023492804527498, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 24917 + }, + { + "epoch": 0.24918, + "grad_norm": 1.1014849602529209, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 24918 + }, + { + "epoch": 0.24919, + "grad_norm": 0.9990919715518439, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 24919 + }, + { + "epoch": 0.2492, + "grad_norm": 1.0663615229825434, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 24920 + }, + { + "epoch": 0.24921, + "grad_norm": 1.0528735236315585, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 24921 + }, + { + "epoch": 0.24922, + "grad_norm": 1.199816773733297, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 24922 + }, + { + "epoch": 0.24923, + "grad_norm": 1.1173343403670535, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 24923 + }, + { + "epoch": 0.24924, + "grad_norm": 0.9378391455818269, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 24924 + }, + { + "epoch": 0.24925, + "grad_norm": 1.1048501113335771, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 24925 + }, + { + "epoch": 0.24926, + "grad_norm": 1.0806457920588033, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 24926 + }, + { + "epoch": 0.24927, + "grad_norm": 0.9615726934577841, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 24927 + }, + { + "epoch": 0.24928, + "grad_norm": 1.0818224999195944, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 24928 + }, + { + "epoch": 0.24929, + "grad_norm": 0.9941380664769957, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 24929 + }, + { + "epoch": 0.2493, + "grad_norm": 1.0406468421013546, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 24930 + }, + { + "epoch": 0.24931, + "grad_norm": 1.3384930483683002, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 24931 + }, + { + "epoch": 0.24932, + "grad_norm": 0.8509412258123786, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 24932 + }, + { + "epoch": 0.24933, + "grad_norm": 0.9096628569191039, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 24933 + }, + { + "epoch": 0.24934, + "grad_norm": 0.9831080182663523, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 24934 + }, + { + "epoch": 0.24935, + "grad_norm": 1.1534805726592832, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 24935 + }, + { + "epoch": 0.24936, + "grad_norm": 1.136107275971308, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 24936 + }, + { + "epoch": 0.24937, + "grad_norm": 0.9091946647510657, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 24937 + }, + { + "epoch": 0.24938, + "grad_norm": 0.9497277350686605, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 24938 + }, + { + "epoch": 0.24939, + "grad_norm": 1.1816783263118078, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 24939 + }, + { + "epoch": 0.2494, + "grad_norm": 1.1248676641808746, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 24940 + }, + { + "epoch": 0.24941, + "grad_norm": 1.0250622294462157, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 24941 + }, + { + "epoch": 0.24942, + "grad_norm": 1.0603312906010682, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 24942 + }, + { + "epoch": 0.24943, + "grad_norm": 0.9749429252777092, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 24943 + }, + { + "epoch": 0.24944, + "grad_norm": 1.09788938701373, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 24944 + }, + { + "epoch": 0.24945, + "grad_norm": 1.0175519253372067, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 24945 + }, + { + "epoch": 0.24946, + "grad_norm": 1.0531255375750987, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 24946 + }, + { + "epoch": 0.24947, + "grad_norm": 1.0065123458158343, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 24947 + }, + { + "epoch": 0.24948, + "grad_norm": 1.0736484468839813, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 24948 + }, + { + "epoch": 0.24949, + "grad_norm": 1.079982010560508, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 24949 + }, + { + "epoch": 0.2495, + "grad_norm": 1.068973653856954, + "learning_rate": 0.003, + "loss": 4.08, + "step": 24950 + }, + { + "epoch": 0.24951, + "grad_norm": 1.2626388398961848, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 24951 + }, + { + "epoch": 0.24952, + "grad_norm": 0.99488278768349, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 24952 + }, + { + "epoch": 0.24953, + "grad_norm": 1.1000786473567383, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 24953 + }, + { + "epoch": 0.24954, + "grad_norm": 1.0503755856656511, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 24954 + }, + { + "epoch": 0.24955, + "grad_norm": 1.1891433986257949, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 24955 + }, + { + "epoch": 0.24956, + "grad_norm": 1.064796291610176, + "learning_rate": 0.003, + "loss": 4.069, + "step": 24956 + }, + { + "epoch": 0.24957, + "grad_norm": 1.1135632802588087, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 24957 + }, + { + "epoch": 0.24958, + "grad_norm": 1.1177639421583856, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 24958 + }, + { + "epoch": 0.24959, + "grad_norm": 1.1672542758958973, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 24959 + }, + { + "epoch": 0.2496, + "grad_norm": 1.369431001387121, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 24960 + }, + { + "epoch": 0.24961, + "grad_norm": 0.8716233981691348, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 24961 + }, + { + "epoch": 0.24962, + "grad_norm": 1.0053927398021192, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 24962 + }, + { + "epoch": 0.24963, + "grad_norm": 1.1213751455401229, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 24963 + }, + { + "epoch": 0.24964, + "grad_norm": 0.9045651317962339, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 24964 + }, + { + "epoch": 0.24965, + "grad_norm": 1.0104185411446192, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 24965 + }, + { + "epoch": 0.24966, + "grad_norm": 0.9911219554910186, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 24966 + }, + { + "epoch": 0.24967, + "grad_norm": 1.0690818796825414, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 24967 + }, + { + "epoch": 0.24968, + "grad_norm": 1.1233999917673538, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 24968 + }, + { + "epoch": 0.24969, + "grad_norm": 1.1735138404877197, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 24969 + }, + { + "epoch": 0.2497, + "grad_norm": 1.114482441730397, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 24970 + }, + { + "epoch": 0.24971, + "grad_norm": 1.025911053878739, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 24971 + }, + { + "epoch": 0.24972, + "grad_norm": 1.2334356116473928, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 24972 + }, + { + "epoch": 0.24973, + "grad_norm": 0.8905243739550014, + "learning_rate": 0.003, + "loss": 4.057, + "step": 24973 + }, + { + "epoch": 0.24974, + "grad_norm": 0.8582416844448876, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 24974 + }, + { + "epoch": 0.24975, + "grad_norm": 1.0596423715865078, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 24975 + }, + { + "epoch": 0.24976, + "grad_norm": 1.1550362664962825, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 24976 + }, + { + "epoch": 0.24977, + "grad_norm": 0.8951672268517176, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 24977 + }, + { + "epoch": 0.24978, + "grad_norm": 1.0923331349300307, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 24978 + }, + { + "epoch": 0.24979, + "grad_norm": 1.0872654253412226, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 24979 + }, + { + "epoch": 0.2498, + "grad_norm": 1.0266398804990775, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 24980 + }, + { + "epoch": 0.24981, + "grad_norm": 1.194988444407034, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 24981 + }, + { + "epoch": 0.24982, + "grad_norm": 1.074810382894597, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 24982 + }, + { + "epoch": 0.24983, + "grad_norm": 1.114221546653547, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 24983 + }, + { + "epoch": 0.24984, + "grad_norm": 0.9151409388554572, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 24984 + }, + { + "epoch": 0.24985, + "grad_norm": 0.9845599845434989, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 24985 + }, + { + "epoch": 0.24986, + "grad_norm": 1.3448383481057626, + "learning_rate": 0.003, + "loss": 4.051, + "step": 24986 + }, + { + "epoch": 0.24987, + "grad_norm": 0.8573299500734617, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 24987 + }, + { + "epoch": 0.24988, + "grad_norm": 0.9799795658784689, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 24988 + }, + { + "epoch": 0.24989, + "grad_norm": 1.1559525844128304, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 24989 + }, + { + "epoch": 0.2499, + "grad_norm": 1.001018260661963, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 24990 + }, + { + "epoch": 0.24991, + "grad_norm": 1.305471860678751, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 24991 + }, + { + "epoch": 0.24992, + "grad_norm": 0.9366648052872386, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 24992 + }, + { + "epoch": 0.24993, + "grad_norm": 1.0242900587748804, + "learning_rate": 0.003, + "loss": 4.062, + "step": 24993 + }, + { + "epoch": 0.24994, + "grad_norm": 1.081452235739146, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 24994 + }, + { + "epoch": 0.24995, + "grad_norm": 1.0184984136505242, + "learning_rate": 0.003, + "loss": 4.076, + "step": 24995 + }, + { + "epoch": 0.24996, + "grad_norm": 1.367278645251946, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 24996 + }, + { + "epoch": 0.24997, + "grad_norm": 1.1086780852922398, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 24997 + }, + { + "epoch": 0.24998, + "grad_norm": 1.0146278296463331, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 24998 + }, + { + "epoch": 0.24999, + "grad_norm": 1.2078500008734836, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 24999 + }, + { + "epoch": 0.25, + "grad_norm": 1.207230057391132, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 25000 + }, + { + "epoch": 0.25001, + "grad_norm": 1.1345898826247052, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 25001 + }, + { + "epoch": 0.25002, + "grad_norm": 1.097941971692078, + "learning_rate": 0.003, + "loss": 4.039, + "step": 25002 + }, + { + "epoch": 0.25003, + "grad_norm": 1.0341972365247325, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 25003 + }, + { + "epoch": 0.25004, + "grad_norm": 1.1097941945782377, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 25004 + }, + { + "epoch": 0.25005, + "grad_norm": 1.0897638265701954, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 25005 + }, + { + "epoch": 0.25006, + "grad_norm": 1.0033657918442598, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 25006 + }, + { + "epoch": 0.25007, + "grad_norm": 1.0562752604677648, + "learning_rate": 0.003, + "loss": 4.019, + "step": 25007 + }, + { + "epoch": 0.25008, + "grad_norm": 1.165357206519667, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 25008 + }, + { + "epoch": 0.25009, + "grad_norm": 1.2030805615047455, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 25009 + }, + { + "epoch": 0.2501, + "grad_norm": 0.8254858512727387, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 25010 + }, + { + "epoch": 0.25011, + "grad_norm": 0.9605356295015156, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 25011 + }, + { + "epoch": 0.25012, + "grad_norm": 1.220719204560168, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 25012 + }, + { + "epoch": 0.25013, + "grad_norm": 0.9491572481890724, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 25013 + }, + { + "epoch": 0.25014, + "grad_norm": 0.8579252873782365, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 25014 + }, + { + "epoch": 0.25015, + "grad_norm": 0.9588529869312399, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 25015 + }, + { + "epoch": 0.25016, + "grad_norm": 1.0804612437696341, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 25016 + }, + { + "epoch": 0.25017, + "grad_norm": 1.0593186403470476, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 25017 + }, + { + "epoch": 0.25018, + "grad_norm": 1.074780049572735, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 25018 + }, + { + "epoch": 0.25019, + "grad_norm": 1.0183529251366024, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 25019 + }, + { + "epoch": 0.2502, + "grad_norm": 1.322392539401327, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 25020 + }, + { + "epoch": 0.25021, + "grad_norm": 0.9132224397079756, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 25021 + }, + { + "epoch": 0.25022, + "grad_norm": 0.857644022807906, + "learning_rate": 0.003, + "loss": 4.056, + "step": 25022 + }, + { + "epoch": 0.25023, + "grad_norm": 1.0056891376115031, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 25023 + }, + { + "epoch": 0.25024, + "grad_norm": 1.264178805666726, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 25024 + }, + { + "epoch": 0.25025, + "grad_norm": 0.84405218457244, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 25025 + }, + { + "epoch": 0.25026, + "grad_norm": 0.995907564823086, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 25026 + }, + { + "epoch": 0.25027, + "grad_norm": 1.1297685620933566, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 25027 + }, + { + "epoch": 0.25028, + "grad_norm": 1.0616747652410368, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 25028 + }, + { + "epoch": 0.25029, + "grad_norm": 1.0851390768438514, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 25029 + }, + { + "epoch": 0.2503, + "grad_norm": 0.9834843078507347, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 25030 + }, + { + "epoch": 0.25031, + "grad_norm": 1.151580607269873, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 25031 + }, + { + "epoch": 0.25032, + "grad_norm": 0.8965496663643644, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 25032 + }, + { + "epoch": 0.25033, + "grad_norm": 1.0938551844959532, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 25033 + }, + { + "epoch": 0.25034, + "grad_norm": 1.2097529021952849, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 25034 + }, + { + "epoch": 0.25035, + "grad_norm": 1.0245051689604316, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 25035 + }, + { + "epoch": 0.25036, + "grad_norm": 1.191996013936727, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 25036 + }, + { + "epoch": 0.25037, + "grad_norm": 1.0756249981658754, + "learning_rate": 0.003, + "loss": 4.062, + "step": 25037 + }, + { + "epoch": 0.25038, + "grad_norm": 1.1947223202715342, + "learning_rate": 0.003, + "loss": 4.066, + "step": 25038 + }, + { + "epoch": 0.25039, + "grad_norm": 0.7864134532989098, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 25039 + }, + { + "epoch": 0.2504, + "grad_norm": 0.8880030765822636, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 25040 + }, + { + "epoch": 0.25041, + "grad_norm": 0.9458572540332919, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 25041 + }, + { + "epoch": 0.25042, + "grad_norm": 1.0512576603932353, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 25042 + }, + { + "epoch": 0.25043, + "grad_norm": 1.2182534879560263, + "learning_rate": 0.003, + "loss": 4.05, + "step": 25043 + }, + { + "epoch": 0.25044, + "grad_norm": 1.0657658916571664, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 25044 + }, + { + "epoch": 0.25045, + "grad_norm": 1.271665694204878, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 25045 + }, + { + "epoch": 0.25046, + "grad_norm": 1.2366058629163463, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 25046 + }, + { + "epoch": 0.25047, + "grad_norm": 1.2011889419224224, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 25047 + }, + { + "epoch": 0.25048, + "grad_norm": 0.9523268389234487, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 25048 + }, + { + "epoch": 0.25049, + "grad_norm": 1.0399105512054287, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 25049 + }, + { + "epoch": 0.2505, + "grad_norm": 0.9926524423325467, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 25050 + }, + { + "epoch": 0.25051, + "grad_norm": 0.9815803669771158, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 25051 + }, + { + "epoch": 0.25052, + "grad_norm": 1.1543986089481921, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 25052 + }, + { + "epoch": 0.25053, + "grad_norm": 0.928910127074389, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 25053 + }, + { + "epoch": 0.25054, + "grad_norm": 0.8532172734519416, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 25054 + }, + { + "epoch": 0.25055, + "grad_norm": 1.0115400068607534, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 25055 + }, + { + "epoch": 0.25056, + "grad_norm": 1.229764101255921, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 25056 + }, + { + "epoch": 0.25057, + "grad_norm": 1.0007742618203082, + "learning_rate": 0.003, + "loss": 4.063, + "step": 25057 + }, + { + "epoch": 0.25058, + "grad_norm": 1.2342956908425882, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 25058 + }, + { + "epoch": 0.25059, + "grad_norm": 1.1623283659871413, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 25059 + }, + { + "epoch": 0.2506, + "grad_norm": 1.0319354854796359, + "learning_rate": 0.003, + "loss": 4.038, + "step": 25060 + }, + { + "epoch": 0.25061, + "grad_norm": 1.1747937390307646, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 25061 + }, + { + "epoch": 0.25062, + "grad_norm": 1.0385681925525865, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 25062 + }, + { + "epoch": 0.25063, + "grad_norm": 1.1284852297177683, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 25063 + }, + { + "epoch": 0.25064, + "grad_norm": 0.9506627498289891, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 25064 + }, + { + "epoch": 0.25065, + "grad_norm": 1.1464517120779407, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 25065 + }, + { + "epoch": 0.25066, + "grad_norm": 1.0872888302896317, + "learning_rate": 0.003, + "loss": 4.054, + "step": 25066 + }, + { + "epoch": 0.25067, + "grad_norm": 1.1658185696417, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 25067 + }, + { + "epoch": 0.25068, + "grad_norm": 0.955366628128355, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 25068 + }, + { + "epoch": 0.25069, + "grad_norm": 1.085858385607787, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 25069 + }, + { + "epoch": 0.2507, + "grad_norm": 1.1535403352130718, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 25070 + }, + { + "epoch": 0.25071, + "grad_norm": 0.825719713150224, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 25071 + }, + { + "epoch": 0.25072, + "grad_norm": 0.8582014255968449, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 25072 + }, + { + "epoch": 0.25073, + "grad_norm": 1.0569015595204123, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 25073 + }, + { + "epoch": 0.25074, + "grad_norm": 1.2067991248257097, + "learning_rate": 0.003, + "loss": 4.0864, + "step": 25074 + }, + { + "epoch": 0.25075, + "grad_norm": 1.040576803936537, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 25075 + }, + { + "epoch": 0.25076, + "grad_norm": 1.2345777407414955, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 25076 + }, + { + "epoch": 0.25077, + "grad_norm": 0.9496678294252305, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 25077 + }, + { + "epoch": 0.25078, + "grad_norm": 1.0587275197520944, + "learning_rate": 0.003, + "loss": 4.054, + "step": 25078 + }, + { + "epoch": 0.25079, + "grad_norm": 1.1788059642539142, + "learning_rate": 0.003, + "loss": 4.026, + "step": 25079 + }, + { + "epoch": 0.2508, + "grad_norm": 0.8650535009772723, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 25080 + }, + { + "epoch": 0.25081, + "grad_norm": 0.8673987556638504, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 25081 + }, + { + "epoch": 0.25082, + "grad_norm": 0.8966696457188696, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 25082 + }, + { + "epoch": 0.25083, + "grad_norm": 1.0813213412698623, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 25083 + }, + { + "epoch": 0.25084, + "grad_norm": 1.3907587709599023, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 25084 + }, + { + "epoch": 0.25085, + "grad_norm": 0.6865589894300546, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 25085 + }, + { + "epoch": 0.25086, + "grad_norm": 0.9095268471483012, + "learning_rate": 0.003, + "loss": 4.054, + "step": 25086 + }, + { + "epoch": 0.25087, + "grad_norm": 1.0619766934309331, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 25087 + }, + { + "epoch": 0.25088, + "grad_norm": 1.2846037487718802, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 25088 + }, + { + "epoch": 0.25089, + "grad_norm": 1.103189724539586, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 25089 + }, + { + "epoch": 0.2509, + "grad_norm": 1.1508001884556365, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 25090 + }, + { + "epoch": 0.25091, + "grad_norm": 1.0899293772036127, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 25091 + }, + { + "epoch": 0.25092, + "grad_norm": 0.9639041711098145, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 25092 + }, + { + "epoch": 0.25093, + "grad_norm": 1.217702784752262, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 25093 + }, + { + "epoch": 0.25094, + "grad_norm": 0.9698592540440453, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 25094 + }, + { + "epoch": 0.25095, + "grad_norm": 1.0638171417266706, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 25095 + }, + { + "epoch": 0.25096, + "grad_norm": 1.0109062648753147, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 25096 + }, + { + "epoch": 0.25097, + "grad_norm": 1.078464021060432, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 25097 + }, + { + "epoch": 0.25098, + "grad_norm": 1.0957246371653435, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 25098 + }, + { + "epoch": 0.25099, + "grad_norm": 1.2052312080298775, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 25099 + }, + { + "epoch": 0.251, + "grad_norm": 0.9372170567170444, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 25100 + }, + { + "epoch": 0.25101, + "grad_norm": 1.0427183585296895, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 25101 + }, + { + "epoch": 0.25102, + "grad_norm": 1.3933994635335913, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 25102 + }, + { + "epoch": 0.25103, + "grad_norm": 0.8403424885580703, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 25103 + }, + { + "epoch": 0.25104, + "grad_norm": 0.809557858970937, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 25104 + }, + { + "epoch": 0.25105, + "grad_norm": 1.122045653239169, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 25105 + }, + { + "epoch": 0.25106, + "grad_norm": 1.2433279518395108, + "learning_rate": 0.003, + "loss": 4.062, + "step": 25106 + }, + { + "epoch": 0.25107, + "grad_norm": 0.9992050478104696, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 25107 + }, + { + "epoch": 0.25108, + "grad_norm": 1.2664436526934362, + "learning_rate": 0.003, + "loss": 4.042, + "step": 25108 + }, + { + "epoch": 0.25109, + "grad_norm": 0.9653464125470582, + "learning_rate": 0.003, + "loss": 4.047, + "step": 25109 + }, + { + "epoch": 0.2511, + "grad_norm": 1.2137928711004948, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 25110 + }, + { + "epoch": 0.25111, + "grad_norm": 1.0356291733342833, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 25111 + }, + { + "epoch": 0.25112, + "grad_norm": 1.1395462313183402, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 25112 + }, + { + "epoch": 0.25113, + "grad_norm": 0.9162786596007066, + "learning_rate": 0.003, + "loss": 4.084, + "step": 25113 + }, + { + "epoch": 0.25114, + "grad_norm": 1.1864580473246638, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 25114 + }, + { + "epoch": 0.25115, + "grad_norm": 0.9538408926182247, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 25115 + }, + { + "epoch": 0.25116, + "grad_norm": 1.0121392710359185, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 25116 + }, + { + "epoch": 0.25117, + "grad_norm": 1.1305716520495077, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 25117 + }, + { + "epoch": 0.25118, + "grad_norm": 0.9608107506960001, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 25118 + }, + { + "epoch": 0.25119, + "grad_norm": 1.1461279533209112, + "learning_rate": 0.003, + "loss": 4.011, + "step": 25119 + }, + { + "epoch": 0.2512, + "grad_norm": 1.3165853138396786, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 25120 + }, + { + "epoch": 0.25121, + "grad_norm": 1.1155688996836688, + "learning_rate": 0.003, + "loss": 4.075, + "step": 25121 + }, + { + "epoch": 0.25122, + "grad_norm": 1.1066475047832878, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 25122 + }, + { + "epoch": 0.25123, + "grad_norm": 1.1342373552131741, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 25123 + }, + { + "epoch": 0.25124, + "grad_norm": 0.8894700188102743, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 25124 + }, + { + "epoch": 0.25125, + "grad_norm": 1.077502255601287, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 25125 + }, + { + "epoch": 0.25126, + "grad_norm": 1.261271003768833, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 25126 + }, + { + "epoch": 0.25127, + "grad_norm": 0.8229103622806678, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 25127 + }, + { + "epoch": 0.25128, + "grad_norm": 0.8739765216613085, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 25128 + }, + { + "epoch": 0.25129, + "grad_norm": 1.130863011375631, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 25129 + }, + { + "epoch": 0.2513, + "grad_norm": 0.9591863037322964, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 25130 + }, + { + "epoch": 0.25131, + "grad_norm": 1.0206063674282162, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 25131 + }, + { + "epoch": 0.25132, + "grad_norm": 1.147011268339023, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 25132 + }, + { + "epoch": 0.25133, + "grad_norm": 0.9924375592710674, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 25133 + }, + { + "epoch": 0.25134, + "grad_norm": 1.0847734447183706, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 25134 + }, + { + "epoch": 0.25135, + "grad_norm": 1.091468510063335, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 25135 + }, + { + "epoch": 0.25136, + "grad_norm": 1.0865910673102057, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 25136 + }, + { + "epoch": 0.25137, + "grad_norm": 1.01360095099507, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 25137 + }, + { + "epoch": 0.25138, + "grad_norm": 1.1314526107796197, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 25138 + }, + { + "epoch": 0.25139, + "grad_norm": 1.2160792055968817, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 25139 + }, + { + "epoch": 0.2514, + "grad_norm": 0.9228547763255682, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 25140 + }, + { + "epoch": 0.25141, + "grad_norm": 1.05140448840119, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 25141 + }, + { + "epoch": 0.25142, + "grad_norm": 1.0213623839410477, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 25142 + }, + { + "epoch": 0.25143, + "grad_norm": 1.1177872370951922, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 25143 + }, + { + "epoch": 0.25144, + "grad_norm": 1.0039562542417373, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 25144 + }, + { + "epoch": 0.25145, + "grad_norm": 1.009194546328439, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 25145 + }, + { + "epoch": 0.25146, + "grad_norm": 1.1733367050156376, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 25146 + }, + { + "epoch": 0.25147, + "grad_norm": 1.0759854430481026, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 25147 + }, + { + "epoch": 0.25148, + "grad_norm": 0.9999854669950944, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 25148 + }, + { + "epoch": 0.25149, + "grad_norm": 1.077386782900836, + "learning_rate": 0.003, + "loss": 4.03, + "step": 25149 + }, + { + "epoch": 0.2515, + "grad_norm": 1.143796199176022, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 25150 + }, + { + "epoch": 0.25151, + "grad_norm": 1.1607856171250277, + "learning_rate": 0.003, + "loss": 4.065, + "step": 25151 + }, + { + "epoch": 0.25152, + "grad_norm": 1.0208460646189477, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 25152 + }, + { + "epoch": 0.25153, + "grad_norm": 1.5178591256487506, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 25153 + }, + { + "epoch": 0.25154, + "grad_norm": 0.7859865243332907, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 25154 + }, + { + "epoch": 0.25155, + "grad_norm": 0.8012136954525042, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 25155 + }, + { + "epoch": 0.25156, + "grad_norm": 1.0306451745001384, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 25156 + }, + { + "epoch": 0.25157, + "grad_norm": 1.090930406234507, + "learning_rate": 0.003, + "loss": 4.041, + "step": 25157 + }, + { + "epoch": 0.25158, + "grad_norm": 1.1278212852482086, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 25158 + }, + { + "epoch": 0.25159, + "grad_norm": 1.2350859648672199, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 25159 + }, + { + "epoch": 0.2516, + "grad_norm": 1.270355255716984, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 25160 + }, + { + "epoch": 0.25161, + "grad_norm": 0.9995662124468079, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 25161 + }, + { + "epoch": 0.25162, + "grad_norm": 1.0346066712070647, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 25162 + }, + { + "epoch": 0.25163, + "grad_norm": 1.1989958433448504, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 25163 + }, + { + "epoch": 0.25164, + "grad_norm": 0.9236617220226859, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 25164 + }, + { + "epoch": 0.25165, + "grad_norm": 1.0153980854104974, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 25165 + }, + { + "epoch": 0.25166, + "grad_norm": 1.2314533202211262, + "learning_rate": 0.003, + "loss": 4.049, + "step": 25166 + }, + { + "epoch": 0.25167, + "grad_norm": 0.8254389176482468, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 25167 + }, + { + "epoch": 0.25168, + "grad_norm": 0.8839581187124218, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 25168 + }, + { + "epoch": 0.25169, + "grad_norm": 1.0894862852047218, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 25169 + }, + { + "epoch": 0.2517, + "grad_norm": 1.1394776135406652, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 25170 + }, + { + "epoch": 0.25171, + "grad_norm": 0.9253599637149859, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 25171 + }, + { + "epoch": 0.25172, + "grad_norm": 1.1314968433583064, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 25172 + }, + { + "epoch": 0.25173, + "grad_norm": 1.1704878427281782, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 25173 + }, + { + "epoch": 0.25174, + "grad_norm": 0.9034289577249186, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 25174 + }, + { + "epoch": 0.25175, + "grad_norm": 1.0653551774317684, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 25175 + }, + { + "epoch": 0.25176, + "grad_norm": 0.9852229383300832, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 25176 + }, + { + "epoch": 0.25177, + "grad_norm": 1.0690878090526426, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 25177 + }, + { + "epoch": 0.25178, + "grad_norm": 0.9238929848197971, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 25178 + }, + { + "epoch": 0.25179, + "grad_norm": 1.2739452559564493, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 25179 + }, + { + "epoch": 0.2518, + "grad_norm": 1.0291331787922058, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 25180 + }, + { + "epoch": 0.25181, + "grad_norm": 0.9730724387439237, + "learning_rate": 0.003, + "loss": 4.03, + "step": 25181 + }, + { + "epoch": 0.25182, + "grad_norm": 1.1896887488552825, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 25182 + }, + { + "epoch": 0.25183, + "grad_norm": 1.106842806159237, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 25183 + }, + { + "epoch": 0.25184, + "grad_norm": 1.1029409253093339, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 25184 + }, + { + "epoch": 0.25185, + "grad_norm": 1.028603533870993, + "learning_rate": 0.003, + "loss": 4.031, + "step": 25185 + }, + { + "epoch": 0.25186, + "grad_norm": 0.9820968749704664, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 25186 + }, + { + "epoch": 0.25187, + "grad_norm": 0.923278125026581, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 25187 + }, + { + "epoch": 0.25188, + "grad_norm": 1.0680040520930079, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 25188 + }, + { + "epoch": 0.25189, + "grad_norm": 1.1615239878549934, + "learning_rate": 0.003, + "loss": 4.0956, + "step": 25189 + }, + { + "epoch": 0.2519, + "grad_norm": 1.2518505996705331, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 25190 + }, + { + "epoch": 0.25191, + "grad_norm": 0.9702966396679568, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 25191 + }, + { + "epoch": 0.25192, + "grad_norm": 1.105760268587983, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 25192 + }, + { + "epoch": 0.25193, + "grad_norm": 1.2299366737301394, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 25193 + }, + { + "epoch": 0.25194, + "grad_norm": 1.0339399138255887, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 25194 + }, + { + "epoch": 0.25195, + "grad_norm": 0.9899160133812825, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 25195 + }, + { + "epoch": 0.25196, + "grad_norm": 1.0432873035974988, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 25196 + }, + { + "epoch": 0.25197, + "grad_norm": 1.2533372493791837, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 25197 + }, + { + "epoch": 0.25198, + "grad_norm": 1.0722745673789618, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 25198 + }, + { + "epoch": 0.25199, + "grad_norm": 1.1465439265677482, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 25199 + }, + { + "epoch": 0.252, + "grad_norm": 1.0434961378290775, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 25200 + }, + { + "epoch": 0.25201, + "grad_norm": 1.0524321263177803, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 25201 + }, + { + "epoch": 0.25202, + "grad_norm": 1.0520890186614278, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 25202 + }, + { + "epoch": 0.25203, + "grad_norm": 1.0341354383889303, + "learning_rate": 0.003, + "loss": 4.019, + "step": 25203 + }, + { + "epoch": 0.25204, + "grad_norm": 1.080368353949088, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 25204 + }, + { + "epoch": 0.25205, + "grad_norm": 1.0008221121511507, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 25205 + }, + { + "epoch": 0.25206, + "grad_norm": 1.1608027523240836, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 25206 + }, + { + "epoch": 0.25207, + "grad_norm": 1.0174051354611995, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 25207 + }, + { + "epoch": 0.25208, + "grad_norm": 1.382669469305539, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 25208 + }, + { + "epoch": 0.25209, + "grad_norm": 1.0663674087836241, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 25209 + }, + { + "epoch": 0.2521, + "grad_norm": 1.058707853961933, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 25210 + }, + { + "epoch": 0.25211, + "grad_norm": 1.095317123833428, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 25211 + }, + { + "epoch": 0.25212, + "grad_norm": 1.278245720479363, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 25212 + }, + { + "epoch": 0.25213, + "grad_norm": 0.8403443384541186, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 25213 + }, + { + "epoch": 0.25214, + "grad_norm": 0.7439628200528042, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 25214 + }, + { + "epoch": 0.25215, + "grad_norm": 0.8252456074686969, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 25215 + }, + { + "epoch": 0.25216, + "grad_norm": 1.0974863067469216, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 25216 + }, + { + "epoch": 0.25217, + "grad_norm": 1.148782514363991, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 25217 + }, + { + "epoch": 0.25218, + "grad_norm": 0.8378678084204353, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 25218 + }, + { + "epoch": 0.25219, + "grad_norm": 0.8121727056459919, + "learning_rate": 0.003, + "loss": 4.039, + "step": 25219 + }, + { + "epoch": 0.2522, + "grad_norm": 0.905429146854106, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 25220 + }, + { + "epoch": 0.25221, + "grad_norm": 0.97484168491915, + "learning_rate": 0.003, + "loss": 4.0942, + "step": 25221 + }, + { + "epoch": 0.25222, + "grad_norm": 1.2915073053679207, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 25222 + }, + { + "epoch": 0.25223, + "grad_norm": 1.3089538218054597, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 25223 + }, + { + "epoch": 0.25224, + "grad_norm": 1.1909598413361546, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 25224 + }, + { + "epoch": 0.25225, + "grad_norm": 0.9766728159412146, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 25225 + }, + { + "epoch": 0.25226, + "grad_norm": 1.0144247936213326, + "learning_rate": 0.003, + "loss": 4.0864, + "step": 25226 + }, + { + "epoch": 0.25227, + "grad_norm": 1.1306109522381051, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 25227 + }, + { + "epoch": 0.25228, + "grad_norm": 0.8569998916553501, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 25228 + }, + { + "epoch": 0.25229, + "grad_norm": 0.9063253013670521, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 25229 + }, + { + "epoch": 0.2523, + "grad_norm": 1.032837036667242, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 25230 + }, + { + "epoch": 0.25231, + "grad_norm": 1.0755108942062237, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 25231 + }, + { + "epoch": 0.25232, + "grad_norm": 1.3346616535550742, + "learning_rate": 0.003, + "loss": 4.063, + "step": 25232 + }, + { + "epoch": 0.25233, + "grad_norm": 0.8861641979348999, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 25233 + }, + { + "epoch": 0.25234, + "grad_norm": 0.9427746842406144, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 25234 + }, + { + "epoch": 0.25235, + "grad_norm": 1.064553198020745, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 25235 + }, + { + "epoch": 0.25236, + "grad_norm": 1.1192419880901403, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 25236 + }, + { + "epoch": 0.25237, + "grad_norm": 0.8865611191796824, + "learning_rate": 0.003, + "loss": 4.034, + "step": 25237 + }, + { + "epoch": 0.25238, + "grad_norm": 1.0242760733644591, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 25238 + }, + { + "epoch": 0.25239, + "grad_norm": 1.078364378229095, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 25239 + }, + { + "epoch": 0.2524, + "grad_norm": 1.036660659800211, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 25240 + }, + { + "epoch": 0.25241, + "grad_norm": 0.9200308578326332, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 25241 + }, + { + "epoch": 0.25242, + "grad_norm": 1.078001141814157, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 25242 + }, + { + "epoch": 0.25243, + "grad_norm": 1.0615710799315676, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 25243 + }, + { + "epoch": 0.25244, + "grad_norm": 1.1960050928517731, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 25244 + }, + { + "epoch": 0.25245, + "grad_norm": 1.404469252244973, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 25245 + }, + { + "epoch": 0.25246, + "grad_norm": 1.0312606580665766, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 25246 + }, + { + "epoch": 0.25247, + "grad_norm": 1.3086810522957923, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 25247 + }, + { + "epoch": 0.25248, + "grad_norm": 1.0076530189648665, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 25248 + }, + { + "epoch": 0.25249, + "grad_norm": 1.1067784982361148, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 25249 + }, + { + "epoch": 0.2525, + "grad_norm": 1.1537513089809583, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 25250 + }, + { + "epoch": 0.25251, + "grad_norm": 0.9029079910180186, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 25251 + }, + { + "epoch": 0.25252, + "grad_norm": 1.0295978978299947, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 25252 + }, + { + "epoch": 0.25253, + "grad_norm": 1.2409204956417448, + "learning_rate": 0.003, + "loss": 4.0947, + "step": 25253 + }, + { + "epoch": 0.25254, + "grad_norm": 0.7827385293976917, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 25254 + }, + { + "epoch": 0.25255, + "grad_norm": 0.8242001421916544, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 25255 + }, + { + "epoch": 0.25256, + "grad_norm": 0.8752241112589209, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 25256 + }, + { + "epoch": 0.25257, + "grad_norm": 0.8599826448303941, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 25257 + }, + { + "epoch": 0.25258, + "grad_norm": 0.9434673714530487, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 25258 + }, + { + "epoch": 0.25259, + "grad_norm": 1.1918023187577267, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 25259 + }, + { + "epoch": 0.2526, + "grad_norm": 1.1489311092887493, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 25260 + }, + { + "epoch": 0.25261, + "grad_norm": 1.3893418606827428, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 25261 + }, + { + "epoch": 0.25262, + "grad_norm": 0.981343594074749, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 25262 + }, + { + "epoch": 0.25263, + "grad_norm": 1.1022935696488914, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 25263 + }, + { + "epoch": 0.25264, + "grad_norm": 1.2727643508697914, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 25264 + }, + { + "epoch": 0.25265, + "grad_norm": 1.1133496927623232, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 25265 + }, + { + "epoch": 0.25266, + "grad_norm": 1.240047733228059, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 25266 + }, + { + "epoch": 0.25267, + "grad_norm": 0.8541439399828976, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 25267 + }, + { + "epoch": 0.25268, + "grad_norm": 1.009563184500924, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 25268 + }, + { + "epoch": 0.25269, + "grad_norm": 1.070120823276679, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 25269 + }, + { + "epoch": 0.2527, + "grad_norm": 1.0876039751518398, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 25270 + }, + { + "epoch": 0.25271, + "grad_norm": 1.3119588144617038, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 25271 + }, + { + "epoch": 0.25272, + "grad_norm": 1.232600417452226, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 25272 + }, + { + "epoch": 0.25273, + "grad_norm": 0.8270392556802257, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 25273 + }, + { + "epoch": 0.25274, + "grad_norm": 0.8640430968952804, + "learning_rate": 0.003, + "loss": 4.028, + "step": 25274 + }, + { + "epoch": 0.25275, + "grad_norm": 1.0120955654205386, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 25275 + }, + { + "epoch": 0.25276, + "grad_norm": 1.1685999939910412, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 25276 + }, + { + "epoch": 0.25277, + "grad_norm": 0.9745780028980217, + "learning_rate": 0.003, + "loss": 4.064, + "step": 25277 + }, + { + "epoch": 0.25278, + "grad_norm": 1.131012781590332, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 25278 + }, + { + "epoch": 0.25279, + "grad_norm": 1.080790321414541, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 25279 + }, + { + "epoch": 0.2528, + "grad_norm": 1.1030484817056916, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 25280 + }, + { + "epoch": 0.25281, + "grad_norm": 0.9935141231569282, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 25281 + }, + { + "epoch": 0.25282, + "grad_norm": 1.282226566146276, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 25282 + }, + { + "epoch": 0.25283, + "grad_norm": 0.9818332767846524, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 25283 + }, + { + "epoch": 0.25284, + "grad_norm": 1.0302106055079412, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 25284 + }, + { + "epoch": 0.25285, + "grad_norm": 1.1920225955571557, + "learning_rate": 0.003, + "loss": 4.072, + "step": 25285 + }, + { + "epoch": 0.25286, + "grad_norm": 1.0187322745075051, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 25286 + }, + { + "epoch": 0.25287, + "grad_norm": 0.9432252703317576, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 25287 + }, + { + "epoch": 0.25288, + "grad_norm": 1.0167953275414365, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 25288 + }, + { + "epoch": 0.25289, + "grad_norm": 1.0652784187372233, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 25289 + }, + { + "epoch": 0.2529, + "grad_norm": 0.9910825004436132, + "learning_rate": 0.003, + "loss": 4.04, + "step": 25290 + }, + { + "epoch": 0.25291, + "grad_norm": 1.0906377814735368, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 25291 + }, + { + "epoch": 0.25292, + "grad_norm": 1.0415852860425519, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 25292 + }, + { + "epoch": 0.25293, + "grad_norm": 1.0508537020307487, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 25293 + }, + { + "epoch": 0.25294, + "grad_norm": 1.1099876965364919, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 25294 + }, + { + "epoch": 0.25295, + "grad_norm": 1.060385248708587, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 25295 + }, + { + "epoch": 0.25296, + "grad_norm": 1.0700704764639823, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 25296 + }, + { + "epoch": 0.25297, + "grad_norm": 1.0566020148464017, + "learning_rate": 0.003, + "loss": 4.0871, + "step": 25297 + }, + { + "epoch": 0.25298, + "grad_norm": 1.2988438370321724, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 25298 + }, + { + "epoch": 0.25299, + "grad_norm": 1.2164074608103923, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 25299 + }, + { + "epoch": 0.253, + "grad_norm": 1.1024407276579824, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 25300 + }, + { + "epoch": 0.25301, + "grad_norm": 0.9357073725751937, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 25301 + }, + { + "epoch": 0.25302, + "grad_norm": 1.103323952355201, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 25302 + }, + { + "epoch": 0.25303, + "grad_norm": 1.089893075480791, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 25303 + }, + { + "epoch": 0.25304, + "grad_norm": 1.03718863980005, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 25304 + }, + { + "epoch": 0.25305, + "grad_norm": 1.0924939228466999, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 25305 + }, + { + "epoch": 0.25306, + "grad_norm": 0.9739820243704387, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 25306 + }, + { + "epoch": 0.25307, + "grad_norm": 0.9805864802140074, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 25307 + }, + { + "epoch": 0.25308, + "grad_norm": 1.0245302116132284, + "learning_rate": 0.003, + "loss": 4.038, + "step": 25308 + }, + { + "epoch": 0.25309, + "grad_norm": 1.1312621331739925, + "learning_rate": 0.003, + "loss": 4.063, + "step": 25309 + }, + { + "epoch": 0.2531, + "grad_norm": 0.9280568089027946, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 25310 + }, + { + "epoch": 0.25311, + "grad_norm": 0.9705990172885777, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 25311 + }, + { + "epoch": 0.25312, + "grad_norm": 1.0269656711264628, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 25312 + }, + { + "epoch": 0.25313, + "grad_norm": 1.1978756456049005, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 25313 + }, + { + "epoch": 0.25314, + "grad_norm": 1.2142360001075765, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 25314 + }, + { + "epoch": 0.25315, + "grad_norm": 1.0033532181032645, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 25315 + }, + { + "epoch": 0.25316, + "grad_norm": 1.0992272319566343, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 25316 + }, + { + "epoch": 0.25317, + "grad_norm": 1.013955577029886, + "learning_rate": 0.003, + "loss": 4.066, + "step": 25317 + }, + { + "epoch": 0.25318, + "grad_norm": 1.2789825308794667, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 25318 + }, + { + "epoch": 0.25319, + "grad_norm": 1.1110352603064015, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 25319 + }, + { + "epoch": 0.2532, + "grad_norm": 1.131070748324791, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 25320 + }, + { + "epoch": 0.25321, + "grad_norm": 1.106490432856189, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 25321 + }, + { + "epoch": 0.25322, + "grad_norm": 1.1558099843517327, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 25322 + }, + { + "epoch": 0.25323, + "grad_norm": 1.115990790673505, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 25323 + }, + { + "epoch": 0.25324, + "grad_norm": 0.9026688988556, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 25324 + }, + { + "epoch": 0.25325, + "grad_norm": 0.9383385201746441, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 25325 + }, + { + "epoch": 0.25326, + "grad_norm": 0.9537282477247824, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 25326 + }, + { + "epoch": 0.25327, + "grad_norm": 1.0804600184594706, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 25327 + }, + { + "epoch": 0.25328, + "grad_norm": 1.1462655510773163, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 25328 + }, + { + "epoch": 0.25329, + "grad_norm": 1.3594755430267387, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 25329 + }, + { + "epoch": 0.2533, + "grad_norm": 0.7731576866406855, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 25330 + }, + { + "epoch": 0.25331, + "grad_norm": 0.80358500720989, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 25331 + }, + { + "epoch": 0.25332, + "grad_norm": 0.926510428223034, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 25332 + }, + { + "epoch": 0.25333, + "grad_norm": 1.3375319850009542, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 25333 + }, + { + "epoch": 0.25334, + "grad_norm": 0.9341210802941049, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 25334 + }, + { + "epoch": 0.25335, + "grad_norm": 0.9593045586554779, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 25335 + }, + { + "epoch": 0.25336, + "grad_norm": 1.1354624792589323, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 25336 + }, + { + "epoch": 0.25337, + "grad_norm": 1.0896323551909735, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 25337 + }, + { + "epoch": 0.25338, + "grad_norm": 1.0879061797304714, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 25338 + }, + { + "epoch": 0.25339, + "grad_norm": 0.8724956441640512, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 25339 + }, + { + "epoch": 0.2534, + "grad_norm": 1.0564963131198284, + "learning_rate": 0.003, + "loss": 4.05, + "step": 25340 + }, + { + "epoch": 0.25341, + "grad_norm": 1.335347576770556, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 25341 + }, + { + "epoch": 0.25342, + "grad_norm": 1.1208877139310376, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 25342 + }, + { + "epoch": 0.25343, + "grad_norm": 1.074531161920392, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 25343 + }, + { + "epoch": 0.25344, + "grad_norm": 1.0821923754063587, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 25344 + }, + { + "epoch": 0.25345, + "grad_norm": 1.1634673306544963, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 25345 + }, + { + "epoch": 0.25346, + "grad_norm": 1.0351826748621826, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 25346 + }, + { + "epoch": 0.25347, + "grad_norm": 1.117496630961491, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 25347 + }, + { + "epoch": 0.25348, + "grad_norm": 0.9718142101822199, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 25348 + }, + { + "epoch": 0.25349, + "grad_norm": 1.247528434230533, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 25349 + }, + { + "epoch": 0.2535, + "grad_norm": 0.9289010352418264, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 25350 + }, + { + "epoch": 0.25351, + "grad_norm": 1.0203033850683356, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 25351 + }, + { + "epoch": 0.25352, + "grad_norm": 1.142335853930613, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 25352 + }, + { + "epoch": 0.25353, + "grad_norm": 1.006752887660181, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 25353 + }, + { + "epoch": 0.25354, + "grad_norm": 1.0828876725507073, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 25354 + }, + { + "epoch": 0.25355, + "grad_norm": 1.1047687385394456, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 25355 + }, + { + "epoch": 0.25356, + "grad_norm": 1.1356967773393407, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 25356 + }, + { + "epoch": 0.25357, + "grad_norm": 0.9144792985208664, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 25357 + }, + { + "epoch": 0.25358, + "grad_norm": 0.9109186743053219, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 25358 + }, + { + "epoch": 0.25359, + "grad_norm": 1.345288736876884, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 25359 + }, + { + "epoch": 0.2536, + "grad_norm": 1.0298366156604293, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 25360 + }, + { + "epoch": 0.25361, + "grad_norm": 1.0656290561145978, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 25361 + }, + { + "epoch": 0.25362, + "grad_norm": 1.004886434867466, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 25362 + }, + { + "epoch": 0.25363, + "grad_norm": 1.2233268680822535, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 25363 + }, + { + "epoch": 0.25364, + "grad_norm": 1.0332912305513835, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 25364 + }, + { + "epoch": 0.25365, + "grad_norm": 1.0126984559172574, + "learning_rate": 0.003, + "loss": 4.0873, + "step": 25365 + }, + { + "epoch": 0.25366, + "grad_norm": 1.1209282294509557, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 25366 + }, + { + "epoch": 0.25367, + "grad_norm": 0.9084865673330188, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 25367 + }, + { + "epoch": 0.25368, + "grad_norm": 0.989460958061078, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 25368 + }, + { + "epoch": 0.25369, + "grad_norm": 1.0464266974467533, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 25369 + }, + { + "epoch": 0.2537, + "grad_norm": 1.2398630567637812, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 25370 + }, + { + "epoch": 0.25371, + "grad_norm": 1.4001939580063134, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 25371 + }, + { + "epoch": 0.25372, + "grad_norm": 0.9304552737753433, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 25372 + }, + { + "epoch": 0.25373, + "grad_norm": 0.9941543353358434, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 25373 + }, + { + "epoch": 0.25374, + "grad_norm": 1.3719534823650326, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 25374 + }, + { + "epoch": 0.25375, + "grad_norm": 0.8857553266874272, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 25375 + }, + { + "epoch": 0.25376, + "grad_norm": 0.9409299836231735, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 25376 + }, + { + "epoch": 0.25377, + "grad_norm": 1.1757596258906116, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 25377 + }, + { + "epoch": 0.25378, + "grad_norm": 1.0450922674644367, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 25378 + }, + { + "epoch": 0.25379, + "grad_norm": 0.9742328698099912, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 25379 + }, + { + "epoch": 0.2538, + "grad_norm": 1.073434878223569, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 25380 + }, + { + "epoch": 0.25381, + "grad_norm": 0.9678184024530108, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 25381 + }, + { + "epoch": 0.25382, + "grad_norm": 1.240118554233363, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 25382 + }, + { + "epoch": 0.25383, + "grad_norm": 1.0491761576599021, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 25383 + }, + { + "epoch": 0.25384, + "grad_norm": 1.011453567458961, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 25384 + }, + { + "epoch": 0.25385, + "grad_norm": 1.0621521138639507, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 25385 + }, + { + "epoch": 0.25386, + "grad_norm": 1.2192297735671282, + "learning_rate": 0.003, + "loss": 4.045, + "step": 25386 + }, + { + "epoch": 0.25387, + "grad_norm": 1.300993188925232, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 25387 + }, + { + "epoch": 0.25388, + "grad_norm": 1.0458538657367398, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 25388 + }, + { + "epoch": 0.25389, + "grad_norm": 1.1514999846378908, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 25389 + }, + { + "epoch": 0.2539, + "grad_norm": 1.0430464207052446, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 25390 + }, + { + "epoch": 0.25391, + "grad_norm": 1.3768647809082635, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 25391 + }, + { + "epoch": 0.25392, + "grad_norm": 0.9685474731845205, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 25392 + }, + { + "epoch": 0.25393, + "grad_norm": 0.9765825115582891, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 25393 + }, + { + "epoch": 0.25394, + "grad_norm": 1.1412503446985185, + "learning_rate": 0.003, + "loss": 4.1017, + "step": 25394 + }, + { + "epoch": 0.25395, + "grad_norm": 1.124158172781498, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 25395 + }, + { + "epoch": 0.25396, + "grad_norm": 0.8744446307397139, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 25396 + }, + { + "epoch": 0.25397, + "grad_norm": 0.9878567062462399, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 25397 + }, + { + "epoch": 0.25398, + "grad_norm": 1.0588319227928873, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 25398 + }, + { + "epoch": 0.25399, + "grad_norm": 0.7982080768002091, + "learning_rate": 0.003, + "loss": 4.04, + "step": 25399 + }, + { + "epoch": 0.254, + "grad_norm": 0.8054274945611255, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 25400 + }, + { + "epoch": 0.25401, + "grad_norm": 1.123753539113704, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 25401 + }, + { + "epoch": 0.25402, + "grad_norm": 1.188369073740504, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 25402 + }, + { + "epoch": 0.25403, + "grad_norm": 0.966003603247998, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 25403 + }, + { + "epoch": 0.25404, + "grad_norm": 0.9934729118510304, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 25404 + }, + { + "epoch": 0.25405, + "grad_norm": 1.0346459288380576, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 25405 + }, + { + "epoch": 0.25406, + "grad_norm": 1.3710242686619716, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 25406 + }, + { + "epoch": 0.25407, + "grad_norm": 0.9004164883194008, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 25407 + }, + { + "epoch": 0.25408, + "grad_norm": 1.014114105928067, + "learning_rate": 0.003, + "loss": 4.046, + "step": 25408 + }, + { + "epoch": 0.25409, + "grad_norm": 1.2618740168148952, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 25409 + }, + { + "epoch": 0.2541, + "grad_norm": 1.0076755534233977, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 25410 + }, + { + "epoch": 0.25411, + "grad_norm": 1.185502124418848, + "learning_rate": 0.003, + "loss": 4.06, + "step": 25411 + }, + { + "epoch": 0.25412, + "grad_norm": 0.988698834557606, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 25412 + }, + { + "epoch": 0.25413, + "grad_norm": 1.1178783835642205, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 25413 + }, + { + "epoch": 0.25414, + "grad_norm": 0.9641657107522036, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 25414 + }, + { + "epoch": 0.25415, + "grad_norm": 1.2226174711266002, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 25415 + }, + { + "epoch": 0.25416, + "grad_norm": 1.0145562128854304, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 25416 + }, + { + "epoch": 0.25417, + "grad_norm": 1.247632899195283, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 25417 + }, + { + "epoch": 0.25418, + "grad_norm": 0.995871107436543, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 25418 + }, + { + "epoch": 0.25419, + "grad_norm": 1.118923946267722, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 25419 + }, + { + "epoch": 0.2542, + "grad_norm": 1.0289647820757026, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 25420 + }, + { + "epoch": 0.25421, + "grad_norm": 1.0793068721172716, + "learning_rate": 0.003, + "loss": 4.017, + "step": 25421 + }, + { + "epoch": 0.25422, + "grad_norm": 1.4029239422895274, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 25422 + }, + { + "epoch": 0.25423, + "grad_norm": 0.8652486831014439, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 25423 + }, + { + "epoch": 0.25424, + "grad_norm": 0.9651877272220147, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 25424 + }, + { + "epoch": 0.25425, + "grad_norm": 1.0847302012549556, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 25425 + }, + { + "epoch": 0.25426, + "grad_norm": 1.1698617368243918, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 25426 + }, + { + "epoch": 0.25427, + "grad_norm": 1.0293645870159536, + "learning_rate": 0.003, + "loss": 4.039, + "step": 25427 + }, + { + "epoch": 0.25428, + "grad_norm": 0.9502539540717984, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 25428 + }, + { + "epoch": 0.25429, + "grad_norm": 1.1083626797084722, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 25429 + }, + { + "epoch": 0.2543, + "grad_norm": 0.9652205758280874, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 25430 + }, + { + "epoch": 0.25431, + "grad_norm": 1.086912172939567, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 25431 + }, + { + "epoch": 0.25432, + "grad_norm": 1.2245400583463923, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 25432 + }, + { + "epoch": 0.25433, + "grad_norm": 0.9924750823911994, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 25433 + }, + { + "epoch": 0.25434, + "grad_norm": 1.071692994714624, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 25434 + }, + { + "epoch": 0.25435, + "grad_norm": 1.2183349329463664, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 25435 + }, + { + "epoch": 0.25436, + "grad_norm": 0.94248802887086, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 25436 + }, + { + "epoch": 0.25437, + "grad_norm": 0.8981862481955479, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 25437 + }, + { + "epoch": 0.25438, + "grad_norm": 1.052262116164484, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 25438 + }, + { + "epoch": 0.25439, + "grad_norm": 1.0147398726340924, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 25439 + }, + { + "epoch": 0.2544, + "grad_norm": 0.9543293073453261, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 25440 + }, + { + "epoch": 0.25441, + "grad_norm": 1.0539805706442553, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 25441 + }, + { + "epoch": 0.25442, + "grad_norm": 1.167661536033742, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 25442 + }, + { + "epoch": 0.25443, + "grad_norm": 1.0755573693247327, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 25443 + }, + { + "epoch": 0.25444, + "grad_norm": 1.114054733573648, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 25444 + }, + { + "epoch": 0.25445, + "grad_norm": 1.1885728578725232, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 25445 + }, + { + "epoch": 0.25446, + "grad_norm": 1.0613628925370688, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 25446 + }, + { + "epoch": 0.25447, + "grad_norm": 1.3070957997651138, + "learning_rate": 0.003, + "loss": 4.0857, + "step": 25447 + }, + { + "epoch": 0.25448, + "grad_norm": 1.1986538402106321, + "learning_rate": 0.003, + "loss": 4.029, + "step": 25448 + }, + { + "epoch": 0.25449, + "grad_norm": 0.9961598072702109, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 25449 + }, + { + "epoch": 0.2545, + "grad_norm": 1.145891661787272, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 25450 + }, + { + "epoch": 0.25451, + "grad_norm": 1.010950434181964, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 25451 + }, + { + "epoch": 0.25452, + "grad_norm": 0.9532095909875067, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 25452 + }, + { + "epoch": 0.25453, + "grad_norm": 0.9704725757741797, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 25453 + }, + { + "epoch": 0.25454, + "grad_norm": 0.9379905421142546, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 25454 + }, + { + "epoch": 0.25455, + "grad_norm": 1.0187790127591874, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 25455 + }, + { + "epoch": 0.25456, + "grad_norm": 1.1327782621319578, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 25456 + }, + { + "epoch": 0.25457, + "grad_norm": 1.1097637510687608, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 25457 + }, + { + "epoch": 0.25458, + "grad_norm": 1.182743629835161, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 25458 + }, + { + "epoch": 0.25459, + "grad_norm": 0.9511745577844828, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 25459 + }, + { + "epoch": 0.2546, + "grad_norm": 1.010645737750688, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 25460 + }, + { + "epoch": 0.25461, + "grad_norm": 1.2547951794994188, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 25461 + }, + { + "epoch": 0.25462, + "grad_norm": 0.7637937119440498, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 25462 + }, + { + "epoch": 0.25463, + "grad_norm": 0.6720642829898615, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 25463 + }, + { + "epoch": 0.25464, + "grad_norm": 0.7360349487082786, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 25464 + }, + { + "epoch": 0.25465, + "grad_norm": 0.7982356737143727, + "learning_rate": 0.003, + "loss": 4.046, + "step": 25465 + }, + { + "epoch": 0.25466, + "grad_norm": 0.9982095189717287, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 25466 + }, + { + "epoch": 0.25467, + "grad_norm": 1.411668369518374, + "learning_rate": 0.003, + "loss": 4.049, + "step": 25467 + }, + { + "epoch": 0.25468, + "grad_norm": 0.936289217277403, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 25468 + }, + { + "epoch": 0.25469, + "grad_norm": 1.3454855075154581, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 25469 + }, + { + "epoch": 0.2547, + "grad_norm": 0.9998231142611824, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 25470 + }, + { + "epoch": 0.25471, + "grad_norm": 1.0441867857455918, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 25471 + }, + { + "epoch": 0.25472, + "grad_norm": 1.250452122013893, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 25472 + }, + { + "epoch": 0.25473, + "grad_norm": 1.1601568615025755, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 25473 + }, + { + "epoch": 0.25474, + "grad_norm": 1.1033639618211188, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 25474 + }, + { + "epoch": 0.25475, + "grad_norm": 1.1098978796544139, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 25475 + }, + { + "epoch": 0.25476, + "grad_norm": 1.014338368487571, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 25476 + }, + { + "epoch": 0.25477, + "grad_norm": 0.9462015460403654, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 25477 + }, + { + "epoch": 0.25478, + "grad_norm": 1.0253965009946904, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 25478 + }, + { + "epoch": 0.25479, + "grad_norm": 1.1441568211961004, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 25479 + }, + { + "epoch": 0.2548, + "grad_norm": 1.1713423045706806, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 25480 + }, + { + "epoch": 0.25481, + "grad_norm": 0.9206899839722749, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 25481 + }, + { + "epoch": 0.25482, + "grad_norm": 1.0634825826908725, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 25482 + }, + { + "epoch": 0.25483, + "grad_norm": 1.179177509099683, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 25483 + }, + { + "epoch": 0.25484, + "grad_norm": 1.0483680807019429, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 25484 + }, + { + "epoch": 0.25485, + "grad_norm": 1.0416365464003097, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 25485 + }, + { + "epoch": 0.25486, + "grad_norm": 1.038522829553971, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 25486 + }, + { + "epoch": 0.25487, + "grad_norm": 1.0977570736651059, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 25487 + }, + { + "epoch": 0.25488, + "grad_norm": 1.0063232806075975, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 25488 + }, + { + "epoch": 0.25489, + "grad_norm": 1.1378693277089162, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 25489 + }, + { + "epoch": 0.2549, + "grad_norm": 1.1372752633271328, + "learning_rate": 0.003, + "loss": 4.0935, + "step": 25490 + }, + { + "epoch": 0.25491, + "grad_norm": 1.1057467523463982, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 25491 + }, + { + "epoch": 0.25492, + "grad_norm": 0.9452754636024293, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 25492 + }, + { + "epoch": 0.25493, + "grad_norm": 0.9652452267477947, + "learning_rate": 0.003, + "loss": 4.02, + "step": 25493 + }, + { + "epoch": 0.25494, + "grad_norm": 1.1527012935007988, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 25494 + }, + { + "epoch": 0.25495, + "grad_norm": 1.2410167663663763, + "learning_rate": 0.003, + "loss": 4.076, + "step": 25495 + }, + { + "epoch": 0.25496, + "grad_norm": 0.9390500347810612, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 25496 + }, + { + "epoch": 0.25497, + "grad_norm": 0.9836318206595412, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 25497 + }, + { + "epoch": 0.25498, + "grad_norm": 1.1094512355391062, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 25498 + }, + { + "epoch": 0.25499, + "grad_norm": 0.9997342483482291, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 25499 + }, + { + "epoch": 0.255, + "grad_norm": 1.0694182558596588, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 25500 + }, + { + "epoch": 0.25501, + "grad_norm": 1.1236221587251023, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 25501 + }, + { + "epoch": 0.25502, + "grad_norm": 1.1650948565144001, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 25502 + }, + { + "epoch": 0.25503, + "grad_norm": 0.991502951982773, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 25503 + }, + { + "epoch": 0.25504, + "grad_norm": 1.1899273548698877, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 25504 + }, + { + "epoch": 0.25505, + "grad_norm": 1.065104142505017, + "learning_rate": 0.003, + "loss": 4.034, + "step": 25505 + }, + { + "epoch": 0.25506, + "grad_norm": 1.0789212073678438, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 25506 + }, + { + "epoch": 0.25507, + "grad_norm": 1.0477468981934364, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 25507 + }, + { + "epoch": 0.25508, + "grad_norm": 1.2459495594226975, + "learning_rate": 0.003, + "loss": 4.1006, + "step": 25508 + }, + { + "epoch": 0.25509, + "grad_norm": 1.119901948446813, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 25509 + }, + { + "epoch": 0.2551, + "grad_norm": 1.0936543640634946, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 25510 + }, + { + "epoch": 0.25511, + "grad_norm": 1.1513881922516744, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 25511 + }, + { + "epoch": 0.25512, + "grad_norm": 0.9600585168677739, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 25512 + }, + { + "epoch": 0.25513, + "grad_norm": 1.0480154361122629, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 25513 + }, + { + "epoch": 0.25514, + "grad_norm": 1.0470406396971197, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 25514 + }, + { + "epoch": 0.25515, + "grad_norm": 0.9511427363564062, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 25515 + }, + { + "epoch": 0.25516, + "grad_norm": 1.239348097148028, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 25516 + }, + { + "epoch": 0.25517, + "grad_norm": 1.1642254805770396, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 25517 + }, + { + "epoch": 0.25518, + "grad_norm": 1.097736548315423, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 25518 + }, + { + "epoch": 0.25519, + "grad_norm": 1.092197444736002, + "learning_rate": 0.003, + "loss": 4.076, + "step": 25519 + }, + { + "epoch": 0.2552, + "grad_norm": 1.0099974916166534, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 25520 + }, + { + "epoch": 0.25521, + "grad_norm": 1.1154637653618857, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 25521 + }, + { + "epoch": 0.25522, + "grad_norm": 0.9836071178220482, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 25522 + }, + { + "epoch": 0.25523, + "grad_norm": 1.0610963735421417, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 25523 + }, + { + "epoch": 0.25524, + "grad_norm": 1.1457985500188876, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 25524 + }, + { + "epoch": 0.25525, + "grad_norm": 0.914181318605516, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 25525 + }, + { + "epoch": 0.25526, + "grad_norm": 0.9783565637454834, + "learning_rate": 0.003, + "loss": 4.027, + "step": 25526 + }, + { + "epoch": 0.25527, + "grad_norm": 1.0500412485119297, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 25527 + }, + { + "epoch": 0.25528, + "grad_norm": 1.2574711903455622, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 25528 + }, + { + "epoch": 0.25529, + "grad_norm": 0.907133523113672, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 25529 + }, + { + "epoch": 0.2553, + "grad_norm": 1.147749192491452, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 25530 + }, + { + "epoch": 0.25531, + "grad_norm": 1.2776166193898695, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 25531 + }, + { + "epoch": 0.25532, + "grad_norm": 1.0847660596682827, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 25532 + }, + { + "epoch": 0.25533, + "grad_norm": 1.162049265149529, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 25533 + }, + { + "epoch": 0.25534, + "grad_norm": 0.9335518415358439, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 25534 + }, + { + "epoch": 0.25535, + "grad_norm": 1.1227862950731464, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 25535 + }, + { + "epoch": 0.25536, + "grad_norm": 1.2817697966724912, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 25536 + }, + { + "epoch": 0.25537, + "grad_norm": 0.9828059653492026, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 25537 + }, + { + "epoch": 0.25538, + "grad_norm": 1.0088744361636004, + "learning_rate": 0.003, + "loss": 4.038, + "step": 25538 + }, + { + "epoch": 0.25539, + "grad_norm": 1.0064186511739717, + "learning_rate": 0.003, + "loss": 4.055, + "step": 25539 + }, + { + "epoch": 0.2554, + "grad_norm": 1.1921321969781742, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 25540 + }, + { + "epoch": 0.25541, + "grad_norm": 1.0429432527769111, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 25541 + }, + { + "epoch": 0.25542, + "grad_norm": 1.4404546314247881, + "learning_rate": 0.003, + "loss": 4.082, + "step": 25542 + }, + { + "epoch": 0.25543, + "grad_norm": 0.801643141802338, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 25543 + }, + { + "epoch": 0.25544, + "grad_norm": 0.8737621989161353, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 25544 + }, + { + "epoch": 0.25545, + "grad_norm": 1.0198431669336232, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 25545 + }, + { + "epoch": 0.25546, + "grad_norm": 1.3091057992180881, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 25546 + }, + { + "epoch": 0.25547, + "grad_norm": 1.0118308683258004, + "learning_rate": 0.003, + "loss": 4.0926, + "step": 25547 + }, + { + "epoch": 0.25548, + "grad_norm": 0.9559239083204156, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 25548 + }, + { + "epoch": 0.25549, + "grad_norm": 1.209839288202283, + "learning_rate": 0.003, + "loss": 4.0901, + "step": 25549 + }, + { + "epoch": 0.2555, + "grad_norm": 1.3866657522536683, + "learning_rate": 0.003, + "loss": 4.0912, + "step": 25550 + }, + { + "epoch": 0.25551, + "grad_norm": 0.7857814202905293, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 25551 + }, + { + "epoch": 0.25552, + "grad_norm": 0.7022120645041122, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 25552 + }, + { + "epoch": 0.25553, + "grad_norm": 0.8319616696849653, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 25553 + }, + { + "epoch": 0.25554, + "grad_norm": 0.9550096820055958, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 25554 + }, + { + "epoch": 0.25555, + "grad_norm": 1.2743065464756331, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 25555 + }, + { + "epoch": 0.25556, + "grad_norm": 1.1508896966774198, + "learning_rate": 0.003, + "loss": 4.059, + "step": 25556 + }, + { + "epoch": 0.25557, + "grad_norm": 1.1190564697026235, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 25557 + }, + { + "epoch": 0.25558, + "grad_norm": 0.9561446426829947, + "learning_rate": 0.003, + "loss": 4.018, + "step": 25558 + }, + { + "epoch": 0.25559, + "grad_norm": 1.075113050122302, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 25559 + }, + { + "epoch": 0.2556, + "grad_norm": 0.948908627564799, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 25560 + }, + { + "epoch": 0.25561, + "grad_norm": 0.9376243025120791, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 25561 + }, + { + "epoch": 0.25562, + "grad_norm": 1.0623750857890364, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 25562 + }, + { + "epoch": 0.25563, + "grad_norm": 1.1161695287217515, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 25563 + }, + { + "epoch": 0.25564, + "grad_norm": 0.9438129110655388, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 25564 + }, + { + "epoch": 0.25565, + "grad_norm": 0.8451488015836276, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 25565 + }, + { + "epoch": 0.25566, + "grad_norm": 0.9404825565338486, + "learning_rate": 0.003, + "loss": 4.063, + "step": 25566 + }, + { + "epoch": 0.25567, + "grad_norm": 1.184753143325313, + "learning_rate": 0.003, + "loss": 4.075, + "step": 25567 + }, + { + "epoch": 0.25568, + "grad_norm": 1.2239961822618268, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 25568 + }, + { + "epoch": 0.25569, + "grad_norm": 0.8024522186837126, + "learning_rate": 0.003, + "loss": 4.0857, + "step": 25569 + }, + { + "epoch": 0.2557, + "grad_norm": 0.8674409323962663, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 25570 + }, + { + "epoch": 0.25571, + "grad_norm": 1.3268510683360781, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 25571 + }, + { + "epoch": 0.25572, + "grad_norm": 1.1289635465196903, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 25572 + }, + { + "epoch": 0.25573, + "grad_norm": 1.2248639204793568, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 25573 + }, + { + "epoch": 0.25574, + "grad_norm": 1.0355262857204486, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 25574 + }, + { + "epoch": 0.25575, + "grad_norm": 1.2686299937578585, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 25575 + }, + { + "epoch": 0.25576, + "grad_norm": 0.9100166646583047, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 25576 + }, + { + "epoch": 0.25577, + "grad_norm": 0.9604707172574886, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 25577 + }, + { + "epoch": 0.25578, + "grad_norm": 1.0472260908141178, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 25578 + }, + { + "epoch": 0.25579, + "grad_norm": 1.1766448848335904, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 25579 + }, + { + "epoch": 0.2558, + "grad_norm": 1.2988121391691725, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 25580 + }, + { + "epoch": 0.25581, + "grad_norm": 0.9943813890459962, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 25581 + }, + { + "epoch": 0.25582, + "grad_norm": 1.0145488445820123, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 25582 + }, + { + "epoch": 0.25583, + "grad_norm": 1.217603671336377, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 25583 + }, + { + "epoch": 0.25584, + "grad_norm": 1.1386689037576228, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 25584 + }, + { + "epoch": 0.25585, + "grad_norm": 0.8694036108232888, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 25585 + }, + { + "epoch": 0.25586, + "grad_norm": 0.8025316069538043, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 25586 + }, + { + "epoch": 0.25587, + "grad_norm": 0.944419727690762, + "learning_rate": 0.003, + "loss": 4.051, + "step": 25587 + }, + { + "epoch": 0.25588, + "grad_norm": 1.2218153026907868, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 25588 + }, + { + "epoch": 0.25589, + "grad_norm": 1.3129107878604411, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 25589 + }, + { + "epoch": 0.2559, + "grad_norm": 1.08585564810458, + "learning_rate": 0.003, + "loss": 4.072, + "step": 25590 + }, + { + "epoch": 0.25591, + "grad_norm": 1.1459790208448541, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 25591 + }, + { + "epoch": 0.25592, + "grad_norm": 1.1348511355590987, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 25592 + }, + { + "epoch": 0.25593, + "grad_norm": 1.11597799263945, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 25593 + }, + { + "epoch": 0.25594, + "grad_norm": 0.8790866857208846, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 25594 + }, + { + "epoch": 0.25595, + "grad_norm": 0.7033758565960845, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 25595 + }, + { + "epoch": 0.25596, + "grad_norm": 0.949021125343743, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 25596 + }, + { + "epoch": 0.25597, + "grad_norm": 1.3027905815956455, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 25597 + }, + { + "epoch": 0.25598, + "grad_norm": 0.9835865605696512, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 25598 + }, + { + "epoch": 0.25599, + "grad_norm": 1.0651791014691014, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 25599 + }, + { + "epoch": 0.256, + "grad_norm": 1.060825594954697, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 25600 + }, + { + "epoch": 0.25601, + "grad_norm": 1.0928264858214258, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 25601 + }, + { + "epoch": 0.25602, + "grad_norm": 0.9833200670909624, + "learning_rate": 0.003, + "loss": 4.054, + "step": 25602 + }, + { + "epoch": 0.25603, + "grad_norm": 0.8989624465492014, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 25603 + }, + { + "epoch": 0.25604, + "grad_norm": 0.8981293760190842, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 25604 + }, + { + "epoch": 0.25605, + "grad_norm": 1.1996445590370604, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 25605 + }, + { + "epoch": 0.25606, + "grad_norm": 1.0349436819530429, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 25606 + }, + { + "epoch": 0.25607, + "grad_norm": 1.0279348308548644, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 25607 + }, + { + "epoch": 0.25608, + "grad_norm": 1.201050622155861, + "learning_rate": 0.003, + "loss": 4.059, + "step": 25608 + }, + { + "epoch": 0.25609, + "grad_norm": 1.1005225172404622, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 25609 + }, + { + "epoch": 0.2561, + "grad_norm": 1.0833798043362903, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 25610 + }, + { + "epoch": 0.25611, + "grad_norm": 1.0077969894837762, + "learning_rate": 0.003, + "loss": 4.062, + "step": 25611 + }, + { + "epoch": 0.25612, + "grad_norm": 1.046088636709356, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 25612 + }, + { + "epoch": 0.25613, + "grad_norm": 1.1868145207600966, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 25613 + }, + { + "epoch": 0.25614, + "grad_norm": 1.022617904150116, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 25614 + }, + { + "epoch": 0.25615, + "grad_norm": 1.097832524026796, + "learning_rate": 0.003, + "loss": 4.086, + "step": 25615 + }, + { + "epoch": 0.25616, + "grad_norm": 1.182952146065001, + "learning_rate": 0.003, + "loss": 4.0927, + "step": 25616 + }, + { + "epoch": 0.25617, + "grad_norm": 0.9981533188474536, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 25617 + }, + { + "epoch": 0.25618, + "grad_norm": 1.3194558927422124, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 25618 + }, + { + "epoch": 0.25619, + "grad_norm": 0.8787526209689945, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 25619 + }, + { + "epoch": 0.2562, + "grad_norm": 1.1251294592160426, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 25620 + }, + { + "epoch": 0.25621, + "grad_norm": 1.2534617974366615, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 25621 + }, + { + "epoch": 0.25622, + "grad_norm": 1.1917942843929672, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 25622 + }, + { + "epoch": 0.25623, + "grad_norm": 1.1475438090612515, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 25623 + }, + { + "epoch": 0.25624, + "grad_norm": 1.042200338839964, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 25624 + }, + { + "epoch": 0.25625, + "grad_norm": 1.3425817335225245, + "learning_rate": 0.003, + "loss": 4.0874, + "step": 25625 + }, + { + "epoch": 0.25626, + "grad_norm": 0.999418838104237, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 25626 + }, + { + "epoch": 0.25627, + "grad_norm": 1.2265077811656666, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 25627 + }, + { + "epoch": 0.25628, + "grad_norm": 0.8455311716645031, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 25628 + }, + { + "epoch": 0.25629, + "grad_norm": 0.9646969261984563, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 25629 + }, + { + "epoch": 0.2563, + "grad_norm": 1.2672256316502346, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 25630 + }, + { + "epoch": 0.25631, + "grad_norm": 0.8684332955581371, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 25631 + }, + { + "epoch": 0.25632, + "grad_norm": 0.8453480596427102, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 25632 + }, + { + "epoch": 0.25633, + "grad_norm": 0.8473640554836361, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 25633 + }, + { + "epoch": 0.25634, + "grad_norm": 1.1231730137670128, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 25634 + }, + { + "epoch": 0.25635, + "grad_norm": 0.9067697617297554, + "learning_rate": 0.003, + "loss": 4.053, + "step": 25635 + }, + { + "epoch": 0.25636, + "grad_norm": 0.8406121704992755, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 25636 + }, + { + "epoch": 0.25637, + "grad_norm": 0.8218084697224448, + "learning_rate": 0.003, + "loss": 4.077, + "step": 25637 + }, + { + "epoch": 0.25638, + "grad_norm": 1.088479080382667, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 25638 + }, + { + "epoch": 0.25639, + "grad_norm": 1.3840905465239537, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 25639 + }, + { + "epoch": 0.2564, + "grad_norm": 0.9829979670302257, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 25640 + }, + { + "epoch": 0.25641, + "grad_norm": 1.1917231138474744, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 25641 + }, + { + "epoch": 0.25642, + "grad_norm": 1.2224781266295026, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 25642 + }, + { + "epoch": 0.25643, + "grad_norm": 1.055943901279658, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 25643 + }, + { + "epoch": 0.25644, + "grad_norm": 1.0804273422712007, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 25644 + }, + { + "epoch": 0.25645, + "grad_norm": 1.0061554905080095, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 25645 + }, + { + "epoch": 0.25646, + "grad_norm": 1.2274736221116225, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 25646 + }, + { + "epoch": 0.25647, + "grad_norm": 0.8292761853966156, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 25647 + }, + { + "epoch": 0.25648, + "grad_norm": 0.8847069571851746, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 25648 + }, + { + "epoch": 0.25649, + "grad_norm": 1.0469260962626792, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 25649 + }, + { + "epoch": 0.2565, + "grad_norm": 1.036796264396191, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 25650 + }, + { + "epoch": 0.25651, + "grad_norm": 1.1341392878315593, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 25651 + }, + { + "epoch": 0.25652, + "grad_norm": 1.1749457507162533, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 25652 + }, + { + "epoch": 0.25653, + "grad_norm": 1.1055635716856005, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 25653 + }, + { + "epoch": 0.25654, + "grad_norm": 1.134132181262467, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 25654 + }, + { + "epoch": 0.25655, + "grad_norm": 0.8813215036702801, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 25655 + }, + { + "epoch": 0.25656, + "grad_norm": 0.8075918092176976, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 25656 + }, + { + "epoch": 0.25657, + "grad_norm": 0.9895940696632407, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 25657 + }, + { + "epoch": 0.25658, + "grad_norm": 1.3140291303752103, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 25658 + }, + { + "epoch": 0.25659, + "grad_norm": 0.902165505026691, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 25659 + }, + { + "epoch": 0.2566, + "grad_norm": 1.042627450446038, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 25660 + }, + { + "epoch": 0.25661, + "grad_norm": 1.0857723245220965, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 25661 + }, + { + "epoch": 0.25662, + "grad_norm": 1.2369933404277502, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 25662 + }, + { + "epoch": 0.25663, + "grad_norm": 0.9510657975502086, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 25663 + }, + { + "epoch": 0.25664, + "grad_norm": 1.1025910382381616, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 25664 + }, + { + "epoch": 0.25665, + "grad_norm": 1.3097344128041744, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 25665 + }, + { + "epoch": 0.25666, + "grad_norm": 1.0832671859260379, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 25666 + }, + { + "epoch": 0.25667, + "grad_norm": 0.9732624216566056, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 25667 + }, + { + "epoch": 0.25668, + "grad_norm": 1.0245853950801067, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 25668 + }, + { + "epoch": 0.25669, + "grad_norm": 1.1662876351878786, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 25669 + }, + { + "epoch": 0.2567, + "grad_norm": 0.9375907854172567, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 25670 + }, + { + "epoch": 0.25671, + "grad_norm": 1.212697529849734, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 25671 + }, + { + "epoch": 0.25672, + "grad_norm": 0.9632536577982578, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 25672 + }, + { + "epoch": 0.25673, + "grad_norm": 1.252990475811103, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 25673 + }, + { + "epoch": 0.25674, + "grad_norm": 1.318341121221907, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 25674 + }, + { + "epoch": 0.25675, + "grad_norm": 0.8543079458779989, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 25675 + }, + { + "epoch": 0.25676, + "grad_norm": 0.9979555127375588, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 25676 + }, + { + "epoch": 0.25677, + "grad_norm": 1.0132514470511327, + "learning_rate": 0.003, + "loss": 4.0934, + "step": 25677 + }, + { + "epoch": 0.25678, + "grad_norm": 1.180098645271529, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 25678 + }, + { + "epoch": 0.25679, + "grad_norm": 1.167375945197378, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 25679 + }, + { + "epoch": 0.2568, + "grad_norm": 1.311548812122579, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 25680 + }, + { + "epoch": 0.25681, + "grad_norm": 0.9542106898459158, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 25681 + }, + { + "epoch": 0.25682, + "grad_norm": 1.2529067630655062, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 25682 + }, + { + "epoch": 0.25683, + "grad_norm": 1.0549529134508324, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 25683 + }, + { + "epoch": 0.25684, + "grad_norm": 1.0807856857118523, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 25684 + }, + { + "epoch": 0.25685, + "grad_norm": 0.9705384544914369, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 25685 + }, + { + "epoch": 0.25686, + "grad_norm": 1.0258260514353967, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 25686 + }, + { + "epoch": 0.25687, + "grad_norm": 1.1277146015939647, + "learning_rate": 0.003, + "loss": 4.038, + "step": 25687 + }, + { + "epoch": 0.25688, + "grad_norm": 0.9710026595888851, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 25688 + }, + { + "epoch": 0.25689, + "grad_norm": 1.4187541192173552, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 25689 + }, + { + "epoch": 0.2569, + "grad_norm": 1.0346008064267, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 25690 + }, + { + "epoch": 0.25691, + "grad_norm": 1.1754636560542622, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 25691 + }, + { + "epoch": 0.25692, + "grad_norm": 1.0623338513940137, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 25692 + }, + { + "epoch": 0.25693, + "grad_norm": 1.0255652230394081, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 25693 + }, + { + "epoch": 0.25694, + "grad_norm": 1.1396136891027056, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 25694 + }, + { + "epoch": 0.25695, + "grad_norm": 0.9671932926165683, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 25695 + }, + { + "epoch": 0.25696, + "grad_norm": 1.2167843000661807, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 25696 + }, + { + "epoch": 0.25697, + "grad_norm": 0.8799353872950394, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 25697 + }, + { + "epoch": 0.25698, + "grad_norm": 1.0630586231773005, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 25698 + }, + { + "epoch": 0.25699, + "grad_norm": 0.9862715096399022, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 25699 + }, + { + "epoch": 0.257, + "grad_norm": 1.3067841809072716, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 25700 + }, + { + "epoch": 0.25701, + "grad_norm": 1.0981013071842665, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 25701 + }, + { + "epoch": 0.25702, + "grad_norm": 0.9838258063611638, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 25702 + }, + { + "epoch": 0.25703, + "grad_norm": 1.0617215041651102, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 25703 + }, + { + "epoch": 0.25704, + "grad_norm": 0.9880309961531575, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 25704 + }, + { + "epoch": 0.25705, + "grad_norm": 1.1057822786404978, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 25705 + }, + { + "epoch": 0.25706, + "grad_norm": 0.9967156900642223, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 25706 + }, + { + "epoch": 0.25707, + "grad_norm": 1.0211234829691738, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 25707 + }, + { + "epoch": 0.25708, + "grad_norm": 0.9656761652444358, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 25708 + }, + { + "epoch": 0.25709, + "grad_norm": 1.2428248193459799, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 25709 + }, + { + "epoch": 0.2571, + "grad_norm": 1.1438943942224868, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 25710 + }, + { + "epoch": 0.25711, + "grad_norm": 0.9341758661731504, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 25711 + }, + { + "epoch": 0.25712, + "grad_norm": 1.136311636191151, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 25712 + }, + { + "epoch": 0.25713, + "grad_norm": 1.1514316897467902, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 25713 + }, + { + "epoch": 0.25714, + "grad_norm": 1.0856397185719937, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 25714 + }, + { + "epoch": 0.25715, + "grad_norm": 1.080609811636853, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 25715 + }, + { + "epoch": 0.25716, + "grad_norm": 1.0064742930883364, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 25716 + }, + { + "epoch": 0.25717, + "grad_norm": 1.018893973710554, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 25717 + }, + { + "epoch": 0.25718, + "grad_norm": 1.1457648137531307, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 25718 + }, + { + "epoch": 0.25719, + "grad_norm": 0.9906475984637149, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 25719 + }, + { + "epoch": 0.2572, + "grad_norm": 1.0666500310613463, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 25720 + }, + { + "epoch": 0.25721, + "grad_norm": 1.3220289101297236, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 25721 + }, + { + "epoch": 0.25722, + "grad_norm": 1.0727271305894006, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 25722 + }, + { + "epoch": 0.25723, + "grad_norm": 1.1055765047810278, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 25723 + }, + { + "epoch": 0.25724, + "grad_norm": 1.2612361951954516, + "learning_rate": 0.003, + "loss": 4.05, + "step": 25724 + }, + { + "epoch": 0.25725, + "grad_norm": 1.0849331172022771, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 25725 + }, + { + "epoch": 0.25726, + "grad_norm": 1.184564074562522, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 25726 + }, + { + "epoch": 0.25727, + "grad_norm": 1.0460166018671333, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 25727 + }, + { + "epoch": 0.25728, + "grad_norm": 1.0951619301801157, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 25728 + }, + { + "epoch": 0.25729, + "grad_norm": 1.0342489389085094, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 25729 + }, + { + "epoch": 0.2573, + "grad_norm": 1.1941903998755041, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 25730 + }, + { + "epoch": 0.25731, + "grad_norm": 0.9826402379062007, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 25731 + }, + { + "epoch": 0.25732, + "grad_norm": 1.1573958168580842, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 25732 + }, + { + "epoch": 0.25733, + "grad_norm": 0.8450128706162078, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 25733 + }, + { + "epoch": 0.25734, + "grad_norm": 0.9946793246297824, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 25734 + }, + { + "epoch": 0.25735, + "grad_norm": 1.5114566508462421, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 25735 + }, + { + "epoch": 0.25736, + "grad_norm": 0.944883796130798, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 25736 + }, + { + "epoch": 0.25737, + "grad_norm": 0.9518025486915036, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 25737 + }, + { + "epoch": 0.25738, + "grad_norm": 1.1294867464773342, + "learning_rate": 0.003, + "loss": 4.005, + "step": 25738 + }, + { + "epoch": 0.25739, + "grad_norm": 1.0714365472559562, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 25739 + }, + { + "epoch": 0.2574, + "grad_norm": 1.0853569301916408, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 25740 + }, + { + "epoch": 0.25741, + "grad_norm": 1.0472064947516986, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 25741 + }, + { + "epoch": 0.25742, + "grad_norm": 0.9603539628113091, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 25742 + }, + { + "epoch": 0.25743, + "grad_norm": 0.9492848725555092, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 25743 + }, + { + "epoch": 0.25744, + "grad_norm": 1.1115876002904876, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 25744 + }, + { + "epoch": 0.25745, + "grad_norm": 1.0855057401587855, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 25745 + }, + { + "epoch": 0.25746, + "grad_norm": 1.095623168222791, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 25746 + }, + { + "epoch": 0.25747, + "grad_norm": 0.9614252073036532, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 25747 + }, + { + "epoch": 0.25748, + "grad_norm": 1.039013546327199, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 25748 + }, + { + "epoch": 0.25749, + "grad_norm": 1.1199341458962166, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 25749 + }, + { + "epoch": 0.2575, + "grad_norm": 0.7836595266128187, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 25750 + }, + { + "epoch": 0.25751, + "grad_norm": 0.7534084268809405, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 25751 + }, + { + "epoch": 0.25752, + "grad_norm": 0.9183802861915605, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 25752 + }, + { + "epoch": 0.25753, + "grad_norm": 1.129543720879673, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 25753 + }, + { + "epoch": 0.25754, + "grad_norm": 1.00594320052028, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 25754 + }, + { + "epoch": 0.25755, + "grad_norm": 1.4489191784148938, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 25755 + }, + { + "epoch": 0.25756, + "grad_norm": 0.8011840022367472, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 25756 + }, + { + "epoch": 0.25757, + "grad_norm": 0.7650790673120871, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 25757 + }, + { + "epoch": 0.25758, + "grad_norm": 0.9738132811008723, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 25758 + }, + { + "epoch": 0.25759, + "grad_norm": 1.1804563736088935, + "learning_rate": 0.003, + "loss": 4.048, + "step": 25759 + }, + { + "epoch": 0.2576, + "grad_norm": 0.8523297012721865, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 25760 + }, + { + "epoch": 0.25761, + "grad_norm": 0.9130033782681095, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 25761 + }, + { + "epoch": 0.25762, + "grad_norm": 1.525033916098443, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 25762 + }, + { + "epoch": 0.25763, + "grad_norm": 0.9801472108734197, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 25763 + }, + { + "epoch": 0.25764, + "grad_norm": 1.152849281564869, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 25764 + }, + { + "epoch": 0.25765, + "grad_norm": 1.1982882781894448, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 25765 + }, + { + "epoch": 0.25766, + "grad_norm": 0.8691111815165771, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 25766 + }, + { + "epoch": 0.25767, + "grad_norm": 0.9919674391369467, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 25767 + }, + { + "epoch": 0.25768, + "grad_norm": 1.1586583679562736, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 25768 + }, + { + "epoch": 0.25769, + "grad_norm": 1.009931470866332, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 25769 + }, + { + "epoch": 0.2577, + "grad_norm": 1.0222385749068303, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 25770 + }, + { + "epoch": 0.25771, + "grad_norm": 1.1862074923849488, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 25771 + }, + { + "epoch": 0.25772, + "grad_norm": 0.9249618412560741, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 25772 + }, + { + "epoch": 0.25773, + "grad_norm": 1.0608559753797289, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 25773 + }, + { + "epoch": 0.25774, + "grad_norm": 1.0772899971170375, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 25774 + }, + { + "epoch": 0.25775, + "grad_norm": 1.2445530321352858, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 25775 + }, + { + "epoch": 0.25776, + "grad_norm": 0.8725864813470995, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 25776 + }, + { + "epoch": 0.25777, + "grad_norm": 0.8846941012289384, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 25777 + }, + { + "epoch": 0.25778, + "grad_norm": 1.061765836610767, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 25778 + }, + { + "epoch": 0.25779, + "grad_norm": 1.4286490243603118, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 25779 + }, + { + "epoch": 0.2578, + "grad_norm": 1.2364683058547146, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 25780 + }, + { + "epoch": 0.25781, + "grad_norm": 0.8072780411223044, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 25781 + }, + { + "epoch": 0.25782, + "grad_norm": 0.9406067807665851, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 25782 + }, + { + "epoch": 0.25783, + "grad_norm": 1.0983483282107553, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 25783 + }, + { + "epoch": 0.25784, + "grad_norm": 1.1012472348253204, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 25784 + }, + { + "epoch": 0.25785, + "grad_norm": 0.9930393414159552, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 25785 + }, + { + "epoch": 0.25786, + "grad_norm": 1.1254170836041697, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 25786 + }, + { + "epoch": 0.25787, + "grad_norm": 0.9371880830071575, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 25787 + }, + { + "epoch": 0.25788, + "grad_norm": 0.9416403870839284, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 25788 + }, + { + "epoch": 0.25789, + "grad_norm": 1.1222246455586509, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 25789 + }, + { + "epoch": 0.2579, + "grad_norm": 1.1021651051643118, + "learning_rate": 0.003, + "loss": 4.053, + "step": 25790 + }, + { + "epoch": 0.25791, + "grad_norm": 1.0031339107140536, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 25791 + }, + { + "epoch": 0.25792, + "grad_norm": 1.049963522371582, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 25792 + }, + { + "epoch": 0.25793, + "grad_norm": 0.9473244802400285, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 25793 + }, + { + "epoch": 0.25794, + "grad_norm": 0.9235533291124012, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 25794 + }, + { + "epoch": 0.25795, + "grad_norm": 1.0256238843941574, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 25795 + }, + { + "epoch": 0.25796, + "grad_norm": 1.1630431470537288, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 25796 + }, + { + "epoch": 0.25797, + "grad_norm": 1.1483519684165782, + "learning_rate": 0.003, + "loss": 4.05, + "step": 25797 + }, + { + "epoch": 0.25798, + "grad_norm": 1.000145540105286, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 25798 + }, + { + "epoch": 0.25799, + "grad_norm": 1.0943665458748868, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 25799 + }, + { + "epoch": 0.258, + "grad_norm": 1.28066986166805, + "learning_rate": 0.003, + "loss": 4.062, + "step": 25800 + }, + { + "epoch": 0.25801, + "grad_norm": 0.9341328979357717, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 25801 + }, + { + "epoch": 0.25802, + "grad_norm": 1.0096608164600553, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 25802 + }, + { + "epoch": 0.25803, + "grad_norm": 1.2372430807747388, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 25803 + }, + { + "epoch": 0.25804, + "grad_norm": 1.2997199119298262, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 25804 + }, + { + "epoch": 0.25805, + "grad_norm": 0.9935335197002388, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 25805 + }, + { + "epoch": 0.25806, + "grad_norm": 1.2558848050139557, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 25806 + }, + { + "epoch": 0.25807, + "grad_norm": 0.975991886611135, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 25807 + }, + { + "epoch": 0.25808, + "grad_norm": 1.0892950073786667, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 25808 + }, + { + "epoch": 0.25809, + "grad_norm": 1.016620135390295, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 25809 + }, + { + "epoch": 0.2581, + "grad_norm": 1.0869226378712127, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 25810 + }, + { + "epoch": 0.25811, + "grad_norm": 1.0518858747990107, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 25811 + }, + { + "epoch": 0.25812, + "grad_norm": 1.2574930086464873, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 25812 + }, + { + "epoch": 0.25813, + "grad_norm": 0.9636585346736337, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 25813 + }, + { + "epoch": 0.25814, + "grad_norm": 1.1581133003778774, + "learning_rate": 0.003, + "loss": 4.0898, + "step": 25814 + }, + { + "epoch": 0.25815, + "grad_norm": 0.9727219165425257, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 25815 + }, + { + "epoch": 0.25816, + "grad_norm": 1.0083981142373024, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 25816 + }, + { + "epoch": 0.25817, + "grad_norm": 1.0430244122453551, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 25817 + }, + { + "epoch": 0.25818, + "grad_norm": 1.0020569363404175, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 25818 + }, + { + "epoch": 0.25819, + "grad_norm": 1.1012380261803516, + "learning_rate": 0.003, + "loss": 4.0929, + "step": 25819 + }, + { + "epoch": 0.2582, + "grad_norm": 0.8947681777021732, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 25820 + }, + { + "epoch": 0.25821, + "grad_norm": 0.9726545902090727, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 25821 + }, + { + "epoch": 0.25822, + "grad_norm": 1.043757698524712, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 25822 + }, + { + "epoch": 0.25823, + "grad_norm": 1.2849745726133472, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 25823 + }, + { + "epoch": 0.25824, + "grad_norm": 1.3716293058350026, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 25824 + }, + { + "epoch": 0.25825, + "grad_norm": 0.929846117280229, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 25825 + }, + { + "epoch": 0.25826, + "grad_norm": 1.0147093870496144, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 25826 + }, + { + "epoch": 0.25827, + "grad_norm": 1.1853253864232858, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 25827 + }, + { + "epoch": 0.25828, + "grad_norm": 1.0689975955639959, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 25828 + }, + { + "epoch": 0.25829, + "grad_norm": 1.0807136233142007, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 25829 + }, + { + "epoch": 0.2583, + "grad_norm": 1.1450206055141208, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 25830 + }, + { + "epoch": 0.25831, + "grad_norm": 1.1175813376192265, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 25831 + }, + { + "epoch": 0.25832, + "grad_norm": 1.0896329073788273, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 25832 + }, + { + "epoch": 0.25833, + "grad_norm": 1.133154781283191, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 25833 + }, + { + "epoch": 0.25834, + "grad_norm": 1.003530478078473, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 25834 + }, + { + "epoch": 0.25835, + "grad_norm": 1.0349218186735323, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 25835 + }, + { + "epoch": 0.25836, + "grad_norm": 1.2155706305687488, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 25836 + }, + { + "epoch": 0.25837, + "grad_norm": 0.9094293169030665, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 25837 + }, + { + "epoch": 0.25838, + "grad_norm": 1.0100508067003469, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 25838 + }, + { + "epoch": 0.25839, + "grad_norm": 1.0402577876647603, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 25839 + }, + { + "epoch": 0.2584, + "grad_norm": 1.0569635968648334, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 25840 + }, + { + "epoch": 0.25841, + "grad_norm": 1.0366760155712904, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 25841 + }, + { + "epoch": 0.25842, + "grad_norm": 0.9698411359193844, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 25842 + }, + { + "epoch": 0.25843, + "grad_norm": 1.158329921732603, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 25843 + }, + { + "epoch": 0.25844, + "grad_norm": 1.119152338519284, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 25844 + }, + { + "epoch": 0.25845, + "grad_norm": 0.9691009041415688, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 25845 + }, + { + "epoch": 0.25846, + "grad_norm": 1.0048052733602666, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 25846 + }, + { + "epoch": 0.25847, + "grad_norm": 1.3269970913473077, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 25847 + }, + { + "epoch": 0.25848, + "grad_norm": 1.0953166148832563, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 25848 + }, + { + "epoch": 0.25849, + "grad_norm": 1.0109241721418762, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 25849 + }, + { + "epoch": 0.2585, + "grad_norm": 1.086209537029173, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 25850 + }, + { + "epoch": 0.25851, + "grad_norm": 1.0413841483290815, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 25851 + }, + { + "epoch": 0.25852, + "grad_norm": 1.0683101719199635, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 25852 + }, + { + "epoch": 0.25853, + "grad_norm": 1.0084206511388236, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 25853 + }, + { + "epoch": 0.25854, + "grad_norm": 1.216733407200061, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 25854 + }, + { + "epoch": 0.25855, + "grad_norm": 0.9167964454900894, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 25855 + }, + { + "epoch": 0.25856, + "grad_norm": 1.0622714485513325, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 25856 + }, + { + "epoch": 0.25857, + "grad_norm": 1.1482196268397125, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 25857 + }, + { + "epoch": 0.25858, + "grad_norm": 1.1205970498947753, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 25858 + }, + { + "epoch": 0.25859, + "grad_norm": 1.0414324362334058, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 25859 + }, + { + "epoch": 0.2586, + "grad_norm": 0.9446889050465628, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 25860 + }, + { + "epoch": 0.25861, + "grad_norm": 0.9290512834409133, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 25861 + }, + { + "epoch": 0.25862, + "grad_norm": 1.003622616687256, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 25862 + }, + { + "epoch": 0.25863, + "grad_norm": 1.084110657013388, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 25863 + }, + { + "epoch": 0.25864, + "grad_norm": 0.9671642479654068, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 25864 + }, + { + "epoch": 0.25865, + "grad_norm": 1.0716972366814943, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 25865 + }, + { + "epoch": 0.25866, + "grad_norm": 1.065533784126673, + "learning_rate": 0.003, + "loss": 4.057, + "step": 25866 + }, + { + "epoch": 0.25867, + "grad_norm": 1.362529191607796, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 25867 + }, + { + "epoch": 0.25868, + "grad_norm": 0.82525893122982, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 25868 + }, + { + "epoch": 0.25869, + "grad_norm": 0.9726805629663962, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 25869 + }, + { + "epoch": 0.2587, + "grad_norm": 1.3037162003738316, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 25870 + }, + { + "epoch": 0.25871, + "grad_norm": 1.3373934941176486, + "learning_rate": 0.003, + "loss": 4.032, + "step": 25871 + }, + { + "epoch": 0.25872, + "grad_norm": 0.7352924692955987, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 25872 + }, + { + "epoch": 0.25873, + "grad_norm": 0.7951117405234358, + "learning_rate": 0.003, + "loss": 4.026, + "step": 25873 + }, + { + "epoch": 0.25874, + "grad_norm": 0.9134721830161254, + "learning_rate": 0.003, + "loss": 4.044, + "step": 25874 + }, + { + "epoch": 0.25875, + "grad_norm": 1.1029733004846642, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 25875 + }, + { + "epoch": 0.25876, + "grad_norm": 1.1557000660305765, + "learning_rate": 0.003, + "loss": 4.058, + "step": 25876 + }, + { + "epoch": 0.25877, + "grad_norm": 1.0455265150380357, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 25877 + }, + { + "epoch": 0.25878, + "grad_norm": 1.2665574795434675, + "learning_rate": 0.003, + "loss": 4.0868, + "step": 25878 + }, + { + "epoch": 0.25879, + "grad_norm": 0.8410159143473711, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 25879 + }, + { + "epoch": 0.2588, + "grad_norm": 0.8953081972855947, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 25880 + }, + { + "epoch": 0.25881, + "grad_norm": 1.0160249980479927, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 25881 + }, + { + "epoch": 0.25882, + "grad_norm": 0.9893718556831915, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 25882 + }, + { + "epoch": 0.25883, + "grad_norm": 1.0507087747956299, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 25883 + }, + { + "epoch": 0.25884, + "grad_norm": 1.1340197843720985, + "learning_rate": 0.003, + "loss": 4.034, + "step": 25884 + }, + { + "epoch": 0.25885, + "grad_norm": 1.2204653449485965, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 25885 + }, + { + "epoch": 0.25886, + "grad_norm": 1.1604463120357653, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 25886 + }, + { + "epoch": 0.25887, + "grad_norm": 1.2651787181245189, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 25887 + }, + { + "epoch": 0.25888, + "grad_norm": 0.7928120061522667, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 25888 + }, + { + "epoch": 0.25889, + "grad_norm": 0.8068979629756909, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 25889 + }, + { + "epoch": 0.2589, + "grad_norm": 0.9981109918051627, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 25890 + }, + { + "epoch": 0.25891, + "grad_norm": 1.171197087454086, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 25891 + }, + { + "epoch": 0.25892, + "grad_norm": 0.9758141516070132, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 25892 + }, + { + "epoch": 0.25893, + "grad_norm": 1.1528464500334674, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 25893 + }, + { + "epoch": 0.25894, + "grad_norm": 1.2504304757220557, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 25894 + }, + { + "epoch": 0.25895, + "grad_norm": 0.9493002845718653, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 25895 + }, + { + "epoch": 0.25896, + "grad_norm": 1.006169334688126, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 25896 + }, + { + "epoch": 0.25897, + "grad_norm": 1.296493321383995, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 25897 + }, + { + "epoch": 0.25898, + "grad_norm": 1.1076520050809249, + "learning_rate": 0.003, + "loss": 4.042, + "step": 25898 + }, + { + "epoch": 0.25899, + "grad_norm": 0.9673460965612726, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 25899 + }, + { + "epoch": 0.259, + "grad_norm": 0.9363715731507622, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 25900 + }, + { + "epoch": 0.25901, + "grad_norm": 1.0987222925815163, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 25901 + }, + { + "epoch": 0.25902, + "grad_norm": 1.1265024147257423, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 25902 + }, + { + "epoch": 0.25903, + "grad_norm": 0.8948963777112059, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 25903 + }, + { + "epoch": 0.25904, + "grad_norm": 0.9553118414940905, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 25904 + }, + { + "epoch": 0.25905, + "grad_norm": 1.1674181147951026, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 25905 + }, + { + "epoch": 0.25906, + "grad_norm": 0.8671538842582174, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 25906 + }, + { + "epoch": 0.25907, + "grad_norm": 0.78375792678563, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 25907 + }, + { + "epoch": 0.25908, + "grad_norm": 0.7581315034072473, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 25908 + }, + { + "epoch": 0.25909, + "grad_norm": 0.9791830681883529, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 25909 + }, + { + "epoch": 0.2591, + "grad_norm": 1.3143119499827565, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 25910 + }, + { + "epoch": 0.25911, + "grad_norm": 0.79653804389322, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 25911 + }, + { + "epoch": 0.25912, + "grad_norm": 0.8538878992821394, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 25912 + }, + { + "epoch": 0.25913, + "grad_norm": 0.954057732648047, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 25913 + }, + { + "epoch": 0.25914, + "grad_norm": 1.2223918277127328, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 25914 + }, + { + "epoch": 0.25915, + "grad_norm": 1.1894375963111778, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 25915 + }, + { + "epoch": 0.25916, + "grad_norm": 1.164275995123677, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 25916 + }, + { + "epoch": 0.25917, + "grad_norm": 1.0185763343721514, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 25917 + }, + { + "epoch": 0.25918, + "grad_norm": 1.0927216443127767, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 25918 + }, + { + "epoch": 0.25919, + "grad_norm": 1.1961635169719433, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 25919 + }, + { + "epoch": 0.2592, + "grad_norm": 1.0499280195877978, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 25920 + }, + { + "epoch": 0.25921, + "grad_norm": 1.08673267101805, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 25921 + }, + { + "epoch": 0.25922, + "grad_norm": 0.8736959164273458, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 25922 + }, + { + "epoch": 0.25923, + "grad_norm": 0.9440196897215265, + "learning_rate": 0.003, + "loss": 4.053, + "step": 25923 + }, + { + "epoch": 0.25924, + "grad_norm": 1.0300402836123204, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 25924 + }, + { + "epoch": 0.25925, + "grad_norm": 1.2345017096508024, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 25925 + }, + { + "epoch": 0.25926, + "grad_norm": 1.2428314729538215, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 25926 + }, + { + "epoch": 0.25927, + "grad_norm": 1.143153565160509, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 25927 + }, + { + "epoch": 0.25928, + "grad_norm": 1.0051060060934671, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 25928 + }, + { + "epoch": 0.25929, + "grad_norm": 1.305677658352798, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 25929 + }, + { + "epoch": 0.2593, + "grad_norm": 1.0367380904487091, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 25930 + }, + { + "epoch": 0.25931, + "grad_norm": 1.1476097268046723, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 25931 + }, + { + "epoch": 0.25932, + "grad_norm": 1.0367257817040065, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 25932 + }, + { + "epoch": 0.25933, + "grad_norm": 0.9431340047755115, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 25933 + }, + { + "epoch": 0.25934, + "grad_norm": 1.2951472055219038, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 25934 + }, + { + "epoch": 0.25935, + "grad_norm": 0.9077521942174333, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 25935 + }, + { + "epoch": 0.25936, + "grad_norm": 0.7802502331716366, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 25936 + }, + { + "epoch": 0.25937, + "grad_norm": 0.7122817667500361, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 25937 + }, + { + "epoch": 0.25938, + "grad_norm": 0.847142240838496, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 25938 + }, + { + "epoch": 0.25939, + "grad_norm": 1.1085673008097567, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 25939 + }, + { + "epoch": 0.2594, + "grad_norm": 1.2143421164946004, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 25940 + }, + { + "epoch": 0.25941, + "grad_norm": 0.907581035900119, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 25941 + }, + { + "epoch": 0.25942, + "grad_norm": 1.1024258140355936, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 25942 + }, + { + "epoch": 0.25943, + "grad_norm": 1.0998851571906874, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 25943 + }, + { + "epoch": 0.25944, + "grad_norm": 1.0722166616369602, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 25944 + }, + { + "epoch": 0.25945, + "grad_norm": 1.0761659436234488, + "learning_rate": 0.003, + "loss": 4.058, + "step": 25945 + }, + { + "epoch": 0.25946, + "grad_norm": 0.9584065696251614, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 25946 + }, + { + "epoch": 0.25947, + "grad_norm": 1.215005195156983, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 25947 + }, + { + "epoch": 0.25948, + "grad_norm": 1.0845617430290317, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 25948 + }, + { + "epoch": 0.25949, + "grad_norm": 0.9895410232401181, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 25949 + }, + { + "epoch": 0.2595, + "grad_norm": 1.238158763250355, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 25950 + }, + { + "epoch": 0.25951, + "grad_norm": 1.0347449790524341, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 25951 + }, + { + "epoch": 0.25952, + "grad_norm": 1.1194775632996208, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 25952 + }, + { + "epoch": 0.25953, + "grad_norm": 1.0900002650345249, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 25953 + }, + { + "epoch": 0.25954, + "grad_norm": 1.0972652213534488, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 25954 + }, + { + "epoch": 0.25955, + "grad_norm": 1.2482208100783485, + "learning_rate": 0.003, + "loss": 4.067, + "step": 25955 + }, + { + "epoch": 0.25956, + "grad_norm": 1.0849795310607224, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 25956 + }, + { + "epoch": 0.25957, + "grad_norm": 1.2347054115708929, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 25957 + }, + { + "epoch": 0.25958, + "grad_norm": 1.0193056626550547, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 25958 + }, + { + "epoch": 0.25959, + "grad_norm": 1.163202476850608, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 25959 + }, + { + "epoch": 0.2596, + "grad_norm": 1.0532366625802276, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 25960 + }, + { + "epoch": 0.25961, + "grad_norm": 1.1906005827442387, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 25961 + }, + { + "epoch": 0.25962, + "grad_norm": 0.9490531737892268, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 25962 + }, + { + "epoch": 0.25963, + "grad_norm": 1.202334505435022, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 25963 + }, + { + "epoch": 0.25964, + "grad_norm": 1.3716678933914364, + "learning_rate": 0.003, + "loss": 4.07, + "step": 25964 + }, + { + "epoch": 0.25965, + "grad_norm": 0.9475166140182826, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 25965 + }, + { + "epoch": 0.25966, + "grad_norm": 0.915666784661665, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 25966 + }, + { + "epoch": 0.25967, + "grad_norm": 1.1232731188451828, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 25967 + }, + { + "epoch": 0.25968, + "grad_norm": 0.9631946475978255, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 25968 + }, + { + "epoch": 0.25969, + "grad_norm": 1.192963823810518, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 25969 + }, + { + "epoch": 0.2597, + "grad_norm": 1.01556307760195, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 25970 + }, + { + "epoch": 0.25971, + "grad_norm": 1.1306462490520974, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 25971 + }, + { + "epoch": 0.25972, + "grad_norm": 1.0283034517424707, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 25972 + }, + { + "epoch": 0.25973, + "grad_norm": 1.1551696184278282, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 25973 + }, + { + "epoch": 0.25974, + "grad_norm": 0.9529895801856582, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 25974 + }, + { + "epoch": 0.25975, + "grad_norm": 1.0412784330948726, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 25975 + }, + { + "epoch": 0.25976, + "grad_norm": 1.1943944927846624, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 25976 + }, + { + "epoch": 0.25977, + "grad_norm": 1.1116855265632075, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 25977 + }, + { + "epoch": 0.25978, + "grad_norm": 1.104044659569365, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 25978 + }, + { + "epoch": 0.25979, + "grad_norm": 1.1214255122790668, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 25979 + }, + { + "epoch": 0.2598, + "grad_norm": 1.107395324384743, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 25980 + }, + { + "epoch": 0.25981, + "grad_norm": 1.1871155397022457, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 25981 + }, + { + "epoch": 0.25982, + "grad_norm": 0.8371816147214286, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 25982 + }, + { + "epoch": 0.25983, + "grad_norm": 0.8422698225973774, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 25983 + }, + { + "epoch": 0.25984, + "grad_norm": 1.0762669957284143, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 25984 + }, + { + "epoch": 0.25985, + "grad_norm": 1.2456900063220075, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 25985 + }, + { + "epoch": 0.25986, + "grad_norm": 0.9806056409661795, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 25986 + }, + { + "epoch": 0.25987, + "grad_norm": 1.3448706123126073, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 25987 + }, + { + "epoch": 0.25988, + "grad_norm": 0.8980485375180552, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 25988 + }, + { + "epoch": 0.25989, + "grad_norm": 0.8171432056234686, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 25989 + }, + { + "epoch": 0.2599, + "grad_norm": 1.0793062539074607, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 25990 + }, + { + "epoch": 0.25991, + "grad_norm": 1.4470175998051922, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 25991 + }, + { + "epoch": 0.25992, + "grad_norm": 0.9870359462551042, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 25992 + }, + { + "epoch": 0.25993, + "grad_norm": 1.0977684256829063, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 25993 + }, + { + "epoch": 0.25994, + "grad_norm": 1.1682323926035165, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 25994 + }, + { + "epoch": 0.25995, + "grad_norm": 1.1650233520687738, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 25995 + }, + { + "epoch": 0.25996, + "grad_norm": 1.0352098611862488, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 25996 + }, + { + "epoch": 0.25997, + "grad_norm": 0.9233543665303275, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 25997 + }, + { + "epoch": 0.25998, + "grad_norm": 0.953821714272774, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 25998 + }, + { + "epoch": 0.25999, + "grad_norm": 1.125727734606913, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 25999 + }, + { + "epoch": 0.26, + "grad_norm": 1.064511405800055, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 26000 + }, + { + "epoch": 0.26001, + "grad_norm": 1.1335621640500742, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 26001 + }, + { + "epoch": 0.26002, + "grad_norm": 1.1331678117809258, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 26002 + }, + { + "epoch": 0.26003, + "grad_norm": 1.0991188963888168, + "learning_rate": 0.003, + "loss": 4.042, + "step": 26003 + }, + { + "epoch": 0.26004, + "grad_norm": 0.9897318399202192, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 26004 + }, + { + "epoch": 0.26005, + "grad_norm": 1.1523666559622163, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 26005 + }, + { + "epoch": 0.26006, + "grad_norm": 1.0729570740157508, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 26006 + }, + { + "epoch": 0.26007, + "grad_norm": 1.394403373153391, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 26007 + }, + { + "epoch": 0.26008, + "grad_norm": 1.0482370026706231, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 26008 + }, + { + "epoch": 0.26009, + "grad_norm": 0.9624169538512305, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 26009 + }, + { + "epoch": 0.2601, + "grad_norm": 1.0999889441281097, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 26010 + }, + { + "epoch": 0.26011, + "grad_norm": 1.1338842216406582, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 26011 + }, + { + "epoch": 0.26012, + "grad_norm": 0.9978650237009042, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 26012 + }, + { + "epoch": 0.26013, + "grad_norm": 1.0800267843541236, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 26013 + }, + { + "epoch": 0.26014, + "grad_norm": 1.0528542429435046, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 26014 + }, + { + "epoch": 0.26015, + "grad_norm": 0.8881855794276029, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 26015 + }, + { + "epoch": 0.26016, + "grad_norm": 1.0669752612753836, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 26016 + }, + { + "epoch": 0.26017, + "grad_norm": 0.972722873087851, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 26017 + }, + { + "epoch": 0.26018, + "grad_norm": 1.024369453914771, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 26018 + }, + { + "epoch": 0.26019, + "grad_norm": 1.2218678997915073, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 26019 + }, + { + "epoch": 0.2602, + "grad_norm": 1.0798579414382983, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 26020 + }, + { + "epoch": 0.26021, + "grad_norm": 1.017724261981842, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 26021 + }, + { + "epoch": 0.26022, + "grad_norm": 1.0822523968773754, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 26022 + }, + { + "epoch": 0.26023, + "grad_norm": 1.1108427923768542, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 26023 + }, + { + "epoch": 0.26024, + "grad_norm": 1.0937477852080641, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 26024 + }, + { + "epoch": 0.26025, + "grad_norm": 1.1919537548123043, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 26025 + }, + { + "epoch": 0.26026, + "grad_norm": 1.0427225249194796, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 26026 + }, + { + "epoch": 0.26027, + "grad_norm": 1.2033863185479134, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 26027 + }, + { + "epoch": 0.26028, + "grad_norm": 0.9994186684830635, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 26028 + }, + { + "epoch": 0.26029, + "grad_norm": 1.0903589869210073, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 26029 + }, + { + "epoch": 0.2603, + "grad_norm": 1.1033935992256287, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 26030 + }, + { + "epoch": 0.26031, + "grad_norm": 1.0879832037239292, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 26031 + }, + { + "epoch": 0.26032, + "grad_norm": 0.9306196950044299, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 26032 + }, + { + "epoch": 0.26033, + "grad_norm": 0.9044262881775635, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 26033 + }, + { + "epoch": 0.26034, + "grad_norm": 0.947039458876025, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 26034 + }, + { + "epoch": 0.26035, + "grad_norm": 1.2620315582944313, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 26035 + }, + { + "epoch": 0.26036, + "grad_norm": 1.169182993741157, + "learning_rate": 0.003, + "loss": 4.049, + "step": 26036 + }, + { + "epoch": 0.26037, + "grad_norm": 1.056147386484613, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 26037 + }, + { + "epoch": 0.26038, + "grad_norm": 1.1253606594266699, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 26038 + }, + { + "epoch": 0.26039, + "grad_norm": 1.1074347561007272, + "learning_rate": 0.003, + "loss": 4.074, + "step": 26039 + }, + { + "epoch": 0.2604, + "grad_norm": 0.9520281602176537, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 26040 + }, + { + "epoch": 0.26041, + "grad_norm": 1.1108383555613228, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 26041 + }, + { + "epoch": 0.26042, + "grad_norm": 1.0058973308910966, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 26042 + }, + { + "epoch": 0.26043, + "grad_norm": 0.9816617473682364, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 26043 + }, + { + "epoch": 0.26044, + "grad_norm": 1.0820503846271143, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 26044 + }, + { + "epoch": 0.26045, + "grad_norm": 1.3469938538755892, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 26045 + }, + { + "epoch": 0.26046, + "grad_norm": 0.9810043499380554, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 26046 + }, + { + "epoch": 0.26047, + "grad_norm": 1.0013208938630371, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 26047 + }, + { + "epoch": 0.26048, + "grad_norm": 1.221282010842073, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 26048 + }, + { + "epoch": 0.26049, + "grad_norm": 1.06908315085462, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 26049 + }, + { + "epoch": 0.2605, + "grad_norm": 0.9801304526645832, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 26050 + }, + { + "epoch": 0.26051, + "grad_norm": 1.1733225505824922, + "learning_rate": 0.003, + "loss": 4.044, + "step": 26051 + }, + { + "epoch": 0.26052, + "grad_norm": 0.995321210940488, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 26052 + }, + { + "epoch": 0.26053, + "grad_norm": 0.9502192462993398, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 26053 + }, + { + "epoch": 0.26054, + "grad_norm": 0.955300941180278, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 26054 + }, + { + "epoch": 0.26055, + "grad_norm": 1.126253678154865, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 26055 + }, + { + "epoch": 0.26056, + "grad_norm": 1.06864186508917, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 26056 + }, + { + "epoch": 0.26057, + "grad_norm": 1.0247786519491184, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 26057 + }, + { + "epoch": 0.26058, + "grad_norm": 1.2190017172301697, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 26058 + }, + { + "epoch": 0.26059, + "grad_norm": 0.9887576283146983, + "learning_rate": 0.003, + "loss": 4.0878, + "step": 26059 + }, + { + "epoch": 0.2606, + "grad_norm": 1.342076973833636, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 26060 + }, + { + "epoch": 0.26061, + "grad_norm": 0.9295618084800656, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 26061 + }, + { + "epoch": 0.26062, + "grad_norm": 1.082385284454233, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 26062 + }, + { + "epoch": 0.26063, + "grad_norm": 1.074954195719047, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 26063 + }, + { + "epoch": 0.26064, + "grad_norm": 1.0476708971478284, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 26064 + }, + { + "epoch": 0.26065, + "grad_norm": 1.1750398779289213, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 26065 + }, + { + "epoch": 0.26066, + "grad_norm": 0.9742853380250971, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 26066 + }, + { + "epoch": 0.26067, + "grad_norm": 1.09233674422325, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 26067 + }, + { + "epoch": 0.26068, + "grad_norm": 1.2446423874060917, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 26068 + }, + { + "epoch": 0.26069, + "grad_norm": 1.0114986592502315, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 26069 + }, + { + "epoch": 0.2607, + "grad_norm": 1.1578515982517532, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 26070 + }, + { + "epoch": 0.26071, + "grad_norm": 0.9524399263040055, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 26071 + }, + { + "epoch": 0.26072, + "grad_norm": 1.0623270496543222, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 26072 + }, + { + "epoch": 0.26073, + "grad_norm": 1.1003151972265957, + "learning_rate": 0.003, + "loss": 4.048, + "step": 26073 + }, + { + "epoch": 0.26074, + "grad_norm": 0.9841962212049318, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 26074 + }, + { + "epoch": 0.26075, + "grad_norm": 1.14696349745626, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 26075 + }, + { + "epoch": 0.26076, + "grad_norm": 0.8719215728063716, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 26076 + }, + { + "epoch": 0.26077, + "grad_norm": 0.8541752835730702, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 26077 + }, + { + "epoch": 0.26078, + "grad_norm": 0.9101802938915563, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 26078 + }, + { + "epoch": 0.26079, + "grad_norm": 1.186982360248966, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 26079 + }, + { + "epoch": 0.2608, + "grad_norm": 1.5806067394331607, + "learning_rate": 0.003, + "loss": 4.1036, + "step": 26080 + }, + { + "epoch": 0.26081, + "grad_norm": 0.6148597504435358, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 26081 + }, + { + "epoch": 0.26082, + "grad_norm": 0.9149769171286626, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 26082 + }, + { + "epoch": 0.26083, + "grad_norm": 1.3751581666222845, + "learning_rate": 0.003, + "loss": 4.068, + "step": 26083 + }, + { + "epoch": 0.26084, + "grad_norm": 0.8455651095757869, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 26084 + }, + { + "epoch": 0.26085, + "grad_norm": 1.0468027719578268, + "learning_rate": 0.003, + "loss": 4.031, + "step": 26085 + }, + { + "epoch": 0.26086, + "grad_norm": 1.2129304962620846, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 26086 + }, + { + "epoch": 0.26087, + "grad_norm": 0.9250842185861752, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 26087 + }, + { + "epoch": 0.26088, + "grad_norm": 1.0350763653975876, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 26088 + }, + { + "epoch": 0.26089, + "grad_norm": 1.1496832719418997, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 26089 + }, + { + "epoch": 0.2609, + "grad_norm": 0.9304456067743476, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 26090 + }, + { + "epoch": 0.26091, + "grad_norm": 1.000534207024584, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 26091 + }, + { + "epoch": 0.26092, + "grad_norm": 1.0745200748262116, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 26092 + }, + { + "epoch": 0.26093, + "grad_norm": 0.9746406908603433, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 26093 + }, + { + "epoch": 0.26094, + "grad_norm": 1.1960474075313952, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 26094 + }, + { + "epoch": 0.26095, + "grad_norm": 1.1495579776858893, + "learning_rate": 0.003, + "loss": 4.056, + "step": 26095 + }, + { + "epoch": 0.26096, + "grad_norm": 0.8167046206708348, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 26096 + }, + { + "epoch": 0.26097, + "grad_norm": 0.8579133140194526, + "learning_rate": 0.003, + "loss": 4.048, + "step": 26097 + }, + { + "epoch": 0.26098, + "grad_norm": 0.9383100318012834, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 26098 + }, + { + "epoch": 0.26099, + "grad_norm": 0.9596055652826698, + "learning_rate": 0.003, + "loss": 4.061, + "step": 26099 + }, + { + "epoch": 0.261, + "grad_norm": 1.0162900047541008, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 26100 + }, + { + "epoch": 0.26101, + "grad_norm": 1.0328129921297928, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 26101 + }, + { + "epoch": 0.26102, + "grad_norm": 1.251406149565709, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 26102 + }, + { + "epoch": 0.26103, + "grad_norm": 0.8596650709776507, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 26103 + }, + { + "epoch": 0.26104, + "grad_norm": 0.9881488040990255, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 26104 + }, + { + "epoch": 0.26105, + "grad_norm": 1.3488195569564938, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 26105 + }, + { + "epoch": 0.26106, + "grad_norm": 0.7829338246005859, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 26106 + }, + { + "epoch": 0.26107, + "grad_norm": 1.0510356863415915, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 26107 + }, + { + "epoch": 0.26108, + "grad_norm": 0.9775209176519705, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 26108 + }, + { + "epoch": 0.26109, + "grad_norm": 1.093405918876141, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 26109 + }, + { + "epoch": 0.2611, + "grad_norm": 1.26969590456516, + "learning_rate": 0.003, + "loss": 4.0962, + "step": 26110 + }, + { + "epoch": 0.26111, + "grad_norm": 1.11755084211255, + "learning_rate": 0.003, + "loss": 4.024, + "step": 26111 + }, + { + "epoch": 0.26112, + "grad_norm": 1.320478703374044, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 26112 + }, + { + "epoch": 0.26113, + "grad_norm": 1.0474030104876424, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 26113 + }, + { + "epoch": 0.26114, + "grad_norm": 1.1408881675350542, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 26114 + }, + { + "epoch": 0.26115, + "grad_norm": 1.024523876985758, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 26115 + }, + { + "epoch": 0.26116, + "grad_norm": 1.135926205637625, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 26116 + }, + { + "epoch": 0.26117, + "grad_norm": 0.9800491252538537, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 26117 + }, + { + "epoch": 0.26118, + "grad_norm": 0.8953361564151356, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 26118 + }, + { + "epoch": 0.26119, + "grad_norm": 0.8793778208589106, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 26119 + }, + { + "epoch": 0.2612, + "grad_norm": 0.8902583011429954, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 26120 + }, + { + "epoch": 0.26121, + "grad_norm": 1.1938848970222877, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 26121 + }, + { + "epoch": 0.26122, + "grad_norm": 1.2351594379033757, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 26122 + }, + { + "epoch": 0.26123, + "grad_norm": 1.0496282160114283, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 26123 + }, + { + "epoch": 0.26124, + "grad_norm": 1.067290726061188, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 26124 + }, + { + "epoch": 0.26125, + "grad_norm": 1.2504680437823361, + "learning_rate": 0.003, + "loss": 4.056, + "step": 26125 + }, + { + "epoch": 0.26126, + "grad_norm": 1.0408541891392138, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 26126 + }, + { + "epoch": 0.26127, + "grad_norm": 1.0372023559217889, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 26127 + }, + { + "epoch": 0.26128, + "grad_norm": 1.1454546166918216, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 26128 + }, + { + "epoch": 0.26129, + "grad_norm": 1.1386539266987374, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 26129 + }, + { + "epoch": 0.2613, + "grad_norm": 0.9835458318562196, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 26130 + }, + { + "epoch": 0.26131, + "grad_norm": 0.9946359230181678, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 26131 + }, + { + "epoch": 0.26132, + "grad_norm": 1.1294119709546375, + "learning_rate": 0.003, + "loss": 4.049, + "step": 26132 + }, + { + "epoch": 0.26133, + "grad_norm": 0.8210088465456514, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 26133 + }, + { + "epoch": 0.26134, + "grad_norm": 0.91594715792922, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 26134 + }, + { + "epoch": 0.26135, + "grad_norm": 0.8897337751642205, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 26135 + }, + { + "epoch": 0.26136, + "grad_norm": 1.0096331079992946, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 26136 + }, + { + "epoch": 0.26137, + "grad_norm": 1.2049718851525184, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 26137 + }, + { + "epoch": 0.26138, + "grad_norm": 0.9455259686275669, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 26138 + }, + { + "epoch": 0.26139, + "grad_norm": 1.1775833892081815, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 26139 + }, + { + "epoch": 0.2614, + "grad_norm": 0.947738398974754, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 26140 + }, + { + "epoch": 0.26141, + "grad_norm": 1.1617819077389906, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 26141 + }, + { + "epoch": 0.26142, + "grad_norm": 1.4443914897626375, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 26142 + }, + { + "epoch": 0.26143, + "grad_norm": 1.1832073487756765, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 26143 + }, + { + "epoch": 0.26144, + "grad_norm": 1.0090110142450215, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 26144 + }, + { + "epoch": 0.26145, + "grad_norm": 1.3701551392969533, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 26145 + }, + { + "epoch": 0.26146, + "grad_norm": 0.921289519056914, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 26146 + }, + { + "epoch": 0.26147, + "grad_norm": 0.9567045396117821, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 26147 + }, + { + "epoch": 0.26148, + "grad_norm": 1.0597465328236029, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 26148 + }, + { + "epoch": 0.26149, + "grad_norm": 0.9938889730663114, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 26149 + }, + { + "epoch": 0.2615, + "grad_norm": 1.1729223131846354, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 26150 + }, + { + "epoch": 0.26151, + "grad_norm": 1.2873987169161762, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 26151 + }, + { + "epoch": 0.26152, + "grad_norm": 0.9329175018292187, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 26152 + }, + { + "epoch": 0.26153, + "grad_norm": 0.9568442726946275, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 26153 + }, + { + "epoch": 0.26154, + "grad_norm": 1.2564355318858302, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 26154 + }, + { + "epoch": 0.26155, + "grad_norm": 0.8614031750008186, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 26155 + }, + { + "epoch": 0.26156, + "grad_norm": 0.93804131415788, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 26156 + }, + { + "epoch": 0.26157, + "grad_norm": 1.2917185503267947, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 26157 + }, + { + "epoch": 0.26158, + "grad_norm": 0.8790167273078529, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 26158 + }, + { + "epoch": 0.26159, + "grad_norm": 0.9488284369110906, + "learning_rate": 0.003, + "loss": 4.0988, + "step": 26159 + }, + { + "epoch": 0.2616, + "grad_norm": 1.029911317544546, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 26160 + }, + { + "epoch": 0.26161, + "grad_norm": 1.1715089078869216, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 26161 + }, + { + "epoch": 0.26162, + "grad_norm": 1.0625569772099452, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 26162 + }, + { + "epoch": 0.26163, + "grad_norm": 1.3188071251024729, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 26163 + }, + { + "epoch": 0.26164, + "grad_norm": 0.8965030841005472, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 26164 + }, + { + "epoch": 0.26165, + "grad_norm": 0.9775132476475559, + "learning_rate": 0.003, + "loss": 4.056, + "step": 26165 + }, + { + "epoch": 0.26166, + "grad_norm": 1.3631614811573376, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 26166 + }, + { + "epoch": 0.26167, + "grad_norm": 0.903071182860524, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 26167 + }, + { + "epoch": 0.26168, + "grad_norm": 0.9249647786734159, + "learning_rate": 0.003, + "loss": 4.024, + "step": 26168 + }, + { + "epoch": 0.26169, + "grad_norm": 1.0680619726145026, + "learning_rate": 0.003, + "loss": 4.065, + "step": 26169 + }, + { + "epoch": 0.2617, + "grad_norm": 1.2717934456698015, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 26170 + }, + { + "epoch": 0.26171, + "grad_norm": 0.8192912225939855, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 26171 + }, + { + "epoch": 0.26172, + "grad_norm": 0.7675948376752249, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 26172 + }, + { + "epoch": 0.26173, + "grad_norm": 0.8300216575636915, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 26173 + }, + { + "epoch": 0.26174, + "grad_norm": 0.9076003733442711, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 26174 + }, + { + "epoch": 0.26175, + "grad_norm": 0.8862093355240493, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 26175 + }, + { + "epoch": 0.26176, + "grad_norm": 1.0255538692748607, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 26176 + }, + { + "epoch": 0.26177, + "grad_norm": 1.0067565105009533, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 26177 + }, + { + "epoch": 0.26178, + "grad_norm": 1.2971932866434415, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 26178 + }, + { + "epoch": 0.26179, + "grad_norm": 0.8920359515134298, + "learning_rate": 0.003, + "loss": 4.04, + "step": 26179 + }, + { + "epoch": 0.2618, + "grad_norm": 0.9624260452957794, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 26180 + }, + { + "epoch": 0.26181, + "grad_norm": 1.234820096203331, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 26181 + }, + { + "epoch": 0.26182, + "grad_norm": 1.149588459551877, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 26182 + }, + { + "epoch": 0.26183, + "grad_norm": 1.235234529380895, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 26183 + }, + { + "epoch": 0.26184, + "grad_norm": 1.0589809442563447, + "learning_rate": 0.003, + "loss": 4.049, + "step": 26184 + }, + { + "epoch": 0.26185, + "grad_norm": 1.145749739377389, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 26185 + }, + { + "epoch": 0.26186, + "grad_norm": 0.966821523321488, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 26186 + }, + { + "epoch": 0.26187, + "grad_norm": 1.1433151374590982, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 26187 + }, + { + "epoch": 0.26188, + "grad_norm": 1.0877603594870193, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 26188 + }, + { + "epoch": 0.26189, + "grad_norm": 1.0378532781794847, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 26189 + }, + { + "epoch": 0.2619, + "grad_norm": 1.11523105897899, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 26190 + }, + { + "epoch": 0.26191, + "grad_norm": 1.0657639935042311, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 26191 + }, + { + "epoch": 0.26192, + "grad_norm": 1.0158613802754959, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 26192 + }, + { + "epoch": 0.26193, + "grad_norm": 1.1518852625407092, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 26193 + }, + { + "epoch": 0.26194, + "grad_norm": 1.1712089546003877, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 26194 + }, + { + "epoch": 0.26195, + "grad_norm": 1.0845856291510112, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 26195 + }, + { + "epoch": 0.26196, + "grad_norm": 0.9561841567791208, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 26196 + }, + { + "epoch": 0.26197, + "grad_norm": 1.092970778226398, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 26197 + }, + { + "epoch": 0.26198, + "grad_norm": 1.3133732277796795, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 26198 + }, + { + "epoch": 0.26199, + "grad_norm": 1.0913548553339392, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 26199 + }, + { + "epoch": 0.262, + "grad_norm": 1.0585095556690471, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 26200 + }, + { + "epoch": 0.26201, + "grad_norm": 0.9947150952618748, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 26201 + }, + { + "epoch": 0.26202, + "grad_norm": 1.0703770930037348, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 26202 + }, + { + "epoch": 0.26203, + "grad_norm": 0.9838845342562488, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 26203 + }, + { + "epoch": 0.26204, + "grad_norm": 0.8902144462047856, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 26204 + }, + { + "epoch": 0.26205, + "grad_norm": 0.8409661363625096, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 26205 + }, + { + "epoch": 0.26206, + "grad_norm": 0.9060009516788524, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 26206 + }, + { + "epoch": 0.26207, + "grad_norm": 1.0686066479295726, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 26207 + }, + { + "epoch": 0.26208, + "grad_norm": 1.129800020870765, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 26208 + }, + { + "epoch": 0.26209, + "grad_norm": 1.0713055027590468, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 26209 + }, + { + "epoch": 0.2621, + "grad_norm": 1.2063513306338993, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 26210 + }, + { + "epoch": 0.26211, + "grad_norm": 1.1020101062974286, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 26211 + }, + { + "epoch": 0.26212, + "grad_norm": 1.2868226427211837, + "learning_rate": 0.003, + "loss": 4.0891, + "step": 26212 + }, + { + "epoch": 0.26213, + "grad_norm": 1.1106615941819031, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 26213 + }, + { + "epoch": 0.26214, + "grad_norm": 0.9823883634601519, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 26214 + }, + { + "epoch": 0.26215, + "grad_norm": 1.2116187211221408, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 26215 + }, + { + "epoch": 0.26216, + "grad_norm": 0.906415713807938, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 26216 + }, + { + "epoch": 0.26217, + "grad_norm": 1.0738799174347267, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 26217 + }, + { + "epoch": 0.26218, + "grad_norm": 1.015496120658486, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 26218 + }, + { + "epoch": 0.26219, + "grad_norm": 0.9581662489097157, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 26219 + }, + { + "epoch": 0.2622, + "grad_norm": 1.196620577725661, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 26220 + }, + { + "epoch": 0.26221, + "grad_norm": 1.0756746182953323, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 26221 + }, + { + "epoch": 0.26222, + "grad_norm": 1.0820419773748908, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 26222 + }, + { + "epoch": 0.26223, + "grad_norm": 1.110244654089877, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 26223 + }, + { + "epoch": 0.26224, + "grad_norm": 1.104757207149956, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 26224 + }, + { + "epoch": 0.26225, + "grad_norm": 1.1400250731470036, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 26225 + }, + { + "epoch": 0.26226, + "grad_norm": 1.1788508333330618, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 26226 + }, + { + "epoch": 0.26227, + "grad_norm": 1.11373503316036, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 26227 + }, + { + "epoch": 0.26228, + "grad_norm": 0.9042746417975286, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 26228 + }, + { + "epoch": 0.26229, + "grad_norm": 1.1890948800390757, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 26229 + }, + { + "epoch": 0.2623, + "grad_norm": 1.1496493792881317, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 26230 + }, + { + "epoch": 0.26231, + "grad_norm": 0.9667409952271416, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 26231 + }, + { + "epoch": 0.26232, + "grad_norm": 1.0980660336188393, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 26232 + }, + { + "epoch": 0.26233, + "grad_norm": 1.3428044927666247, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 26233 + }, + { + "epoch": 0.26234, + "grad_norm": 0.9576783913587845, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 26234 + }, + { + "epoch": 0.26235, + "grad_norm": 1.091279712008159, + "learning_rate": 0.003, + "loss": 4.079, + "step": 26235 + }, + { + "epoch": 0.26236, + "grad_norm": 1.1291454354024837, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 26236 + }, + { + "epoch": 0.26237, + "grad_norm": 1.1732339032423138, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 26237 + }, + { + "epoch": 0.26238, + "grad_norm": 0.9762480117274377, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 26238 + }, + { + "epoch": 0.26239, + "grad_norm": 0.9870575762790379, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 26239 + }, + { + "epoch": 0.2624, + "grad_norm": 1.1024900193019118, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 26240 + }, + { + "epoch": 0.26241, + "grad_norm": 1.0010105507881628, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 26241 + }, + { + "epoch": 0.26242, + "grad_norm": 1.1829765489908772, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 26242 + }, + { + "epoch": 0.26243, + "grad_norm": 1.10757694851085, + "learning_rate": 0.003, + "loss": 4.038, + "step": 26243 + }, + { + "epoch": 0.26244, + "grad_norm": 0.9921756710330537, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 26244 + }, + { + "epoch": 0.26245, + "grad_norm": 1.1077197818704196, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 26245 + }, + { + "epoch": 0.26246, + "grad_norm": 1.124487434369494, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 26246 + }, + { + "epoch": 0.26247, + "grad_norm": 0.9767526177909048, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 26247 + }, + { + "epoch": 0.26248, + "grad_norm": 1.0267839007792816, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 26248 + }, + { + "epoch": 0.26249, + "grad_norm": 1.1746365172568674, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 26249 + }, + { + "epoch": 0.2625, + "grad_norm": 1.1445437487249472, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 26250 + }, + { + "epoch": 0.26251, + "grad_norm": 1.0001812676315047, + "learning_rate": 0.003, + "loss": 4.073, + "step": 26251 + }, + { + "epoch": 0.26252, + "grad_norm": 1.0097463661611195, + "learning_rate": 0.003, + "loss": 4.061, + "step": 26252 + }, + { + "epoch": 0.26253, + "grad_norm": 1.0575842055695934, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 26253 + }, + { + "epoch": 0.26254, + "grad_norm": 1.292422425462145, + "learning_rate": 0.003, + "loss": 4.053, + "step": 26254 + }, + { + "epoch": 0.26255, + "grad_norm": 1.0520814652685953, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 26255 + }, + { + "epoch": 0.26256, + "grad_norm": 0.9346485110389737, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 26256 + }, + { + "epoch": 0.26257, + "grad_norm": 0.9565867283762317, + "learning_rate": 0.003, + "loss": 4.05, + "step": 26257 + }, + { + "epoch": 0.26258, + "grad_norm": 1.159786411372917, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 26258 + }, + { + "epoch": 0.26259, + "grad_norm": 1.0748367616695589, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 26259 + }, + { + "epoch": 0.2626, + "grad_norm": 1.1280505673554646, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 26260 + }, + { + "epoch": 0.26261, + "grad_norm": 1.2466409198439847, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 26261 + }, + { + "epoch": 0.26262, + "grad_norm": 1.001067787729949, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 26262 + }, + { + "epoch": 0.26263, + "grad_norm": 1.0237921368929943, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 26263 + }, + { + "epoch": 0.26264, + "grad_norm": 0.9470870027454916, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 26264 + }, + { + "epoch": 0.26265, + "grad_norm": 0.9377345899405329, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 26265 + }, + { + "epoch": 0.26266, + "grad_norm": 1.055196565322768, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 26266 + }, + { + "epoch": 0.26267, + "grad_norm": 1.0249142743280097, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 26267 + }, + { + "epoch": 0.26268, + "grad_norm": 1.328548373792517, + "learning_rate": 0.003, + "loss": 4.032, + "step": 26268 + }, + { + "epoch": 0.26269, + "grad_norm": 0.9051805258300913, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 26269 + }, + { + "epoch": 0.2627, + "grad_norm": 0.9649860736278666, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 26270 + }, + { + "epoch": 0.26271, + "grad_norm": 1.0527579680378254, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 26271 + }, + { + "epoch": 0.26272, + "grad_norm": 1.0791735515522753, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 26272 + }, + { + "epoch": 0.26273, + "grad_norm": 1.0639031347274863, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 26273 + }, + { + "epoch": 0.26274, + "grad_norm": 1.0081038436769019, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 26274 + }, + { + "epoch": 0.26275, + "grad_norm": 1.1018249187250604, + "learning_rate": 0.003, + "loss": 4.017, + "step": 26275 + }, + { + "epoch": 0.26276, + "grad_norm": 1.1546286061982816, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 26276 + }, + { + "epoch": 0.26277, + "grad_norm": 1.0037400231666824, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 26277 + }, + { + "epoch": 0.26278, + "grad_norm": 1.1566584342837427, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 26278 + }, + { + "epoch": 0.26279, + "grad_norm": 1.1099350118494697, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 26279 + }, + { + "epoch": 0.2628, + "grad_norm": 0.9091642122553973, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 26280 + }, + { + "epoch": 0.26281, + "grad_norm": 1.0201480845127322, + "learning_rate": 0.003, + "loss": 4.039, + "step": 26281 + }, + { + "epoch": 0.26282, + "grad_norm": 1.15799490754445, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 26282 + }, + { + "epoch": 0.26283, + "grad_norm": 1.0843064537891487, + "learning_rate": 0.003, + "loss": 4.044, + "step": 26283 + }, + { + "epoch": 0.26284, + "grad_norm": 1.0309631140401265, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 26284 + }, + { + "epoch": 0.26285, + "grad_norm": 1.0662946825360806, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 26285 + }, + { + "epoch": 0.26286, + "grad_norm": 1.0221970229234467, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 26286 + }, + { + "epoch": 0.26287, + "grad_norm": 1.1196330696583912, + "learning_rate": 0.003, + "loss": 4.0992, + "step": 26287 + }, + { + "epoch": 0.26288, + "grad_norm": 1.0698827719808826, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 26288 + }, + { + "epoch": 0.26289, + "grad_norm": 1.1291347271961878, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 26289 + }, + { + "epoch": 0.2629, + "grad_norm": 1.2358070936017624, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 26290 + }, + { + "epoch": 0.26291, + "grad_norm": 0.9873519905136712, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 26291 + }, + { + "epoch": 0.26292, + "grad_norm": 1.1259163322100698, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 26292 + }, + { + "epoch": 0.26293, + "grad_norm": 1.224557903888818, + "learning_rate": 0.003, + "loss": 4.026, + "step": 26293 + }, + { + "epoch": 0.26294, + "grad_norm": 0.9663206059690524, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 26294 + }, + { + "epoch": 0.26295, + "grad_norm": 1.1745656274767975, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 26295 + }, + { + "epoch": 0.26296, + "grad_norm": 0.9618661153455741, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 26296 + }, + { + "epoch": 0.26297, + "grad_norm": 1.1712360304434812, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 26297 + }, + { + "epoch": 0.26298, + "grad_norm": 1.0975719050794064, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 26298 + }, + { + "epoch": 0.26299, + "grad_norm": 1.1533095901986068, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 26299 + }, + { + "epoch": 0.263, + "grad_norm": 1.0382871006141199, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 26300 + }, + { + "epoch": 0.26301, + "grad_norm": 1.0622621069309224, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 26301 + }, + { + "epoch": 0.26302, + "grad_norm": 1.1507452421970203, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 26302 + }, + { + "epoch": 0.26303, + "grad_norm": 0.8351108923056995, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 26303 + }, + { + "epoch": 0.26304, + "grad_norm": 0.7655215629445693, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 26304 + }, + { + "epoch": 0.26305, + "grad_norm": 0.9341418605257483, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 26305 + }, + { + "epoch": 0.26306, + "grad_norm": 1.0653161894764132, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 26306 + }, + { + "epoch": 0.26307, + "grad_norm": 1.0804996382519607, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 26307 + }, + { + "epoch": 0.26308, + "grad_norm": 1.2867026520608698, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 26308 + }, + { + "epoch": 0.26309, + "grad_norm": 0.851108632299064, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 26309 + }, + { + "epoch": 0.2631, + "grad_norm": 0.7972370763805793, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 26310 + }, + { + "epoch": 0.26311, + "grad_norm": 0.9281654121182649, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 26311 + }, + { + "epoch": 0.26312, + "grad_norm": 1.1531371502599104, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 26312 + }, + { + "epoch": 0.26313, + "grad_norm": 1.0396049975643926, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 26313 + }, + { + "epoch": 0.26314, + "grad_norm": 1.2390903557933626, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 26314 + }, + { + "epoch": 0.26315, + "grad_norm": 0.939478845291326, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 26315 + }, + { + "epoch": 0.26316, + "grad_norm": 1.1312826521008341, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 26316 + }, + { + "epoch": 0.26317, + "grad_norm": 1.2575209384093768, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 26317 + }, + { + "epoch": 0.26318, + "grad_norm": 0.9854274096331384, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 26318 + }, + { + "epoch": 0.26319, + "grad_norm": 1.2107148986350063, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 26319 + }, + { + "epoch": 0.2632, + "grad_norm": 1.1747191326929378, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 26320 + }, + { + "epoch": 0.26321, + "grad_norm": 1.091901825220608, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 26321 + }, + { + "epoch": 0.26322, + "grad_norm": 1.1086425080979196, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 26322 + }, + { + "epoch": 0.26323, + "grad_norm": 1.0765762583277185, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 26323 + }, + { + "epoch": 0.26324, + "grad_norm": 1.2285713663916968, + "learning_rate": 0.003, + "loss": 4.0854, + "step": 26324 + }, + { + "epoch": 0.26325, + "grad_norm": 0.8709233175219202, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 26325 + }, + { + "epoch": 0.26326, + "grad_norm": 0.7584852166952665, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 26326 + }, + { + "epoch": 0.26327, + "grad_norm": 0.854028669340203, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 26327 + }, + { + "epoch": 0.26328, + "grad_norm": 0.9020121085756826, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 26328 + }, + { + "epoch": 0.26329, + "grad_norm": 1.0982171397702587, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 26329 + }, + { + "epoch": 0.2633, + "grad_norm": 1.1701068250906916, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 26330 + }, + { + "epoch": 0.26331, + "grad_norm": 1.181782005297754, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 26331 + }, + { + "epoch": 0.26332, + "grad_norm": 1.1080355461955191, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 26332 + }, + { + "epoch": 0.26333, + "grad_norm": 1.0645276170478204, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 26333 + }, + { + "epoch": 0.26334, + "grad_norm": 0.8426255925803584, + "learning_rate": 0.003, + "loss": 4.058, + "step": 26334 + }, + { + "epoch": 0.26335, + "grad_norm": 0.8397628708706064, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 26335 + }, + { + "epoch": 0.26336, + "grad_norm": 0.9285460120869985, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 26336 + }, + { + "epoch": 0.26337, + "grad_norm": 1.0497722285119653, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 26337 + }, + { + "epoch": 0.26338, + "grad_norm": 1.093056276470593, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 26338 + }, + { + "epoch": 0.26339, + "grad_norm": 1.010944558976674, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 26339 + }, + { + "epoch": 0.2634, + "grad_norm": 1.1141246702051164, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 26340 + }, + { + "epoch": 0.26341, + "grad_norm": 0.9453408381372298, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 26341 + }, + { + "epoch": 0.26342, + "grad_norm": 1.0146157000064926, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 26342 + }, + { + "epoch": 0.26343, + "grad_norm": 1.2140296047284231, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 26343 + }, + { + "epoch": 0.26344, + "grad_norm": 1.0551383708639777, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 26344 + }, + { + "epoch": 0.26345, + "grad_norm": 1.2432970949088198, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 26345 + }, + { + "epoch": 0.26346, + "grad_norm": 0.8079924255173467, + "learning_rate": 0.003, + "loss": 4.043, + "step": 26346 + }, + { + "epoch": 0.26347, + "grad_norm": 1.0017713643543515, + "learning_rate": 0.003, + "loss": 4.0868, + "step": 26347 + }, + { + "epoch": 0.26348, + "grad_norm": 1.341403508740543, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 26348 + }, + { + "epoch": 0.26349, + "grad_norm": 0.83653019755823, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 26349 + }, + { + "epoch": 0.2635, + "grad_norm": 1.0604338729391645, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 26350 + }, + { + "epoch": 0.26351, + "grad_norm": 1.3467735201362623, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 26351 + }, + { + "epoch": 0.26352, + "grad_norm": 0.9487580020012472, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 26352 + }, + { + "epoch": 0.26353, + "grad_norm": 1.2066153989616655, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 26353 + }, + { + "epoch": 0.26354, + "grad_norm": 1.1241939293992513, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 26354 + }, + { + "epoch": 0.26355, + "grad_norm": 1.166699432561236, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 26355 + }, + { + "epoch": 0.26356, + "grad_norm": 0.8593544871520914, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 26356 + }, + { + "epoch": 0.26357, + "grad_norm": 0.923313079787164, + "learning_rate": 0.003, + "loss": 4.086, + "step": 26357 + }, + { + "epoch": 0.26358, + "grad_norm": 1.0449486504703118, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 26358 + }, + { + "epoch": 0.26359, + "grad_norm": 0.9046243965209673, + "learning_rate": 0.003, + "loss": 4.055, + "step": 26359 + }, + { + "epoch": 0.2636, + "grad_norm": 0.9711881419836772, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 26360 + }, + { + "epoch": 0.26361, + "grad_norm": 1.1620998210077347, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 26361 + }, + { + "epoch": 0.26362, + "grad_norm": 1.1454849864299113, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 26362 + }, + { + "epoch": 0.26363, + "grad_norm": 0.9864121634238664, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 26363 + }, + { + "epoch": 0.26364, + "grad_norm": 1.1098844156581404, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 26364 + }, + { + "epoch": 0.26365, + "grad_norm": 1.0385851261935468, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 26365 + }, + { + "epoch": 0.26366, + "grad_norm": 1.0525489294632244, + "learning_rate": 0.003, + "loss": 4.052, + "step": 26366 + }, + { + "epoch": 0.26367, + "grad_norm": 1.1277862833254206, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 26367 + }, + { + "epoch": 0.26368, + "grad_norm": 1.0214468916822526, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 26368 + }, + { + "epoch": 0.26369, + "grad_norm": 1.1943000112499857, + "learning_rate": 0.003, + "loss": 4.055, + "step": 26369 + }, + { + "epoch": 0.2637, + "grad_norm": 1.1760184407519005, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 26370 + }, + { + "epoch": 0.26371, + "grad_norm": 1.190575076644339, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 26371 + }, + { + "epoch": 0.26372, + "grad_norm": 0.9144643132848189, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 26372 + }, + { + "epoch": 0.26373, + "grad_norm": 1.0009777934944855, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 26373 + }, + { + "epoch": 0.26374, + "grad_norm": 1.2128296814165311, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 26374 + }, + { + "epoch": 0.26375, + "grad_norm": 0.9931473693083682, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 26375 + }, + { + "epoch": 0.26376, + "grad_norm": 1.1136920558541383, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 26376 + }, + { + "epoch": 0.26377, + "grad_norm": 1.1806640398051662, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 26377 + }, + { + "epoch": 0.26378, + "grad_norm": 1.0011073945369025, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 26378 + }, + { + "epoch": 0.26379, + "grad_norm": 1.0576836720038414, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 26379 + }, + { + "epoch": 0.2638, + "grad_norm": 0.8886346219244987, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 26380 + }, + { + "epoch": 0.26381, + "grad_norm": 0.8692721447336678, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 26381 + }, + { + "epoch": 0.26382, + "grad_norm": 0.8891373639235298, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 26382 + }, + { + "epoch": 0.26383, + "grad_norm": 0.8368501253860819, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 26383 + }, + { + "epoch": 0.26384, + "grad_norm": 1.0812017331518107, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 26384 + }, + { + "epoch": 0.26385, + "grad_norm": 1.271655021676624, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 26385 + }, + { + "epoch": 0.26386, + "grad_norm": 1.0111017787755534, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 26386 + }, + { + "epoch": 0.26387, + "grad_norm": 1.2735923060598984, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 26387 + }, + { + "epoch": 0.26388, + "grad_norm": 0.9547229230644638, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 26388 + }, + { + "epoch": 0.26389, + "grad_norm": 0.8991077163919485, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 26389 + }, + { + "epoch": 0.2639, + "grad_norm": 0.9795266218512588, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 26390 + }, + { + "epoch": 0.26391, + "grad_norm": 1.08366657253546, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 26391 + }, + { + "epoch": 0.26392, + "grad_norm": 1.2671687578815907, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 26392 + }, + { + "epoch": 0.26393, + "grad_norm": 1.1032857152738018, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 26393 + }, + { + "epoch": 0.26394, + "grad_norm": 1.1159757356024909, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 26394 + }, + { + "epoch": 0.26395, + "grad_norm": 1.079539885992238, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 26395 + }, + { + "epoch": 0.26396, + "grad_norm": 1.3776208712741373, + "learning_rate": 0.003, + "loss": 4.082, + "step": 26396 + }, + { + "epoch": 0.26397, + "grad_norm": 0.9633993506721334, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 26397 + }, + { + "epoch": 0.26398, + "grad_norm": 0.9937670963036491, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 26398 + }, + { + "epoch": 0.26399, + "grad_norm": 1.119596787139305, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 26399 + }, + { + "epoch": 0.264, + "grad_norm": 0.9371911309170787, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 26400 + }, + { + "epoch": 0.26401, + "grad_norm": 1.0643539751931925, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 26401 + }, + { + "epoch": 0.26402, + "grad_norm": 0.8511545939124537, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 26402 + }, + { + "epoch": 0.26403, + "grad_norm": 0.9113791893172801, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 26403 + }, + { + "epoch": 0.26404, + "grad_norm": 1.0597637932935384, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 26404 + }, + { + "epoch": 0.26405, + "grad_norm": 1.2335332838195263, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 26405 + }, + { + "epoch": 0.26406, + "grad_norm": 1.1639048248102213, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 26406 + }, + { + "epoch": 0.26407, + "grad_norm": 0.8963913962270109, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 26407 + }, + { + "epoch": 0.26408, + "grad_norm": 0.9542635683137879, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 26408 + }, + { + "epoch": 0.26409, + "grad_norm": 1.0155840340173994, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 26409 + }, + { + "epoch": 0.2641, + "grad_norm": 1.1056931289394318, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 26410 + }, + { + "epoch": 0.26411, + "grad_norm": 0.9474202589901622, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 26411 + }, + { + "epoch": 0.26412, + "grad_norm": 1.0596613496129803, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 26412 + }, + { + "epoch": 0.26413, + "grad_norm": 1.2448276919520906, + "learning_rate": 0.003, + "loss": 4.0956, + "step": 26413 + }, + { + "epoch": 0.26414, + "grad_norm": 1.1329552040915047, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 26414 + }, + { + "epoch": 0.26415, + "grad_norm": 1.1599786766745839, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 26415 + }, + { + "epoch": 0.26416, + "grad_norm": 1.1506301649946278, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 26416 + }, + { + "epoch": 0.26417, + "grad_norm": 1.1678363807262424, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 26417 + }, + { + "epoch": 0.26418, + "grad_norm": 0.86641466768313, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 26418 + }, + { + "epoch": 0.26419, + "grad_norm": 0.9657929609497748, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 26419 + }, + { + "epoch": 0.2642, + "grad_norm": 1.121421463197832, + "learning_rate": 0.003, + "loss": 4.0878, + "step": 26420 + }, + { + "epoch": 0.26421, + "grad_norm": 1.3205257420786987, + "learning_rate": 0.003, + "loss": 4.07, + "step": 26421 + }, + { + "epoch": 0.26422, + "grad_norm": 0.9272962052365282, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 26422 + }, + { + "epoch": 0.26423, + "grad_norm": 1.033876686465868, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 26423 + }, + { + "epoch": 0.26424, + "grad_norm": 1.3586671383489994, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 26424 + }, + { + "epoch": 0.26425, + "grad_norm": 1.019168159752844, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 26425 + }, + { + "epoch": 0.26426, + "grad_norm": 1.0859916636935394, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 26426 + }, + { + "epoch": 0.26427, + "grad_norm": 1.1048253790272033, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 26427 + }, + { + "epoch": 0.26428, + "grad_norm": 1.1022864441945532, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 26428 + }, + { + "epoch": 0.26429, + "grad_norm": 1.1724218779648474, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 26429 + }, + { + "epoch": 0.2643, + "grad_norm": 0.8580818894643248, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 26430 + }, + { + "epoch": 0.26431, + "grad_norm": 1.0338888104920387, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 26431 + }, + { + "epoch": 0.26432, + "grad_norm": 1.05794114123531, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 26432 + }, + { + "epoch": 0.26433, + "grad_norm": 0.9114554381460644, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 26433 + }, + { + "epoch": 0.26434, + "grad_norm": 1.1368290319021161, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 26434 + }, + { + "epoch": 0.26435, + "grad_norm": 0.9000618660440478, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 26435 + }, + { + "epoch": 0.26436, + "grad_norm": 1.1275886418572698, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 26436 + }, + { + "epoch": 0.26437, + "grad_norm": 0.9468630258738344, + "learning_rate": 0.003, + "loss": 4.087, + "step": 26437 + }, + { + "epoch": 0.26438, + "grad_norm": 0.9205237739964749, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 26438 + }, + { + "epoch": 0.26439, + "grad_norm": 0.9754018541984176, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 26439 + }, + { + "epoch": 0.2644, + "grad_norm": 1.1591444482441076, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 26440 + }, + { + "epoch": 0.26441, + "grad_norm": 1.437219795594992, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 26441 + }, + { + "epoch": 0.26442, + "grad_norm": 0.7789779273179234, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 26442 + }, + { + "epoch": 0.26443, + "grad_norm": 0.6647319817450432, + "learning_rate": 0.003, + "loss": 4.055, + "step": 26443 + }, + { + "epoch": 0.26444, + "grad_norm": 0.8026667619280662, + "learning_rate": 0.003, + "loss": 4.037, + "step": 26444 + }, + { + "epoch": 0.26445, + "grad_norm": 0.9876695518423152, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 26445 + }, + { + "epoch": 0.26446, + "grad_norm": 1.1665430058224522, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 26446 + }, + { + "epoch": 0.26447, + "grad_norm": 1.0302619629468694, + "learning_rate": 0.003, + "loss": 4.086, + "step": 26447 + }, + { + "epoch": 0.26448, + "grad_norm": 1.1135261131145562, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 26448 + }, + { + "epoch": 0.26449, + "grad_norm": 0.9126930205496252, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 26449 + }, + { + "epoch": 0.2645, + "grad_norm": 1.3058575204863545, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 26450 + }, + { + "epoch": 0.26451, + "grad_norm": 1.4064142994185236, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 26451 + }, + { + "epoch": 0.26452, + "grad_norm": 0.8419595892646138, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 26452 + }, + { + "epoch": 0.26453, + "grad_norm": 0.7833406719345647, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 26453 + }, + { + "epoch": 0.26454, + "grad_norm": 0.9464988865427642, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 26454 + }, + { + "epoch": 0.26455, + "grad_norm": 1.2523126463102674, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 26455 + }, + { + "epoch": 0.26456, + "grad_norm": 1.0993448503861503, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 26456 + }, + { + "epoch": 0.26457, + "grad_norm": 1.0684019918713812, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 26457 + }, + { + "epoch": 0.26458, + "grad_norm": 0.7789141334171557, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 26458 + }, + { + "epoch": 0.26459, + "grad_norm": 0.8347718618151811, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 26459 + }, + { + "epoch": 0.2646, + "grad_norm": 1.0222026372423125, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 26460 + }, + { + "epoch": 0.26461, + "grad_norm": 1.267075814495188, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 26461 + }, + { + "epoch": 0.26462, + "grad_norm": 0.8929377959339344, + "learning_rate": 0.003, + "loss": 4.02, + "step": 26462 + }, + { + "epoch": 0.26463, + "grad_norm": 1.1002800077645736, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 26463 + }, + { + "epoch": 0.26464, + "grad_norm": 1.2241408140415941, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 26464 + }, + { + "epoch": 0.26465, + "grad_norm": 0.9199021361350643, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 26465 + }, + { + "epoch": 0.26466, + "grad_norm": 1.141260860780698, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 26466 + }, + { + "epoch": 0.26467, + "grad_norm": 1.290021862684046, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 26467 + }, + { + "epoch": 0.26468, + "grad_norm": 1.0969269549040377, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 26468 + }, + { + "epoch": 0.26469, + "grad_norm": 1.0513444177776083, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 26469 + }, + { + "epoch": 0.2647, + "grad_norm": 0.9429236098518875, + "learning_rate": 0.003, + "loss": 4.042, + "step": 26470 + }, + { + "epoch": 0.26471, + "grad_norm": 0.9337441714378916, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 26471 + }, + { + "epoch": 0.26472, + "grad_norm": 1.0156469188388038, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 26472 + }, + { + "epoch": 0.26473, + "grad_norm": 1.0585814444597301, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 26473 + }, + { + "epoch": 0.26474, + "grad_norm": 0.8468784617641089, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 26474 + }, + { + "epoch": 0.26475, + "grad_norm": 1.0115324888209636, + "learning_rate": 0.003, + "loss": 4.058, + "step": 26475 + }, + { + "epoch": 0.26476, + "grad_norm": 1.1680445375189459, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 26476 + }, + { + "epoch": 0.26477, + "grad_norm": 0.9868278355053323, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 26477 + }, + { + "epoch": 0.26478, + "grad_norm": 1.0258266495284805, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 26478 + }, + { + "epoch": 0.26479, + "grad_norm": 1.163428723800181, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 26479 + }, + { + "epoch": 0.2648, + "grad_norm": 1.0101502272424137, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 26480 + }, + { + "epoch": 0.26481, + "grad_norm": 1.2045910778600708, + "learning_rate": 0.003, + "loss": 4.055, + "step": 26481 + }, + { + "epoch": 0.26482, + "grad_norm": 1.0900159839632282, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 26482 + }, + { + "epoch": 0.26483, + "grad_norm": 1.3221563227427378, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 26483 + }, + { + "epoch": 0.26484, + "grad_norm": 0.9299087286019401, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 26484 + }, + { + "epoch": 0.26485, + "grad_norm": 0.8066317756905538, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 26485 + }, + { + "epoch": 0.26486, + "grad_norm": 1.059592191200587, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 26486 + }, + { + "epoch": 0.26487, + "grad_norm": 1.3993359650349824, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 26487 + }, + { + "epoch": 0.26488, + "grad_norm": 1.1606784454711832, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 26488 + }, + { + "epoch": 0.26489, + "grad_norm": 1.039233304894505, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 26489 + }, + { + "epoch": 0.2649, + "grad_norm": 1.0012061103932528, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 26490 + }, + { + "epoch": 0.26491, + "grad_norm": 1.0922095339235758, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 26491 + }, + { + "epoch": 0.26492, + "grad_norm": 1.2962297888775403, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 26492 + }, + { + "epoch": 0.26493, + "grad_norm": 0.8915301046938275, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 26493 + }, + { + "epoch": 0.26494, + "grad_norm": 1.0821620839435961, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 26494 + }, + { + "epoch": 0.26495, + "grad_norm": 1.1825204907031703, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 26495 + }, + { + "epoch": 0.26496, + "grad_norm": 0.8889313985042615, + "learning_rate": 0.003, + "loss": 4.044, + "step": 26496 + }, + { + "epoch": 0.26497, + "grad_norm": 0.922755667407196, + "learning_rate": 0.003, + "loss": 4.06, + "step": 26497 + }, + { + "epoch": 0.26498, + "grad_norm": 1.1142897359064308, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 26498 + }, + { + "epoch": 0.26499, + "grad_norm": 1.1538337114526218, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 26499 + }, + { + "epoch": 0.265, + "grad_norm": 0.941712221323319, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 26500 + }, + { + "epoch": 0.26501, + "grad_norm": 1.0457282621023696, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 26501 + }, + { + "epoch": 0.26502, + "grad_norm": 1.2415693389764972, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 26502 + }, + { + "epoch": 0.26503, + "grad_norm": 0.8938835052338983, + "learning_rate": 0.003, + "loss": 4.053, + "step": 26503 + }, + { + "epoch": 0.26504, + "grad_norm": 0.9641581345251794, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 26504 + }, + { + "epoch": 0.26505, + "grad_norm": 1.015222276776737, + "learning_rate": 0.003, + "loss": 4.059, + "step": 26505 + }, + { + "epoch": 0.26506, + "grad_norm": 1.0581007100447894, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 26506 + }, + { + "epoch": 0.26507, + "grad_norm": 1.0903494069478616, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 26507 + }, + { + "epoch": 0.26508, + "grad_norm": 1.2064499137477, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 26508 + }, + { + "epoch": 0.26509, + "grad_norm": 1.1586353002516179, + "learning_rate": 0.003, + "loss": 4.035, + "step": 26509 + }, + { + "epoch": 0.2651, + "grad_norm": 0.8974029070477083, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 26510 + }, + { + "epoch": 0.26511, + "grad_norm": 0.8669707212444355, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 26511 + }, + { + "epoch": 0.26512, + "grad_norm": 0.8457846101487962, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 26512 + }, + { + "epoch": 0.26513, + "grad_norm": 1.0579655769537484, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 26513 + }, + { + "epoch": 0.26514, + "grad_norm": 1.1069730209238946, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 26514 + }, + { + "epoch": 0.26515, + "grad_norm": 1.030780308938152, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 26515 + }, + { + "epoch": 0.26516, + "grad_norm": 1.0536418557532783, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 26516 + }, + { + "epoch": 0.26517, + "grad_norm": 0.9528299700372573, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 26517 + }, + { + "epoch": 0.26518, + "grad_norm": 0.976984312752043, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 26518 + }, + { + "epoch": 0.26519, + "grad_norm": 1.1163262317785538, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 26519 + }, + { + "epoch": 0.2652, + "grad_norm": 1.2832757516251052, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 26520 + }, + { + "epoch": 0.26521, + "grad_norm": 1.176020196628434, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 26521 + }, + { + "epoch": 0.26522, + "grad_norm": 1.1121517924398197, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 26522 + }, + { + "epoch": 0.26523, + "grad_norm": 1.1099243345272196, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 26523 + }, + { + "epoch": 0.26524, + "grad_norm": 0.9154179905254216, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 26524 + }, + { + "epoch": 0.26525, + "grad_norm": 1.1366627901492605, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 26525 + }, + { + "epoch": 0.26526, + "grad_norm": 1.046974820275566, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 26526 + }, + { + "epoch": 0.26527, + "grad_norm": 0.8866323221964547, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 26527 + }, + { + "epoch": 0.26528, + "grad_norm": 1.009622619742977, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 26528 + }, + { + "epoch": 0.26529, + "grad_norm": 1.2129881510540383, + "learning_rate": 0.003, + "loss": 4.067, + "step": 26529 + }, + { + "epoch": 0.2653, + "grad_norm": 1.151834447992916, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 26530 + }, + { + "epoch": 0.26531, + "grad_norm": 1.02304145240852, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 26531 + }, + { + "epoch": 0.26532, + "grad_norm": 0.9984349017341874, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 26532 + }, + { + "epoch": 0.26533, + "grad_norm": 1.301705389621584, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 26533 + }, + { + "epoch": 0.26534, + "grad_norm": 0.9990359484960285, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 26534 + }, + { + "epoch": 0.26535, + "grad_norm": 1.0600980314954265, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 26535 + }, + { + "epoch": 0.26536, + "grad_norm": 1.076985445792377, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 26536 + }, + { + "epoch": 0.26537, + "grad_norm": 0.910752932466144, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 26537 + }, + { + "epoch": 0.26538, + "grad_norm": 1.2229898886910413, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 26538 + }, + { + "epoch": 0.26539, + "grad_norm": 1.0048871772368, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 26539 + }, + { + "epoch": 0.2654, + "grad_norm": 0.9442538901189876, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 26540 + }, + { + "epoch": 0.26541, + "grad_norm": 1.182059139333234, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 26541 + }, + { + "epoch": 0.26542, + "grad_norm": 1.1009045171824035, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 26542 + }, + { + "epoch": 0.26543, + "grad_norm": 1.3237535855988376, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 26543 + }, + { + "epoch": 0.26544, + "grad_norm": 0.9445069471641298, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 26544 + }, + { + "epoch": 0.26545, + "grad_norm": 1.1250879870583759, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 26545 + }, + { + "epoch": 0.26546, + "grad_norm": 1.0253821022339225, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 26546 + }, + { + "epoch": 0.26547, + "grad_norm": 1.1742984762490043, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 26547 + }, + { + "epoch": 0.26548, + "grad_norm": 0.8323026267823054, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 26548 + }, + { + "epoch": 0.26549, + "grad_norm": 0.8138759185553106, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 26549 + }, + { + "epoch": 0.2655, + "grad_norm": 0.8830609354811194, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 26550 + }, + { + "epoch": 0.26551, + "grad_norm": 1.314851157346909, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 26551 + }, + { + "epoch": 0.26552, + "grad_norm": 1.0623518075911265, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 26552 + }, + { + "epoch": 0.26553, + "grad_norm": 0.9826212480139138, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 26553 + }, + { + "epoch": 0.26554, + "grad_norm": 1.1210588651054556, + "learning_rate": 0.003, + "loss": 4.04, + "step": 26554 + }, + { + "epoch": 0.26555, + "grad_norm": 1.2743458031655899, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 26555 + }, + { + "epoch": 0.26556, + "grad_norm": 1.08657903585367, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 26556 + }, + { + "epoch": 0.26557, + "grad_norm": 0.9336045549197405, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 26557 + }, + { + "epoch": 0.26558, + "grad_norm": 1.0444334651678138, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 26558 + }, + { + "epoch": 0.26559, + "grad_norm": 1.1115386340329354, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 26559 + }, + { + "epoch": 0.2656, + "grad_norm": 1.119522660672067, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 26560 + }, + { + "epoch": 0.26561, + "grad_norm": 1.1266339526798783, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 26561 + }, + { + "epoch": 0.26562, + "grad_norm": 0.9321654539433611, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 26562 + }, + { + "epoch": 0.26563, + "grad_norm": 1.2448831359180768, + "learning_rate": 0.003, + "loss": 4.073, + "step": 26563 + }, + { + "epoch": 0.26564, + "grad_norm": 1.061095759972243, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 26564 + }, + { + "epoch": 0.26565, + "grad_norm": 1.0757032253382366, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 26565 + }, + { + "epoch": 0.26566, + "grad_norm": 1.080646125951733, + "learning_rate": 0.003, + "loss": 4.04, + "step": 26566 + }, + { + "epoch": 0.26567, + "grad_norm": 0.9548696801184711, + "learning_rate": 0.003, + "loss": 4.044, + "step": 26567 + }, + { + "epoch": 0.26568, + "grad_norm": 0.9892847204284675, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 26568 + }, + { + "epoch": 0.26569, + "grad_norm": 1.0470425433423378, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 26569 + }, + { + "epoch": 0.2657, + "grad_norm": 1.1667058066074532, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 26570 + }, + { + "epoch": 0.26571, + "grad_norm": 1.106855115532888, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 26571 + }, + { + "epoch": 0.26572, + "grad_norm": 1.0660706135323998, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 26572 + }, + { + "epoch": 0.26573, + "grad_norm": 1.104356019901685, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 26573 + }, + { + "epoch": 0.26574, + "grad_norm": 0.9859347501678153, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 26574 + }, + { + "epoch": 0.26575, + "grad_norm": 1.2739515763137674, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 26575 + }, + { + "epoch": 0.26576, + "grad_norm": 1.0449129342368042, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 26576 + }, + { + "epoch": 0.26577, + "grad_norm": 1.0669125719946082, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 26577 + }, + { + "epoch": 0.26578, + "grad_norm": 1.1477133454021928, + "learning_rate": 0.003, + "loss": 4.0857, + "step": 26578 + }, + { + "epoch": 0.26579, + "grad_norm": 1.04003658978568, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 26579 + }, + { + "epoch": 0.2658, + "grad_norm": 1.2777283217359545, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 26580 + }, + { + "epoch": 0.26581, + "grad_norm": 0.927384514267429, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 26581 + }, + { + "epoch": 0.26582, + "grad_norm": 0.9360770875704449, + "learning_rate": 0.003, + "loss": 4.036, + "step": 26582 + }, + { + "epoch": 0.26583, + "grad_norm": 0.9780087857046821, + "learning_rate": 0.003, + "loss": 4.058, + "step": 26583 + }, + { + "epoch": 0.26584, + "grad_norm": 1.0979666938761103, + "learning_rate": 0.003, + "loss": 4.064, + "step": 26584 + }, + { + "epoch": 0.26585, + "grad_norm": 0.9571131716016358, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 26585 + }, + { + "epoch": 0.26586, + "grad_norm": 1.1022332719047396, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 26586 + }, + { + "epoch": 0.26587, + "grad_norm": 0.9008057575958516, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 26587 + }, + { + "epoch": 0.26588, + "grad_norm": 1.1751398578058847, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 26588 + }, + { + "epoch": 0.26589, + "grad_norm": 1.3137671608458132, + "learning_rate": 0.003, + "loss": 4.057, + "step": 26589 + }, + { + "epoch": 0.2659, + "grad_norm": 0.9208840751935631, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 26590 + }, + { + "epoch": 0.26591, + "grad_norm": 1.02712004902227, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 26591 + }, + { + "epoch": 0.26592, + "grad_norm": 1.359286713537915, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 26592 + }, + { + "epoch": 0.26593, + "grad_norm": 0.9955906671596694, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 26593 + }, + { + "epoch": 0.26594, + "grad_norm": 1.0865571816932207, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 26594 + }, + { + "epoch": 0.26595, + "grad_norm": 1.1310658974863954, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 26595 + }, + { + "epoch": 0.26596, + "grad_norm": 1.1738920600548421, + "learning_rate": 0.003, + "loss": 4.063, + "step": 26596 + }, + { + "epoch": 0.26597, + "grad_norm": 0.9375163903909862, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 26597 + }, + { + "epoch": 0.26598, + "grad_norm": 1.1100619922862665, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 26598 + }, + { + "epoch": 0.26599, + "grad_norm": 1.0227711314670989, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 26599 + }, + { + "epoch": 0.266, + "grad_norm": 1.1330588125375272, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 26600 + }, + { + "epoch": 0.26601, + "grad_norm": 0.9462924834343879, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 26601 + }, + { + "epoch": 0.26602, + "grad_norm": 0.9440740060009417, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 26602 + }, + { + "epoch": 0.26603, + "grad_norm": 1.0190497824962148, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 26603 + }, + { + "epoch": 0.26604, + "grad_norm": 1.2639398122291226, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 26604 + }, + { + "epoch": 0.26605, + "grad_norm": 0.8710009242381153, + "learning_rate": 0.003, + "loss": 4.059, + "step": 26605 + }, + { + "epoch": 0.26606, + "grad_norm": 1.0837002353373733, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 26606 + }, + { + "epoch": 0.26607, + "grad_norm": 0.9193327173361947, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 26607 + }, + { + "epoch": 0.26608, + "grad_norm": 1.1902247221753457, + "learning_rate": 0.003, + "loss": 4.1018, + "step": 26608 + }, + { + "epoch": 0.26609, + "grad_norm": 1.1747537915916422, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 26609 + }, + { + "epoch": 0.2661, + "grad_norm": 1.065147592051408, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 26610 + }, + { + "epoch": 0.26611, + "grad_norm": 1.1142627532570986, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 26611 + }, + { + "epoch": 0.26612, + "grad_norm": 1.1730616989269698, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 26612 + }, + { + "epoch": 0.26613, + "grad_norm": 1.0321545866788913, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 26613 + }, + { + "epoch": 0.26614, + "grad_norm": 1.0818015650084585, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 26614 + }, + { + "epoch": 0.26615, + "grad_norm": 1.2181065474360082, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 26615 + }, + { + "epoch": 0.26616, + "grad_norm": 1.1542873846914805, + "learning_rate": 0.003, + "loss": 4.06, + "step": 26616 + }, + { + "epoch": 0.26617, + "grad_norm": 1.1200730347610255, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 26617 + }, + { + "epoch": 0.26618, + "grad_norm": 1.075920864364899, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 26618 + }, + { + "epoch": 0.26619, + "grad_norm": 1.157884718052505, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 26619 + }, + { + "epoch": 0.2662, + "grad_norm": 0.924596994894297, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 26620 + }, + { + "epoch": 0.26621, + "grad_norm": 0.9054754380013018, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 26621 + }, + { + "epoch": 0.26622, + "grad_norm": 1.065344713908043, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 26622 + }, + { + "epoch": 0.26623, + "grad_norm": 0.9977083345255137, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 26623 + }, + { + "epoch": 0.26624, + "grad_norm": 1.13484859601715, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 26624 + }, + { + "epoch": 0.26625, + "grad_norm": 0.9779977772016306, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 26625 + }, + { + "epoch": 0.26626, + "grad_norm": 1.1516833483737694, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 26626 + }, + { + "epoch": 0.26627, + "grad_norm": 1.0002474774577597, + "learning_rate": 0.003, + "loss": 4.0903, + "step": 26627 + }, + { + "epoch": 0.26628, + "grad_norm": 1.1008463262694481, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 26628 + }, + { + "epoch": 0.26629, + "grad_norm": 1.092853213526667, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 26629 + }, + { + "epoch": 0.2663, + "grad_norm": 1.1060575474956733, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 26630 + }, + { + "epoch": 0.26631, + "grad_norm": 0.9488857969067628, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 26631 + }, + { + "epoch": 0.26632, + "grad_norm": 1.0123767121074083, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 26632 + }, + { + "epoch": 0.26633, + "grad_norm": 1.1892419712300435, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 26633 + }, + { + "epoch": 0.26634, + "grad_norm": 0.8205767472409904, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 26634 + }, + { + "epoch": 0.26635, + "grad_norm": 0.804551517511788, + "learning_rate": 0.003, + "loss": 4.039, + "step": 26635 + }, + { + "epoch": 0.26636, + "grad_norm": 0.9971796044696801, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 26636 + }, + { + "epoch": 0.26637, + "grad_norm": 1.3172838688375939, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 26637 + }, + { + "epoch": 0.26638, + "grad_norm": 1.06718311528191, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 26638 + }, + { + "epoch": 0.26639, + "grad_norm": 1.179070733229589, + "learning_rate": 0.003, + "loss": 4.0902, + "step": 26639 + }, + { + "epoch": 0.2664, + "grad_norm": 0.923666056364222, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 26640 + }, + { + "epoch": 0.26641, + "grad_norm": 1.0190013686266943, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 26641 + }, + { + "epoch": 0.26642, + "grad_norm": 1.2063986434934129, + "learning_rate": 0.003, + "loss": 4.04, + "step": 26642 + }, + { + "epoch": 0.26643, + "grad_norm": 1.070623926873051, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 26643 + }, + { + "epoch": 0.26644, + "grad_norm": 1.0224244753225666, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 26644 + }, + { + "epoch": 0.26645, + "grad_norm": 1.1094420870000559, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 26645 + }, + { + "epoch": 0.26646, + "grad_norm": 1.0547732204995643, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 26646 + }, + { + "epoch": 0.26647, + "grad_norm": 1.0419730562363436, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 26647 + }, + { + "epoch": 0.26648, + "grad_norm": 1.2318975526493612, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 26648 + }, + { + "epoch": 0.26649, + "grad_norm": 0.9774017960010851, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 26649 + }, + { + "epoch": 0.2665, + "grad_norm": 1.0102380077229103, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 26650 + }, + { + "epoch": 0.26651, + "grad_norm": 0.8975767821334405, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 26651 + }, + { + "epoch": 0.26652, + "grad_norm": 1.014314510058257, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 26652 + }, + { + "epoch": 0.26653, + "grad_norm": 0.9893888470442641, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 26653 + }, + { + "epoch": 0.26654, + "grad_norm": 0.9536964938644839, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 26654 + }, + { + "epoch": 0.26655, + "grad_norm": 1.0807736905006478, + "learning_rate": 0.003, + "loss": 4.058, + "step": 26655 + }, + { + "epoch": 0.26656, + "grad_norm": 1.2415493723623883, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 26656 + }, + { + "epoch": 0.26657, + "grad_norm": 1.1816751637996226, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 26657 + }, + { + "epoch": 0.26658, + "grad_norm": 1.2251003055896381, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 26658 + }, + { + "epoch": 0.26659, + "grad_norm": 1.1513458872032738, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 26659 + }, + { + "epoch": 0.2666, + "grad_norm": 0.9846698341873625, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 26660 + }, + { + "epoch": 0.26661, + "grad_norm": 1.0468548588160427, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 26661 + }, + { + "epoch": 0.26662, + "grad_norm": 1.1478923551362754, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 26662 + }, + { + "epoch": 0.26663, + "grad_norm": 1.042536294893675, + "learning_rate": 0.003, + "loss": 4.039, + "step": 26663 + }, + { + "epoch": 0.26664, + "grad_norm": 1.3147187257575226, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 26664 + }, + { + "epoch": 0.26665, + "grad_norm": 1.2370762874247105, + "learning_rate": 0.003, + "loss": 4.055, + "step": 26665 + }, + { + "epoch": 0.26666, + "grad_norm": 1.0800765519868702, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 26666 + }, + { + "epoch": 0.26667, + "grad_norm": 1.230484182223522, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 26667 + }, + { + "epoch": 0.26668, + "grad_norm": 0.9122006863974208, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 26668 + }, + { + "epoch": 0.26669, + "grad_norm": 0.9984606466257071, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 26669 + }, + { + "epoch": 0.2667, + "grad_norm": 1.1724329398674433, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 26670 + }, + { + "epoch": 0.26671, + "grad_norm": 0.8694513899418249, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 26671 + }, + { + "epoch": 0.26672, + "grad_norm": 0.8306078881337416, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 26672 + }, + { + "epoch": 0.26673, + "grad_norm": 0.8974939418060455, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 26673 + }, + { + "epoch": 0.26674, + "grad_norm": 1.0555406292899419, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 26674 + }, + { + "epoch": 0.26675, + "grad_norm": 1.0064229367749922, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 26675 + }, + { + "epoch": 0.26676, + "grad_norm": 1.1947491973328856, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 26676 + }, + { + "epoch": 0.26677, + "grad_norm": 1.3323143631883922, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 26677 + }, + { + "epoch": 0.26678, + "grad_norm": 1.015167661729525, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 26678 + }, + { + "epoch": 0.26679, + "grad_norm": 1.0731510733757315, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 26679 + }, + { + "epoch": 0.2668, + "grad_norm": 1.167583630224231, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 26680 + }, + { + "epoch": 0.26681, + "grad_norm": 0.9675811419913829, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 26681 + }, + { + "epoch": 0.26682, + "grad_norm": 1.108853703108265, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 26682 + }, + { + "epoch": 0.26683, + "grad_norm": 1.0060787492738053, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 26683 + }, + { + "epoch": 0.26684, + "grad_norm": 1.171072218112588, + "learning_rate": 0.003, + "loss": 4.046, + "step": 26684 + }, + { + "epoch": 0.26685, + "grad_norm": 1.1299220261245222, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 26685 + }, + { + "epoch": 0.26686, + "grad_norm": 1.056114655055812, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 26686 + }, + { + "epoch": 0.26687, + "grad_norm": 0.9604247844828804, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 26687 + }, + { + "epoch": 0.26688, + "grad_norm": 1.166541773544023, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 26688 + }, + { + "epoch": 0.26689, + "grad_norm": 0.9254042045079468, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 26689 + }, + { + "epoch": 0.2669, + "grad_norm": 1.1449893067328476, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 26690 + }, + { + "epoch": 0.26691, + "grad_norm": 1.213682191359099, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 26691 + }, + { + "epoch": 0.26692, + "grad_norm": 1.0740065649307486, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 26692 + }, + { + "epoch": 0.26693, + "grad_norm": 0.9875050122124862, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 26693 + }, + { + "epoch": 0.26694, + "grad_norm": 1.2337232742064126, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 26694 + }, + { + "epoch": 0.26695, + "grad_norm": 0.883922379482893, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 26695 + }, + { + "epoch": 0.26696, + "grad_norm": 0.820980915144422, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 26696 + }, + { + "epoch": 0.26697, + "grad_norm": 0.854892113962849, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 26697 + }, + { + "epoch": 0.26698, + "grad_norm": 0.885779892249008, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 26698 + }, + { + "epoch": 0.26699, + "grad_norm": 1.056756155248427, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 26699 + }, + { + "epoch": 0.267, + "grad_norm": 1.0392557528671587, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 26700 + }, + { + "epoch": 0.26701, + "grad_norm": 1.0881859575377082, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 26701 + }, + { + "epoch": 0.26702, + "grad_norm": 1.0424486914466138, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 26702 + }, + { + "epoch": 0.26703, + "grad_norm": 1.126486958906004, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 26703 + }, + { + "epoch": 0.26704, + "grad_norm": 1.2820725484660402, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 26704 + }, + { + "epoch": 0.26705, + "grad_norm": 1.1042896325775917, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 26705 + }, + { + "epoch": 0.26706, + "grad_norm": 1.0872431715820183, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 26706 + }, + { + "epoch": 0.26707, + "grad_norm": 0.9117285560790032, + "learning_rate": 0.003, + "loss": 4.038, + "step": 26707 + }, + { + "epoch": 0.26708, + "grad_norm": 0.9503424849109091, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 26708 + }, + { + "epoch": 0.26709, + "grad_norm": 1.0468737365744245, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 26709 + }, + { + "epoch": 0.2671, + "grad_norm": 1.0862347972270558, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 26710 + }, + { + "epoch": 0.26711, + "grad_norm": 1.1980162660772926, + "learning_rate": 0.003, + "loss": 4.053, + "step": 26711 + }, + { + "epoch": 0.26712, + "grad_norm": 1.0877412051087938, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 26712 + }, + { + "epoch": 0.26713, + "grad_norm": 1.0915680594639652, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 26713 + }, + { + "epoch": 0.26714, + "grad_norm": 1.1313681053352007, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 26714 + }, + { + "epoch": 0.26715, + "grad_norm": 1.1085663424828875, + "learning_rate": 0.003, + "loss": 4.048, + "step": 26715 + }, + { + "epoch": 0.26716, + "grad_norm": 1.0587943797345314, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 26716 + }, + { + "epoch": 0.26717, + "grad_norm": 1.0902373523094848, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 26717 + }, + { + "epoch": 0.26718, + "grad_norm": 1.1228721633785397, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 26718 + }, + { + "epoch": 0.26719, + "grad_norm": 0.9062964474538887, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 26719 + }, + { + "epoch": 0.2672, + "grad_norm": 0.9973972192435997, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 26720 + }, + { + "epoch": 0.26721, + "grad_norm": 1.267416849302713, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 26721 + }, + { + "epoch": 0.26722, + "grad_norm": 1.0245971918913632, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 26722 + }, + { + "epoch": 0.26723, + "grad_norm": 1.0691644688096333, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 26723 + }, + { + "epoch": 0.26724, + "grad_norm": 1.0988505586021222, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 26724 + }, + { + "epoch": 0.26725, + "grad_norm": 0.9852894499003052, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 26725 + }, + { + "epoch": 0.26726, + "grad_norm": 0.9471311514219352, + "learning_rate": 0.003, + "loss": 4.077, + "step": 26726 + }, + { + "epoch": 0.26727, + "grad_norm": 1.2526811090532797, + "learning_rate": 0.003, + "loss": 4.055, + "step": 26727 + }, + { + "epoch": 0.26728, + "grad_norm": 1.1791966549506703, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 26728 + }, + { + "epoch": 0.26729, + "grad_norm": 1.1493646141247256, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 26729 + }, + { + "epoch": 0.2673, + "grad_norm": 0.9363480972558187, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 26730 + }, + { + "epoch": 0.26731, + "grad_norm": 1.0739110986556857, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 26731 + }, + { + "epoch": 0.26732, + "grad_norm": 1.1588305212805645, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 26732 + }, + { + "epoch": 0.26733, + "grad_norm": 0.8759749386484998, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 26733 + }, + { + "epoch": 0.26734, + "grad_norm": 0.9804079635357517, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 26734 + }, + { + "epoch": 0.26735, + "grad_norm": 1.0781250841510372, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 26735 + }, + { + "epoch": 0.26736, + "grad_norm": 1.3363382522185936, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 26736 + }, + { + "epoch": 0.26737, + "grad_norm": 0.933610908985995, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 26737 + }, + { + "epoch": 0.26738, + "grad_norm": 1.1122106882831753, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 26738 + }, + { + "epoch": 0.26739, + "grad_norm": 1.057509929918589, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 26739 + }, + { + "epoch": 0.2674, + "grad_norm": 1.2276952628104605, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 26740 + }, + { + "epoch": 0.26741, + "grad_norm": 1.06971175355527, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 26741 + }, + { + "epoch": 0.26742, + "grad_norm": 1.1786001703266455, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 26742 + }, + { + "epoch": 0.26743, + "grad_norm": 1.1026429013979633, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 26743 + }, + { + "epoch": 0.26744, + "grad_norm": 1.154642306526144, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 26744 + }, + { + "epoch": 0.26745, + "grad_norm": 1.0672879837777107, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 26745 + }, + { + "epoch": 0.26746, + "grad_norm": 1.1393507872841764, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 26746 + }, + { + "epoch": 0.26747, + "grad_norm": 1.009466012383764, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 26747 + }, + { + "epoch": 0.26748, + "grad_norm": 1.160446313477599, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 26748 + }, + { + "epoch": 0.26749, + "grad_norm": 0.9672233453560191, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 26749 + }, + { + "epoch": 0.2675, + "grad_norm": 0.9416178364060769, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 26750 + }, + { + "epoch": 0.26751, + "grad_norm": 0.9329108401775541, + "learning_rate": 0.003, + "loss": 4.056, + "step": 26751 + }, + { + "epoch": 0.26752, + "grad_norm": 0.9246494154173391, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 26752 + }, + { + "epoch": 0.26753, + "grad_norm": 1.0257749306765482, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 26753 + }, + { + "epoch": 0.26754, + "grad_norm": 1.2706044370436589, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 26754 + }, + { + "epoch": 0.26755, + "grad_norm": 1.1022846133582012, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 26755 + }, + { + "epoch": 0.26756, + "grad_norm": 1.0883148113479117, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 26756 + }, + { + "epoch": 0.26757, + "grad_norm": 1.0115616595706358, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 26757 + }, + { + "epoch": 0.26758, + "grad_norm": 1.0056776550686617, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 26758 + }, + { + "epoch": 0.26759, + "grad_norm": 0.968966387418883, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 26759 + }, + { + "epoch": 0.2676, + "grad_norm": 1.0002664477154204, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 26760 + }, + { + "epoch": 0.26761, + "grad_norm": 1.0692157618012972, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 26761 + }, + { + "epoch": 0.26762, + "grad_norm": 1.0261510330812786, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 26762 + }, + { + "epoch": 0.26763, + "grad_norm": 1.004580041022209, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 26763 + }, + { + "epoch": 0.26764, + "grad_norm": 1.1228339824829938, + "learning_rate": 0.003, + "loss": 4.0905, + "step": 26764 + }, + { + "epoch": 0.26765, + "grad_norm": 0.8331741135406662, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 26765 + }, + { + "epoch": 0.26766, + "grad_norm": 0.8001862459288873, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 26766 + }, + { + "epoch": 0.26767, + "grad_norm": 0.948080886002724, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 26767 + }, + { + "epoch": 0.26768, + "grad_norm": 1.4065757860164954, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 26768 + }, + { + "epoch": 0.26769, + "grad_norm": 1.2471851511180012, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 26769 + }, + { + "epoch": 0.2677, + "grad_norm": 0.9826319773466633, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 26770 + }, + { + "epoch": 0.26771, + "grad_norm": 1.1518803584237747, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 26771 + }, + { + "epoch": 0.26772, + "grad_norm": 1.3455707603241194, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 26772 + }, + { + "epoch": 0.26773, + "grad_norm": 0.8380558056168175, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 26773 + }, + { + "epoch": 0.26774, + "grad_norm": 0.7859366655504083, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 26774 + }, + { + "epoch": 0.26775, + "grad_norm": 0.8896745146365694, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 26775 + }, + { + "epoch": 0.26776, + "grad_norm": 1.1843409363124329, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 26776 + }, + { + "epoch": 0.26777, + "grad_norm": 1.1904654340105227, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 26777 + }, + { + "epoch": 0.26778, + "grad_norm": 1.094724299777232, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 26778 + }, + { + "epoch": 0.26779, + "grad_norm": 1.1517481290500604, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 26779 + }, + { + "epoch": 0.2678, + "grad_norm": 0.9192352074061683, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 26780 + }, + { + "epoch": 0.26781, + "grad_norm": 0.922445205927564, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 26781 + }, + { + "epoch": 0.26782, + "grad_norm": 1.0070528033020345, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 26782 + }, + { + "epoch": 0.26783, + "grad_norm": 0.9955833016689616, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 26783 + }, + { + "epoch": 0.26784, + "grad_norm": 1.019459665619818, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 26784 + }, + { + "epoch": 0.26785, + "grad_norm": 1.0012165667996145, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 26785 + }, + { + "epoch": 0.26786, + "grad_norm": 1.061816619489918, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 26786 + }, + { + "epoch": 0.26787, + "grad_norm": 1.036575833500414, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 26787 + }, + { + "epoch": 0.26788, + "grad_norm": 1.2231706509941416, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 26788 + }, + { + "epoch": 0.26789, + "grad_norm": 1.0784385703074106, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 26789 + }, + { + "epoch": 0.2679, + "grad_norm": 1.0691516364127822, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 26790 + }, + { + "epoch": 0.26791, + "grad_norm": 1.029978674295027, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 26791 + }, + { + "epoch": 0.26792, + "grad_norm": 1.1797574607716657, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 26792 + }, + { + "epoch": 0.26793, + "grad_norm": 0.9935730247867368, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 26793 + }, + { + "epoch": 0.26794, + "grad_norm": 1.2497413664328314, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 26794 + }, + { + "epoch": 0.26795, + "grad_norm": 0.8752007744132113, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 26795 + }, + { + "epoch": 0.26796, + "grad_norm": 0.9920630921070501, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 26796 + }, + { + "epoch": 0.26797, + "grad_norm": 1.0481474193654892, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 26797 + }, + { + "epoch": 0.26798, + "grad_norm": 1.1023093733573888, + "learning_rate": 0.003, + "loss": 4.044, + "step": 26798 + }, + { + "epoch": 0.26799, + "grad_norm": 1.2508940618457378, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 26799 + }, + { + "epoch": 0.268, + "grad_norm": 0.8872164161752644, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 26800 + }, + { + "epoch": 0.26801, + "grad_norm": 1.076272605516021, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 26801 + }, + { + "epoch": 0.26802, + "grad_norm": 1.1516148774284887, + "learning_rate": 0.003, + "loss": 4.033, + "step": 26802 + }, + { + "epoch": 0.26803, + "grad_norm": 1.0284725305922229, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 26803 + }, + { + "epoch": 0.26804, + "grad_norm": 1.1140950260435714, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 26804 + }, + { + "epoch": 0.26805, + "grad_norm": 1.1121934608013093, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 26805 + }, + { + "epoch": 0.26806, + "grad_norm": 1.1903148987109626, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 26806 + }, + { + "epoch": 0.26807, + "grad_norm": 0.8530707579249848, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 26807 + }, + { + "epoch": 0.26808, + "grad_norm": 1.2479326609657162, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 26808 + }, + { + "epoch": 0.26809, + "grad_norm": 1.4966208472140052, + "learning_rate": 0.003, + "loss": 4.067, + "step": 26809 + }, + { + "epoch": 0.2681, + "grad_norm": 0.709271087981489, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 26810 + }, + { + "epoch": 0.26811, + "grad_norm": 0.8568317620885233, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 26811 + }, + { + "epoch": 0.26812, + "grad_norm": 0.9715796228541116, + "learning_rate": 0.003, + "loss": 4.047, + "step": 26812 + }, + { + "epoch": 0.26813, + "grad_norm": 0.9835689129672457, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 26813 + }, + { + "epoch": 0.26814, + "grad_norm": 1.3271541121026407, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 26814 + }, + { + "epoch": 0.26815, + "grad_norm": 0.9072648063227641, + "learning_rate": 0.003, + "loss": 4.062, + "step": 26815 + }, + { + "epoch": 0.26816, + "grad_norm": 0.9928163140148173, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 26816 + }, + { + "epoch": 0.26817, + "grad_norm": 1.1463048865898406, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 26817 + }, + { + "epoch": 0.26818, + "grad_norm": 1.1165920047226283, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 26818 + }, + { + "epoch": 0.26819, + "grad_norm": 1.078979626055762, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 26819 + }, + { + "epoch": 0.2682, + "grad_norm": 1.014026171836722, + "learning_rate": 0.003, + "loss": 4.05, + "step": 26820 + }, + { + "epoch": 0.26821, + "grad_norm": 1.169209933888162, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 26821 + }, + { + "epoch": 0.26822, + "grad_norm": 0.9500198279588572, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 26822 + }, + { + "epoch": 0.26823, + "grad_norm": 0.9190413717452715, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 26823 + }, + { + "epoch": 0.26824, + "grad_norm": 1.0858553926566017, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 26824 + }, + { + "epoch": 0.26825, + "grad_norm": 0.9926146406656006, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 26825 + }, + { + "epoch": 0.26826, + "grad_norm": 1.0693726750599382, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 26826 + }, + { + "epoch": 0.26827, + "grad_norm": 0.9867254407311024, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 26827 + }, + { + "epoch": 0.26828, + "grad_norm": 1.2248391966824101, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 26828 + }, + { + "epoch": 0.26829, + "grad_norm": 1.1515887137693166, + "learning_rate": 0.003, + "loss": 4.05, + "step": 26829 + }, + { + "epoch": 0.2683, + "grad_norm": 1.2565676588074208, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 26830 + }, + { + "epoch": 0.26831, + "grad_norm": 0.8656605102236276, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 26831 + }, + { + "epoch": 0.26832, + "grad_norm": 0.8732187008053546, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 26832 + }, + { + "epoch": 0.26833, + "grad_norm": 1.0111523803027445, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 26833 + }, + { + "epoch": 0.26834, + "grad_norm": 0.9645941607291166, + "learning_rate": 0.003, + "loss": 4.073, + "step": 26834 + }, + { + "epoch": 0.26835, + "grad_norm": 1.1864513026234307, + "learning_rate": 0.003, + "loss": 4.061, + "step": 26835 + }, + { + "epoch": 0.26836, + "grad_norm": 1.1097615013137792, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 26836 + }, + { + "epoch": 0.26837, + "grad_norm": 0.9530125843767526, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 26837 + }, + { + "epoch": 0.26838, + "grad_norm": 0.9016399674609793, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 26838 + }, + { + "epoch": 0.26839, + "grad_norm": 0.7767524421166492, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 26839 + }, + { + "epoch": 0.2684, + "grad_norm": 0.8774730914939122, + "learning_rate": 0.003, + "loss": 4.063, + "step": 26840 + }, + { + "epoch": 0.26841, + "grad_norm": 0.9798609451735619, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 26841 + }, + { + "epoch": 0.26842, + "grad_norm": 1.0494847359740636, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 26842 + }, + { + "epoch": 0.26843, + "grad_norm": 1.3048670148206494, + "learning_rate": 0.003, + "loss": 4.067, + "step": 26843 + }, + { + "epoch": 0.26844, + "grad_norm": 1.6286680112187852, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 26844 + }, + { + "epoch": 0.26845, + "grad_norm": 0.7683375874830267, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 26845 + }, + { + "epoch": 0.26846, + "grad_norm": 1.1432094534463233, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 26846 + }, + { + "epoch": 0.26847, + "grad_norm": 1.1807250526118498, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 26847 + }, + { + "epoch": 0.26848, + "grad_norm": 1.2253658848184996, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 26848 + }, + { + "epoch": 0.26849, + "grad_norm": 1.2267486277672484, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 26849 + }, + { + "epoch": 0.2685, + "grad_norm": 1.0534199729533695, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 26850 + }, + { + "epoch": 0.26851, + "grad_norm": 0.8331915134416985, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 26851 + }, + { + "epoch": 0.26852, + "grad_norm": 0.9827193995987478, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 26852 + }, + { + "epoch": 0.26853, + "grad_norm": 1.2363583433200258, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 26853 + }, + { + "epoch": 0.26854, + "grad_norm": 0.9491056137598552, + "learning_rate": 0.003, + "loss": 4.029, + "step": 26854 + }, + { + "epoch": 0.26855, + "grad_norm": 1.1247589882729319, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 26855 + }, + { + "epoch": 0.26856, + "grad_norm": 1.0459429376918308, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 26856 + }, + { + "epoch": 0.26857, + "grad_norm": 1.0900866515615701, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 26857 + }, + { + "epoch": 0.26858, + "grad_norm": 0.939068670384764, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 26858 + }, + { + "epoch": 0.26859, + "grad_norm": 1.038602938558867, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 26859 + }, + { + "epoch": 0.2686, + "grad_norm": 1.1664331902074467, + "learning_rate": 0.003, + "loss": 4.029, + "step": 26860 + }, + { + "epoch": 0.26861, + "grad_norm": 1.1032564723540408, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 26861 + }, + { + "epoch": 0.26862, + "grad_norm": 0.9034396134007411, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 26862 + }, + { + "epoch": 0.26863, + "grad_norm": 0.8387537719208761, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 26863 + }, + { + "epoch": 0.26864, + "grad_norm": 0.9844407337379462, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 26864 + }, + { + "epoch": 0.26865, + "grad_norm": 1.1966395218261725, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 26865 + }, + { + "epoch": 0.26866, + "grad_norm": 0.7212679425390363, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 26866 + }, + { + "epoch": 0.26867, + "grad_norm": 0.6387705844229508, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 26867 + }, + { + "epoch": 0.26868, + "grad_norm": 0.8124986708880754, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 26868 + }, + { + "epoch": 0.26869, + "grad_norm": 1.0199729101364285, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 26869 + }, + { + "epoch": 0.2687, + "grad_norm": 1.1510015739778998, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 26870 + }, + { + "epoch": 0.26871, + "grad_norm": 0.8808307108123189, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 26871 + }, + { + "epoch": 0.26872, + "grad_norm": 1.0243352659962677, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 26872 + }, + { + "epoch": 0.26873, + "grad_norm": 1.0973430057405524, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 26873 + }, + { + "epoch": 0.26874, + "grad_norm": 1.0853668063792399, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 26874 + }, + { + "epoch": 0.26875, + "grad_norm": 1.2195153645232064, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 26875 + }, + { + "epoch": 0.26876, + "grad_norm": 1.2769079669102001, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 26876 + }, + { + "epoch": 0.26877, + "grad_norm": 0.9964651450941394, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 26877 + }, + { + "epoch": 0.26878, + "grad_norm": 0.9029781704705343, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 26878 + }, + { + "epoch": 0.26879, + "grad_norm": 1.0775329963247582, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 26879 + }, + { + "epoch": 0.2688, + "grad_norm": 1.2100715231223007, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 26880 + }, + { + "epoch": 0.26881, + "grad_norm": 1.0296931428160687, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 26881 + }, + { + "epoch": 0.26882, + "grad_norm": 1.2435657060063205, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 26882 + }, + { + "epoch": 0.26883, + "grad_norm": 0.9907672872160684, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 26883 + }, + { + "epoch": 0.26884, + "grad_norm": 1.0819619934724567, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 26884 + }, + { + "epoch": 0.26885, + "grad_norm": 1.259986856328494, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 26885 + }, + { + "epoch": 0.26886, + "grad_norm": 0.9655979431238378, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 26886 + }, + { + "epoch": 0.26887, + "grad_norm": 1.089685512960752, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 26887 + }, + { + "epoch": 0.26888, + "grad_norm": 1.046028626826351, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 26888 + }, + { + "epoch": 0.26889, + "grad_norm": 1.1934954498726378, + "learning_rate": 0.003, + "loss": 4.047, + "step": 26889 + }, + { + "epoch": 0.2689, + "grad_norm": 1.1023532614539762, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 26890 + }, + { + "epoch": 0.26891, + "grad_norm": 1.1178748505268714, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 26891 + }, + { + "epoch": 0.26892, + "grad_norm": 0.9998134214059791, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 26892 + }, + { + "epoch": 0.26893, + "grad_norm": 1.1496082242722625, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 26893 + }, + { + "epoch": 0.26894, + "grad_norm": 0.9817176311417707, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 26894 + }, + { + "epoch": 0.26895, + "grad_norm": 1.0427067964596017, + "learning_rate": 0.003, + "loss": 4.021, + "step": 26895 + }, + { + "epoch": 0.26896, + "grad_norm": 1.0230939287176182, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 26896 + }, + { + "epoch": 0.26897, + "grad_norm": 1.0534828746511065, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 26897 + }, + { + "epoch": 0.26898, + "grad_norm": 0.9626246512959473, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 26898 + }, + { + "epoch": 0.26899, + "grad_norm": 1.0233692922570679, + "learning_rate": 0.003, + "loss": 4.06, + "step": 26899 + }, + { + "epoch": 0.269, + "grad_norm": 0.9443072043540855, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 26900 + }, + { + "epoch": 0.26901, + "grad_norm": 1.0543151227661465, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 26901 + }, + { + "epoch": 0.26902, + "grad_norm": 1.002292986612039, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 26902 + }, + { + "epoch": 0.26903, + "grad_norm": 1.2611503309152357, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 26903 + }, + { + "epoch": 0.26904, + "grad_norm": 1.0893508783664168, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 26904 + }, + { + "epoch": 0.26905, + "grad_norm": 1.0914974866733815, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 26905 + }, + { + "epoch": 0.26906, + "grad_norm": 0.9869016078985033, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 26906 + }, + { + "epoch": 0.26907, + "grad_norm": 0.9465523328861647, + "learning_rate": 0.003, + "loss": 4.023, + "step": 26907 + }, + { + "epoch": 0.26908, + "grad_norm": 1.0480097003540365, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 26908 + }, + { + "epoch": 0.26909, + "grad_norm": 1.0546388032213472, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 26909 + }, + { + "epoch": 0.2691, + "grad_norm": 1.1182120722120392, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 26910 + }, + { + "epoch": 0.26911, + "grad_norm": 1.000084466544105, + "learning_rate": 0.003, + "loss": 4.0903, + "step": 26911 + }, + { + "epoch": 0.26912, + "grad_norm": 1.1338187513689155, + "learning_rate": 0.003, + "loss": 4.051, + "step": 26912 + }, + { + "epoch": 0.26913, + "grad_norm": 1.0657255057071753, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 26913 + }, + { + "epoch": 0.26914, + "grad_norm": 1.2467110769254637, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 26914 + }, + { + "epoch": 0.26915, + "grad_norm": 0.8444306530155046, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 26915 + }, + { + "epoch": 0.26916, + "grad_norm": 0.8854145537923664, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 26916 + }, + { + "epoch": 0.26917, + "grad_norm": 1.1625883279066864, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 26917 + }, + { + "epoch": 0.26918, + "grad_norm": 1.3352909160424011, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 26918 + }, + { + "epoch": 0.26919, + "grad_norm": 0.7901703075807266, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 26919 + }, + { + "epoch": 0.2692, + "grad_norm": 0.8732023771739535, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 26920 + }, + { + "epoch": 0.26921, + "grad_norm": 1.0045342992980415, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 26921 + }, + { + "epoch": 0.26922, + "grad_norm": 1.2822645165775968, + "learning_rate": 0.003, + "loss": 4.0985, + "step": 26922 + }, + { + "epoch": 0.26923, + "grad_norm": 0.9285745246831076, + "learning_rate": 0.003, + "loss": 4.065, + "step": 26923 + }, + { + "epoch": 0.26924, + "grad_norm": 1.1645326510629241, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 26924 + }, + { + "epoch": 0.26925, + "grad_norm": 1.1286594386668396, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 26925 + }, + { + "epoch": 0.26926, + "grad_norm": 1.1746488924631804, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 26926 + }, + { + "epoch": 0.26927, + "grad_norm": 1.0967441282052584, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 26927 + }, + { + "epoch": 0.26928, + "grad_norm": 1.0427929377204348, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 26928 + }, + { + "epoch": 0.26929, + "grad_norm": 1.0754783276388384, + "learning_rate": 0.003, + "loss": 4.075, + "step": 26929 + }, + { + "epoch": 0.2693, + "grad_norm": 1.111793081708011, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 26930 + }, + { + "epoch": 0.26931, + "grad_norm": 1.353077306582076, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 26931 + }, + { + "epoch": 0.26932, + "grad_norm": 0.8915338394326969, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 26932 + }, + { + "epoch": 0.26933, + "grad_norm": 1.114353473117297, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 26933 + }, + { + "epoch": 0.26934, + "grad_norm": 1.1211211797275693, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 26934 + }, + { + "epoch": 0.26935, + "grad_norm": 1.0321046525675253, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 26935 + }, + { + "epoch": 0.26936, + "grad_norm": 1.0381781745384666, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 26936 + }, + { + "epoch": 0.26937, + "grad_norm": 1.2472399282575746, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 26937 + }, + { + "epoch": 0.26938, + "grad_norm": 1.022052804836978, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 26938 + }, + { + "epoch": 0.26939, + "grad_norm": 0.9490371613709448, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 26939 + }, + { + "epoch": 0.2694, + "grad_norm": 1.0724717336959508, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 26940 + }, + { + "epoch": 0.26941, + "grad_norm": 1.1068045177833339, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 26941 + }, + { + "epoch": 0.26942, + "grad_norm": 0.7354262627176481, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 26942 + }, + { + "epoch": 0.26943, + "grad_norm": 0.7418310857452417, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 26943 + }, + { + "epoch": 0.26944, + "grad_norm": 0.8631947206238432, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 26944 + }, + { + "epoch": 0.26945, + "grad_norm": 1.0921755012897556, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 26945 + }, + { + "epoch": 0.26946, + "grad_norm": 1.1121940725881616, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 26946 + }, + { + "epoch": 0.26947, + "grad_norm": 0.9546527438180605, + "learning_rate": 0.003, + "loss": 4.049, + "step": 26947 + }, + { + "epoch": 0.26948, + "grad_norm": 1.1192965091117841, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 26948 + }, + { + "epoch": 0.26949, + "grad_norm": 0.9545633103940946, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 26949 + }, + { + "epoch": 0.2695, + "grad_norm": 1.0976227127062619, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 26950 + }, + { + "epoch": 0.26951, + "grad_norm": 1.026211788518591, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 26951 + }, + { + "epoch": 0.26952, + "grad_norm": 1.1397554875597278, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 26952 + }, + { + "epoch": 0.26953, + "grad_norm": 1.0367186338383363, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 26953 + }, + { + "epoch": 0.26954, + "grad_norm": 1.191641608053688, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 26954 + }, + { + "epoch": 0.26955, + "grad_norm": 1.067771735907857, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 26955 + }, + { + "epoch": 0.26956, + "grad_norm": 0.9346288690904685, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 26956 + }, + { + "epoch": 0.26957, + "grad_norm": 0.9368058222145376, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 26957 + }, + { + "epoch": 0.26958, + "grad_norm": 1.3840400506358859, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 26958 + }, + { + "epoch": 0.26959, + "grad_norm": 1.136332736607424, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 26959 + }, + { + "epoch": 0.2696, + "grad_norm": 0.8558284590777315, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 26960 + }, + { + "epoch": 0.26961, + "grad_norm": 0.9440852406099733, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 26961 + }, + { + "epoch": 0.26962, + "grad_norm": 1.2241955292195281, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 26962 + }, + { + "epoch": 0.26963, + "grad_norm": 0.9813647061778265, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 26963 + }, + { + "epoch": 0.26964, + "grad_norm": 1.1862437727590776, + "learning_rate": 0.003, + "loss": 4.09, + "step": 26964 + }, + { + "epoch": 0.26965, + "grad_norm": 0.959046773465775, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 26965 + }, + { + "epoch": 0.26966, + "grad_norm": 0.9504142556918675, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 26966 + }, + { + "epoch": 0.26967, + "grad_norm": 0.9803547182211063, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 26967 + }, + { + "epoch": 0.26968, + "grad_norm": 1.2006126050448411, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 26968 + }, + { + "epoch": 0.26969, + "grad_norm": 0.8603349183139753, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 26969 + }, + { + "epoch": 0.2697, + "grad_norm": 0.9234168888855087, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 26970 + }, + { + "epoch": 0.26971, + "grad_norm": 0.9619791802128066, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 26971 + }, + { + "epoch": 0.26972, + "grad_norm": 0.9940319321544319, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 26972 + }, + { + "epoch": 0.26973, + "grad_norm": 1.2401016655285249, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 26973 + }, + { + "epoch": 0.26974, + "grad_norm": 0.9184799919649014, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 26974 + }, + { + "epoch": 0.26975, + "grad_norm": 0.9683062269552627, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 26975 + }, + { + "epoch": 0.26976, + "grad_norm": 1.3072180926484196, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 26976 + }, + { + "epoch": 0.26977, + "grad_norm": 1.1392030887303215, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 26977 + }, + { + "epoch": 0.26978, + "grad_norm": 0.968495754738615, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 26978 + }, + { + "epoch": 0.26979, + "grad_norm": 1.0495837741061531, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 26979 + }, + { + "epoch": 0.2698, + "grad_norm": 1.196275549388806, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 26980 + }, + { + "epoch": 0.26981, + "grad_norm": 0.7676250831269414, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 26981 + }, + { + "epoch": 0.26982, + "grad_norm": 0.8980305576254827, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 26982 + }, + { + "epoch": 0.26983, + "grad_norm": 0.8807224845563445, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 26983 + }, + { + "epoch": 0.26984, + "grad_norm": 1.0208489750442309, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 26984 + }, + { + "epoch": 0.26985, + "grad_norm": 1.4239110858099697, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 26985 + }, + { + "epoch": 0.26986, + "grad_norm": 1.0988152834689524, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 26986 + }, + { + "epoch": 0.26987, + "grad_norm": 1.023881866178955, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 26987 + }, + { + "epoch": 0.26988, + "grad_norm": 1.0490747411985941, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 26988 + }, + { + "epoch": 0.26989, + "grad_norm": 1.1541222605225236, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 26989 + }, + { + "epoch": 0.2699, + "grad_norm": 1.1429728475525023, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 26990 + }, + { + "epoch": 0.26991, + "grad_norm": 0.8827047747316787, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 26991 + }, + { + "epoch": 0.26992, + "grad_norm": 1.1088109831762936, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 26992 + }, + { + "epoch": 0.26993, + "grad_norm": 1.272239463069594, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 26993 + }, + { + "epoch": 0.26994, + "grad_norm": 0.8319618247889919, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 26994 + }, + { + "epoch": 0.26995, + "grad_norm": 0.8212398884025085, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 26995 + }, + { + "epoch": 0.26996, + "grad_norm": 0.8172360622818606, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 26996 + }, + { + "epoch": 0.26997, + "grad_norm": 1.0228394411258888, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 26997 + }, + { + "epoch": 0.26998, + "grad_norm": 1.2145082460016485, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 26998 + }, + { + "epoch": 0.26999, + "grad_norm": 1.2321694320453611, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 26999 + }, + { + "epoch": 0.27, + "grad_norm": 1.100739230189542, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 27000 + }, + { + "epoch": 0.27001, + "grad_norm": 1.0833931634272886, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 27001 + }, + { + "epoch": 0.27002, + "grad_norm": 1.203143520576742, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 27002 + }, + { + "epoch": 0.27003, + "grad_norm": 0.935598485392021, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 27003 + }, + { + "epoch": 0.27004, + "grad_norm": 1.0733447497125437, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 27004 + }, + { + "epoch": 0.27005, + "grad_norm": 1.162480746548299, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 27005 + }, + { + "epoch": 0.27006, + "grad_norm": 0.9795762555573075, + "learning_rate": 0.003, + "loss": 4.041, + "step": 27006 + }, + { + "epoch": 0.27007, + "grad_norm": 1.0696971027463236, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 27007 + }, + { + "epoch": 0.27008, + "grad_norm": 1.1306728171377893, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 27008 + }, + { + "epoch": 0.27009, + "grad_norm": 1.0848975982331621, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 27009 + }, + { + "epoch": 0.2701, + "grad_norm": 0.9495886106445924, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 27010 + }, + { + "epoch": 0.27011, + "grad_norm": 0.8892904331226078, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 27011 + }, + { + "epoch": 0.27012, + "grad_norm": 0.968872766932089, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 27012 + }, + { + "epoch": 0.27013, + "grad_norm": 1.2951287864998708, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 27013 + }, + { + "epoch": 0.27014, + "grad_norm": 1.1952923528564927, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 27014 + }, + { + "epoch": 0.27015, + "grad_norm": 1.080322215306843, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 27015 + }, + { + "epoch": 0.27016, + "grad_norm": 1.2801267297587986, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 27016 + }, + { + "epoch": 0.27017, + "grad_norm": 1.0321630320186324, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 27017 + }, + { + "epoch": 0.27018, + "grad_norm": 1.3133854728594305, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 27018 + }, + { + "epoch": 0.27019, + "grad_norm": 0.9106868740668254, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 27019 + }, + { + "epoch": 0.2702, + "grad_norm": 0.9194716825498901, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 27020 + }, + { + "epoch": 0.27021, + "grad_norm": 0.9008649340642071, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 27021 + }, + { + "epoch": 0.27022, + "grad_norm": 1.1458023741257566, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 27022 + }, + { + "epoch": 0.27023, + "grad_norm": 1.1459954825626657, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 27023 + }, + { + "epoch": 0.27024, + "grad_norm": 1.0632942226382438, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 27024 + }, + { + "epoch": 0.27025, + "grad_norm": 1.138752778410995, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 27025 + }, + { + "epoch": 0.27026, + "grad_norm": 1.1596732352296144, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 27026 + }, + { + "epoch": 0.27027, + "grad_norm": 0.9601346472296097, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 27027 + }, + { + "epoch": 0.27028, + "grad_norm": 1.020753976452519, + "learning_rate": 0.003, + "loss": 4.045, + "step": 27028 + }, + { + "epoch": 0.27029, + "grad_norm": 1.227116465350458, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 27029 + }, + { + "epoch": 0.2703, + "grad_norm": 1.0051487961313788, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 27030 + }, + { + "epoch": 0.27031, + "grad_norm": 1.0974653749745154, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 27031 + }, + { + "epoch": 0.27032, + "grad_norm": 1.0527056061330233, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 27032 + }, + { + "epoch": 0.27033, + "grad_norm": 1.157996951653347, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 27033 + }, + { + "epoch": 0.27034, + "grad_norm": 0.8911541105846905, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 27034 + }, + { + "epoch": 0.27035, + "grad_norm": 1.0237580353684461, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 27035 + }, + { + "epoch": 0.27036, + "grad_norm": 1.0891576701558114, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 27036 + }, + { + "epoch": 0.27037, + "grad_norm": 1.190460891921895, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 27037 + }, + { + "epoch": 0.27038, + "grad_norm": 1.0703344590559152, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 27038 + }, + { + "epoch": 0.27039, + "grad_norm": 1.2517599113569569, + "learning_rate": 0.003, + "loss": 4.057, + "step": 27039 + }, + { + "epoch": 0.2704, + "grad_norm": 1.0375219840904055, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 27040 + }, + { + "epoch": 0.27041, + "grad_norm": 1.061965871537981, + "learning_rate": 0.003, + "loss": 4.04, + "step": 27041 + }, + { + "epoch": 0.27042, + "grad_norm": 1.0088335324188864, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 27042 + }, + { + "epoch": 0.27043, + "grad_norm": 0.8974689296749301, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 27043 + }, + { + "epoch": 0.27044, + "grad_norm": 0.9934620472910828, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 27044 + }, + { + "epoch": 0.27045, + "grad_norm": 1.271614120460686, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 27045 + }, + { + "epoch": 0.27046, + "grad_norm": 0.9366371035965921, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 27046 + }, + { + "epoch": 0.27047, + "grad_norm": 0.9237633700095224, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 27047 + }, + { + "epoch": 0.27048, + "grad_norm": 0.9436312633534497, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 27048 + }, + { + "epoch": 0.27049, + "grad_norm": 1.2447613969587554, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 27049 + }, + { + "epoch": 0.2705, + "grad_norm": 0.8917361350097879, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 27050 + }, + { + "epoch": 0.27051, + "grad_norm": 0.9992760658215937, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 27051 + }, + { + "epoch": 0.27052, + "grad_norm": 1.303386635715294, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 27052 + }, + { + "epoch": 0.27053, + "grad_norm": 0.9830300940995337, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 27053 + }, + { + "epoch": 0.27054, + "grad_norm": 1.2660959187807415, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 27054 + }, + { + "epoch": 0.27055, + "grad_norm": 0.9872915505858267, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 27055 + }, + { + "epoch": 0.27056, + "grad_norm": 1.3308076761200345, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 27056 + }, + { + "epoch": 0.27057, + "grad_norm": 0.9070697482786427, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 27057 + }, + { + "epoch": 0.27058, + "grad_norm": 0.8805133936375608, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 27058 + }, + { + "epoch": 0.27059, + "grad_norm": 0.9504905241199003, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 27059 + }, + { + "epoch": 0.2706, + "grad_norm": 1.2201694341388478, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 27060 + }, + { + "epoch": 0.27061, + "grad_norm": 1.075359807501724, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 27061 + }, + { + "epoch": 0.27062, + "grad_norm": 1.075856416136065, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 27062 + }, + { + "epoch": 0.27063, + "grad_norm": 1.1191602519230897, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 27063 + }, + { + "epoch": 0.27064, + "grad_norm": 1.198313944018119, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 27064 + }, + { + "epoch": 0.27065, + "grad_norm": 0.9746665446790534, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 27065 + }, + { + "epoch": 0.27066, + "grad_norm": 0.9432309929960785, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 27066 + }, + { + "epoch": 0.27067, + "grad_norm": 1.3562051972221247, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 27067 + }, + { + "epoch": 0.27068, + "grad_norm": 1.2268301714604364, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 27068 + }, + { + "epoch": 0.27069, + "grad_norm": 0.8664542391320205, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 27069 + }, + { + "epoch": 0.2707, + "grad_norm": 0.8084375452387125, + "learning_rate": 0.003, + "loss": 4.075, + "step": 27070 + }, + { + "epoch": 0.27071, + "grad_norm": 0.9203801883160762, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 27071 + }, + { + "epoch": 0.27072, + "grad_norm": 0.9411009616937467, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 27072 + }, + { + "epoch": 0.27073, + "grad_norm": 1.0260475278020875, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 27073 + }, + { + "epoch": 0.27074, + "grad_norm": 1.246964325559167, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 27074 + }, + { + "epoch": 0.27075, + "grad_norm": 0.9256219867430165, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 27075 + }, + { + "epoch": 0.27076, + "grad_norm": 1.270628549184203, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 27076 + }, + { + "epoch": 0.27077, + "grad_norm": 1.0113651775070502, + "learning_rate": 0.003, + "loss": 4.04, + "step": 27077 + }, + { + "epoch": 0.27078, + "grad_norm": 1.0425770763104698, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 27078 + }, + { + "epoch": 0.27079, + "grad_norm": 1.1220179138224724, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 27079 + }, + { + "epoch": 0.2708, + "grad_norm": 1.0367942149015041, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 27080 + }, + { + "epoch": 0.27081, + "grad_norm": 1.271928973978982, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 27081 + }, + { + "epoch": 0.27082, + "grad_norm": 0.9128296245204055, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 27082 + }, + { + "epoch": 0.27083, + "grad_norm": 1.1577705171210306, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 27083 + }, + { + "epoch": 0.27084, + "grad_norm": 1.2185472777058801, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 27084 + }, + { + "epoch": 0.27085, + "grad_norm": 1.2097272808398454, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 27085 + }, + { + "epoch": 0.27086, + "grad_norm": 1.3814600317824917, + "learning_rate": 0.003, + "loss": 4.066, + "step": 27086 + }, + { + "epoch": 0.27087, + "grad_norm": 1.0358985577852773, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 27087 + }, + { + "epoch": 0.27088, + "grad_norm": 0.9805948789909763, + "learning_rate": 0.003, + "loss": 4.085, + "step": 27088 + }, + { + "epoch": 0.27089, + "grad_norm": 0.9101679632117872, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 27089 + }, + { + "epoch": 0.2709, + "grad_norm": 0.9402535728646509, + "learning_rate": 0.003, + "loss": 4.056, + "step": 27090 + }, + { + "epoch": 0.27091, + "grad_norm": 1.0500722080679645, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 27091 + }, + { + "epoch": 0.27092, + "grad_norm": 1.1495587749126444, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 27092 + }, + { + "epoch": 0.27093, + "grad_norm": 1.1129126640673446, + "learning_rate": 0.003, + "loss": 4.035, + "step": 27093 + }, + { + "epoch": 0.27094, + "grad_norm": 1.125943106553163, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 27094 + }, + { + "epoch": 0.27095, + "grad_norm": 1.0661162796252037, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 27095 + }, + { + "epoch": 0.27096, + "grad_norm": 1.1907373887345851, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 27096 + }, + { + "epoch": 0.27097, + "grad_norm": 1.1010596510122879, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 27097 + }, + { + "epoch": 0.27098, + "grad_norm": 1.1401255206192598, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 27098 + }, + { + "epoch": 0.27099, + "grad_norm": 0.8947038145105125, + "learning_rate": 0.003, + "loss": 4.071, + "step": 27099 + }, + { + "epoch": 0.271, + "grad_norm": 0.9353861339838417, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 27100 + }, + { + "epoch": 0.27101, + "grad_norm": 1.0581719248959909, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 27101 + }, + { + "epoch": 0.27102, + "grad_norm": 1.1915222013262798, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 27102 + }, + { + "epoch": 0.27103, + "grad_norm": 0.9348925813580471, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 27103 + }, + { + "epoch": 0.27104, + "grad_norm": 1.2006195284866927, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 27104 + }, + { + "epoch": 0.27105, + "grad_norm": 0.8829152632029077, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 27105 + }, + { + "epoch": 0.27106, + "grad_norm": 0.90477444129757, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 27106 + }, + { + "epoch": 0.27107, + "grad_norm": 1.142220235983961, + "learning_rate": 0.003, + "loss": 4.073, + "step": 27107 + }, + { + "epoch": 0.27108, + "grad_norm": 1.027874724076729, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 27108 + }, + { + "epoch": 0.27109, + "grad_norm": 1.2014210139612016, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 27109 + }, + { + "epoch": 0.2711, + "grad_norm": 0.9862165346349923, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 27110 + }, + { + "epoch": 0.27111, + "grad_norm": 1.1321891319384025, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 27111 + }, + { + "epoch": 0.27112, + "grad_norm": 1.1192143720274472, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 27112 + }, + { + "epoch": 0.27113, + "grad_norm": 0.9089375006477014, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 27113 + }, + { + "epoch": 0.27114, + "grad_norm": 0.9671288727531214, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 27114 + }, + { + "epoch": 0.27115, + "grad_norm": 1.1833937529914376, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 27115 + }, + { + "epoch": 0.27116, + "grad_norm": 0.9907882390195536, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 27116 + }, + { + "epoch": 0.27117, + "grad_norm": 0.9793968808726624, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 27117 + }, + { + "epoch": 0.27118, + "grad_norm": 1.091427438971601, + "learning_rate": 0.003, + "loss": 4.052, + "step": 27118 + }, + { + "epoch": 0.27119, + "grad_norm": 0.9685140650030735, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 27119 + }, + { + "epoch": 0.2712, + "grad_norm": 1.18643049159646, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 27120 + }, + { + "epoch": 0.27121, + "grad_norm": 0.976388865335396, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 27121 + }, + { + "epoch": 0.27122, + "grad_norm": 0.968872474874116, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 27122 + }, + { + "epoch": 0.27123, + "grad_norm": 1.1244312146474933, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 27123 + }, + { + "epoch": 0.27124, + "grad_norm": 1.0476260047463657, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 27124 + }, + { + "epoch": 0.27125, + "grad_norm": 1.1654492121263356, + "learning_rate": 0.003, + "loss": 4.057, + "step": 27125 + }, + { + "epoch": 0.27126, + "grad_norm": 1.013115617047971, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 27126 + }, + { + "epoch": 0.27127, + "grad_norm": 1.3289936855473292, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 27127 + }, + { + "epoch": 0.27128, + "grad_norm": 0.9856442193792712, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 27128 + }, + { + "epoch": 0.27129, + "grad_norm": 1.1654627928877612, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 27129 + }, + { + "epoch": 0.2713, + "grad_norm": 1.3196862800391267, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 27130 + }, + { + "epoch": 0.27131, + "grad_norm": 1.2110573621115845, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 27131 + }, + { + "epoch": 0.27132, + "grad_norm": 1.0356187094267397, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 27132 + }, + { + "epoch": 0.27133, + "grad_norm": 0.9884175488607941, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 27133 + }, + { + "epoch": 0.27134, + "grad_norm": 1.2272504603520693, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 27134 + }, + { + "epoch": 0.27135, + "grad_norm": 1.162005331188801, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 27135 + }, + { + "epoch": 0.27136, + "grad_norm": 1.1019366864457831, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 27136 + }, + { + "epoch": 0.27137, + "grad_norm": 1.0103677627336478, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 27137 + }, + { + "epoch": 0.27138, + "grad_norm": 1.068083025213866, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 27138 + }, + { + "epoch": 0.27139, + "grad_norm": 0.9875805298933429, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 27139 + }, + { + "epoch": 0.2714, + "grad_norm": 1.0411104248387766, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 27140 + }, + { + "epoch": 0.27141, + "grad_norm": 1.003099042461938, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 27141 + }, + { + "epoch": 0.27142, + "grad_norm": 1.0309190189261879, + "learning_rate": 0.003, + "loss": 4.044, + "step": 27142 + }, + { + "epoch": 0.27143, + "grad_norm": 1.1664362770171566, + "learning_rate": 0.003, + "loss": 4.038, + "step": 27143 + }, + { + "epoch": 0.27144, + "grad_norm": 0.956920726761389, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 27144 + }, + { + "epoch": 0.27145, + "grad_norm": 0.9076921217463881, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 27145 + }, + { + "epoch": 0.27146, + "grad_norm": 1.0624307278117935, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 27146 + }, + { + "epoch": 0.27147, + "grad_norm": 1.2849007216092507, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 27147 + }, + { + "epoch": 0.27148, + "grad_norm": 0.6895266519258929, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 27148 + }, + { + "epoch": 0.27149, + "grad_norm": 0.7750044934889079, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 27149 + }, + { + "epoch": 0.2715, + "grad_norm": 0.8284356596274748, + "learning_rate": 0.003, + "loss": 4.02, + "step": 27150 + }, + { + "epoch": 0.27151, + "grad_norm": 1.172499876850328, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 27151 + }, + { + "epoch": 0.27152, + "grad_norm": 1.0918603931603759, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 27152 + }, + { + "epoch": 0.27153, + "grad_norm": 1.0856384782018769, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 27153 + }, + { + "epoch": 0.27154, + "grad_norm": 1.127105938546042, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 27154 + }, + { + "epoch": 0.27155, + "grad_norm": 0.9977333692606654, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 27155 + }, + { + "epoch": 0.27156, + "grad_norm": 1.2079555228499121, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 27156 + }, + { + "epoch": 0.27157, + "grad_norm": 1.008492010495567, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 27157 + }, + { + "epoch": 0.27158, + "grad_norm": 1.0238110033404633, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 27158 + }, + { + "epoch": 0.27159, + "grad_norm": 1.2159272450805418, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 27159 + }, + { + "epoch": 0.2716, + "grad_norm": 0.8693072232091914, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 27160 + }, + { + "epoch": 0.27161, + "grad_norm": 0.9450108823641081, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 27161 + }, + { + "epoch": 0.27162, + "grad_norm": 1.0874719890913827, + "learning_rate": 0.003, + "loss": 4.028, + "step": 27162 + }, + { + "epoch": 0.27163, + "grad_norm": 1.147314016037995, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 27163 + }, + { + "epoch": 0.27164, + "grad_norm": 1.0552670052095194, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 27164 + }, + { + "epoch": 0.27165, + "grad_norm": 1.1529492364368759, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 27165 + }, + { + "epoch": 0.27166, + "grad_norm": 0.9894276800800359, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 27166 + }, + { + "epoch": 0.27167, + "grad_norm": 1.0109478279853585, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 27167 + }, + { + "epoch": 0.27168, + "grad_norm": 1.2155433823372226, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 27168 + }, + { + "epoch": 0.27169, + "grad_norm": 0.9771232902236732, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 27169 + }, + { + "epoch": 0.2717, + "grad_norm": 1.1869196190060642, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 27170 + }, + { + "epoch": 0.27171, + "grad_norm": 1.0290609954555252, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 27171 + }, + { + "epoch": 0.27172, + "grad_norm": 1.2692788558967636, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 27172 + }, + { + "epoch": 0.27173, + "grad_norm": 0.9490833405106597, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 27173 + }, + { + "epoch": 0.27174, + "grad_norm": 0.9801546064819542, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 27174 + }, + { + "epoch": 0.27175, + "grad_norm": 1.1698392968373772, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 27175 + }, + { + "epoch": 0.27176, + "grad_norm": 1.107605774785907, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 27176 + }, + { + "epoch": 0.27177, + "grad_norm": 0.9818987737473076, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 27177 + }, + { + "epoch": 0.27178, + "grad_norm": 1.1908890566303072, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 27178 + }, + { + "epoch": 0.27179, + "grad_norm": 1.0203604406114495, + "learning_rate": 0.003, + "loss": 4.0884, + "step": 27179 + }, + { + "epoch": 0.2718, + "grad_norm": 1.0738624938426193, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 27180 + }, + { + "epoch": 0.27181, + "grad_norm": 1.0054197701583756, + "learning_rate": 0.003, + "loss": 4.076, + "step": 27181 + }, + { + "epoch": 0.27182, + "grad_norm": 1.3020694790589058, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 27182 + }, + { + "epoch": 0.27183, + "grad_norm": 0.9145410822730063, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 27183 + }, + { + "epoch": 0.27184, + "grad_norm": 1.1467779705392402, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 27184 + }, + { + "epoch": 0.27185, + "grad_norm": 1.1470439418589344, + "learning_rate": 0.003, + "loss": 4.041, + "step": 27185 + }, + { + "epoch": 0.27186, + "grad_norm": 0.899104755350675, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 27186 + }, + { + "epoch": 0.27187, + "grad_norm": 0.9970156133738393, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 27187 + }, + { + "epoch": 0.27188, + "grad_norm": 1.1325446775343755, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 27188 + }, + { + "epoch": 0.27189, + "grad_norm": 1.060579703988351, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 27189 + }, + { + "epoch": 0.2719, + "grad_norm": 1.1140258089372967, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 27190 + }, + { + "epoch": 0.27191, + "grad_norm": 1.092912179735995, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 27191 + }, + { + "epoch": 0.27192, + "grad_norm": 0.9362235997632866, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 27192 + }, + { + "epoch": 0.27193, + "grad_norm": 1.0681922040151532, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 27193 + }, + { + "epoch": 0.27194, + "grad_norm": 0.9984365032190907, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 27194 + }, + { + "epoch": 0.27195, + "grad_norm": 1.0356659831085335, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 27195 + }, + { + "epoch": 0.27196, + "grad_norm": 1.168731981148666, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 27196 + }, + { + "epoch": 0.27197, + "grad_norm": 0.9437569715854225, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 27197 + }, + { + "epoch": 0.27198, + "grad_norm": 1.222310694920855, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 27198 + }, + { + "epoch": 0.27199, + "grad_norm": 0.9747569851755005, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 27199 + }, + { + "epoch": 0.272, + "grad_norm": 1.039264916423929, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 27200 + }, + { + "epoch": 0.27201, + "grad_norm": 1.1171826971429064, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 27201 + }, + { + "epoch": 0.27202, + "grad_norm": 0.9951133694680617, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 27202 + }, + { + "epoch": 0.27203, + "grad_norm": 1.2239470783644673, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 27203 + }, + { + "epoch": 0.27204, + "grad_norm": 0.9294741897844149, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 27204 + }, + { + "epoch": 0.27205, + "grad_norm": 1.0130726422909109, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 27205 + }, + { + "epoch": 0.27206, + "grad_norm": 1.3154316020637706, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 27206 + }, + { + "epoch": 0.27207, + "grad_norm": 0.8897237937576832, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 27207 + }, + { + "epoch": 0.27208, + "grad_norm": 0.9081506710441876, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 27208 + }, + { + "epoch": 0.27209, + "grad_norm": 0.8836519418233132, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 27209 + }, + { + "epoch": 0.2721, + "grad_norm": 0.8501091531956082, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 27210 + }, + { + "epoch": 0.27211, + "grad_norm": 1.1724227522266035, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 27211 + }, + { + "epoch": 0.27212, + "grad_norm": 1.2576399080449452, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 27212 + }, + { + "epoch": 0.27213, + "grad_norm": 1.0263745799542046, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 27213 + }, + { + "epoch": 0.27214, + "grad_norm": 1.043817244792075, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 27214 + }, + { + "epoch": 0.27215, + "grad_norm": 1.1172593895136156, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 27215 + }, + { + "epoch": 0.27216, + "grad_norm": 1.0258368899449555, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 27216 + }, + { + "epoch": 0.27217, + "grad_norm": 1.169629522995478, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 27217 + }, + { + "epoch": 0.27218, + "grad_norm": 1.081036978082153, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 27218 + }, + { + "epoch": 0.27219, + "grad_norm": 0.9602408170843139, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 27219 + }, + { + "epoch": 0.2722, + "grad_norm": 1.082711168496573, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 27220 + }, + { + "epoch": 0.27221, + "grad_norm": 1.1408259463019235, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 27221 + }, + { + "epoch": 0.27222, + "grad_norm": 0.9734761606739013, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 27222 + }, + { + "epoch": 0.27223, + "grad_norm": 0.9518190634411554, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 27223 + }, + { + "epoch": 0.27224, + "grad_norm": 1.0808759983005414, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 27224 + }, + { + "epoch": 0.27225, + "grad_norm": 1.0643878744529514, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 27225 + }, + { + "epoch": 0.27226, + "grad_norm": 1.1378287271884346, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 27226 + }, + { + "epoch": 0.27227, + "grad_norm": 1.0456754178706589, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 27227 + }, + { + "epoch": 0.27228, + "grad_norm": 1.1227584769383339, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 27228 + }, + { + "epoch": 0.27229, + "grad_norm": 0.9198448509509805, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 27229 + }, + { + "epoch": 0.2723, + "grad_norm": 0.9403612915994095, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 27230 + }, + { + "epoch": 0.27231, + "grad_norm": 1.0264997357837178, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 27231 + }, + { + "epoch": 0.27232, + "grad_norm": 1.0193422946053128, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 27232 + }, + { + "epoch": 0.27233, + "grad_norm": 1.0039209599069796, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 27233 + }, + { + "epoch": 0.27234, + "grad_norm": 0.9426145064283538, + "learning_rate": 0.003, + "loss": 4.042, + "step": 27234 + }, + { + "epoch": 0.27235, + "grad_norm": 1.0898605418594727, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 27235 + }, + { + "epoch": 0.27236, + "grad_norm": 0.8954769074371087, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 27236 + }, + { + "epoch": 0.27237, + "grad_norm": 1.1365897204499027, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 27237 + }, + { + "epoch": 0.27238, + "grad_norm": 1.2672406110960066, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 27238 + }, + { + "epoch": 0.27239, + "grad_norm": 0.9713119733906583, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 27239 + }, + { + "epoch": 0.2724, + "grad_norm": 1.1911638062915029, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 27240 + }, + { + "epoch": 0.27241, + "grad_norm": 1.225456519393676, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 27241 + }, + { + "epoch": 0.27242, + "grad_norm": 0.9669364749809242, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 27242 + }, + { + "epoch": 0.27243, + "grad_norm": 1.022817436462372, + "learning_rate": 0.003, + "loss": 4.057, + "step": 27243 + }, + { + "epoch": 0.27244, + "grad_norm": 1.2116295872863083, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 27244 + }, + { + "epoch": 0.27245, + "grad_norm": 0.8368475012750117, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 27245 + }, + { + "epoch": 0.27246, + "grad_norm": 0.9287566182015128, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 27246 + }, + { + "epoch": 0.27247, + "grad_norm": 1.0643131475184595, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 27247 + }, + { + "epoch": 0.27248, + "grad_norm": 0.9727380063881659, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 27248 + }, + { + "epoch": 0.27249, + "grad_norm": 1.00199529968325, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 27249 + }, + { + "epoch": 0.2725, + "grad_norm": 1.1084196612878303, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 27250 + }, + { + "epoch": 0.27251, + "grad_norm": 0.924368503935326, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 27251 + }, + { + "epoch": 0.27252, + "grad_norm": 0.9529986100637771, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 27252 + }, + { + "epoch": 0.27253, + "grad_norm": 1.026791301367341, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 27253 + }, + { + "epoch": 0.27254, + "grad_norm": 1.1187955541436918, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 27254 + }, + { + "epoch": 0.27255, + "grad_norm": 1.3324504894694966, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 27255 + }, + { + "epoch": 0.27256, + "grad_norm": 1.0363607252077196, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 27256 + }, + { + "epoch": 0.27257, + "grad_norm": 1.3578988209643403, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 27257 + }, + { + "epoch": 0.27258, + "grad_norm": 0.8757670296719375, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 27258 + }, + { + "epoch": 0.27259, + "grad_norm": 0.8770819521388789, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 27259 + }, + { + "epoch": 0.2726, + "grad_norm": 0.9661615378895537, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 27260 + }, + { + "epoch": 0.27261, + "grad_norm": 1.0129213991691208, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 27261 + }, + { + "epoch": 0.27262, + "grad_norm": 1.0963854856046404, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 27262 + }, + { + "epoch": 0.27263, + "grad_norm": 1.144234741460212, + "learning_rate": 0.003, + "loss": 4.079, + "step": 27263 + }, + { + "epoch": 0.27264, + "grad_norm": 0.9459423976084361, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 27264 + }, + { + "epoch": 0.27265, + "grad_norm": 0.9883342479906416, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 27265 + }, + { + "epoch": 0.27266, + "grad_norm": 1.1666078091405885, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 27266 + }, + { + "epoch": 0.27267, + "grad_norm": 1.167826942924025, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 27267 + }, + { + "epoch": 0.27268, + "grad_norm": 1.1418406979599667, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 27268 + }, + { + "epoch": 0.27269, + "grad_norm": 1.0342861665242147, + "learning_rate": 0.003, + "loss": 4.0874, + "step": 27269 + }, + { + "epoch": 0.2727, + "grad_norm": 1.1857788876634612, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 27270 + }, + { + "epoch": 0.27271, + "grad_norm": 1.1047647915267078, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 27271 + }, + { + "epoch": 0.27272, + "grad_norm": 1.057759886479363, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 27272 + }, + { + "epoch": 0.27273, + "grad_norm": 1.1354802414435878, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 27273 + }, + { + "epoch": 0.27274, + "grad_norm": 1.170148064640876, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 27274 + }, + { + "epoch": 0.27275, + "grad_norm": 0.9315550922269127, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 27275 + }, + { + "epoch": 0.27276, + "grad_norm": 0.9101951093462436, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 27276 + }, + { + "epoch": 0.27277, + "grad_norm": 0.9671242860652375, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 27277 + }, + { + "epoch": 0.27278, + "grad_norm": 1.0528542759382589, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 27278 + }, + { + "epoch": 0.27279, + "grad_norm": 1.027658762573179, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 27279 + }, + { + "epoch": 0.2728, + "grad_norm": 1.0863942126482689, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 27280 + }, + { + "epoch": 0.27281, + "grad_norm": 1.0759960266876532, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 27281 + }, + { + "epoch": 0.27282, + "grad_norm": 1.1643883679441525, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 27282 + }, + { + "epoch": 0.27283, + "grad_norm": 1.0474772365239489, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 27283 + }, + { + "epoch": 0.27284, + "grad_norm": 0.9876630137371339, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 27284 + }, + { + "epoch": 0.27285, + "grad_norm": 1.396228563565394, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 27285 + }, + { + "epoch": 0.27286, + "grad_norm": 0.7500639065347421, + "learning_rate": 0.003, + "loss": 4.047, + "step": 27286 + }, + { + "epoch": 0.27287, + "grad_norm": 0.786550786737386, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 27287 + }, + { + "epoch": 0.27288, + "grad_norm": 0.8383404301943667, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 27288 + }, + { + "epoch": 0.27289, + "grad_norm": 1.1364885232962982, + "learning_rate": 0.003, + "loss": 4.064, + "step": 27289 + }, + { + "epoch": 0.2729, + "grad_norm": 1.2551871777663843, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 27290 + }, + { + "epoch": 0.27291, + "grad_norm": 0.9789228439773696, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 27291 + }, + { + "epoch": 0.27292, + "grad_norm": 1.0087599011219832, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 27292 + }, + { + "epoch": 0.27293, + "grad_norm": 1.6352886190908713, + "learning_rate": 0.003, + "loss": 4.043, + "step": 27293 + }, + { + "epoch": 0.27294, + "grad_norm": 0.9657937876682489, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 27294 + }, + { + "epoch": 0.27295, + "grad_norm": 1.0189057509719421, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 27295 + }, + { + "epoch": 0.27296, + "grad_norm": 1.1217469955027763, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 27296 + }, + { + "epoch": 0.27297, + "grad_norm": 1.0247044594540125, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 27297 + }, + { + "epoch": 0.27298, + "grad_norm": 1.12198740730256, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 27298 + }, + { + "epoch": 0.27299, + "grad_norm": 1.0672711197827813, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 27299 + }, + { + "epoch": 0.273, + "grad_norm": 1.2568176607712243, + "learning_rate": 0.003, + "loss": 4.051, + "step": 27300 + }, + { + "epoch": 0.27301, + "grad_norm": 0.8528683637533285, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 27301 + }, + { + "epoch": 0.27302, + "grad_norm": 0.9465963961956881, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 27302 + }, + { + "epoch": 0.27303, + "grad_norm": 1.1878498692122805, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 27303 + }, + { + "epoch": 0.27304, + "grad_norm": 1.0693193230268203, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 27304 + }, + { + "epoch": 0.27305, + "grad_norm": 1.416422333817786, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 27305 + }, + { + "epoch": 0.27306, + "grad_norm": 0.9855152585968002, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 27306 + }, + { + "epoch": 0.27307, + "grad_norm": 1.1336326998508905, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 27307 + }, + { + "epoch": 0.27308, + "grad_norm": 1.1179138766487668, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 27308 + }, + { + "epoch": 0.27309, + "grad_norm": 0.969066714276414, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 27309 + }, + { + "epoch": 0.2731, + "grad_norm": 0.9800320016884712, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 27310 + }, + { + "epoch": 0.27311, + "grad_norm": 0.9674736997233329, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 27311 + }, + { + "epoch": 0.27312, + "grad_norm": 1.041359792630894, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 27312 + }, + { + "epoch": 0.27313, + "grad_norm": 1.2518309364547264, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 27313 + }, + { + "epoch": 0.27314, + "grad_norm": 0.8842777483980588, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 27314 + }, + { + "epoch": 0.27315, + "grad_norm": 0.9082162420938303, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 27315 + }, + { + "epoch": 0.27316, + "grad_norm": 1.042387041107201, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 27316 + }, + { + "epoch": 0.27317, + "grad_norm": 1.106068891853591, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 27317 + }, + { + "epoch": 0.27318, + "grad_norm": 1.0253515865776697, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 27318 + }, + { + "epoch": 0.27319, + "grad_norm": 1.1341091870291793, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 27319 + }, + { + "epoch": 0.2732, + "grad_norm": 0.9294682824552896, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 27320 + }, + { + "epoch": 0.27321, + "grad_norm": 1.3327789589593337, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 27321 + }, + { + "epoch": 0.27322, + "grad_norm": 0.889175504841587, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 27322 + }, + { + "epoch": 0.27323, + "grad_norm": 0.9448476707550837, + "learning_rate": 0.003, + "loss": 4.0874, + "step": 27323 + }, + { + "epoch": 0.27324, + "grad_norm": 1.1645511875571128, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 27324 + }, + { + "epoch": 0.27325, + "grad_norm": 0.906325865670123, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 27325 + }, + { + "epoch": 0.27326, + "grad_norm": 1.031340725055981, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 27326 + }, + { + "epoch": 0.27327, + "grad_norm": 1.1580936682649077, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 27327 + }, + { + "epoch": 0.27328, + "grad_norm": 1.075128254237074, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 27328 + }, + { + "epoch": 0.27329, + "grad_norm": 0.9654088625754963, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 27329 + }, + { + "epoch": 0.2733, + "grad_norm": 1.0561004406040635, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 27330 + }, + { + "epoch": 0.27331, + "grad_norm": 1.236790009860851, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 27331 + }, + { + "epoch": 0.27332, + "grad_norm": 1.038036714289782, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 27332 + }, + { + "epoch": 0.27333, + "grad_norm": 1.2253812286282666, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 27333 + }, + { + "epoch": 0.27334, + "grad_norm": 1.002531042629791, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 27334 + }, + { + "epoch": 0.27335, + "grad_norm": 1.169930305841294, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 27335 + }, + { + "epoch": 0.27336, + "grad_norm": 1.097711240824955, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 27336 + }, + { + "epoch": 0.27337, + "grad_norm": 0.9537625686404074, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 27337 + }, + { + "epoch": 0.27338, + "grad_norm": 1.1384632173712554, + "learning_rate": 0.003, + "loss": 4.049, + "step": 27338 + }, + { + "epoch": 0.27339, + "grad_norm": 1.232436568624425, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 27339 + }, + { + "epoch": 0.2734, + "grad_norm": 1.0227173081896188, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 27340 + }, + { + "epoch": 0.27341, + "grad_norm": 1.2026086171962094, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 27341 + }, + { + "epoch": 0.27342, + "grad_norm": 0.9537655361273377, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 27342 + }, + { + "epoch": 0.27343, + "grad_norm": 0.961707555120919, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 27343 + }, + { + "epoch": 0.27344, + "grad_norm": 1.2028996012115887, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 27344 + }, + { + "epoch": 0.27345, + "grad_norm": 0.9805595480491385, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 27345 + }, + { + "epoch": 0.27346, + "grad_norm": 1.116840756616587, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 27346 + }, + { + "epoch": 0.27347, + "grad_norm": 1.1059014716925368, + "learning_rate": 0.003, + "loss": 4.018, + "step": 27347 + }, + { + "epoch": 0.27348, + "grad_norm": 1.1829235415290582, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 27348 + }, + { + "epoch": 0.27349, + "grad_norm": 0.9832298361286687, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 27349 + }, + { + "epoch": 0.2735, + "grad_norm": 1.1284327036435433, + "learning_rate": 0.003, + "loss": 4.036, + "step": 27350 + }, + { + "epoch": 0.27351, + "grad_norm": 1.1825005884379047, + "learning_rate": 0.003, + "loss": 4.072, + "step": 27351 + }, + { + "epoch": 0.27352, + "grad_norm": 0.9680203200504887, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 27352 + }, + { + "epoch": 0.27353, + "grad_norm": 0.9361493621229655, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 27353 + }, + { + "epoch": 0.27354, + "grad_norm": 0.9902759453125812, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 27354 + }, + { + "epoch": 0.27355, + "grad_norm": 1.20305107861521, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 27355 + }, + { + "epoch": 0.27356, + "grad_norm": 0.895667545104383, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 27356 + }, + { + "epoch": 0.27357, + "grad_norm": 0.9076859840455643, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 27357 + }, + { + "epoch": 0.27358, + "grad_norm": 1.0363929384933055, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 27358 + }, + { + "epoch": 0.27359, + "grad_norm": 1.334425923203938, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 27359 + }, + { + "epoch": 0.2736, + "grad_norm": 0.8111046340832536, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 27360 + }, + { + "epoch": 0.27361, + "grad_norm": 0.7773616390373682, + "learning_rate": 0.003, + "loss": 4.035, + "step": 27361 + }, + { + "epoch": 0.27362, + "grad_norm": 0.8633030566953422, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 27362 + }, + { + "epoch": 0.27363, + "grad_norm": 0.9169295296198525, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 27363 + }, + { + "epoch": 0.27364, + "grad_norm": 1.1363412572417038, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 27364 + }, + { + "epoch": 0.27365, + "grad_norm": 1.287926838469997, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 27365 + }, + { + "epoch": 0.27366, + "grad_norm": 0.9463754582375279, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 27366 + }, + { + "epoch": 0.27367, + "grad_norm": 1.0691033633900509, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 27367 + }, + { + "epoch": 0.27368, + "grad_norm": 1.1962940334116998, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 27368 + }, + { + "epoch": 0.27369, + "grad_norm": 1.0139005754836126, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 27369 + }, + { + "epoch": 0.2737, + "grad_norm": 1.0952496531509608, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 27370 + }, + { + "epoch": 0.27371, + "grad_norm": 1.0174700732098672, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 27371 + }, + { + "epoch": 0.27372, + "grad_norm": 0.9075286056923774, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 27372 + }, + { + "epoch": 0.27373, + "grad_norm": 1.0978523799041016, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 27373 + }, + { + "epoch": 0.27374, + "grad_norm": 1.0639468793692468, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 27374 + }, + { + "epoch": 0.27375, + "grad_norm": 1.149037518323945, + "learning_rate": 0.003, + "loss": 4.0881, + "step": 27375 + }, + { + "epoch": 0.27376, + "grad_norm": 1.0537438600373803, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 27376 + }, + { + "epoch": 0.27377, + "grad_norm": 1.1189841524808992, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 27377 + }, + { + "epoch": 0.27378, + "grad_norm": 1.1595371949385203, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 27378 + }, + { + "epoch": 0.27379, + "grad_norm": 0.948649039278111, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 27379 + }, + { + "epoch": 0.2738, + "grad_norm": 1.2190820134178675, + "learning_rate": 0.003, + "loss": 4.076, + "step": 27380 + }, + { + "epoch": 0.27381, + "grad_norm": 1.0005334793763243, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 27381 + }, + { + "epoch": 0.27382, + "grad_norm": 1.0739429440935444, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 27382 + }, + { + "epoch": 0.27383, + "grad_norm": 1.2041483689754358, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 27383 + }, + { + "epoch": 0.27384, + "grad_norm": 1.1609708152113332, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 27384 + }, + { + "epoch": 0.27385, + "grad_norm": 1.102833106565972, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 27385 + }, + { + "epoch": 0.27386, + "grad_norm": 1.1595796353806094, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 27386 + }, + { + "epoch": 0.27387, + "grad_norm": 1.087262263028011, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 27387 + }, + { + "epoch": 0.27388, + "grad_norm": 1.1157334635372378, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 27388 + }, + { + "epoch": 0.27389, + "grad_norm": 0.9082374352587687, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 27389 + }, + { + "epoch": 0.2739, + "grad_norm": 0.9248594145016937, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 27390 + }, + { + "epoch": 0.27391, + "grad_norm": 0.9220927099263657, + "learning_rate": 0.003, + "loss": 4.033, + "step": 27391 + }, + { + "epoch": 0.27392, + "grad_norm": 1.1024647341114564, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 27392 + }, + { + "epoch": 0.27393, + "grad_norm": 1.2212609246217603, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 27393 + }, + { + "epoch": 0.27394, + "grad_norm": 1.0734500671991785, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 27394 + }, + { + "epoch": 0.27395, + "grad_norm": 1.2257318893933808, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 27395 + }, + { + "epoch": 0.27396, + "grad_norm": 0.9236543239351526, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 27396 + }, + { + "epoch": 0.27397, + "grad_norm": 0.9687319329892784, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 27397 + }, + { + "epoch": 0.27398, + "grad_norm": 1.2142296264817407, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 27398 + }, + { + "epoch": 0.27399, + "grad_norm": 0.9868589740092933, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 27399 + }, + { + "epoch": 0.274, + "grad_norm": 1.0466820838031863, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 27400 + }, + { + "epoch": 0.27401, + "grad_norm": 1.1268250675422566, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 27401 + }, + { + "epoch": 0.27402, + "grad_norm": 0.9805817503810408, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 27402 + }, + { + "epoch": 0.27403, + "grad_norm": 1.1284064918204955, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 27403 + }, + { + "epoch": 0.27404, + "grad_norm": 0.9995810626022, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 27404 + }, + { + "epoch": 0.27405, + "grad_norm": 0.97920096855065, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 27405 + }, + { + "epoch": 0.27406, + "grad_norm": 0.9130992617511767, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 27406 + }, + { + "epoch": 0.27407, + "grad_norm": 1.0997881695629863, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 27407 + }, + { + "epoch": 0.27408, + "grad_norm": 1.2545771003849697, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 27408 + }, + { + "epoch": 0.27409, + "grad_norm": 0.8302824530895875, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 27409 + }, + { + "epoch": 0.2741, + "grad_norm": 0.8010437041706123, + "learning_rate": 0.003, + "loss": 4.048, + "step": 27410 + }, + { + "epoch": 0.27411, + "grad_norm": 1.0425000137368567, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 27411 + }, + { + "epoch": 0.27412, + "grad_norm": 1.3489811200005914, + "learning_rate": 0.003, + "loss": 4.036, + "step": 27412 + }, + { + "epoch": 0.27413, + "grad_norm": 0.8882173257573349, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 27413 + }, + { + "epoch": 0.27414, + "grad_norm": 0.8678970152322028, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 27414 + }, + { + "epoch": 0.27415, + "grad_norm": 1.0413787565832058, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 27415 + }, + { + "epoch": 0.27416, + "grad_norm": 1.389559392570043, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 27416 + }, + { + "epoch": 0.27417, + "grad_norm": 0.9218292095646536, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 27417 + }, + { + "epoch": 0.27418, + "grad_norm": 1.0575902718947046, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 27418 + }, + { + "epoch": 0.27419, + "grad_norm": 1.1997188405717039, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 27419 + }, + { + "epoch": 0.2742, + "grad_norm": 1.1173746361657544, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 27420 + }, + { + "epoch": 0.27421, + "grad_norm": 1.2161547832536703, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 27421 + }, + { + "epoch": 0.27422, + "grad_norm": 0.943889751702832, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 27422 + }, + { + "epoch": 0.27423, + "grad_norm": 1.2483320483238425, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 27423 + }, + { + "epoch": 0.27424, + "grad_norm": 1.056658979176417, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 27424 + }, + { + "epoch": 0.27425, + "grad_norm": 0.9724304759592508, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 27425 + }, + { + "epoch": 0.27426, + "grad_norm": 0.9574805571291035, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 27426 + }, + { + "epoch": 0.27427, + "grad_norm": 1.0718186501856661, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 27427 + }, + { + "epoch": 0.27428, + "grad_norm": 1.1407833823783031, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 27428 + }, + { + "epoch": 0.27429, + "grad_norm": 1.2091166434775522, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 27429 + }, + { + "epoch": 0.2743, + "grad_norm": 0.9671321056895757, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 27430 + }, + { + "epoch": 0.27431, + "grad_norm": 1.1739589408313191, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 27431 + }, + { + "epoch": 0.27432, + "grad_norm": 1.0844736622246927, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 27432 + }, + { + "epoch": 0.27433, + "grad_norm": 1.1330965574429344, + "learning_rate": 0.003, + "loss": 4.061, + "step": 27433 + }, + { + "epoch": 0.27434, + "grad_norm": 0.9288879780695354, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 27434 + }, + { + "epoch": 0.27435, + "grad_norm": 0.8652212949090874, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 27435 + }, + { + "epoch": 0.27436, + "grad_norm": 1.01502846617141, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 27436 + }, + { + "epoch": 0.27437, + "grad_norm": 1.0692853993711242, + "learning_rate": 0.003, + "loss": 4.074, + "step": 27437 + }, + { + "epoch": 0.27438, + "grad_norm": 1.197964895370257, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 27438 + }, + { + "epoch": 0.27439, + "grad_norm": 1.12535745094421, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 27439 + }, + { + "epoch": 0.2744, + "grad_norm": 1.1533955443753192, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 27440 + }, + { + "epoch": 0.27441, + "grad_norm": 1.2114309337927947, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 27441 + }, + { + "epoch": 0.27442, + "grad_norm": 1.063985318315013, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 27442 + }, + { + "epoch": 0.27443, + "grad_norm": 1.241298688925437, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 27443 + }, + { + "epoch": 0.27444, + "grad_norm": 0.8662390822464002, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 27444 + }, + { + "epoch": 0.27445, + "grad_norm": 0.8528133073872791, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 27445 + }, + { + "epoch": 0.27446, + "grad_norm": 1.0724847717351886, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 27446 + }, + { + "epoch": 0.27447, + "grad_norm": 1.193629316870422, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 27447 + }, + { + "epoch": 0.27448, + "grad_norm": 0.9774757484883196, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 27448 + }, + { + "epoch": 0.27449, + "grad_norm": 1.186889807601644, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 27449 + }, + { + "epoch": 0.2745, + "grad_norm": 0.9427521740636089, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 27450 + }, + { + "epoch": 0.27451, + "grad_norm": 0.9153615304204664, + "learning_rate": 0.003, + "loss": 4.02, + "step": 27451 + }, + { + "epoch": 0.27452, + "grad_norm": 1.2694146857991533, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 27452 + }, + { + "epoch": 0.27453, + "grad_norm": 0.9969360891965531, + "learning_rate": 0.003, + "loss": 4.07, + "step": 27453 + }, + { + "epoch": 0.27454, + "grad_norm": 1.3395333217751644, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 27454 + }, + { + "epoch": 0.27455, + "grad_norm": 0.7788879221312709, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 27455 + }, + { + "epoch": 0.27456, + "grad_norm": 0.7851320203483015, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 27456 + }, + { + "epoch": 0.27457, + "grad_norm": 0.9145166773578715, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 27457 + }, + { + "epoch": 0.27458, + "grad_norm": 1.1564105340865134, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 27458 + }, + { + "epoch": 0.27459, + "grad_norm": 1.1633346435111427, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 27459 + }, + { + "epoch": 0.2746, + "grad_norm": 1.2203564588176097, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 27460 + }, + { + "epoch": 0.27461, + "grad_norm": 0.9203199320369134, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 27461 + }, + { + "epoch": 0.27462, + "grad_norm": 0.8642702520474596, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 27462 + }, + { + "epoch": 0.27463, + "grad_norm": 0.9935000928972275, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 27463 + }, + { + "epoch": 0.27464, + "grad_norm": 1.0387320821332215, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 27464 + }, + { + "epoch": 0.27465, + "grad_norm": 1.082025089026283, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 27465 + }, + { + "epoch": 0.27466, + "grad_norm": 1.1526581502243738, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 27466 + }, + { + "epoch": 0.27467, + "grad_norm": 1.2476459104607949, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 27467 + }, + { + "epoch": 0.27468, + "grad_norm": 1.2202446469927906, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 27468 + }, + { + "epoch": 0.27469, + "grad_norm": 1.1549768652686894, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 27469 + }, + { + "epoch": 0.2747, + "grad_norm": 1.0051662419434941, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 27470 + }, + { + "epoch": 0.27471, + "grad_norm": 1.165602961888166, + "learning_rate": 0.003, + "loss": 4.051, + "step": 27471 + }, + { + "epoch": 0.27472, + "grad_norm": 1.022723943154893, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 27472 + }, + { + "epoch": 0.27473, + "grad_norm": 1.1204402420266089, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 27473 + }, + { + "epoch": 0.27474, + "grad_norm": 1.2540874837186249, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 27474 + }, + { + "epoch": 0.27475, + "grad_norm": 0.9030629493367209, + "learning_rate": 0.003, + "loss": 4.051, + "step": 27475 + }, + { + "epoch": 0.27476, + "grad_norm": 0.8367326238211357, + "learning_rate": 0.003, + "loss": 4.072, + "step": 27476 + }, + { + "epoch": 0.27477, + "grad_norm": 0.9972587536483758, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 27477 + }, + { + "epoch": 0.27478, + "grad_norm": 1.4344063348165526, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 27478 + }, + { + "epoch": 0.27479, + "grad_norm": 0.8197124069518865, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 27479 + }, + { + "epoch": 0.2748, + "grad_norm": 0.9408193029457637, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 27480 + }, + { + "epoch": 0.27481, + "grad_norm": 1.1884039408909355, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 27481 + }, + { + "epoch": 0.27482, + "grad_norm": 1.0463724873757674, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 27482 + }, + { + "epoch": 0.27483, + "grad_norm": 0.9579009927438651, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 27483 + }, + { + "epoch": 0.27484, + "grad_norm": 1.0883257192625793, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 27484 + }, + { + "epoch": 0.27485, + "grad_norm": 1.1599325969229413, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 27485 + }, + { + "epoch": 0.27486, + "grad_norm": 0.9036993655342397, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 27486 + }, + { + "epoch": 0.27487, + "grad_norm": 1.256009871826562, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 27487 + }, + { + "epoch": 0.27488, + "grad_norm": 1.144180586734586, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 27488 + }, + { + "epoch": 0.27489, + "grad_norm": 1.2508854447083892, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 27489 + }, + { + "epoch": 0.2749, + "grad_norm": 1.2140722340426517, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 27490 + }, + { + "epoch": 0.27491, + "grad_norm": 1.0041767953900056, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 27491 + }, + { + "epoch": 0.27492, + "grad_norm": 1.2510687141124188, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 27492 + }, + { + "epoch": 0.27493, + "grad_norm": 1.0139977837339016, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 27493 + }, + { + "epoch": 0.27494, + "grad_norm": 1.1336179662532797, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 27494 + }, + { + "epoch": 0.27495, + "grad_norm": 1.0153397234312822, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 27495 + }, + { + "epoch": 0.27496, + "grad_norm": 1.0404164934148437, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 27496 + }, + { + "epoch": 0.27497, + "grad_norm": 1.1230774494282412, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 27497 + }, + { + "epoch": 0.27498, + "grad_norm": 1.0321168311694013, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 27498 + }, + { + "epoch": 0.27499, + "grad_norm": 1.1440671205921467, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 27499 + }, + { + "epoch": 0.275, + "grad_norm": 0.877853736869828, + "learning_rate": 0.003, + "loss": 4.032, + "step": 27500 + }, + { + "epoch": 0.27501, + "grad_norm": 0.9224944420571735, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 27501 + }, + { + "epoch": 0.27502, + "grad_norm": 1.1433345180564611, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 27502 + }, + { + "epoch": 0.27503, + "grad_norm": 1.0614633637969502, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 27503 + }, + { + "epoch": 0.27504, + "grad_norm": 1.0298737158160116, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 27504 + }, + { + "epoch": 0.27505, + "grad_norm": 1.0970715461727953, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 27505 + }, + { + "epoch": 0.27506, + "grad_norm": 1.0125640866339667, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 27506 + }, + { + "epoch": 0.27507, + "grad_norm": 1.0975431435323573, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 27507 + }, + { + "epoch": 0.27508, + "grad_norm": 1.177731304922212, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 27508 + }, + { + "epoch": 0.27509, + "grad_norm": 1.1749507249355067, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 27509 + }, + { + "epoch": 0.2751, + "grad_norm": 1.0440216669160847, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 27510 + }, + { + "epoch": 0.27511, + "grad_norm": 1.0755783795085843, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 27511 + }, + { + "epoch": 0.27512, + "grad_norm": 0.9981680985604524, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 27512 + }, + { + "epoch": 0.27513, + "grad_norm": 1.0149513075045422, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 27513 + }, + { + "epoch": 0.27514, + "grad_norm": 1.0787703165135076, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 27514 + }, + { + "epoch": 0.27515, + "grad_norm": 0.9827358961242949, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 27515 + }, + { + "epoch": 0.27516, + "grad_norm": 1.0175887394791912, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 27516 + }, + { + "epoch": 0.27517, + "grad_norm": 1.1133674499794897, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 27517 + }, + { + "epoch": 0.27518, + "grad_norm": 0.9634699247088123, + "learning_rate": 0.003, + "loss": 4.091, + "step": 27518 + }, + { + "epoch": 0.27519, + "grad_norm": 0.9629930692249145, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 27519 + }, + { + "epoch": 0.2752, + "grad_norm": 1.093670663413353, + "learning_rate": 0.003, + "loss": 4.063, + "step": 27520 + }, + { + "epoch": 0.27521, + "grad_norm": 1.2358525599960797, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 27521 + }, + { + "epoch": 0.27522, + "grad_norm": 1.2547042363181, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 27522 + }, + { + "epoch": 0.27523, + "grad_norm": 0.8561119852333825, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 27523 + }, + { + "epoch": 0.27524, + "grad_norm": 0.865666122701375, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 27524 + }, + { + "epoch": 0.27525, + "grad_norm": 1.0222951535814264, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 27525 + }, + { + "epoch": 0.27526, + "grad_norm": 1.3466266915860532, + "learning_rate": 0.003, + "loss": 4.042, + "step": 27526 + }, + { + "epoch": 0.27527, + "grad_norm": 0.8198306846182449, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 27527 + }, + { + "epoch": 0.27528, + "grad_norm": 0.9047753837432967, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 27528 + }, + { + "epoch": 0.27529, + "grad_norm": 0.9865214692014286, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 27529 + }, + { + "epoch": 0.2753, + "grad_norm": 1.042068116041343, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 27530 + }, + { + "epoch": 0.27531, + "grad_norm": 1.1492154514849293, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 27531 + }, + { + "epoch": 0.27532, + "grad_norm": 1.0342104045312333, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 27532 + }, + { + "epoch": 0.27533, + "grad_norm": 1.053764570303467, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 27533 + }, + { + "epoch": 0.27534, + "grad_norm": 1.1817901477284265, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 27534 + }, + { + "epoch": 0.27535, + "grad_norm": 0.9806820763030493, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 27535 + }, + { + "epoch": 0.27536, + "grad_norm": 1.0073489383087293, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 27536 + }, + { + "epoch": 0.27537, + "grad_norm": 1.1339388314451226, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 27537 + }, + { + "epoch": 0.27538, + "grad_norm": 1.2029404577681577, + "learning_rate": 0.003, + "loss": 4.028, + "step": 27538 + }, + { + "epoch": 0.27539, + "grad_norm": 1.077799717417379, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 27539 + }, + { + "epoch": 0.2754, + "grad_norm": 1.1104401678865192, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 27540 + }, + { + "epoch": 0.27541, + "grad_norm": 1.1520591931402906, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 27541 + }, + { + "epoch": 0.27542, + "grad_norm": 1.2088443113578105, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 27542 + }, + { + "epoch": 0.27543, + "grad_norm": 1.0395973664842972, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 27543 + }, + { + "epoch": 0.27544, + "grad_norm": 1.0098348014122598, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 27544 + }, + { + "epoch": 0.27545, + "grad_norm": 1.158319807667759, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 27545 + }, + { + "epoch": 0.27546, + "grad_norm": 1.11805077922114, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 27546 + }, + { + "epoch": 0.27547, + "grad_norm": 1.0061437343912358, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 27547 + }, + { + "epoch": 0.27548, + "grad_norm": 0.8831993258967081, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 27548 + }, + { + "epoch": 0.27549, + "grad_norm": 1.0044061595980134, + "learning_rate": 0.003, + "loss": 4.046, + "step": 27549 + }, + { + "epoch": 0.2755, + "grad_norm": 1.1813316542860612, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 27550 + }, + { + "epoch": 0.27551, + "grad_norm": 1.0992829628798821, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 27551 + }, + { + "epoch": 0.27552, + "grad_norm": 1.4253623559673028, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 27552 + }, + { + "epoch": 0.27553, + "grad_norm": 0.8696635410510692, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 27553 + }, + { + "epoch": 0.27554, + "grad_norm": 1.0001594341227569, + "learning_rate": 0.003, + "loss": 4.0912, + "step": 27554 + }, + { + "epoch": 0.27555, + "grad_norm": 1.0729068857668704, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 27555 + }, + { + "epoch": 0.27556, + "grad_norm": 1.0780165652282017, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 27556 + }, + { + "epoch": 0.27557, + "grad_norm": 0.8545786563029867, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 27557 + }, + { + "epoch": 0.27558, + "grad_norm": 1.050367669674419, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 27558 + }, + { + "epoch": 0.27559, + "grad_norm": 1.4595576396072458, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 27559 + }, + { + "epoch": 0.2756, + "grad_norm": 0.8045900528468389, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 27560 + }, + { + "epoch": 0.27561, + "grad_norm": 0.8432418240710794, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 27561 + }, + { + "epoch": 0.27562, + "grad_norm": 0.8616909457061412, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 27562 + }, + { + "epoch": 0.27563, + "grad_norm": 0.9713909367395194, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 27563 + }, + { + "epoch": 0.27564, + "grad_norm": 1.4064672535849538, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 27564 + }, + { + "epoch": 0.27565, + "grad_norm": 0.970373788147466, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 27565 + }, + { + "epoch": 0.27566, + "grad_norm": 1.215063848942346, + "learning_rate": 0.003, + "loss": 4.083, + "step": 27566 + }, + { + "epoch": 0.27567, + "grad_norm": 1.0456224206550548, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 27567 + }, + { + "epoch": 0.27568, + "grad_norm": 0.9789008261055285, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 27568 + }, + { + "epoch": 0.27569, + "grad_norm": 1.0843414981814181, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 27569 + }, + { + "epoch": 0.2757, + "grad_norm": 1.1479696548745135, + "learning_rate": 0.003, + "loss": 4.035, + "step": 27570 + }, + { + "epoch": 0.27571, + "grad_norm": 1.0941372038687087, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 27571 + }, + { + "epoch": 0.27572, + "grad_norm": 1.0841330909634086, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 27572 + }, + { + "epoch": 0.27573, + "grad_norm": 1.0259031514175678, + "learning_rate": 0.003, + "loss": 4.017, + "step": 27573 + }, + { + "epoch": 0.27574, + "grad_norm": 1.2579917987415856, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 27574 + }, + { + "epoch": 0.27575, + "grad_norm": 0.8975389931319211, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 27575 + }, + { + "epoch": 0.27576, + "grad_norm": 0.9163206951423775, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 27576 + }, + { + "epoch": 0.27577, + "grad_norm": 1.110602032486097, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 27577 + }, + { + "epoch": 0.27578, + "grad_norm": 1.2013106978230368, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 27578 + }, + { + "epoch": 0.27579, + "grad_norm": 0.8771720594414198, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 27579 + }, + { + "epoch": 0.2758, + "grad_norm": 1.0233957673147989, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 27580 + }, + { + "epoch": 0.27581, + "grad_norm": 1.1865952913224895, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 27581 + }, + { + "epoch": 0.27582, + "grad_norm": 0.9475092697450046, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 27582 + }, + { + "epoch": 0.27583, + "grad_norm": 1.005578558163522, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 27583 + }, + { + "epoch": 0.27584, + "grad_norm": 1.1427899246845425, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 27584 + }, + { + "epoch": 0.27585, + "grad_norm": 0.90516209662194, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 27585 + }, + { + "epoch": 0.27586, + "grad_norm": 1.1031749411217908, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 27586 + }, + { + "epoch": 0.27587, + "grad_norm": 1.1097071901302398, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 27587 + }, + { + "epoch": 0.27588, + "grad_norm": 1.2831685523069596, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 27588 + }, + { + "epoch": 0.27589, + "grad_norm": 0.9982339816330889, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 27589 + }, + { + "epoch": 0.2759, + "grad_norm": 1.0637216707338875, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 27590 + }, + { + "epoch": 0.27591, + "grad_norm": 1.2285100632835582, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 27591 + }, + { + "epoch": 0.27592, + "grad_norm": 1.1718701966952656, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 27592 + }, + { + "epoch": 0.27593, + "grad_norm": 0.8903860694178592, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 27593 + }, + { + "epoch": 0.27594, + "grad_norm": 1.010453950233723, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 27594 + }, + { + "epoch": 0.27595, + "grad_norm": 1.1642974130406936, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 27595 + }, + { + "epoch": 0.27596, + "grad_norm": 0.9121295682290402, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 27596 + }, + { + "epoch": 0.27597, + "grad_norm": 0.8637293890561403, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 27597 + }, + { + "epoch": 0.27598, + "grad_norm": 0.9878346147099564, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 27598 + }, + { + "epoch": 0.27599, + "grad_norm": 0.9524223952688488, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 27599 + }, + { + "epoch": 0.276, + "grad_norm": 0.8611740561770004, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 27600 + }, + { + "epoch": 0.27601, + "grad_norm": 0.9592883050756861, + "learning_rate": 0.003, + "loss": 4.047, + "step": 27601 + }, + { + "epoch": 0.27602, + "grad_norm": 1.072818315508533, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 27602 + }, + { + "epoch": 0.27603, + "grad_norm": 1.280037052631408, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 27603 + }, + { + "epoch": 0.27604, + "grad_norm": 0.8399161417830676, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 27604 + }, + { + "epoch": 0.27605, + "grad_norm": 0.9990145971528384, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 27605 + }, + { + "epoch": 0.27606, + "grad_norm": 1.208019592985164, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 27606 + }, + { + "epoch": 0.27607, + "grad_norm": 0.9823845081212192, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 27607 + }, + { + "epoch": 0.27608, + "grad_norm": 1.1211029689508105, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 27608 + }, + { + "epoch": 0.27609, + "grad_norm": 1.2363433749311208, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 27609 + }, + { + "epoch": 0.2761, + "grad_norm": 0.8890850911935946, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 27610 + }, + { + "epoch": 0.27611, + "grad_norm": 0.9261973955005387, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 27611 + }, + { + "epoch": 0.27612, + "grad_norm": 1.1564414858861303, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 27612 + }, + { + "epoch": 0.27613, + "grad_norm": 0.9088504468348749, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 27613 + }, + { + "epoch": 0.27614, + "grad_norm": 0.9707110431302266, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 27614 + }, + { + "epoch": 0.27615, + "grad_norm": 1.0183806392792678, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 27615 + }, + { + "epoch": 0.27616, + "grad_norm": 1.0870081049952767, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 27616 + }, + { + "epoch": 0.27617, + "grad_norm": 1.078464381675632, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 27617 + }, + { + "epoch": 0.27618, + "grad_norm": 1.2852356422642826, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 27618 + }, + { + "epoch": 0.27619, + "grad_norm": 0.9573162887761163, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 27619 + }, + { + "epoch": 0.2762, + "grad_norm": 1.1324589960121811, + "learning_rate": 0.003, + "loss": 4.053, + "step": 27620 + }, + { + "epoch": 0.27621, + "grad_norm": 0.8509918949966426, + "learning_rate": 0.003, + "loss": 4.061, + "step": 27621 + }, + { + "epoch": 0.27622, + "grad_norm": 1.1007643840007213, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 27622 + }, + { + "epoch": 0.27623, + "grad_norm": 1.4140762248579557, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 27623 + }, + { + "epoch": 0.27624, + "grad_norm": 0.9339813629218627, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 27624 + }, + { + "epoch": 0.27625, + "grad_norm": 1.231397856456483, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 27625 + }, + { + "epoch": 0.27626, + "grad_norm": 1.3581870120922155, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 27626 + }, + { + "epoch": 0.27627, + "grad_norm": 0.8765974403250064, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 27627 + }, + { + "epoch": 0.27628, + "grad_norm": 0.9160450189311867, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 27628 + }, + { + "epoch": 0.27629, + "grad_norm": 1.0516779280371351, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 27629 + }, + { + "epoch": 0.2763, + "grad_norm": 1.169342739518736, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 27630 + }, + { + "epoch": 0.27631, + "grad_norm": 1.1879412200355166, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 27631 + }, + { + "epoch": 0.27632, + "grad_norm": 0.92465895711438, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 27632 + }, + { + "epoch": 0.27633, + "grad_norm": 0.9190190415871333, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 27633 + }, + { + "epoch": 0.27634, + "grad_norm": 1.1461371155016489, + "learning_rate": 0.003, + "loss": 4.029, + "step": 27634 + }, + { + "epoch": 0.27635, + "grad_norm": 1.1185058487961559, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 27635 + }, + { + "epoch": 0.27636, + "grad_norm": 0.9404153493658614, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 27636 + }, + { + "epoch": 0.27637, + "grad_norm": 0.9327935773804074, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 27637 + }, + { + "epoch": 0.27638, + "grad_norm": 0.9840223738637565, + "learning_rate": 0.003, + "loss": 4.052, + "step": 27638 + }, + { + "epoch": 0.27639, + "grad_norm": 1.2587986347756206, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 27639 + }, + { + "epoch": 0.2764, + "grad_norm": 0.8345766216560714, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 27640 + }, + { + "epoch": 0.27641, + "grad_norm": 0.8390597539165693, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 27641 + }, + { + "epoch": 0.27642, + "grad_norm": 0.91606324803415, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 27642 + }, + { + "epoch": 0.27643, + "grad_norm": 1.19529807332904, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 27643 + }, + { + "epoch": 0.27644, + "grad_norm": 1.0018180565550712, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 27644 + }, + { + "epoch": 0.27645, + "grad_norm": 1.2773985528264231, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 27645 + }, + { + "epoch": 0.27646, + "grad_norm": 0.9166240751989776, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 27646 + }, + { + "epoch": 0.27647, + "grad_norm": 0.9303236466924674, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 27647 + }, + { + "epoch": 0.27648, + "grad_norm": 0.919531916795222, + "learning_rate": 0.003, + "loss": 4.051, + "step": 27648 + }, + { + "epoch": 0.27649, + "grad_norm": 1.0900641092265158, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 27649 + }, + { + "epoch": 0.2765, + "grad_norm": 1.3416274541575572, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 27650 + }, + { + "epoch": 0.27651, + "grad_norm": 0.8686396064967578, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 27651 + }, + { + "epoch": 0.27652, + "grad_norm": 1.0077961796620898, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 27652 + }, + { + "epoch": 0.27653, + "grad_norm": 1.4208060805393194, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 27653 + }, + { + "epoch": 0.27654, + "grad_norm": 0.9637117264644572, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 27654 + }, + { + "epoch": 0.27655, + "grad_norm": 1.004080465055311, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 27655 + }, + { + "epoch": 0.27656, + "grad_norm": 1.271397721685905, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 27656 + }, + { + "epoch": 0.27657, + "grad_norm": 0.9976353934061721, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 27657 + }, + { + "epoch": 0.27658, + "grad_norm": 1.1508450870768054, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 27658 + }, + { + "epoch": 0.27659, + "grad_norm": 1.0873594073888917, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 27659 + }, + { + "epoch": 0.2766, + "grad_norm": 1.032247951546685, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 27660 + }, + { + "epoch": 0.27661, + "grad_norm": 1.138943725073879, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 27661 + }, + { + "epoch": 0.27662, + "grad_norm": 1.0918672379932763, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 27662 + }, + { + "epoch": 0.27663, + "grad_norm": 0.9597238577096918, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 27663 + }, + { + "epoch": 0.27664, + "grad_norm": 0.8569833862495566, + "learning_rate": 0.003, + "loss": 4.028, + "step": 27664 + }, + { + "epoch": 0.27665, + "grad_norm": 0.9631527407975773, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 27665 + }, + { + "epoch": 0.27666, + "grad_norm": 1.0775161354216423, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 27666 + }, + { + "epoch": 0.27667, + "grad_norm": 0.9468890305253572, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 27667 + }, + { + "epoch": 0.27668, + "grad_norm": 1.0261903265541668, + "learning_rate": 0.003, + "loss": 4.055, + "step": 27668 + }, + { + "epoch": 0.27669, + "grad_norm": 1.1917886133289, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 27669 + }, + { + "epoch": 0.2767, + "grad_norm": 0.9637820455499518, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 27670 + }, + { + "epoch": 0.27671, + "grad_norm": 1.1746185246184675, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 27671 + }, + { + "epoch": 0.27672, + "grad_norm": 1.1223309718262224, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 27672 + }, + { + "epoch": 0.27673, + "grad_norm": 1.1078791254969664, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 27673 + }, + { + "epoch": 0.27674, + "grad_norm": 1.0744583027747443, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 27674 + }, + { + "epoch": 0.27675, + "grad_norm": 1.0847826155888256, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 27675 + }, + { + "epoch": 0.27676, + "grad_norm": 1.2964109810422124, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 27676 + }, + { + "epoch": 0.27677, + "grad_norm": 0.7975862711131871, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 27677 + }, + { + "epoch": 0.27678, + "grad_norm": 0.8198174562003071, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 27678 + }, + { + "epoch": 0.27679, + "grad_norm": 1.049035548523677, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 27679 + }, + { + "epoch": 0.2768, + "grad_norm": 1.340363245386692, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 27680 + }, + { + "epoch": 0.27681, + "grad_norm": 0.7678654818062912, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 27681 + }, + { + "epoch": 0.27682, + "grad_norm": 0.711235305418412, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 27682 + }, + { + "epoch": 0.27683, + "grad_norm": 0.7914857333034059, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 27683 + }, + { + "epoch": 0.27684, + "grad_norm": 1.0264198642465754, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 27684 + }, + { + "epoch": 0.27685, + "grad_norm": 1.1652803627944357, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 27685 + }, + { + "epoch": 0.27686, + "grad_norm": 0.8700523529183385, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 27686 + }, + { + "epoch": 0.27687, + "grad_norm": 1.0912810598488127, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 27687 + }, + { + "epoch": 0.27688, + "grad_norm": 1.2756668603393315, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 27688 + }, + { + "epoch": 0.27689, + "grad_norm": 1.0196804171476066, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 27689 + }, + { + "epoch": 0.2769, + "grad_norm": 1.1802917143246119, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 27690 + }, + { + "epoch": 0.27691, + "grad_norm": 1.251007444220524, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 27691 + }, + { + "epoch": 0.27692, + "grad_norm": 1.0128258790887352, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 27692 + }, + { + "epoch": 0.27693, + "grad_norm": 1.0283899218691028, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 27693 + }, + { + "epoch": 0.27694, + "grad_norm": 1.1276449958539507, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 27694 + }, + { + "epoch": 0.27695, + "grad_norm": 1.0072253025597058, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 27695 + }, + { + "epoch": 0.27696, + "grad_norm": 1.0775886323306108, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 27696 + }, + { + "epoch": 0.27697, + "grad_norm": 1.0296525030987094, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 27697 + }, + { + "epoch": 0.27698, + "grad_norm": 1.1268124576633876, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 27698 + }, + { + "epoch": 0.27699, + "grad_norm": 0.9891695924570644, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 27699 + }, + { + "epoch": 0.277, + "grad_norm": 1.2021084254510348, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 27700 + }, + { + "epoch": 0.27701, + "grad_norm": 1.0508184900196895, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 27701 + }, + { + "epoch": 0.27702, + "grad_norm": 1.075164501845793, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 27702 + }, + { + "epoch": 0.27703, + "grad_norm": 1.1252031484097562, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 27703 + }, + { + "epoch": 0.27704, + "grad_norm": 0.9501868145151212, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 27704 + }, + { + "epoch": 0.27705, + "grad_norm": 1.1256859779812893, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 27705 + }, + { + "epoch": 0.27706, + "grad_norm": 1.0890786279061206, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 27706 + }, + { + "epoch": 0.27707, + "grad_norm": 1.0007105098988882, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 27707 + }, + { + "epoch": 0.27708, + "grad_norm": 1.0651565263604659, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 27708 + }, + { + "epoch": 0.27709, + "grad_norm": 1.274070193890888, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 27709 + }, + { + "epoch": 0.2771, + "grad_norm": 0.958368350045831, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 27710 + }, + { + "epoch": 0.27711, + "grad_norm": 1.2055991614190378, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 27711 + }, + { + "epoch": 0.27712, + "grad_norm": 1.1232910930199813, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 27712 + }, + { + "epoch": 0.27713, + "grad_norm": 1.150730157502445, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 27713 + }, + { + "epoch": 0.27714, + "grad_norm": 1.1679394557219744, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 27714 + }, + { + "epoch": 0.27715, + "grad_norm": 0.849972582878212, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 27715 + }, + { + "epoch": 0.27716, + "grad_norm": 0.8964450179078949, + "learning_rate": 0.003, + "loss": 4.048, + "step": 27716 + }, + { + "epoch": 0.27717, + "grad_norm": 1.0477830413328482, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 27717 + }, + { + "epoch": 0.27718, + "grad_norm": 1.2145751601247181, + "learning_rate": 0.003, + "loss": 4.042, + "step": 27718 + }, + { + "epoch": 0.27719, + "grad_norm": 0.853716961201871, + "learning_rate": 0.003, + "loss": 4.029, + "step": 27719 + }, + { + "epoch": 0.2772, + "grad_norm": 1.032434085794419, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 27720 + }, + { + "epoch": 0.27721, + "grad_norm": 1.0675484387586747, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 27721 + }, + { + "epoch": 0.27722, + "grad_norm": 1.1622552623077909, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 27722 + }, + { + "epoch": 0.27723, + "grad_norm": 0.8558614077107, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 27723 + }, + { + "epoch": 0.27724, + "grad_norm": 0.7270097545705969, + "learning_rate": 0.003, + "loss": 4.057, + "step": 27724 + }, + { + "epoch": 0.27725, + "grad_norm": 0.8685621363580076, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 27725 + }, + { + "epoch": 0.27726, + "grad_norm": 1.0187255863937972, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 27726 + }, + { + "epoch": 0.27727, + "grad_norm": 1.171922813973846, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 27727 + }, + { + "epoch": 0.27728, + "grad_norm": 1.1227428392904524, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 27728 + }, + { + "epoch": 0.27729, + "grad_norm": 0.9580204961037899, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 27729 + }, + { + "epoch": 0.2773, + "grad_norm": 1.0339187129424188, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 27730 + }, + { + "epoch": 0.27731, + "grad_norm": 0.9848077218062271, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 27731 + }, + { + "epoch": 0.27732, + "grad_norm": 1.0327032554157123, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 27732 + }, + { + "epoch": 0.27733, + "grad_norm": 1.2350611860160503, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 27733 + }, + { + "epoch": 0.27734, + "grad_norm": 1.0417820643028914, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 27734 + }, + { + "epoch": 0.27735, + "grad_norm": 1.0247321566913445, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 27735 + }, + { + "epoch": 0.27736, + "grad_norm": 1.0259602125156448, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 27736 + }, + { + "epoch": 0.27737, + "grad_norm": 1.2061676582210488, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 27737 + }, + { + "epoch": 0.27738, + "grad_norm": 1.1538143574803421, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 27738 + }, + { + "epoch": 0.27739, + "grad_norm": 1.1103075919928975, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 27739 + }, + { + "epoch": 0.2774, + "grad_norm": 1.082838695316885, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 27740 + }, + { + "epoch": 0.27741, + "grad_norm": 0.9845739052238514, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 27741 + }, + { + "epoch": 0.27742, + "grad_norm": 1.0699414682830612, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 27742 + }, + { + "epoch": 0.27743, + "grad_norm": 1.0308325699580096, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 27743 + }, + { + "epoch": 0.27744, + "grad_norm": 1.036556619144357, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 27744 + }, + { + "epoch": 0.27745, + "grad_norm": 1.0416631391457487, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 27745 + }, + { + "epoch": 0.27746, + "grad_norm": 1.1900567490996734, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 27746 + }, + { + "epoch": 0.27747, + "grad_norm": 1.1568806952436301, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 27747 + }, + { + "epoch": 0.27748, + "grad_norm": 0.9803334954147341, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 27748 + }, + { + "epoch": 0.27749, + "grad_norm": 1.170757448211153, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 27749 + }, + { + "epoch": 0.2775, + "grad_norm": 1.1119657066497675, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 27750 + }, + { + "epoch": 0.27751, + "grad_norm": 1.0250317069739256, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 27751 + }, + { + "epoch": 0.27752, + "grad_norm": 1.2727751568618826, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 27752 + }, + { + "epoch": 0.27753, + "grad_norm": 0.8310520259170798, + "learning_rate": 0.003, + "loss": 4.047, + "step": 27753 + }, + { + "epoch": 0.27754, + "grad_norm": 0.93213935857922, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 27754 + }, + { + "epoch": 0.27755, + "grad_norm": 1.278617568066189, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 27755 + }, + { + "epoch": 0.27756, + "grad_norm": 1.2171757607989995, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 27756 + }, + { + "epoch": 0.27757, + "grad_norm": 1.1073201993997037, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 27757 + }, + { + "epoch": 0.27758, + "grad_norm": 0.9247319429877513, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 27758 + }, + { + "epoch": 0.27759, + "grad_norm": 0.9739825548253913, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 27759 + }, + { + "epoch": 0.2776, + "grad_norm": 1.2942231662172432, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 27760 + }, + { + "epoch": 0.27761, + "grad_norm": 0.8946654191675318, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 27761 + }, + { + "epoch": 0.27762, + "grad_norm": 1.098424083861399, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 27762 + }, + { + "epoch": 0.27763, + "grad_norm": 1.1193088544187833, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 27763 + }, + { + "epoch": 0.27764, + "grad_norm": 1.201843858445773, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 27764 + }, + { + "epoch": 0.27765, + "grad_norm": 0.8716399316509339, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 27765 + }, + { + "epoch": 0.27766, + "grad_norm": 0.940898045932022, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 27766 + }, + { + "epoch": 0.27767, + "grad_norm": 1.2042827933078661, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 27767 + }, + { + "epoch": 0.27768, + "grad_norm": 1.105358299469714, + "learning_rate": 0.003, + "loss": 4.048, + "step": 27768 + }, + { + "epoch": 0.27769, + "grad_norm": 0.9328591894187857, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 27769 + }, + { + "epoch": 0.2777, + "grad_norm": 0.8449503718152275, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 27770 + }, + { + "epoch": 0.27771, + "grad_norm": 1.0111530995335543, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 27771 + }, + { + "epoch": 0.27772, + "grad_norm": 1.3398723228312694, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 27772 + }, + { + "epoch": 0.27773, + "grad_norm": 0.7509971481258538, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 27773 + }, + { + "epoch": 0.27774, + "grad_norm": 0.889923740559553, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 27774 + }, + { + "epoch": 0.27775, + "grad_norm": 1.1133562054817896, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 27775 + }, + { + "epoch": 0.27776, + "grad_norm": 1.3123791504238915, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 27776 + }, + { + "epoch": 0.27777, + "grad_norm": 0.9521453495499592, + "learning_rate": 0.003, + "loss": 4.054, + "step": 27777 + }, + { + "epoch": 0.27778, + "grad_norm": 1.0782730833366077, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 27778 + }, + { + "epoch": 0.27779, + "grad_norm": 1.3822406079402287, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 27779 + }, + { + "epoch": 0.2778, + "grad_norm": 0.7133285770931417, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 27780 + }, + { + "epoch": 0.27781, + "grad_norm": 0.7234148038193512, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 27781 + }, + { + "epoch": 0.27782, + "grad_norm": 0.7812688046013403, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 27782 + }, + { + "epoch": 0.27783, + "grad_norm": 1.104183861736075, + "learning_rate": 0.003, + "loss": 4.058, + "step": 27783 + }, + { + "epoch": 0.27784, + "grad_norm": 1.2662187216404468, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 27784 + }, + { + "epoch": 0.27785, + "grad_norm": 1.0228398069160083, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 27785 + }, + { + "epoch": 0.27786, + "grad_norm": 1.1509018836173894, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 27786 + }, + { + "epoch": 0.27787, + "grad_norm": 0.8205062392115873, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 27787 + }, + { + "epoch": 0.27788, + "grad_norm": 0.867446098861149, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 27788 + }, + { + "epoch": 0.27789, + "grad_norm": 0.9180749957671928, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 27789 + }, + { + "epoch": 0.2779, + "grad_norm": 0.8249583137280374, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 27790 + }, + { + "epoch": 0.27791, + "grad_norm": 0.864039870870856, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 27791 + }, + { + "epoch": 0.27792, + "grad_norm": 0.9071279026456049, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 27792 + }, + { + "epoch": 0.27793, + "grad_norm": 1.243426713777616, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 27793 + }, + { + "epoch": 0.27794, + "grad_norm": 1.095992430729363, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 27794 + }, + { + "epoch": 0.27795, + "grad_norm": 0.9686337780581192, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 27795 + }, + { + "epoch": 0.27796, + "grad_norm": 1.2109023437775517, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 27796 + }, + { + "epoch": 0.27797, + "grad_norm": 1.1046222215532902, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 27797 + }, + { + "epoch": 0.27798, + "grad_norm": 1.061962424499125, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 27798 + }, + { + "epoch": 0.27799, + "grad_norm": 1.0384109849756296, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 27799 + }, + { + "epoch": 0.278, + "grad_norm": 1.1900352711270734, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 27800 + }, + { + "epoch": 0.27801, + "grad_norm": 1.1057103703985522, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 27801 + }, + { + "epoch": 0.27802, + "grad_norm": 1.107935475775994, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 27802 + }, + { + "epoch": 0.27803, + "grad_norm": 1.1417878604355842, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 27803 + }, + { + "epoch": 0.27804, + "grad_norm": 0.921891423183917, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 27804 + }, + { + "epoch": 0.27805, + "grad_norm": 1.0712839884579108, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 27805 + }, + { + "epoch": 0.27806, + "grad_norm": 1.1265822981129252, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 27806 + }, + { + "epoch": 0.27807, + "grad_norm": 1.1495329257396778, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 27807 + }, + { + "epoch": 0.27808, + "grad_norm": 1.0874512703892714, + "learning_rate": 0.003, + "loss": 4.043, + "step": 27808 + }, + { + "epoch": 0.27809, + "grad_norm": 1.0672407740140037, + "learning_rate": 0.003, + "loss": 4.029, + "step": 27809 + }, + { + "epoch": 0.2781, + "grad_norm": 0.9756930951561703, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 27810 + }, + { + "epoch": 0.27811, + "grad_norm": 1.25514789640384, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 27811 + }, + { + "epoch": 0.27812, + "grad_norm": 1.0341899002580532, + "learning_rate": 0.003, + "loss": 4.041, + "step": 27812 + }, + { + "epoch": 0.27813, + "grad_norm": 1.0842938846987573, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 27813 + }, + { + "epoch": 0.27814, + "grad_norm": 0.9381467989904105, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 27814 + }, + { + "epoch": 0.27815, + "grad_norm": 1.0340579332134863, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 27815 + }, + { + "epoch": 0.27816, + "grad_norm": 1.2419886040549921, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 27816 + }, + { + "epoch": 0.27817, + "grad_norm": 0.9870403170179475, + "learning_rate": 0.003, + "loss": 4.075, + "step": 27817 + }, + { + "epoch": 0.27818, + "grad_norm": 1.1098079967815109, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 27818 + }, + { + "epoch": 0.27819, + "grad_norm": 1.0357111439590128, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 27819 + }, + { + "epoch": 0.2782, + "grad_norm": 1.340230052860021, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 27820 + }, + { + "epoch": 0.27821, + "grad_norm": 1.146272439984818, + "learning_rate": 0.003, + "loss": 4.072, + "step": 27821 + }, + { + "epoch": 0.27822, + "grad_norm": 1.1011107090901648, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 27822 + }, + { + "epoch": 0.27823, + "grad_norm": 1.1151952283471671, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 27823 + }, + { + "epoch": 0.27824, + "grad_norm": 0.9985407113592684, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 27824 + }, + { + "epoch": 0.27825, + "grad_norm": 1.1865212847925117, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 27825 + }, + { + "epoch": 0.27826, + "grad_norm": 1.1031840111986362, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 27826 + }, + { + "epoch": 0.27827, + "grad_norm": 1.1988583262065917, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 27827 + }, + { + "epoch": 0.27828, + "grad_norm": 1.0607423437715364, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 27828 + }, + { + "epoch": 0.27829, + "grad_norm": 0.931179612174383, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 27829 + }, + { + "epoch": 0.2783, + "grad_norm": 1.2399862693090782, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 27830 + }, + { + "epoch": 0.27831, + "grad_norm": 0.9773450957216953, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 27831 + }, + { + "epoch": 0.27832, + "grad_norm": 1.249050885802265, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 27832 + }, + { + "epoch": 0.27833, + "grad_norm": 0.9418279473154356, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 27833 + }, + { + "epoch": 0.27834, + "grad_norm": 0.945784432087064, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 27834 + }, + { + "epoch": 0.27835, + "grad_norm": 0.9740656457475453, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 27835 + }, + { + "epoch": 0.27836, + "grad_norm": 0.8854840938742919, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 27836 + }, + { + "epoch": 0.27837, + "grad_norm": 0.8852609157506234, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 27837 + }, + { + "epoch": 0.27838, + "grad_norm": 0.9781121495918179, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 27838 + }, + { + "epoch": 0.27839, + "grad_norm": 1.0797326372146747, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 27839 + }, + { + "epoch": 0.2784, + "grad_norm": 1.1110582473632504, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 27840 + }, + { + "epoch": 0.27841, + "grad_norm": 1.3213323904083756, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 27841 + }, + { + "epoch": 0.27842, + "grad_norm": 1.0543466695894466, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 27842 + }, + { + "epoch": 0.27843, + "grad_norm": 1.1788096880919468, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 27843 + }, + { + "epoch": 0.27844, + "grad_norm": 1.028192974253234, + "learning_rate": 0.003, + "loss": 4.074, + "step": 27844 + }, + { + "epoch": 0.27845, + "grad_norm": 1.1474969321637156, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 27845 + }, + { + "epoch": 0.27846, + "grad_norm": 1.0447425717058538, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 27846 + }, + { + "epoch": 0.27847, + "grad_norm": 1.095235676404283, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 27847 + }, + { + "epoch": 0.27848, + "grad_norm": 0.985936492166445, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 27848 + }, + { + "epoch": 0.27849, + "grad_norm": 1.0445694038731859, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 27849 + }, + { + "epoch": 0.2785, + "grad_norm": 1.1068791196081271, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 27850 + }, + { + "epoch": 0.27851, + "grad_norm": 1.206190271472032, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 27851 + }, + { + "epoch": 0.27852, + "grad_norm": 0.8665077821122273, + "learning_rate": 0.003, + "loss": 4.075, + "step": 27852 + }, + { + "epoch": 0.27853, + "grad_norm": 0.8832815687461788, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 27853 + }, + { + "epoch": 0.27854, + "grad_norm": 1.0846623665750195, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 27854 + }, + { + "epoch": 0.27855, + "grad_norm": 0.9415946466834489, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 27855 + }, + { + "epoch": 0.27856, + "grad_norm": 1.122904451276162, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 27856 + }, + { + "epoch": 0.27857, + "grad_norm": 1.0976931742435656, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 27857 + }, + { + "epoch": 0.27858, + "grad_norm": 1.012276269261339, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 27858 + }, + { + "epoch": 0.27859, + "grad_norm": 1.351570281048585, + "learning_rate": 0.003, + "loss": 4.058, + "step": 27859 + }, + { + "epoch": 0.2786, + "grad_norm": 0.7680030844535797, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 27860 + }, + { + "epoch": 0.27861, + "grad_norm": 0.7789835644437625, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 27861 + }, + { + "epoch": 0.27862, + "grad_norm": 0.9211544831786801, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 27862 + }, + { + "epoch": 0.27863, + "grad_norm": 1.1564911448581225, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 27863 + }, + { + "epoch": 0.27864, + "grad_norm": 1.0814530503472268, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 27864 + }, + { + "epoch": 0.27865, + "grad_norm": 1.2195204521461884, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 27865 + }, + { + "epoch": 0.27866, + "grad_norm": 1.0568986200265218, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 27866 + }, + { + "epoch": 0.27867, + "grad_norm": 1.1090657019926526, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 27867 + }, + { + "epoch": 0.27868, + "grad_norm": 1.07664665949004, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 27868 + }, + { + "epoch": 0.27869, + "grad_norm": 1.1658521326065054, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 27869 + }, + { + "epoch": 0.2787, + "grad_norm": 1.159073629408047, + "learning_rate": 0.003, + "loss": 4.041, + "step": 27870 + }, + { + "epoch": 0.27871, + "grad_norm": 0.9140448622974242, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 27871 + }, + { + "epoch": 0.27872, + "grad_norm": 0.9465021864211179, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 27872 + }, + { + "epoch": 0.27873, + "grad_norm": 1.048739573180355, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 27873 + }, + { + "epoch": 0.27874, + "grad_norm": 0.9214649370270157, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 27874 + }, + { + "epoch": 0.27875, + "grad_norm": 1.2061402662206004, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 27875 + }, + { + "epoch": 0.27876, + "grad_norm": 1.034559121631152, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 27876 + }, + { + "epoch": 0.27877, + "grad_norm": 1.2013038940361287, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 27877 + }, + { + "epoch": 0.27878, + "grad_norm": 1.1554880671712326, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 27878 + }, + { + "epoch": 0.27879, + "grad_norm": 1.0559923692234454, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 27879 + }, + { + "epoch": 0.2788, + "grad_norm": 1.1995836675927023, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 27880 + }, + { + "epoch": 0.27881, + "grad_norm": 0.9253383627728623, + "learning_rate": 0.003, + "loss": 4.025, + "step": 27881 + }, + { + "epoch": 0.27882, + "grad_norm": 0.986248662571534, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 27882 + }, + { + "epoch": 0.27883, + "grad_norm": 1.0755513625781066, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 27883 + }, + { + "epoch": 0.27884, + "grad_norm": 1.1604000363090161, + "learning_rate": 0.003, + "loss": 4.05, + "step": 27884 + }, + { + "epoch": 0.27885, + "grad_norm": 0.965783674014117, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 27885 + }, + { + "epoch": 0.27886, + "grad_norm": 1.1311031884909324, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 27886 + }, + { + "epoch": 0.27887, + "grad_norm": 1.2313713145863983, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 27887 + }, + { + "epoch": 0.27888, + "grad_norm": 1.251081242093084, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 27888 + }, + { + "epoch": 0.27889, + "grad_norm": 1.0513326837790344, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 27889 + }, + { + "epoch": 0.2789, + "grad_norm": 1.1722606988689757, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 27890 + }, + { + "epoch": 0.27891, + "grad_norm": 1.099705736587624, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 27891 + }, + { + "epoch": 0.27892, + "grad_norm": 0.9598086396125327, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 27892 + }, + { + "epoch": 0.27893, + "grad_norm": 1.0227546142808028, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 27893 + }, + { + "epoch": 0.27894, + "grad_norm": 1.0522692748158191, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 27894 + }, + { + "epoch": 0.27895, + "grad_norm": 1.1079596432130772, + "learning_rate": 0.003, + "loss": 4.057, + "step": 27895 + }, + { + "epoch": 0.27896, + "grad_norm": 1.0137218171209539, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 27896 + }, + { + "epoch": 0.27897, + "grad_norm": 0.9937102771303371, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 27897 + }, + { + "epoch": 0.27898, + "grad_norm": 0.9432653288409969, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 27898 + }, + { + "epoch": 0.27899, + "grad_norm": 1.1229449648798913, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 27899 + }, + { + "epoch": 0.279, + "grad_norm": 1.0559082593780196, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 27900 + }, + { + "epoch": 0.27901, + "grad_norm": 1.3439548380195898, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 27901 + }, + { + "epoch": 0.27902, + "grad_norm": 1.2359823956879492, + "learning_rate": 0.003, + "loss": 4.038, + "step": 27902 + }, + { + "epoch": 0.27903, + "grad_norm": 1.0587684007938818, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 27903 + }, + { + "epoch": 0.27904, + "grad_norm": 1.0707012112086576, + "learning_rate": 0.003, + "loss": 4.048, + "step": 27904 + }, + { + "epoch": 0.27905, + "grad_norm": 1.1553099516588785, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 27905 + }, + { + "epoch": 0.27906, + "grad_norm": 0.8464650534739766, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 27906 + }, + { + "epoch": 0.27907, + "grad_norm": 0.9590927845021954, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 27907 + }, + { + "epoch": 0.27908, + "grad_norm": 1.109622038969436, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 27908 + }, + { + "epoch": 0.27909, + "grad_norm": 1.0926618470053786, + "learning_rate": 0.003, + "loss": 4.055, + "step": 27909 + }, + { + "epoch": 0.2791, + "grad_norm": 1.143627556149577, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 27910 + }, + { + "epoch": 0.27911, + "grad_norm": 1.0635601643802728, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 27911 + }, + { + "epoch": 0.27912, + "grad_norm": 1.0193176365245906, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 27912 + }, + { + "epoch": 0.27913, + "grad_norm": 1.1103216991025757, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 27913 + }, + { + "epoch": 0.27914, + "grad_norm": 1.4177000184945587, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 27914 + }, + { + "epoch": 0.27915, + "grad_norm": 0.798670847626472, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 27915 + }, + { + "epoch": 0.27916, + "grad_norm": 0.9373279817558774, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 27916 + }, + { + "epoch": 0.27917, + "grad_norm": 1.5227500735522121, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 27917 + }, + { + "epoch": 0.27918, + "grad_norm": 0.9800531478489402, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 27918 + }, + { + "epoch": 0.27919, + "grad_norm": 0.7726858696715303, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 27919 + }, + { + "epoch": 0.2792, + "grad_norm": 0.7871351904434918, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 27920 + }, + { + "epoch": 0.27921, + "grad_norm": 0.9539750276567104, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 27921 + }, + { + "epoch": 0.27922, + "grad_norm": 1.618699715834157, + "learning_rate": 0.003, + "loss": 4.1015, + "step": 27922 + }, + { + "epoch": 0.27923, + "grad_norm": 0.718679538507499, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 27923 + }, + { + "epoch": 0.27924, + "grad_norm": 0.9877015244329262, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 27924 + }, + { + "epoch": 0.27925, + "grad_norm": 1.255794280595779, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 27925 + }, + { + "epoch": 0.27926, + "grad_norm": 0.7752498837870255, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 27926 + }, + { + "epoch": 0.27927, + "grad_norm": 0.9249320522883676, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 27927 + }, + { + "epoch": 0.27928, + "grad_norm": 0.994372296610955, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 27928 + }, + { + "epoch": 0.27929, + "grad_norm": 1.2444584077816572, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 27929 + }, + { + "epoch": 0.2793, + "grad_norm": 1.0328511151476127, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 27930 + }, + { + "epoch": 0.27931, + "grad_norm": 1.1860765253229841, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 27931 + }, + { + "epoch": 0.27932, + "grad_norm": 0.9981001337206675, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 27932 + }, + { + "epoch": 0.27933, + "grad_norm": 1.3267453580489177, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 27933 + }, + { + "epoch": 0.27934, + "grad_norm": 1.080149784179221, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 27934 + }, + { + "epoch": 0.27935, + "grad_norm": 1.0042688323028484, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 27935 + }, + { + "epoch": 0.27936, + "grad_norm": 1.2805293401552291, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 27936 + }, + { + "epoch": 0.27937, + "grad_norm": 1.049100384269636, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 27937 + }, + { + "epoch": 0.27938, + "grad_norm": 1.148530792833864, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 27938 + }, + { + "epoch": 0.27939, + "grad_norm": 1.1729613509241559, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 27939 + }, + { + "epoch": 0.2794, + "grad_norm": 0.9176364019210935, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 27940 + }, + { + "epoch": 0.27941, + "grad_norm": 0.8239721243481692, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 27941 + }, + { + "epoch": 0.27942, + "grad_norm": 0.838513682623255, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 27942 + }, + { + "epoch": 0.27943, + "grad_norm": 1.0166227960503786, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 27943 + }, + { + "epoch": 0.27944, + "grad_norm": 0.994092522773955, + "learning_rate": 0.003, + "loss": 4.065, + "step": 27944 + }, + { + "epoch": 0.27945, + "grad_norm": 1.0725596322823336, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 27945 + }, + { + "epoch": 0.27946, + "grad_norm": 1.202575147418277, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 27946 + }, + { + "epoch": 0.27947, + "grad_norm": 0.9675241686141636, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 27947 + }, + { + "epoch": 0.27948, + "grad_norm": 1.2267941961575033, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 27948 + }, + { + "epoch": 0.27949, + "grad_norm": 1.1059874179157054, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 27949 + }, + { + "epoch": 0.2795, + "grad_norm": 1.0217766495703042, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 27950 + }, + { + "epoch": 0.27951, + "grad_norm": 1.1529341105428252, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 27951 + }, + { + "epoch": 0.27952, + "grad_norm": 0.9905619203995053, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 27952 + }, + { + "epoch": 0.27953, + "grad_norm": 1.121192752447898, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 27953 + }, + { + "epoch": 0.27954, + "grad_norm": 0.8988972132343938, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 27954 + }, + { + "epoch": 0.27955, + "grad_norm": 0.9611193163441749, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 27955 + }, + { + "epoch": 0.27956, + "grad_norm": 1.26062806075441, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 27956 + }, + { + "epoch": 0.27957, + "grad_norm": 0.9421559724339583, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 27957 + }, + { + "epoch": 0.27958, + "grad_norm": 0.878112930628147, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 27958 + }, + { + "epoch": 0.27959, + "grad_norm": 1.0433967747215498, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 27959 + }, + { + "epoch": 0.2796, + "grad_norm": 1.131617472525823, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 27960 + }, + { + "epoch": 0.27961, + "grad_norm": 1.0601225117588624, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 27961 + }, + { + "epoch": 0.27962, + "grad_norm": 1.2450587317476423, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 27962 + }, + { + "epoch": 0.27963, + "grad_norm": 1.0757855902950724, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 27963 + }, + { + "epoch": 0.27964, + "grad_norm": 1.339722991986603, + "learning_rate": 0.003, + "loss": 4.065, + "step": 27964 + }, + { + "epoch": 0.27965, + "grad_norm": 1.1301227188058827, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 27965 + }, + { + "epoch": 0.27966, + "grad_norm": 1.0657452706185477, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 27966 + }, + { + "epoch": 0.27967, + "grad_norm": 1.2269110977702125, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 27967 + }, + { + "epoch": 0.27968, + "grad_norm": 0.8730959877145401, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 27968 + }, + { + "epoch": 0.27969, + "grad_norm": 0.9524791132429211, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 27969 + }, + { + "epoch": 0.2797, + "grad_norm": 1.4120248062542957, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 27970 + }, + { + "epoch": 0.27971, + "grad_norm": 0.8915553551072096, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 27971 + }, + { + "epoch": 0.27972, + "grad_norm": 0.9767230290024855, + "learning_rate": 0.003, + "loss": 4.056, + "step": 27972 + }, + { + "epoch": 0.27973, + "grad_norm": 1.305975433456087, + "learning_rate": 0.003, + "loss": 4.05, + "step": 27973 + }, + { + "epoch": 0.27974, + "grad_norm": 0.8896931350937097, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 27974 + }, + { + "epoch": 0.27975, + "grad_norm": 0.8346525733906603, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 27975 + }, + { + "epoch": 0.27976, + "grad_norm": 0.9170569027648564, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 27976 + }, + { + "epoch": 0.27977, + "grad_norm": 1.349979728257226, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 27977 + }, + { + "epoch": 0.27978, + "grad_norm": 1.0330732662522888, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 27978 + }, + { + "epoch": 0.27979, + "grad_norm": 1.0718886185554581, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 27979 + }, + { + "epoch": 0.2798, + "grad_norm": 0.9071158774108823, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 27980 + }, + { + "epoch": 0.27981, + "grad_norm": 1.250726877080188, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 27981 + }, + { + "epoch": 0.27982, + "grad_norm": 1.2189519652399883, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 27982 + }, + { + "epoch": 0.27983, + "grad_norm": 0.8160694863619969, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 27983 + }, + { + "epoch": 0.27984, + "grad_norm": 0.7173966853578413, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 27984 + }, + { + "epoch": 0.27985, + "grad_norm": 0.9034324981497762, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 27985 + }, + { + "epoch": 0.27986, + "grad_norm": 1.1278929163028222, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 27986 + }, + { + "epoch": 0.27987, + "grad_norm": 1.1137545406280158, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 27987 + }, + { + "epoch": 0.27988, + "grad_norm": 1.2833865486502651, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 27988 + }, + { + "epoch": 0.27989, + "grad_norm": 0.974358052774076, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 27989 + }, + { + "epoch": 0.2799, + "grad_norm": 0.9462638688898973, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 27990 + }, + { + "epoch": 0.27991, + "grad_norm": 1.2197574330206191, + "learning_rate": 0.003, + "loss": 4.05, + "step": 27991 + }, + { + "epoch": 0.27992, + "grad_norm": 1.1198763054242733, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 27992 + }, + { + "epoch": 0.27993, + "grad_norm": 1.1826566304690784, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 27993 + }, + { + "epoch": 0.27994, + "grad_norm": 0.8907489242639597, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 27994 + }, + { + "epoch": 0.27995, + "grad_norm": 0.9853110632144124, + "learning_rate": 0.003, + "loss": 4.1036, + "step": 27995 + }, + { + "epoch": 0.27996, + "grad_norm": 1.0859166899037578, + "learning_rate": 0.003, + "loss": 4.065, + "step": 27996 + }, + { + "epoch": 0.27997, + "grad_norm": 0.9782734805394204, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 27997 + }, + { + "epoch": 0.27998, + "grad_norm": 1.1501106560100778, + "learning_rate": 0.003, + "loss": 4.046, + "step": 27998 + }, + { + "epoch": 0.27999, + "grad_norm": 0.9915939866817599, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 27999 + }, + { + "epoch": 0.28, + "grad_norm": 1.1979487036961554, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 28000 + }, + { + "epoch": 0.28001, + "grad_norm": 1.114182856311641, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 28001 + }, + { + "epoch": 0.28002, + "grad_norm": 1.0743538338788186, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 28002 + }, + { + "epoch": 0.28003, + "grad_norm": 1.1388801421657857, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 28003 + }, + { + "epoch": 0.28004, + "grad_norm": 1.1858129820181855, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 28004 + }, + { + "epoch": 0.28005, + "grad_norm": 0.8691179564146596, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 28005 + }, + { + "epoch": 0.28006, + "grad_norm": 1.0319220437329086, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 28006 + }, + { + "epoch": 0.28007, + "grad_norm": 1.2057425550650578, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 28007 + }, + { + "epoch": 0.28008, + "grad_norm": 0.9701549388666415, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 28008 + }, + { + "epoch": 0.28009, + "grad_norm": 1.183509753544753, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 28009 + }, + { + "epoch": 0.2801, + "grad_norm": 1.1113765812780292, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 28010 + }, + { + "epoch": 0.28011, + "grad_norm": 0.9069772595398233, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 28011 + }, + { + "epoch": 0.28012, + "grad_norm": 0.9590939586837778, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 28012 + }, + { + "epoch": 0.28013, + "grad_norm": 1.0907199092979438, + "learning_rate": 0.003, + "loss": 4.044, + "step": 28013 + }, + { + "epoch": 0.28014, + "grad_norm": 0.8391370732239554, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 28014 + }, + { + "epoch": 0.28015, + "grad_norm": 0.8533808180521466, + "learning_rate": 0.003, + "loss": 4.045, + "step": 28015 + }, + { + "epoch": 0.28016, + "grad_norm": 0.8963963958445766, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 28016 + }, + { + "epoch": 0.28017, + "grad_norm": 0.9344338155742637, + "learning_rate": 0.003, + "loss": 4.077, + "step": 28017 + }, + { + "epoch": 0.28018, + "grad_norm": 0.908168948763108, + "learning_rate": 0.003, + "loss": 4.055, + "step": 28018 + }, + { + "epoch": 0.28019, + "grad_norm": 1.1130591867732542, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 28019 + }, + { + "epoch": 0.2802, + "grad_norm": 1.2749763351890182, + "learning_rate": 0.003, + "loss": 4.053, + "step": 28020 + }, + { + "epoch": 0.28021, + "grad_norm": 1.1257712219261962, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 28021 + }, + { + "epoch": 0.28022, + "grad_norm": 1.0502432434081928, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 28022 + }, + { + "epoch": 0.28023, + "grad_norm": 1.2628223613116036, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 28023 + }, + { + "epoch": 0.28024, + "grad_norm": 1.0922751581607313, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 28024 + }, + { + "epoch": 0.28025, + "grad_norm": 1.206501179983521, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 28025 + }, + { + "epoch": 0.28026, + "grad_norm": 1.0488547504867953, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 28026 + }, + { + "epoch": 0.28027, + "grad_norm": 0.8683122411782506, + "learning_rate": 0.003, + "loss": 4.049, + "step": 28027 + }, + { + "epoch": 0.28028, + "grad_norm": 0.8034658211823814, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 28028 + }, + { + "epoch": 0.28029, + "grad_norm": 0.7979080028000409, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 28029 + }, + { + "epoch": 0.2803, + "grad_norm": 0.8662784026188782, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 28030 + }, + { + "epoch": 0.28031, + "grad_norm": 1.003048738787509, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 28031 + }, + { + "epoch": 0.28032, + "grad_norm": 1.1555369868763679, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 28032 + }, + { + "epoch": 0.28033, + "grad_norm": 0.9099136546676041, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 28033 + }, + { + "epoch": 0.28034, + "grad_norm": 1.1603807362022038, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 28034 + }, + { + "epoch": 0.28035, + "grad_norm": 1.1702779943290118, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 28035 + }, + { + "epoch": 0.28036, + "grad_norm": 0.8498924238860622, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 28036 + }, + { + "epoch": 0.28037, + "grad_norm": 0.9445453571336643, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 28037 + }, + { + "epoch": 0.28038, + "grad_norm": 1.0937638382098624, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 28038 + }, + { + "epoch": 0.28039, + "grad_norm": 1.166060407352606, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 28039 + }, + { + "epoch": 0.2804, + "grad_norm": 1.187082207413196, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 28040 + }, + { + "epoch": 0.28041, + "grad_norm": 1.13375942612313, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 28041 + }, + { + "epoch": 0.28042, + "grad_norm": 1.1766117596215966, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 28042 + }, + { + "epoch": 0.28043, + "grad_norm": 1.086323769462954, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 28043 + }, + { + "epoch": 0.28044, + "grad_norm": 1.1451580044432692, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 28044 + }, + { + "epoch": 0.28045, + "grad_norm": 1.000989508692486, + "learning_rate": 0.003, + "loss": 4.046, + "step": 28045 + }, + { + "epoch": 0.28046, + "grad_norm": 1.1393796407311603, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 28046 + }, + { + "epoch": 0.28047, + "grad_norm": 1.085659313306965, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 28047 + }, + { + "epoch": 0.28048, + "grad_norm": 1.1514662306606251, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 28048 + }, + { + "epoch": 0.28049, + "grad_norm": 1.1801769939723072, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 28049 + }, + { + "epoch": 0.2805, + "grad_norm": 0.9563663050396374, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 28050 + }, + { + "epoch": 0.28051, + "grad_norm": 1.0290177519921484, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 28051 + }, + { + "epoch": 0.28052, + "grad_norm": 1.0432141809403437, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 28052 + }, + { + "epoch": 0.28053, + "grad_norm": 1.1436550131590186, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 28053 + }, + { + "epoch": 0.28054, + "grad_norm": 0.9953149168408209, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 28054 + }, + { + "epoch": 0.28055, + "grad_norm": 1.0750180462829513, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 28055 + }, + { + "epoch": 0.28056, + "grad_norm": 1.0141134878857279, + "learning_rate": 0.003, + "loss": 4.073, + "step": 28056 + }, + { + "epoch": 0.28057, + "grad_norm": 1.0389763569584853, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 28057 + }, + { + "epoch": 0.28058, + "grad_norm": 1.0455291979991104, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 28058 + }, + { + "epoch": 0.28059, + "grad_norm": 1.0679999724809706, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 28059 + }, + { + "epoch": 0.2806, + "grad_norm": 1.0725475607600317, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 28060 + }, + { + "epoch": 0.28061, + "grad_norm": 0.9985879614010028, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 28061 + }, + { + "epoch": 0.28062, + "grad_norm": 1.1796899389358586, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 28062 + }, + { + "epoch": 0.28063, + "grad_norm": 1.0199969955404704, + "learning_rate": 0.003, + "loss": 4.068, + "step": 28063 + }, + { + "epoch": 0.28064, + "grad_norm": 1.1466169917940559, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 28064 + }, + { + "epoch": 0.28065, + "grad_norm": 1.065200796692506, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 28065 + }, + { + "epoch": 0.28066, + "grad_norm": 1.3954782856737884, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 28066 + }, + { + "epoch": 0.28067, + "grad_norm": 1.0381783794875656, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 28067 + }, + { + "epoch": 0.28068, + "grad_norm": 1.2202840602245801, + "learning_rate": 0.003, + "loss": 4.05, + "step": 28068 + }, + { + "epoch": 0.28069, + "grad_norm": 1.2057422666916606, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 28069 + }, + { + "epoch": 0.2807, + "grad_norm": 1.0177991330493104, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 28070 + }, + { + "epoch": 0.28071, + "grad_norm": 1.1905559615198478, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 28071 + }, + { + "epoch": 0.28072, + "grad_norm": 1.1409238616153563, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 28072 + }, + { + "epoch": 0.28073, + "grad_norm": 1.0465518724479896, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 28073 + }, + { + "epoch": 0.28074, + "grad_norm": 1.036165482022467, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 28074 + }, + { + "epoch": 0.28075, + "grad_norm": 1.0192199642707274, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 28075 + }, + { + "epoch": 0.28076, + "grad_norm": 1.0474168669659096, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 28076 + }, + { + "epoch": 0.28077, + "grad_norm": 1.0189815718045283, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 28077 + }, + { + "epoch": 0.28078, + "grad_norm": 1.1391396257927284, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 28078 + }, + { + "epoch": 0.28079, + "grad_norm": 0.8635575293156389, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 28079 + }, + { + "epoch": 0.2808, + "grad_norm": 1.1535014583277308, + "learning_rate": 0.003, + "loss": 4.059, + "step": 28080 + }, + { + "epoch": 0.28081, + "grad_norm": 1.0325122132082942, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 28081 + }, + { + "epoch": 0.28082, + "grad_norm": 1.1022941523944323, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 28082 + }, + { + "epoch": 0.28083, + "grad_norm": 0.9048117228880886, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 28083 + }, + { + "epoch": 0.28084, + "grad_norm": 0.9767040363702567, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 28084 + }, + { + "epoch": 0.28085, + "grad_norm": 0.9987428263169941, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 28085 + }, + { + "epoch": 0.28086, + "grad_norm": 1.2014956507328889, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 28086 + }, + { + "epoch": 0.28087, + "grad_norm": 1.086993858324205, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 28087 + }, + { + "epoch": 0.28088, + "grad_norm": 1.0670346851721577, + "learning_rate": 0.003, + "loss": 4.045, + "step": 28088 + }, + { + "epoch": 0.28089, + "grad_norm": 1.282804646920521, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 28089 + }, + { + "epoch": 0.2809, + "grad_norm": 1.178199755905427, + "learning_rate": 0.003, + "loss": 4.031, + "step": 28090 + }, + { + "epoch": 0.28091, + "grad_norm": 1.347969432977879, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 28091 + }, + { + "epoch": 0.28092, + "grad_norm": 0.9520233105402853, + "learning_rate": 0.003, + "loss": 4.0855, + "step": 28092 + }, + { + "epoch": 0.28093, + "grad_norm": 1.044939966703711, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 28093 + }, + { + "epoch": 0.28094, + "grad_norm": 1.302977007200758, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 28094 + }, + { + "epoch": 0.28095, + "grad_norm": 0.9240325459351697, + "learning_rate": 0.003, + "loss": 4.051, + "step": 28095 + }, + { + "epoch": 0.28096, + "grad_norm": 1.0574982699832045, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 28096 + }, + { + "epoch": 0.28097, + "grad_norm": 1.0122746038187553, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 28097 + }, + { + "epoch": 0.28098, + "grad_norm": 1.0164218903149915, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 28098 + }, + { + "epoch": 0.28099, + "grad_norm": 1.1079455353141734, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 28099 + }, + { + "epoch": 0.281, + "grad_norm": 1.2568442376127298, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 28100 + }, + { + "epoch": 0.28101, + "grad_norm": 0.9407488736544098, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 28101 + }, + { + "epoch": 0.28102, + "grad_norm": 0.9017962166001826, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 28102 + }, + { + "epoch": 0.28103, + "grad_norm": 1.0390691365741407, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 28103 + }, + { + "epoch": 0.28104, + "grad_norm": 1.0868556641975315, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 28104 + }, + { + "epoch": 0.28105, + "grad_norm": 0.7945912973529994, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 28105 + }, + { + "epoch": 0.28106, + "grad_norm": 0.8112436487763289, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 28106 + }, + { + "epoch": 0.28107, + "grad_norm": 0.8650605261163612, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 28107 + }, + { + "epoch": 0.28108, + "grad_norm": 0.8852291226350714, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 28108 + }, + { + "epoch": 0.28109, + "grad_norm": 1.0015737266403317, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 28109 + }, + { + "epoch": 0.2811, + "grad_norm": 1.3937481021124647, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 28110 + }, + { + "epoch": 0.28111, + "grad_norm": 1.0297746803046646, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 28111 + }, + { + "epoch": 0.28112, + "grad_norm": 1.1048204828866965, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 28112 + }, + { + "epoch": 0.28113, + "grad_norm": 1.110059470242111, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 28113 + }, + { + "epoch": 0.28114, + "grad_norm": 1.007083987333921, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 28114 + }, + { + "epoch": 0.28115, + "grad_norm": 1.2600214914523415, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 28115 + }, + { + "epoch": 0.28116, + "grad_norm": 0.990559271018209, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 28116 + }, + { + "epoch": 0.28117, + "grad_norm": 1.1530973301168477, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 28117 + }, + { + "epoch": 0.28118, + "grad_norm": 0.955635063148254, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 28118 + }, + { + "epoch": 0.28119, + "grad_norm": 0.9912824948580561, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 28119 + }, + { + "epoch": 0.2812, + "grad_norm": 1.0809725176212388, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 28120 + }, + { + "epoch": 0.28121, + "grad_norm": 1.170427915989388, + "learning_rate": 0.003, + "loss": 4.046, + "step": 28121 + }, + { + "epoch": 0.28122, + "grad_norm": 1.0814015261449652, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 28122 + }, + { + "epoch": 0.28123, + "grad_norm": 1.1177956711994093, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 28123 + }, + { + "epoch": 0.28124, + "grad_norm": 1.1052944418934578, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 28124 + }, + { + "epoch": 0.28125, + "grad_norm": 1.0013776522791564, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 28125 + }, + { + "epoch": 0.28126, + "grad_norm": 1.0552117283520968, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 28126 + }, + { + "epoch": 0.28127, + "grad_norm": 1.0759669534104366, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 28127 + }, + { + "epoch": 0.28128, + "grad_norm": 0.981253747331319, + "learning_rate": 0.003, + "loss": 4.041, + "step": 28128 + }, + { + "epoch": 0.28129, + "grad_norm": 0.9341853831456706, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 28129 + }, + { + "epoch": 0.2813, + "grad_norm": 1.0412293379083855, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 28130 + }, + { + "epoch": 0.28131, + "grad_norm": 1.2420537290662481, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 28131 + }, + { + "epoch": 0.28132, + "grad_norm": 0.9838847133914558, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 28132 + }, + { + "epoch": 0.28133, + "grad_norm": 0.952441648139469, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 28133 + }, + { + "epoch": 0.28134, + "grad_norm": 1.1679095801291768, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 28134 + }, + { + "epoch": 0.28135, + "grad_norm": 0.9022484178297141, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 28135 + }, + { + "epoch": 0.28136, + "grad_norm": 1.0451403649681186, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 28136 + }, + { + "epoch": 0.28137, + "grad_norm": 1.2232969423053512, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 28137 + }, + { + "epoch": 0.28138, + "grad_norm": 1.051223374795255, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 28138 + }, + { + "epoch": 0.28139, + "grad_norm": 1.1660735290333264, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 28139 + }, + { + "epoch": 0.2814, + "grad_norm": 0.801785661163332, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 28140 + }, + { + "epoch": 0.28141, + "grad_norm": 0.9951835591898865, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 28141 + }, + { + "epoch": 0.28142, + "grad_norm": 1.3012414292886334, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 28142 + }, + { + "epoch": 0.28143, + "grad_norm": 0.9484512407278357, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 28143 + }, + { + "epoch": 0.28144, + "grad_norm": 1.1184533377651906, + "learning_rate": 0.003, + "loss": 4.1095, + "step": 28144 + }, + { + "epoch": 0.28145, + "grad_norm": 1.0929017658060343, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 28145 + }, + { + "epoch": 0.28146, + "grad_norm": 1.4397817575869798, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 28146 + }, + { + "epoch": 0.28147, + "grad_norm": 0.9855681830394056, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 28147 + }, + { + "epoch": 0.28148, + "grad_norm": 1.163015188281615, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 28148 + }, + { + "epoch": 0.28149, + "grad_norm": 1.1751564872186115, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 28149 + }, + { + "epoch": 0.2815, + "grad_norm": 1.1238404588396051, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 28150 + }, + { + "epoch": 0.28151, + "grad_norm": 1.0968972263365007, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 28151 + }, + { + "epoch": 0.28152, + "grad_norm": 1.1987033437623744, + "learning_rate": 0.003, + "loss": 4.035, + "step": 28152 + }, + { + "epoch": 0.28153, + "grad_norm": 0.9383980493050469, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 28153 + }, + { + "epoch": 0.28154, + "grad_norm": 1.0693812798210416, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 28154 + }, + { + "epoch": 0.28155, + "grad_norm": 1.0107242037309594, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 28155 + }, + { + "epoch": 0.28156, + "grad_norm": 1.0064214402445364, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 28156 + }, + { + "epoch": 0.28157, + "grad_norm": 1.1348958298532257, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 28157 + }, + { + "epoch": 0.28158, + "grad_norm": 1.191580657984124, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 28158 + }, + { + "epoch": 0.28159, + "grad_norm": 0.8823134699247049, + "learning_rate": 0.003, + "loss": 4.044, + "step": 28159 + }, + { + "epoch": 0.2816, + "grad_norm": 0.8370141433552193, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 28160 + }, + { + "epoch": 0.28161, + "grad_norm": 0.978128327461552, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 28161 + }, + { + "epoch": 0.28162, + "grad_norm": 1.3310426795025694, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 28162 + }, + { + "epoch": 0.28163, + "grad_norm": 0.8009458150920411, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 28163 + }, + { + "epoch": 0.28164, + "grad_norm": 0.7660080950132938, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 28164 + }, + { + "epoch": 0.28165, + "grad_norm": 0.9034337555573274, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 28165 + }, + { + "epoch": 0.28166, + "grad_norm": 0.9239846256794597, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 28166 + }, + { + "epoch": 0.28167, + "grad_norm": 1.2294069217831989, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 28167 + }, + { + "epoch": 0.28168, + "grad_norm": 1.243795959301877, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 28168 + }, + { + "epoch": 0.28169, + "grad_norm": 0.914043423742592, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 28169 + }, + { + "epoch": 0.2817, + "grad_norm": 0.8872266688807496, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 28170 + }, + { + "epoch": 0.28171, + "grad_norm": 1.0890740300338435, + "learning_rate": 0.003, + "loss": 4.067, + "step": 28171 + }, + { + "epoch": 0.28172, + "grad_norm": 1.4264943698391492, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 28172 + }, + { + "epoch": 0.28173, + "grad_norm": 0.8474195729272797, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 28173 + }, + { + "epoch": 0.28174, + "grad_norm": 0.8607961755476737, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 28174 + }, + { + "epoch": 0.28175, + "grad_norm": 0.9638586815606993, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 28175 + }, + { + "epoch": 0.28176, + "grad_norm": 1.0653491490251605, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 28176 + }, + { + "epoch": 0.28177, + "grad_norm": 1.0317909164971861, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 28177 + }, + { + "epoch": 0.28178, + "grad_norm": 0.9461302223070777, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 28178 + }, + { + "epoch": 0.28179, + "grad_norm": 0.9853783573391884, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 28179 + }, + { + "epoch": 0.2818, + "grad_norm": 1.1182738907715017, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 28180 + }, + { + "epoch": 0.28181, + "grad_norm": 1.23022575470754, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 28181 + }, + { + "epoch": 0.28182, + "grad_norm": 1.3014295793952553, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 28182 + }, + { + "epoch": 0.28183, + "grad_norm": 1.0866320489178587, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 28183 + }, + { + "epoch": 0.28184, + "grad_norm": 0.9650865810751301, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 28184 + }, + { + "epoch": 0.28185, + "grad_norm": 1.0643652121153104, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 28185 + }, + { + "epoch": 0.28186, + "grad_norm": 1.0693197387749633, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 28186 + }, + { + "epoch": 0.28187, + "grad_norm": 1.2110201686894864, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 28187 + }, + { + "epoch": 0.28188, + "grad_norm": 1.0066884932574096, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 28188 + }, + { + "epoch": 0.28189, + "grad_norm": 1.2252777069750052, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 28189 + }, + { + "epoch": 0.2819, + "grad_norm": 1.0677379440555081, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 28190 + }, + { + "epoch": 0.28191, + "grad_norm": 1.0433083727810526, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 28191 + }, + { + "epoch": 0.28192, + "grad_norm": 1.0873929622273952, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 28192 + }, + { + "epoch": 0.28193, + "grad_norm": 1.536097370283987, + "learning_rate": 0.003, + "loss": 4.026, + "step": 28193 + }, + { + "epoch": 0.28194, + "grad_norm": 0.9040081612517019, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 28194 + }, + { + "epoch": 0.28195, + "grad_norm": 0.9346769024674172, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 28195 + }, + { + "epoch": 0.28196, + "grad_norm": 1.052818893071318, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 28196 + }, + { + "epoch": 0.28197, + "grad_norm": 1.281047655014127, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 28197 + }, + { + "epoch": 0.28198, + "grad_norm": 0.8298899166388599, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 28198 + }, + { + "epoch": 0.28199, + "grad_norm": 0.8501892640083643, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 28199 + }, + { + "epoch": 0.282, + "grad_norm": 1.0320305651519852, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 28200 + }, + { + "epoch": 0.28201, + "grad_norm": 1.086923180443239, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 28201 + }, + { + "epoch": 0.28202, + "grad_norm": 0.9103571463139288, + "learning_rate": 0.003, + "loss": 4.032, + "step": 28202 + }, + { + "epoch": 0.28203, + "grad_norm": 1.0614561929359572, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 28203 + }, + { + "epoch": 0.28204, + "grad_norm": 1.3104355849068865, + "learning_rate": 0.003, + "loss": 4.065, + "step": 28204 + }, + { + "epoch": 0.28205, + "grad_norm": 1.0619275830885846, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 28205 + }, + { + "epoch": 0.28206, + "grad_norm": 0.9668589136819111, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 28206 + }, + { + "epoch": 0.28207, + "grad_norm": 0.974535681225248, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 28207 + }, + { + "epoch": 0.28208, + "grad_norm": 1.2060440192103798, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 28208 + }, + { + "epoch": 0.28209, + "grad_norm": 1.0382005799974814, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 28209 + }, + { + "epoch": 0.2821, + "grad_norm": 1.0710402820048333, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 28210 + }, + { + "epoch": 0.28211, + "grad_norm": 0.9684389514850921, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 28211 + }, + { + "epoch": 0.28212, + "grad_norm": 1.1122567952995537, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 28212 + }, + { + "epoch": 0.28213, + "grad_norm": 1.0969398011084033, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 28213 + }, + { + "epoch": 0.28214, + "grad_norm": 1.0252694984933486, + "learning_rate": 0.003, + "loss": 4.041, + "step": 28214 + }, + { + "epoch": 0.28215, + "grad_norm": 1.0429130421035098, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 28215 + }, + { + "epoch": 0.28216, + "grad_norm": 1.2292100093445246, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 28216 + }, + { + "epoch": 0.28217, + "grad_norm": 0.8992928662234442, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 28217 + }, + { + "epoch": 0.28218, + "grad_norm": 1.081824117923006, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 28218 + }, + { + "epoch": 0.28219, + "grad_norm": 1.0798425161753589, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 28219 + }, + { + "epoch": 0.2822, + "grad_norm": 0.9765230806264232, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 28220 + }, + { + "epoch": 0.28221, + "grad_norm": 1.1557725647591197, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 28221 + }, + { + "epoch": 0.28222, + "grad_norm": 1.0416021208677633, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 28222 + }, + { + "epoch": 0.28223, + "grad_norm": 1.264676772685101, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 28223 + }, + { + "epoch": 0.28224, + "grad_norm": 1.0258850293022328, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 28224 + }, + { + "epoch": 0.28225, + "grad_norm": 1.0975212102810166, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 28225 + }, + { + "epoch": 0.28226, + "grad_norm": 1.1189269923286336, + "learning_rate": 0.003, + "loss": 4.033, + "step": 28226 + }, + { + "epoch": 0.28227, + "grad_norm": 1.0901727994705908, + "learning_rate": 0.003, + "loss": 4.036, + "step": 28227 + }, + { + "epoch": 0.28228, + "grad_norm": 1.3034622457289968, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 28228 + }, + { + "epoch": 0.28229, + "grad_norm": 0.8438998375734027, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 28229 + }, + { + "epoch": 0.2823, + "grad_norm": 0.7524942942522401, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 28230 + }, + { + "epoch": 0.28231, + "grad_norm": 0.8576374408560764, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 28231 + }, + { + "epoch": 0.28232, + "grad_norm": 1.1976159715542707, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 28232 + }, + { + "epoch": 0.28233, + "grad_norm": 1.1007839195311198, + "learning_rate": 0.003, + "loss": 4.048, + "step": 28233 + }, + { + "epoch": 0.28234, + "grad_norm": 1.0158482125785153, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 28234 + }, + { + "epoch": 0.28235, + "grad_norm": 1.1811763101112704, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 28235 + }, + { + "epoch": 0.28236, + "grad_norm": 1.014753320540679, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 28236 + }, + { + "epoch": 0.28237, + "grad_norm": 1.0887813293322766, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 28237 + }, + { + "epoch": 0.28238, + "grad_norm": 1.1252527856344128, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 28238 + }, + { + "epoch": 0.28239, + "grad_norm": 0.9864298385728518, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 28239 + }, + { + "epoch": 0.2824, + "grad_norm": 1.1055276424907559, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 28240 + }, + { + "epoch": 0.28241, + "grad_norm": 0.9618151846020927, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 28241 + }, + { + "epoch": 0.28242, + "grad_norm": 0.9524421458672689, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 28242 + }, + { + "epoch": 0.28243, + "grad_norm": 0.9233009450565519, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 28243 + }, + { + "epoch": 0.28244, + "grad_norm": 1.057958227117754, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 28244 + }, + { + "epoch": 0.28245, + "grad_norm": 1.0413345973302333, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 28245 + }, + { + "epoch": 0.28246, + "grad_norm": 1.2420656970963393, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 28246 + }, + { + "epoch": 0.28247, + "grad_norm": 1.2720490612762634, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 28247 + }, + { + "epoch": 0.28248, + "grad_norm": 1.118007426026991, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 28248 + }, + { + "epoch": 0.28249, + "grad_norm": 1.0447103480911777, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 28249 + }, + { + "epoch": 0.2825, + "grad_norm": 1.1523589773680654, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 28250 + }, + { + "epoch": 0.28251, + "grad_norm": 1.119352866390661, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 28251 + }, + { + "epoch": 0.28252, + "grad_norm": 1.011144114505086, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 28252 + }, + { + "epoch": 0.28253, + "grad_norm": 1.2043372545041888, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 28253 + }, + { + "epoch": 0.28254, + "grad_norm": 0.9270039157257257, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 28254 + }, + { + "epoch": 0.28255, + "grad_norm": 1.025664130053844, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 28255 + }, + { + "epoch": 0.28256, + "grad_norm": 1.1020860927721463, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 28256 + }, + { + "epoch": 0.28257, + "grad_norm": 1.0228379758533521, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 28257 + }, + { + "epoch": 0.28258, + "grad_norm": 1.0467268915062646, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 28258 + }, + { + "epoch": 0.28259, + "grad_norm": 1.333229613020862, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 28259 + }, + { + "epoch": 0.2826, + "grad_norm": 1.0533389448915238, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 28260 + }, + { + "epoch": 0.28261, + "grad_norm": 1.0688766282633644, + "learning_rate": 0.003, + "loss": 4.005, + "step": 28261 + }, + { + "epoch": 0.28262, + "grad_norm": 0.9476592770398798, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 28262 + }, + { + "epoch": 0.28263, + "grad_norm": 1.0624263945775094, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 28263 + }, + { + "epoch": 0.28264, + "grad_norm": 1.241216825716021, + "learning_rate": 0.003, + "loss": 4.025, + "step": 28264 + }, + { + "epoch": 0.28265, + "grad_norm": 0.7619472335541417, + "learning_rate": 0.003, + "loss": 4.054, + "step": 28265 + }, + { + "epoch": 0.28266, + "grad_norm": 0.6950006764232824, + "learning_rate": 0.003, + "loss": 4.06, + "step": 28266 + }, + { + "epoch": 0.28267, + "grad_norm": 0.9063899438592498, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 28267 + }, + { + "epoch": 0.28268, + "grad_norm": 1.141566940116779, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 28268 + }, + { + "epoch": 0.28269, + "grad_norm": 0.9303264800304936, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 28269 + }, + { + "epoch": 0.2827, + "grad_norm": 1.1506579872248521, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 28270 + }, + { + "epoch": 0.28271, + "grad_norm": 1.0991736860014907, + "learning_rate": 0.003, + "loss": 4.0961, + "step": 28271 + }, + { + "epoch": 0.28272, + "grad_norm": 1.1763315956396536, + "learning_rate": 0.003, + "loss": 4.012, + "step": 28272 + }, + { + "epoch": 0.28273, + "grad_norm": 1.0882494907285427, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 28273 + }, + { + "epoch": 0.28274, + "grad_norm": 1.1798621762756956, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 28274 + }, + { + "epoch": 0.28275, + "grad_norm": 1.0049143832376641, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 28275 + }, + { + "epoch": 0.28276, + "grad_norm": 0.9869179753919637, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 28276 + }, + { + "epoch": 0.28277, + "grad_norm": 1.1696530864262678, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 28277 + }, + { + "epoch": 0.28278, + "grad_norm": 1.0788422928774632, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 28278 + }, + { + "epoch": 0.28279, + "grad_norm": 1.0014424693032098, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 28279 + }, + { + "epoch": 0.2828, + "grad_norm": 1.2927814453530364, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 28280 + }, + { + "epoch": 0.28281, + "grad_norm": 0.9424752473091897, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 28281 + }, + { + "epoch": 0.28282, + "grad_norm": 1.1516176462512568, + "learning_rate": 0.003, + "loss": 4.078, + "step": 28282 + }, + { + "epoch": 0.28283, + "grad_norm": 1.3425830597436335, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 28283 + }, + { + "epoch": 0.28284, + "grad_norm": 0.8549269830846674, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 28284 + }, + { + "epoch": 0.28285, + "grad_norm": 0.8761600350662694, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 28285 + }, + { + "epoch": 0.28286, + "grad_norm": 1.151945906340714, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 28286 + }, + { + "epoch": 0.28287, + "grad_norm": 1.2700856989088531, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 28287 + }, + { + "epoch": 0.28288, + "grad_norm": 1.013103010190782, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 28288 + }, + { + "epoch": 0.28289, + "grad_norm": 1.048326918978169, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 28289 + }, + { + "epoch": 0.2829, + "grad_norm": 1.0397799813608235, + "learning_rate": 0.003, + "loss": 4.023, + "step": 28290 + }, + { + "epoch": 0.28291, + "grad_norm": 0.969860565690893, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 28291 + }, + { + "epoch": 0.28292, + "grad_norm": 1.0817598824304053, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 28292 + }, + { + "epoch": 0.28293, + "grad_norm": 1.036404075202368, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 28293 + }, + { + "epoch": 0.28294, + "grad_norm": 1.197272639471192, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 28294 + }, + { + "epoch": 0.28295, + "grad_norm": 1.0304935398197532, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 28295 + }, + { + "epoch": 0.28296, + "grad_norm": 1.186185113786614, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 28296 + }, + { + "epoch": 0.28297, + "grad_norm": 1.0606054999523948, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 28297 + }, + { + "epoch": 0.28298, + "grad_norm": 1.0361422945952865, + "learning_rate": 0.003, + "loss": 4.057, + "step": 28298 + }, + { + "epoch": 0.28299, + "grad_norm": 1.1766799011801075, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 28299 + }, + { + "epoch": 0.283, + "grad_norm": 1.1206929202965663, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 28300 + }, + { + "epoch": 0.28301, + "grad_norm": 1.4331758457252008, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 28301 + }, + { + "epoch": 0.28302, + "grad_norm": 0.9576089348374642, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 28302 + }, + { + "epoch": 0.28303, + "grad_norm": 0.8938145980387027, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 28303 + }, + { + "epoch": 0.28304, + "grad_norm": 1.0027516317879572, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 28304 + }, + { + "epoch": 0.28305, + "grad_norm": 1.522939024690305, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 28305 + }, + { + "epoch": 0.28306, + "grad_norm": 1.0181503714116176, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 28306 + }, + { + "epoch": 0.28307, + "grad_norm": 1.1144399598831498, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 28307 + }, + { + "epoch": 0.28308, + "grad_norm": 0.915617510099284, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 28308 + }, + { + "epoch": 0.28309, + "grad_norm": 0.8843905395128456, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 28309 + }, + { + "epoch": 0.2831, + "grad_norm": 1.0393617016102128, + "learning_rate": 0.003, + "loss": 4.046, + "step": 28310 + }, + { + "epoch": 0.28311, + "grad_norm": 1.2363214467579828, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 28311 + }, + { + "epoch": 0.28312, + "grad_norm": 0.9661753524454273, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 28312 + }, + { + "epoch": 0.28313, + "grad_norm": 1.143577273528379, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 28313 + }, + { + "epoch": 0.28314, + "grad_norm": 1.2211421849707043, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 28314 + }, + { + "epoch": 0.28315, + "grad_norm": 0.906290945546779, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 28315 + }, + { + "epoch": 0.28316, + "grad_norm": 1.1518965447404754, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 28316 + }, + { + "epoch": 0.28317, + "grad_norm": 1.1268824759086766, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 28317 + }, + { + "epoch": 0.28318, + "grad_norm": 0.9568669427277248, + "learning_rate": 0.003, + "loss": 4.054, + "step": 28318 + }, + { + "epoch": 0.28319, + "grad_norm": 0.9079376424706331, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 28319 + }, + { + "epoch": 0.2832, + "grad_norm": 1.0477449968745096, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 28320 + }, + { + "epoch": 0.28321, + "grad_norm": 1.1989135961508626, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 28321 + }, + { + "epoch": 0.28322, + "grad_norm": 1.0711698857845413, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 28322 + }, + { + "epoch": 0.28323, + "grad_norm": 1.2341940058502108, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 28323 + }, + { + "epoch": 0.28324, + "grad_norm": 0.8953767203079651, + "learning_rate": 0.003, + "loss": 4.04, + "step": 28324 + }, + { + "epoch": 0.28325, + "grad_norm": 0.9975942456926636, + "learning_rate": 0.003, + "loss": 4.0881, + "step": 28325 + }, + { + "epoch": 0.28326, + "grad_norm": 1.431194951933343, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 28326 + }, + { + "epoch": 0.28327, + "grad_norm": 0.8246811986181679, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 28327 + }, + { + "epoch": 0.28328, + "grad_norm": 0.7981685704746961, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 28328 + }, + { + "epoch": 0.28329, + "grad_norm": 0.9478460067282135, + "learning_rate": 0.003, + "loss": 4.038, + "step": 28329 + }, + { + "epoch": 0.2833, + "grad_norm": 1.0862208581953836, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 28330 + }, + { + "epoch": 0.28331, + "grad_norm": 1.1558841023477697, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 28331 + }, + { + "epoch": 0.28332, + "grad_norm": 1.147286978527018, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 28332 + }, + { + "epoch": 0.28333, + "grad_norm": 1.2928230400359952, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 28333 + }, + { + "epoch": 0.28334, + "grad_norm": 0.9558989371929919, + "learning_rate": 0.003, + "loss": 4.0845, + "step": 28334 + }, + { + "epoch": 0.28335, + "grad_norm": 0.8685798280199282, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 28335 + }, + { + "epoch": 0.28336, + "grad_norm": 1.0191460762225468, + "learning_rate": 0.003, + "loss": 4.03, + "step": 28336 + }, + { + "epoch": 0.28337, + "grad_norm": 1.15715567218559, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 28337 + }, + { + "epoch": 0.28338, + "grad_norm": 0.9696049935350747, + "learning_rate": 0.003, + "loss": 4.047, + "step": 28338 + }, + { + "epoch": 0.28339, + "grad_norm": 1.1593006120806522, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 28339 + }, + { + "epoch": 0.2834, + "grad_norm": 1.2546676531668346, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 28340 + }, + { + "epoch": 0.28341, + "grad_norm": 1.0313251123575493, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 28341 + }, + { + "epoch": 0.28342, + "grad_norm": 1.1254827441484725, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 28342 + }, + { + "epoch": 0.28343, + "grad_norm": 1.19314419915888, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 28343 + }, + { + "epoch": 0.28344, + "grad_norm": 0.8211805445039736, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 28344 + }, + { + "epoch": 0.28345, + "grad_norm": 0.9899582559599546, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 28345 + }, + { + "epoch": 0.28346, + "grad_norm": 1.231913523084874, + "learning_rate": 0.003, + "loss": 4.077, + "step": 28346 + }, + { + "epoch": 0.28347, + "grad_norm": 0.9038901024431586, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 28347 + }, + { + "epoch": 0.28348, + "grad_norm": 1.2684185460794994, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 28348 + }, + { + "epoch": 0.28349, + "grad_norm": 1.0401124577474237, + "learning_rate": 0.003, + "loss": 4.05, + "step": 28349 + }, + { + "epoch": 0.2835, + "grad_norm": 1.229915956882832, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 28350 + }, + { + "epoch": 0.28351, + "grad_norm": 0.9173193880857025, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 28351 + }, + { + "epoch": 0.28352, + "grad_norm": 0.9787707881306342, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 28352 + }, + { + "epoch": 0.28353, + "grad_norm": 1.1028283453302952, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 28353 + }, + { + "epoch": 0.28354, + "grad_norm": 1.0464010022377028, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 28354 + }, + { + "epoch": 0.28355, + "grad_norm": 1.0466328680743693, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 28355 + }, + { + "epoch": 0.28356, + "grad_norm": 1.1101041052881206, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 28356 + }, + { + "epoch": 0.28357, + "grad_norm": 0.9617995262030797, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 28357 + }, + { + "epoch": 0.28358, + "grad_norm": 0.8803415534128805, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 28358 + }, + { + "epoch": 0.28359, + "grad_norm": 0.9240717430783809, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 28359 + }, + { + "epoch": 0.2836, + "grad_norm": 1.1582807663690458, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 28360 + }, + { + "epoch": 0.28361, + "grad_norm": 1.0458876321162958, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 28361 + }, + { + "epoch": 0.28362, + "grad_norm": 1.1004283557481633, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 28362 + }, + { + "epoch": 0.28363, + "grad_norm": 1.1381415839454103, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 28363 + }, + { + "epoch": 0.28364, + "grad_norm": 1.0054255120075266, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 28364 + }, + { + "epoch": 0.28365, + "grad_norm": 1.1826856653069198, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 28365 + }, + { + "epoch": 0.28366, + "grad_norm": 1.1783793338970854, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 28366 + }, + { + "epoch": 0.28367, + "grad_norm": 0.9343491259232387, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 28367 + }, + { + "epoch": 0.28368, + "grad_norm": 0.927092609620482, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 28368 + }, + { + "epoch": 0.28369, + "grad_norm": 0.9928450303748831, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 28369 + }, + { + "epoch": 0.2837, + "grad_norm": 1.1672759276419904, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 28370 + }, + { + "epoch": 0.28371, + "grad_norm": 0.9513400142152404, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 28371 + }, + { + "epoch": 0.28372, + "grad_norm": 1.14500912567787, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 28372 + }, + { + "epoch": 0.28373, + "grad_norm": 1.019937491163225, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 28373 + }, + { + "epoch": 0.28374, + "grad_norm": 1.2860466732385891, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 28374 + }, + { + "epoch": 0.28375, + "grad_norm": 0.9220110546886067, + "learning_rate": 0.003, + "loss": 4.057, + "step": 28375 + }, + { + "epoch": 0.28376, + "grad_norm": 1.0182727716602025, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 28376 + }, + { + "epoch": 0.28377, + "grad_norm": 1.1005388291545544, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 28377 + }, + { + "epoch": 0.28378, + "grad_norm": 1.0389404649211718, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 28378 + }, + { + "epoch": 0.28379, + "grad_norm": 1.2945771911985238, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 28379 + }, + { + "epoch": 0.2838, + "grad_norm": 0.8648064848107098, + "learning_rate": 0.003, + "loss": 4.035, + "step": 28380 + }, + { + "epoch": 0.28381, + "grad_norm": 1.0438407072560247, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 28381 + }, + { + "epoch": 0.28382, + "grad_norm": 1.4708561266598363, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 28382 + }, + { + "epoch": 0.28383, + "grad_norm": 0.9298952675255923, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 28383 + }, + { + "epoch": 0.28384, + "grad_norm": 0.991125988936176, + "learning_rate": 0.003, + "loss": 4.046, + "step": 28384 + }, + { + "epoch": 0.28385, + "grad_norm": 1.0806293759563728, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 28385 + }, + { + "epoch": 0.28386, + "grad_norm": 1.106198706438275, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 28386 + }, + { + "epoch": 0.28387, + "grad_norm": 1.047474750376261, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 28387 + }, + { + "epoch": 0.28388, + "grad_norm": 1.0791213773275001, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 28388 + }, + { + "epoch": 0.28389, + "grad_norm": 0.968282860693668, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 28389 + }, + { + "epoch": 0.2839, + "grad_norm": 1.0446191979413944, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 28390 + }, + { + "epoch": 0.28391, + "grad_norm": 1.0956702895555805, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 28391 + }, + { + "epoch": 0.28392, + "grad_norm": 1.1363661192460763, + "learning_rate": 0.003, + "loss": 4.036, + "step": 28392 + }, + { + "epoch": 0.28393, + "grad_norm": 1.073168242163969, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 28393 + }, + { + "epoch": 0.28394, + "grad_norm": 1.1419840428929897, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 28394 + }, + { + "epoch": 0.28395, + "grad_norm": 1.0766783460612437, + "learning_rate": 0.003, + "loss": 4.0941, + "step": 28395 + }, + { + "epoch": 0.28396, + "grad_norm": 1.221406414609383, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 28396 + }, + { + "epoch": 0.28397, + "grad_norm": 0.9873835199728597, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 28397 + }, + { + "epoch": 0.28398, + "grad_norm": 1.1695219309111249, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 28398 + }, + { + "epoch": 0.28399, + "grad_norm": 1.1014307314535356, + "learning_rate": 0.003, + "loss": 4.035, + "step": 28399 + }, + { + "epoch": 0.284, + "grad_norm": 1.0339579865013526, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 28400 + }, + { + "epoch": 0.28401, + "grad_norm": 1.1472160971590784, + "learning_rate": 0.003, + "loss": 4.048, + "step": 28401 + }, + { + "epoch": 0.28402, + "grad_norm": 0.9672425932641868, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 28402 + }, + { + "epoch": 0.28403, + "grad_norm": 1.155010863767048, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 28403 + }, + { + "epoch": 0.28404, + "grad_norm": 1.0528011405545494, + "learning_rate": 0.003, + "loss": 4.042, + "step": 28404 + }, + { + "epoch": 0.28405, + "grad_norm": 1.096032073887284, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 28405 + }, + { + "epoch": 0.28406, + "grad_norm": 0.9998744871533132, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 28406 + }, + { + "epoch": 0.28407, + "grad_norm": 1.19454751286097, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 28407 + }, + { + "epoch": 0.28408, + "grad_norm": 1.0100365749933629, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 28408 + }, + { + "epoch": 0.28409, + "grad_norm": 1.1646024422958763, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 28409 + }, + { + "epoch": 0.2841, + "grad_norm": 0.8941324851235812, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 28410 + }, + { + "epoch": 0.28411, + "grad_norm": 0.9007002497642388, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 28411 + }, + { + "epoch": 0.28412, + "grad_norm": 0.998857235963949, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 28412 + }, + { + "epoch": 0.28413, + "grad_norm": 1.1840991981579811, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 28413 + }, + { + "epoch": 0.28414, + "grad_norm": 0.9664105201213299, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 28414 + }, + { + "epoch": 0.28415, + "grad_norm": 1.0181617272865289, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 28415 + }, + { + "epoch": 0.28416, + "grad_norm": 1.0746069700551615, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 28416 + }, + { + "epoch": 0.28417, + "grad_norm": 0.9325511315030978, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 28417 + }, + { + "epoch": 0.28418, + "grad_norm": 1.0710292895629243, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 28418 + }, + { + "epoch": 0.28419, + "grad_norm": 1.0169646724571675, + "learning_rate": 0.003, + "loss": 4.0918, + "step": 28419 + }, + { + "epoch": 0.2842, + "grad_norm": 1.1995319602897447, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 28420 + }, + { + "epoch": 0.28421, + "grad_norm": 1.417334119369154, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 28421 + }, + { + "epoch": 0.28422, + "grad_norm": 1.0519873390965624, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 28422 + }, + { + "epoch": 0.28423, + "grad_norm": 1.0861784605360363, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 28423 + }, + { + "epoch": 0.28424, + "grad_norm": 0.9028705969992703, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 28424 + }, + { + "epoch": 0.28425, + "grad_norm": 1.0307130638895354, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 28425 + }, + { + "epoch": 0.28426, + "grad_norm": 1.5232923163387582, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 28426 + }, + { + "epoch": 0.28427, + "grad_norm": 0.7890594314659162, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 28427 + }, + { + "epoch": 0.28428, + "grad_norm": 0.8176876681169359, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 28428 + }, + { + "epoch": 0.28429, + "grad_norm": 0.97684429781671, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 28429 + }, + { + "epoch": 0.2843, + "grad_norm": 1.3591061723341022, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 28430 + }, + { + "epoch": 0.28431, + "grad_norm": 0.830382765890675, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 28431 + }, + { + "epoch": 0.28432, + "grad_norm": 0.8367829375967778, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 28432 + }, + { + "epoch": 0.28433, + "grad_norm": 0.9432146613728569, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 28433 + }, + { + "epoch": 0.28434, + "grad_norm": 1.1119794952098891, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 28434 + }, + { + "epoch": 0.28435, + "grad_norm": 1.1792544449347726, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 28435 + }, + { + "epoch": 0.28436, + "grad_norm": 1.272796797670015, + "learning_rate": 0.003, + "loss": 4.083, + "step": 28436 + }, + { + "epoch": 0.28437, + "grad_norm": 0.9531957435762014, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 28437 + }, + { + "epoch": 0.28438, + "grad_norm": 0.9799134803915126, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 28438 + }, + { + "epoch": 0.28439, + "grad_norm": 1.1170918458906118, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 28439 + }, + { + "epoch": 0.2844, + "grad_norm": 1.1563370544149487, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 28440 + }, + { + "epoch": 0.28441, + "grad_norm": 1.1123675669602344, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 28441 + }, + { + "epoch": 0.28442, + "grad_norm": 0.9617219872562799, + "learning_rate": 0.003, + "loss": 4.055, + "step": 28442 + }, + { + "epoch": 0.28443, + "grad_norm": 1.0370889812096107, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 28443 + }, + { + "epoch": 0.28444, + "grad_norm": 1.2213993241761738, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 28444 + }, + { + "epoch": 0.28445, + "grad_norm": 0.9768992626600986, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 28445 + }, + { + "epoch": 0.28446, + "grad_norm": 1.0055352358723195, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 28446 + }, + { + "epoch": 0.28447, + "grad_norm": 1.063628854058651, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 28447 + }, + { + "epoch": 0.28448, + "grad_norm": 1.290086331084947, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 28448 + }, + { + "epoch": 0.28449, + "grad_norm": 0.8705481084773036, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 28449 + }, + { + "epoch": 0.2845, + "grad_norm": 0.919196652972387, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 28450 + }, + { + "epoch": 0.28451, + "grad_norm": 1.165430839147859, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 28451 + }, + { + "epoch": 0.28452, + "grad_norm": 1.2260832688045944, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 28452 + }, + { + "epoch": 0.28453, + "grad_norm": 0.8540531932273426, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 28453 + }, + { + "epoch": 0.28454, + "grad_norm": 0.9100446722194215, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 28454 + }, + { + "epoch": 0.28455, + "grad_norm": 1.3281087638479054, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 28455 + }, + { + "epoch": 0.28456, + "grad_norm": 1.1906394701698628, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 28456 + }, + { + "epoch": 0.28457, + "grad_norm": 1.0919611652600272, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 28457 + }, + { + "epoch": 0.28458, + "grad_norm": 1.01621525298455, + "learning_rate": 0.003, + "loss": 4.046, + "step": 28458 + }, + { + "epoch": 0.28459, + "grad_norm": 1.4110017714787517, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 28459 + }, + { + "epoch": 0.2846, + "grad_norm": 0.764268318138597, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 28460 + }, + { + "epoch": 0.28461, + "grad_norm": 0.8212992224139142, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 28461 + }, + { + "epoch": 0.28462, + "grad_norm": 0.8737851595805332, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 28462 + }, + { + "epoch": 0.28463, + "grad_norm": 0.9941555244781204, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 28463 + }, + { + "epoch": 0.28464, + "grad_norm": 1.0158750658274251, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 28464 + }, + { + "epoch": 0.28465, + "grad_norm": 1.2691381732386309, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 28465 + }, + { + "epoch": 0.28466, + "grad_norm": 1.2199465265932867, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 28466 + }, + { + "epoch": 0.28467, + "grad_norm": 0.8446011055846866, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 28467 + }, + { + "epoch": 0.28468, + "grad_norm": 0.8120786765540742, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 28468 + }, + { + "epoch": 0.28469, + "grad_norm": 0.9897858489855795, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 28469 + }, + { + "epoch": 0.2847, + "grad_norm": 1.207512715045942, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 28470 + }, + { + "epoch": 0.28471, + "grad_norm": 0.8850394992610031, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 28471 + }, + { + "epoch": 0.28472, + "grad_norm": 0.9784942065085347, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 28472 + }, + { + "epoch": 0.28473, + "grad_norm": 1.4224674370967625, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 28473 + }, + { + "epoch": 0.28474, + "grad_norm": 1.0797833181701468, + "learning_rate": 0.003, + "loss": 4.045, + "step": 28474 + }, + { + "epoch": 0.28475, + "grad_norm": 0.9266250482879705, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 28475 + }, + { + "epoch": 0.28476, + "grad_norm": 1.0088094713384916, + "learning_rate": 0.003, + "loss": 4.047, + "step": 28476 + }, + { + "epoch": 0.28477, + "grad_norm": 1.265037285854835, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 28477 + }, + { + "epoch": 0.28478, + "grad_norm": 0.8024935716617593, + "learning_rate": 0.003, + "loss": 4.037, + "step": 28478 + }, + { + "epoch": 0.28479, + "grad_norm": 0.8999612352230348, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 28479 + }, + { + "epoch": 0.2848, + "grad_norm": 1.2384685185329827, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 28480 + }, + { + "epoch": 0.28481, + "grad_norm": 1.040864874221417, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 28481 + }, + { + "epoch": 0.28482, + "grad_norm": 1.0335084172127929, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 28482 + }, + { + "epoch": 0.28483, + "grad_norm": 1.3240506679533706, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 28483 + }, + { + "epoch": 0.28484, + "grad_norm": 0.815825809430056, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 28484 + }, + { + "epoch": 0.28485, + "grad_norm": 0.8531617588582067, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 28485 + }, + { + "epoch": 0.28486, + "grad_norm": 0.8690304226055514, + "learning_rate": 0.003, + "loss": 4.034, + "step": 28486 + }, + { + "epoch": 0.28487, + "grad_norm": 1.088460786045591, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 28487 + }, + { + "epoch": 0.28488, + "grad_norm": 1.2301396090690822, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 28488 + }, + { + "epoch": 0.28489, + "grad_norm": 0.9479824473478253, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 28489 + }, + { + "epoch": 0.2849, + "grad_norm": 1.1753943764212715, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 28490 + }, + { + "epoch": 0.28491, + "grad_norm": 1.1004144958944715, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 28491 + }, + { + "epoch": 0.28492, + "grad_norm": 1.2248662133288857, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 28492 + }, + { + "epoch": 0.28493, + "grad_norm": 1.083297422425785, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 28493 + }, + { + "epoch": 0.28494, + "grad_norm": 1.0946965553894261, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 28494 + }, + { + "epoch": 0.28495, + "grad_norm": 0.8683511543464277, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 28495 + }, + { + "epoch": 0.28496, + "grad_norm": 0.8376513794186725, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 28496 + }, + { + "epoch": 0.28497, + "grad_norm": 0.7959292534350491, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 28497 + }, + { + "epoch": 0.28498, + "grad_norm": 0.8446152937723328, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 28498 + }, + { + "epoch": 0.28499, + "grad_norm": 1.0463269134115267, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 28499 + }, + { + "epoch": 0.285, + "grad_norm": 1.0810713989311518, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 28500 + }, + { + "epoch": 0.28501, + "grad_norm": 0.9282065678070713, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 28501 + }, + { + "epoch": 0.28502, + "grad_norm": 1.0262790289955428, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 28502 + }, + { + "epoch": 0.28503, + "grad_norm": 1.1636909371408948, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 28503 + }, + { + "epoch": 0.28504, + "grad_norm": 1.186447312403609, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 28504 + }, + { + "epoch": 0.28505, + "grad_norm": 1.0731570232563765, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 28505 + }, + { + "epoch": 0.28506, + "grad_norm": 1.4190798524479764, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 28506 + }, + { + "epoch": 0.28507, + "grad_norm": 1.0200856624542387, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 28507 + }, + { + "epoch": 0.28508, + "grad_norm": 1.0836258569393682, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 28508 + }, + { + "epoch": 0.28509, + "grad_norm": 0.9834920692500091, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 28509 + }, + { + "epoch": 0.2851, + "grad_norm": 1.1103723636519116, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 28510 + }, + { + "epoch": 0.28511, + "grad_norm": 1.1798407963323452, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 28511 + }, + { + "epoch": 0.28512, + "grad_norm": 1.0870077246213081, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 28512 + }, + { + "epoch": 0.28513, + "grad_norm": 0.9235922592753771, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 28513 + }, + { + "epoch": 0.28514, + "grad_norm": 1.023069742321689, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 28514 + }, + { + "epoch": 0.28515, + "grad_norm": 1.019877608258244, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 28515 + }, + { + "epoch": 0.28516, + "grad_norm": 0.9792746591753052, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 28516 + }, + { + "epoch": 0.28517, + "grad_norm": 1.2069600384169086, + "learning_rate": 0.003, + "loss": 4.0898, + "step": 28517 + }, + { + "epoch": 0.28518, + "grad_norm": 1.0520792016473446, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 28518 + }, + { + "epoch": 0.28519, + "grad_norm": 1.1974580350712776, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 28519 + }, + { + "epoch": 0.2852, + "grad_norm": 0.8839680225579309, + "learning_rate": 0.003, + "loss": 4.066, + "step": 28520 + }, + { + "epoch": 0.28521, + "grad_norm": 1.1665294437713425, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 28521 + }, + { + "epoch": 0.28522, + "grad_norm": 1.2413348644691453, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 28522 + }, + { + "epoch": 0.28523, + "grad_norm": 0.909132940845315, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 28523 + }, + { + "epoch": 0.28524, + "grad_norm": 0.9788890356402395, + "learning_rate": 0.003, + "loss": 4.065, + "step": 28524 + }, + { + "epoch": 0.28525, + "grad_norm": 1.324522784438691, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 28525 + }, + { + "epoch": 0.28526, + "grad_norm": 0.970729113888255, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 28526 + }, + { + "epoch": 0.28527, + "grad_norm": 1.0935801133955896, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 28527 + }, + { + "epoch": 0.28528, + "grad_norm": 1.1893590452011937, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 28528 + }, + { + "epoch": 0.28529, + "grad_norm": 1.1769611200932573, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 28529 + }, + { + "epoch": 0.2853, + "grad_norm": 0.8794748982996977, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 28530 + }, + { + "epoch": 0.28531, + "grad_norm": 0.8859261702307876, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 28531 + }, + { + "epoch": 0.28532, + "grad_norm": 1.1537085160905174, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 28532 + }, + { + "epoch": 0.28533, + "grad_norm": 1.304759850882072, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 28533 + }, + { + "epoch": 0.28534, + "grad_norm": 1.0487001400652816, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 28534 + }, + { + "epoch": 0.28535, + "grad_norm": 1.1464259966873718, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 28535 + }, + { + "epoch": 0.28536, + "grad_norm": 1.0909639706593734, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 28536 + }, + { + "epoch": 0.28537, + "grad_norm": 0.9983941128613821, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 28537 + }, + { + "epoch": 0.28538, + "grad_norm": 0.9950904993055258, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 28538 + }, + { + "epoch": 0.28539, + "grad_norm": 0.9740346410003009, + "learning_rate": 0.003, + "loss": 4.024, + "step": 28539 + }, + { + "epoch": 0.2854, + "grad_norm": 0.9231776833121283, + "learning_rate": 0.003, + "loss": 4.049, + "step": 28540 + }, + { + "epoch": 0.28541, + "grad_norm": 0.9613173522763104, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 28541 + }, + { + "epoch": 0.28542, + "grad_norm": 1.0612976059214487, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 28542 + }, + { + "epoch": 0.28543, + "grad_norm": 1.169618451281373, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 28543 + }, + { + "epoch": 0.28544, + "grad_norm": 1.0745304849179804, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 28544 + }, + { + "epoch": 0.28545, + "grad_norm": 1.3438156974221394, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 28545 + }, + { + "epoch": 0.28546, + "grad_norm": 0.7168926985429324, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 28546 + }, + { + "epoch": 0.28547, + "grad_norm": 0.7729374409792619, + "learning_rate": 0.003, + "loss": 4.057, + "step": 28547 + }, + { + "epoch": 0.28548, + "grad_norm": 0.942433340695535, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 28548 + }, + { + "epoch": 0.28549, + "grad_norm": 1.2963787505779338, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 28549 + }, + { + "epoch": 0.2855, + "grad_norm": 1.052068043087559, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 28550 + }, + { + "epoch": 0.28551, + "grad_norm": 1.200671587944267, + "learning_rate": 0.003, + "loss": 4.053, + "step": 28551 + }, + { + "epoch": 0.28552, + "grad_norm": 1.0451978980901175, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 28552 + }, + { + "epoch": 0.28553, + "grad_norm": 0.993232864680296, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 28553 + }, + { + "epoch": 0.28554, + "grad_norm": 1.0904605855956855, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 28554 + }, + { + "epoch": 0.28555, + "grad_norm": 0.9583856862167444, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 28555 + }, + { + "epoch": 0.28556, + "grad_norm": 1.0903504171481644, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 28556 + }, + { + "epoch": 0.28557, + "grad_norm": 1.0542519191838697, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 28557 + }, + { + "epoch": 0.28558, + "grad_norm": 1.012434848334973, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 28558 + }, + { + "epoch": 0.28559, + "grad_norm": 1.0471037054072, + "learning_rate": 0.003, + "loss": 4.083, + "step": 28559 + }, + { + "epoch": 0.2856, + "grad_norm": 1.3307574835601776, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 28560 + }, + { + "epoch": 0.28561, + "grad_norm": 1.0801404419111824, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 28561 + }, + { + "epoch": 0.28562, + "grad_norm": 1.1726621429081907, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 28562 + }, + { + "epoch": 0.28563, + "grad_norm": 1.1144955346727583, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 28563 + }, + { + "epoch": 0.28564, + "grad_norm": 1.1251240012957267, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 28564 + }, + { + "epoch": 0.28565, + "grad_norm": 1.0373566271294208, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 28565 + }, + { + "epoch": 0.28566, + "grad_norm": 1.2337347685084645, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 28566 + }, + { + "epoch": 0.28567, + "grad_norm": 0.9391948966276945, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 28567 + }, + { + "epoch": 0.28568, + "grad_norm": 0.9134895851452427, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 28568 + }, + { + "epoch": 0.28569, + "grad_norm": 1.1554050168142358, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 28569 + }, + { + "epoch": 0.2857, + "grad_norm": 1.2055599772146823, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 28570 + }, + { + "epoch": 0.28571, + "grad_norm": 1.102175164592626, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 28571 + }, + { + "epoch": 0.28572, + "grad_norm": 1.1689992567052934, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 28572 + }, + { + "epoch": 0.28573, + "grad_norm": 1.002997825056329, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 28573 + }, + { + "epoch": 0.28574, + "grad_norm": 0.9979366358750109, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 28574 + }, + { + "epoch": 0.28575, + "grad_norm": 1.0914431274736822, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 28575 + }, + { + "epoch": 0.28576, + "grad_norm": 1.076323899538689, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 28576 + }, + { + "epoch": 0.28577, + "grad_norm": 1.196146167859645, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 28577 + }, + { + "epoch": 0.28578, + "grad_norm": 1.0084795168698857, + "learning_rate": 0.003, + "loss": 4.033, + "step": 28578 + }, + { + "epoch": 0.28579, + "grad_norm": 1.1426074988864712, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 28579 + }, + { + "epoch": 0.2858, + "grad_norm": 0.9186800693794664, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 28580 + }, + { + "epoch": 0.28581, + "grad_norm": 1.006766587820355, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 28581 + }, + { + "epoch": 0.28582, + "grad_norm": 1.2007124855265583, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 28582 + }, + { + "epoch": 0.28583, + "grad_norm": 1.094962241920677, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 28583 + }, + { + "epoch": 0.28584, + "grad_norm": 1.1930276413047665, + "learning_rate": 0.003, + "loss": 4.012, + "step": 28584 + }, + { + "epoch": 0.28585, + "grad_norm": 0.937079639647783, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 28585 + }, + { + "epoch": 0.28586, + "grad_norm": 0.8501892528231876, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 28586 + }, + { + "epoch": 0.28587, + "grad_norm": 0.9117337007887149, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 28587 + }, + { + "epoch": 0.28588, + "grad_norm": 1.0386135144853568, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 28588 + }, + { + "epoch": 0.28589, + "grad_norm": 1.0715499104942463, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 28589 + }, + { + "epoch": 0.2859, + "grad_norm": 1.0661979780569892, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 28590 + }, + { + "epoch": 0.28591, + "grad_norm": 1.2075998687072131, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 28591 + }, + { + "epoch": 0.28592, + "grad_norm": 0.9613126045981987, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 28592 + }, + { + "epoch": 0.28593, + "grad_norm": 1.2229391207737994, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 28593 + }, + { + "epoch": 0.28594, + "grad_norm": 0.9834844123426557, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 28594 + }, + { + "epoch": 0.28595, + "grad_norm": 0.8924932799881957, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 28595 + }, + { + "epoch": 0.28596, + "grad_norm": 0.9335932915752639, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 28596 + }, + { + "epoch": 0.28597, + "grad_norm": 1.3163570321216085, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 28597 + }, + { + "epoch": 0.28598, + "grad_norm": 0.955868896749576, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 28598 + }, + { + "epoch": 0.28599, + "grad_norm": 0.9411674927492346, + "learning_rate": 0.003, + "loss": 4.02, + "step": 28599 + }, + { + "epoch": 0.286, + "grad_norm": 1.018360384203518, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 28600 + }, + { + "epoch": 0.28601, + "grad_norm": 1.214515076157159, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 28601 + }, + { + "epoch": 0.28602, + "grad_norm": 1.0293926808557305, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 28602 + }, + { + "epoch": 0.28603, + "grad_norm": 1.2849534928951338, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 28603 + }, + { + "epoch": 0.28604, + "grad_norm": 1.0855331499460903, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 28604 + }, + { + "epoch": 0.28605, + "grad_norm": 1.0260836830898763, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 28605 + }, + { + "epoch": 0.28606, + "grad_norm": 1.018652730774459, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 28606 + }, + { + "epoch": 0.28607, + "grad_norm": 1.107746554707529, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 28607 + }, + { + "epoch": 0.28608, + "grad_norm": 1.0792097936637284, + "learning_rate": 0.003, + "loss": 4.051, + "step": 28608 + }, + { + "epoch": 0.28609, + "grad_norm": 0.8394855813123078, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 28609 + }, + { + "epoch": 0.2861, + "grad_norm": 0.870150809743468, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 28610 + }, + { + "epoch": 0.28611, + "grad_norm": 0.9502141660689155, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 28611 + }, + { + "epoch": 0.28612, + "grad_norm": 0.9314550108793223, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 28612 + }, + { + "epoch": 0.28613, + "grad_norm": 1.2291873991119722, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 28613 + }, + { + "epoch": 0.28614, + "grad_norm": 0.9980985743597343, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 28614 + }, + { + "epoch": 0.28615, + "grad_norm": 1.525183654433964, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 28615 + }, + { + "epoch": 0.28616, + "grad_norm": 1.0979544093581235, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 28616 + }, + { + "epoch": 0.28617, + "grad_norm": 0.9823291137878902, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 28617 + }, + { + "epoch": 0.28618, + "grad_norm": 1.0390008324558777, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 28618 + }, + { + "epoch": 0.28619, + "grad_norm": 1.2494921774735555, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 28619 + }, + { + "epoch": 0.2862, + "grad_norm": 1.1594670402171612, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 28620 + }, + { + "epoch": 0.28621, + "grad_norm": 1.040147501766103, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 28621 + }, + { + "epoch": 0.28622, + "grad_norm": 1.2726937443599118, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 28622 + }, + { + "epoch": 0.28623, + "grad_norm": 0.9018077597443906, + "learning_rate": 0.003, + "loss": 4.029, + "step": 28623 + }, + { + "epoch": 0.28624, + "grad_norm": 0.803891761855308, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 28624 + }, + { + "epoch": 0.28625, + "grad_norm": 0.8215555161330542, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 28625 + }, + { + "epoch": 0.28626, + "grad_norm": 0.735129214133083, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 28626 + }, + { + "epoch": 0.28627, + "grad_norm": 0.6486156918634486, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 28627 + }, + { + "epoch": 0.28628, + "grad_norm": 0.8619736065035976, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 28628 + }, + { + "epoch": 0.28629, + "grad_norm": 1.2809795266289008, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 28629 + }, + { + "epoch": 0.2863, + "grad_norm": 1.0455599467507077, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 28630 + }, + { + "epoch": 0.28631, + "grad_norm": 0.9816252760133752, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 28631 + }, + { + "epoch": 0.28632, + "grad_norm": 1.1069288894308122, + "learning_rate": 0.003, + "loss": 4.029, + "step": 28632 + }, + { + "epoch": 0.28633, + "grad_norm": 1.156535389836581, + "learning_rate": 0.003, + "loss": 4.058, + "step": 28633 + }, + { + "epoch": 0.28634, + "grad_norm": 1.0442196175866156, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 28634 + }, + { + "epoch": 0.28635, + "grad_norm": 1.1438936086434273, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 28635 + }, + { + "epoch": 0.28636, + "grad_norm": 0.8858075274841799, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 28636 + }, + { + "epoch": 0.28637, + "grad_norm": 0.9474840863831209, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 28637 + }, + { + "epoch": 0.28638, + "grad_norm": 1.249520515428568, + "learning_rate": 0.003, + "loss": 4.0923, + "step": 28638 + }, + { + "epoch": 0.28639, + "grad_norm": 1.0309583413066006, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 28639 + }, + { + "epoch": 0.2864, + "grad_norm": 1.5350033693603498, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 28640 + }, + { + "epoch": 0.28641, + "grad_norm": 0.7988564748906887, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 28641 + }, + { + "epoch": 0.28642, + "grad_norm": 0.8919864841052212, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 28642 + }, + { + "epoch": 0.28643, + "grad_norm": 0.9067611589166907, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 28643 + }, + { + "epoch": 0.28644, + "grad_norm": 1.0866019560067706, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 28644 + }, + { + "epoch": 0.28645, + "grad_norm": 1.4425695241448897, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 28645 + }, + { + "epoch": 0.28646, + "grad_norm": 0.9193942779260135, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 28646 + }, + { + "epoch": 0.28647, + "grad_norm": 1.013527683075799, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 28647 + }, + { + "epoch": 0.28648, + "grad_norm": 1.3248696605831194, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 28648 + }, + { + "epoch": 0.28649, + "grad_norm": 1.0316156715855336, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 28649 + }, + { + "epoch": 0.2865, + "grad_norm": 1.116304940822961, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 28650 + }, + { + "epoch": 0.28651, + "grad_norm": 0.9617917335782055, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 28651 + }, + { + "epoch": 0.28652, + "grad_norm": 1.2600467432419735, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 28652 + }, + { + "epoch": 0.28653, + "grad_norm": 1.2037602080348104, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 28653 + }, + { + "epoch": 0.28654, + "grad_norm": 1.0802013562768116, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 28654 + }, + { + "epoch": 0.28655, + "grad_norm": 1.057957860082967, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 28655 + }, + { + "epoch": 0.28656, + "grad_norm": 1.0344053091334657, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 28656 + }, + { + "epoch": 0.28657, + "grad_norm": 1.140337240175989, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 28657 + }, + { + "epoch": 0.28658, + "grad_norm": 1.1849133606351483, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 28658 + }, + { + "epoch": 0.28659, + "grad_norm": 0.9976340380894032, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 28659 + }, + { + "epoch": 0.2866, + "grad_norm": 1.1006634091978138, + "learning_rate": 0.003, + "loss": 3.991, + "step": 28660 + }, + { + "epoch": 0.28661, + "grad_norm": 1.0271414077925782, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 28661 + }, + { + "epoch": 0.28662, + "grad_norm": 1.0144676431496091, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 28662 + }, + { + "epoch": 0.28663, + "grad_norm": 1.0250514131656152, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 28663 + }, + { + "epoch": 0.28664, + "grad_norm": 1.1636415429483384, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 28664 + }, + { + "epoch": 0.28665, + "grad_norm": 0.8981371695916635, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 28665 + }, + { + "epoch": 0.28666, + "grad_norm": 1.0089486801315857, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 28666 + }, + { + "epoch": 0.28667, + "grad_norm": 1.1712182535467741, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 28667 + }, + { + "epoch": 0.28668, + "grad_norm": 0.9373945512933428, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 28668 + }, + { + "epoch": 0.28669, + "grad_norm": 0.9494320265834704, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 28669 + }, + { + "epoch": 0.2867, + "grad_norm": 0.9860254282616898, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 28670 + }, + { + "epoch": 0.28671, + "grad_norm": 1.1917740117356919, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 28671 + }, + { + "epoch": 0.28672, + "grad_norm": 1.0889710378138397, + "learning_rate": 0.003, + "loss": 4.039, + "step": 28672 + }, + { + "epoch": 0.28673, + "grad_norm": 0.9881089933359626, + "learning_rate": 0.003, + "loss": 4.05, + "step": 28673 + }, + { + "epoch": 0.28674, + "grad_norm": 0.9880729147222439, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 28674 + }, + { + "epoch": 0.28675, + "grad_norm": 1.2372021576211996, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 28675 + }, + { + "epoch": 0.28676, + "grad_norm": 1.2490691971570622, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 28676 + }, + { + "epoch": 0.28677, + "grad_norm": 0.8331160120917905, + "learning_rate": 0.003, + "loss": 4.068, + "step": 28677 + }, + { + "epoch": 0.28678, + "grad_norm": 0.7636919127384839, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 28678 + }, + { + "epoch": 0.28679, + "grad_norm": 0.8514403961479843, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 28679 + }, + { + "epoch": 0.2868, + "grad_norm": 1.01214493574034, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 28680 + }, + { + "epoch": 0.28681, + "grad_norm": 1.2654626166156002, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 28681 + }, + { + "epoch": 0.28682, + "grad_norm": 1.2614738948918434, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 28682 + }, + { + "epoch": 0.28683, + "grad_norm": 1.0471046415768497, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 28683 + }, + { + "epoch": 0.28684, + "grad_norm": 1.2802217010416748, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 28684 + }, + { + "epoch": 0.28685, + "grad_norm": 1.0316983017330854, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 28685 + }, + { + "epoch": 0.28686, + "grad_norm": 1.0612471274409943, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 28686 + }, + { + "epoch": 0.28687, + "grad_norm": 1.1904390708633985, + "learning_rate": 0.003, + "loss": 4.045, + "step": 28687 + }, + { + "epoch": 0.28688, + "grad_norm": 0.935949457683684, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 28688 + }, + { + "epoch": 0.28689, + "grad_norm": 1.3762615069874657, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 28689 + }, + { + "epoch": 0.2869, + "grad_norm": 0.9758207607895223, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 28690 + }, + { + "epoch": 0.28691, + "grad_norm": 0.92946836528148, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 28691 + }, + { + "epoch": 0.28692, + "grad_norm": 0.9109924343085799, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 28692 + }, + { + "epoch": 0.28693, + "grad_norm": 1.1050762245904573, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 28693 + }, + { + "epoch": 0.28694, + "grad_norm": 1.0458959940849286, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 28694 + }, + { + "epoch": 0.28695, + "grad_norm": 1.1454706574211686, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 28695 + }, + { + "epoch": 0.28696, + "grad_norm": 0.975268445146075, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 28696 + }, + { + "epoch": 0.28697, + "grad_norm": 1.120066659507972, + "learning_rate": 0.003, + "loss": 4.033, + "step": 28697 + }, + { + "epoch": 0.28698, + "grad_norm": 0.9734750191155456, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 28698 + }, + { + "epoch": 0.28699, + "grad_norm": 0.9726089618362278, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 28699 + }, + { + "epoch": 0.287, + "grad_norm": 1.0711441577239957, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 28700 + }, + { + "epoch": 0.28701, + "grad_norm": 1.089986394635112, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 28701 + }, + { + "epoch": 0.28702, + "grad_norm": 0.9910230071547013, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 28702 + }, + { + "epoch": 0.28703, + "grad_norm": 1.0345274952423886, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 28703 + }, + { + "epoch": 0.28704, + "grad_norm": 1.0009640562644422, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 28704 + }, + { + "epoch": 0.28705, + "grad_norm": 1.0047605425011081, + "learning_rate": 0.003, + "loss": 4.054, + "step": 28705 + }, + { + "epoch": 0.28706, + "grad_norm": 1.2205693957571804, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 28706 + }, + { + "epoch": 0.28707, + "grad_norm": 1.1079253255172383, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 28707 + }, + { + "epoch": 0.28708, + "grad_norm": 1.30523629019359, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 28708 + }, + { + "epoch": 0.28709, + "grad_norm": 0.8199860164050703, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 28709 + }, + { + "epoch": 0.2871, + "grad_norm": 0.8416722303885555, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 28710 + }, + { + "epoch": 0.28711, + "grad_norm": 0.9502144828364654, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 28711 + }, + { + "epoch": 0.28712, + "grad_norm": 1.0533501850218652, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 28712 + }, + { + "epoch": 0.28713, + "grad_norm": 1.1980119200908914, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 28713 + }, + { + "epoch": 0.28714, + "grad_norm": 1.0900523751691524, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 28714 + }, + { + "epoch": 0.28715, + "grad_norm": 0.8898064160747344, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 28715 + }, + { + "epoch": 0.28716, + "grad_norm": 0.941289942709176, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 28716 + }, + { + "epoch": 0.28717, + "grad_norm": 1.080667767123961, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 28717 + }, + { + "epoch": 0.28718, + "grad_norm": 1.1076796087825747, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 28718 + }, + { + "epoch": 0.28719, + "grad_norm": 1.121811181781137, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 28719 + }, + { + "epoch": 0.2872, + "grad_norm": 1.1255048943219945, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 28720 + }, + { + "epoch": 0.28721, + "grad_norm": 1.1963215784916668, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 28721 + }, + { + "epoch": 0.28722, + "grad_norm": 1.0715297925510523, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 28722 + }, + { + "epoch": 0.28723, + "grad_norm": 1.1029846725676413, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 28723 + }, + { + "epoch": 0.28724, + "grad_norm": 1.2253474476672126, + "learning_rate": 0.003, + "loss": 4.04, + "step": 28724 + }, + { + "epoch": 0.28725, + "grad_norm": 0.8390915002716094, + "learning_rate": 0.003, + "loss": 4.026, + "step": 28725 + }, + { + "epoch": 0.28726, + "grad_norm": 0.752038942431805, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 28726 + }, + { + "epoch": 0.28727, + "grad_norm": 0.7489227501241332, + "learning_rate": 0.003, + "loss": 4.028, + "step": 28727 + }, + { + "epoch": 0.28728, + "grad_norm": 0.9703748762581752, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 28728 + }, + { + "epoch": 0.28729, + "grad_norm": 1.3174959840404894, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 28729 + }, + { + "epoch": 0.2873, + "grad_norm": 0.9886746801190063, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 28730 + }, + { + "epoch": 0.28731, + "grad_norm": 1.5247390587407592, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 28731 + }, + { + "epoch": 0.28732, + "grad_norm": 0.8592797256766662, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 28732 + }, + { + "epoch": 0.28733, + "grad_norm": 0.9172831771055792, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 28733 + }, + { + "epoch": 0.28734, + "grad_norm": 1.3124554342529393, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 28734 + }, + { + "epoch": 0.28735, + "grad_norm": 0.9974330653765844, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 28735 + }, + { + "epoch": 0.28736, + "grad_norm": 1.1416388716377999, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 28736 + }, + { + "epoch": 0.28737, + "grad_norm": 1.139008938070121, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 28737 + }, + { + "epoch": 0.28738, + "grad_norm": 1.0960854366836517, + "learning_rate": 0.003, + "loss": 4.045, + "step": 28738 + }, + { + "epoch": 0.28739, + "grad_norm": 0.9445952192377916, + "learning_rate": 0.003, + "loss": 4.062, + "step": 28739 + }, + { + "epoch": 0.2874, + "grad_norm": 0.9933035153994677, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 28740 + }, + { + "epoch": 0.28741, + "grad_norm": 1.2045160141093971, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 28741 + }, + { + "epoch": 0.28742, + "grad_norm": 1.0564096675078511, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 28742 + }, + { + "epoch": 0.28743, + "grad_norm": 1.1882277681983395, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 28743 + }, + { + "epoch": 0.28744, + "grad_norm": 1.060877818903612, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 28744 + }, + { + "epoch": 0.28745, + "grad_norm": 1.043313186693969, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 28745 + }, + { + "epoch": 0.28746, + "grad_norm": 1.210354473833106, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 28746 + }, + { + "epoch": 0.28747, + "grad_norm": 0.776668624651301, + "learning_rate": 0.003, + "loss": 4.033, + "step": 28747 + }, + { + "epoch": 0.28748, + "grad_norm": 0.8862244888938946, + "learning_rate": 0.003, + "loss": 4.03, + "step": 28748 + }, + { + "epoch": 0.28749, + "grad_norm": 1.1627801828896331, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 28749 + }, + { + "epoch": 0.2875, + "grad_norm": 1.2636496355325637, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 28750 + }, + { + "epoch": 0.28751, + "grad_norm": 1.0557137942722548, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 28751 + }, + { + "epoch": 0.28752, + "grad_norm": 1.1602437102065406, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 28752 + }, + { + "epoch": 0.28753, + "grad_norm": 1.1214469678199133, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 28753 + }, + { + "epoch": 0.28754, + "grad_norm": 1.0920558771132576, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 28754 + }, + { + "epoch": 0.28755, + "grad_norm": 1.1135087477789767, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 28755 + }, + { + "epoch": 0.28756, + "grad_norm": 1.2081615519437918, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 28756 + }, + { + "epoch": 0.28757, + "grad_norm": 0.9446121138534607, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 28757 + }, + { + "epoch": 0.28758, + "grad_norm": 1.1256198874381664, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 28758 + }, + { + "epoch": 0.28759, + "grad_norm": 1.2039006177637857, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 28759 + }, + { + "epoch": 0.2876, + "grad_norm": 0.87776025562572, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 28760 + }, + { + "epoch": 0.28761, + "grad_norm": 0.8393200744046563, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 28761 + }, + { + "epoch": 0.28762, + "grad_norm": 1.0044486491248361, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 28762 + }, + { + "epoch": 0.28763, + "grad_norm": 1.0733054816069842, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 28763 + }, + { + "epoch": 0.28764, + "grad_norm": 1.1396504587950058, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 28764 + }, + { + "epoch": 0.28765, + "grad_norm": 1.0621075232621913, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 28765 + }, + { + "epoch": 0.28766, + "grad_norm": 1.2695494177386384, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 28766 + }, + { + "epoch": 0.28767, + "grad_norm": 1.073737214515063, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 28767 + }, + { + "epoch": 0.28768, + "grad_norm": 1.1305781879762706, + "learning_rate": 0.003, + "loss": 4.067, + "step": 28768 + }, + { + "epoch": 0.28769, + "grad_norm": 0.9943162738431134, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 28769 + }, + { + "epoch": 0.2877, + "grad_norm": 1.147746858695975, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 28770 + }, + { + "epoch": 0.28771, + "grad_norm": 1.1080317520296294, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 28771 + }, + { + "epoch": 0.28772, + "grad_norm": 1.0614157050409685, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 28772 + }, + { + "epoch": 0.28773, + "grad_norm": 1.3026116570286683, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 28773 + }, + { + "epoch": 0.28774, + "grad_norm": 1.0051378337669643, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 28774 + }, + { + "epoch": 0.28775, + "grad_norm": 1.105122770464552, + "learning_rate": 0.003, + "loss": 4.044, + "step": 28775 + }, + { + "epoch": 0.28776, + "grad_norm": 0.993989037512551, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 28776 + }, + { + "epoch": 0.28777, + "grad_norm": 1.3659091093968478, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 28777 + }, + { + "epoch": 0.28778, + "grad_norm": 1.210803359903117, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 28778 + }, + { + "epoch": 0.28779, + "grad_norm": 1.0661658440513295, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 28779 + }, + { + "epoch": 0.2878, + "grad_norm": 0.9577265526060561, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 28780 + }, + { + "epoch": 0.28781, + "grad_norm": 1.048654704450843, + "learning_rate": 0.003, + "loss": 4.044, + "step": 28781 + }, + { + "epoch": 0.28782, + "grad_norm": 1.202196987272661, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 28782 + }, + { + "epoch": 0.28783, + "grad_norm": 1.0906276980552847, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 28783 + }, + { + "epoch": 0.28784, + "grad_norm": 1.109681917360073, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 28784 + }, + { + "epoch": 0.28785, + "grad_norm": 1.1745173217483014, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 28785 + }, + { + "epoch": 0.28786, + "grad_norm": 1.1661512810142847, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 28786 + }, + { + "epoch": 0.28787, + "grad_norm": 1.3400720327910058, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 28787 + }, + { + "epoch": 0.28788, + "grad_norm": 0.952191158378184, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 28788 + }, + { + "epoch": 0.28789, + "grad_norm": 0.9040099075268389, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 28789 + }, + { + "epoch": 0.2879, + "grad_norm": 0.8189104323229701, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 28790 + }, + { + "epoch": 0.28791, + "grad_norm": 0.7516065582367091, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 28791 + }, + { + "epoch": 0.28792, + "grad_norm": 0.7652721791420264, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 28792 + }, + { + "epoch": 0.28793, + "grad_norm": 0.8264815231445409, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 28793 + }, + { + "epoch": 0.28794, + "grad_norm": 1.0008099922930465, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 28794 + }, + { + "epoch": 0.28795, + "grad_norm": 1.3689928609880384, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 28795 + }, + { + "epoch": 0.28796, + "grad_norm": 0.8343942549929909, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 28796 + }, + { + "epoch": 0.28797, + "grad_norm": 0.9115921398465102, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 28797 + }, + { + "epoch": 0.28798, + "grad_norm": 0.9391094673082053, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 28798 + }, + { + "epoch": 0.28799, + "grad_norm": 1.1195131422271238, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 28799 + }, + { + "epoch": 0.288, + "grad_norm": 1.2507419185463349, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 28800 + }, + { + "epoch": 0.28801, + "grad_norm": 1.1390215742771959, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 28801 + }, + { + "epoch": 0.28802, + "grad_norm": 1.0133009366222079, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 28802 + }, + { + "epoch": 0.28803, + "grad_norm": 1.2513602722639265, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 28803 + }, + { + "epoch": 0.28804, + "grad_norm": 0.9485317649097389, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 28804 + }, + { + "epoch": 0.28805, + "grad_norm": 0.865577633482965, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 28805 + }, + { + "epoch": 0.28806, + "grad_norm": 0.9564743755039128, + "learning_rate": 0.003, + "loss": 4.037, + "step": 28806 + }, + { + "epoch": 0.28807, + "grad_norm": 1.1845444873972435, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 28807 + }, + { + "epoch": 0.28808, + "grad_norm": 0.8588954717165209, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 28808 + }, + { + "epoch": 0.28809, + "grad_norm": 1.0345409193022637, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 28809 + }, + { + "epoch": 0.2881, + "grad_norm": 1.1213631868204743, + "learning_rate": 0.003, + "loss": 4.041, + "step": 28810 + }, + { + "epoch": 0.28811, + "grad_norm": 1.344864914729251, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 28811 + }, + { + "epoch": 0.28812, + "grad_norm": 0.8551941927844171, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 28812 + }, + { + "epoch": 0.28813, + "grad_norm": 1.0415723155933794, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 28813 + }, + { + "epoch": 0.28814, + "grad_norm": 1.0750580548424997, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 28814 + }, + { + "epoch": 0.28815, + "grad_norm": 1.2626640690577544, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 28815 + }, + { + "epoch": 0.28816, + "grad_norm": 0.9491921987359638, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 28816 + }, + { + "epoch": 0.28817, + "grad_norm": 1.1478945387391888, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 28817 + }, + { + "epoch": 0.28818, + "grad_norm": 1.1265277232233437, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 28818 + }, + { + "epoch": 0.28819, + "grad_norm": 0.9890918359475447, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 28819 + }, + { + "epoch": 0.2882, + "grad_norm": 0.9841143204781324, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 28820 + }, + { + "epoch": 0.28821, + "grad_norm": 1.0955814668872996, + "learning_rate": 0.003, + "loss": 4.035, + "step": 28821 + }, + { + "epoch": 0.28822, + "grad_norm": 0.8680235639852562, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 28822 + }, + { + "epoch": 0.28823, + "grad_norm": 0.9011932140360052, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 28823 + }, + { + "epoch": 0.28824, + "grad_norm": 1.138161095352757, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 28824 + }, + { + "epoch": 0.28825, + "grad_norm": 1.2756568890801292, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 28825 + }, + { + "epoch": 0.28826, + "grad_norm": 1.2206252220328846, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 28826 + }, + { + "epoch": 0.28827, + "grad_norm": 1.1095579145608883, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 28827 + }, + { + "epoch": 0.28828, + "grad_norm": 1.0569428740448112, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 28828 + }, + { + "epoch": 0.28829, + "grad_norm": 1.2154104012285218, + "learning_rate": 0.003, + "loss": 4.05, + "step": 28829 + }, + { + "epoch": 0.2883, + "grad_norm": 0.9844288550431994, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 28830 + }, + { + "epoch": 0.28831, + "grad_norm": 1.4548796151697734, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 28831 + }, + { + "epoch": 0.28832, + "grad_norm": 0.780851759453261, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 28832 + }, + { + "epoch": 0.28833, + "grad_norm": 0.7213837515423824, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 28833 + }, + { + "epoch": 0.28834, + "grad_norm": 0.9223697326046878, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 28834 + }, + { + "epoch": 0.28835, + "grad_norm": 1.1431686378793675, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 28835 + }, + { + "epoch": 0.28836, + "grad_norm": 1.100959384190114, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 28836 + }, + { + "epoch": 0.28837, + "grad_norm": 1.0235133663086382, + "learning_rate": 0.003, + "loss": 4.033, + "step": 28837 + }, + { + "epoch": 0.28838, + "grad_norm": 1.0791081745078464, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 28838 + }, + { + "epoch": 0.28839, + "grad_norm": 1.1999501068288718, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 28839 + }, + { + "epoch": 0.2884, + "grad_norm": 1.3492069531632715, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 28840 + }, + { + "epoch": 0.28841, + "grad_norm": 0.9086805473103813, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 28841 + }, + { + "epoch": 0.28842, + "grad_norm": 0.9066079700463744, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 28842 + }, + { + "epoch": 0.28843, + "grad_norm": 1.0936281442518438, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 28843 + }, + { + "epoch": 0.28844, + "grad_norm": 1.2942429559447501, + "learning_rate": 0.003, + "loss": 4.047, + "step": 28844 + }, + { + "epoch": 0.28845, + "grad_norm": 1.1223138872679177, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 28845 + }, + { + "epoch": 0.28846, + "grad_norm": 1.2341291534183934, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 28846 + }, + { + "epoch": 0.28847, + "grad_norm": 1.0072645846145039, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 28847 + }, + { + "epoch": 0.28848, + "grad_norm": 1.1193356979449591, + "learning_rate": 0.003, + "loss": 4.012, + "step": 28848 + }, + { + "epoch": 0.28849, + "grad_norm": 1.1680415625965592, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 28849 + }, + { + "epoch": 0.2885, + "grad_norm": 1.0198655064899194, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 28850 + }, + { + "epoch": 0.28851, + "grad_norm": 1.1919186278979843, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 28851 + }, + { + "epoch": 0.28852, + "grad_norm": 1.1382608846601345, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 28852 + }, + { + "epoch": 0.28853, + "grad_norm": 0.9205315643630243, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 28853 + }, + { + "epoch": 0.28854, + "grad_norm": 1.0851404069850823, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 28854 + }, + { + "epoch": 0.28855, + "grad_norm": 1.1197073886900164, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 28855 + }, + { + "epoch": 0.28856, + "grad_norm": 0.9986907832744208, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 28856 + }, + { + "epoch": 0.28857, + "grad_norm": 0.9796568686597364, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 28857 + }, + { + "epoch": 0.28858, + "grad_norm": 1.0862907975040637, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 28858 + }, + { + "epoch": 0.28859, + "grad_norm": 1.19342086130678, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 28859 + }, + { + "epoch": 0.2886, + "grad_norm": 0.9806669490342005, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 28860 + }, + { + "epoch": 0.28861, + "grad_norm": 0.925775413768787, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 28861 + }, + { + "epoch": 0.28862, + "grad_norm": 1.1914600145702094, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 28862 + }, + { + "epoch": 0.28863, + "grad_norm": 1.3099321660411192, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 28863 + }, + { + "epoch": 0.28864, + "grad_norm": 0.9334236910020909, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 28864 + }, + { + "epoch": 0.28865, + "grad_norm": 1.1057489114720696, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 28865 + }, + { + "epoch": 0.28866, + "grad_norm": 1.2962629638635286, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 28866 + }, + { + "epoch": 0.28867, + "grad_norm": 1.0143070936485283, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 28867 + }, + { + "epoch": 0.28868, + "grad_norm": 1.113846900016663, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 28868 + }, + { + "epoch": 0.28869, + "grad_norm": 1.022734076841889, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 28869 + }, + { + "epoch": 0.2887, + "grad_norm": 1.1573276933536873, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 28870 + }, + { + "epoch": 0.28871, + "grad_norm": 0.9605407100347976, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 28871 + }, + { + "epoch": 0.28872, + "grad_norm": 0.916563676236475, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 28872 + }, + { + "epoch": 0.28873, + "grad_norm": 0.9190258157549843, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 28873 + }, + { + "epoch": 0.28874, + "grad_norm": 1.0153620134770482, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 28874 + }, + { + "epoch": 0.28875, + "grad_norm": 1.1383977568056713, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 28875 + }, + { + "epoch": 0.28876, + "grad_norm": 1.0928821687335337, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 28876 + }, + { + "epoch": 0.28877, + "grad_norm": 1.1774868144047925, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 28877 + }, + { + "epoch": 0.28878, + "grad_norm": 0.9928788448793388, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 28878 + }, + { + "epoch": 0.28879, + "grad_norm": 1.1837008936082434, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 28879 + }, + { + "epoch": 0.2888, + "grad_norm": 0.9742687895483421, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 28880 + }, + { + "epoch": 0.28881, + "grad_norm": 0.9811776308184476, + "learning_rate": 0.003, + "loss": 4.047, + "step": 28881 + }, + { + "epoch": 0.28882, + "grad_norm": 1.1896444341655876, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 28882 + }, + { + "epoch": 0.28883, + "grad_norm": 0.9511313325118811, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 28883 + }, + { + "epoch": 0.28884, + "grad_norm": 0.9781822991223896, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 28884 + }, + { + "epoch": 0.28885, + "grad_norm": 1.0455293767314902, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 28885 + }, + { + "epoch": 0.28886, + "grad_norm": 0.9403231937345885, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 28886 + }, + { + "epoch": 0.28887, + "grad_norm": 1.00279419032404, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 28887 + }, + { + "epoch": 0.28888, + "grad_norm": 1.0758855227924502, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 28888 + }, + { + "epoch": 0.28889, + "grad_norm": 1.1523299425072575, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 28889 + }, + { + "epoch": 0.2889, + "grad_norm": 1.1651797744914287, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 28890 + }, + { + "epoch": 0.28891, + "grad_norm": 1.0055283633019647, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 28891 + }, + { + "epoch": 0.28892, + "grad_norm": 1.3287331628907282, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 28892 + }, + { + "epoch": 0.28893, + "grad_norm": 0.77917611783554, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 28893 + }, + { + "epoch": 0.28894, + "grad_norm": 0.7794694938343664, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 28894 + }, + { + "epoch": 0.28895, + "grad_norm": 1.0709341286494862, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 28895 + }, + { + "epoch": 0.28896, + "grad_norm": 1.281376606878187, + "learning_rate": 0.003, + "loss": 4.052, + "step": 28896 + }, + { + "epoch": 0.28897, + "grad_norm": 0.9917816623770753, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 28897 + }, + { + "epoch": 0.28898, + "grad_norm": 1.299822825750924, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 28898 + }, + { + "epoch": 0.28899, + "grad_norm": 1.0081439229370617, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 28899 + }, + { + "epoch": 0.289, + "grad_norm": 1.239523484249895, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 28900 + }, + { + "epoch": 0.28901, + "grad_norm": 0.8208084690902573, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 28901 + }, + { + "epoch": 0.28902, + "grad_norm": 0.817554464605387, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 28902 + }, + { + "epoch": 0.28903, + "grad_norm": 1.05658882162586, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 28903 + }, + { + "epoch": 0.28904, + "grad_norm": 1.029494482282177, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 28904 + }, + { + "epoch": 0.28905, + "grad_norm": 1.0641463919246585, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 28905 + }, + { + "epoch": 0.28906, + "grad_norm": 1.0605055343289862, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 28906 + }, + { + "epoch": 0.28907, + "grad_norm": 1.1273219824175411, + "learning_rate": 0.003, + "loss": 4.029, + "step": 28907 + }, + { + "epoch": 0.28908, + "grad_norm": 1.0946044237463044, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 28908 + }, + { + "epoch": 0.28909, + "grad_norm": 1.002636644944635, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 28909 + }, + { + "epoch": 0.2891, + "grad_norm": 1.1229344795380838, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 28910 + }, + { + "epoch": 0.28911, + "grad_norm": 1.1741132271731547, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 28911 + }, + { + "epoch": 0.28912, + "grad_norm": 1.1428745672652456, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 28912 + }, + { + "epoch": 0.28913, + "grad_norm": 1.0115089058497908, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 28913 + }, + { + "epoch": 0.28914, + "grad_norm": 1.1394722194236246, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 28914 + }, + { + "epoch": 0.28915, + "grad_norm": 0.8466099588923972, + "learning_rate": 0.003, + "loss": 4.058, + "step": 28915 + }, + { + "epoch": 0.28916, + "grad_norm": 0.9970401714768062, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 28916 + }, + { + "epoch": 0.28917, + "grad_norm": 1.086191175305031, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 28917 + }, + { + "epoch": 0.28918, + "grad_norm": 0.9129150176820635, + "learning_rate": 0.003, + "loss": 4.063, + "step": 28918 + }, + { + "epoch": 0.28919, + "grad_norm": 1.1291061776385471, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 28919 + }, + { + "epoch": 0.2892, + "grad_norm": 1.1571843483694875, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 28920 + }, + { + "epoch": 0.28921, + "grad_norm": 0.9423134712594691, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 28921 + }, + { + "epoch": 0.28922, + "grad_norm": 1.078296156538658, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 28922 + }, + { + "epoch": 0.28923, + "grad_norm": 1.328262820094468, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 28923 + }, + { + "epoch": 0.28924, + "grad_norm": 1.0616767256072388, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 28924 + }, + { + "epoch": 0.28925, + "grad_norm": 1.1331740900063678, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 28925 + }, + { + "epoch": 0.28926, + "grad_norm": 1.191693448429308, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 28926 + }, + { + "epoch": 0.28927, + "grad_norm": 1.0362606870783309, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 28927 + }, + { + "epoch": 0.28928, + "grad_norm": 1.0551648802439018, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 28928 + }, + { + "epoch": 0.28929, + "grad_norm": 1.2600164111542425, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 28929 + }, + { + "epoch": 0.2893, + "grad_norm": 0.9140590209989776, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 28930 + }, + { + "epoch": 0.28931, + "grad_norm": 0.828786928765684, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 28931 + }, + { + "epoch": 0.28932, + "grad_norm": 1.028983907349773, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 28932 + }, + { + "epoch": 0.28933, + "grad_norm": 1.3153698532189813, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 28933 + }, + { + "epoch": 0.28934, + "grad_norm": 0.979603361742441, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 28934 + }, + { + "epoch": 0.28935, + "grad_norm": 0.9888895371714124, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 28935 + }, + { + "epoch": 0.28936, + "grad_norm": 1.0117720151894067, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 28936 + }, + { + "epoch": 0.28937, + "grad_norm": 1.3527001188626255, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 28937 + }, + { + "epoch": 0.28938, + "grad_norm": 0.9826641603706237, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 28938 + }, + { + "epoch": 0.28939, + "grad_norm": 0.9494825731628928, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 28939 + }, + { + "epoch": 0.2894, + "grad_norm": 1.0198655946327757, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 28940 + }, + { + "epoch": 0.28941, + "grad_norm": 1.4548592305397956, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 28941 + }, + { + "epoch": 0.28942, + "grad_norm": 0.8560447224985646, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 28942 + }, + { + "epoch": 0.28943, + "grad_norm": 0.9944302945218119, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 28943 + }, + { + "epoch": 0.28944, + "grad_norm": 0.9936352915014907, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 28944 + }, + { + "epoch": 0.28945, + "grad_norm": 1.1362193146847812, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 28945 + }, + { + "epoch": 0.28946, + "grad_norm": 0.9431113998113264, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 28946 + }, + { + "epoch": 0.28947, + "grad_norm": 1.2511423873362102, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 28947 + }, + { + "epoch": 0.28948, + "grad_norm": 1.0645796451651541, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 28948 + }, + { + "epoch": 0.28949, + "grad_norm": 0.8598062213162405, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 28949 + }, + { + "epoch": 0.2895, + "grad_norm": 0.9817681790489631, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 28950 + }, + { + "epoch": 0.28951, + "grad_norm": 1.4232187928309463, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 28951 + }, + { + "epoch": 0.28952, + "grad_norm": 0.9618337665246809, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 28952 + }, + { + "epoch": 0.28953, + "grad_norm": 0.9414656401699871, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 28953 + }, + { + "epoch": 0.28954, + "grad_norm": 0.9863887161843297, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 28954 + }, + { + "epoch": 0.28955, + "grad_norm": 1.0681598347369998, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 28955 + }, + { + "epoch": 0.28956, + "grad_norm": 1.1678420884116283, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 28956 + }, + { + "epoch": 0.28957, + "grad_norm": 1.0115799789472446, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 28957 + }, + { + "epoch": 0.28958, + "grad_norm": 1.0423032667697603, + "learning_rate": 0.003, + "loss": 4.06, + "step": 28958 + }, + { + "epoch": 0.28959, + "grad_norm": 1.120959507587759, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 28959 + }, + { + "epoch": 0.2896, + "grad_norm": 1.0501742827539438, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 28960 + }, + { + "epoch": 0.28961, + "grad_norm": 1.1034959172929981, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 28961 + }, + { + "epoch": 0.28962, + "grad_norm": 0.8605102034228963, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 28962 + }, + { + "epoch": 0.28963, + "grad_norm": 0.9714302320186714, + "learning_rate": 0.003, + "loss": 4.073, + "step": 28963 + }, + { + "epoch": 0.28964, + "grad_norm": 1.0806403823463282, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 28964 + }, + { + "epoch": 0.28965, + "grad_norm": 1.1455406885595616, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 28965 + }, + { + "epoch": 0.28966, + "grad_norm": 1.0348398103395824, + "learning_rate": 0.003, + "loss": 4.062, + "step": 28966 + }, + { + "epoch": 0.28967, + "grad_norm": 1.159315947953932, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 28967 + }, + { + "epoch": 0.28968, + "grad_norm": 1.0110129833592818, + "learning_rate": 0.003, + "loss": 4.048, + "step": 28968 + }, + { + "epoch": 0.28969, + "grad_norm": 1.3333068889300534, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 28969 + }, + { + "epoch": 0.2897, + "grad_norm": 1.0810879669702036, + "learning_rate": 0.003, + "loss": 4.024, + "step": 28970 + }, + { + "epoch": 0.28971, + "grad_norm": 1.1782448536976786, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 28971 + }, + { + "epoch": 0.28972, + "grad_norm": 0.8977925814440247, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 28972 + }, + { + "epoch": 0.28973, + "grad_norm": 1.1886787599055113, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 28973 + }, + { + "epoch": 0.28974, + "grad_norm": 1.1981168092825316, + "learning_rate": 0.003, + "loss": 4.073, + "step": 28974 + }, + { + "epoch": 0.28975, + "grad_norm": 1.1555679722790566, + "learning_rate": 0.003, + "loss": 4.041, + "step": 28975 + }, + { + "epoch": 0.28976, + "grad_norm": 1.0955212467607491, + "learning_rate": 0.003, + "loss": 4.047, + "step": 28976 + }, + { + "epoch": 0.28977, + "grad_norm": 1.186947879562948, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 28977 + }, + { + "epoch": 0.28978, + "grad_norm": 0.9590999373541235, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 28978 + }, + { + "epoch": 0.28979, + "grad_norm": 1.0948061522224575, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 28979 + }, + { + "epoch": 0.2898, + "grad_norm": 1.0274547476636966, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 28980 + }, + { + "epoch": 0.28981, + "grad_norm": 1.084943860170522, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 28981 + }, + { + "epoch": 0.28982, + "grad_norm": 0.9761557749604715, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 28982 + }, + { + "epoch": 0.28983, + "grad_norm": 1.3051300564084947, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 28983 + }, + { + "epoch": 0.28984, + "grad_norm": 1.0900378765644634, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 28984 + }, + { + "epoch": 0.28985, + "grad_norm": 1.0706423240896352, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 28985 + }, + { + "epoch": 0.28986, + "grad_norm": 0.862454627091812, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 28986 + }, + { + "epoch": 0.28987, + "grad_norm": 1.0110906504733999, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 28987 + }, + { + "epoch": 0.28988, + "grad_norm": 1.2407833367888048, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 28988 + }, + { + "epoch": 0.28989, + "grad_norm": 0.9208675981929476, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 28989 + }, + { + "epoch": 0.2899, + "grad_norm": 1.0931669256727503, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 28990 + }, + { + "epoch": 0.28991, + "grad_norm": 1.2257121311845778, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 28991 + }, + { + "epoch": 0.28992, + "grad_norm": 0.9628558281352907, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 28992 + }, + { + "epoch": 0.28993, + "grad_norm": 1.1326643050787826, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 28993 + }, + { + "epoch": 0.28994, + "grad_norm": 0.8875668481278278, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 28994 + }, + { + "epoch": 0.28995, + "grad_norm": 0.7959961608022296, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 28995 + }, + { + "epoch": 0.28996, + "grad_norm": 1.0792057431995137, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 28996 + }, + { + "epoch": 0.28997, + "grad_norm": 0.9799534897658739, + "learning_rate": 0.003, + "loss": 4.06, + "step": 28997 + }, + { + "epoch": 0.28998, + "grad_norm": 1.1105347483935806, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 28998 + }, + { + "epoch": 0.28999, + "grad_norm": 1.2692683082970198, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 28999 + }, + { + "epoch": 0.29, + "grad_norm": 1.0298229262708627, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 29000 + }, + { + "epoch": 0.29001, + "grad_norm": 1.0613225591576119, + "learning_rate": 0.003, + "loss": 4.0907, + "step": 29001 + }, + { + "epoch": 0.29002, + "grad_norm": 1.1143597201530426, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 29002 + }, + { + "epoch": 0.29003, + "grad_norm": 1.4052274053886213, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 29003 + }, + { + "epoch": 0.29004, + "grad_norm": 1.2925324405634133, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 29004 + }, + { + "epoch": 0.29005, + "grad_norm": 0.8399377551235675, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 29005 + }, + { + "epoch": 0.29006, + "grad_norm": 0.7684778553737899, + "learning_rate": 0.003, + "loss": 4.059, + "step": 29006 + }, + { + "epoch": 0.29007, + "grad_norm": 0.8477215488398882, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 29007 + }, + { + "epoch": 0.29008, + "grad_norm": 0.880410295364316, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 29008 + }, + { + "epoch": 0.29009, + "grad_norm": 1.0721588710416758, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 29009 + }, + { + "epoch": 0.2901, + "grad_norm": 1.1594926825260918, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 29010 + }, + { + "epoch": 0.29011, + "grad_norm": 1.2614281424326783, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 29011 + }, + { + "epoch": 0.29012, + "grad_norm": 1.1523221513180288, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 29012 + }, + { + "epoch": 0.29013, + "grad_norm": 0.8857981439818393, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 29013 + }, + { + "epoch": 0.29014, + "grad_norm": 0.8942331678599078, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 29014 + }, + { + "epoch": 0.29015, + "grad_norm": 1.0968779420794221, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 29015 + }, + { + "epoch": 0.29016, + "grad_norm": 1.1513168430436203, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 29016 + }, + { + "epoch": 0.29017, + "grad_norm": 0.8559946160502427, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 29017 + }, + { + "epoch": 0.29018, + "grad_norm": 0.8953367842692732, + "learning_rate": 0.003, + "loss": 4.034, + "step": 29018 + }, + { + "epoch": 0.29019, + "grad_norm": 1.054293116490111, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 29019 + }, + { + "epoch": 0.2902, + "grad_norm": 1.316620266042009, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 29020 + }, + { + "epoch": 0.29021, + "grad_norm": 1.0256632467043452, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 29021 + }, + { + "epoch": 0.29022, + "grad_norm": 1.0199244435911587, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 29022 + }, + { + "epoch": 0.29023, + "grad_norm": 1.0077105764696346, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 29023 + }, + { + "epoch": 0.29024, + "grad_norm": 1.1581914155761925, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 29024 + }, + { + "epoch": 0.29025, + "grad_norm": 0.9664962165131844, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 29025 + }, + { + "epoch": 0.29026, + "grad_norm": 1.0808938621276294, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 29026 + }, + { + "epoch": 0.29027, + "grad_norm": 1.0055103895844757, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 29027 + }, + { + "epoch": 0.29028, + "grad_norm": 0.9676453464301665, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 29028 + }, + { + "epoch": 0.29029, + "grad_norm": 1.085026104720533, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 29029 + }, + { + "epoch": 0.2903, + "grad_norm": 0.9912603991894178, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 29030 + }, + { + "epoch": 0.29031, + "grad_norm": 1.019975165674966, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 29031 + }, + { + "epoch": 0.29032, + "grad_norm": 1.157160274681798, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 29032 + }, + { + "epoch": 0.29033, + "grad_norm": 1.115756483934538, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 29033 + }, + { + "epoch": 0.29034, + "grad_norm": 1.4810187209243326, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 29034 + }, + { + "epoch": 0.29035, + "grad_norm": 0.9064949944053241, + "learning_rate": 0.003, + "loss": 4.045, + "step": 29035 + }, + { + "epoch": 0.29036, + "grad_norm": 0.9816554810319781, + "learning_rate": 0.003, + "loss": 4.052, + "step": 29036 + }, + { + "epoch": 0.29037, + "grad_norm": 1.4845666937998017, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 29037 + }, + { + "epoch": 0.29038, + "grad_norm": 0.9650700974093946, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 29038 + }, + { + "epoch": 0.29039, + "grad_norm": 1.09394890805738, + "learning_rate": 0.003, + "loss": 4.078, + "step": 29039 + }, + { + "epoch": 0.2904, + "grad_norm": 1.1375018670485764, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 29040 + }, + { + "epoch": 0.29041, + "grad_norm": 1.080706349799095, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 29041 + }, + { + "epoch": 0.29042, + "grad_norm": 1.250950247078964, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 29042 + }, + { + "epoch": 0.29043, + "grad_norm": 0.8516997888988722, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 29043 + }, + { + "epoch": 0.29044, + "grad_norm": 0.7653139184725098, + "learning_rate": 0.003, + "loss": 4.051, + "step": 29044 + }, + { + "epoch": 0.29045, + "grad_norm": 0.9013764860425266, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 29045 + }, + { + "epoch": 0.29046, + "grad_norm": 0.9023266924062829, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 29046 + }, + { + "epoch": 0.29047, + "grad_norm": 1.0964783515711702, + "learning_rate": 0.003, + "loss": 4.065, + "step": 29047 + }, + { + "epoch": 0.29048, + "grad_norm": 0.9441905093766498, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 29048 + }, + { + "epoch": 0.29049, + "grad_norm": 1.0438226301796303, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 29049 + }, + { + "epoch": 0.2905, + "grad_norm": 1.290060717919776, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 29050 + }, + { + "epoch": 0.29051, + "grad_norm": 0.9930833507183314, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 29051 + }, + { + "epoch": 0.29052, + "grad_norm": 1.2215686620087725, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 29052 + }, + { + "epoch": 0.29053, + "grad_norm": 0.9178478440055774, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 29053 + }, + { + "epoch": 0.29054, + "grad_norm": 0.8985181648541861, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 29054 + }, + { + "epoch": 0.29055, + "grad_norm": 1.0152498432559514, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 29055 + }, + { + "epoch": 0.29056, + "grad_norm": 1.1471518098866842, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 29056 + }, + { + "epoch": 0.29057, + "grad_norm": 0.9673601227963061, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 29057 + }, + { + "epoch": 0.29058, + "grad_norm": 0.97254852278344, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 29058 + }, + { + "epoch": 0.29059, + "grad_norm": 1.2836183655450946, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 29059 + }, + { + "epoch": 0.2906, + "grad_norm": 0.9774765696466567, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 29060 + }, + { + "epoch": 0.29061, + "grad_norm": 1.2291303004965768, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 29061 + }, + { + "epoch": 0.29062, + "grad_norm": 1.0959765639236858, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 29062 + }, + { + "epoch": 0.29063, + "grad_norm": 1.2008207517246814, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 29063 + }, + { + "epoch": 0.29064, + "grad_norm": 1.0737761784348976, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 29064 + }, + { + "epoch": 0.29065, + "grad_norm": 1.0467441253876175, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 29065 + }, + { + "epoch": 0.29066, + "grad_norm": 1.0825219317235026, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 29066 + }, + { + "epoch": 0.29067, + "grad_norm": 1.165111463993616, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 29067 + }, + { + "epoch": 0.29068, + "grad_norm": 1.0369697669375477, + "learning_rate": 0.003, + "loss": 4.045, + "step": 29068 + }, + { + "epoch": 0.29069, + "grad_norm": 0.9815804501369969, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 29069 + }, + { + "epoch": 0.2907, + "grad_norm": 0.9345501486196909, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 29070 + }, + { + "epoch": 0.29071, + "grad_norm": 1.0665026295379065, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 29071 + }, + { + "epoch": 0.29072, + "grad_norm": 1.1284948673331363, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 29072 + }, + { + "epoch": 0.29073, + "grad_norm": 0.9394313664252618, + "learning_rate": 0.003, + "loss": 4.008, + "step": 29073 + }, + { + "epoch": 0.29074, + "grad_norm": 0.9752886139370084, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 29074 + }, + { + "epoch": 0.29075, + "grad_norm": 1.024121912061539, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 29075 + }, + { + "epoch": 0.29076, + "grad_norm": 1.1513229186873688, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 29076 + }, + { + "epoch": 0.29077, + "grad_norm": 0.9290763066620953, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 29077 + }, + { + "epoch": 0.29078, + "grad_norm": 1.1315465990752247, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 29078 + }, + { + "epoch": 0.29079, + "grad_norm": 1.1238994464730225, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 29079 + }, + { + "epoch": 0.2908, + "grad_norm": 1.1254410169633924, + "learning_rate": 0.003, + "loss": 4.078, + "step": 29080 + }, + { + "epoch": 0.29081, + "grad_norm": 0.9529285537381482, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 29081 + }, + { + "epoch": 0.29082, + "grad_norm": 0.9327462436405851, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 29082 + }, + { + "epoch": 0.29083, + "grad_norm": 1.0187105018232439, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 29083 + }, + { + "epoch": 0.29084, + "grad_norm": 1.1574861594672063, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 29084 + }, + { + "epoch": 0.29085, + "grad_norm": 1.0454381824498413, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 29085 + }, + { + "epoch": 0.29086, + "grad_norm": 1.228211937415368, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 29086 + }, + { + "epoch": 0.29087, + "grad_norm": 0.9154031064533922, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 29087 + }, + { + "epoch": 0.29088, + "grad_norm": 0.863369020397376, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 29088 + }, + { + "epoch": 0.29089, + "grad_norm": 0.8990715644994332, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 29089 + }, + { + "epoch": 0.2909, + "grad_norm": 1.3441346290134615, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 29090 + }, + { + "epoch": 0.29091, + "grad_norm": 1.266155187826164, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 29091 + }, + { + "epoch": 0.29092, + "grad_norm": 1.0338654811631427, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 29092 + }, + { + "epoch": 0.29093, + "grad_norm": 1.2139237705732346, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 29093 + }, + { + "epoch": 0.29094, + "grad_norm": 1.0876700830361974, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 29094 + }, + { + "epoch": 0.29095, + "grad_norm": 1.0684289178140083, + "learning_rate": 0.003, + "loss": 4.054, + "step": 29095 + }, + { + "epoch": 0.29096, + "grad_norm": 0.881388073265663, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 29096 + }, + { + "epoch": 0.29097, + "grad_norm": 0.9184274725318734, + "learning_rate": 0.003, + "loss": 4.076, + "step": 29097 + }, + { + "epoch": 0.29098, + "grad_norm": 0.9925023601664581, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 29098 + }, + { + "epoch": 0.29099, + "grad_norm": 1.1380604374872096, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 29099 + }, + { + "epoch": 0.291, + "grad_norm": 1.1867845615223211, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 29100 + }, + { + "epoch": 0.29101, + "grad_norm": 1.0187677641897879, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 29101 + }, + { + "epoch": 0.29102, + "grad_norm": 1.1098415666907104, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 29102 + }, + { + "epoch": 0.29103, + "grad_norm": 1.0354237343850674, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 29103 + }, + { + "epoch": 0.29104, + "grad_norm": 1.0750173274721078, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 29104 + }, + { + "epoch": 0.29105, + "grad_norm": 1.0014064609200575, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 29105 + }, + { + "epoch": 0.29106, + "grad_norm": 1.0242543597303806, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 29106 + }, + { + "epoch": 0.29107, + "grad_norm": 1.3358834009669174, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 29107 + }, + { + "epoch": 0.29108, + "grad_norm": 1.1910777457154995, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 29108 + }, + { + "epoch": 0.29109, + "grad_norm": 1.0415314304072518, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 29109 + }, + { + "epoch": 0.2911, + "grad_norm": 0.9573583882374793, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 29110 + }, + { + "epoch": 0.29111, + "grad_norm": 0.9697738858835683, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 29111 + }, + { + "epoch": 0.29112, + "grad_norm": 1.349129910128441, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 29112 + }, + { + "epoch": 0.29113, + "grad_norm": 0.8895451979380197, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 29113 + }, + { + "epoch": 0.29114, + "grad_norm": 1.0535119843120009, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 29114 + }, + { + "epoch": 0.29115, + "grad_norm": 1.092640316848432, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 29115 + }, + { + "epoch": 0.29116, + "grad_norm": 1.1304235193354426, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 29116 + }, + { + "epoch": 0.29117, + "grad_norm": 1.3380999269205829, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 29117 + }, + { + "epoch": 0.29118, + "grad_norm": 0.9371429818621941, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 29118 + }, + { + "epoch": 0.29119, + "grad_norm": 0.8154314001464738, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 29119 + }, + { + "epoch": 0.2912, + "grad_norm": 0.905676864835108, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 29120 + }, + { + "epoch": 0.29121, + "grad_norm": 1.1406809164046614, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 29121 + }, + { + "epoch": 0.29122, + "grad_norm": 1.3627581606268757, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 29122 + }, + { + "epoch": 0.29123, + "grad_norm": 1.0044361548272727, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 29123 + }, + { + "epoch": 0.29124, + "grad_norm": 1.2740033742030912, + "learning_rate": 0.003, + "loss": 4.046, + "step": 29124 + }, + { + "epoch": 0.29125, + "grad_norm": 1.0466387050880643, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 29125 + }, + { + "epoch": 0.29126, + "grad_norm": 1.268707619718581, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 29126 + }, + { + "epoch": 0.29127, + "grad_norm": 0.9873031031581314, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 29127 + }, + { + "epoch": 0.29128, + "grad_norm": 0.9403493018602056, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 29128 + }, + { + "epoch": 0.29129, + "grad_norm": 1.1594505203386354, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 29129 + }, + { + "epoch": 0.2913, + "grad_norm": 0.9586813380841828, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 29130 + }, + { + "epoch": 0.29131, + "grad_norm": 1.0384165506545593, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 29131 + }, + { + "epoch": 0.29132, + "grad_norm": 1.146718640303647, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 29132 + }, + { + "epoch": 0.29133, + "grad_norm": 0.896164312081771, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 29133 + }, + { + "epoch": 0.29134, + "grad_norm": 0.8804389154331278, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 29134 + }, + { + "epoch": 0.29135, + "grad_norm": 1.0223593622836682, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 29135 + }, + { + "epoch": 0.29136, + "grad_norm": 1.0945303834939317, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 29136 + }, + { + "epoch": 0.29137, + "grad_norm": 0.9098649249890415, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 29137 + }, + { + "epoch": 0.29138, + "grad_norm": 1.0738444211367042, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 29138 + }, + { + "epoch": 0.29139, + "grad_norm": 1.4614513034005063, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 29139 + }, + { + "epoch": 0.2914, + "grad_norm": 1.1124797206739034, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 29140 + }, + { + "epoch": 0.29141, + "grad_norm": 0.9786591892427519, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 29141 + }, + { + "epoch": 0.29142, + "grad_norm": 0.9951930283850062, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 29142 + }, + { + "epoch": 0.29143, + "grad_norm": 1.2270661197789503, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 29143 + }, + { + "epoch": 0.29144, + "grad_norm": 1.0616959615420378, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 29144 + }, + { + "epoch": 0.29145, + "grad_norm": 0.9771694187328386, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 29145 + }, + { + "epoch": 0.29146, + "grad_norm": 0.9458925540109012, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 29146 + }, + { + "epoch": 0.29147, + "grad_norm": 1.0206710064959816, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 29147 + }, + { + "epoch": 0.29148, + "grad_norm": 1.0001839439731615, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 29148 + }, + { + "epoch": 0.29149, + "grad_norm": 1.2312929730964604, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 29149 + }, + { + "epoch": 0.2915, + "grad_norm": 0.8262558882669492, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 29150 + }, + { + "epoch": 0.29151, + "grad_norm": 0.8953555519609631, + "learning_rate": 0.003, + "loss": 4.069, + "step": 29151 + }, + { + "epoch": 0.29152, + "grad_norm": 1.0671348912476684, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 29152 + }, + { + "epoch": 0.29153, + "grad_norm": 1.3082052195216498, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 29153 + }, + { + "epoch": 0.29154, + "grad_norm": 1.0274196707729675, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 29154 + }, + { + "epoch": 0.29155, + "grad_norm": 1.252829654616139, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 29155 + }, + { + "epoch": 0.29156, + "grad_norm": 0.8798913662879674, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 29156 + }, + { + "epoch": 0.29157, + "grad_norm": 0.9530368185853929, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 29157 + }, + { + "epoch": 0.29158, + "grad_norm": 1.2554758708276554, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 29158 + }, + { + "epoch": 0.29159, + "grad_norm": 1.2708148970844273, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 29159 + }, + { + "epoch": 0.2916, + "grad_norm": 0.972463828493336, + "learning_rate": 0.003, + "loss": 4.056, + "step": 29160 + }, + { + "epoch": 0.29161, + "grad_norm": 0.9943885312215908, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 29161 + }, + { + "epoch": 0.29162, + "grad_norm": 1.2374546898363186, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 29162 + }, + { + "epoch": 0.29163, + "grad_norm": 1.2755533907961447, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 29163 + }, + { + "epoch": 0.29164, + "grad_norm": 0.9136844119850526, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 29164 + }, + { + "epoch": 0.29165, + "grad_norm": 0.9049554100272833, + "learning_rate": 0.003, + "loss": 4.016, + "step": 29165 + }, + { + "epoch": 0.29166, + "grad_norm": 1.1367769585961627, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 29166 + }, + { + "epoch": 0.29167, + "grad_norm": 0.9079465120294479, + "learning_rate": 0.003, + "loss": 4.037, + "step": 29167 + }, + { + "epoch": 0.29168, + "grad_norm": 0.9514976040025075, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 29168 + }, + { + "epoch": 0.29169, + "grad_norm": 1.096138378591896, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 29169 + }, + { + "epoch": 0.2917, + "grad_norm": 1.0700939589830503, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 29170 + }, + { + "epoch": 0.29171, + "grad_norm": 1.1963423116976322, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 29171 + }, + { + "epoch": 0.29172, + "grad_norm": 0.8832235369625953, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 29172 + }, + { + "epoch": 0.29173, + "grad_norm": 1.127189164282233, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 29173 + }, + { + "epoch": 0.29174, + "grad_norm": 0.9362616033786677, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 29174 + }, + { + "epoch": 0.29175, + "grad_norm": 0.94707593057727, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 29175 + }, + { + "epoch": 0.29176, + "grad_norm": 1.2417263653508703, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 29176 + }, + { + "epoch": 0.29177, + "grad_norm": 0.9519473091465448, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 29177 + }, + { + "epoch": 0.29178, + "grad_norm": 0.973787477015651, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 29178 + }, + { + "epoch": 0.29179, + "grad_norm": 1.294938623093467, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 29179 + }, + { + "epoch": 0.2918, + "grad_norm": 1.03650836108577, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 29180 + }, + { + "epoch": 0.29181, + "grad_norm": 1.2407830175218195, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 29181 + }, + { + "epoch": 0.29182, + "grad_norm": 0.9098592635354799, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 29182 + }, + { + "epoch": 0.29183, + "grad_norm": 1.0331211059238736, + "learning_rate": 0.003, + "loss": 4.04, + "step": 29183 + }, + { + "epoch": 0.29184, + "grad_norm": 1.1481753437014575, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 29184 + }, + { + "epoch": 0.29185, + "grad_norm": 0.8792814747390454, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 29185 + }, + { + "epoch": 0.29186, + "grad_norm": 1.0599602319114705, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 29186 + }, + { + "epoch": 0.29187, + "grad_norm": 1.4112218970524668, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 29187 + }, + { + "epoch": 0.29188, + "grad_norm": 0.9748814117665104, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 29188 + }, + { + "epoch": 0.29189, + "grad_norm": 1.0762433556998985, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 29189 + }, + { + "epoch": 0.2919, + "grad_norm": 1.1871184245970965, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 29190 + }, + { + "epoch": 0.29191, + "grad_norm": 1.0056802876792532, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 29191 + }, + { + "epoch": 0.29192, + "grad_norm": 1.1574196492860604, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 29192 + }, + { + "epoch": 0.29193, + "grad_norm": 1.0830206647819351, + "learning_rate": 0.003, + "loss": 4.049, + "step": 29193 + }, + { + "epoch": 0.29194, + "grad_norm": 1.0376476979179663, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 29194 + }, + { + "epoch": 0.29195, + "grad_norm": 1.1551605473549584, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 29195 + }, + { + "epoch": 0.29196, + "grad_norm": 1.0498787558651541, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 29196 + }, + { + "epoch": 0.29197, + "grad_norm": 1.1750556924504076, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 29197 + }, + { + "epoch": 0.29198, + "grad_norm": 0.9280894865374645, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 29198 + }, + { + "epoch": 0.29199, + "grad_norm": 0.9813945056192752, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 29199 + }, + { + "epoch": 0.292, + "grad_norm": 1.080821661077971, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 29200 + }, + { + "epoch": 0.29201, + "grad_norm": 0.9990349675833224, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 29201 + }, + { + "epoch": 0.29202, + "grad_norm": 1.081992155774376, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 29202 + }, + { + "epoch": 0.29203, + "grad_norm": 1.2790656863567669, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 29203 + }, + { + "epoch": 0.29204, + "grad_norm": 0.9197425849517732, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 29204 + }, + { + "epoch": 0.29205, + "grad_norm": 1.1236322137977106, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 29205 + }, + { + "epoch": 0.29206, + "grad_norm": 1.376085049088111, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 29206 + }, + { + "epoch": 0.29207, + "grad_norm": 0.9591903401619662, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 29207 + }, + { + "epoch": 0.29208, + "grad_norm": 1.0186397810557253, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 29208 + }, + { + "epoch": 0.29209, + "grad_norm": 1.1968639749684873, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 29209 + }, + { + "epoch": 0.2921, + "grad_norm": 0.935410359869728, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 29210 + }, + { + "epoch": 0.29211, + "grad_norm": 1.1231260763687005, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 29211 + }, + { + "epoch": 0.29212, + "grad_norm": 1.082312603032422, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 29212 + }, + { + "epoch": 0.29213, + "grad_norm": 1.036232081020505, + "learning_rate": 0.003, + "loss": 4.044, + "step": 29213 + }, + { + "epoch": 0.29214, + "grad_norm": 1.004407901882765, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 29214 + }, + { + "epoch": 0.29215, + "grad_norm": 1.1136643755614575, + "learning_rate": 0.003, + "loss": 4.043, + "step": 29215 + }, + { + "epoch": 0.29216, + "grad_norm": 0.9384026559898834, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 29216 + }, + { + "epoch": 0.29217, + "grad_norm": 1.0689204614527534, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 29217 + }, + { + "epoch": 0.29218, + "grad_norm": 0.8790378130528568, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 29218 + }, + { + "epoch": 0.29219, + "grad_norm": 1.0314114025947034, + "learning_rate": 0.003, + "loss": 4.1019, + "step": 29219 + }, + { + "epoch": 0.2922, + "grad_norm": 1.0879490651573223, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 29220 + }, + { + "epoch": 0.29221, + "grad_norm": 1.0234307229696311, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 29221 + }, + { + "epoch": 0.29222, + "grad_norm": 1.0192558018452063, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 29222 + }, + { + "epoch": 0.29223, + "grad_norm": 1.304787167662759, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 29223 + }, + { + "epoch": 0.29224, + "grad_norm": 0.9821831273812625, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 29224 + }, + { + "epoch": 0.29225, + "grad_norm": 1.1492517639404314, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 29225 + }, + { + "epoch": 0.29226, + "grad_norm": 1.144249721274636, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 29226 + }, + { + "epoch": 0.29227, + "grad_norm": 1.2208147572586363, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 29227 + }, + { + "epoch": 0.29228, + "grad_norm": 0.896221174293353, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 29228 + }, + { + "epoch": 0.29229, + "grad_norm": 0.9604158018429966, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 29229 + }, + { + "epoch": 0.2923, + "grad_norm": 1.0747303061009812, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 29230 + }, + { + "epoch": 0.29231, + "grad_norm": 1.1452567053532932, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 29231 + }, + { + "epoch": 0.29232, + "grad_norm": 1.0813183590682012, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 29232 + }, + { + "epoch": 0.29233, + "grad_norm": 0.9563458727556441, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 29233 + }, + { + "epoch": 0.29234, + "grad_norm": 1.0606853128096199, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 29234 + }, + { + "epoch": 0.29235, + "grad_norm": 0.8081022761585719, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 29235 + }, + { + "epoch": 0.29236, + "grad_norm": 0.8788798971618882, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 29236 + }, + { + "epoch": 0.29237, + "grad_norm": 0.9923237604208585, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 29237 + }, + { + "epoch": 0.29238, + "grad_norm": 1.1364122724935455, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 29238 + }, + { + "epoch": 0.29239, + "grad_norm": 1.0294932728340374, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 29239 + }, + { + "epoch": 0.2924, + "grad_norm": 1.3004699807212599, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 29240 + }, + { + "epoch": 0.29241, + "grad_norm": 0.8691425655725709, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 29241 + }, + { + "epoch": 0.29242, + "grad_norm": 0.9574372646820628, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 29242 + }, + { + "epoch": 0.29243, + "grad_norm": 1.0917669296220993, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 29243 + }, + { + "epoch": 0.29244, + "grad_norm": 1.0712407288115922, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 29244 + }, + { + "epoch": 0.29245, + "grad_norm": 1.0951127262878781, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 29245 + }, + { + "epoch": 0.29246, + "grad_norm": 1.0245301163493863, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 29246 + }, + { + "epoch": 0.29247, + "grad_norm": 1.3763556561610786, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 29247 + }, + { + "epoch": 0.29248, + "grad_norm": 1.2494966092964142, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 29248 + }, + { + "epoch": 0.29249, + "grad_norm": 1.004269584770182, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 29249 + }, + { + "epoch": 0.2925, + "grad_norm": 1.2665356636711098, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 29250 + }, + { + "epoch": 0.29251, + "grad_norm": 0.8870387354568242, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 29251 + }, + { + "epoch": 0.29252, + "grad_norm": 1.002723435364974, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 29252 + }, + { + "epoch": 0.29253, + "grad_norm": 1.2806630241469483, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 29253 + }, + { + "epoch": 0.29254, + "grad_norm": 0.8528522699995147, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 29254 + }, + { + "epoch": 0.29255, + "grad_norm": 0.9762387758737637, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 29255 + }, + { + "epoch": 0.29256, + "grad_norm": 1.307835663041549, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 29256 + }, + { + "epoch": 0.29257, + "grad_norm": 1.0096632633226579, + "learning_rate": 0.003, + "loss": 4.049, + "step": 29257 + }, + { + "epoch": 0.29258, + "grad_norm": 1.1577731271378031, + "learning_rate": 0.003, + "loss": 4.06, + "step": 29258 + }, + { + "epoch": 0.29259, + "grad_norm": 0.9984698788622808, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 29259 + }, + { + "epoch": 0.2926, + "grad_norm": 1.182478347380431, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 29260 + }, + { + "epoch": 0.29261, + "grad_norm": 1.0171253969415521, + "learning_rate": 0.003, + "loss": 4.088, + "step": 29261 + }, + { + "epoch": 0.29262, + "grad_norm": 1.1193415261177038, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 29262 + }, + { + "epoch": 0.29263, + "grad_norm": 0.8880240992611034, + "learning_rate": 0.003, + "loss": 4.069, + "step": 29263 + }, + { + "epoch": 0.29264, + "grad_norm": 0.8858997058246613, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 29264 + }, + { + "epoch": 0.29265, + "grad_norm": 0.9421848551784586, + "learning_rate": 0.003, + "loss": 4.073, + "step": 29265 + }, + { + "epoch": 0.29266, + "grad_norm": 1.1211845540566783, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 29266 + }, + { + "epoch": 0.29267, + "grad_norm": 1.0858007022064455, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 29267 + }, + { + "epoch": 0.29268, + "grad_norm": 1.0400210457336307, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 29268 + }, + { + "epoch": 0.29269, + "grad_norm": 0.9573567310727708, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 29269 + }, + { + "epoch": 0.2927, + "grad_norm": 1.1263874620414194, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 29270 + }, + { + "epoch": 0.29271, + "grad_norm": 0.9281797946731708, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 29271 + }, + { + "epoch": 0.29272, + "grad_norm": 0.8941371252290468, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 29272 + }, + { + "epoch": 0.29273, + "grad_norm": 0.8025960825369988, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 29273 + }, + { + "epoch": 0.29274, + "grad_norm": 1.0202984776696238, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 29274 + }, + { + "epoch": 0.29275, + "grad_norm": 1.511706945808506, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 29275 + }, + { + "epoch": 0.29276, + "grad_norm": 1.0069791966579404, + "learning_rate": 0.003, + "loss": 4.045, + "step": 29276 + }, + { + "epoch": 0.29277, + "grad_norm": 1.2594653044502042, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 29277 + }, + { + "epoch": 0.29278, + "grad_norm": 1.2949454856185767, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 29278 + }, + { + "epoch": 0.29279, + "grad_norm": 1.0003976796193528, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 29279 + }, + { + "epoch": 0.2928, + "grad_norm": 0.9924711094776009, + "learning_rate": 0.003, + "loss": 4.046, + "step": 29280 + }, + { + "epoch": 0.29281, + "grad_norm": 1.0504669953860457, + "learning_rate": 0.003, + "loss": 4.056, + "step": 29281 + }, + { + "epoch": 0.29282, + "grad_norm": 1.1523417014066948, + "learning_rate": 0.003, + "loss": 4.029, + "step": 29282 + }, + { + "epoch": 0.29283, + "grad_norm": 1.190391527136254, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 29283 + }, + { + "epoch": 0.29284, + "grad_norm": 0.9562637786806929, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 29284 + }, + { + "epoch": 0.29285, + "grad_norm": 0.999490079329279, + "learning_rate": 0.003, + "loss": 4.048, + "step": 29285 + }, + { + "epoch": 0.29286, + "grad_norm": 1.3184906462660806, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 29286 + }, + { + "epoch": 0.29287, + "grad_norm": 0.8135450147800392, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 29287 + }, + { + "epoch": 0.29288, + "grad_norm": 0.7807222576104342, + "learning_rate": 0.003, + "loss": 4.027, + "step": 29288 + }, + { + "epoch": 0.29289, + "grad_norm": 0.7883317427435306, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 29289 + }, + { + "epoch": 0.2929, + "grad_norm": 0.7829322626958297, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 29290 + }, + { + "epoch": 0.29291, + "grad_norm": 0.7933808809484754, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 29291 + }, + { + "epoch": 0.29292, + "grad_norm": 1.0877918095147536, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 29292 + }, + { + "epoch": 0.29293, + "grad_norm": 1.3333696247194626, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 29293 + }, + { + "epoch": 0.29294, + "grad_norm": 0.8605638998933983, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 29294 + }, + { + "epoch": 0.29295, + "grad_norm": 0.7561381374454009, + "learning_rate": 0.003, + "loss": 4.027, + "step": 29295 + }, + { + "epoch": 0.29296, + "grad_norm": 0.7670639106225743, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 29296 + }, + { + "epoch": 0.29297, + "grad_norm": 0.9495245958551363, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 29297 + }, + { + "epoch": 0.29298, + "grad_norm": 1.220921211605004, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 29298 + }, + { + "epoch": 0.29299, + "grad_norm": 0.9278398835540781, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 29299 + }, + { + "epoch": 0.293, + "grad_norm": 0.9996405678356316, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 29300 + }, + { + "epoch": 0.29301, + "grad_norm": 1.0548658105473185, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 29301 + }, + { + "epoch": 0.29302, + "grad_norm": 1.0767119735687851, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 29302 + }, + { + "epoch": 0.29303, + "grad_norm": 1.1332481675888313, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 29303 + }, + { + "epoch": 0.29304, + "grad_norm": 1.135751299721351, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 29304 + }, + { + "epoch": 0.29305, + "grad_norm": 1.2362785137110406, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 29305 + }, + { + "epoch": 0.29306, + "grad_norm": 0.9115212165226684, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 29306 + }, + { + "epoch": 0.29307, + "grad_norm": 1.0283752664011643, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 29307 + }, + { + "epoch": 0.29308, + "grad_norm": 1.2496039843921336, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 29308 + }, + { + "epoch": 0.29309, + "grad_norm": 1.198846481302419, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 29309 + }, + { + "epoch": 0.2931, + "grad_norm": 1.2161142319018856, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 29310 + }, + { + "epoch": 0.29311, + "grad_norm": 1.0929975527690743, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 29311 + }, + { + "epoch": 0.29312, + "grad_norm": 1.1729348594894455, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 29312 + }, + { + "epoch": 0.29313, + "grad_norm": 1.063441054930478, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 29313 + }, + { + "epoch": 0.29314, + "grad_norm": 1.1927328100945445, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 29314 + }, + { + "epoch": 0.29315, + "grad_norm": 1.0777290967948814, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 29315 + }, + { + "epoch": 0.29316, + "grad_norm": 1.1471299031436821, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 29316 + }, + { + "epoch": 0.29317, + "grad_norm": 0.9912697644732641, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 29317 + }, + { + "epoch": 0.29318, + "grad_norm": 1.1783399852549779, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 29318 + }, + { + "epoch": 0.29319, + "grad_norm": 0.9468684608538515, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 29319 + }, + { + "epoch": 0.2932, + "grad_norm": 0.899567893617548, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 29320 + }, + { + "epoch": 0.29321, + "grad_norm": 1.0053795532184402, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 29321 + }, + { + "epoch": 0.29322, + "grad_norm": 1.0502813949796417, + "learning_rate": 0.003, + "loss": 4.036, + "step": 29322 + }, + { + "epoch": 0.29323, + "grad_norm": 0.8849490970332289, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 29323 + }, + { + "epoch": 0.29324, + "grad_norm": 0.8840707030168373, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 29324 + }, + { + "epoch": 0.29325, + "grad_norm": 1.1829973851884528, + "learning_rate": 0.003, + "loss": 4.046, + "step": 29325 + }, + { + "epoch": 0.29326, + "grad_norm": 1.4114721346553265, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 29326 + }, + { + "epoch": 0.29327, + "grad_norm": 0.8955694222489897, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 29327 + }, + { + "epoch": 0.29328, + "grad_norm": 1.100885973317773, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 29328 + }, + { + "epoch": 0.29329, + "grad_norm": 1.21185537197748, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 29329 + }, + { + "epoch": 0.2933, + "grad_norm": 0.8111318823129664, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 29330 + }, + { + "epoch": 0.29331, + "grad_norm": 0.7847056610903264, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 29331 + }, + { + "epoch": 0.29332, + "grad_norm": 0.9293199058790814, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 29332 + }, + { + "epoch": 0.29333, + "grad_norm": 1.0232801636704845, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 29333 + }, + { + "epoch": 0.29334, + "grad_norm": 0.9862829232137806, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 29334 + }, + { + "epoch": 0.29335, + "grad_norm": 1.0493769408215918, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 29335 + }, + { + "epoch": 0.29336, + "grad_norm": 1.0031675965818772, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 29336 + }, + { + "epoch": 0.29337, + "grad_norm": 1.14658291128205, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 29337 + }, + { + "epoch": 0.29338, + "grad_norm": 1.2151989301594477, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 29338 + }, + { + "epoch": 0.29339, + "grad_norm": 1.2179351149916298, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 29339 + }, + { + "epoch": 0.2934, + "grad_norm": 1.0649806240205466, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 29340 + }, + { + "epoch": 0.29341, + "grad_norm": 1.2395092052630157, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 29341 + }, + { + "epoch": 0.29342, + "grad_norm": 1.1281230326770728, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 29342 + }, + { + "epoch": 0.29343, + "grad_norm": 1.2135036190688056, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 29343 + }, + { + "epoch": 0.29344, + "grad_norm": 0.902611350758833, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 29344 + }, + { + "epoch": 0.29345, + "grad_norm": 0.8209685084201805, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 29345 + }, + { + "epoch": 0.29346, + "grad_norm": 1.0450353323193082, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 29346 + }, + { + "epoch": 0.29347, + "grad_norm": 1.4916788204010312, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 29347 + }, + { + "epoch": 0.29348, + "grad_norm": 0.7475476189457558, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 29348 + }, + { + "epoch": 0.29349, + "grad_norm": 0.8821610530952078, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 29349 + }, + { + "epoch": 0.2935, + "grad_norm": 1.0517078618462752, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 29350 + }, + { + "epoch": 0.29351, + "grad_norm": 1.3851289740876287, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 29351 + }, + { + "epoch": 0.29352, + "grad_norm": 0.7714294048225221, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 29352 + }, + { + "epoch": 0.29353, + "grad_norm": 0.7652842208483686, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 29353 + }, + { + "epoch": 0.29354, + "grad_norm": 0.8062555919057249, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 29354 + }, + { + "epoch": 0.29355, + "grad_norm": 0.8020335186263488, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 29355 + }, + { + "epoch": 0.29356, + "grad_norm": 0.8306838826208753, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 29356 + }, + { + "epoch": 0.29357, + "grad_norm": 0.815495241857203, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 29357 + }, + { + "epoch": 0.29358, + "grad_norm": 0.8334354503347282, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 29358 + }, + { + "epoch": 0.29359, + "grad_norm": 0.9451933820574103, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 29359 + }, + { + "epoch": 0.2936, + "grad_norm": 1.3862068984297256, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 29360 + }, + { + "epoch": 0.29361, + "grad_norm": 1.2507278257463599, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 29361 + }, + { + "epoch": 0.29362, + "grad_norm": 0.9651834768292706, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 29362 + }, + { + "epoch": 0.29363, + "grad_norm": 1.1548179599017663, + "learning_rate": 0.003, + "loss": 4.058, + "step": 29363 + }, + { + "epoch": 0.29364, + "grad_norm": 1.2214713858875526, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 29364 + }, + { + "epoch": 0.29365, + "grad_norm": 1.1133867505957527, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 29365 + }, + { + "epoch": 0.29366, + "grad_norm": 1.0294918092782497, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 29366 + }, + { + "epoch": 0.29367, + "grad_norm": 1.0695077768696115, + "learning_rate": 0.003, + "loss": 4.062, + "step": 29367 + }, + { + "epoch": 0.29368, + "grad_norm": 1.0418694803635005, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 29368 + }, + { + "epoch": 0.29369, + "grad_norm": 1.0431991803761653, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 29369 + }, + { + "epoch": 0.2937, + "grad_norm": 1.0314225697073567, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 29370 + }, + { + "epoch": 0.29371, + "grad_norm": 1.205206436756868, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 29371 + }, + { + "epoch": 0.29372, + "grad_norm": 1.2914824621497316, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 29372 + }, + { + "epoch": 0.29373, + "grad_norm": 1.036972568418131, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 29373 + }, + { + "epoch": 0.29374, + "grad_norm": 1.072335909207184, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 29374 + }, + { + "epoch": 0.29375, + "grad_norm": 0.9686846999751549, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 29375 + }, + { + "epoch": 0.29376, + "grad_norm": 1.0176364214779068, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 29376 + }, + { + "epoch": 0.29377, + "grad_norm": 1.1388933000206105, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 29377 + }, + { + "epoch": 0.29378, + "grad_norm": 1.0722652062940377, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 29378 + }, + { + "epoch": 0.29379, + "grad_norm": 1.1014947635050656, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 29379 + }, + { + "epoch": 0.2938, + "grad_norm": 0.9384851241906019, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 29380 + }, + { + "epoch": 0.29381, + "grad_norm": 0.9715022107029464, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 29381 + }, + { + "epoch": 0.29382, + "grad_norm": 1.0269209777736312, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 29382 + }, + { + "epoch": 0.29383, + "grad_norm": 1.0070934914932488, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 29383 + }, + { + "epoch": 0.29384, + "grad_norm": 1.130434545285956, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 29384 + }, + { + "epoch": 0.29385, + "grad_norm": 1.0792405843488027, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 29385 + }, + { + "epoch": 0.29386, + "grad_norm": 1.2980841121783726, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 29386 + }, + { + "epoch": 0.29387, + "grad_norm": 1.1430814572186239, + "learning_rate": 0.003, + "loss": 4.057, + "step": 29387 + }, + { + "epoch": 0.29388, + "grad_norm": 1.336837946582919, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 29388 + }, + { + "epoch": 0.29389, + "grad_norm": 0.9143810942888401, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 29389 + }, + { + "epoch": 0.2939, + "grad_norm": 1.0032645164265728, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 29390 + }, + { + "epoch": 0.29391, + "grad_norm": 1.1267851011382746, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 29391 + }, + { + "epoch": 0.29392, + "grad_norm": 1.0195612491522101, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 29392 + }, + { + "epoch": 0.29393, + "grad_norm": 1.1807664276300849, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 29393 + }, + { + "epoch": 0.29394, + "grad_norm": 0.9841012451285469, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 29394 + }, + { + "epoch": 0.29395, + "grad_norm": 1.1779291633889097, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 29395 + }, + { + "epoch": 0.29396, + "grad_norm": 1.1322939884315601, + "learning_rate": 0.003, + "loss": 4.037, + "step": 29396 + }, + { + "epoch": 0.29397, + "grad_norm": 1.0920089040826495, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 29397 + }, + { + "epoch": 0.29398, + "grad_norm": 1.1875591505461465, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 29398 + }, + { + "epoch": 0.29399, + "grad_norm": 0.8669898056891145, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 29399 + }, + { + "epoch": 0.294, + "grad_norm": 0.8727008853355489, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 29400 + }, + { + "epoch": 0.29401, + "grad_norm": 1.0027755311988265, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 29401 + }, + { + "epoch": 0.29402, + "grad_norm": 1.232720538025115, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 29402 + }, + { + "epoch": 0.29403, + "grad_norm": 0.9419278725049276, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 29403 + }, + { + "epoch": 0.29404, + "grad_norm": 1.1242477418630457, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 29404 + }, + { + "epoch": 0.29405, + "grad_norm": 1.1350732843235538, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 29405 + }, + { + "epoch": 0.29406, + "grad_norm": 1.1216964320596001, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 29406 + }, + { + "epoch": 0.29407, + "grad_norm": 1.0254664536453495, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 29407 + }, + { + "epoch": 0.29408, + "grad_norm": 1.1569497720273554, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 29408 + }, + { + "epoch": 0.29409, + "grad_norm": 1.0718124157061633, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 29409 + }, + { + "epoch": 0.2941, + "grad_norm": 1.0089034907703802, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 29410 + }, + { + "epoch": 0.29411, + "grad_norm": 1.2183941845218196, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 29411 + }, + { + "epoch": 0.29412, + "grad_norm": 1.057039627040901, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 29412 + }, + { + "epoch": 0.29413, + "grad_norm": 1.0220271456860401, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 29413 + }, + { + "epoch": 0.29414, + "grad_norm": 0.9914788775073956, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 29414 + }, + { + "epoch": 0.29415, + "grad_norm": 1.0976456835133206, + "learning_rate": 0.003, + "loss": 4.041, + "step": 29415 + }, + { + "epoch": 0.29416, + "grad_norm": 1.1200516387047847, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 29416 + }, + { + "epoch": 0.29417, + "grad_norm": 1.0500191506004182, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 29417 + }, + { + "epoch": 0.29418, + "grad_norm": 1.1398337548701223, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 29418 + }, + { + "epoch": 0.29419, + "grad_norm": 1.1115481385484618, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 29419 + }, + { + "epoch": 0.2942, + "grad_norm": 1.1740174376591284, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 29420 + }, + { + "epoch": 0.29421, + "grad_norm": 0.9437266844921209, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 29421 + }, + { + "epoch": 0.29422, + "grad_norm": 1.234971429918723, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 29422 + }, + { + "epoch": 0.29423, + "grad_norm": 1.1640766055820222, + "learning_rate": 0.003, + "loss": 4.042, + "step": 29423 + }, + { + "epoch": 0.29424, + "grad_norm": 1.076140508718708, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 29424 + }, + { + "epoch": 0.29425, + "grad_norm": 0.9794285555331601, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 29425 + }, + { + "epoch": 0.29426, + "grad_norm": 1.048798330512327, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 29426 + }, + { + "epoch": 0.29427, + "grad_norm": 1.2580890437421608, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 29427 + }, + { + "epoch": 0.29428, + "grad_norm": 0.8015830640323922, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 29428 + }, + { + "epoch": 0.29429, + "grad_norm": 0.83344310890065, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 29429 + }, + { + "epoch": 0.2943, + "grad_norm": 0.9299562028571626, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 29430 + }, + { + "epoch": 0.29431, + "grad_norm": 1.029545071931151, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 29431 + }, + { + "epoch": 0.29432, + "grad_norm": 1.2413337042159012, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 29432 + }, + { + "epoch": 0.29433, + "grad_norm": 0.9360446342182975, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 29433 + }, + { + "epoch": 0.29434, + "grad_norm": 0.848488706010888, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 29434 + }, + { + "epoch": 0.29435, + "grad_norm": 0.7565538331100565, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 29435 + }, + { + "epoch": 0.29436, + "grad_norm": 0.8295430844893986, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 29436 + }, + { + "epoch": 0.29437, + "grad_norm": 1.0922253687430268, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 29437 + }, + { + "epoch": 0.29438, + "grad_norm": 1.3910870970014741, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 29438 + }, + { + "epoch": 0.29439, + "grad_norm": 1.043059408647494, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 29439 + }, + { + "epoch": 0.2944, + "grad_norm": 1.0981960285775894, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 29440 + }, + { + "epoch": 0.29441, + "grad_norm": 1.099931566728234, + "learning_rate": 0.003, + "loss": 4.042, + "step": 29441 + }, + { + "epoch": 0.29442, + "grad_norm": 0.9621934461896732, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 29442 + }, + { + "epoch": 0.29443, + "grad_norm": 1.1078331317094439, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 29443 + }, + { + "epoch": 0.29444, + "grad_norm": 1.1688203236538843, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 29444 + }, + { + "epoch": 0.29445, + "grad_norm": 1.0724363857203463, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 29445 + }, + { + "epoch": 0.29446, + "grad_norm": 1.0988925252408706, + "learning_rate": 0.003, + "loss": 4.042, + "step": 29446 + }, + { + "epoch": 0.29447, + "grad_norm": 1.006451421703592, + "learning_rate": 0.003, + "loss": 4.025, + "step": 29447 + }, + { + "epoch": 0.29448, + "grad_norm": 1.0730271991811955, + "learning_rate": 0.003, + "loss": 4.036, + "step": 29448 + }, + { + "epoch": 0.29449, + "grad_norm": 0.9907347322983976, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 29449 + }, + { + "epoch": 0.2945, + "grad_norm": 1.0308842009996295, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 29450 + }, + { + "epoch": 0.29451, + "grad_norm": 1.300166242297691, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 29451 + }, + { + "epoch": 0.29452, + "grad_norm": 0.8656276209579702, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 29452 + }, + { + "epoch": 0.29453, + "grad_norm": 0.9445419352449046, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 29453 + }, + { + "epoch": 0.29454, + "grad_norm": 1.2557354082184222, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 29454 + }, + { + "epoch": 0.29455, + "grad_norm": 1.2395649775375326, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 29455 + }, + { + "epoch": 0.29456, + "grad_norm": 0.8939253763999209, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 29456 + }, + { + "epoch": 0.29457, + "grad_norm": 0.9367722958056519, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 29457 + }, + { + "epoch": 0.29458, + "grad_norm": 1.1713017660604697, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 29458 + }, + { + "epoch": 0.29459, + "grad_norm": 1.3046524122287408, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 29459 + }, + { + "epoch": 0.2946, + "grad_norm": 0.917754448359289, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 29460 + }, + { + "epoch": 0.29461, + "grad_norm": 0.8881608248133053, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 29461 + }, + { + "epoch": 0.29462, + "grad_norm": 1.0587279261660638, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 29462 + }, + { + "epoch": 0.29463, + "grad_norm": 1.2291970431121322, + "learning_rate": 0.003, + "loss": 4.052, + "step": 29463 + }, + { + "epoch": 0.29464, + "grad_norm": 1.2553161322649888, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 29464 + }, + { + "epoch": 0.29465, + "grad_norm": 1.0899709397842123, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 29465 + }, + { + "epoch": 0.29466, + "grad_norm": 1.1827824420003499, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 29466 + }, + { + "epoch": 0.29467, + "grad_norm": 0.8360804904227456, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 29467 + }, + { + "epoch": 0.29468, + "grad_norm": 0.9107261107796969, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 29468 + }, + { + "epoch": 0.29469, + "grad_norm": 1.0323632317321514, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 29469 + }, + { + "epoch": 0.2947, + "grad_norm": 1.2675068423600635, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 29470 + }, + { + "epoch": 0.29471, + "grad_norm": 1.137241960136992, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 29471 + }, + { + "epoch": 0.29472, + "grad_norm": 1.0542426794068667, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 29472 + }, + { + "epoch": 0.29473, + "grad_norm": 0.9352505603504143, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 29473 + }, + { + "epoch": 0.29474, + "grad_norm": 1.05512169363281, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 29474 + }, + { + "epoch": 0.29475, + "grad_norm": 1.0668391628845746, + "learning_rate": 0.003, + "loss": 4.054, + "step": 29475 + }, + { + "epoch": 0.29476, + "grad_norm": 1.0367333087462742, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 29476 + }, + { + "epoch": 0.29477, + "grad_norm": 1.1517505296870059, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 29477 + }, + { + "epoch": 0.29478, + "grad_norm": 1.1539157018815704, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 29478 + }, + { + "epoch": 0.29479, + "grad_norm": 1.6730712572956759, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 29479 + }, + { + "epoch": 0.2948, + "grad_norm": 0.8345980753370956, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 29480 + }, + { + "epoch": 0.29481, + "grad_norm": 0.9803328981987058, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 29481 + }, + { + "epoch": 0.29482, + "grad_norm": 1.5490196224714456, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 29482 + }, + { + "epoch": 0.29483, + "grad_norm": 0.981576280814537, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 29483 + }, + { + "epoch": 0.29484, + "grad_norm": 1.3510777812712902, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 29484 + }, + { + "epoch": 0.29485, + "grad_norm": 0.8963531313249701, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 29485 + }, + { + "epoch": 0.29486, + "grad_norm": 0.9404939039053283, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 29486 + }, + { + "epoch": 0.29487, + "grad_norm": 1.2000375159017478, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 29487 + }, + { + "epoch": 0.29488, + "grad_norm": 0.9685277938153033, + "learning_rate": 0.003, + "loss": 4.043, + "step": 29488 + }, + { + "epoch": 0.29489, + "grad_norm": 0.843862327125721, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 29489 + }, + { + "epoch": 0.2949, + "grad_norm": 0.9231720218438576, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 29490 + }, + { + "epoch": 0.29491, + "grad_norm": 0.9453580080105386, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 29491 + }, + { + "epoch": 0.29492, + "grad_norm": 1.0985552914020111, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 29492 + }, + { + "epoch": 0.29493, + "grad_norm": 1.1932991590250015, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 29493 + }, + { + "epoch": 0.29494, + "grad_norm": 1.2054795322979739, + "learning_rate": 0.003, + "loss": 4.0991, + "step": 29494 + }, + { + "epoch": 0.29495, + "grad_norm": 1.1427408286123426, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 29495 + }, + { + "epoch": 0.29496, + "grad_norm": 1.057281733628086, + "learning_rate": 0.003, + "loss": 4.051, + "step": 29496 + }, + { + "epoch": 0.29497, + "grad_norm": 1.0500502886601926, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 29497 + }, + { + "epoch": 0.29498, + "grad_norm": 1.3890001502031675, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 29498 + }, + { + "epoch": 0.29499, + "grad_norm": 1.2754569903564925, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 29499 + }, + { + "epoch": 0.295, + "grad_norm": 0.9555444734878511, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 29500 + }, + { + "epoch": 0.29501, + "grad_norm": 1.081272604983355, + "learning_rate": 0.003, + "loss": 4.034, + "step": 29501 + }, + { + "epoch": 0.29502, + "grad_norm": 1.042398547383382, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 29502 + }, + { + "epoch": 0.29503, + "grad_norm": 1.0001043297047605, + "learning_rate": 0.003, + "loss": 4.081, + "step": 29503 + }, + { + "epoch": 0.29504, + "grad_norm": 0.9650304127890678, + "learning_rate": 0.003, + "loss": 4.046, + "step": 29504 + }, + { + "epoch": 0.29505, + "grad_norm": 0.9537702048533052, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 29505 + }, + { + "epoch": 0.29506, + "grad_norm": 1.018801074818957, + "learning_rate": 0.003, + "loss": 4.027, + "step": 29506 + }, + { + "epoch": 0.29507, + "grad_norm": 1.0168821893757625, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 29507 + }, + { + "epoch": 0.29508, + "grad_norm": 1.0082071767787757, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 29508 + }, + { + "epoch": 0.29509, + "grad_norm": 1.0354158069964332, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 29509 + }, + { + "epoch": 0.2951, + "grad_norm": 1.070670230671644, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 29510 + }, + { + "epoch": 0.29511, + "grad_norm": 1.2564363416057942, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 29511 + }, + { + "epoch": 0.29512, + "grad_norm": 1.0443067005739164, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 29512 + }, + { + "epoch": 0.29513, + "grad_norm": 1.3871545653881443, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 29513 + }, + { + "epoch": 0.29514, + "grad_norm": 1.0902508703036586, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 29514 + }, + { + "epoch": 0.29515, + "grad_norm": 0.958320867667851, + "learning_rate": 0.003, + "loss": 4.031, + "step": 29515 + }, + { + "epoch": 0.29516, + "grad_norm": 0.9400414199245267, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 29516 + }, + { + "epoch": 0.29517, + "grad_norm": 0.9953859865763187, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 29517 + }, + { + "epoch": 0.29518, + "grad_norm": 1.2805241312962126, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 29518 + }, + { + "epoch": 0.29519, + "grad_norm": 1.0907302012184323, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 29519 + }, + { + "epoch": 0.2952, + "grad_norm": 1.07321139609848, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 29520 + }, + { + "epoch": 0.29521, + "grad_norm": 1.0058793029802315, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 29521 + }, + { + "epoch": 0.29522, + "grad_norm": 1.1946775127306504, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 29522 + }, + { + "epoch": 0.29523, + "grad_norm": 1.0444091993440967, + "learning_rate": 0.003, + "loss": 4.062, + "step": 29523 + }, + { + "epoch": 0.29524, + "grad_norm": 0.9868573485788861, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 29524 + }, + { + "epoch": 0.29525, + "grad_norm": 1.1115449087184885, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 29525 + }, + { + "epoch": 0.29526, + "grad_norm": 1.1309568196729414, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 29526 + }, + { + "epoch": 0.29527, + "grad_norm": 0.929237620066838, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 29527 + }, + { + "epoch": 0.29528, + "grad_norm": 1.26793862611713, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 29528 + }, + { + "epoch": 0.29529, + "grad_norm": 1.0095050850004816, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 29529 + }, + { + "epoch": 0.2953, + "grad_norm": 1.0990812663364813, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 29530 + }, + { + "epoch": 0.29531, + "grad_norm": 1.0149673226994038, + "learning_rate": 0.003, + "loss": 4.047, + "step": 29531 + }, + { + "epoch": 0.29532, + "grad_norm": 1.0148986301023517, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 29532 + }, + { + "epoch": 0.29533, + "grad_norm": 1.0231693803869144, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 29533 + }, + { + "epoch": 0.29534, + "grad_norm": 1.1212453838293623, + "learning_rate": 0.003, + "loss": 4.042, + "step": 29534 + }, + { + "epoch": 0.29535, + "grad_norm": 0.9576912115722782, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 29535 + }, + { + "epoch": 0.29536, + "grad_norm": 0.9371887502515966, + "learning_rate": 0.003, + "loss": 4.002, + "step": 29536 + }, + { + "epoch": 0.29537, + "grad_norm": 0.9795205963560573, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 29537 + }, + { + "epoch": 0.29538, + "grad_norm": 1.2985033146855738, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 29538 + }, + { + "epoch": 0.29539, + "grad_norm": 0.9958353690162672, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 29539 + }, + { + "epoch": 0.2954, + "grad_norm": 1.1884474779756111, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 29540 + }, + { + "epoch": 0.29541, + "grad_norm": 1.1425293187658832, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 29541 + }, + { + "epoch": 0.29542, + "grad_norm": 1.2237945928905727, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 29542 + }, + { + "epoch": 0.29543, + "grad_norm": 1.2949751522616573, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 29543 + }, + { + "epoch": 0.29544, + "grad_norm": 0.9781743511840189, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 29544 + }, + { + "epoch": 0.29545, + "grad_norm": 1.0479583889913981, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 29545 + }, + { + "epoch": 0.29546, + "grad_norm": 1.2755204285059218, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 29546 + }, + { + "epoch": 0.29547, + "grad_norm": 1.0600627709517303, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 29547 + }, + { + "epoch": 0.29548, + "grad_norm": 1.0792276028528855, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 29548 + }, + { + "epoch": 0.29549, + "grad_norm": 0.9252133429813597, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 29549 + }, + { + "epoch": 0.2955, + "grad_norm": 1.055165152388345, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 29550 + }, + { + "epoch": 0.29551, + "grad_norm": 1.1615160583826967, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 29551 + }, + { + "epoch": 0.29552, + "grad_norm": 1.0073531428068907, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 29552 + }, + { + "epoch": 0.29553, + "grad_norm": 1.126472528631102, + "learning_rate": 0.003, + "loss": 4.06, + "step": 29553 + }, + { + "epoch": 0.29554, + "grad_norm": 1.0989485987086196, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 29554 + }, + { + "epoch": 0.29555, + "grad_norm": 1.0282994134612213, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 29555 + }, + { + "epoch": 0.29556, + "grad_norm": 1.1579997127402686, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 29556 + }, + { + "epoch": 0.29557, + "grad_norm": 1.1058299585634497, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 29557 + }, + { + "epoch": 0.29558, + "grad_norm": 1.049121118112991, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 29558 + }, + { + "epoch": 0.29559, + "grad_norm": 0.9176175247826909, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 29559 + }, + { + "epoch": 0.2956, + "grad_norm": 0.8290225575375315, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 29560 + }, + { + "epoch": 0.29561, + "grad_norm": 0.8862256691507348, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 29561 + }, + { + "epoch": 0.29562, + "grad_norm": 0.9710469648254914, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 29562 + }, + { + "epoch": 0.29563, + "grad_norm": 0.9580858711897061, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 29563 + }, + { + "epoch": 0.29564, + "grad_norm": 1.1219354176585237, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 29564 + }, + { + "epoch": 0.29565, + "grad_norm": 1.0753873429137002, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 29565 + }, + { + "epoch": 0.29566, + "grad_norm": 1.2638922741465894, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 29566 + }, + { + "epoch": 0.29567, + "grad_norm": 0.972541228548322, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 29567 + }, + { + "epoch": 0.29568, + "grad_norm": 1.4156726518775355, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 29568 + }, + { + "epoch": 0.29569, + "grad_norm": 1.039518558495795, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 29569 + }, + { + "epoch": 0.2957, + "grad_norm": 1.0881150966580624, + "learning_rate": 0.003, + "loss": 3.994, + "step": 29570 + }, + { + "epoch": 0.29571, + "grad_norm": 1.232013077542583, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 29571 + }, + { + "epoch": 0.29572, + "grad_norm": 1.0811580632222053, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 29572 + }, + { + "epoch": 0.29573, + "grad_norm": 1.2780465892813446, + "learning_rate": 0.003, + "loss": 4.047, + "step": 29573 + }, + { + "epoch": 0.29574, + "grad_norm": 1.0892957134436319, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 29574 + }, + { + "epoch": 0.29575, + "grad_norm": 1.124886428769311, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 29575 + }, + { + "epoch": 0.29576, + "grad_norm": 0.9244527384256437, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 29576 + }, + { + "epoch": 0.29577, + "grad_norm": 0.9289449460210788, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 29577 + }, + { + "epoch": 0.29578, + "grad_norm": 1.1245613498294535, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 29578 + }, + { + "epoch": 0.29579, + "grad_norm": 1.1008110589060824, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 29579 + }, + { + "epoch": 0.2958, + "grad_norm": 1.0194317937097388, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 29580 + }, + { + "epoch": 0.29581, + "grad_norm": 1.1494089619232304, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 29581 + }, + { + "epoch": 0.29582, + "grad_norm": 1.3024081019759028, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 29582 + }, + { + "epoch": 0.29583, + "grad_norm": 0.9384977509851888, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 29583 + }, + { + "epoch": 0.29584, + "grad_norm": 0.9942781145671697, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 29584 + }, + { + "epoch": 0.29585, + "grad_norm": 1.1121333420251487, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 29585 + }, + { + "epoch": 0.29586, + "grad_norm": 0.9679392478628098, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 29586 + }, + { + "epoch": 0.29587, + "grad_norm": 1.0319722498502324, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 29587 + }, + { + "epoch": 0.29588, + "grad_norm": 1.0126667024852742, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 29588 + }, + { + "epoch": 0.29589, + "grad_norm": 1.1826125978538222, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 29589 + }, + { + "epoch": 0.2959, + "grad_norm": 0.8135914945848276, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 29590 + }, + { + "epoch": 0.29591, + "grad_norm": 0.9022469036165698, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 29591 + }, + { + "epoch": 0.29592, + "grad_norm": 1.006721042460114, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 29592 + }, + { + "epoch": 0.29593, + "grad_norm": 1.1070734444627492, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 29593 + }, + { + "epoch": 0.29594, + "grad_norm": 1.0268414889430229, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 29594 + }, + { + "epoch": 0.29595, + "grad_norm": 1.1956144435200635, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 29595 + }, + { + "epoch": 0.29596, + "grad_norm": 1.1957095521985925, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 29596 + }, + { + "epoch": 0.29597, + "grad_norm": 0.9651298815408665, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 29597 + }, + { + "epoch": 0.29598, + "grad_norm": 1.1217646212766585, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 29598 + }, + { + "epoch": 0.29599, + "grad_norm": 1.0293861623616827, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 29599 + }, + { + "epoch": 0.296, + "grad_norm": 1.057558684910815, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 29600 + }, + { + "epoch": 0.29601, + "grad_norm": 1.0013515141068638, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 29601 + }, + { + "epoch": 0.29602, + "grad_norm": 1.1725983469084589, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 29602 + }, + { + "epoch": 0.29603, + "grad_norm": 0.93431025051427, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 29603 + }, + { + "epoch": 0.29604, + "grad_norm": 1.0041843574230402, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 29604 + }, + { + "epoch": 0.29605, + "grad_norm": 1.2806578516777452, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 29605 + }, + { + "epoch": 0.29606, + "grad_norm": 0.885370122197077, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 29606 + }, + { + "epoch": 0.29607, + "grad_norm": 0.9531846651176699, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 29607 + }, + { + "epoch": 0.29608, + "grad_norm": 1.0306962324000666, + "learning_rate": 0.003, + "loss": 4.018, + "step": 29608 + }, + { + "epoch": 0.29609, + "grad_norm": 0.9397281380224946, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 29609 + }, + { + "epoch": 0.2961, + "grad_norm": 1.008938767721999, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 29610 + }, + { + "epoch": 0.29611, + "grad_norm": 1.2531734140716448, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 29611 + }, + { + "epoch": 0.29612, + "grad_norm": 0.9969500520698812, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 29612 + }, + { + "epoch": 0.29613, + "grad_norm": 1.2983388904227784, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 29613 + }, + { + "epoch": 0.29614, + "grad_norm": 0.9492474005442227, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 29614 + }, + { + "epoch": 0.29615, + "grad_norm": 0.9722869950086285, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 29615 + }, + { + "epoch": 0.29616, + "grad_norm": 1.1159748274647991, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 29616 + }, + { + "epoch": 0.29617, + "grad_norm": 1.0045861888183605, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 29617 + }, + { + "epoch": 0.29618, + "grad_norm": 1.1985053368160041, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 29618 + }, + { + "epoch": 0.29619, + "grad_norm": 1.0763185426741175, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 29619 + }, + { + "epoch": 0.2962, + "grad_norm": 1.1317351103399864, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 29620 + }, + { + "epoch": 0.29621, + "grad_norm": 0.8037656301986305, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 29621 + }, + { + "epoch": 0.29622, + "grad_norm": 1.0706854534649248, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 29622 + }, + { + "epoch": 0.29623, + "grad_norm": 1.1671747683061793, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 29623 + }, + { + "epoch": 0.29624, + "grad_norm": 1.1400806265922683, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 29624 + }, + { + "epoch": 0.29625, + "grad_norm": 1.1113426334613954, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 29625 + }, + { + "epoch": 0.29626, + "grad_norm": 1.0700729348841342, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 29626 + }, + { + "epoch": 0.29627, + "grad_norm": 1.0131203754373408, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 29627 + }, + { + "epoch": 0.29628, + "grad_norm": 1.093282571708716, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 29628 + }, + { + "epoch": 0.29629, + "grad_norm": 1.0998358731964513, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 29629 + }, + { + "epoch": 0.2963, + "grad_norm": 1.1323235142217865, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 29630 + }, + { + "epoch": 0.29631, + "grad_norm": 1.0408327587771868, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 29631 + }, + { + "epoch": 0.29632, + "grad_norm": 0.9822547574407193, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 29632 + }, + { + "epoch": 0.29633, + "grad_norm": 1.20109053133986, + "learning_rate": 0.003, + "loss": 4.051, + "step": 29633 + }, + { + "epoch": 0.29634, + "grad_norm": 0.8132360212982945, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 29634 + }, + { + "epoch": 0.29635, + "grad_norm": 0.8479643302689079, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 29635 + }, + { + "epoch": 0.29636, + "grad_norm": 1.0231350665343872, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 29636 + }, + { + "epoch": 0.29637, + "grad_norm": 1.2015436414343987, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 29637 + }, + { + "epoch": 0.29638, + "grad_norm": 1.0663851071871824, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 29638 + }, + { + "epoch": 0.29639, + "grad_norm": 1.0487422523353145, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 29639 + }, + { + "epoch": 0.2964, + "grad_norm": 1.160288109833599, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 29640 + }, + { + "epoch": 0.29641, + "grad_norm": 1.5372039860502698, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 29641 + }, + { + "epoch": 0.29642, + "grad_norm": 0.8550186666017944, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 29642 + }, + { + "epoch": 0.29643, + "grad_norm": 0.8605071251365365, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 29643 + }, + { + "epoch": 0.29644, + "grad_norm": 1.1396615701922364, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 29644 + }, + { + "epoch": 0.29645, + "grad_norm": 1.15449512714996, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 29645 + }, + { + "epoch": 0.29646, + "grad_norm": 0.9369286356951559, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 29646 + }, + { + "epoch": 0.29647, + "grad_norm": 1.0660131339459993, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 29647 + }, + { + "epoch": 0.29648, + "grad_norm": 1.0753463739843694, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 29648 + }, + { + "epoch": 0.29649, + "grad_norm": 1.0687584857800776, + "learning_rate": 0.003, + "loss": 4.062, + "step": 29649 + }, + { + "epoch": 0.2965, + "grad_norm": 1.1641651980342176, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 29650 + }, + { + "epoch": 0.29651, + "grad_norm": 1.0467458097227886, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 29651 + }, + { + "epoch": 0.29652, + "grad_norm": 1.0823493973291856, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 29652 + }, + { + "epoch": 0.29653, + "grad_norm": 1.0141039750489016, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 29653 + }, + { + "epoch": 0.29654, + "grad_norm": 1.052203535046186, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 29654 + }, + { + "epoch": 0.29655, + "grad_norm": 1.205824212060383, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 29655 + }, + { + "epoch": 0.29656, + "grad_norm": 1.051562873653984, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 29656 + }, + { + "epoch": 0.29657, + "grad_norm": 0.9577608323946376, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 29657 + }, + { + "epoch": 0.29658, + "grad_norm": 0.7695167362117536, + "learning_rate": 0.003, + "loss": 4.039, + "step": 29658 + }, + { + "epoch": 0.29659, + "grad_norm": 0.8477633993120015, + "learning_rate": 0.003, + "loss": 4.034, + "step": 29659 + }, + { + "epoch": 0.2966, + "grad_norm": 1.0944668660692365, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 29660 + }, + { + "epoch": 0.29661, + "grad_norm": 1.1923025525497657, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 29661 + }, + { + "epoch": 0.29662, + "grad_norm": 1.0427229116147547, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 29662 + }, + { + "epoch": 0.29663, + "grad_norm": 1.124619099829402, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 29663 + }, + { + "epoch": 0.29664, + "grad_norm": 0.9170500018217663, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 29664 + }, + { + "epoch": 0.29665, + "grad_norm": 0.9606865783686139, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 29665 + }, + { + "epoch": 0.29666, + "grad_norm": 1.0822115012083533, + "learning_rate": 0.003, + "loss": 4.05, + "step": 29666 + }, + { + "epoch": 0.29667, + "grad_norm": 1.0520774398950143, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 29667 + }, + { + "epoch": 0.29668, + "grad_norm": 1.3619152308171105, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 29668 + }, + { + "epoch": 0.29669, + "grad_norm": 0.9793323812939678, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 29669 + }, + { + "epoch": 0.2967, + "grad_norm": 1.1329131562312789, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 29670 + }, + { + "epoch": 0.29671, + "grad_norm": 1.0735059502576223, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 29671 + }, + { + "epoch": 0.29672, + "grad_norm": 1.1156925210981494, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 29672 + }, + { + "epoch": 0.29673, + "grad_norm": 1.0197095864999075, + "learning_rate": 0.003, + "loss": 4.065, + "step": 29673 + }, + { + "epoch": 0.29674, + "grad_norm": 1.2361632958202768, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 29674 + }, + { + "epoch": 0.29675, + "grad_norm": 1.16014195199349, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 29675 + }, + { + "epoch": 0.29676, + "grad_norm": 1.0500083181020026, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 29676 + }, + { + "epoch": 0.29677, + "grad_norm": 1.2709573288708662, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 29677 + }, + { + "epoch": 0.29678, + "grad_norm": 0.9047926193435155, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 29678 + }, + { + "epoch": 0.29679, + "grad_norm": 1.179631366940322, + "learning_rate": 0.003, + "loss": 4.02, + "step": 29679 + }, + { + "epoch": 0.2968, + "grad_norm": 0.9689768641638936, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 29680 + }, + { + "epoch": 0.29681, + "grad_norm": 1.0936860890490652, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 29681 + }, + { + "epoch": 0.29682, + "grad_norm": 1.251381788796184, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 29682 + }, + { + "epoch": 0.29683, + "grad_norm": 0.9583136355054199, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 29683 + }, + { + "epoch": 0.29684, + "grad_norm": 1.0941588320946476, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 29684 + }, + { + "epoch": 0.29685, + "grad_norm": 0.9152982847245137, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 29685 + }, + { + "epoch": 0.29686, + "grad_norm": 0.9862872207354623, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 29686 + }, + { + "epoch": 0.29687, + "grad_norm": 1.0899020339109022, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 29687 + }, + { + "epoch": 0.29688, + "grad_norm": 0.9364220954873899, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 29688 + }, + { + "epoch": 0.29689, + "grad_norm": 0.9786157589896375, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 29689 + }, + { + "epoch": 0.2969, + "grad_norm": 1.1443348844256012, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 29690 + }, + { + "epoch": 0.29691, + "grad_norm": 1.0185516942338873, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 29691 + }, + { + "epoch": 0.29692, + "grad_norm": 0.9319836990580325, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 29692 + }, + { + "epoch": 0.29693, + "grad_norm": 0.8890743346334762, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 29693 + }, + { + "epoch": 0.29694, + "grad_norm": 1.0582128934075186, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 29694 + }, + { + "epoch": 0.29695, + "grad_norm": 1.410593412735711, + "learning_rate": 0.003, + "loss": 4.054, + "step": 29695 + }, + { + "epoch": 0.29696, + "grad_norm": 0.9941280066229681, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 29696 + }, + { + "epoch": 0.29697, + "grad_norm": 1.0650098281913336, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 29697 + }, + { + "epoch": 0.29698, + "grad_norm": 1.099663115046292, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 29698 + }, + { + "epoch": 0.29699, + "grad_norm": 1.014333277089527, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 29699 + }, + { + "epoch": 0.297, + "grad_norm": 1.1850947033570112, + "learning_rate": 0.003, + "loss": 4.056, + "step": 29700 + }, + { + "epoch": 0.29701, + "grad_norm": 0.9661356590872284, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 29701 + }, + { + "epoch": 0.29702, + "grad_norm": 1.2339462377510262, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 29702 + }, + { + "epoch": 0.29703, + "grad_norm": 0.9454772767712343, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 29703 + }, + { + "epoch": 0.29704, + "grad_norm": 1.0426758471034452, + "learning_rate": 0.003, + "loss": 4.0888, + "step": 29704 + }, + { + "epoch": 0.29705, + "grad_norm": 1.2136060985990187, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 29705 + }, + { + "epoch": 0.29706, + "grad_norm": 1.2944040622731057, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 29706 + }, + { + "epoch": 0.29707, + "grad_norm": 0.9819247992482449, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 29707 + }, + { + "epoch": 0.29708, + "grad_norm": 1.1074411639523707, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 29708 + }, + { + "epoch": 0.29709, + "grad_norm": 1.1279402676703782, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 29709 + }, + { + "epoch": 0.2971, + "grad_norm": 1.0282463085525755, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 29710 + }, + { + "epoch": 0.29711, + "grad_norm": 0.9556622966537149, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 29711 + }, + { + "epoch": 0.29712, + "grad_norm": 0.9713311835193971, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 29712 + }, + { + "epoch": 0.29713, + "grad_norm": 1.087367296212545, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 29713 + }, + { + "epoch": 0.29714, + "grad_norm": 1.131177436050584, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 29714 + }, + { + "epoch": 0.29715, + "grad_norm": 1.0249308557628172, + "learning_rate": 0.003, + "loss": 4.035, + "step": 29715 + }, + { + "epoch": 0.29716, + "grad_norm": 1.292260434365323, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 29716 + }, + { + "epoch": 0.29717, + "grad_norm": 0.9525486117595119, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 29717 + }, + { + "epoch": 0.29718, + "grad_norm": 1.1006818761476522, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 29718 + }, + { + "epoch": 0.29719, + "grad_norm": 0.9756730930930695, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 29719 + }, + { + "epoch": 0.2972, + "grad_norm": 0.9700810429089282, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 29720 + }, + { + "epoch": 0.29721, + "grad_norm": 1.1899848988113109, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 29721 + }, + { + "epoch": 0.29722, + "grad_norm": 1.022719970207354, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 29722 + }, + { + "epoch": 0.29723, + "grad_norm": 1.3048430768326627, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 29723 + }, + { + "epoch": 0.29724, + "grad_norm": 1.0182550325428976, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 29724 + }, + { + "epoch": 0.29725, + "grad_norm": 1.0074681805784815, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 29725 + }, + { + "epoch": 0.29726, + "grad_norm": 1.2219577897357348, + "learning_rate": 0.003, + "loss": 4.06, + "step": 29726 + }, + { + "epoch": 0.29727, + "grad_norm": 0.9455393076662831, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 29727 + }, + { + "epoch": 0.29728, + "grad_norm": 0.9530938532533614, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 29728 + }, + { + "epoch": 0.29729, + "grad_norm": 1.041203725754412, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 29729 + }, + { + "epoch": 0.2973, + "grad_norm": 1.1260230837291136, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 29730 + }, + { + "epoch": 0.29731, + "grad_norm": 0.9673501917355647, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 29731 + }, + { + "epoch": 0.29732, + "grad_norm": 1.0434245239811033, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 29732 + }, + { + "epoch": 0.29733, + "grad_norm": 0.9523873440617202, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 29733 + }, + { + "epoch": 0.29734, + "grad_norm": 1.0326521383891025, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 29734 + }, + { + "epoch": 0.29735, + "grad_norm": 1.2011617731545703, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 29735 + }, + { + "epoch": 0.29736, + "grad_norm": 1.0177904473219916, + "learning_rate": 0.003, + "loss": 4.082, + "step": 29736 + }, + { + "epoch": 0.29737, + "grad_norm": 1.023763824097497, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 29737 + }, + { + "epoch": 0.29738, + "grad_norm": 1.198301028284381, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 29738 + }, + { + "epoch": 0.29739, + "grad_norm": 0.9183556860356475, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 29739 + }, + { + "epoch": 0.2974, + "grad_norm": 1.179782146473203, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 29740 + }, + { + "epoch": 0.29741, + "grad_norm": 1.18825379223564, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 29741 + }, + { + "epoch": 0.29742, + "grad_norm": 1.1263462159151465, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 29742 + }, + { + "epoch": 0.29743, + "grad_norm": 1.0919207049064696, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 29743 + }, + { + "epoch": 0.29744, + "grad_norm": 1.1002723821254925, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 29744 + }, + { + "epoch": 0.29745, + "grad_norm": 0.9654777499427724, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 29745 + }, + { + "epoch": 0.29746, + "grad_norm": 1.0168682027880769, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 29746 + }, + { + "epoch": 0.29747, + "grad_norm": 1.2045333572392773, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 29747 + }, + { + "epoch": 0.29748, + "grad_norm": 1.02779784853404, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 29748 + }, + { + "epoch": 0.29749, + "grad_norm": 1.0516996095348141, + "learning_rate": 0.003, + "loss": 4.052, + "step": 29749 + }, + { + "epoch": 0.2975, + "grad_norm": 1.2504622417363085, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 29750 + }, + { + "epoch": 0.29751, + "grad_norm": 1.0500514166087314, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 29751 + }, + { + "epoch": 0.29752, + "grad_norm": 1.1883382295214868, + "learning_rate": 0.003, + "loss": 4.027, + "step": 29752 + }, + { + "epoch": 0.29753, + "grad_norm": 0.966796641798219, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 29753 + }, + { + "epoch": 0.29754, + "grad_norm": 1.136013910627941, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 29754 + }, + { + "epoch": 0.29755, + "grad_norm": 0.9518676042898114, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 29755 + }, + { + "epoch": 0.29756, + "grad_norm": 1.0151336523418852, + "learning_rate": 0.003, + "loss": 4.039, + "step": 29756 + }, + { + "epoch": 0.29757, + "grad_norm": 1.0956810304806728, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 29757 + }, + { + "epoch": 0.29758, + "grad_norm": 1.1281177286153383, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 29758 + }, + { + "epoch": 0.29759, + "grad_norm": 1.227096284093403, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 29759 + }, + { + "epoch": 0.2976, + "grad_norm": 0.8964270373256443, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 29760 + }, + { + "epoch": 0.29761, + "grad_norm": 0.876178044057259, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 29761 + }, + { + "epoch": 0.29762, + "grad_norm": 0.9450747385877432, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 29762 + }, + { + "epoch": 0.29763, + "grad_norm": 1.1405429268165905, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 29763 + }, + { + "epoch": 0.29764, + "grad_norm": 1.0145826999815468, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 29764 + }, + { + "epoch": 0.29765, + "grad_norm": 1.2686070420213529, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 29765 + }, + { + "epoch": 0.29766, + "grad_norm": 0.8065223070191168, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 29766 + }, + { + "epoch": 0.29767, + "grad_norm": 0.7933375043857676, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 29767 + }, + { + "epoch": 0.29768, + "grad_norm": 0.9072066717851851, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 29768 + }, + { + "epoch": 0.29769, + "grad_norm": 1.0923652328292612, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 29769 + }, + { + "epoch": 0.2977, + "grad_norm": 1.0476985681610214, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 29770 + }, + { + "epoch": 0.29771, + "grad_norm": 1.0467956766887252, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 29771 + }, + { + "epoch": 0.29772, + "grad_norm": 1.244655563518728, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 29772 + }, + { + "epoch": 0.29773, + "grad_norm": 1.1766361968952461, + "learning_rate": 0.003, + "loss": 4.052, + "step": 29773 + }, + { + "epoch": 0.29774, + "grad_norm": 0.8787358127516284, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 29774 + }, + { + "epoch": 0.29775, + "grad_norm": 0.7860517458218924, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 29775 + }, + { + "epoch": 0.29776, + "grad_norm": 0.952502935492005, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 29776 + }, + { + "epoch": 0.29777, + "grad_norm": 1.376161437213204, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 29777 + }, + { + "epoch": 0.29778, + "grad_norm": 0.8982536665561907, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 29778 + }, + { + "epoch": 0.29779, + "grad_norm": 1.161720180387011, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 29779 + }, + { + "epoch": 0.2978, + "grad_norm": 1.0442455484299675, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 29780 + }, + { + "epoch": 0.29781, + "grad_norm": 1.2267413985373579, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 29781 + }, + { + "epoch": 0.29782, + "grad_norm": 0.958209048246162, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 29782 + }, + { + "epoch": 0.29783, + "grad_norm": 1.0991877066664777, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 29783 + }, + { + "epoch": 0.29784, + "grad_norm": 1.1930541735449007, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 29784 + }, + { + "epoch": 0.29785, + "grad_norm": 1.301484373803896, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 29785 + }, + { + "epoch": 0.29786, + "grad_norm": 0.8257520906811301, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 29786 + }, + { + "epoch": 0.29787, + "grad_norm": 0.8793012444698889, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 29787 + }, + { + "epoch": 0.29788, + "grad_norm": 1.090098526792619, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 29788 + }, + { + "epoch": 0.29789, + "grad_norm": 1.305860183336439, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 29789 + }, + { + "epoch": 0.2979, + "grad_norm": 0.9130443709326086, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 29790 + }, + { + "epoch": 0.29791, + "grad_norm": 1.1375411172181564, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 29791 + }, + { + "epoch": 0.29792, + "grad_norm": 1.2091128447016135, + "learning_rate": 0.003, + "loss": 4.0926, + "step": 29792 + }, + { + "epoch": 0.29793, + "grad_norm": 0.9554205569165426, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 29793 + }, + { + "epoch": 0.29794, + "grad_norm": 0.9575974998769506, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 29794 + }, + { + "epoch": 0.29795, + "grad_norm": 1.0632820328893082, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 29795 + }, + { + "epoch": 0.29796, + "grad_norm": 1.0986699350782883, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 29796 + }, + { + "epoch": 0.29797, + "grad_norm": 1.119265566905063, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 29797 + }, + { + "epoch": 0.29798, + "grad_norm": 1.026859428435591, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 29798 + }, + { + "epoch": 0.29799, + "grad_norm": 1.049096069066015, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 29799 + }, + { + "epoch": 0.298, + "grad_norm": 1.1287486745716215, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 29800 + }, + { + "epoch": 0.29801, + "grad_norm": 1.2308571588024104, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 29801 + }, + { + "epoch": 0.29802, + "grad_norm": 1.0349874300194446, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 29802 + }, + { + "epoch": 0.29803, + "grad_norm": 1.2884451757783797, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 29803 + }, + { + "epoch": 0.29804, + "grad_norm": 1.0584062170518664, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 29804 + }, + { + "epoch": 0.29805, + "grad_norm": 1.235796319131392, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 29805 + }, + { + "epoch": 0.29806, + "grad_norm": 0.9724179294746428, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 29806 + }, + { + "epoch": 0.29807, + "grad_norm": 1.1159359230796615, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 29807 + }, + { + "epoch": 0.29808, + "grad_norm": 1.1839687396933594, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 29808 + }, + { + "epoch": 0.29809, + "grad_norm": 0.9301823930135413, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 29809 + }, + { + "epoch": 0.2981, + "grad_norm": 0.9170229099001045, + "learning_rate": 0.003, + "loss": 4.043, + "step": 29810 + }, + { + "epoch": 0.29811, + "grad_norm": 1.0134042574054254, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 29811 + }, + { + "epoch": 0.29812, + "grad_norm": 1.4585720720052118, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 29812 + }, + { + "epoch": 0.29813, + "grad_norm": 0.924857349750193, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 29813 + }, + { + "epoch": 0.29814, + "grad_norm": 0.9176772046611664, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 29814 + }, + { + "epoch": 0.29815, + "grad_norm": 1.0518545923638414, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 29815 + }, + { + "epoch": 0.29816, + "grad_norm": 1.0614935292266952, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 29816 + }, + { + "epoch": 0.29817, + "grad_norm": 1.10762315546244, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 29817 + }, + { + "epoch": 0.29818, + "grad_norm": 1.2561362315577502, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 29818 + }, + { + "epoch": 0.29819, + "grad_norm": 1.3713348396347784, + "learning_rate": 0.003, + "loss": 4.052, + "step": 29819 + }, + { + "epoch": 0.2982, + "grad_norm": 0.7611543232748034, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 29820 + }, + { + "epoch": 0.29821, + "grad_norm": 0.8223766091561833, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 29821 + }, + { + "epoch": 0.29822, + "grad_norm": 1.05472462987539, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 29822 + }, + { + "epoch": 0.29823, + "grad_norm": 1.1802880467653505, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 29823 + }, + { + "epoch": 0.29824, + "grad_norm": 1.014873051156813, + "learning_rate": 0.003, + "loss": 4.04, + "step": 29824 + }, + { + "epoch": 0.29825, + "grad_norm": 1.3378715206121734, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 29825 + }, + { + "epoch": 0.29826, + "grad_norm": 1.0158861558752936, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 29826 + }, + { + "epoch": 0.29827, + "grad_norm": 1.392463849579065, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 29827 + }, + { + "epoch": 0.29828, + "grad_norm": 0.8095466595537, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 29828 + }, + { + "epoch": 0.29829, + "grad_norm": 0.7443032193149398, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 29829 + }, + { + "epoch": 0.2983, + "grad_norm": 0.8667553407845245, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 29830 + }, + { + "epoch": 0.29831, + "grad_norm": 1.0276403527360238, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 29831 + }, + { + "epoch": 0.29832, + "grad_norm": 1.2555995415696737, + "learning_rate": 0.003, + "loss": 4.0912, + "step": 29832 + }, + { + "epoch": 0.29833, + "grad_norm": 0.967544236181092, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 29833 + }, + { + "epoch": 0.29834, + "grad_norm": 1.1553141395611248, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 29834 + }, + { + "epoch": 0.29835, + "grad_norm": 1.2273654255378783, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 29835 + }, + { + "epoch": 0.29836, + "grad_norm": 0.9398718458724801, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 29836 + }, + { + "epoch": 0.29837, + "grad_norm": 1.0690942405742505, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 29837 + }, + { + "epoch": 0.29838, + "grad_norm": 1.2621487639927975, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 29838 + }, + { + "epoch": 0.29839, + "grad_norm": 1.086842253159147, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 29839 + }, + { + "epoch": 0.2984, + "grad_norm": 0.9130132098440644, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 29840 + }, + { + "epoch": 0.29841, + "grad_norm": 0.9292508227032181, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 29841 + }, + { + "epoch": 0.29842, + "grad_norm": 0.9758291134835305, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 29842 + }, + { + "epoch": 0.29843, + "grad_norm": 1.0953241936753666, + "learning_rate": 0.003, + "loss": 4.038, + "step": 29843 + }, + { + "epoch": 0.29844, + "grad_norm": 1.061639803986594, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 29844 + }, + { + "epoch": 0.29845, + "grad_norm": 1.2028891190689786, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 29845 + }, + { + "epoch": 0.29846, + "grad_norm": 0.8642475045832874, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 29846 + }, + { + "epoch": 0.29847, + "grad_norm": 0.7187849414905737, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 29847 + }, + { + "epoch": 0.29848, + "grad_norm": 0.8298612045662153, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 29848 + }, + { + "epoch": 0.29849, + "grad_norm": 1.1779565289880645, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 29849 + }, + { + "epoch": 0.2985, + "grad_norm": 1.308690174751314, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 29850 + }, + { + "epoch": 0.29851, + "grad_norm": 0.9140443073630007, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 29851 + }, + { + "epoch": 0.29852, + "grad_norm": 1.0905408374792205, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 29852 + }, + { + "epoch": 0.29853, + "grad_norm": 1.1100905133687224, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 29853 + }, + { + "epoch": 0.29854, + "grad_norm": 1.218517467327708, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 29854 + }, + { + "epoch": 0.29855, + "grad_norm": 0.939731526632792, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 29855 + }, + { + "epoch": 0.29856, + "grad_norm": 0.9338203415272067, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 29856 + }, + { + "epoch": 0.29857, + "grad_norm": 0.9871966261774011, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 29857 + }, + { + "epoch": 0.29858, + "grad_norm": 1.1963727711861312, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 29858 + }, + { + "epoch": 0.29859, + "grad_norm": 1.0114791472074258, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 29859 + }, + { + "epoch": 0.2986, + "grad_norm": 1.2588655094733538, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 29860 + }, + { + "epoch": 0.29861, + "grad_norm": 1.17916319144672, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 29861 + }, + { + "epoch": 0.29862, + "grad_norm": 1.079353112092283, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 29862 + }, + { + "epoch": 0.29863, + "grad_norm": 1.398773841741194, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 29863 + }, + { + "epoch": 0.29864, + "grad_norm": 0.9534368518787966, + "learning_rate": 0.003, + "loss": 4.04, + "step": 29864 + }, + { + "epoch": 0.29865, + "grad_norm": 0.9428690063173859, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 29865 + }, + { + "epoch": 0.29866, + "grad_norm": 1.0595484355128213, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 29866 + }, + { + "epoch": 0.29867, + "grad_norm": 1.2796580273914662, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 29867 + }, + { + "epoch": 0.29868, + "grad_norm": 0.9722051344683549, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 29868 + }, + { + "epoch": 0.29869, + "grad_norm": 1.179588355668522, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 29869 + }, + { + "epoch": 0.2987, + "grad_norm": 1.0866762779458443, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 29870 + }, + { + "epoch": 0.29871, + "grad_norm": 1.0987038644274882, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 29871 + }, + { + "epoch": 0.29872, + "grad_norm": 1.033651567897002, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 29872 + }, + { + "epoch": 0.29873, + "grad_norm": 1.0458190627843287, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 29873 + }, + { + "epoch": 0.29874, + "grad_norm": 1.062492539232403, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 29874 + }, + { + "epoch": 0.29875, + "grad_norm": 1.3657981740525824, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 29875 + }, + { + "epoch": 0.29876, + "grad_norm": 0.8674088317631138, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 29876 + }, + { + "epoch": 0.29877, + "grad_norm": 0.9396434232094001, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 29877 + }, + { + "epoch": 0.29878, + "grad_norm": 1.2797595705576608, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 29878 + }, + { + "epoch": 0.29879, + "grad_norm": 1.1553374046350635, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 29879 + }, + { + "epoch": 0.2988, + "grad_norm": 0.9081782230143359, + "learning_rate": 0.003, + "loss": 4.033, + "step": 29880 + }, + { + "epoch": 0.29881, + "grad_norm": 0.9098811110217757, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 29881 + }, + { + "epoch": 0.29882, + "grad_norm": 1.176671274657057, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 29882 + }, + { + "epoch": 0.29883, + "grad_norm": 0.9941816760372478, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 29883 + }, + { + "epoch": 0.29884, + "grad_norm": 1.1067261732260107, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 29884 + }, + { + "epoch": 0.29885, + "grad_norm": 0.9498358898263871, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 29885 + }, + { + "epoch": 0.29886, + "grad_norm": 1.1467334667891176, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 29886 + }, + { + "epoch": 0.29887, + "grad_norm": 1.2115597758463859, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 29887 + }, + { + "epoch": 0.29888, + "grad_norm": 0.9260960807880517, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 29888 + }, + { + "epoch": 0.29889, + "grad_norm": 0.95792791385796, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 29889 + }, + { + "epoch": 0.2989, + "grad_norm": 1.1585878201399218, + "learning_rate": 0.003, + "loss": 4.061, + "step": 29890 + }, + { + "epoch": 0.29891, + "grad_norm": 1.0268243911534218, + "learning_rate": 0.003, + "loss": 4.067, + "step": 29891 + }, + { + "epoch": 0.29892, + "grad_norm": 1.2294821143637544, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 29892 + }, + { + "epoch": 0.29893, + "grad_norm": 1.1966642863418362, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 29893 + }, + { + "epoch": 0.29894, + "grad_norm": 0.978062426145778, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 29894 + }, + { + "epoch": 0.29895, + "grad_norm": 1.0676773325089142, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 29895 + }, + { + "epoch": 0.29896, + "grad_norm": 1.1157500529931743, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 29896 + }, + { + "epoch": 0.29897, + "grad_norm": 1.146832416194817, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 29897 + }, + { + "epoch": 0.29898, + "grad_norm": 1.141878523622955, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 29898 + }, + { + "epoch": 0.29899, + "grad_norm": 1.0641830865608557, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 29899 + }, + { + "epoch": 0.299, + "grad_norm": 0.9498800880786231, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 29900 + }, + { + "epoch": 0.29901, + "grad_norm": 0.9724224668826824, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 29901 + }, + { + "epoch": 0.29902, + "grad_norm": 1.1584366672850466, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 29902 + }, + { + "epoch": 0.29903, + "grad_norm": 1.0839372295964456, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 29903 + }, + { + "epoch": 0.29904, + "grad_norm": 1.2575063112031666, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 29904 + }, + { + "epoch": 0.29905, + "grad_norm": 0.8439546981875998, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 29905 + }, + { + "epoch": 0.29906, + "grad_norm": 0.832322495169208, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 29906 + }, + { + "epoch": 0.29907, + "grad_norm": 0.8912857252018596, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 29907 + }, + { + "epoch": 0.29908, + "grad_norm": 1.146357715674646, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 29908 + }, + { + "epoch": 0.29909, + "grad_norm": 1.124794417946039, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 29909 + }, + { + "epoch": 0.2991, + "grad_norm": 0.8130739226431539, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 29910 + }, + { + "epoch": 0.29911, + "grad_norm": 0.8244267065702038, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 29911 + }, + { + "epoch": 0.29912, + "grad_norm": 0.9387463880059892, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 29912 + }, + { + "epoch": 0.29913, + "grad_norm": 1.076454878425281, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 29913 + }, + { + "epoch": 0.29914, + "grad_norm": 1.0404822905727447, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 29914 + }, + { + "epoch": 0.29915, + "grad_norm": 1.0166586223806646, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 29915 + }, + { + "epoch": 0.29916, + "grad_norm": 1.0451345801115857, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 29916 + }, + { + "epoch": 0.29917, + "grad_norm": 1.1369997562853202, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 29917 + }, + { + "epoch": 0.29918, + "grad_norm": 1.1061966695890741, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 29918 + }, + { + "epoch": 0.29919, + "grad_norm": 1.1305131619346989, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 29919 + }, + { + "epoch": 0.2992, + "grad_norm": 0.9199645397530428, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 29920 + }, + { + "epoch": 0.29921, + "grad_norm": 1.0119670032790695, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 29921 + }, + { + "epoch": 0.29922, + "grad_norm": 1.2163009018184272, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 29922 + }, + { + "epoch": 0.29923, + "grad_norm": 1.1723790202097457, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 29923 + }, + { + "epoch": 0.29924, + "grad_norm": 1.0845580323503992, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 29924 + }, + { + "epoch": 0.29925, + "grad_norm": 1.015098625734287, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 29925 + }, + { + "epoch": 0.29926, + "grad_norm": 1.143505113538933, + "learning_rate": 0.003, + "loss": 4.035, + "step": 29926 + }, + { + "epoch": 0.29927, + "grad_norm": 0.9510202323429991, + "learning_rate": 0.003, + "loss": 4.03, + "step": 29927 + }, + { + "epoch": 0.29928, + "grad_norm": 1.0003160552253407, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 29928 + }, + { + "epoch": 0.29929, + "grad_norm": 1.0463788453280742, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 29929 + }, + { + "epoch": 0.2993, + "grad_norm": 1.087837053459514, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 29930 + }, + { + "epoch": 0.29931, + "grad_norm": 1.1343597590176595, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 29931 + }, + { + "epoch": 0.29932, + "grad_norm": 1.0000900781765794, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 29932 + }, + { + "epoch": 0.29933, + "grad_norm": 1.0853269834661128, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 29933 + }, + { + "epoch": 0.29934, + "grad_norm": 1.2837541098392264, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 29934 + }, + { + "epoch": 0.29935, + "grad_norm": 1.020936238532086, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 29935 + }, + { + "epoch": 0.29936, + "grad_norm": 1.237728133286655, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 29936 + }, + { + "epoch": 0.29937, + "grad_norm": 0.9584012765243217, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 29937 + }, + { + "epoch": 0.29938, + "grad_norm": 1.1717675972407873, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 29938 + }, + { + "epoch": 0.29939, + "grad_norm": 0.9965028602697712, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 29939 + }, + { + "epoch": 0.2994, + "grad_norm": 1.148537755588104, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 29940 + }, + { + "epoch": 0.29941, + "grad_norm": 1.0032093315958086, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 29941 + }, + { + "epoch": 0.29942, + "grad_norm": 1.0274784534787533, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 29942 + }, + { + "epoch": 0.29943, + "grad_norm": 1.139870794164059, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 29943 + }, + { + "epoch": 0.29944, + "grad_norm": 1.1129314032614073, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 29944 + }, + { + "epoch": 0.29945, + "grad_norm": 0.9949531095756404, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 29945 + }, + { + "epoch": 0.29946, + "grad_norm": 1.2074655792333846, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 29946 + }, + { + "epoch": 0.29947, + "grad_norm": 1.0507320950506551, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 29947 + }, + { + "epoch": 0.29948, + "grad_norm": 1.0647818755305023, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 29948 + }, + { + "epoch": 0.29949, + "grad_norm": 1.1568292485944647, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 29949 + }, + { + "epoch": 0.2995, + "grad_norm": 1.095989522906087, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 29950 + }, + { + "epoch": 0.29951, + "grad_norm": 0.9411739847087198, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 29951 + }, + { + "epoch": 0.29952, + "grad_norm": 1.0324677099567359, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 29952 + }, + { + "epoch": 0.29953, + "grad_norm": 1.230556479280951, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 29953 + }, + { + "epoch": 0.29954, + "grad_norm": 1.0666609456854992, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 29954 + }, + { + "epoch": 0.29955, + "grad_norm": 1.4215434976183177, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 29955 + }, + { + "epoch": 0.29956, + "grad_norm": 0.8177252666718423, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 29956 + }, + { + "epoch": 0.29957, + "grad_norm": 0.886257256453193, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 29957 + }, + { + "epoch": 0.29958, + "grad_norm": 1.0756438145713254, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 29958 + }, + { + "epoch": 0.29959, + "grad_norm": 0.9560892470403191, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 29959 + }, + { + "epoch": 0.2996, + "grad_norm": 1.0048252678384717, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 29960 + }, + { + "epoch": 0.29961, + "grad_norm": 0.9016361905498271, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 29961 + }, + { + "epoch": 0.29962, + "grad_norm": 0.965844723049094, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 29962 + }, + { + "epoch": 0.29963, + "grad_norm": 1.0380172318924765, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 29963 + }, + { + "epoch": 0.29964, + "grad_norm": 1.1807932997083568, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 29964 + }, + { + "epoch": 0.29965, + "grad_norm": 0.925919897865295, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 29965 + }, + { + "epoch": 0.29966, + "grad_norm": 1.2264172681603411, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 29966 + }, + { + "epoch": 0.29967, + "grad_norm": 1.2678576378920925, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 29967 + }, + { + "epoch": 0.29968, + "grad_norm": 0.8902253916285043, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 29968 + }, + { + "epoch": 0.29969, + "grad_norm": 0.8305891681545309, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 29969 + }, + { + "epoch": 0.2997, + "grad_norm": 0.9913006528350308, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 29970 + }, + { + "epoch": 0.29971, + "grad_norm": 1.6571229646938497, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 29971 + }, + { + "epoch": 0.29972, + "grad_norm": 0.6358213402615391, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 29972 + }, + { + "epoch": 0.29973, + "grad_norm": 0.925146236579952, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 29973 + }, + { + "epoch": 0.29974, + "grad_norm": 1.071919958246788, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 29974 + }, + { + "epoch": 0.29975, + "grad_norm": 1.191373517377357, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 29975 + }, + { + "epoch": 0.29976, + "grad_norm": 1.0732924660079206, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 29976 + }, + { + "epoch": 0.29977, + "grad_norm": 1.2111110989829723, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 29977 + }, + { + "epoch": 0.29978, + "grad_norm": 0.9601405769941326, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 29978 + }, + { + "epoch": 0.29979, + "grad_norm": 1.2025637314543172, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 29979 + }, + { + "epoch": 0.2998, + "grad_norm": 1.0171428316062157, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 29980 + }, + { + "epoch": 0.29981, + "grad_norm": 1.133516850848833, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 29981 + }, + { + "epoch": 0.29982, + "grad_norm": 1.26200935099796, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 29982 + }, + { + "epoch": 0.29983, + "grad_norm": 1.129971522055585, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 29983 + }, + { + "epoch": 0.29984, + "grad_norm": 0.9546511613348924, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 29984 + }, + { + "epoch": 0.29985, + "grad_norm": 0.9813927454259967, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 29985 + }, + { + "epoch": 0.29986, + "grad_norm": 1.1717065490645278, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 29986 + }, + { + "epoch": 0.29987, + "grad_norm": 1.134425816672821, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 29987 + }, + { + "epoch": 0.29988, + "grad_norm": 1.128500072527028, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 29988 + }, + { + "epoch": 0.29989, + "grad_norm": 1.107992968584869, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 29989 + }, + { + "epoch": 0.2999, + "grad_norm": 0.989316459663728, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 29990 + }, + { + "epoch": 0.29991, + "grad_norm": 1.1505009877486567, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 29991 + }, + { + "epoch": 0.29992, + "grad_norm": 1.0326328743927449, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 29992 + }, + { + "epoch": 0.29993, + "grad_norm": 1.0264218168053256, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 29993 + }, + { + "epoch": 0.29994, + "grad_norm": 1.2023396189319606, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 29994 + }, + { + "epoch": 0.29995, + "grad_norm": 1.1264086555908226, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 29995 + }, + { + "epoch": 0.29996, + "grad_norm": 1.2410930184598936, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 29996 + }, + { + "epoch": 0.29997, + "grad_norm": 0.9874637644282724, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 29997 + }, + { + "epoch": 0.29998, + "grad_norm": 1.1699661139724207, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 29998 + }, + { + "epoch": 0.29999, + "grad_norm": 0.97601615645263, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 29999 + }, + { + "epoch": 0.3, + "grad_norm": 0.9753067005264043, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 30000 + }, + { + "epoch": 0.30001, + "grad_norm": 1.4062239332843118, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 30001 + }, + { + "epoch": 0.30002, + "grad_norm": 0.8144708984709705, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 30002 + }, + { + "epoch": 0.30003, + "grad_norm": 0.893253049790433, + "learning_rate": 0.003, + "loss": 4.026, + "step": 30003 + }, + { + "epoch": 0.30004, + "grad_norm": 1.221330048457432, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 30004 + }, + { + "epoch": 0.30005, + "grad_norm": 0.8990297520626109, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 30005 + }, + { + "epoch": 0.30006, + "grad_norm": 1.0627818921798253, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 30006 + }, + { + "epoch": 0.30007, + "grad_norm": 1.2383986062684056, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 30007 + }, + { + "epoch": 0.30008, + "grad_norm": 0.8313395016419542, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 30008 + }, + { + "epoch": 0.30009, + "grad_norm": 0.968789119091997, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 30009 + }, + { + "epoch": 0.3001, + "grad_norm": 1.1614252915429584, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 30010 + }, + { + "epoch": 0.30011, + "grad_norm": 1.0309891868806393, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 30011 + }, + { + "epoch": 0.30012, + "grad_norm": 1.1388788893015045, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 30012 + }, + { + "epoch": 0.30013, + "grad_norm": 1.0258114600048893, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 30013 + }, + { + "epoch": 0.30014, + "grad_norm": 1.2088586926323324, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 30014 + }, + { + "epoch": 0.30015, + "grad_norm": 0.8220090273153844, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 30015 + }, + { + "epoch": 0.30016, + "grad_norm": 0.8150200857191822, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 30016 + }, + { + "epoch": 0.30017, + "grad_norm": 0.9728953043120029, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 30017 + }, + { + "epoch": 0.30018, + "grad_norm": 1.1094421200317353, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 30018 + }, + { + "epoch": 0.30019, + "grad_norm": 1.0053358602007616, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 30019 + }, + { + "epoch": 0.3002, + "grad_norm": 1.0826053904426503, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 30020 + }, + { + "epoch": 0.30021, + "grad_norm": 1.0478505056781728, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 30021 + }, + { + "epoch": 0.30022, + "grad_norm": 1.1765777475549097, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 30022 + }, + { + "epoch": 0.30023, + "grad_norm": 1.1359500785766705, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 30023 + }, + { + "epoch": 0.30024, + "grad_norm": 1.1247178978146593, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 30024 + }, + { + "epoch": 0.30025, + "grad_norm": 1.137496088225742, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 30025 + }, + { + "epoch": 0.30026, + "grad_norm": 1.2309894331108817, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 30026 + }, + { + "epoch": 0.30027, + "grad_norm": 1.1937245132981003, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 30027 + }, + { + "epoch": 0.30028, + "grad_norm": 1.0188568038477865, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 30028 + }, + { + "epoch": 0.30029, + "grad_norm": 1.0762412235943895, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 30029 + }, + { + "epoch": 0.3003, + "grad_norm": 1.2335714649388316, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 30030 + }, + { + "epoch": 0.30031, + "grad_norm": 0.9210384669410004, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 30031 + }, + { + "epoch": 0.30032, + "grad_norm": 1.0408247328085183, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 30032 + }, + { + "epoch": 0.30033, + "grad_norm": 1.2927148589666397, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 30033 + }, + { + "epoch": 0.30034, + "grad_norm": 0.8879070239700921, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 30034 + }, + { + "epoch": 0.30035, + "grad_norm": 0.9485186864411721, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 30035 + }, + { + "epoch": 0.30036, + "grad_norm": 1.0884435155275831, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 30036 + }, + { + "epoch": 0.30037, + "grad_norm": 1.246440286953759, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 30037 + }, + { + "epoch": 0.30038, + "grad_norm": 1.080588743395153, + "learning_rate": 0.003, + "loss": 4.052, + "step": 30038 + }, + { + "epoch": 0.30039, + "grad_norm": 0.9024036140736544, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 30039 + }, + { + "epoch": 0.3004, + "grad_norm": 0.7990579806231957, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 30040 + }, + { + "epoch": 0.30041, + "grad_norm": 0.8346898397877797, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 30041 + }, + { + "epoch": 0.30042, + "grad_norm": 1.0894350284007097, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 30042 + }, + { + "epoch": 0.30043, + "grad_norm": 1.3509793532775887, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 30043 + }, + { + "epoch": 0.30044, + "grad_norm": 0.7588147481886602, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 30044 + }, + { + "epoch": 0.30045, + "grad_norm": 0.7560091256284373, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 30045 + }, + { + "epoch": 0.30046, + "grad_norm": 0.8452992542764027, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 30046 + }, + { + "epoch": 0.30047, + "grad_norm": 1.040861158820246, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 30047 + }, + { + "epoch": 0.30048, + "grad_norm": 1.349138131573085, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 30048 + }, + { + "epoch": 0.30049, + "grad_norm": 0.9023700064776918, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 30049 + }, + { + "epoch": 0.3005, + "grad_norm": 1.106623836668821, + "learning_rate": 0.003, + "loss": 4.034, + "step": 30050 + }, + { + "epoch": 0.30051, + "grad_norm": 0.9523574501858864, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 30051 + }, + { + "epoch": 0.30052, + "grad_norm": 1.1273862739997533, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 30052 + }, + { + "epoch": 0.30053, + "grad_norm": 0.9765186804991316, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 30053 + }, + { + "epoch": 0.30054, + "grad_norm": 1.256010815075085, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 30054 + }, + { + "epoch": 0.30055, + "grad_norm": 1.133537281097802, + "learning_rate": 0.003, + "loss": 4.031, + "step": 30055 + }, + { + "epoch": 0.30056, + "grad_norm": 1.0946461886002892, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 30056 + }, + { + "epoch": 0.30057, + "grad_norm": 1.1054376749219037, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 30057 + }, + { + "epoch": 0.30058, + "grad_norm": 1.2112760303428831, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 30058 + }, + { + "epoch": 0.30059, + "grad_norm": 1.0198858530939607, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 30059 + }, + { + "epoch": 0.3006, + "grad_norm": 1.3317619908530487, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 30060 + }, + { + "epoch": 0.30061, + "grad_norm": 0.895676998337477, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 30061 + }, + { + "epoch": 0.30062, + "grad_norm": 0.9546969960781494, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 30062 + }, + { + "epoch": 0.30063, + "grad_norm": 1.1394845377749139, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 30063 + }, + { + "epoch": 0.30064, + "grad_norm": 1.030725885204288, + "learning_rate": 0.003, + "loss": 4.052, + "step": 30064 + }, + { + "epoch": 0.30065, + "grad_norm": 0.9742715047888156, + "learning_rate": 0.003, + "loss": 4.026, + "step": 30065 + }, + { + "epoch": 0.30066, + "grad_norm": 0.9762241084453231, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 30066 + }, + { + "epoch": 0.30067, + "grad_norm": 1.0827553679464987, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 30067 + }, + { + "epoch": 0.30068, + "grad_norm": 1.0601781836214401, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 30068 + }, + { + "epoch": 0.30069, + "grad_norm": 1.1686951175676774, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 30069 + }, + { + "epoch": 0.3007, + "grad_norm": 1.0310601672735085, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 30070 + }, + { + "epoch": 0.30071, + "grad_norm": 1.1798785181575369, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 30071 + }, + { + "epoch": 0.30072, + "grad_norm": 1.202015280697754, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 30072 + }, + { + "epoch": 0.30073, + "grad_norm": 0.7973772776677926, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 30073 + }, + { + "epoch": 0.30074, + "grad_norm": 0.8124147553953158, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 30074 + }, + { + "epoch": 0.30075, + "grad_norm": 1.0880403241494516, + "learning_rate": 0.003, + "loss": 4.048, + "step": 30075 + }, + { + "epoch": 0.30076, + "grad_norm": 1.1992857495165548, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 30076 + }, + { + "epoch": 0.30077, + "grad_norm": 0.9734428876991728, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 30077 + }, + { + "epoch": 0.30078, + "grad_norm": 1.005676602058751, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 30078 + }, + { + "epoch": 0.30079, + "grad_norm": 0.9760689262139679, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 30079 + }, + { + "epoch": 0.3008, + "grad_norm": 1.1543856793989475, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 30080 + }, + { + "epoch": 0.30081, + "grad_norm": 0.963415102858283, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 30081 + }, + { + "epoch": 0.30082, + "grad_norm": 1.3622171326267523, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 30082 + }, + { + "epoch": 0.30083, + "grad_norm": 0.9736020659430044, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 30083 + }, + { + "epoch": 0.30084, + "grad_norm": 1.0782757332717814, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 30084 + }, + { + "epoch": 0.30085, + "grad_norm": 1.3508552102469087, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 30085 + }, + { + "epoch": 0.30086, + "grad_norm": 1.2361956163963799, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 30086 + }, + { + "epoch": 0.30087, + "grad_norm": 0.9483343682004833, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 30087 + }, + { + "epoch": 0.30088, + "grad_norm": 0.9757900547955183, + "learning_rate": 0.003, + "loss": 4.072, + "step": 30088 + }, + { + "epoch": 0.30089, + "grad_norm": 1.2170518696854535, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 30089 + }, + { + "epoch": 0.3009, + "grad_norm": 0.9034650278224796, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 30090 + }, + { + "epoch": 0.30091, + "grad_norm": 1.0442539259006167, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 30091 + }, + { + "epoch": 0.30092, + "grad_norm": 1.1303937937968127, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 30092 + }, + { + "epoch": 0.30093, + "grad_norm": 1.0237965929106938, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 30093 + }, + { + "epoch": 0.30094, + "grad_norm": 1.019238833627632, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 30094 + }, + { + "epoch": 0.30095, + "grad_norm": 1.228601131068322, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 30095 + }, + { + "epoch": 0.30096, + "grad_norm": 0.9970088193993968, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 30096 + }, + { + "epoch": 0.30097, + "grad_norm": 1.0777845286808008, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 30097 + }, + { + "epoch": 0.30098, + "grad_norm": 1.0353477502749249, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 30098 + }, + { + "epoch": 0.30099, + "grad_norm": 1.546452514119611, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 30099 + }, + { + "epoch": 0.301, + "grad_norm": 0.8412477905709287, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 30100 + }, + { + "epoch": 0.30101, + "grad_norm": 0.768361962394437, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 30101 + }, + { + "epoch": 0.30102, + "grad_norm": 0.8043182140849372, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 30102 + }, + { + "epoch": 0.30103, + "grad_norm": 0.9584649652916183, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 30103 + }, + { + "epoch": 0.30104, + "grad_norm": 1.019146341603254, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 30104 + }, + { + "epoch": 0.30105, + "grad_norm": 1.3713763738903673, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 30105 + }, + { + "epoch": 0.30106, + "grad_norm": 1.1053931004494015, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 30106 + }, + { + "epoch": 0.30107, + "grad_norm": 1.0324175980416506, + "learning_rate": 0.003, + "loss": 4.049, + "step": 30107 + }, + { + "epoch": 0.30108, + "grad_norm": 1.119040637844338, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 30108 + }, + { + "epoch": 0.30109, + "grad_norm": 1.1334752443938578, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 30109 + }, + { + "epoch": 0.3011, + "grad_norm": 0.8537298506869394, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 30110 + }, + { + "epoch": 0.30111, + "grad_norm": 0.9030267853613423, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 30111 + }, + { + "epoch": 0.30112, + "grad_norm": 0.9376756167599487, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 30112 + }, + { + "epoch": 0.30113, + "grad_norm": 1.2185670010873693, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 30113 + }, + { + "epoch": 0.30114, + "grad_norm": 1.1700167821570295, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 30114 + }, + { + "epoch": 0.30115, + "grad_norm": 1.1521754277490426, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 30115 + }, + { + "epoch": 0.30116, + "grad_norm": 1.0830440825038412, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 30116 + }, + { + "epoch": 0.30117, + "grad_norm": 1.0759733150022195, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 30117 + }, + { + "epoch": 0.30118, + "grad_norm": 1.0851766459451977, + "learning_rate": 0.003, + "loss": 4.07, + "step": 30118 + }, + { + "epoch": 0.30119, + "grad_norm": 1.0460999065001475, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 30119 + }, + { + "epoch": 0.3012, + "grad_norm": 1.0818034106069128, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 30120 + }, + { + "epoch": 0.30121, + "grad_norm": 0.9928860103960752, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 30121 + }, + { + "epoch": 0.30122, + "grad_norm": 1.0729247498642258, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 30122 + }, + { + "epoch": 0.30123, + "grad_norm": 0.9818688314495403, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 30123 + }, + { + "epoch": 0.30124, + "grad_norm": 1.2352519037502674, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 30124 + }, + { + "epoch": 0.30125, + "grad_norm": 0.9838005700522446, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 30125 + }, + { + "epoch": 0.30126, + "grad_norm": 1.1334131884965972, + "learning_rate": 0.003, + "loss": 4.066, + "step": 30126 + }, + { + "epoch": 0.30127, + "grad_norm": 1.1539094751138936, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 30127 + }, + { + "epoch": 0.30128, + "grad_norm": 0.9085509113314439, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 30128 + }, + { + "epoch": 0.30129, + "grad_norm": 0.9587730093329753, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 30129 + }, + { + "epoch": 0.3013, + "grad_norm": 1.0208234246294425, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 30130 + }, + { + "epoch": 0.30131, + "grad_norm": 0.9531122550797614, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 30131 + }, + { + "epoch": 0.30132, + "grad_norm": 1.1441570435898152, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 30132 + }, + { + "epoch": 0.30133, + "grad_norm": 0.9251074375572165, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 30133 + }, + { + "epoch": 0.30134, + "grad_norm": 0.9754656423909108, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 30134 + }, + { + "epoch": 0.30135, + "grad_norm": 1.1635262988597064, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 30135 + }, + { + "epoch": 0.30136, + "grad_norm": 1.0755664955525548, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 30136 + }, + { + "epoch": 0.30137, + "grad_norm": 1.2708639597830567, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 30137 + }, + { + "epoch": 0.30138, + "grad_norm": 1.1119995911353442, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 30138 + }, + { + "epoch": 0.30139, + "grad_norm": 1.0222272815290703, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 30139 + }, + { + "epoch": 0.3014, + "grad_norm": 0.9979513940699515, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 30140 + }, + { + "epoch": 0.30141, + "grad_norm": 1.0288859550429112, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 30141 + }, + { + "epoch": 0.30142, + "grad_norm": 1.2610898532076211, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 30142 + }, + { + "epoch": 0.30143, + "grad_norm": 0.9983546178101397, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 30143 + }, + { + "epoch": 0.30144, + "grad_norm": 1.328124495246968, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 30144 + }, + { + "epoch": 0.30145, + "grad_norm": 1.0016594370210594, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 30145 + }, + { + "epoch": 0.30146, + "grad_norm": 1.1266121016613087, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 30146 + }, + { + "epoch": 0.30147, + "grad_norm": 0.9424465246229069, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 30147 + }, + { + "epoch": 0.30148, + "grad_norm": 1.0431415027583488, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 30148 + }, + { + "epoch": 0.30149, + "grad_norm": 1.1010375879757788, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 30149 + }, + { + "epoch": 0.3015, + "grad_norm": 1.141760145980947, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 30150 + }, + { + "epoch": 0.30151, + "grad_norm": 0.9884378443630473, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 30151 + }, + { + "epoch": 0.30152, + "grad_norm": 0.9497167171391041, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 30152 + }, + { + "epoch": 0.30153, + "grad_norm": 1.050203863465832, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 30153 + }, + { + "epoch": 0.30154, + "grad_norm": 1.2873073730296225, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 30154 + }, + { + "epoch": 0.30155, + "grad_norm": 0.960418664751485, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 30155 + }, + { + "epoch": 0.30156, + "grad_norm": 1.0738555921378796, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 30156 + }, + { + "epoch": 0.30157, + "grad_norm": 1.171350711834754, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 30157 + }, + { + "epoch": 0.30158, + "grad_norm": 1.0770374851304518, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 30158 + }, + { + "epoch": 0.30159, + "grad_norm": 1.1743358059931124, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 30159 + }, + { + "epoch": 0.3016, + "grad_norm": 0.9926750132110479, + "learning_rate": 0.003, + "loss": 4.071, + "step": 30160 + }, + { + "epoch": 0.30161, + "grad_norm": 1.114204758144242, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 30161 + }, + { + "epoch": 0.30162, + "grad_norm": 0.9701686886310977, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 30162 + }, + { + "epoch": 0.30163, + "grad_norm": 1.1907451581829183, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 30163 + }, + { + "epoch": 0.30164, + "grad_norm": 1.1067591324406874, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 30164 + }, + { + "epoch": 0.30165, + "grad_norm": 1.3117610812140135, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 30165 + }, + { + "epoch": 0.30166, + "grad_norm": 1.1625125063233905, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 30166 + }, + { + "epoch": 0.30167, + "grad_norm": 1.0180588215779627, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 30167 + }, + { + "epoch": 0.30168, + "grad_norm": 1.1544556555781036, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 30168 + }, + { + "epoch": 0.30169, + "grad_norm": 0.9807437607716807, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 30169 + }, + { + "epoch": 0.3017, + "grad_norm": 1.0105327131023893, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 30170 + }, + { + "epoch": 0.30171, + "grad_norm": 1.250186971782343, + "learning_rate": 0.003, + "loss": 4.05, + "step": 30171 + }, + { + "epoch": 0.30172, + "grad_norm": 0.9427719447750631, + "learning_rate": 0.003, + "loss": 4.035, + "step": 30172 + }, + { + "epoch": 0.30173, + "grad_norm": 1.2247345575078956, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 30173 + }, + { + "epoch": 0.30174, + "grad_norm": 1.0471598794624195, + "learning_rate": 0.003, + "loss": 4.022, + "step": 30174 + }, + { + "epoch": 0.30175, + "grad_norm": 1.06857463905667, + "learning_rate": 0.003, + "loss": 4.0955, + "step": 30175 + }, + { + "epoch": 0.30176, + "grad_norm": 1.0122603981739609, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 30176 + }, + { + "epoch": 0.30177, + "grad_norm": 1.1351430482573508, + "learning_rate": 0.003, + "loss": 4.018, + "step": 30177 + }, + { + "epoch": 0.30178, + "grad_norm": 0.9659382947496855, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 30178 + }, + { + "epoch": 0.30179, + "grad_norm": 1.1288031084693488, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 30179 + }, + { + "epoch": 0.3018, + "grad_norm": 1.2710547693890284, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 30180 + }, + { + "epoch": 0.30181, + "grad_norm": 0.9498795153051696, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 30181 + }, + { + "epoch": 0.30182, + "grad_norm": 1.0291392413934406, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 30182 + }, + { + "epoch": 0.30183, + "grad_norm": 1.0332297225357585, + "learning_rate": 0.003, + "loss": 4.028, + "step": 30183 + }, + { + "epoch": 0.30184, + "grad_norm": 1.1291092040653077, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 30184 + }, + { + "epoch": 0.30185, + "grad_norm": 1.0302462570265143, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 30185 + }, + { + "epoch": 0.30186, + "grad_norm": 1.0363034096209056, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 30186 + }, + { + "epoch": 0.30187, + "grad_norm": 1.0687029476780106, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 30187 + }, + { + "epoch": 0.30188, + "grad_norm": 1.145539137029252, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 30188 + }, + { + "epoch": 0.30189, + "grad_norm": 0.9529419287755079, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 30189 + }, + { + "epoch": 0.3019, + "grad_norm": 1.1126849689640999, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 30190 + }, + { + "epoch": 0.30191, + "grad_norm": 0.8746066336249793, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 30191 + }, + { + "epoch": 0.30192, + "grad_norm": 0.9470772112864352, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 30192 + }, + { + "epoch": 0.30193, + "grad_norm": 1.00922010774416, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 30193 + }, + { + "epoch": 0.30194, + "grad_norm": 1.494411090173074, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 30194 + }, + { + "epoch": 0.30195, + "grad_norm": 1.0591789495097503, + "learning_rate": 0.003, + "loss": 4.072, + "step": 30195 + }, + { + "epoch": 0.30196, + "grad_norm": 1.2134446761265747, + "learning_rate": 0.003, + "loss": 4.074, + "step": 30196 + }, + { + "epoch": 0.30197, + "grad_norm": 1.0451119854911526, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 30197 + }, + { + "epoch": 0.30198, + "grad_norm": 1.209640202409125, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 30198 + }, + { + "epoch": 0.30199, + "grad_norm": 1.0551719268713382, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 30199 + }, + { + "epoch": 0.302, + "grad_norm": 1.0643538455991264, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 30200 + }, + { + "epoch": 0.30201, + "grad_norm": 1.2467320669925137, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 30201 + }, + { + "epoch": 0.30202, + "grad_norm": 0.9911901965400157, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 30202 + }, + { + "epoch": 0.30203, + "grad_norm": 0.9594687634034816, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 30203 + }, + { + "epoch": 0.30204, + "grad_norm": 1.208288240992083, + "learning_rate": 0.003, + "loss": 4.049, + "step": 30204 + }, + { + "epoch": 0.30205, + "grad_norm": 1.147272547726401, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 30205 + }, + { + "epoch": 0.30206, + "grad_norm": 0.9761412827306974, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 30206 + }, + { + "epoch": 0.30207, + "grad_norm": 1.1503573985818787, + "learning_rate": 0.003, + "loss": 4.065, + "step": 30207 + }, + { + "epoch": 0.30208, + "grad_norm": 0.942934857063543, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 30208 + }, + { + "epoch": 0.30209, + "grad_norm": 1.188288738829705, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 30209 + }, + { + "epoch": 0.3021, + "grad_norm": 1.0034724137618263, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 30210 + }, + { + "epoch": 0.30211, + "grad_norm": 1.117286361655862, + "learning_rate": 0.003, + "loss": 4.042, + "step": 30211 + }, + { + "epoch": 0.30212, + "grad_norm": 0.9259404278959955, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 30212 + }, + { + "epoch": 0.30213, + "grad_norm": 1.0307081040609545, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 30213 + }, + { + "epoch": 0.30214, + "grad_norm": 1.2374948652525266, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 30214 + }, + { + "epoch": 0.30215, + "grad_norm": 1.2776881873294639, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 30215 + }, + { + "epoch": 0.30216, + "grad_norm": 0.8066718479914387, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 30216 + }, + { + "epoch": 0.30217, + "grad_norm": 0.9404991983053924, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 30217 + }, + { + "epoch": 0.30218, + "grad_norm": 1.2091673186434857, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 30218 + }, + { + "epoch": 0.30219, + "grad_norm": 1.042153436109762, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 30219 + }, + { + "epoch": 0.3022, + "grad_norm": 1.3165486279239245, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 30220 + }, + { + "epoch": 0.30221, + "grad_norm": 0.7975562775977223, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 30221 + }, + { + "epoch": 0.30222, + "grad_norm": 0.8623102134388201, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 30222 + }, + { + "epoch": 0.30223, + "grad_norm": 1.1109598774223004, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 30223 + }, + { + "epoch": 0.30224, + "grad_norm": 1.1846239061010002, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 30224 + }, + { + "epoch": 0.30225, + "grad_norm": 0.9482143383686622, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 30225 + }, + { + "epoch": 0.30226, + "grad_norm": 1.0312781991634157, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 30226 + }, + { + "epoch": 0.30227, + "grad_norm": 1.2461519865662611, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 30227 + }, + { + "epoch": 0.30228, + "grad_norm": 0.8738091727242387, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 30228 + }, + { + "epoch": 0.30229, + "grad_norm": 1.0167464894464464, + "learning_rate": 0.003, + "loss": 4.04, + "step": 30229 + }, + { + "epoch": 0.3023, + "grad_norm": 1.1341869981098827, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 30230 + }, + { + "epoch": 0.30231, + "grad_norm": 1.1124847744520456, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 30231 + }, + { + "epoch": 0.30232, + "grad_norm": 1.2723068255065415, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 30232 + }, + { + "epoch": 0.30233, + "grad_norm": 1.1701762789741699, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 30233 + }, + { + "epoch": 0.30234, + "grad_norm": 0.885175437502927, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 30234 + }, + { + "epoch": 0.30235, + "grad_norm": 0.8426812119287402, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 30235 + }, + { + "epoch": 0.30236, + "grad_norm": 0.887110493872594, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 30236 + }, + { + "epoch": 0.30237, + "grad_norm": 0.9625975905668169, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 30237 + }, + { + "epoch": 0.30238, + "grad_norm": 1.0139292655335286, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 30238 + }, + { + "epoch": 0.30239, + "grad_norm": 1.1353138808538417, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 30239 + }, + { + "epoch": 0.3024, + "grad_norm": 0.9104569053633249, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 30240 + }, + { + "epoch": 0.30241, + "grad_norm": 0.9171895590805165, + "learning_rate": 0.003, + "loss": 4.052, + "step": 30241 + }, + { + "epoch": 0.30242, + "grad_norm": 1.1006662252664392, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 30242 + }, + { + "epoch": 0.30243, + "grad_norm": 1.1490126567742982, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 30243 + }, + { + "epoch": 0.30244, + "grad_norm": 1.0489457010937995, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 30244 + }, + { + "epoch": 0.30245, + "grad_norm": 1.2283884840806854, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 30245 + }, + { + "epoch": 0.30246, + "grad_norm": 1.145376409795685, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 30246 + }, + { + "epoch": 0.30247, + "grad_norm": 0.9219079755562846, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 30247 + }, + { + "epoch": 0.30248, + "grad_norm": 0.9352559231134049, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 30248 + }, + { + "epoch": 0.30249, + "grad_norm": 1.081408088955862, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 30249 + }, + { + "epoch": 0.3025, + "grad_norm": 1.1312946895675493, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 30250 + }, + { + "epoch": 0.30251, + "grad_norm": 1.0043428095101252, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 30251 + }, + { + "epoch": 0.30252, + "grad_norm": 1.5170136836726469, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 30252 + }, + { + "epoch": 0.30253, + "grad_norm": 1.0502736950526625, + "learning_rate": 0.003, + "loss": 4.039, + "step": 30253 + }, + { + "epoch": 0.30254, + "grad_norm": 1.1261525340444385, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 30254 + }, + { + "epoch": 0.30255, + "grad_norm": 1.498562788421431, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 30255 + }, + { + "epoch": 0.30256, + "grad_norm": 0.9286882004430491, + "learning_rate": 0.003, + "loss": 4.066, + "step": 30256 + }, + { + "epoch": 0.30257, + "grad_norm": 0.9249933124348515, + "learning_rate": 0.003, + "loss": 4.052, + "step": 30257 + }, + { + "epoch": 0.30258, + "grad_norm": 1.3376674008388616, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 30258 + }, + { + "epoch": 0.30259, + "grad_norm": 1.2067502467198241, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 30259 + }, + { + "epoch": 0.3026, + "grad_norm": 1.1176691335562745, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 30260 + }, + { + "epoch": 0.30261, + "grad_norm": 1.0816585821300706, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 30261 + }, + { + "epoch": 0.30262, + "grad_norm": 0.8854878742704033, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 30262 + }, + { + "epoch": 0.30263, + "grad_norm": 1.0088215497566624, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 30263 + }, + { + "epoch": 0.30264, + "grad_norm": 1.1022956415887109, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 30264 + }, + { + "epoch": 0.30265, + "grad_norm": 0.8665169699760473, + "learning_rate": 0.003, + "loss": 4.097, + "step": 30265 + }, + { + "epoch": 0.30266, + "grad_norm": 0.7985642853207682, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 30266 + }, + { + "epoch": 0.30267, + "grad_norm": 0.9351609003277931, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 30267 + }, + { + "epoch": 0.30268, + "grad_norm": 0.9834448374234167, + "learning_rate": 0.003, + "loss": 4.076, + "step": 30268 + }, + { + "epoch": 0.30269, + "grad_norm": 1.2936172161341923, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 30269 + }, + { + "epoch": 0.3027, + "grad_norm": 1.0364770698290207, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 30270 + }, + { + "epoch": 0.30271, + "grad_norm": 1.2584216681242248, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 30271 + }, + { + "epoch": 0.30272, + "grad_norm": 0.9460781552286397, + "learning_rate": 0.003, + "loss": 4.0991, + "step": 30272 + }, + { + "epoch": 0.30273, + "grad_norm": 1.3130233217313065, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 30273 + }, + { + "epoch": 0.30274, + "grad_norm": 1.1827655976312785, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 30274 + }, + { + "epoch": 0.30275, + "grad_norm": 0.9295118749241327, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 30275 + }, + { + "epoch": 0.30276, + "grad_norm": 1.1147205532225564, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 30276 + }, + { + "epoch": 0.30277, + "grad_norm": 1.3493065599836314, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 30277 + }, + { + "epoch": 0.30278, + "grad_norm": 0.7015517114203431, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 30278 + }, + { + "epoch": 0.30279, + "grad_norm": 0.6861364488948158, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 30279 + }, + { + "epoch": 0.3028, + "grad_norm": 0.7569772412393646, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 30280 + }, + { + "epoch": 0.30281, + "grad_norm": 0.8138334953926168, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 30281 + }, + { + "epoch": 0.30282, + "grad_norm": 0.8534883281387481, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 30282 + }, + { + "epoch": 0.30283, + "grad_norm": 1.2013186663169362, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 30283 + }, + { + "epoch": 0.30284, + "grad_norm": 1.1112917247309226, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 30284 + }, + { + "epoch": 0.30285, + "grad_norm": 1.1329086703030926, + "learning_rate": 0.003, + "loss": 4.065, + "step": 30285 + }, + { + "epoch": 0.30286, + "grad_norm": 1.1773603059904967, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 30286 + }, + { + "epoch": 0.30287, + "grad_norm": 1.1776262173595464, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 30287 + }, + { + "epoch": 0.30288, + "grad_norm": 0.9109305169940162, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 30288 + }, + { + "epoch": 0.30289, + "grad_norm": 0.9563942247534718, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 30289 + }, + { + "epoch": 0.3029, + "grad_norm": 1.3136513618769985, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 30290 + }, + { + "epoch": 0.30291, + "grad_norm": 0.9310130898322987, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 30291 + }, + { + "epoch": 0.30292, + "grad_norm": 0.8820457679552767, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 30292 + }, + { + "epoch": 0.30293, + "grad_norm": 0.8918280034123334, + "learning_rate": 0.003, + "loss": 4.046, + "step": 30293 + }, + { + "epoch": 0.30294, + "grad_norm": 1.0440123678267859, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 30294 + }, + { + "epoch": 0.30295, + "grad_norm": 1.3116007030882875, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 30295 + }, + { + "epoch": 0.30296, + "grad_norm": 0.9553496369833663, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 30296 + }, + { + "epoch": 0.30297, + "grad_norm": 1.1457560667675695, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 30297 + }, + { + "epoch": 0.30298, + "grad_norm": 1.0202153304896209, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 30298 + }, + { + "epoch": 0.30299, + "grad_norm": 0.9466211047046305, + "learning_rate": 0.003, + "loss": 4.057, + "step": 30299 + }, + { + "epoch": 0.303, + "grad_norm": 1.0289143537182615, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 30300 + }, + { + "epoch": 0.30301, + "grad_norm": 1.0810102964967752, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 30301 + }, + { + "epoch": 0.30302, + "grad_norm": 0.8870030833798794, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 30302 + }, + { + "epoch": 0.30303, + "grad_norm": 1.0342801401029462, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 30303 + }, + { + "epoch": 0.30304, + "grad_norm": 1.0603328963219723, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 30304 + }, + { + "epoch": 0.30305, + "grad_norm": 0.9490423981463911, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 30305 + }, + { + "epoch": 0.30306, + "grad_norm": 1.1380897293744077, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 30306 + }, + { + "epoch": 0.30307, + "grad_norm": 1.1780725876145373, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 30307 + }, + { + "epoch": 0.30308, + "grad_norm": 1.1074117977509044, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 30308 + }, + { + "epoch": 0.30309, + "grad_norm": 1.0735294720502557, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 30309 + }, + { + "epoch": 0.3031, + "grad_norm": 1.079912866560762, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 30310 + }, + { + "epoch": 0.30311, + "grad_norm": 0.9045066777302633, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 30311 + }, + { + "epoch": 0.30312, + "grad_norm": 0.8925177116538712, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 30312 + }, + { + "epoch": 0.30313, + "grad_norm": 1.1177386530783728, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 30313 + }, + { + "epoch": 0.30314, + "grad_norm": 1.0381998415570548, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 30314 + }, + { + "epoch": 0.30315, + "grad_norm": 0.855003573083474, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 30315 + }, + { + "epoch": 0.30316, + "grad_norm": 0.9239351759918542, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 30316 + }, + { + "epoch": 0.30317, + "grad_norm": 1.0972212232313974, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 30317 + }, + { + "epoch": 0.30318, + "grad_norm": 0.9747867645614446, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 30318 + }, + { + "epoch": 0.30319, + "grad_norm": 0.9976890465906224, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 30319 + }, + { + "epoch": 0.3032, + "grad_norm": 1.1204993340311382, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 30320 + }, + { + "epoch": 0.30321, + "grad_norm": 1.1421547633422025, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 30321 + }, + { + "epoch": 0.30322, + "grad_norm": 1.1917021820821052, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 30322 + }, + { + "epoch": 0.30323, + "grad_norm": 1.1380519338989459, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 30323 + }, + { + "epoch": 0.30324, + "grad_norm": 1.2105818917521145, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 30324 + }, + { + "epoch": 0.30325, + "grad_norm": 1.1002194315122027, + "learning_rate": 0.003, + "loss": 4.054, + "step": 30325 + }, + { + "epoch": 0.30326, + "grad_norm": 1.1764575343939287, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 30326 + }, + { + "epoch": 0.30327, + "grad_norm": 1.1003792937832222, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 30327 + }, + { + "epoch": 0.30328, + "grad_norm": 1.0945262522822636, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 30328 + }, + { + "epoch": 0.30329, + "grad_norm": 1.0523884754282073, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 30329 + }, + { + "epoch": 0.3033, + "grad_norm": 1.1314987377989256, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 30330 + }, + { + "epoch": 0.30331, + "grad_norm": 1.025818085695948, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 30331 + }, + { + "epoch": 0.30332, + "grad_norm": 1.1194363522283497, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 30332 + }, + { + "epoch": 0.30333, + "grad_norm": 0.9622595013427874, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 30333 + }, + { + "epoch": 0.30334, + "grad_norm": 1.04205342928563, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 30334 + }, + { + "epoch": 0.30335, + "grad_norm": 1.4117427137346792, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 30335 + }, + { + "epoch": 0.30336, + "grad_norm": 0.7495007297037893, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 30336 + }, + { + "epoch": 0.30337, + "grad_norm": 0.9113356433678218, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 30337 + }, + { + "epoch": 0.30338, + "grad_norm": 1.032928682139034, + "learning_rate": 0.003, + "loss": 4.0856, + "step": 30338 + }, + { + "epoch": 0.30339, + "grad_norm": 1.2115741117298362, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 30339 + }, + { + "epoch": 0.3034, + "grad_norm": 1.315988280084033, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 30340 + }, + { + "epoch": 0.30341, + "grad_norm": 1.095036159783137, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 30341 + }, + { + "epoch": 0.30342, + "grad_norm": 1.2454449470171134, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 30342 + }, + { + "epoch": 0.30343, + "grad_norm": 1.0918799133170065, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 30343 + }, + { + "epoch": 0.30344, + "grad_norm": 1.186054352131005, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 30344 + }, + { + "epoch": 0.30345, + "grad_norm": 0.9325969388268784, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 30345 + }, + { + "epoch": 0.30346, + "grad_norm": 1.023742826007479, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 30346 + }, + { + "epoch": 0.30347, + "grad_norm": 1.1920444486034683, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 30347 + }, + { + "epoch": 0.30348, + "grad_norm": 0.9389076480666501, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 30348 + }, + { + "epoch": 0.30349, + "grad_norm": 0.9636949026133552, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 30349 + }, + { + "epoch": 0.3035, + "grad_norm": 1.0517361768801146, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 30350 + }, + { + "epoch": 0.30351, + "grad_norm": 1.0513005659032792, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 30351 + }, + { + "epoch": 0.30352, + "grad_norm": 1.0525797720887968, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 30352 + }, + { + "epoch": 0.30353, + "grad_norm": 1.1547757547341908, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 30353 + }, + { + "epoch": 0.30354, + "grad_norm": 1.101135839550288, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 30354 + }, + { + "epoch": 0.30355, + "grad_norm": 0.9912696362040494, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 30355 + }, + { + "epoch": 0.30356, + "grad_norm": 1.1212620620485034, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 30356 + }, + { + "epoch": 0.30357, + "grad_norm": 0.9682504162802814, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 30357 + }, + { + "epoch": 0.30358, + "grad_norm": 1.200164400617681, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 30358 + }, + { + "epoch": 0.30359, + "grad_norm": 0.9564011735508393, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 30359 + }, + { + "epoch": 0.3036, + "grad_norm": 1.0366754218590053, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 30360 + }, + { + "epoch": 0.30361, + "grad_norm": 1.2535847755014824, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 30361 + }, + { + "epoch": 0.30362, + "grad_norm": 0.972520369320125, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 30362 + }, + { + "epoch": 0.30363, + "grad_norm": 1.109587820315968, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 30363 + }, + { + "epoch": 0.30364, + "grad_norm": 1.1822249363455326, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 30364 + }, + { + "epoch": 0.30365, + "grad_norm": 1.0820557488522793, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 30365 + }, + { + "epoch": 0.30366, + "grad_norm": 1.2088796677725924, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 30366 + }, + { + "epoch": 0.30367, + "grad_norm": 1.0022683501240233, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 30367 + }, + { + "epoch": 0.30368, + "grad_norm": 1.165506745355414, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 30368 + }, + { + "epoch": 0.30369, + "grad_norm": 0.7672898332263036, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 30369 + }, + { + "epoch": 0.3037, + "grad_norm": 0.7536532476532641, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 30370 + }, + { + "epoch": 0.30371, + "grad_norm": 0.905048260593655, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 30371 + }, + { + "epoch": 0.30372, + "grad_norm": 1.2236118450547533, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 30372 + }, + { + "epoch": 0.30373, + "grad_norm": 1.1257839201555764, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 30373 + }, + { + "epoch": 0.30374, + "grad_norm": 1.0123903906798433, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 30374 + }, + { + "epoch": 0.30375, + "grad_norm": 1.0319020904521117, + "learning_rate": 0.003, + "loss": 4.039, + "step": 30375 + }, + { + "epoch": 0.30376, + "grad_norm": 1.2178106873260894, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 30376 + }, + { + "epoch": 0.30377, + "grad_norm": 1.0077809150534978, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 30377 + }, + { + "epoch": 0.30378, + "grad_norm": 1.114448682413398, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 30378 + }, + { + "epoch": 0.30379, + "grad_norm": 1.0810482028582642, + "learning_rate": 0.003, + "loss": 4.041, + "step": 30379 + }, + { + "epoch": 0.3038, + "grad_norm": 1.0848061218531593, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 30380 + }, + { + "epoch": 0.30381, + "grad_norm": 1.2018619539821787, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 30381 + }, + { + "epoch": 0.30382, + "grad_norm": 0.9888081945218341, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 30382 + }, + { + "epoch": 0.30383, + "grad_norm": 1.0644722069323354, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 30383 + }, + { + "epoch": 0.30384, + "grad_norm": 1.0906069155963254, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 30384 + }, + { + "epoch": 0.30385, + "grad_norm": 1.23029272527647, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 30385 + }, + { + "epoch": 0.30386, + "grad_norm": 1.1422598963982142, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 30386 + }, + { + "epoch": 0.30387, + "grad_norm": 1.1377542745298275, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 30387 + }, + { + "epoch": 0.30388, + "grad_norm": 1.018539742911096, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 30388 + }, + { + "epoch": 0.30389, + "grad_norm": 0.9717288722975393, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 30389 + }, + { + "epoch": 0.3039, + "grad_norm": 1.0661498500272304, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 30390 + }, + { + "epoch": 0.30391, + "grad_norm": 1.0854686534959463, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 30391 + }, + { + "epoch": 0.30392, + "grad_norm": 1.0951470775766614, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 30392 + }, + { + "epoch": 0.30393, + "grad_norm": 0.9230587319915085, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 30393 + }, + { + "epoch": 0.30394, + "grad_norm": 0.9722116845033133, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 30394 + }, + { + "epoch": 0.30395, + "grad_norm": 1.1372569244435438, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 30395 + }, + { + "epoch": 0.30396, + "grad_norm": 1.0388487206633104, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 30396 + }, + { + "epoch": 0.30397, + "grad_norm": 1.0406097388268785, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 30397 + }, + { + "epoch": 0.30398, + "grad_norm": 0.9584266050963852, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 30398 + }, + { + "epoch": 0.30399, + "grad_norm": 1.2027121518435082, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 30399 + }, + { + "epoch": 0.304, + "grad_norm": 1.2884432250173903, + "learning_rate": 0.003, + "loss": 4.04, + "step": 30400 + }, + { + "epoch": 0.30401, + "grad_norm": 1.311615393631646, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 30401 + }, + { + "epoch": 0.30402, + "grad_norm": 0.847367653694476, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 30402 + }, + { + "epoch": 0.30403, + "grad_norm": 0.8516283327996628, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 30403 + }, + { + "epoch": 0.30404, + "grad_norm": 0.9888312949725602, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 30404 + }, + { + "epoch": 0.30405, + "grad_norm": 1.4274624512257876, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 30405 + }, + { + "epoch": 0.30406, + "grad_norm": 0.9356428402088864, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 30406 + }, + { + "epoch": 0.30407, + "grad_norm": 1.0464000162976317, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 30407 + }, + { + "epoch": 0.30408, + "grad_norm": 1.0020521215359917, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 30408 + }, + { + "epoch": 0.30409, + "grad_norm": 1.0806953495326754, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 30409 + }, + { + "epoch": 0.3041, + "grad_norm": 1.1751808024001842, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 30410 + }, + { + "epoch": 0.30411, + "grad_norm": 0.9444846174009358, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 30411 + }, + { + "epoch": 0.30412, + "grad_norm": 1.1483415229652478, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 30412 + }, + { + "epoch": 0.30413, + "grad_norm": 1.149643597017814, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 30413 + }, + { + "epoch": 0.30414, + "grad_norm": 1.2223256849905153, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 30414 + }, + { + "epoch": 0.30415, + "grad_norm": 0.8783951588487505, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 30415 + }, + { + "epoch": 0.30416, + "grad_norm": 0.9401936161465178, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 30416 + }, + { + "epoch": 0.30417, + "grad_norm": 1.0635310115180165, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 30417 + }, + { + "epoch": 0.30418, + "grad_norm": 1.022992444038248, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 30418 + }, + { + "epoch": 0.30419, + "grad_norm": 1.146123484455338, + "learning_rate": 0.003, + "loss": 4.035, + "step": 30419 + }, + { + "epoch": 0.3042, + "grad_norm": 1.1092653978327207, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 30420 + }, + { + "epoch": 0.30421, + "grad_norm": 1.045171098862859, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 30421 + }, + { + "epoch": 0.30422, + "grad_norm": 1.1709234140656575, + "learning_rate": 0.003, + "loss": 4.047, + "step": 30422 + }, + { + "epoch": 0.30423, + "grad_norm": 1.0270254923564528, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 30423 + }, + { + "epoch": 0.30424, + "grad_norm": 1.100377063018833, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 30424 + }, + { + "epoch": 0.30425, + "grad_norm": 1.0087475281806124, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 30425 + }, + { + "epoch": 0.30426, + "grad_norm": 1.2784705117898176, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 30426 + }, + { + "epoch": 0.30427, + "grad_norm": 0.9579436642300803, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 30427 + }, + { + "epoch": 0.30428, + "grad_norm": 1.1087581187595317, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 30428 + }, + { + "epoch": 0.30429, + "grad_norm": 1.0522121384174392, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 30429 + }, + { + "epoch": 0.3043, + "grad_norm": 1.0412067194190728, + "learning_rate": 0.003, + "loss": 4.04, + "step": 30430 + }, + { + "epoch": 0.30431, + "grad_norm": 1.073894859056245, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 30431 + }, + { + "epoch": 0.30432, + "grad_norm": 0.977381239014256, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 30432 + }, + { + "epoch": 0.30433, + "grad_norm": 1.0562827956330822, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 30433 + }, + { + "epoch": 0.30434, + "grad_norm": 0.9721605496668575, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 30434 + }, + { + "epoch": 0.30435, + "grad_norm": 1.0334328383320868, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 30435 + }, + { + "epoch": 0.30436, + "grad_norm": 1.0365126671639036, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 30436 + }, + { + "epoch": 0.30437, + "grad_norm": 1.0728488674855172, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 30437 + }, + { + "epoch": 0.30438, + "grad_norm": 1.139865284783114, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 30438 + }, + { + "epoch": 0.30439, + "grad_norm": 1.027521671968784, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 30439 + }, + { + "epoch": 0.3044, + "grad_norm": 0.9850214575073486, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 30440 + }, + { + "epoch": 0.30441, + "grad_norm": 0.9689619096180261, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 30441 + }, + { + "epoch": 0.30442, + "grad_norm": 1.0741348463987221, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 30442 + }, + { + "epoch": 0.30443, + "grad_norm": 1.032593233352022, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 30443 + }, + { + "epoch": 0.30444, + "grad_norm": 1.050251123783945, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 30444 + }, + { + "epoch": 0.30445, + "grad_norm": 1.358985740484578, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 30445 + }, + { + "epoch": 0.30446, + "grad_norm": 0.9025027433184625, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 30446 + }, + { + "epoch": 0.30447, + "grad_norm": 1.1368541195405168, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 30447 + }, + { + "epoch": 0.30448, + "grad_norm": 1.1749478194164715, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 30448 + }, + { + "epoch": 0.30449, + "grad_norm": 1.192788227443152, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 30449 + }, + { + "epoch": 0.3045, + "grad_norm": 1.1085657554308184, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 30450 + }, + { + "epoch": 0.30451, + "grad_norm": 1.0606892089251652, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 30451 + }, + { + "epoch": 0.30452, + "grad_norm": 1.0416244916061215, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 30452 + }, + { + "epoch": 0.30453, + "grad_norm": 1.0610685539073321, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 30453 + }, + { + "epoch": 0.30454, + "grad_norm": 1.1316160506123167, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 30454 + }, + { + "epoch": 0.30455, + "grad_norm": 1.0397237067272176, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 30455 + }, + { + "epoch": 0.30456, + "grad_norm": 1.0228133779303445, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 30456 + }, + { + "epoch": 0.30457, + "grad_norm": 1.2104430503827384, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 30457 + }, + { + "epoch": 0.30458, + "grad_norm": 1.1105666852456955, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 30458 + }, + { + "epoch": 0.30459, + "grad_norm": 1.2542652402146381, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 30459 + }, + { + "epoch": 0.3046, + "grad_norm": 0.9245833153413253, + "learning_rate": 0.003, + "loss": 4.009, + "step": 30460 + }, + { + "epoch": 0.30461, + "grad_norm": 0.9072233118260092, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 30461 + }, + { + "epoch": 0.30462, + "grad_norm": 1.2595427207424739, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 30462 + }, + { + "epoch": 0.30463, + "grad_norm": 1.2380543226401108, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 30463 + }, + { + "epoch": 0.30464, + "grad_norm": 0.8918998660090336, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 30464 + }, + { + "epoch": 0.30465, + "grad_norm": 0.8713382796973171, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 30465 + }, + { + "epoch": 0.30466, + "grad_norm": 0.9191057816791182, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 30466 + }, + { + "epoch": 0.30467, + "grad_norm": 0.8908377908765204, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 30467 + }, + { + "epoch": 0.30468, + "grad_norm": 1.0664256071106288, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 30468 + }, + { + "epoch": 0.30469, + "grad_norm": 1.3698286265028752, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 30469 + }, + { + "epoch": 0.3047, + "grad_norm": 0.8783819092832725, + "learning_rate": 0.003, + "loss": 4.045, + "step": 30470 + }, + { + "epoch": 0.30471, + "grad_norm": 0.9424774143926616, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 30471 + }, + { + "epoch": 0.30472, + "grad_norm": 1.2460264328023674, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 30472 + }, + { + "epoch": 0.30473, + "grad_norm": 0.8372687432909803, + "learning_rate": 0.003, + "loss": 4.035, + "step": 30473 + }, + { + "epoch": 0.30474, + "grad_norm": 0.8531275674710624, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 30474 + }, + { + "epoch": 0.30475, + "grad_norm": 0.9843638762676862, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 30475 + }, + { + "epoch": 0.30476, + "grad_norm": 1.1735211382158348, + "learning_rate": 0.003, + "loss": 4.027, + "step": 30476 + }, + { + "epoch": 0.30477, + "grad_norm": 1.1071431364239581, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 30477 + }, + { + "epoch": 0.30478, + "grad_norm": 1.207405092672969, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 30478 + }, + { + "epoch": 0.30479, + "grad_norm": 0.9915565923799083, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 30479 + }, + { + "epoch": 0.3048, + "grad_norm": 1.0199489070203591, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 30480 + }, + { + "epoch": 0.30481, + "grad_norm": 1.1100103563533725, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 30481 + }, + { + "epoch": 0.30482, + "grad_norm": 1.04994125418937, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 30482 + }, + { + "epoch": 0.30483, + "grad_norm": 1.2946560797755466, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 30483 + }, + { + "epoch": 0.30484, + "grad_norm": 0.8131225891637383, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 30484 + }, + { + "epoch": 0.30485, + "grad_norm": 0.8382000078182112, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 30485 + }, + { + "epoch": 0.30486, + "grad_norm": 0.9786503253455021, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 30486 + }, + { + "epoch": 0.30487, + "grad_norm": 1.061588178880223, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 30487 + }, + { + "epoch": 0.30488, + "grad_norm": 0.969105139037679, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 30488 + }, + { + "epoch": 0.30489, + "grad_norm": 1.12406181523246, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 30489 + }, + { + "epoch": 0.3049, + "grad_norm": 1.2368922563032476, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 30490 + }, + { + "epoch": 0.30491, + "grad_norm": 0.9974099684707606, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 30491 + }, + { + "epoch": 0.30492, + "grad_norm": 1.1406598243754558, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 30492 + }, + { + "epoch": 0.30493, + "grad_norm": 1.0790013592185599, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 30493 + }, + { + "epoch": 0.30494, + "grad_norm": 1.2358588229624212, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 30494 + }, + { + "epoch": 0.30495, + "grad_norm": 0.9455586882067857, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 30495 + }, + { + "epoch": 0.30496, + "grad_norm": 0.9630882396085011, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 30496 + }, + { + "epoch": 0.30497, + "grad_norm": 1.1156506655359413, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 30497 + }, + { + "epoch": 0.30498, + "grad_norm": 0.9282015532411302, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 30498 + }, + { + "epoch": 0.30499, + "grad_norm": 0.9599792959631057, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 30499 + }, + { + "epoch": 0.305, + "grad_norm": 1.1711383354014342, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 30500 + }, + { + "epoch": 0.30501, + "grad_norm": 0.9867371647718833, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 30501 + }, + { + "epoch": 0.30502, + "grad_norm": 1.2028278518567663, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 30502 + }, + { + "epoch": 0.30503, + "grad_norm": 0.9192655673672407, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 30503 + }, + { + "epoch": 0.30504, + "grad_norm": 1.019061559126934, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 30504 + }, + { + "epoch": 0.30505, + "grad_norm": 1.2597124901077483, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 30505 + }, + { + "epoch": 0.30506, + "grad_norm": 1.2917408466194145, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 30506 + }, + { + "epoch": 0.30507, + "grad_norm": 0.9206814130782586, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 30507 + }, + { + "epoch": 0.30508, + "grad_norm": 1.0024046482876992, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 30508 + }, + { + "epoch": 0.30509, + "grad_norm": 1.382279289481995, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 30509 + }, + { + "epoch": 0.3051, + "grad_norm": 0.9656812337125492, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 30510 + }, + { + "epoch": 0.30511, + "grad_norm": 1.0377865695358885, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 30511 + }, + { + "epoch": 0.30512, + "grad_norm": 1.2524283116253438, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 30512 + }, + { + "epoch": 0.30513, + "grad_norm": 0.98004721864189, + "learning_rate": 0.003, + "loss": 4.034, + "step": 30513 + }, + { + "epoch": 0.30514, + "grad_norm": 1.0771680437847977, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 30514 + }, + { + "epoch": 0.30515, + "grad_norm": 1.1810557477967794, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 30515 + }, + { + "epoch": 0.30516, + "grad_norm": 0.9927457814990783, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 30516 + }, + { + "epoch": 0.30517, + "grad_norm": 0.9977798633249453, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 30517 + }, + { + "epoch": 0.30518, + "grad_norm": 1.0038320304298602, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 30518 + }, + { + "epoch": 0.30519, + "grad_norm": 1.1642262683286226, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 30519 + }, + { + "epoch": 0.3052, + "grad_norm": 1.0514237317385209, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 30520 + }, + { + "epoch": 0.30521, + "grad_norm": 1.0879548618247672, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 30521 + }, + { + "epoch": 0.30522, + "grad_norm": 1.0275886254885171, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 30522 + }, + { + "epoch": 0.30523, + "grad_norm": 1.1454236072516575, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 30523 + }, + { + "epoch": 0.30524, + "grad_norm": 0.9572303461666369, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 30524 + }, + { + "epoch": 0.30525, + "grad_norm": 0.9465976102584759, + "learning_rate": 0.003, + "loss": 4.04, + "step": 30525 + }, + { + "epoch": 0.30526, + "grad_norm": 1.072606621325838, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 30526 + }, + { + "epoch": 0.30527, + "grad_norm": 1.16623234198557, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 30527 + }, + { + "epoch": 0.30528, + "grad_norm": 1.0515273247548105, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 30528 + }, + { + "epoch": 0.30529, + "grad_norm": 1.2764194439921068, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 30529 + }, + { + "epoch": 0.3053, + "grad_norm": 1.0004677674376385, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 30530 + }, + { + "epoch": 0.30531, + "grad_norm": 1.2441714539964661, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 30531 + }, + { + "epoch": 0.30532, + "grad_norm": 1.0294198697795094, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 30532 + }, + { + "epoch": 0.30533, + "grad_norm": 0.9915247141276678, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 30533 + }, + { + "epoch": 0.30534, + "grad_norm": 0.9383650656733137, + "learning_rate": 0.003, + "loss": 4.006, + "step": 30534 + }, + { + "epoch": 0.30535, + "grad_norm": 0.9733656400515208, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 30535 + }, + { + "epoch": 0.30536, + "grad_norm": 1.344006999806203, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 30536 + }, + { + "epoch": 0.30537, + "grad_norm": 0.8725428753686821, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 30537 + }, + { + "epoch": 0.30538, + "grad_norm": 0.9282978833704557, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 30538 + }, + { + "epoch": 0.30539, + "grad_norm": 1.1597468816308634, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 30539 + }, + { + "epoch": 0.3054, + "grad_norm": 1.146716311323764, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 30540 + }, + { + "epoch": 0.30541, + "grad_norm": 1.0804177299937074, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 30541 + }, + { + "epoch": 0.30542, + "grad_norm": 0.9647026325607819, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 30542 + }, + { + "epoch": 0.30543, + "grad_norm": 1.331519619692959, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 30543 + }, + { + "epoch": 0.30544, + "grad_norm": 0.896034046436172, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 30544 + }, + { + "epoch": 0.30545, + "grad_norm": 0.8505608544043247, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 30545 + }, + { + "epoch": 0.30546, + "grad_norm": 0.9829955549019783, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 30546 + }, + { + "epoch": 0.30547, + "grad_norm": 1.4719528143236948, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 30547 + }, + { + "epoch": 0.30548, + "grad_norm": 0.730529947762878, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 30548 + }, + { + "epoch": 0.30549, + "grad_norm": 0.7896909866628387, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 30549 + }, + { + "epoch": 0.3055, + "grad_norm": 0.8905052610541337, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 30550 + }, + { + "epoch": 0.30551, + "grad_norm": 1.0981318274017864, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 30551 + }, + { + "epoch": 0.30552, + "grad_norm": 1.1577187887354745, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 30552 + }, + { + "epoch": 0.30553, + "grad_norm": 1.1570148240128562, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 30553 + }, + { + "epoch": 0.30554, + "grad_norm": 1.086987793425318, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 30554 + }, + { + "epoch": 0.30555, + "grad_norm": 1.1117017819583124, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 30555 + }, + { + "epoch": 0.30556, + "grad_norm": 1.2051278668650445, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 30556 + }, + { + "epoch": 0.30557, + "grad_norm": 1.0529527335250182, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 30557 + }, + { + "epoch": 0.30558, + "grad_norm": 1.0907689786380872, + "learning_rate": 0.003, + "loss": 4.057, + "step": 30558 + }, + { + "epoch": 0.30559, + "grad_norm": 1.121537294387749, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 30559 + }, + { + "epoch": 0.3056, + "grad_norm": 1.2009778933815825, + "learning_rate": 0.003, + "loss": 4.0934, + "step": 30560 + }, + { + "epoch": 0.30561, + "grad_norm": 1.019935938417264, + "learning_rate": 0.003, + "loss": 4.049, + "step": 30561 + }, + { + "epoch": 0.30562, + "grad_norm": 1.051708831382468, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 30562 + }, + { + "epoch": 0.30563, + "grad_norm": 1.1400451156519515, + "learning_rate": 0.003, + "loss": 4.043, + "step": 30563 + }, + { + "epoch": 0.30564, + "grad_norm": 0.9696246324048103, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 30564 + }, + { + "epoch": 0.30565, + "grad_norm": 1.0234262709022, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 30565 + }, + { + "epoch": 0.30566, + "grad_norm": 0.9644846374840682, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 30566 + }, + { + "epoch": 0.30567, + "grad_norm": 0.9312565504840613, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 30567 + }, + { + "epoch": 0.30568, + "grad_norm": 1.130766015694547, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 30568 + }, + { + "epoch": 0.30569, + "grad_norm": 0.9076289399745278, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 30569 + }, + { + "epoch": 0.3057, + "grad_norm": 1.0530329693054719, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 30570 + }, + { + "epoch": 0.30571, + "grad_norm": 1.0431509269590515, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 30571 + }, + { + "epoch": 0.30572, + "grad_norm": 1.0151384414628495, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 30572 + }, + { + "epoch": 0.30573, + "grad_norm": 1.1302356345747928, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 30573 + }, + { + "epoch": 0.30574, + "grad_norm": 1.2478389685822788, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 30574 + }, + { + "epoch": 0.30575, + "grad_norm": 1.3207023335898445, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 30575 + }, + { + "epoch": 0.30576, + "grad_norm": 0.9081236318683575, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 30576 + }, + { + "epoch": 0.30577, + "grad_norm": 0.9526649750046916, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 30577 + }, + { + "epoch": 0.30578, + "grad_norm": 1.1578268783640882, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 30578 + }, + { + "epoch": 0.30579, + "grad_norm": 1.0149277495341662, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 30579 + }, + { + "epoch": 0.3058, + "grad_norm": 1.4313957736254477, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 30580 + }, + { + "epoch": 0.30581, + "grad_norm": 0.8983451028415044, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 30581 + }, + { + "epoch": 0.30582, + "grad_norm": 0.803884352771098, + "learning_rate": 0.003, + "loss": 4.045, + "step": 30582 + }, + { + "epoch": 0.30583, + "grad_norm": 0.8383057205099089, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 30583 + }, + { + "epoch": 0.30584, + "grad_norm": 0.9813923466579688, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 30584 + }, + { + "epoch": 0.30585, + "grad_norm": 1.469783669421222, + "learning_rate": 0.003, + "loss": 4.057, + "step": 30585 + }, + { + "epoch": 0.30586, + "grad_norm": 0.8659844810853655, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 30586 + }, + { + "epoch": 0.30587, + "grad_norm": 1.0861494382222372, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 30587 + }, + { + "epoch": 0.30588, + "grad_norm": 1.2475357101059714, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 30588 + }, + { + "epoch": 0.30589, + "grad_norm": 0.8890891797761767, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 30589 + }, + { + "epoch": 0.3059, + "grad_norm": 1.14517231732202, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 30590 + }, + { + "epoch": 0.30591, + "grad_norm": 1.1542728979854284, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 30591 + }, + { + "epoch": 0.30592, + "grad_norm": 0.9284006745827619, + "learning_rate": 0.003, + "loss": 4.036, + "step": 30592 + }, + { + "epoch": 0.30593, + "grad_norm": 1.096833335932734, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 30593 + }, + { + "epoch": 0.30594, + "grad_norm": 1.4801975049559306, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 30594 + }, + { + "epoch": 0.30595, + "grad_norm": 0.7095845399027192, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 30595 + }, + { + "epoch": 0.30596, + "grad_norm": 0.66897127597773, + "learning_rate": 0.003, + "loss": 4.071, + "step": 30596 + }, + { + "epoch": 0.30597, + "grad_norm": 0.6969688196539284, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 30597 + }, + { + "epoch": 0.30598, + "grad_norm": 0.9093096873826184, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 30598 + }, + { + "epoch": 0.30599, + "grad_norm": 1.3220328123821403, + "learning_rate": 0.003, + "loss": 4.056, + "step": 30599 + }, + { + "epoch": 0.306, + "grad_norm": 1.1378375476651137, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 30600 + }, + { + "epoch": 0.30601, + "grad_norm": 1.0719524766611128, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 30601 + }, + { + "epoch": 0.30602, + "grad_norm": 1.0366050285898942, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 30602 + }, + { + "epoch": 0.30603, + "grad_norm": 1.5203691826684622, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 30603 + }, + { + "epoch": 0.30604, + "grad_norm": 0.8300611437773836, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 30604 + }, + { + "epoch": 0.30605, + "grad_norm": 0.8521316461318635, + "learning_rate": 0.003, + "loss": 4.062, + "step": 30605 + }, + { + "epoch": 0.30606, + "grad_norm": 1.036886967834563, + "learning_rate": 0.003, + "loss": 4.0873, + "step": 30606 + }, + { + "epoch": 0.30607, + "grad_norm": 1.3419210047835075, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 30607 + }, + { + "epoch": 0.30608, + "grad_norm": 0.9892606653445759, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 30608 + }, + { + "epoch": 0.30609, + "grad_norm": 1.0545195741942122, + "learning_rate": 0.003, + "loss": 4.072, + "step": 30609 + }, + { + "epoch": 0.3061, + "grad_norm": 0.8633736490343505, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 30610 + }, + { + "epoch": 0.30611, + "grad_norm": 0.9688379565041904, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 30611 + }, + { + "epoch": 0.30612, + "grad_norm": 1.1492291709511289, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 30612 + }, + { + "epoch": 0.30613, + "grad_norm": 1.1325264193723144, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 30613 + }, + { + "epoch": 0.30614, + "grad_norm": 1.3321005824764247, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 30614 + }, + { + "epoch": 0.30615, + "grad_norm": 1.068839792727922, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 30615 + }, + { + "epoch": 0.30616, + "grad_norm": 1.0996174202213564, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 30616 + }, + { + "epoch": 0.30617, + "grad_norm": 1.0807183846856168, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 30617 + }, + { + "epoch": 0.30618, + "grad_norm": 1.1815913884236673, + "learning_rate": 0.003, + "loss": 4.032, + "step": 30618 + }, + { + "epoch": 0.30619, + "grad_norm": 1.0378072820050896, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 30619 + }, + { + "epoch": 0.3062, + "grad_norm": 1.2459216018824515, + "learning_rate": 0.003, + "loss": 4.029, + "step": 30620 + }, + { + "epoch": 0.30621, + "grad_norm": 0.9117532356511944, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 30621 + }, + { + "epoch": 0.30622, + "grad_norm": 1.0180072560203628, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 30622 + }, + { + "epoch": 0.30623, + "grad_norm": 1.1198687824684967, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 30623 + }, + { + "epoch": 0.30624, + "grad_norm": 1.0366087008069387, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 30624 + }, + { + "epoch": 0.30625, + "grad_norm": 1.0017417100907258, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 30625 + }, + { + "epoch": 0.30626, + "grad_norm": 1.2840398713602397, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 30626 + }, + { + "epoch": 0.30627, + "grad_norm": 1.0257172870343219, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 30627 + }, + { + "epoch": 0.30628, + "grad_norm": 1.1775064459705933, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 30628 + }, + { + "epoch": 0.30629, + "grad_norm": 1.1612464927946717, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 30629 + }, + { + "epoch": 0.3063, + "grad_norm": 0.9719693797331322, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 30630 + }, + { + "epoch": 0.30631, + "grad_norm": 1.350942762617211, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 30631 + }, + { + "epoch": 0.30632, + "grad_norm": 1.02900533476507, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 30632 + }, + { + "epoch": 0.30633, + "grad_norm": 1.258877742986612, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 30633 + }, + { + "epoch": 0.30634, + "grad_norm": 1.0610665283135445, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 30634 + }, + { + "epoch": 0.30635, + "grad_norm": 1.2413144151272526, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 30635 + }, + { + "epoch": 0.30636, + "grad_norm": 1.076105724492369, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 30636 + }, + { + "epoch": 0.30637, + "grad_norm": 1.299046244868638, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 30637 + }, + { + "epoch": 0.30638, + "grad_norm": 0.7840022825233836, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 30638 + }, + { + "epoch": 0.30639, + "grad_norm": 0.751735356537494, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 30639 + }, + { + "epoch": 0.3064, + "grad_norm": 0.8016760089388002, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 30640 + }, + { + "epoch": 0.30641, + "grad_norm": 1.0027407416385363, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 30641 + }, + { + "epoch": 0.30642, + "grad_norm": 1.4986222943601522, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 30642 + }, + { + "epoch": 0.30643, + "grad_norm": 0.7488749624858153, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 30643 + }, + { + "epoch": 0.30644, + "grad_norm": 0.8603703172022126, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 30644 + }, + { + "epoch": 0.30645, + "grad_norm": 0.9388772509171236, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 30645 + }, + { + "epoch": 0.30646, + "grad_norm": 1.1180065300252493, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 30646 + }, + { + "epoch": 0.30647, + "grad_norm": 1.4392780161902052, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 30647 + }, + { + "epoch": 0.30648, + "grad_norm": 0.9557157635374809, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 30648 + }, + { + "epoch": 0.30649, + "grad_norm": 0.867042238939909, + "learning_rate": 0.003, + "loss": 4.063, + "step": 30649 + }, + { + "epoch": 0.3065, + "grad_norm": 0.8973221150010638, + "learning_rate": 0.003, + "loss": 4.043, + "step": 30650 + }, + { + "epoch": 0.30651, + "grad_norm": 0.9971047522994158, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 30651 + }, + { + "epoch": 0.30652, + "grad_norm": 1.198278920789335, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 30652 + }, + { + "epoch": 0.30653, + "grad_norm": 0.9587300022648826, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 30653 + }, + { + "epoch": 0.30654, + "grad_norm": 1.1162361062376802, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 30654 + }, + { + "epoch": 0.30655, + "grad_norm": 1.3090684375494777, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 30655 + }, + { + "epoch": 0.30656, + "grad_norm": 1.0228741763637426, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 30656 + }, + { + "epoch": 0.30657, + "grad_norm": 1.1708174671910432, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 30657 + }, + { + "epoch": 0.30658, + "grad_norm": 0.9671586267858334, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 30658 + }, + { + "epoch": 0.30659, + "grad_norm": 1.1341285241454082, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 30659 + }, + { + "epoch": 0.3066, + "grad_norm": 1.125867409327303, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 30660 + }, + { + "epoch": 0.30661, + "grad_norm": 1.3788907776362165, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 30661 + }, + { + "epoch": 0.30662, + "grad_norm": 1.0602381489492985, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 30662 + }, + { + "epoch": 0.30663, + "grad_norm": 1.0519290310218208, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 30663 + }, + { + "epoch": 0.30664, + "grad_norm": 1.203958259086647, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 30664 + }, + { + "epoch": 0.30665, + "grad_norm": 1.0418220430720437, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 30665 + }, + { + "epoch": 0.30666, + "grad_norm": 1.065516432138485, + "learning_rate": 0.003, + "loss": 4.033, + "step": 30666 + }, + { + "epoch": 0.30667, + "grad_norm": 1.0158761615106866, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 30667 + }, + { + "epoch": 0.30668, + "grad_norm": 1.1321669684335776, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 30668 + }, + { + "epoch": 0.30669, + "grad_norm": 1.0681732895391316, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 30669 + }, + { + "epoch": 0.3067, + "grad_norm": 0.9332501987892512, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 30670 + }, + { + "epoch": 0.30671, + "grad_norm": 0.8637519375882785, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 30671 + }, + { + "epoch": 0.30672, + "grad_norm": 0.9387926893148005, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 30672 + }, + { + "epoch": 0.30673, + "grad_norm": 1.1911145355678534, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 30673 + }, + { + "epoch": 0.30674, + "grad_norm": 1.0332643753649087, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 30674 + }, + { + "epoch": 0.30675, + "grad_norm": 1.139687755334975, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 30675 + }, + { + "epoch": 0.30676, + "grad_norm": 1.1126209798044557, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 30676 + }, + { + "epoch": 0.30677, + "grad_norm": 1.2897789924381973, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 30677 + }, + { + "epoch": 0.30678, + "grad_norm": 0.9392118393756254, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 30678 + }, + { + "epoch": 0.30679, + "grad_norm": 1.010782725979931, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 30679 + }, + { + "epoch": 0.3068, + "grad_norm": 1.3096325566219982, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 30680 + }, + { + "epoch": 0.30681, + "grad_norm": 1.0935795626986886, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 30681 + }, + { + "epoch": 0.30682, + "grad_norm": 1.162409725269421, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 30682 + }, + { + "epoch": 0.30683, + "grad_norm": 1.1538076357202132, + "learning_rate": 0.003, + "loss": 4.048, + "step": 30683 + }, + { + "epoch": 0.30684, + "grad_norm": 1.140084039066211, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 30684 + }, + { + "epoch": 0.30685, + "grad_norm": 0.8459595857182383, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 30685 + }, + { + "epoch": 0.30686, + "grad_norm": 1.1087139331512863, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 30686 + }, + { + "epoch": 0.30687, + "grad_norm": 1.2660852991224094, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 30687 + }, + { + "epoch": 0.30688, + "grad_norm": 1.0861581157682063, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 30688 + }, + { + "epoch": 0.30689, + "grad_norm": 1.121606298918994, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 30689 + }, + { + "epoch": 0.3069, + "grad_norm": 1.2463377880891016, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 30690 + }, + { + "epoch": 0.30691, + "grad_norm": 0.9781770050434623, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 30691 + }, + { + "epoch": 0.30692, + "grad_norm": 0.9412798653895256, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 30692 + }, + { + "epoch": 0.30693, + "grad_norm": 1.1105640955253118, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 30693 + }, + { + "epoch": 0.30694, + "grad_norm": 1.1162271844177045, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 30694 + }, + { + "epoch": 0.30695, + "grad_norm": 0.907936186429479, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 30695 + }, + { + "epoch": 0.30696, + "grad_norm": 0.8923659057873307, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 30696 + }, + { + "epoch": 0.30697, + "grad_norm": 1.0758821732621549, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 30697 + }, + { + "epoch": 0.30698, + "grad_norm": 1.0340296140756027, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 30698 + }, + { + "epoch": 0.30699, + "grad_norm": 1.0914874949768778, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 30699 + }, + { + "epoch": 0.307, + "grad_norm": 1.1481823158517588, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 30700 + }, + { + "epoch": 0.30701, + "grad_norm": 0.9565869083279117, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 30701 + }, + { + "epoch": 0.30702, + "grad_norm": 1.345453666548939, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 30702 + }, + { + "epoch": 0.30703, + "grad_norm": 1.110439980724489, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 30703 + }, + { + "epoch": 0.30704, + "grad_norm": 1.064963544972446, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 30704 + }, + { + "epoch": 0.30705, + "grad_norm": 1.0977469690800483, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 30705 + }, + { + "epoch": 0.30706, + "grad_norm": 1.145319019340794, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 30706 + }, + { + "epoch": 0.30707, + "grad_norm": 0.9655195040878481, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 30707 + }, + { + "epoch": 0.30708, + "grad_norm": 1.0053469213264299, + "learning_rate": 0.003, + "loss": 4.028, + "step": 30708 + }, + { + "epoch": 0.30709, + "grad_norm": 1.0550258854302956, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 30709 + }, + { + "epoch": 0.3071, + "grad_norm": 1.1192722265679167, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 30710 + }, + { + "epoch": 0.30711, + "grad_norm": 0.994908330180915, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 30711 + }, + { + "epoch": 0.30712, + "grad_norm": 1.2155482445903485, + "learning_rate": 0.003, + "loss": 4.047, + "step": 30712 + }, + { + "epoch": 0.30713, + "grad_norm": 0.9733474194608114, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 30713 + }, + { + "epoch": 0.30714, + "grad_norm": 1.1953800887765127, + "learning_rate": 0.003, + "loss": 4.063, + "step": 30714 + }, + { + "epoch": 0.30715, + "grad_norm": 0.9148611007090184, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 30715 + }, + { + "epoch": 0.30716, + "grad_norm": 1.065201836885301, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 30716 + }, + { + "epoch": 0.30717, + "grad_norm": 1.105425357886939, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 30717 + }, + { + "epoch": 0.30718, + "grad_norm": 1.0701447040868957, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 30718 + }, + { + "epoch": 0.30719, + "grad_norm": 1.3061230805783948, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 30719 + }, + { + "epoch": 0.3072, + "grad_norm": 0.9429905711865972, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 30720 + }, + { + "epoch": 0.30721, + "grad_norm": 1.1539037531514567, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 30721 + }, + { + "epoch": 0.30722, + "grad_norm": 1.1965466507615354, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 30722 + }, + { + "epoch": 0.30723, + "grad_norm": 0.7968405708589955, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 30723 + }, + { + "epoch": 0.30724, + "grad_norm": 0.8512385099123516, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 30724 + }, + { + "epoch": 0.30725, + "grad_norm": 1.2502782145312474, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 30725 + }, + { + "epoch": 0.30726, + "grad_norm": 1.240742856064479, + "learning_rate": 0.003, + "loss": 4.068, + "step": 30726 + }, + { + "epoch": 0.30727, + "grad_norm": 0.7098919226599232, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 30727 + }, + { + "epoch": 0.30728, + "grad_norm": 0.6898622718444725, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 30728 + }, + { + "epoch": 0.30729, + "grad_norm": 0.8168605313196332, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 30729 + }, + { + "epoch": 0.3073, + "grad_norm": 0.8436731201622794, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 30730 + }, + { + "epoch": 0.30731, + "grad_norm": 0.9430902798535648, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 30731 + }, + { + "epoch": 0.30732, + "grad_norm": 1.1373570069906676, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 30732 + }, + { + "epoch": 0.30733, + "grad_norm": 1.0483625624458754, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 30733 + }, + { + "epoch": 0.30734, + "grad_norm": 1.0605816319471564, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 30734 + }, + { + "epoch": 0.30735, + "grad_norm": 1.178445653636009, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 30735 + }, + { + "epoch": 0.30736, + "grad_norm": 0.9494476719546036, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 30736 + }, + { + "epoch": 0.30737, + "grad_norm": 1.0791594833128515, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 30737 + }, + { + "epoch": 0.30738, + "grad_norm": 1.0548160803320716, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 30738 + }, + { + "epoch": 0.30739, + "grad_norm": 1.022000704947474, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 30739 + }, + { + "epoch": 0.3074, + "grad_norm": 1.097413353488873, + "learning_rate": 0.003, + "loss": 4.0854, + "step": 30740 + }, + { + "epoch": 0.30741, + "grad_norm": 1.031525371204509, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 30741 + }, + { + "epoch": 0.30742, + "grad_norm": 1.1772099420213336, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 30742 + }, + { + "epoch": 0.30743, + "grad_norm": 0.897297461911717, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 30743 + }, + { + "epoch": 0.30744, + "grad_norm": 0.8203911402764243, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 30744 + }, + { + "epoch": 0.30745, + "grad_norm": 1.0409582345341895, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 30745 + }, + { + "epoch": 0.30746, + "grad_norm": 1.1825718838055834, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 30746 + }, + { + "epoch": 0.30747, + "grad_norm": 0.8222884658118422, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 30747 + }, + { + "epoch": 0.30748, + "grad_norm": 0.833551920161384, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 30748 + }, + { + "epoch": 0.30749, + "grad_norm": 0.956739573731604, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 30749 + }, + { + "epoch": 0.3075, + "grad_norm": 1.1244575113700621, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 30750 + }, + { + "epoch": 0.30751, + "grad_norm": 1.1239729017212023, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 30751 + }, + { + "epoch": 0.30752, + "grad_norm": 1.1001436672736475, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 30752 + }, + { + "epoch": 0.30753, + "grad_norm": 1.083151919888603, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 30753 + }, + { + "epoch": 0.30754, + "grad_norm": 0.955718232945757, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 30754 + }, + { + "epoch": 0.30755, + "grad_norm": 1.2055197039011918, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 30755 + }, + { + "epoch": 0.30756, + "grad_norm": 1.122077742338935, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 30756 + }, + { + "epoch": 0.30757, + "grad_norm": 1.4830924442799744, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 30757 + }, + { + "epoch": 0.30758, + "grad_norm": 1.2179597037822187, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 30758 + }, + { + "epoch": 0.30759, + "grad_norm": 1.1630476870549409, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 30759 + }, + { + "epoch": 0.3076, + "grad_norm": 0.9925008306811987, + "learning_rate": 0.003, + "loss": 4.026, + "step": 30760 + }, + { + "epoch": 0.30761, + "grad_norm": 1.0037724629684226, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 30761 + }, + { + "epoch": 0.30762, + "grad_norm": 1.4268808401107154, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 30762 + }, + { + "epoch": 0.30763, + "grad_norm": 0.969828450382593, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 30763 + }, + { + "epoch": 0.30764, + "grad_norm": 1.1013190683193552, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 30764 + }, + { + "epoch": 0.30765, + "grad_norm": 1.3607651803104694, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 30765 + }, + { + "epoch": 0.30766, + "grad_norm": 0.7931143595229547, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 30766 + }, + { + "epoch": 0.30767, + "grad_norm": 0.7625249603773778, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 30767 + }, + { + "epoch": 0.30768, + "grad_norm": 0.7702922869995767, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 30768 + }, + { + "epoch": 0.30769, + "grad_norm": 0.9724917127548264, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 30769 + }, + { + "epoch": 0.3077, + "grad_norm": 1.2357707974692607, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 30770 + }, + { + "epoch": 0.30771, + "grad_norm": 0.9589557773206338, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 30771 + }, + { + "epoch": 0.30772, + "grad_norm": 1.121391583697353, + "learning_rate": 0.003, + "loss": 4.1013, + "step": 30772 + }, + { + "epoch": 0.30773, + "grad_norm": 1.0169096329741796, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 30773 + }, + { + "epoch": 0.30774, + "grad_norm": 0.9792376214763163, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 30774 + }, + { + "epoch": 0.30775, + "grad_norm": 1.1843734481998862, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 30775 + }, + { + "epoch": 0.30776, + "grad_norm": 0.9326917868344994, + "learning_rate": 0.003, + "loss": 4.026, + "step": 30776 + }, + { + "epoch": 0.30777, + "grad_norm": 0.9879111973071002, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 30777 + }, + { + "epoch": 0.30778, + "grad_norm": 1.3809926231130152, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 30778 + }, + { + "epoch": 0.30779, + "grad_norm": 1.0080191744623026, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 30779 + }, + { + "epoch": 0.3078, + "grad_norm": 1.4294559435706775, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 30780 + }, + { + "epoch": 0.30781, + "grad_norm": 1.0812436326116859, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 30781 + }, + { + "epoch": 0.30782, + "grad_norm": 0.9203855650490519, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 30782 + }, + { + "epoch": 0.30783, + "grad_norm": 0.9356600979526251, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 30783 + }, + { + "epoch": 0.30784, + "grad_norm": 1.1256643557592825, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 30784 + }, + { + "epoch": 0.30785, + "grad_norm": 0.9952084112870181, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 30785 + }, + { + "epoch": 0.30786, + "grad_norm": 1.1675409907506418, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 30786 + }, + { + "epoch": 0.30787, + "grad_norm": 1.1227816645995696, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 30787 + }, + { + "epoch": 0.30788, + "grad_norm": 1.085748909295769, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 30788 + }, + { + "epoch": 0.30789, + "grad_norm": 1.2240506865729694, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 30789 + }, + { + "epoch": 0.3079, + "grad_norm": 0.9019403674027773, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 30790 + }, + { + "epoch": 0.30791, + "grad_norm": 1.1682142106325453, + "learning_rate": 0.003, + "loss": 4.066, + "step": 30791 + }, + { + "epoch": 0.30792, + "grad_norm": 1.2683671985804392, + "learning_rate": 0.003, + "loss": 4.086, + "step": 30792 + }, + { + "epoch": 0.30793, + "grad_norm": 1.1246708740041742, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 30793 + }, + { + "epoch": 0.30794, + "grad_norm": 0.871864952160847, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 30794 + }, + { + "epoch": 0.30795, + "grad_norm": 0.9643190270717139, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 30795 + }, + { + "epoch": 0.30796, + "grad_norm": 1.1620284029137204, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 30796 + }, + { + "epoch": 0.30797, + "grad_norm": 0.9531670372032761, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 30797 + }, + { + "epoch": 0.30798, + "grad_norm": 1.2396261873255963, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 30798 + }, + { + "epoch": 0.30799, + "grad_norm": 1.1601171291188141, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 30799 + }, + { + "epoch": 0.308, + "grad_norm": 1.054738414465745, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 30800 + }, + { + "epoch": 0.30801, + "grad_norm": 1.082904089798655, + "learning_rate": 0.003, + "loss": 4.045, + "step": 30801 + }, + { + "epoch": 0.30802, + "grad_norm": 1.0950773843007033, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 30802 + }, + { + "epoch": 0.30803, + "grad_norm": 1.2730193832350667, + "learning_rate": 0.003, + "loss": 4.031, + "step": 30803 + }, + { + "epoch": 0.30804, + "grad_norm": 0.9280740166956931, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 30804 + }, + { + "epoch": 0.30805, + "grad_norm": 0.9636774457330433, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 30805 + }, + { + "epoch": 0.30806, + "grad_norm": 1.1931059508018425, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 30806 + }, + { + "epoch": 0.30807, + "grad_norm": 0.9210518082620353, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 30807 + }, + { + "epoch": 0.30808, + "grad_norm": 0.9152190021029716, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 30808 + }, + { + "epoch": 0.30809, + "grad_norm": 0.9700039464958855, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 30809 + }, + { + "epoch": 0.3081, + "grad_norm": 1.1691125007091903, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 30810 + }, + { + "epoch": 0.30811, + "grad_norm": 1.160564317830835, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 30811 + }, + { + "epoch": 0.30812, + "grad_norm": 1.0366554135139066, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 30812 + }, + { + "epoch": 0.30813, + "grad_norm": 1.1189834511763166, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 30813 + }, + { + "epoch": 0.30814, + "grad_norm": 0.969524808337459, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 30814 + }, + { + "epoch": 0.30815, + "grad_norm": 1.0044292161811847, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 30815 + }, + { + "epoch": 0.30816, + "grad_norm": 0.8984677669691795, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 30816 + }, + { + "epoch": 0.30817, + "grad_norm": 0.8798005772628609, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 30817 + }, + { + "epoch": 0.30818, + "grad_norm": 0.9037533166691997, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 30818 + }, + { + "epoch": 0.30819, + "grad_norm": 0.9847234663653543, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 30819 + }, + { + "epoch": 0.3082, + "grad_norm": 1.099226383446129, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 30820 + }, + { + "epoch": 0.30821, + "grad_norm": 1.1760564619427725, + "learning_rate": 0.003, + "loss": 4.061, + "step": 30821 + }, + { + "epoch": 0.30822, + "grad_norm": 1.1214911565485273, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 30822 + }, + { + "epoch": 0.30823, + "grad_norm": 0.9321865256629343, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 30823 + }, + { + "epoch": 0.30824, + "grad_norm": 1.2201254616687698, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 30824 + }, + { + "epoch": 0.30825, + "grad_norm": 1.2118257671672878, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 30825 + }, + { + "epoch": 0.30826, + "grad_norm": 0.9787940950640808, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 30826 + }, + { + "epoch": 0.30827, + "grad_norm": 1.2139435476787244, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 30827 + }, + { + "epoch": 0.30828, + "grad_norm": 1.1946019830523216, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 30828 + }, + { + "epoch": 0.30829, + "grad_norm": 1.227936060536281, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 30829 + }, + { + "epoch": 0.3083, + "grad_norm": 0.8677923504924017, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 30830 + }, + { + "epoch": 0.30831, + "grad_norm": 0.9673968400891712, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 30831 + }, + { + "epoch": 0.30832, + "grad_norm": 1.19313266611585, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 30832 + }, + { + "epoch": 0.30833, + "grad_norm": 1.1548398752574218, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 30833 + }, + { + "epoch": 0.30834, + "grad_norm": 1.1049497804310884, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 30834 + }, + { + "epoch": 0.30835, + "grad_norm": 1.2735440584570803, + "learning_rate": 0.003, + "loss": 4.052, + "step": 30835 + }, + { + "epoch": 0.30836, + "grad_norm": 1.1472792064727146, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 30836 + }, + { + "epoch": 0.30837, + "grad_norm": 1.0485947856106153, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 30837 + }, + { + "epoch": 0.30838, + "grad_norm": 0.9545202993825905, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 30838 + }, + { + "epoch": 0.30839, + "grad_norm": 1.1541907332146377, + "learning_rate": 0.003, + "loss": 4.066, + "step": 30839 + }, + { + "epoch": 0.3084, + "grad_norm": 1.4007304640235538, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 30840 + }, + { + "epoch": 0.30841, + "grad_norm": 1.0145181146325892, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 30841 + }, + { + "epoch": 0.30842, + "grad_norm": 1.1044572197137572, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 30842 + }, + { + "epoch": 0.30843, + "grad_norm": 1.005506652616188, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 30843 + }, + { + "epoch": 0.30844, + "grad_norm": 1.0572705886232423, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 30844 + }, + { + "epoch": 0.30845, + "grad_norm": 1.0397380033702504, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 30845 + }, + { + "epoch": 0.30846, + "grad_norm": 0.9261875519954323, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 30846 + }, + { + "epoch": 0.30847, + "grad_norm": 1.1168651961409028, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 30847 + }, + { + "epoch": 0.30848, + "grad_norm": 1.1081750282733682, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 30848 + }, + { + "epoch": 0.30849, + "grad_norm": 0.9955326679350346, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 30849 + }, + { + "epoch": 0.3085, + "grad_norm": 0.9311746183896447, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 30850 + }, + { + "epoch": 0.30851, + "grad_norm": 0.9799744673800718, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 30851 + }, + { + "epoch": 0.30852, + "grad_norm": 1.0514395408137664, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 30852 + }, + { + "epoch": 0.30853, + "grad_norm": 1.2052014048586774, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 30853 + }, + { + "epoch": 0.30854, + "grad_norm": 1.0463247567286986, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 30854 + }, + { + "epoch": 0.30855, + "grad_norm": 1.1443472413638982, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 30855 + }, + { + "epoch": 0.30856, + "grad_norm": 1.2632849665133177, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 30856 + }, + { + "epoch": 0.30857, + "grad_norm": 1.1961182879630556, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 30857 + }, + { + "epoch": 0.30858, + "grad_norm": 0.9550410562505468, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 30858 + }, + { + "epoch": 0.30859, + "grad_norm": 1.0535862793438238, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 30859 + }, + { + "epoch": 0.3086, + "grad_norm": 1.0835933030397806, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 30860 + }, + { + "epoch": 0.30861, + "grad_norm": 1.070231819414741, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 30861 + }, + { + "epoch": 0.30862, + "grad_norm": 1.1753394415306, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 30862 + }, + { + "epoch": 0.30863, + "grad_norm": 0.85703348112408, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 30863 + }, + { + "epoch": 0.30864, + "grad_norm": 1.0326073915154985, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 30864 + }, + { + "epoch": 0.30865, + "grad_norm": 1.1346735264163834, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 30865 + }, + { + "epoch": 0.30866, + "grad_norm": 0.978257902488856, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 30866 + }, + { + "epoch": 0.30867, + "grad_norm": 1.1599168596887601, + "learning_rate": 0.003, + "loss": 4.043, + "step": 30867 + }, + { + "epoch": 0.30868, + "grad_norm": 1.198731525276581, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 30868 + }, + { + "epoch": 0.30869, + "grad_norm": 1.2626087767040612, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 30869 + }, + { + "epoch": 0.3087, + "grad_norm": 0.9521939886953328, + "learning_rate": 0.003, + "loss": 4.032, + "step": 30870 + }, + { + "epoch": 0.30871, + "grad_norm": 1.0214148380818453, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 30871 + }, + { + "epoch": 0.30872, + "grad_norm": 1.089888259319064, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 30872 + }, + { + "epoch": 0.30873, + "grad_norm": 1.057333244830114, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 30873 + }, + { + "epoch": 0.30874, + "grad_norm": 1.0272904926678859, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 30874 + }, + { + "epoch": 0.30875, + "grad_norm": 1.1928972807892666, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 30875 + }, + { + "epoch": 0.30876, + "grad_norm": 1.0698287643284878, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 30876 + }, + { + "epoch": 0.30877, + "grad_norm": 0.950648978794975, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 30877 + }, + { + "epoch": 0.30878, + "grad_norm": 1.0184036146932232, + "learning_rate": 0.003, + "loss": 4.03, + "step": 30878 + }, + { + "epoch": 0.30879, + "grad_norm": 1.2826543021684718, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 30879 + }, + { + "epoch": 0.3088, + "grad_norm": 1.071510033601312, + "learning_rate": 0.003, + "loss": 4.037, + "step": 30880 + }, + { + "epoch": 0.30881, + "grad_norm": 1.0798259433469182, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 30881 + }, + { + "epoch": 0.30882, + "grad_norm": 1.010870999413398, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 30882 + }, + { + "epoch": 0.30883, + "grad_norm": 1.2815284237920133, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 30883 + }, + { + "epoch": 0.30884, + "grad_norm": 1.019624119927432, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 30884 + }, + { + "epoch": 0.30885, + "grad_norm": 1.1011348470653495, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 30885 + }, + { + "epoch": 0.30886, + "grad_norm": 0.990368613711357, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 30886 + }, + { + "epoch": 0.30887, + "grad_norm": 1.1404618085589815, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 30887 + }, + { + "epoch": 0.30888, + "grad_norm": 1.0756295863462242, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 30888 + }, + { + "epoch": 0.30889, + "grad_norm": 0.9656378190912621, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 30889 + }, + { + "epoch": 0.3089, + "grad_norm": 1.1478234515781316, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 30890 + }, + { + "epoch": 0.30891, + "grad_norm": 1.0187246045678369, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 30891 + }, + { + "epoch": 0.30892, + "grad_norm": 1.3342614547674374, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 30892 + }, + { + "epoch": 0.30893, + "grad_norm": 1.0901780071217881, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 30893 + }, + { + "epoch": 0.30894, + "grad_norm": 1.178408496983452, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 30894 + }, + { + "epoch": 0.30895, + "grad_norm": 1.0595382233916704, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 30895 + }, + { + "epoch": 0.30896, + "grad_norm": 1.2298444276296576, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 30896 + }, + { + "epoch": 0.30897, + "grad_norm": 0.8738461238474936, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 30897 + }, + { + "epoch": 0.30898, + "grad_norm": 0.9132943909318295, + "learning_rate": 0.003, + "loss": 4.032, + "step": 30898 + }, + { + "epoch": 0.30899, + "grad_norm": 1.2153200467693246, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 30899 + }, + { + "epoch": 0.309, + "grad_norm": 1.268058244720732, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 30900 + }, + { + "epoch": 0.30901, + "grad_norm": 0.9446585252098992, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 30901 + }, + { + "epoch": 0.30902, + "grad_norm": 0.8659764579894043, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 30902 + }, + { + "epoch": 0.30903, + "grad_norm": 0.9482002451845646, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 30903 + }, + { + "epoch": 0.30904, + "grad_norm": 1.0717957920196426, + "learning_rate": 0.003, + "loss": 4.061, + "step": 30904 + }, + { + "epoch": 0.30905, + "grad_norm": 1.1331841785110497, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 30905 + }, + { + "epoch": 0.30906, + "grad_norm": 1.2096408811786594, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 30906 + }, + { + "epoch": 0.30907, + "grad_norm": 1.0556832555222866, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 30907 + }, + { + "epoch": 0.30908, + "grad_norm": 1.0059945372898842, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 30908 + }, + { + "epoch": 0.30909, + "grad_norm": 1.0379557284787615, + "learning_rate": 0.003, + "loss": 4.017, + "step": 30909 + }, + { + "epoch": 0.3091, + "grad_norm": 1.0202307175640288, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 30910 + }, + { + "epoch": 0.30911, + "grad_norm": 1.0757605444113048, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 30911 + }, + { + "epoch": 0.30912, + "grad_norm": 1.4801825453684987, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 30912 + }, + { + "epoch": 0.30913, + "grad_norm": 0.7847109013622436, + "learning_rate": 0.003, + "loss": 4.038, + "step": 30913 + }, + { + "epoch": 0.30914, + "grad_norm": 0.7964604416944453, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 30914 + }, + { + "epoch": 0.30915, + "grad_norm": 0.9213845438187421, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 30915 + }, + { + "epoch": 0.30916, + "grad_norm": 1.285151146240703, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 30916 + }, + { + "epoch": 0.30917, + "grad_norm": 0.9024147711384594, + "learning_rate": 0.003, + "loss": 4.039, + "step": 30917 + }, + { + "epoch": 0.30918, + "grad_norm": 0.996539488901591, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 30918 + }, + { + "epoch": 0.30919, + "grad_norm": 1.4027935569744514, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 30919 + }, + { + "epoch": 0.3092, + "grad_norm": 0.851860160338816, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 30920 + }, + { + "epoch": 0.30921, + "grad_norm": 0.9618806774582681, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 30921 + }, + { + "epoch": 0.30922, + "grad_norm": 1.090684326852998, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 30922 + }, + { + "epoch": 0.30923, + "grad_norm": 1.0276063084449407, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 30923 + }, + { + "epoch": 0.30924, + "grad_norm": 1.1417261013478164, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 30924 + }, + { + "epoch": 0.30925, + "grad_norm": 1.1917588493601248, + "learning_rate": 0.003, + "loss": 4.0899, + "step": 30925 + }, + { + "epoch": 0.30926, + "grad_norm": 0.9446386898234577, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 30926 + }, + { + "epoch": 0.30927, + "grad_norm": 0.9899210145155206, + "learning_rate": 0.003, + "loss": 4.063, + "step": 30927 + }, + { + "epoch": 0.30928, + "grad_norm": 1.1999084653113203, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 30928 + }, + { + "epoch": 0.30929, + "grad_norm": 1.090021617807192, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 30929 + }, + { + "epoch": 0.3093, + "grad_norm": 1.0482623258343717, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 30930 + }, + { + "epoch": 0.30931, + "grad_norm": 0.9501467173720861, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 30931 + }, + { + "epoch": 0.30932, + "grad_norm": 1.1185927676933625, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 30932 + }, + { + "epoch": 0.30933, + "grad_norm": 1.002796668078201, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 30933 + }, + { + "epoch": 0.30934, + "grad_norm": 1.1412898518464085, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 30934 + }, + { + "epoch": 0.30935, + "grad_norm": 1.0244398626417768, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 30935 + }, + { + "epoch": 0.30936, + "grad_norm": 1.3122057833266314, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 30936 + }, + { + "epoch": 0.30937, + "grad_norm": 1.251926570659136, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 30937 + }, + { + "epoch": 0.30938, + "grad_norm": 0.878910650705691, + "learning_rate": 0.003, + "loss": 4.02, + "step": 30938 + }, + { + "epoch": 0.30939, + "grad_norm": 0.8879979287497642, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 30939 + }, + { + "epoch": 0.3094, + "grad_norm": 1.1005328346747127, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 30940 + }, + { + "epoch": 0.30941, + "grad_norm": 1.2716493650910041, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 30941 + }, + { + "epoch": 0.30942, + "grad_norm": 0.8161758828015745, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 30942 + }, + { + "epoch": 0.30943, + "grad_norm": 0.8807246352426882, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 30943 + }, + { + "epoch": 0.30944, + "grad_norm": 1.1127746381606154, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 30944 + }, + { + "epoch": 0.30945, + "grad_norm": 1.3984314101324804, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 30945 + }, + { + "epoch": 0.30946, + "grad_norm": 0.7369104047426199, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 30946 + }, + { + "epoch": 0.30947, + "grad_norm": 0.7940083945290143, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 30947 + }, + { + "epoch": 0.30948, + "grad_norm": 0.9058334618343157, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 30948 + }, + { + "epoch": 0.30949, + "grad_norm": 1.2087509377903185, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 30949 + }, + { + "epoch": 0.3095, + "grad_norm": 1.132098222664339, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 30950 + }, + { + "epoch": 0.30951, + "grad_norm": 1.1478889711576767, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 30951 + }, + { + "epoch": 0.30952, + "grad_norm": 1.0327770560694074, + "learning_rate": 0.003, + "loss": 4.019, + "step": 30952 + }, + { + "epoch": 0.30953, + "grad_norm": 0.9242802909583098, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 30953 + }, + { + "epoch": 0.30954, + "grad_norm": 1.074260299959251, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 30954 + }, + { + "epoch": 0.30955, + "grad_norm": 1.049686583934577, + "learning_rate": 0.003, + "loss": 4.072, + "step": 30955 + }, + { + "epoch": 0.30956, + "grad_norm": 1.110807462715472, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 30956 + }, + { + "epoch": 0.30957, + "grad_norm": 1.1895719092129278, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 30957 + }, + { + "epoch": 0.30958, + "grad_norm": 1.019857906125548, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 30958 + }, + { + "epoch": 0.30959, + "grad_norm": 1.181862810949238, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 30959 + }, + { + "epoch": 0.3096, + "grad_norm": 1.010008211234792, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 30960 + }, + { + "epoch": 0.30961, + "grad_norm": 1.0586067374133832, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 30961 + }, + { + "epoch": 0.30962, + "grad_norm": 1.0182201896246965, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 30962 + }, + { + "epoch": 0.30963, + "grad_norm": 1.0585437931729205, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 30963 + }, + { + "epoch": 0.30964, + "grad_norm": 0.9829502019838298, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 30964 + }, + { + "epoch": 0.30965, + "grad_norm": 1.3528494274533307, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 30965 + }, + { + "epoch": 0.30966, + "grad_norm": 0.7946703644146212, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 30966 + }, + { + "epoch": 0.30967, + "grad_norm": 1.1426352955906387, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 30967 + }, + { + "epoch": 0.30968, + "grad_norm": 1.2533589026863667, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 30968 + }, + { + "epoch": 0.30969, + "grad_norm": 1.077050617858029, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 30969 + }, + { + "epoch": 0.3097, + "grad_norm": 1.107616035906114, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 30970 + }, + { + "epoch": 0.30971, + "grad_norm": 1.1725615541969838, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 30971 + }, + { + "epoch": 0.30972, + "grad_norm": 1.2231303549034642, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 30972 + }, + { + "epoch": 0.30973, + "grad_norm": 0.9960968448487324, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 30973 + }, + { + "epoch": 0.30974, + "grad_norm": 0.9888716989916406, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 30974 + }, + { + "epoch": 0.30975, + "grad_norm": 1.082226405860246, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 30975 + }, + { + "epoch": 0.30976, + "grad_norm": 1.066301063258497, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 30976 + }, + { + "epoch": 0.30977, + "grad_norm": 0.9607993307245037, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 30977 + }, + { + "epoch": 0.30978, + "grad_norm": 1.0142764110058762, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 30978 + }, + { + "epoch": 0.30979, + "grad_norm": 0.9859613155750677, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 30979 + }, + { + "epoch": 0.3098, + "grad_norm": 1.1971428907169228, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 30980 + }, + { + "epoch": 0.30981, + "grad_norm": 0.9606961474164843, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 30981 + }, + { + "epoch": 0.30982, + "grad_norm": 1.0712672350455648, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 30982 + }, + { + "epoch": 0.30983, + "grad_norm": 1.1683041209824745, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 30983 + }, + { + "epoch": 0.30984, + "grad_norm": 1.1503465962258674, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 30984 + }, + { + "epoch": 0.30985, + "grad_norm": 1.1055611338201574, + "learning_rate": 0.003, + "loss": 4.044, + "step": 30985 + }, + { + "epoch": 0.30986, + "grad_norm": 1.2837251747635812, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 30986 + }, + { + "epoch": 0.30987, + "grad_norm": 1.0324343345313602, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 30987 + }, + { + "epoch": 0.30988, + "grad_norm": 1.0541406717539044, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 30988 + }, + { + "epoch": 0.30989, + "grad_norm": 1.2233490774046614, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 30989 + }, + { + "epoch": 0.3099, + "grad_norm": 1.1119219754870528, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 30990 + }, + { + "epoch": 0.30991, + "grad_norm": 1.080896303599478, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 30991 + }, + { + "epoch": 0.30992, + "grad_norm": 1.2770117693326177, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 30992 + }, + { + "epoch": 0.30993, + "grad_norm": 1.0087696208078465, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 30993 + }, + { + "epoch": 0.30994, + "grad_norm": 0.9920171658150752, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 30994 + }, + { + "epoch": 0.30995, + "grad_norm": 1.1572841727118546, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 30995 + }, + { + "epoch": 0.30996, + "grad_norm": 1.0493038274496496, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 30996 + }, + { + "epoch": 0.30997, + "grad_norm": 1.0060511340526563, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 30997 + }, + { + "epoch": 0.30998, + "grad_norm": 1.1564876149747814, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 30998 + }, + { + "epoch": 0.30999, + "grad_norm": 0.9960520101946883, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 30999 + }, + { + "epoch": 0.31, + "grad_norm": 1.2760607096334333, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 31000 + }, + { + "epoch": 0.31001, + "grad_norm": 1.0555542498666683, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 31001 + }, + { + "epoch": 0.31002, + "grad_norm": 1.2290891317016732, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 31002 + }, + { + "epoch": 0.31003, + "grad_norm": 0.8473529346869545, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 31003 + }, + { + "epoch": 0.31004, + "grad_norm": 0.9120286997454694, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 31004 + }, + { + "epoch": 0.31005, + "grad_norm": 1.0105478610728424, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 31005 + }, + { + "epoch": 0.31006, + "grad_norm": 1.3109570445869796, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 31006 + }, + { + "epoch": 0.31007, + "grad_norm": 0.8528952754564538, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 31007 + }, + { + "epoch": 0.31008, + "grad_norm": 1.0412381212938886, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 31008 + }, + { + "epoch": 0.31009, + "grad_norm": 1.1353684512202071, + "learning_rate": 0.003, + "loss": 4.053, + "step": 31009 + }, + { + "epoch": 0.3101, + "grad_norm": 0.9198186170028776, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 31010 + }, + { + "epoch": 0.31011, + "grad_norm": 1.0912005047258206, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 31011 + }, + { + "epoch": 0.31012, + "grad_norm": 1.186902465599861, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 31012 + }, + { + "epoch": 0.31013, + "grad_norm": 1.1085883136722419, + "learning_rate": 0.003, + "loss": 4.062, + "step": 31013 + }, + { + "epoch": 0.31014, + "grad_norm": 1.1026355353439528, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 31014 + }, + { + "epoch": 0.31015, + "grad_norm": 1.0839324260574505, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 31015 + }, + { + "epoch": 0.31016, + "grad_norm": 1.2835312366024725, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 31016 + }, + { + "epoch": 0.31017, + "grad_norm": 1.0951653489488882, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 31017 + }, + { + "epoch": 0.31018, + "grad_norm": 1.1383745591044938, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 31018 + }, + { + "epoch": 0.31019, + "grad_norm": 0.9015533270920738, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 31019 + }, + { + "epoch": 0.3102, + "grad_norm": 0.9203613596641237, + "learning_rate": 0.003, + "loss": 4.025, + "step": 31020 + }, + { + "epoch": 0.31021, + "grad_norm": 1.144314837622505, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 31021 + }, + { + "epoch": 0.31022, + "grad_norm": 1.087220492636535, + "learning_rate": 0.003, + "loss": 4.037, + "step": 31022 + }, + { + "epoch": 0.31023, + "grad_norm": 1.0961276272522809, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 31023 + }, + { + "epoch": 0.31024, + "grad_norm": 1.0170885136158547, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 31024 + }, + { + "epoch": 0.31025, + "grad_norm": 1.1368556458690324, + "learning_rate": 0.003, + "loss": 4.03, + "step": 31025 + }, + { + "epoch": 0.31026, + "grad_norm": 1.0697493455341114, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 31026 + }, + { + "epoch": 0.31027, + "grad_norm": 1.026521245812156, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 31027 + }, + { + "epoch": 0.31028, + "grad_norm": 1.0892072531787922, + "learning_rate": 0.003, + "loss": 4.036, + "step": 31028 + }, + { + "epoch": 0.31029, + "grad_norm": 0.9960590869690688, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 31029 + }, + { + "epoch": 0.3103, + "grad_norm": 1.0734719295143236, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 31030 + }, + { + "epoch": 0.31031, + "grad_norm": 1.266395705105653, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 31031 + }, + { + "epoch": 0.31032, + "grad_norm": 1.0617868485141615, + "learning_rate": 0.003, + "loss": 4.011, + "step": 31032 + }, + { + "epoch": 0.31033, + "grad_norm": 1.0088355837987713, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 31033 + }, + { + "epoch": 0.31034, + "grad_norm": 1.1940717609215745, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 31034 + }, + { + "epoch": 0.31035, + "grad_norm": 0.9391607600952823, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 31035 + }, + { + "epoch": 0.31036, + "grad_norm": 0.9394671537865527, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 31036 + }, + { + "epoch": 0.31037, + "grad_norm": 1.18239963557344, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 31037 + }, + { + "epoch": 0.31038, + "grad_norm": 1.0700413511689801, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 31038 + }, + { + "epoch": 0.31039, + "grad_norm": 1.0145362000296334, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 31039 + }, + { + "epoch": 0.3104, + "grad_norm": 1.2608144352105684, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 31040 + }, + { + "epoch": 0.31041, + "grad_norm": 0.8705829818518782, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 31041 + }, + { + "epoch": 0.31042, + "grad_norm": 0.8404880259376921, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 31042 + }, + { + "epoch": 0.31043, + "grad_norm": 0.9089820840471261, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 31043 + }, + { + "epoch": 0.31044, + "grad_norm": 1.2957771078766005, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 31044 + }, + { + "epoch": 0.31045, + "grad_norm": 1.0169102902217386, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 31045 + }, + { + "epoch": 0.31046, + "grad_norm": 1.1141795419761915, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 31046 + }, + { + "epoch": 0.31047, + "grad_norm": 0.9754102592123219, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 31047 + }, + { + "epoch": 0.31048, + "grad_norm": 1.042457789868284, + "learning_rate": 0.003, + "loss": 4.042, + "step": 31048 + }, + { + "epoch": 0.31049, + "grad_norm": 1.3215062673256859, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 31049 + }, + { + "epoch": 0.3105, + "grad_norm": 0.8602055305387908, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 31050 + }, + { + "epoch": 0.31051, + "grad_norm": 0.9299802617323926, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 31051 + }, + { + "epoch": 0.31052, + "grad_norm": 1.0010606834013407, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 31052 + }, + { + "epoch": 0.31053, + "grad_norm": 1.1952921378188948, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 31053 + }, + { + "epoch": 0.31054, + "grad_norm": 1.0220824475784533, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 31054 + }, + { + "epoch": 0.31055, + "grad_norm": 1.1545443857430173, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 31055 + }, + { + "epoch": 0.31056, + "grad_norm": 0.9737620636084753, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 31056 + }, + { + "epoch": 0.31057, + "grad_norm": 1.1837184109092476, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 31057 + }, + { + "epoch": 0.31058, + "grad_norm": 1.244367237118735, + "learning_rate": 0.003, + "loss": 4.061, + "step": 31058 + }, + { + "epoch": 0.31059, + "grad_norm": 1.2065651462087055, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 31059 + }, + { + "epoch": 0.3106, + "grad_norm": 0.9940034544691609, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 31060 + }, + { + "epoch": 0.31061, + "grad_norm": 1.2614454168181597, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 31061 + }, + { + "epoch": 0.31062, + "grad_norm": 0.9323911071218031, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 31062 + }, + { + "epoch": 0.31063, + "grad_norm": 1.040646214863591, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 31063 + }, + { + "epoch": 0.31064, + "grad_norm": 1.1838023925543202, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 31064 + }, + { + "epoch": 0.31065, + "grad_norm": 0.9927515787134135, + "learning_rate": 0.003, + "loss": 4.058, + "step": 31065 + }, + { + "epoch": 0.31066, + "grad_norm": 1.1408389296264443, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 31066 + }, + { + "epoch": 0.31067, + "grad_norm": 0.8329741634929821, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 31067 + }, + { + "epoch": 0.31068, + "grad_norm": 0.7819472591888894, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 31068 + }, + { + "epoch": 0.31069, + "grad_norm": 0.8871664952954708, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 31069 + }, + { + "epoch": 0.3107, + "grad_norm": 1.0567993560056672, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 31070 + }, + { + "epoch": 0.31071, + "grad_norm": 1.1131310088148125, + "learning_rate": 0.003, + "loss": 4.033, + "step": 31071 + }, + { + "epoch": 0.31072, + "grad_norm": 0.931374784730663, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 31072 + }, + { + "epoch": 0.31073, + "grad_norm": 1.0076319873572157, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 31073 + }, + { + "epoch": 0.31074, + "grad_norm": 1.3323376139059138, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 31074 + }, + { + "epoch": 0.31075, + "grad_norm": 1.032441674670981, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 31075 + }, + { + "epoch": 0.31076, + "grad_norm": 1.356028785121844, + "learning_rate": 0.003, + "loss": 4.059, + "step": 31076 + }, + { + "epoch": 0.31077, + "grad_norm": 1.0651662890481544, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 31077 + }, + { + "epoch": 0.31078, + "grad_norm": 1.141779170982983, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 31078 + }, + { + "epoch": 0.31079, + "grad_norm": 1.2001188021375413, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 31079 + }, + { + "epoch": 0.3108, + "grad_norm": 1.0761373420600717, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 31080 + }, + { + "epoch": 0.31081, + "grad_norm": 1.0300855451526871, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 31081 + }, + { + "epoch": 0.31082, + "grad_norm": 1.0553798535101278, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 31082 + }, + { + "epoch": 0.31083, + "grad_norm": 1.0507868699380571, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 31083 + }, + { + "epoch": 0.31084, + "grad_norm": 0.9994786754601039, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 31084 + }, + { + "epoch": 0.31085, + "grad_norm": 1.1735845648679575, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 31085 + }, + { + "epoch": 0.31086, + "grad_norm": 1.0527510370704194, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 31086 + }, + { + "epoch": 0.31087, + "grad_norm": 1.1752431248283288, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 31087 + }, + { + "epoch": 0.31088, + "grad_norm": 1.1304244126939345, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 31088 + }, + { + "epoch": 0.31089, + "grad_norm": 0.9012053056276215, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 31089 + }, + { + "epoch": 0.3109, + "grad_norm": 0.9738013981774399, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 31090 + }, + { + "epoch": 0.31091, + "grad_norm": 1.1495498019654458, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 31091 + }, + { + "epoch": 0.31092, + "grad_norm": 1.2140961130792602, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 31092 + }, + { + "epoch": 0.31093, + "grad_norm": 1.0597178394188846, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 31093 + }, + { + "epoch": 0.31094, + "grad_norm": 1.2638366991895462, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 31094 + }, + { + "epoch": 0.31095, + "grad_norm": 0.8679785689694177, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 31095 + }, + { + "epoch": 0.31096, + "grad_norm": 0.88166299764095, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 31096 + }, + { + "epoch": 0.31097, + "grad_norm": 1.0009268507660822, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 31097 + }, + { + "epoch": 0.31098, + "grad_norm": 1.2620048480961026, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 31098 + }, + { + "epoch": 0.31099, + "grad_norm": 0.8854645092759044, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 31099 + }, + { + "epoch": 0.311, + "grad_norm": 0.8612781709998137, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 31100 + }, + { + "epoch": 0.31101, + "grad_norm": 1.040783261444686, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 31101 + }, + { + "epoch": 0.31102, + "grad_norm": 1.3508993001398717, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 31102 + }, + { + "epoch": 0.31103, + "grad_norm": 0.8611263653252598, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 31103 + }, + { + "epoch": 0.31104, + "grad_norm": 1.0513144322675134, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 31104 + }, + { + "epoch": 0.31105, + "grad_norm": 1.0047700277039713, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 31105 + }, + { + "epoch": 0.31106, + "grad_norm": 1.1440661668978078, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 31106 + }, + { + "epoch": 0.31107, + "grad_norm": 1.006178584324847, + "learning_rate": 0.003, + "loss": 4.045, + "step": 31107 + }, + { + "epoch": 0.31108, + "grad_norm": 1.2062358691109427, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 31108 + }, + { + "epoch": 0.31109, + "grad_norm": 1.0252169088320189, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 31109 + }, + { + "epoch": 0.3111, + "grad_norm": 1.2619735047124876, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 31110 + }, + { + "epoch": 0.31111, + "grad_norm": 1.0606496292410104, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 31111 + }, + { + "epoch": 0.31112, + "grad_norm": 1.1430645709196523, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 31112 + }, + { + "epoch": 0.31113, + "grad_norm": 1.1609110386525041, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 31113 + }, + { + "epoch": 0.31114, + "grad_norm": 1.1745880648278748, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 31114 + }, + { + "epoch": 0.31115, + "grad_norm": 1.0301871193618342, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 31115 + }, + { + "epoch": 0.31116, + "grad_norm": 1.3120832953641546, + "learning_rate": 0.003, + "loss": 4.046, + "step": 31116 + }, + { + "epoch": 0.31117, + "grad_norm": 0.9605061123476057, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 31117 + }, + { + "epoch": 0.31118, + "grad_norm": 0.8671429650073651, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 31118 + }, + { + "epoch": 0.31119, + "grad_norm": 0.8792600491871896, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 31119 + }, + { + "epoch": 0.3112, + "grad_norm": 1.2378751116351447, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 31120 + }, + { + "epoch": 0.31121, + "grad_norm": 1.1310641546404685, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 31121 + }, + { + "epoch": 0.31122, + "grad_norm": 0.9822622118416431, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 31122 + }, + { + "epoch": 0.31123, + "grad_norm": 1.107019763779298, + "learning_rate": 0.003, + "loss": 4.056, + "step": 31123 + }, + { + "epoch": 0.31124, + "grad_norm": 1.1811563524659914, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 31124 + }, + { + "epoch": 0.31125, + "grad_norm": 0.9069847229068467, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 31125 + }, + { + "epoch": 0.31126, + "grad_norm": 1.098746375053832, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 31126 + }, + { + "epoch": 0.31127, + "grad_norm": 1.0644167880277882, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 31127 + }, + { + "epoch": 0.31128, + "grad_norm": 1.0170881373447185, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 31128 + }, + { + "epoch": 0.31129, + "grad_norm": 0.9996443216280273, + "learning_rate": 0.003, + "loss": 4.031, + "step": 31129 + }, + { + "epoch": 0.3113, + "grad_norm": 0.9470438892979596, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 31130 + }, + { + "epoch": 0.31131, + "grad_norm": 1.0668899994593042, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 31131 + }, + { + "epoch": 0.31132, + "grad_norm": 1.0330807656003174, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 31132 + }, + { + "epoch": 0.31133, + "grad_norm": 1.0462089527123721, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 31133 + }, + { + "epoch": 0.31134, + "grad_norm": 1.0848690347314152, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 31134 + }, + { + "epoch": 0.31135, + "grad_norm": 1.0640469476503545, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 31135 + }, + { + "epoch": 0.31136, + "grad_norm": 1.287853730531451, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 31136 + }, + { + "epoch": 0.31137, + "grad_norm": 1.0725178259744759, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 31137 + }, + { + "epoch": 0.31138, + "grad_norm": 1.1319647741600662, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 31138 + }, + { + "epoch": 0.31139, + "grad_norm": 0.9687390990502377, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 31139 + }, + { + "epoch": 0.3114, + "grad_norm": 0.9320005185621825, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 31140 + }, + { + "epoch": 0.31141, + "grad_norm": 1.077111047793902, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 31141 + }, + { + "epoch": 0.31142, + "grad_norm": 0.982728017201385, + "learning_rate": 0.003, + "loss": 4.05, + "step": 31142 + }, + { + "epoch": 0.31143, + "grad_norm": 1.0835420617569198, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 31143 + }, + { + "epoch": 0.31144, + "grad_norm": 1.128647170248084, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 31144 + }, + { + "epoch": 0.31145, + "grad_norm": 1.0901780941018022, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 31145 + }, + { + "epoch": 0.31146, + "grad_norm": 1.1187916692880178, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 31146 + }, + { + "epoch": 0.31147, + "grad_norm": 0.9995766343194284, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 31147 + }, + { + "epoch": 0.31148, + "grad_norm": 1.2788343649281282, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 31148 + }, + { + "epoch": 0.31149, + "grad_norm": 1.0963869448534538, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 31149 + }, + { + "epoch": 0.3115, + "grad_norm": 1.3220680721049427, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 31150 + }, + { + "epoch": 0.31151, + "grad_norm": 0.8874910773311713, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 31151 + }, + { + "epoch": 0.31152, + "grad_norm": 0.9687636440842625, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 31152 + }, + { + "epoch": 0.31153, + "grad_norm": 1.231745146237752, + "learning_rate": 0.003, + "loss": 4.0886, + "step": 31153 + }, + { + "epoch": 0.31154, + "grad_norm": 0.8738838839493597, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 31154 + }, + { + "epoch": 0.31155, + "grad_norm": 1.008292591494486, + "learning_rate": 0.003, + "loss": 4.055, + "step": 31155 + }, + { + "epoch": 0.31156, + "grad_norm": 1.1214282929250794, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 31156 + }, + { + "epoch": 0.31157, + "grad_norm": 1.0007625886245206, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 31157 + }, + { + "epoch": 0.31158, + "grad_norm": 1.1656383795660525, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 31158 + }, + { + "epoch": 0.31159, + "grad_norm": 1.1971707269845608, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 31159 + }, + { + "epoch": 0.3116, + "grad_norm": 1.01788574784325, + "learning_rate": 0.003, + "loss": 4.05, + "step": 31160 + }, + { + "epoch": 0.31161, + "grad_norm": 1.2225104651728367, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 31161 + }, + { + "epoch": 0.31162, + "grad_norm": 0.9985429890387931, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 31162 + }, + { + "epoch": 0.31163, + "grad_norm": 1.1507488444493577, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 31163 + }, + { + "epoch": 0.31164, + "grad_norm": 1.0568685495259695, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 31164 + }, + { + "epoch": 0.31165, + "grad_norm": 1.1275961746296084, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 31165 + }, + { + "epoch": 0.31166, + "grad_norm": 0.8570487182286823, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 31166 + }, + { + "epoch": 0.31167, + "grad_norm": 0.9196921162108196, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 31167 + }, + { + "epoch": 0.31168, + "grad_norm": 1.1319176213995752, + "learning_rate": 0.003, + "loss": 4.028, + "step": 31168 + }, + { + "epoch": 0.31169, + "grad_norm": 1.1253900422428758, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 31169 + }, + { + "epoch": 0.3117, + "grad_norm": 1.1903587778784233, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 31170 + }, + { + "epoch": 0.31171, + "grad_norm": 1.0494321465171839, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 31171 + }, + { + "epoch": 0.31172, + "grad_norm": 1.0144177232468017, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 31172 + }, + { + "epoch": 0.31173, + "grad_norm": 1.0583294138412132, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 31173 + }, + { + "epoch": 0.31174, + "grad_norm": 0.9357171418460071, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 31174 + }, + { + "epoch": 0.31175, + "grad_norm": 1.001945080513558, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 31175 + }, + { + "epoch": 0.31176, + "grad_norm": 1.335389912605227, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 31176 + }, + { + "epoch": 0.31177, + "grad_norm": 1.133536534324495, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 31177 + }, + { + "epoch": 0.31178, + "grad_norm": 0.9970672434076979, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 31178 + }, + { + "epoch": 0.31179, + "grad_norm": 1.3858182907041037, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 31179 + }, + { + "epoch": 0.3118, + "grad_norm": 1.0005259702563039, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 31180 + }, + { + "epoch": 0.31181, + "grad_norm": 1.0110639863185473, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 31181 + }, + { + "epoch": 0.31182, + "grad_norm": 1.0558695515304097, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 31182 + }, + { + "epoch": 0.31183, + "grad_norm": 1.0838269182354985, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 31183 + }, + { + "epoch": 0.31184, + "grad_norm": 1.0513494206523009, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 31184 + }, + { + "epoch": 0.31185, + "grad_norm": 0.9823803173088815, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 31185 + }, + { + "epoch": 0.31186, + "grad_norm": 1.3504916669024294, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 31186 + }, + { + "epoch": 0.31187, + "grad_norm": 1.0231560432245956, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 31187 + }, + { + "epoch": 0.31188, + "grad_norm": 1.1131803465949277, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 31188 + }, + { + "epoch": 0.31189, + "grad_norm": 1.1294956341109916, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 31189 + }, + { + "epoch": 0.3119, + "grad_norm": 1.3029542704674424, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 31190 + }, + { + "epoch": 0.31191, + "grad_norm": 0.9654556814906635, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 31191 + }, + { + "epoch": 0.31192, + "grad_norm": 1.023543475471987, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 31192 + }, + { + "epoch": 0.31193, + "grad_norm": 1.0468508221278134, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 31193 + }, + { + "epoch": 0.31194, + "grad_norm": 1.066096209556734, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 31194 + }, + { + "epoch": 0.31195, + "grad_norm": 1.0913651186968027, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 31195 + }, + { + "epoch": 0.31196, + "grad_norm": 1.0237344931579706, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 31196 + }, + { + "epoch": 0.31197, + "grad_norm": 1.1087184376303911, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 31197 + }, + { + "epoch": 0.31198, + "grad_norm": 1.0882781929851806, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 31198 + }, + { + "epoch": 0.31199, + "grad_norm": 1.0100116977229987, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 31199 + }, + { + "epoch": 0.312, + "grad_norm": 1.0554664474158966, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 31200 + }, + { + "epoch": 0.31201, + "grad_norm": 1.055713412129083, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 31201 + }, + { + "epoch": 0.31202, + "grad_norm": 1.0958492190853786, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 31202 + }, + { + "epoch": 0.31203, + "grad_norm": 0.9605903383235717, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 31203 + }, + { + "epoch": 0.31204, + "grad_norm": 1.007574773491165, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 31204 + }, + { + "epoch": 0.31205, + "grad_norm": 1.048848797945928, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 31205 + }, + { + "epoch": 0.31206, + "grad_norm": 1.1070343638099376, + "learning_rate": 0.003, + "loss": 4.055, + "step": 31206 + }, + { + "epoch": 0.31207, + "grad_norm": 1.16147919139706, + "learning_rate": 0.003, + "loss": 4.061, + "step": 31207 + }, + { + "epoch": 0.31208, + "grad_norm": 0.9850215364964423, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 31208 + }, + { + "epoch": 0.31209, + "grad_norm": 1.0904987851516854, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 31209 + }, + { + "epoch": 0.3121, + "grad_norm": 1.3187728230976452, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 31210 + }, + { + "epoch": 0.31211, + "grad_norm": 1.1684375280437753, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 31211 + }, + { + "epoch": 0.31212, + "grad_norm": 1.0234968272645513, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 31212 + }, + { + "epoch": 0.31213, + "grad_norm": 1.0632913180571202, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 31213 + }, + { + "epoch": 0.31214, + "grad_norm": 1.011318975098741, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 31214 + }, + { + "epoch": 0.31215, + "grad_norm": 1.1493004463324574, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 31215 + }, + { + "epoch": 0.31216, + "grad_norm": 0.9451328206472692, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 31216 + }, + { + "epoch": 0.31217, + "grad_norm": 1.1716034784246494, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 31217 + }, + { + "epoch": 0.31218, + "grad_norm": 1.0587314261666423, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 31218 + }, + { + "epoch": 0.31219, + "grad_norm": 0.9574129214398036, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 31219 + }, + { + "epoch": 0.3122, + "grad_norm": 0.8134007623034055, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 31220 + }, + { + "epoch": 0.31221, + "grad_norm": 0.9950701739139122, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 31221 + }, + { + "epoch": 0.31222, + "grad_norm": 1.1785570203984068, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 31222 + }, + { + "epoch": 0.31223, + "grad_norm": 0.9049956799743089, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 31223 + }, + { + "epoch": 0.31224, + "grad_norm": 1.1006958805202836, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 31224 + }, + { + "epoch": 0.31225, + "grad_norm": 0.8936038506560671, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 31225 + }, + { + "epoch": 0.31226, + "grad_norm": 0.9758813989706462, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 31226 + }, + { + "epoch": 0.31227, + "grad_norm": 1.1544539697673915, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 31227 + }, + { + "epoch": 0.31228, + "grad_norm": 0.9499562729768172, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 31228 + }, + { + "epoch": 0.31229, + "grad_norm": 1.1471686315957623, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 31229 + }, + { + "epoch": 0.3123, + "grad_norm": 1.0561044122842325, + "learning_rate": 0.003, + "loss": 4.1019, + "step": 31230 + }, + { + "epoch": 0.31231, + "grad_norm": 1.0262850139737234, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 31231 + }, + { + "epoch": 0.31232, + "grad_norm": 0.9604715111931126, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 31232 + }, + { + "epoch": 0.31233, + "grad_norm": 0.989826502306155, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 31233 + }, + { + "epoch": 0.31234, + "grad_norm": 1.1152441613170667, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 31234 + }, + { + "epoch": 0.31235, + "grad_norm": 1.017068182156696, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 31235 + }, + { + "epoch": 0.31236, + "grad_norm": 1.3279958494134552, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 31236 + }, + { + "epoch": 0.31237, + "grad_norm": 1.2156108631701523, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 31237 + }, + { + "epoch": 0.31238, + "grad_norm": 1.4750225027489616, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 31238 + }, + { + "epoch": 0.31239, + "grad_norm": 0.8273370609929234, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 31239 + }, + { + "epoch": 0.3124, + "grad_norm": 0.7182845851047083, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 31240 + }, + { + "epoch": 0.31241, + "grad_norm": 0.9246500749270103, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 31241 + }, + { + "epoch": 0.31242, + "grad_norm": 1.016206467909732, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 31242 + }, + { + "epoch": 0.31243, + "grad_norm": 1.3414695725410202, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 31243 + }, + { + "epoch": 0.31244, + "grad_norm": 1.056856024573829, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 31244 + }, + { + "epoch": 0.31245, + "grad_norm": 1.0225102397032735, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 31245 + }, + { + "epoch": 0.31246, + "grad_norm": 1.3282568054323103, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 31246 + }, + { + "epoch": 0.31247, + "grad_norm": 1.1230086162946078, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 31247 + }, + { + "epoch": 0.31248, + "grad_norm": 1.1423317171818548, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 31248 + }, + { + "epoch": 0.31249, + "grad_norm": 0.9921608009385177, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 31249 + }, + { + "epoch": 0.3125, + "grad_norm": 0.8895932857812643, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 31250 + }, + { + "epoch": 0.31251, + "grad_norm": 0.8888608389399953, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 31251 + }, + { + "epoch": 0.31252, + "grad_norm": 0.827304598133061, + "learning_rate": 0.003, + "loss": 4.048, + "step": 31252 + }, + { + "epoch": 0.31253, + "grad_norm": 0.9439972414187789, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 31253 + }, + { + "epoch": 0.31254, + "grad_norm": 1.017429967338126, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 31254 + }, + { + "epoch": 0.31255, + "grad_norm": 1.0701918557385874, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 31255 + }, + { + "epoch": 0.31256, + "grad_norm": 1.256234316798439, + "learning_rate": 0.003, + "loss": 4.075, + "step": 31256 + }, + { + "epoch": 0.31257, + "grad_norm": 1.0221384149506383, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 31257 + }, + { + "epoch": 0.31258, + "grad_norm": 1.2315147972002456, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 31258 + }, + { + "epoch": 0.31259, + "grad_norm": 0.919039675677266, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 31259 + }, + { + "epoch": 0.3126, + "grad_norm": 1.0837424821478714, + "learning_rate": 0.003, + "loss": 4.04, + "step": 31260 + }, + { + "epoch": 0.31261, + "grad_norm": 1.070751581345542, + "learning_rate": 0.003, + "loss": 4.072, + "step": 31261 + }, + { + "epoch": 0.31262, + "grad_norm": 1.1216040525722042, + "learning_rate": 0.003, + "loss": 4.064, + "step": 31262 + }, + { + "epoch": 0.31263, + "grad_norm": 1.3739724809925753, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 31263 + }, + { + "epoch": 0.31264, + "grad_norm": 0.9608465714334052, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 31264 + }, + { + "epoch": 0.31265, + "grad_norm": 1.0030826412588514, + "learning_rate": 0.003, + "loss": 4.038, + "step": 31265 + }, + { + "epoch": 0.31266, + "grad_norm": 1.3788709106264132, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 31266 + }, + { + "epoch": 0.31267, + "grad_norm": 0.7879902430781655, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 31267 + }, + { + "epoch": 0.31268, + "grad_norm": 0.8101998478709831, + "learning_rate": 0.003, + "loss": 4.026, + "step": 31268 + }, + { + "epoch": 0.31269, + "grad_norm": 0.9228896325727723, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 31269 + }, + { + "epoch": 0.3127, + "grad_norm": 1.1101706053972489, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 31270 + }, + { + "epoch": 0.31271, + "grad_norm": 0.9852945929659432, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 31271 + }, + { + "epoch": 0.31272, + "grad_norm": 0.9995128334489907, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 31272 + }, + { + "epoch": 0.31273, + "grad_norm": 1.1878265360130849, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 31273 + }, + { + "epoch": 0.31274, + "grad_norm": 1.0230519569914163, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 31274 + }, + { + "epoch": 0.31275, + "grad_norm": 1.021483038342436, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 31275 + }, + { + "epoch": 0.31276, + "grad_norm": 1.1592146512008725, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 31276 + }, + { + "epoch": 0.31277, + "grad_norm": 1.2793092771419827, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 31277 + }, + { + "epoch": 0.31278, + "grad_norm": 1.1893414989692455, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 31278 + }, + { + "epoch": 0.31279, + "grad_norm": 1.0391080334116978, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 31279 + }, + { + "epoch": 0.3128, + "grad_norm": 0.9749854649658818, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 31280 + }, + { + "epoch": 0.31281, + "grad_norm": 1.0141691519929883, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 31281 + }, + { + "epoch": 0.31282, + "grad_norm": 1.1345419297436738, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 31282 + }, + { + "epoch": 0.31283, + "grad_norm": 1.1826685152934544, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 31283 + }, + { + "epoch": 0.31284, + "grad_norm": 0.9822087853867598, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 31284 + }, + { + "epoch": 0.31285, + "grad_norm": 1.2618939138047602, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 31285 + }, + { + "epoch": 0.31286, + "grad_norm": 0.9963515340344505, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 31286 + }, + { + "epoch": 0.31287, + "grad_norm": 1.0835120094661241, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 31287 + }, + { + "epoch": 0.31288, + "grad_norm": 1.1625167064076927, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 31288 + }, + { + "epoch": 0.31289, + "grad_norm": 1.1933723576475364, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 31289 + }, + { + "epoch": 0.3129, + "grad_norm": 0.9996331794978627, + "learning_rate": 0.003, + "loss": 4.052, + "step": 31290 + }, + { + "epoch": 0.31291, + "grad_norm": 1.1114409811344172, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 31291 + }, + { + "epoch": 0.31292, + "grad_norm": 0.9587700495421124, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 31292 + }, + { + "epoch": 0.31293, + "grad_norm": 0.9121595307171488, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 31293 + }, + { + "epoch": 0.31294, + "grad_norm": 0.8983922110863501, + "learning_rate": 0.003, + "loss": 4.078, + "step": 31294 + }, + { + "epoch": 0.31295, + "grad_norm": 1.0626920197604939, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 31295 + }, + { + "epoch": 0.31296, + "grad_norm": 1.2022892726652816, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 31296 + }, + { + "epoch": 0.31297, + "grad_norm": 1.023094961471605, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 31297 + }, + { + "epoch": 0.31298, + "grad_norm": 1.2809997654382104, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 31298 + }, + { + "epoch": 0.31299, + "grad_norm": 0.9551957851312467, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 31299 + }, + { + "epoch": 0.313, + "grad_norm": 1.0895589570841724, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 31300 + }, + { + "epoch": 0.31301, + "grad_norm": 1.0460988093525094, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 31301 + }, + { + "epoch": 0.31302, + "grad_norm": 1.2181371933678917, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 31302 + }, + { + "epoch": 0.31303, + "grad_norm": 1.0459201463639605, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 31303 + }, + { + "epoch": 0.31304, + "grad_norm": 1.18749639154912, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 31304 + }, + { + "epoch": 0.31305, + "grad_norm": 0.8904341658976276, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 31305 + }, + { + "epoch": 0.31306, + "grad_norm": 0.9703228110153403, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 31306 + }, + { + "epoch": 0.31307, + "grad_norm": 1.1731291392891925, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 31307 + }, + { + "epoch": 0.31308, + "grad_norm": 0.8966999876280006, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 31308 + }, + { + "epoch": 0.31309, + "grad_norm": 0.9949235262778449, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 31309 + }, + { + "epoch": 0.3131, + "grad_norm": 1.2251133846542084, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 31310 + }, + { + "epoch": 0.31311, + "grad_norm": 0.9143287273288814, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 31311 + }, + { + "epoch": 0.31312, + "grad_norm": 0.8569141405379002, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 31312 + }, + { + "epoch": 0.31313, + "grad_norm": 0.8804529535924696, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 31313 + }, + { + "epoch": 0.31314, + "grad_norm": 1.1917591685922426, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 31314 + }, + { + "epoch": 0.31315, + "grad_norm": 1.0558748115212944, + "learning_rate": 0.003, + "loss": 4.0967, + "step": 31315 + }, + { + "epoch": 0.31316, + "grad_norm": 1.2174910180837153, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 31316 + }, + { + "epoch": 0.31317, + "grad_norm": 1.2684694553674072, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 31317 + }, + { + "epoch": 0.31318, + "grad_norm": 0.9913832286498128, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 31318 + }, + { + "epoch": 0.31319, + "grad_norm": 1.0769113298095092, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 31319 + }, + { + "epoch": 0.3132, + "grad_norm": 1.1691167545655692, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 31320 + }, + { + "epoch": 0.31321, + "grad_norm": 1.0365699820574183, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 31321 + }, + { + "epoch": 0.31322, + "grad_norm": 1.1688144265719, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 31322 + }, + { + "epoch": 0.31323, + "grad_norm": 0.8986422172422711, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 31323 + }, + { + "epoch": 0.31324, + "grad_norm": 1.0220965834775786, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 31324 + }, + { + "epoch": 0.31325, + "grad_norm": 0.9933904009154347, + "learning_rate": 0.003, + "loss": 4.034, + "step": 31325 + }, + { + "epoch": 0.31326, + "grad_norm": 1.0089106233393044, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 31326 + }, + { + "epoch": 0.31327, + "grad_norm": 1.113187201218263, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 31327 + }, + { + "epoch": 0.31328, + "grad_norm": 0.9408400889315093, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 31328 + }, + { + "epoch": 0.31329, + "grad_norm": 1.0626170552405854, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 31329 + }, + { + "epoch": 0.3133, + "grad_norm": 1.0810972653214361, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 31330 + }, + { + "epoch": 0.31331, + "grad_norm": 1.2433571976107503, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 31331 + }, + { + "epoch": 0.31332, + "grad_norm": 1.2913084448695904, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 31332 + }, + { + "epoch": 0.31333, + "grad_norm": 1.0948596443600782, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 31333 + }, + { + "epoch": 0.31334, + "grad_norm": 1.1230534939792134, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 31334 + }, + { + "epoch": 0.31335, + "grad_norm": 1.0634984908932659, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 31335 + }, + { + "epoch": 0.31336, + "grad_norm": 1.2536737004120921, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 31336 + }, + { + "epoch": 0.31337, + "grad_norm": 1.0120023823804307, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 31337 + }, + { + "epoch": 0.31338, + "grad_norm": 1.149905226620856, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 31338 + }, + { + "epoch": 0.31339, + "grad_norm": 1.0088626543649326, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 31339 + }, + { + "epoch": 0.3134, + "grad_norm": 1.0444360360618516, + "learning_rate": 0.003, + "loss": 4.035, + "step": 31340 + }, + { + "epoch": 0.31341, + "grad_norm": 1.113547349704173, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 31341 + }, + { + "epoch": 0.31342, + "grad_norm": 1.0110084276235007, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 31342 + }, + { + "epoch": 0.31343, + "grad_norm": 1.1523304705768025, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 31343 + }, + { + "epoch": 0.31344, + "grad_norm": 1.0251931468899216, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 31344 + }, + { + "epoch": 0.31345, + "grad_norm": 1.0202179155073494, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 31345 + }, + { + "epoch": 0.31346, + "grad_norm": 1.0213168543014455, + "learning_rate": 0.003, + "loss": 4.063, + "step": 31346 + }, + { + "epoch": 0.31347, + "grad_norm": 1.2198111497163804, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 31347 + }, + { + "epoch": 0.31348, + "grad_norm": 1.160347487194854, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 31348 + }, + { + "epoch": 0.31349, + "grad_norm": 0.9456559547231573, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 31349 + }, + { + "epoch": 0.3135, + "grad_norm": 0.9980262389379452, + "learning_rate": 0.003, + "loss": 4.049, + "step": 31350 + }, + { + "epoch": 0.31351, + "grad_norm": 1.0452757281395981, + "learning_rate": 0.003, + "loss": 4.057, + "step": 31351 + }, + { + "epoch": 0.31352, + "grad_norm": 1.02953443344387, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 31352 + }, + { + "epoch": 0.31353, + "grad_norm": 1.1034820274891988, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 31353 + }, + { + "epoch": 0.31354, + "grad_norm": 0.9955686530074677, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 31354 + }, + { + "epoch": 0.31355, + "grad_norm": 1.264446459669219, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 31355 + }, + { + "epoch": 0.31356, + "grad_norm": 0.9652900683902331, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 31356 + }, + { + "epoch": 0.31357, + "grad_norm": 1.1832607139486984, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 31357 + }, + { + "epoch": 0.31358, + "grad_norm": 1.1132165699793595, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 31358 + }, + { + "epoch": 0.31359, + "grad_norm": 1.0099588609091568, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 31359 + }, + { + "epoch": 0.3136, + "grad_norm": 1.0700552063392073, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 31360 + }, + { + "epoch": 0.31361, + "grad_norm": 1.0290356330676345, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 31361 + }, + { + "epoch": 0.31362, + "grad_norm": 1.131615714489285, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 31362 + }, + { + "epoch": 0.31363, + "grad_norm": 1.0309312419581096, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 31363 + }, + { + "epoch": 0.31364, + "grad_norm": 1.0287853366733137, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 31364 + }, + { + "epoch": 0.31365, + "grad_norm": 1.0549196002158217, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 31365 + }, + { + "epoch": 0.31366, + "grad_norm": 0.9741433303322998, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 31366 + }, + { + "epoch": 0.31367, + "grad_norm": 1.217475033341245, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 31367 + }, + { + "epoch": 0.31368, + "grad_norm": 1.0681081910807106, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 31368 + }, + { + "epoch": 0.31369, + "grad_norm": 1.1665263759503162, + "learning_rate": 0.003, + "loss": 4.031, + "step": 31369 + }, + { + "epoch": 0.3137, + "grad_norm": 1.0060485125362262, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 31370 + }, + { + "epoch": 0.31371, + "grad_norm": 1.209476574690086, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 31371 + }, + { + "epoch": 0.31372, + "grad_norm": 0.988232889999719, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 31372 + }, + { + "epoch": 0.31373, + "grad_norm": 1.1373526629184936, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 31373 + }, + { + "epoch": 0.31374, + "grad_norm": 1.2049531608450719, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 31374 + }, + { + "epoch": 0.31375, + "grad_norm": 1.076013446663324, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 31375 + }, + { + "epoch": 0.31376, + "grad_norm": 1.0514917591785076, + "learning_rate": 0.003, + "loss": 4.051, + "step": 31376 + }, + { + "epoch": 0.31377, + "grad_norm": 1.0331033899223132, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 31377 + }, + { + "epoch": 0.31378, + "grad_norm": 1.193824608144625, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 31378 + }, + { + "epoch": 0.31379, + "grad_norm": 1.0577989713551268, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 31379 + }, + { + "epoch": 0.3138, + "grad_norm": 1.2047516390914352, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 31380 + }, + { + "epoch": 0.31381, + "grad_norm": 1.2253752867013863, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 31381 + }, + { + "epoch": 0.31382, + "grad_norm": 0.868966946769091, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 31382 + }, + { + "epoch": 0.31383, + "grad_norm": 0.9670753241593614, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 31383 + }, + { + "epoch": 0.31384, + "grad_norm": 1.2583232990070807, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 31384 + }, + { + "epoch": 0.31385, + "grad_norm": 1.104629835760925, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 31385 + }, + { + "epoch": 0.31386, + "grad_norm": 0.9692261514580957, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 31386 + }, + { + "epoch": 0.31387, + "grad_norm": 0.9390100559871084, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 31387 + }, + { + "epoch": 0.31388, + "grad_norm": 0.9737005717791583, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 31388 + }, + { + "epoch": 0.31389, + "grad_norm": 1.1556520324097062, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 31389 + }, + { + "epoch": 0.3139, + "grad_norm": 1.1911076833556633, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 31390 + }, + { + "epoch": 0.31391, + "grad_norm": 1.1954043132452912, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 31391 + }, + { + "epoch": 0.31392, + "grad_norm": 0.9298143349283309, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 31392 + }, + { + "epoch": 0.31393, + "grad_norm": 0.8209657655833752, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 31393 + }, + { + "epoch": 0.31394, + "grad_norm": 1.0580418291297802, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 31394 + }, + { + "epoch": 0.31395, + "grad_norm": 1.2413850121199186, + "learning_rate": 0.003, + "loss": 4.035, + "step": 31395 + }, + { + "epoch": 0.31396, + "grad_norm": 0.8911420795218675, + "learning_rate": 0.003, + "loss": 4.042, + "step": 31396 + }, + { + "epoch": 0.31397, + "grad_norm": 1.1117686630505164, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 31397 + }, + { + "epoch": 0.31398, + "grad_norm": 1.2005133300188475, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 31398 + }, + { + "epoch": 0.31399, + "grad_norm": 1.0332268184494093, + "learning_rate": 0.003, + "loss": 4.038, + "step": 31399 + }, + { + "epoch": 0.314, + "grad_norm": 1.0466753122657957, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 31400 + }, + { + "epoch": 0.31401, + "grad_norm": 1.023935368884896, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 31401 + }, + { + "epoch": 0.31402, + "grad_norm": 1.0772877282268072, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 31402 + }, + { + "epoch": 0.31403, + "grad_norm": 0.9739555997164806, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 31403 + }, + { + "epoch": 0.31404, + "grad_norm": 1.3177713829253557, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 31404 + }, + { + "epoch": 0.31405, + "grad_norm": 1.0041383158314068, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 31405 + }, + { + "epoch": 0.31406, + "grad_norm": 1.1869261065730314, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 31406 + }, + { + "epoch": 0.31407, + "grad_norm": 1.0698607548348034, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 31407 + }, + { + "epoch": 0.31408, + "grad_norm": 1.385798225101265, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 31408 + }, + { + "epoch": 0.31409, + "grad_norm": 0.8192381229051299, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 31409 + }, + { + "epoch": 0.3141, + "grad_norm": 0.8337949798137465, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 31410 + }, + { + "epoch": 0.31411, + "grad_norm": 1.091702262762848, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 31411 + }, + { + "epoch": 0.31412, + "grad_norm": 1.0340587243269657, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 31412 + }, + { + "epoch": 0.31413, + "grad_norm": 1.278581535335175, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 31413 + }, + { + "epoch": 0.31414, + "grad_norm": 0.9846815880522126, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 31414 + }, + { + "epoch": 0.31415, + "grad_norm": 1.3227540077466335, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 31415 + }, + { + "epoch": 0.31416, + "grad_norm": 0.9680913725807997, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 31416 + }, + { + "epoch": 0.31417, + "grad_norm": 1.00432015402786, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 31417 + }, + { + "epoch": 0.31418, + "grad_norm": 1.1998684580102001, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 31418 + }, + { + "epoch": 0.31419, + "grad_norm": 1.1114349692696917, + "learning_rate": 0.003, + "loss": 4.06, + "step": 31419 + }, + { + "epoch": 0.3142, + "grad_norm": 0.9761947756520858, + "learning_rate": 0.003, + "loss": 4.024, + "step": 31420 + }, + { + "epoch": 0.31421, + "grad_norm": 0.8899576837175892, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 31421 + }, + { + "epoch": 0.31422, + "grad_norm": 1.0372040111601204, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 31422 + }, + { + "epoch": 0.31423, + "grad_norm": 1.3276529533340018, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 31423 + }, + { + "epoch": 0.31424, + "grad_norm": 0.971621533985379, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 31424 + }, + { + "epoch": 0.31425, + "grad_norm": 1.14240400100827, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 31425 + }, + { + "epoch": 0.31426, + "grad_norm": 1.0865700182227653, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 31426 + }, + { + "epoch": 0.31427, + "grad_norm": 1.069578927029404, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 31427 + }, + { + "epoch": 0.31428, + "grad_norm": 1.2807257624499906, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 31428 + }, + { + "epoch": 0.31429, + "grad_norm": 0.95946059206922, + "learning_rate": 0.003, + "loss": 4.037, + "step": 31429 + }, + { + "epoch": 0.3143, + "grad_norm": 1.0367962146650027, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 31430 + }, + { + "epoch": 0.31431, + "grad_norm": 1.1904230096418587, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 31431 + }, + { + "epoch": 0.31432, + "grad_norm": 0.9006886407604989, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 31432 + }, + { + "epoch": 0.31433, + "grad_norm": 0.9383280121660359, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 31433 + }, + { + "epoch": 0.31434, + "grad_norm": 1.0105981817782284, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 31434 + }, + { + "epoch": 0.31435, + "grad_norm": 1.0955615898651503, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 31435 + }, + { + "epoch": 0.31436, + "grad_norm": 1.128561215178169, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 31436 + }, + { + "epoch": 0.31437, + "grad_norm": 1.033255863430136, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 31437 + }, + { + "epoch": 0.31438, + "grad_norm": 1.0388058639488684, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 31438 + }, + { + "epoch": 0.31439, + "grad_norm": 1.1073908983181122, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 31439 + }, + { + "epoch": 0.3144, + "grad_norm": 1.2125293208495007, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 31440 + }, + { + "epoch": 0.31441, + "grad_norm": 0.864124172678936, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 31441 + }, + { + "epoch": 0.31442, + "grad_norm": 1.1205199493968268, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 31442 + }, + { + "epoch": 0.31443, + "grad_norm": 1.330108780049935, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 31443 + }, + { + "epoch": 0.31444, + "grad_norm": 0.8250898814362734, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 31444 + }, + { + "epoch": 0.31445, + "grad_norm": 0.8190738527373437, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 31445 + }, + { + "epoch": 0.31446, + "grad_norm": 0.9729311814452727, + "learning_rate": 0.003, + "loss": 4.048, + "step": 31446 + }, + { + "epoch": 0.31447, + "grad_norm": 1.255149433876571, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 31447 + }, + { + "epoch": 0.31448, + "grad_norm": 0.9212612494004746, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 31448 + }, + { + "epoch": 0.31449, + "grad_norm": 0.9567689274607274, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 31449 + }, + { + "epoch": 0.3145, + "grad_norm": 1.149749245620219, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 31450 + }, + { + "epoch": 0.31451, + "grad_norm": 1.3714880024099823, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 31451 + }, + { + "epoch": 0.31452, + "grad_norm": 0.8242549975063876, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 31452 + }, + { + "epoch": 0.31453, + "grad_norm": 0.8913123745443219, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 31453 + }, + { + "epoch": 0.31454, + "grad_norm": 0.9537208235892004, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 31454 + }, + { + "epoch": 0.31455, + "grad_norm": 1.342898071069329, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 31455 + }, + { + "epoch": 0.31456, + "grad_norm": 0.977920001794439, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 31456 + }, + { + "epoch": 0.31457, + "grad_norm": 1.1753934360827818, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 31457 + }, + { + "epoch": 0.31458, + "grad_norm": 1.0709180263238773, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 31458 + }, + { + "epoch": 0.31459, + "grad_norm": 1.1531417690969261, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 31459 + }, + { + "epoch": 0.3146, + "grad_norm": 1.2484676908975314, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 31460 + }, + { + "epoch": 0.31461, + "grad_norm": 0.9490029299006396, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 31461 + }, + { + "epoch": 0.31462, + "grad_norm": 1.0316890332378006, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 31462 + }, + { + "epoch": 0.31463, + "grad_norm": 1.251288862968565, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 31463 + }, + { + "epoch": 0.31464, + "grad_norm": 0.8474322440693852, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 31464 + }, + { + "epoch": 0.31465, + "grad_norm": 0.7743604215318405, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 31465 + }, + { + "epoch": 0.31466, + "grad_norm": 0.8568452834624624, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 31466 + }, + { + "epoch": 0.31467, + "grad_norm": 1.0969664873086156, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 31467 + }, + { + "epoch": 0.31468, + "grad_norm": 1.0518681498279405, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 31468 + }, + { + "epoch": 0.31469, + "grad_norm": 1.085745558690428, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 31469 + }, + { + "epoch": 0.3147, + "grad_norm": 1.0389179148433332, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 31470 + }, + { + "epoch": 0.31471, + "grad_norm": 1.0660055117099865, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 31471 + }, + { + "epoch": 0.31472, + "grad_norm": 0.9586277517453398, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 31472 + }, + { + "epoch": 0.31473, + "grad_norm": 1.1644790552922812, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 31473 + }, + { + "epoch": 0.31474, + "grad_norm": 1.01900501837311, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 31474 + }, + { + "epoch": 0.31475, + "grad_norm": 1.1651144195912926, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 31475 + }, + { + "epoch": 0.31476, + "grad_norm": 0.9820703207685064, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 31476 + }, + { + "epoch": 0.31477, + "grad_norm": 1.0547509158592898, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 31477 + }, + { + "epoch": 0.31478, + "grad_norm": 0.9712492637053047, + "learning_rate": 0.003, + "loss": 4.031, + "step": 31478 + }, + { + "epoch": 0.31479, + "grad_norm": 1.1403852808065105, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 31479 + }, + { + "epoch": 0.3148, + "grad_norm": 1.17427406421239, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 31480 + }, + { + "epoch": 0.31481, + "grad_norm": 1.033039011762272, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 31481 + }, + { + "epoch": 0.31482, + "grad_norm": 1.0010002429610256, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 31482 + }, + { + "epoch": 0.31483, + "grad_norm": 1.2537555172353991, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 31483 + }, + { + "epoch": 0.31484, + "grad_norm": 1.0508702666208425, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 31484 + }, + { + "epoch": 0.31485, + "grad_norm": 1.0968211699739991, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 31485 + }, + { + "epoch": 0.31486, + "grad_norm": 1.0143378747194145, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 31486 + }, + { + "epoch": 0.31487, + "grad_norm": 0.8907578800965966, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 31487 + }, + { + "epoch": 0.31488, + "grad_norm": 0.8509518027462518, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 31488 + }, + { + "epoch": 0.31489, + "grad_norm": 1.1457365227285123, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 31489 + }, + { + "epoch": 0.3149, + "grad_norm": 1.3186519985826326, + "learning_rate": 0.003, + "loss": 4.082, + "step": 31490 + }, + { + "epoch": 0.31491, + "grad_norm": 0.7352291015651583, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 31491 + }, + { + "epoch": 0.31492, + "grad_norm": 0.833511420734211, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 31492 + }, + { + "epoch": 0.31493, + "grad_norm": 0.8954976290326812, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 31493 + }, + { + "epoch": 0.31494, + "grad_norm": 1.0925106684464003, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 31494 + }, + { + "epoch": 0.31495, + "grad_norm": 0.9368830591509217, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 31495 + }, + { + "epoch": 0.31496, + "grad_norm": 1.0974643073565953, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 31496 + }, + { + "epoch": 0.31497, + "grad_norm": 1.2148098101513551, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 31497 + }, + { + "epoch": 0.31498, + "grad_norm": 1.0327720338220323, + "learning_rate": 0.003, + "loss": 4.028, + "step": 31498 + }, + { + "epoch": 0.31499, + "grad_norm": 1.2832607950380235, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 31499 + }, + { + "epoch": 0.315, + "grad_norm": 1.0622004890551666, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 31500 + }, + { + "epoch": 0.31501, + "grad_norm": 0.9498638213256548, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 31501 + }, + { + "epoch": 0.31502, + "grad_norm": 0.9091697850666165, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 31502 + }, + { + "epoch": 0.31503, + "grad_norm": 0.9909171775011989, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 31503 + }, + { + "epoch": 0.31504, + "grad_norm": 1.1894757721489442, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 31504 + }, + { + "epoch": 0.31505, + "grad_norm": 1.056143584649335, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 31505 + }, + { + "epoch": 0.31506, + "grad_norm": 1.2979863768616686, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 31506 + }, + { + "epoch": 0.31507, + "grad_norm": 0.7232408117875864, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 31507 + }, + { + "epoch": 0.31508, + "grad_norm": 0.7209449264718542, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 31508 + }, + { + "epoch": 0.31509, + "grad_norm": 0.8642800293758915, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 31509 + }, + { + "epoch": 0.3151, + "grad_norm": 1.3004086007982574, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 31510 + }, + { + "epoch": 0.31511, + "grad_norm": 1.2091642587336315, + "learning_rate": 0.003, + "loss": 4.025, + "step": 31511 + }, + { + "epoch": 0.31512, + "grad_norm": 1.0047379691084688, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 31512 + }, + { + "epoch": 0.31513, + "grad_norm": 1.3292424011860595, + "learning_rate": 0.003, + "loss": 4.042, + "step": 31513 + }, + { + "epoch": 0.31514, + "grad_norm": 1.0974725382841697, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 31514 + }, + { + "epoch": 0.31515, + "grad_norm": 0.8644628960251675, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 31515 + }, + { + "epoch": 0.31516, + "grad_norm": 0.8737609079057641, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 31516 + }, + { + "epoch": 0.31517, + "grad_norm": 1.1085045361609391, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 31517 + }, + { + "epoch": 0.31518, + "grad_norm": 1.0763792922809219, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 31518 + }, + { + "epoch": 0.31519, + "grad_norm": 0.9930418303686558, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 31519 + }, + { + "epoch": 0.3152, + "grad_norm": 1.3440236671593133, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 31520 + }, + { + "epoch": 0.31521, + "grad_norm": 0.9421792684763118, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 31521 + }, + { + "epoch": 0.31522, + "grad_norm": 1.0562719248942074, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 31522 + }, + { + "epoch": 0.31523, + "grad_norm": 1.162458854179107, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 31523 + }, + { + "epoch": 0.31524, + "grad_norm": 1.2233320765615905, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 31524 + }, + { + "epoch": 0.31525, + "grad_norm": 1.0128603557644322, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 31525 + }, + { + "epoch": 0.31526, + "grad_norm": 1.2639598579863125, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 31526 + }, + { + "epoch": 0.31527, + "grad_norm": 1.1016947910065393, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 31527 + }, + { + "epoch": 0.31528, + "grad_norm": 1.241921992417551, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 31528 + }, + { + "epoch": 0.31529, + "grad_norm": 0.9458928059005357, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 31529 + }, + { + "epoch": 0.3153, + "grad_norm": 1.1655975934314908, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 31530 + }, + { + "epoch": 0.31531, + "grad_norm": 1.0746449786858678, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 31531 + }, + { + "epoch": 0.31532, + "grad_norm": 1.1516042723635234, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 31532 + }, + { + "epoch": 0.31533, + "grad_norm": 1.1047877173674097, + "learning_rate": 0.003, + "loss": 4.055, + "step": 31533 + }, + { + "epoch": 0.31534, + "grad_norm": 1.1223428104771302, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 31534 + }, + { + "epoch": 0.31535, + "grad_norm": 1.1564764887850563, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 31535 + }, + { + "epoch": 0.31536, + "grad_norm": 0.8184277462759824, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 31536 + }, + { + "epoch": 0.31537, + "grad_norm": 0.9172306784966124, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 31537 + }, + { + "epoch": 0.31538, + "grad_norm": 0.9557822176055425, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 31538 + }, + { + "epoch": 0.31539, + "grad_norm": 0.9726947340544992, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 31539 + }, + { + "epoch": 0.3154, + "grad_norm": 1.1236428357385901, + "learning_rate": 0.003, + "loss": 4.046, + "step": 31540 + }, + { + "epoch": 0.31541, + "grad_norm": 0.9754789525618154, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 31541 + }, + { + "epoch": 0.31542, + "grad_norm": 1.040671762633709, + "learning_rate": 0.003, + "loss": 3.999, + "step": 31542 + }, + { + "epoch": 0.31543, + "grad_norm": 1.1088954190338995, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 31543 + }, + { + "epoch": 0.31544, + "grad_norm": 0.9246171170101278, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 31544 + }, + { + "epoch": 0.31545, + "grad_norm": 1.057389372259196, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 31545 + }, + { + "epoch": 0.31546, + "grad_norm": 1.1452638991850401, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 31546 + }, + { + "epoch": 0.31547, + "grad_norm": 1.3475962882078103, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 31547 + }, + { + "epoch": 0.31548, + "grad_norm": 1.1977230653654076, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 31548 + }, + { + "epoch": 0.31549, + "grad_norm": 0.9532237011027589, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 31549 + }, + { + "epoch": 0.3155, + "grad_norm": 1.1825304274459467, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 31550 + }, + { + "epoch": 0.31551, + "grad_norm": 1.1894585467050782, + "learning_rate": 0.003, + "loss": 4.058, + "step": 31551 + }, + { + "epoch": 0.31552, + "grad_norm": 0.8747539807100431, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 31552 + }, + { + "epoch": 0.31553, + "grad_norm": 0.8095260384227154, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 31553 + }, + { + "epoch": 0.31554, + "grad_norm": 0.8525732774308465, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 31554 + }, + { + "epoch": 0.31555, + "grad_norm": 1.1494719180132777, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 31555 + }, + { + "epoch": 0.31556, + "grad_norm": 1.3217263980146765, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 31556 + }, + { + "epoch": 0.31557, + "grad_norm": 1.115356905678181, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 31557 + }, + { + "epoch": 0.31558, + "grad_norm": 0.9763004512421191, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 31558 + }, + { + "epoch": 0.31559, + "grad_norm": 1.2787262456907003, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 31559 + }, + { + "epoch": 0.3156, + "grad_norm": 0.9754994588791245, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 31560 + }, + { + "epoch": 0.31561, + "grad_norm": 0.9884504551614549, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 31561 + }, + { + "epoch": 0.31562, + "grad_norm": 1.2071726384105894, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 31562 + }, + { + "epoch": 0.31563, + "grad_norm": 0.8936119840903438, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 31563 + }, + { + "epoch": 0.31564, + "grad_norm": 1.1067876629455544, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 31564 + }, + { + "epoch": 0.31565, + "grad_norm": 0.8433697850142939, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 31565 + }, + { + "epoch": 0.31566, + "grad_norm": 0.9097316181128641, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 31566 + }, + { + "epoch": 0.31567, + "grad_norm": 1.0355250954965296, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 31567 + }, + { + "epoch": 0.31568, + "grad_norm": 1.0601558244791427, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 31568 + }, + { + "epoch": 0.31569, + "grad_norm": 1.1421313649420097, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 31569 + }, + { + "epoch": 0.3157, + "grad_norm": 1.007309256947506, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 31570 + }, + { + "epoch": 0.31571, + "grad_norm": 1.122828648053304, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 31571 + }, + { + "epoch": 0.31572, + "grad_norm": 1.168023713540278, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 31572 + }, + { + "epoch": 0.31573, + "grad_norm": 1.0697598357207434, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 31573 + }, + { + "epoch": 0.31574, + "grad_norm": 0.8745170611050763, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 31574 + }, + { + "epoch": 0.31575, + "grad_norm": 0.9568543891382503, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 31575 + }, + { + "epoch": 0.31576, + "grad_norm": 1.108532314380304, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 31576 + }, + { + "epoch": 0.31577, + "grad_norm": 1.1908691872259844, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 31577 + }, + { + "epoch": 0.31578, + "grad_norm": 1.201165423194132, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 31578 + }, + { + "epoch": 0.31579, + "grad_norm": 1.460191801897932, + "learning_rate": 0.003, + "loss": 4.0858, + "step": 31579 + }, + { + "epoch": 0.3158, + "grad_norm": 0.9211853753203035, + "learning_rate": 0.003, + "loss": 4.076, + "step": 31580 + }, + { + "epoch": 0.31581, + "grad_norm": 0.9766581693044369, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 31581 + }, + { + "epoch": 0.31582, + "grad_norm": 1.365451517619097, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 31582 + }, + { + "epoch": 0.31583, + "grad_norm": 1.0775790947056785, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 31583 + }, + { + "epoch": 0.31584, + "grad_norm": 1.0191119763721264, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 31584 + }, + { + "epoch": 0.31585, + "grad_norm": 1.3784624020014196, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 31585 + }, + { + "epoch": 0.31586, + "grad_norm": 0.9993283264938255, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 31586 + }, + { + "epoch": 0.31587, + "grad_norm": 1.1565134452952186, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 31587 + }, + { + "epoch": 0.31588, + "grad_norm": 0.8413755509712173, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 31588 + }, + { + "epoch": 0.31589, + "grad_norm": 1.0187227695817846, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 31589 + }, + { + "epoch": 0.3159, + "grad_norm": 1.1535535609742626, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 31590 + }, + { + "epoch": 0.31591, + "grad_norm": 1.0593552396613497, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 31591 + }, + { + "epoch": 0.31592, + "grad_norm": 1.1417432224814719, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 31592 + }, + { + "epoch": 0.31593, + "grad_norm": 1.1112578189297886, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 31593 + }, + { + "epoch": 0.31594, + "grad_norm": 1.162676432345353, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 31594 + }, + { + "epoch": 0.31595, + "grad_norm": 1.102449112401589, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 31595 + }, + { + "epoch": 0.31596, + "grad_norm": 1.0144264807217087, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 31596 + }, + { + "epoch": 0.31597, + "grad_norm": 0.9464145259843196, + "learning_rate": 0.003, + "loss": 4.053, + "step": 31597 + }, + { + "epoch": 0.31598, + "grad_norm": 1.031912508244493, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 31598 + }, + { + "epoch": 0.31599, + "grad_norm": 1.361281804536459, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 31599 + }, + { + "epoch": 0.316, + "grad_norm": 0.9063631582126948, + "learning_rate": 0.003, + "loss": 4.036, + "step": 31600 + }, + { + "epoch": 0.31601, + "grad_norm": 0.8898343646251197, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 31601 + }, + { + "epoch": 0.31602, + "grad_norm": 1.098936310365506, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 31602 + }, + { + "epoch": 0.31603, + "grad_norm": 1.1896871751402547, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 31603 + }, + { + "epoch": 0.31604, + "grad_norm": 0.955515573640312, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 31604 + }, + { + "epoch": 0.31605, + "grad_norm": 1.0126465065037056, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 31605 + }, + { + "epoch": 0.31606, + "grad_norm": 1.1794000019355895, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 31606 + }, + { + "epoch": 0.31607, + "grad_norm": 0.9735007091009015, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 31607 + }, + { + "epoch": 0.31608, + "grad_norm": 0.9744976549410456, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 31608 + }, + { + "epoch": 0.31609, + "grad_norm": 1.0107963495361187, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 31609 + }, + { + "epoch": 0.3161, + "grad_norm": 1.0437024204430136, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 31610 + }, + { + "epoch": 0.31611, + "grad_norm": 1.129252853108004, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 31611 + }, + { + "epoch": 0.31612, + "grad_norm": 0.9729756097422139, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 31612 + }, + { + "epoch": 0.31613, + "grad_norm": 1.2180418771245847, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 31613 + }, + { + "epoch": 0.31614, + "grad_norm": 1.440979112232429, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 31614 + }, + { + "epoch": 0.31615, + "grad_norm": 1.1389937177804508, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 31615 + }, + { + "epoch": 0.31616, + "grad_norm": 0.8395555137648465, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 31616 + }, + { + "epoch": 0.31617, + "grad_norm": 0.8546595759690773, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 31617 + }, + { + "epoch": 0.31618, + "grad_norm": 1.092236706902754, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 31618 + }, + { + "epoch": 0.31619, + "grad_norm": 1.182274974253823, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 31619 + }, + { + "epoch": 0.3162, + "grad_norm": 0.8564004817852451, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 31620 + }, + { + "epoch": 0.31621, + "grad_norm": 1.0511006747292964, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 31621 + }, + { + "epoch": 0.31622, + "grad_norm": 1.5312342935216106, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 31622 + }, + { + "epoch": 0.31623, + "grad_norm": 1.0480240530490912, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 31623 + }, + { + "epoch": 0.31624, + "grad_norm": 1.0340299127659072, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 31624 + }, + { + "epoch": 0.31625, + "grad_norm": 1.0355756413699193, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 31625 + }, + { + "epoch": 0.31626, + "grad_norm": 1.0806065666598152, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 31626 + }, + { + "epoch": 0.31627, + "grad_norm": 1.1294545622614751, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 31627 + }, + { + "epoch": 0.31628, + "grad_norm": 1.1337645817935353, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 31628 + }, + { + "epoch": 0.31629, + "grad_norm": 1.0187046439366552, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 31629 + }, + { + "epoch": 0.3163, + "grad_norm": 1.2607482749171286, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 31630 + }, + { + "epoch": 0.31631, + "grad_norm": 1.1085032671583572, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 31631 + }, + { + "epoch": 0.31632, + "grad_norm": 0.9284513419611955, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 31632 + }, + { + "epoch": 0.31633, + "grad_norm": 1.0228066049779567, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 31633 + }, + { + "epoch": 0.31634, + "grad_norm": 1.374305179285254, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 31634 + }, + { + "epoch": 0.31635, + "grad_norm": 0.8099124323533411, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 31635 + }, + { + "epoch": 0.31636, + "grad_norm": 0.7828658511062724, + "learning_rate": 0.003, + "loss": 4.044, + "step": 31636 + }, + { + "epoch": 0.31637, + "grad_norm": 0.8715806579296085, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 31637 + }, + { + "epoch": 0.31638, + "grad_norm": 1.053518431362031, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 31638 + }, + { + "epoch": 0.31639, + "grad_norm": 1.0355994257281436, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 31639 + }, + { + "epoch": 0.3164, + "grad_norm": 1.1737436923785112, + "learning_rate": 0.003, + "loss": 4.017, + "step": 31640 + }, + { + "epoch": 0.31641, + "grad_norm": 1.0976469001869138, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 31641 + }, + { + "epoch": 0.31642, + "grad_norm": 1.015627924097845, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 31642 + }, + { + "epoch": 0.31643, + "grad_norm": 1.098008114515152, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 31643 + }, + { + "epoch": 0.31644, + "grad_norm": 0.9722136900453616, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 31644 + }, + { + "epoch": 0.31645, + "grad_norm": 1.0752314811523396, + "learning_rate": 0.003, + "loss": 4.053, + "step": 31645 + }, + { + "epoch": 0.31646, + "grad_norm": 1.1837253600152289, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 31646 + }, + { + "epoch": 0.31647, + "grad_norm": 0.9831383541405563, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 31647 + }, + { + "epoch": 0.31648, + "grad_norm": 1.1859677698802424, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 31648 + }, + { + "epoch": 0.31649, + "grad_norm": 1.0888531620294746, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 31649 + }, + { + "epoch": 0.3165, + "grad_norm": 0.9621586371534138, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 31650 + }, + { + "epoch": 0.31651, + "grad_norm": 1.0572746326510036, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 31651 + }, + { + "epoch": 0.31652, + "grad_norm": 1.3428342741296835, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 31652 + }, + { + "epoch": 0.31653, + "grad_norm": 0.9683957295502267, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 31653 + }, + { + "epoch": 0.31654, + "grad_norm": 0.9314533925381174, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 31654 + }, + { + "epoch": 0.31655, + "grad_norm": 1.0157852115581731, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 31655 + }, + { + "epoch": 0.31656, + "grad_norm": 1.2047500101291322, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 31656 + }, + { + "epoch": 0.31657, + "grad_norm": 0.9943208847242351, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 31657 + }, + { + "epoch": 0.31658, + "grad_norm": 0.9574457424914088, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 31658 + }, + { + "epoch": 0.31659, + "grad_norm": 1.3474004655147025, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 31659 + }, + { + "epoch": 0.3166, + "grad_norm": 1.1593108201716054, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 31660 + }, + { + "epoch": 0.31661, + "grad_norm": 1.1823186083379498, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 31661 + }, + { + "epoch": 0.31662, + "grad_norm": 1.1736012926323043, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 31662 + }, + { + "epoch": 0.31663, + "grad_norm": 0.9763356694232406, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 31663 + }, + { + "epoch": 0.31664, + "grad_norm": 1.044369263085413, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 31664 + }, + { + "epoch": 0.31665, + "grad_norm": 1.0217448323039433, + "learning_rate": 0.003, + "loss": 4.039, + "step": 31665 + }, + { + "epoch": 0.31666, + "grad_norm": 0.9711550128966944, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 31666 + }, + { + "epoch": 0.31667, + "grad_norm": 1.1262095187707806, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 31667 + }, + { + "epoch": 0.31668, + "grad_norm": 1.2016601694341915, + "learning_rate": 0.003, + "loss": 4.045, + "step": 31668 + }, + { + "epoch": 0.31669, + "grad_norm": 1.2998229389072042, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 31669 + }, + { + "epoch": 0.3167, + "grad_norm": 1.0400457017571636, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 31670 + }, + { + "epoch": 0.31671, + "grad_norm": 1.0004748295592631, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 31671 + }, + { + "epoch": 0.31672, + "grad_norm": 1.0932848431249012, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 31672 + }, + { + "epoch": 0.31673, + "grad_norm": 0.9595587554892328, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 31673 + }, + { + "epoch": 0.31674, + "grad_norm": 0.9317269093611328, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 31674 + }, + { + "epoch": 0.31675, + "grad_norm": 1.0988338360939731, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 31675 + }, + { + "epoch": 0.31676, + "grad_norm": 1.2228756290951746, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 31676 + }, + { + "epoch": 0.31677, + "grad_norm": 0.917244302621194, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 31677 + }, + { + "epoch": 0.31678, + "grad_norm": 0.9639315210880793, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 31678 + }, + { + "epoch": 0.31679, + "grad_norm": 1.0777426946733772, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 31679 + }, + { + "epoch": 0.3168, + "grad_norm": 1.3216994761709844, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 31680 + }, + { + "epoch": 0.31681, + "grad_norm": 0.8152503460617218, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 31681 + }, + { + "epoch": 0.31682, + "grad_norm": 0.8183482028677493, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 31682 + }, + { + "epoch": 0.31683, + "grad_norm": 0.9730221399823724, + "learning_rate": 0.003, + "loss": 4.033, + "step": 31683 + }, + { + "epoch": 0.31684, + "grad_norm": 1.200428092491599, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 31684 + }, + { + "epoch": 0.31685, + "grad_norm": 0.9843980222345076, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 31685 + }, + { + "epoch": 0.31686, + "grad_norm": 1.2752133772701106, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 31686 + }, + { + "epoch": 0.31687, + "grad_norm": 1.0086480021132358, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 31687 + }, + { + "epoch": 0.31688, + "grad_norm": 1.1826224447264333, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 31688 + }, + { + "epoch": 0.31689, + "grad_norm": 0.8563879490423156, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 31689 + }, + { + "epoch": 0.3169, + "grad_norm": 0.9201937565790138, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 31690 + }, + { + "epoch": 0.31691, + "grad_norm": 1.1979815593663543, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 31691 + }, + { + "epoch": 0.31692, + "grad_norm": 1.0238335074520848, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 31692 + }, + { + "epoch": 0.31693, + "grad_norm": 0.9815032200891963, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 31693 + }, + { + "epoch": 0.31694, + "grad_norm": 1.2687887082694602, + "learning_rate": 0.003, + "loss": 4.032, + "step": 31694 + }, + { + "epoch": 0.31695, + "grad_norm": 0.9482782432428378, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 31695 + }, + { + "epoch": 0.31696, + "grad_norm": 1.0457990170084923, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 31696 + }, + { + "epoch": 0.31697, + "grad_norm": 1.0408588548464284, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 31697 + }, + { + "epoch": 0.31698, + "grad_norm": 1.1226482914455078, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 31698 + }, + { + "epoch": 0.31699, + "grad_norm": 1.053063139976024, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 31699 + }, + { + "epoch": 0.317, + "grad_norm": 1.053664156260725, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 31700 + }, + { + "epoch": 0.31701, + "grad_norm": 1.1747229073720837, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 31701 + }, + { + "epoch": 0.31702, + "grad_norm": 0.9397949480224831, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 31702 + }, + { + "epoch": 0.31703, + "grad_norm": 1.2015242458438686, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 31703 + }, + { + "epoch": 0.31704, + "grad_norm": 1.0285616921925438, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 31704 + }, + { + "epoch": 0.31705, + "grad_norm": 1.1215095158917119, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 31705 + }, + { + "epoch": 0.31706, + "grad_norm": 0.9739061003334959, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 31706 + }, + { + "epoch": 0.31707, + "grad_norm": 1.2098115734670192, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 31707 + }, + { + "epoch": 0.31708, + "grad_norm": 0.9161306205718084, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 31708 + }, + { + "epoch": 0.31709, + "grad_norm": 0.9917887285709293, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 31709 + }, + { + "epoch": 0.3171, + "grad_norm": 1.0707588760116242, + "learning_rate": 0.003, + "loss": 4.061, + "step": 31710 + }, + { + "epoch": 0.31711, + "grad_norm": 1.2367716417380414, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 31711 + }, + { + "epoch": 0.31712, + "grad_norm": 1.2508089490160923, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 31712 + }, + { + "epoch": 0.31713, + "grad_norm": 1.2664228630083934, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 31713 + }, + { + "epoch": 0.31714, + "grad_norm": 1.0860742521962634, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 31714 + }, + { + "epoch": 0.31715, + "grad_norm": 1.196219661420946, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 31715 + }, + { + "epoch": 0.31716, + "grad_norm": 1.046883335309301, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 31716 + }, + { + "epoch": 0.31717, + "grad_norm": 1.262801852295055, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 31717 + }, + { + "epoch": 0.31718, + "grad_norm": 0.8769828195515889, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 31718 + }, + { + "epoch": 0.31719, + "grad_norm": 0.8603776828380939, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 31719 + }, + { + "epoch": 0.3172, + "grad_norm": 0.9503886272033572, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 31720 + }, + { + "epoch": 0.31721, + "grad_norm": 0.9167923610781114, + "learning_rate": 0.003, + "loss": 4.053, + "step": 31721 + }, + { + "epoch": 0.31722, + "grad_norm": 1.1175638736074827, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 31722 + }, + { + "epoch": 0.31723, + "grad_norm": 1.0855527885014122, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 31723 + }, + { + "epoch": 0.31724, + "grad_norm": 1.181504543168036, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 31724 + }, + { + "epoch": 0.31725, + "grad_norm": 1.1707076930048408, + "learning_rate": 0.003, + "loss": 4.055, + "step": 31725 + }, + { + "epoch": 0.31726, + "grad_norm": 1.0730227195992275, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 31726 + }, + { + "epoch": 0.31727, + "grad_norm": 1.0730612009265217, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 31727 + }, + { + "epoch": 0.31728, + "grad_norm": 0.9998576081127651, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 31728 + }, + { + "epoch": 0.31729, + "grad_norm": 1.1578092620579425, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 31729 + }, + { + "epoch": 0.3173, + "grad_norm": 0.8442321976464077, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 31730 + }, + { + "epoch": 0.31731, + "grad_norm": 0.9891037688445881, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 31731 + }, + { + "epoch": 0.31732, + "grad_norm": 1.1867288295729352, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 31732 + }, + { + "epoch": 0.31733, + "grad_norm": 1.1441270678376254, + "learning_rate": 0.003, + "loss": 4.058, + "step": 31733 + }, + { + "epoch": 0.31734, + "grad_norm": 1.0785791552963293, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 31734 + }, + { + "epoch": 0.31735, + "grad_norm": 1.2875206028673543, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 31735 + }, + { + "epoch": 0.31736, + "grad_norm": 0.9045543489422752, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 31736 + }, + { + "epoch": 0.31737, + "grad_norm": 0.9355165346631925, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 31737 + }, + { + "epoch": 0.31738, + "grad_norm": 1.109990268634666, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 31738 + }, + { + "epoch": 0.31739, + "grad_norm": 1.3761832939736418, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 31739 + }, + { + "epoch": 0.3174, + "grad_norm": 1.0176065609984861, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 31740 + }, + { + "epoch": 0.31741, + "grad_norm": 1.230501915618836, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 31741 + }, + { + "epoch": 0.31742, + "grad_norm": 1.009398166186859, + "learning_rate": 0.003, + "loss": 4.049, + "step": 31742 + }, + { + "epoch": 0.31743, + "grad_norm": 1.0235997358297235, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 31743 + }, + { + "epoch": 0.31744, + "grad_norm": 0.991735591494973, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 31744 + }, + { + "epoch": 0.31745, + "grad_norm": 1.159804646179269, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 31745 + }, + { + "epoch": 0.31746, + "grad_norm": 0.9941316654366303, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 31746 + }, + { + "epoch": 0.31747, + "grad_norm": 1.1167356512122297, + "learning_rate": 0.003, + "loss": 3.997, + "step": 31747 + }, + { + "epoch": 0.31748, + "grad_norm": 1.3799054242417086, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 31748 + }, + { + "epoch": 0.31749, + "grad_norm": 0.9080896945959637, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 31749 + }, + { + "epoch": 0.3175, + "grad_norm": 0.7279228870444838, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 31750 + }, + { + "epoch": 0.31751, + "grad_norm": 0.7662149046018298, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 31751 + }, + { + "epoch": 0.31752, + "grad_norm": 0.9035786878771825, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 31752 + }, + { + "epoch": 0.31753, + "grad_norm": 1.2081705751247196, + "learning_rate": 0.003, + "loss": 4.058, + "step": 31753 + }, + { + "epoch": 0.31754, + "grad_norm": 0.9122481209578219, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 31754 + }, + { + "epoch": 0.31755, + "grad_norm": 1.0199456126469253, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 31755 + }, + { + "epoch": 0.31756, + "grad_norm": 1.3785578999394692, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 31756 + }, + { + "epoch": 0.31757, + "grad_norm": 0.8048214833073069, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 31757 + }, + { + "epoch": 0.31758, + "grad_norm": 0.8561900104394458, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 31758 + }, + { + "epoch": 0.31759, + "grad_norm": 1.006890570883902, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 31759 + }, + { + "epoch": 0.3176, + "grad_norm": 1.386415764235917, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 31760 + }, + { + "epoch": 0.31761, + "grad_norm": 0.9446105800185973, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 31761 + }, + { + "epoch": 0.31762, + "grad_norm": 1.0521874341751267, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 31762 + }, + { + "epoch": 0.31763, + "grad_norm": 1.2150676112833507, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 31763 + }, + { + "epoch": 0.31764, + "grad_norm": 1.0573906298545648, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 31764 + }, + { + "epoch": 0.31765, + "grad_norm": 1.3313565066521509, + "learning_rate": 0.003, + "loss": 4.075, + "step": 31765 + }, + { + "epoch": 0.31766, + "grad_norm": 1.0477539290827405, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 31766 + }, + { + "epoch": 0.31767, + "grad_norm": 1.184715082523403, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 31767 + }, + { + "epoch": 0.31768, + "grad_norm": 1.253465464770264, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 31768 + }, + { + "epoch": 0.31769, + "grad_norm": 1.1579214634177242, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 31769 + }, + { + "epoch": 0.3177, + "grad_norm": 1.205648940203808, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 31770 + }, + { + "epoch": 0.31771, + "grad_norm": 1.2057572970978239, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 31771 + }, + { + "epoch": 0.31772, + "grad_norm": 1.1163921100657257, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 31772 + }, + { + "epoch": 0.31773, + "grad_norm": 0.8867070653342801, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 31773 + }, + { + "epoch": 0.31774, + "grad_norm": 0.8224580428184314, + "learning_rate": 0.003, + "loss": 4.065, + "step": 31774 + }, + { + "epoch": 0.31775, + "grad_norm": 0.9530705426848101, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 31775 + }, + { + "epoch": 0.31776, + "grad_norm": 1.225395457335267, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 31776 + }, + { + "epoch": 0.31777, + "grad_norm": 1.084807891594216, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 31777 + }, + { + "epoch": 0.31778, + "grad_norm": 1.1270182695530584, + "learning_rate": 0.003, + "loss": 4.043, + "step": 31778 + }, + { + "epoch": 0.31779, + "grad_norm": 1.0839526266937591, + "learning_rate": 0.003, + "loss": 4.055, + "step": 31779 + }, + { + "epoch": 0.3178, + "grad_norm": 1.124435378656918, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 31780 + }, + { + "epoch": 0.31781, + "grad_norm": 0.8973593122383119, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 31781 + }, + { + "epoch": 0.31782, + "grad_norm": 0.824180030904859, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 31782 + }, + { + "epoch": 0.31783, + "grad_norm": 1.044232296077719, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 31783 + }, + { + "epoch": 0.31784, + "grad_norm": 1.3906579890363697, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 31784 + }, + { + "epoch": 0.31785, + "grad_norm": 0.7072869196453162, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 31785 + }, + { + "epoch": 0.31786, + "grad_norm": 0.7690614205209522, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 31786 + }, + { + "epoch": 0.31787, + "grad_norm": 0.7992857464201052, + "learning_rate": 0.003, + "loss": 4.029, + "step": 31787 + }, + { + "epoch": 0.31788, + "grad_norm": 0.9631782852101534, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 31788 + }, + { + "epoch": 0.31789, + "grad_norm": 1.1761520377097598, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 31789 + }, + { + "epoch": 0.3179, + "grad_norm": 1.0537747495741963, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 31790 + }, + { + "epoch": 0.31791, + "grad_norm": 1.1795438569802459, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 31791 + }, + { + "epoch": 0.31792, + "grad_norm": 1.1569193836610059, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 31792 + }, + { + "epoch": 0.31793, + "grad_norm": 1.39785246570019, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 31793 + }, + { + "epoch": 0.31794, + "grad_norm": 1.0052805510750495, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 31794 + }, + { + "epoch": 0.31795, + "grad_norm": 1.0931616091080412, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 31795 + }, + { + "epoch": 0.31796, + "grad_norm": 1.1743259827077384, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 31796 + }, + { + "epoch": 0.31797, + "grad_norm": 1.051254674269716, + "learning_rate": 0.003, + "loss": 4.038, + "step": 31797 + }, + { + "epoch": 0.31798, + "grad_norm": 1.151012923600408, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 31798 + }, + { + "epoch": 0.31799, + "grad_norm": 0.9649144111902124, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 31799 + }, + { + "epoch": 0.318, + "grad_norm": 0.9980172727060264, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 31800 + }, + { + "epoch": 0.31801, + "grad_norm": 0.9675513641602704, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 31801 + }, + { + "epoch": 0.31802, + "grad_norm": 0.8997938482705993, + "learning_rate": 0.003, + "loss": 4.049, + "step": 31802 + }, + { + "epoch": 0.31803, + "grad_norm": 1.0731677948844687, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 31803 + }, + { + "epoch": 0.31804, + "grad_norm": 1.1507708137322465, + "learning_rate": 0.003, + "loss": 4.043, + "step": 31804 + }, + { + "epoch": 0.31805, + "grad_norm": 0.9092014636710631, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 31805 + }, + { + "epoch": 0.31806, + "grad_norm": 1.0714767620679857, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 31806 + }, + { + "epoch": 0.31807, + "grad_norm": 1.3454749309274558, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 31807 + }, + { + "epoch": 0.31808, + "grad_norm": 1.0505883506138882, + "learning_rate": 0.003, + "loss": 4.042, + "step": 31808 + }, + { + "epoch": 0.31809, + "grad_norm": 1.240403320554646, + "learning_rate": 0.003, + "loss": 4.031, + "step": 31809 + }, + { + "epoch": 0.3181, + "grad_norm": 0.9809221337616985, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 31810 + }, + { + "epoch": 0.31811, + "grad_norm": 1.114583543168373, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 31811 + }, + { + "epoch": 0.31812, + "grad_norm": 0.9172436101601256, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 31812 + }, + { + "epoch": 0.31813, + "grad_norm": 0.9197165730142427, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 31813 + }, + { + "epoch": 0.31814, + "grad_norm": 1.3656703905287126, + "learning_rate": 0.003, + "loss": 4.036, + "step": 31814 + }, + { + "epoch": 0.31815, + "grad_norm": 0.8235263274889454, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 31815 + }, + { + "epoch": 0.31816, + "grad_norm": 0.9271557016628661, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 31816 + }, + { + "epoch": 0.31817, + "grad_norm": 1.1636123215425684, + "learning_rate": 0.003, + "loss": 4.055, + "step": 31817 + }, + { + "epoch": 0.31818, + "grad_norm": 1.0320098099994965, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 31818 + }, + { + "epoch": 0.31819, + "grad_norm": 1.2911728463240788, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 31819 + }, + { + "epoch": 0.3182, + "grad_norm": 0.8089084684014567, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 31820 + }, + { + "epoch": 0.31821, + "grad_norm": 0.9315115187788786, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 31821 + }, + { + "epoch": 0.31822, + "grad_norm": 1.0701679713111305, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 31822 + }, + { + "epoch": 0.31823, + "grad_norm": 1.034964971934123, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 31823 + }, + { + "epoch": 0.31824, + "grad_norm": 1.0253517709132303, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 31824 + }, + { + "epoch": 0.31825, + "grad_norm": 1.2594901014643913, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 31825 + }, + { + "epoch": 0.31826, + "grad_norm": 0.9655463869829406, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 31826 + }, + { + "epoch": 0.31827, + "grad_norm": 1.1706741666401204, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 31827 + }, + { + "epoch": 0.31828, + "grad_norm": 1.1741231841227149, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 31828 + }, + { + "epoch": 0.31829, + "grad_norm": 1.2081481042943387, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 31829 + }, + { + "epoch": 0.3183, + "grad_norm": 0.9833199801729804, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 31830 + }, + { + "epoch": 0.31831, + "grad_norm": 1.347828537507272, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 31831 + }, + { + "epoch": 0.31832, + "grad_norm": 0.8463994751768169, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 31832 + }, + { + "epoch": 0.31833, + "grad_norm": 0.7717090400580913, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 31833 + }, + { + "epoch": 0.31834, + "grad_norm": 0.844785566429769, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 31834 + }, + { + "epoch": 0.31835, + "grad_norm": 0.9040775218345254, + "learning_rate": 0.003, + "loss": 4.01, + "step": 31835 + }, + { + "epoch": 0.31836, + "grad_norm": 1.038365960715167, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 31836 + }, + { + "epoch": 0.31837, + "grad_norm": 1.1901353855681414, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 31837 + }, + { + "epoch": 0.31838, + "grad_norm": 0.9095827912456269, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 31838 + }, + { + "epoch": 0.31839, + "grad_norm": 1.0827824279767826, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 31839 + }, + { + "epoch": 0.3184, + "grad_norm": 1.3661853769663066, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 31840 + }, + { + "epoch": 0.31841, + "grad_norm": 0.9667586742116547, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 31841 + }, + { + "epoch": 0.31842, + "grad_norm": 1.2314107020627256, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 31842 + }, + { + "epoch": 0.31843, + "grad_norm": 1.2948351668537141, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 31843 + }, + { + "epoch": 0.31844, + "grad_norm": 0.8153998057925554, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 31844 + }, + { + "epoch": 0.31845, + "grad_norm": 0.7710213836436985, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 31845 + }, + { + "epoch": 0.31846, + "grad_norm": 0.8864660985010172, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 31846 + }, + { + "epoch": 0.31847, + "grad_norm": 1.1436869157188705, + "learning_rate": 0.003, + "loss": 4.029, + "step": 31847 + }, + { + "epoch": 0.31848, + "grad_norm": 1.1747558717074154, + "learning_rate": 0.003, + "loss": 4.072, + "step": 31848 + }, + { + "epoch": 0.31849, + "grad_norm": 1.1032977625607279, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 31849 + }, + { + "epoch": 0.3185, + "grad_norm": 1.0364958871959018, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 31850 + }, + { + "epoch": 0.31851, + "grad_norm": 1.3929831730384195, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 31851 + }, + { + "epoch": 0.31852, + "grad_norm": 0.9426499576866708, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 31852 + }, + { + "epoch": 0.31853, + "grad_norm": 1.0584019840088406, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 31853 + }, + { + "epoch": 0.31854, + "grad_norm": 1.1628669734061776, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 31854 + }, + { + "epoch": 0.31855, + "grad_norm": 0.9910215876111725, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 31855 + }, + { + "epoch": 0.31856, + "grad_norm": 1.1737624688811925, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 31856 + }, + { + "epoch": 0.31857, + "grad_norm": 1.0599446776624348, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 31857 + }, + { + "epoch": 0.31858, + "grad_norm": 1.391726983037745, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 31858 + }, + { + "epoch": 0.31859, + "grad_norm": 0.8594343154465094, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 31859 + }, + { + "epoch": 0.3186, + "grad_norm": 0.9147406467035496, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 31860 + }, + { + "epoch": 0.31861, + "grad_norm": 1.4267147203086084, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 31861 + }, + { + "epoch": 0.31862, + "grad_norm": 1.24073419084833, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 31862 + }, + { + "epoch": 0.31863, + "grad_norm": 1.0332604576266704, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 31863 + }, + { + "epoch": 0.31864, + "grad_norm": 1.0786028045386766, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 31864 + }, + { + "epoch": 0.31865, + "grad_norm": 1.2019661618979032, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 31865 + }, + { + "epoch": 0.31866, + "grad_norm": 0.9247937482894314, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 31866 + }, + { + "epoch": 0.31867, + "grad_norm": 1.1038116569257794, + "learning_rate": 0.003, + "loss": 4.07, + "step": 31867 + }, + { + "epoch": 0.31868, + "grad_norm": 1.2047718331693456, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 31868 + }, + { + "epoch": 0.31869, + "grad_norm": 1.0646008138604233, + "learning_rate": 0.003, + "loss": 4.046, + "step": 31869 + }, + { + "epoch": 0.3187, + "grad_norm": 1.003424750768969, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 31870 + }, + { + "epoch": 0.31871, + "grad_norm": 1.1428417881175212, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 31871 + }, + { + "epoch": 0.31872, + "grad_norm": 0.9174921056134134, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 31872 + }, + { + "epoch": 0.31873, + "grad_norm": 1.0697554976667862, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 31873 + }, + { + "epoch": 0.31874, + "grad_norm": 1.1389263524898725, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 31874 + }, + { + "epoch": 0.31875, + "grad_norm": 1.04261219289545, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 31875 + }, + { + "epoch": 0.31876, + "grad_norm": 0.9193780492274085, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 31876 + }, + { + "epoch": 0.31877, + "grad_norm": 1.056336829489345, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 31877 + }, + { + "epoch": 0.31878, + "grad_norm": 1.121729270865299, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 31878 + }, + { + "epoch": 0.31879, + "grad_norm": 1.151308437602964, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 31879 + }, + { + "epoch": 0.3188, + "grad_norm": 1.1645078005861362, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 31880 + }, + { + "epoch": 0.31881, + "grad_norm": 1.0971098655066773, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 31881 + }, + { + "epoch": 0.31882, + "grad_norm": 1.2047492952169219, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 31882 + }, + { + "epoch": 0.31883, + "grad_norm": 0.9582545800529146, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 31883 + }, + { + "epoch": 0.31884, + "grad_norm": 1.4244209542355661, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 31884 + }, + { + "epoch": 0.31885, + "grad_norm": 1.3100762576868765, + "learning_rate": 0.003, + "loss": 4.042, + "step": 31885 + }, + { + "epoch": 0.31886, + "grad_norm": 0.76898769968431, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 31886 + }, + { + "epoch": 0.31887, + "grad_norm": 0.861172568469878, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 31887 + }, + { + "epoch": 0.31888, + "grad_norm": 1.0842114725504273, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 31888 + }, + { + "epoch": 0.31889, + "grad_norm": 0.9973179232014809, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 31889 + }, + { + "epoch": 0.3189, + "grad_norm": 1.1826370613302222, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 31890 + }, + { + "epoch": 0.31891, + "grad_norm": 1.1612255944784378, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 31891 + }, + { + "epoch": 0.31892, + "grad_norm": 0.864957365720704, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 31892 + }, + { + "epoch": 0.31893, + "grad_norm": 0.9895166783451119, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 31893 + }, + { + "epoch": 0.31894, + "grad_norm": 1.3420851421072226, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 31894 + }, + { + "epoch": 0.31895, + "grad_norm": 0.8623270241756994, + "learning_rate": 0.003, + "loss": 4.052, + "step": 31895 + }, + { + "epoch": 0.31896, + "grad_norm": 0.9196308567995006, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 31896 + }, + { + "epoch": 0.31897, + "grad_norm": 1.1058569356684327, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 31897 + }, + { + "epoch": 0.31898, + "grad_norm": 1.1643247206200156, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 31898 + }, + { + "epoch": 0.31899, + "grad_norm": 0.8965599743646535, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 31899 + }, + { + "epoch": 0.319, + "grad_norm": 1.2220835966745236, + "learning_rate": 0.003, + "loss": 4.007, + "step": 31900 + }, + { + "epoch": 0.31901, + "grad_norm": 1.200280066267465, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 31901 + }, + { + "epoch": 0.31902, + "grad_norm": 0.9751529440972444, + "learning_rate": 0.003, + "loss": 4.031, + "step": 31902 + }, + { + "epoch": 0.31903, + "grad_norm": 1.0545246710858287, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 31903 + }, + { + "epoch": 0.31904, + "grad_norm": 1.1242932449671525, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 31904 + }, + { + "epoch": 0.31905, + "grad_norm": 1.1099925076568147, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 31905 + }, + { + "epoch": 0.31906, + "grad_norm": 1.0833133094091216, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 31906 + }, + { + "epoch": 0.31907, + "grad_norm": 1.1423870403318905, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 31907 + }, + { + "epoch": 0.31908, + "grad_norm": 0.9879843280349828, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 31908 + }, + { + "epoch": 0.31909, + "grad_norm": 1.0723888867145062, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 31909 + }, + { + "epoch": 0.3191, + "grad_norm": 1.0766690084373733, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 31910 + }, + { + "epoch": 0.31911, + "grad_norm": 1.1815746768788251, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 31911 + }, + { + "epoch": 0.31912, + "grad_norm": 1.025347864415083, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 31912 + }, + { + "epoch": 0.31913, + "grad_norm": 1.106270698928186, + "learning_rate": 0.003, + "loss": 4.057, + "step": 31913 + }, + { + "epoch": 0.31914, + "grad_norm": 0.9836643955197196, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 31914 + }, + { + "epoch": 0.31915, + "grad_norm": 1.0826367192297648, + "learning_rate": 0.003, + "loss": 4.029, + "step": 31915 + }, + { + "epoch": 0.31916, + "grad_norm": 0.990240804041412, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 31916 + }, + { + "epoch": 0.31917, + "grad_norm": 0.9787280613861183, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 31917 + }, + { + "epoch": 0.31918, + "grad_norm": 1.1970118485268069, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 31918 + }, + { + "epoch": 0.31919, + "grad_norm": 1.0533353181552034, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 31919 + }, + { + "epoch": 0.3192, + "grad_norm": 1.2245738809971651, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 31920 + }, + { + "epoch": 0.31921, + "grad_norm": 1.2370448322276206, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 31921 + }, + { + "epoch": 0.31922, + "grad_norm": 1.1172194606659087, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 31922 + }, + { + "epoch": 0.31923, + "grad_norm": 1.201482419616384, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 31923 + }, + { + "epoch": 0.31924, + "grad_norm": 1.0817569415749337, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 31924 + }, + { + "epoch": 0.31925, + "grad_norm": 1.018536461719636, + "learning_rate": 0.003, + "loss": 4.049, + "step": 31925 + }, + { + "epoch": 0.31926, + "grad_norm": 1.0439018557483157, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 31926 + }, + { + "epoch": 0.31927, + "grad_norm": 1.2282746450428725, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 31927 + }, + { + "epoch": 0.31928, + "grad_norm": 1.0208531449711578, + "learning_rate": 0.003, + "loss": 4.036, + "step": 31928 + }, + { + "epoch": 0.31929, + "grad_norm": 1.3150001909037323, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 31929 + }, + { + "epoch": 0.3193, + "grad_norm": 1.0869201828136528, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 31930 + }, + { + "epoch": 0.31931, + "grad_norm": 1.0056767960816702, + "learning_rate": 0.003, + "loss": 4.048, + "step": 31931 + }, + { + "epoch": 0.31932, + "grad_norm": 1.0088620228374432, + "learning_rate": 0.003, + "loss": 4.069, + "step": 31932 + }, + { + "epoch": 0.31933, + "grad_norm": 1.098535197597744, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 31933 + }, + { + "epoch": 0.31934, + "grad_norm": 1.1691564112502226, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 31934 + }, + { + "epoch": 0.31935, + "grad_norm": 1.2175242672958235, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 31935 + }, + { + "epoch": 0.31936, + "grad_norm": 1.0766939291469417, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 31936 + }, + { + "epoch": 0.31937, + "grad_norm": 1.0710002760692614, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 31937 + }, + { + "epoch": 0.31938, + "grad_norm": 0.932437366594074, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 31938 + }, + { + "epoch": 0.31939, + "grad_norm": 0.9459612966876225, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 31939 + }, + { + "epoch": 0.3194, + "grad_norm": 1.109431780854956, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 31940 + }, + { + "epoch": 0.31941, + "grad_norm": 1.1120786430967955, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 31941 + }, + { + "epoch": 0.31942, + "grad_norm": 0.8514885322393715, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 31942 + }, + { + "epoch": 0.31943, + "grad_norm": 0.9791490843826585, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 31943 + }, + { + "epoch": 0.31944, + "grad_norm": 1.127365789874664, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 31944 + }, + { + "epoch": 0.31945, + "grad_norm": 1.1142817806775847, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 31945 + }, + { + "epoch": 0.31946, + "grad_norm": 1.1722102710098021, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 31946 + }, + { + "epoch": 0.31947, + "grad_norm": 0.9940368595121066, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 31947 + }, + { + "epoch": 0.31948, + "grad_norm": 1.2140751288411005, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 31948 + }, + { + "epoch": 0.31949, + "grad_norm": 1.2697017221648914, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 31949 + }, + { + "epoch": 0.3195, + "grad_norm": 0.8947042608290683, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 31950 + }, + { + "epoch": 0.31951, + "grad_norm": 0.8939539749933139, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 31951 + }, + { + "epoch": 0.31952, + "grad_norm": 1.0065675667367686, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 31952 + }, + { + "epoch": 0.31953, + "grad_norm": 1.1052947337809724, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 31953 + }, + { + "epoch": 0.31954, + "grad_norm": 1.2002093857106997, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 31954 + }, + { + "epoch": 0.31955, + "grad_norm": 0.968977221415062, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 31955 + }, + { + "epoch": 0.31956, + "grad_norm": 1.3305578338844701, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 31956 + }, + { + "epoch": 0.31957, + "grad_norm": 1.0656759387171777, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 31957 + }, + { + "epoch": 0.31958, + "grad_norm": 1.1579681016475698, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 31958 + }, + { + "epoch": 0.31959, + "grad_norm": 1.1556007684493876, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 31959 + }, + { + "epoch": 0.3196, + "grad_norm": 1.252723728556608, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 31960 + }, + { + "epoch": 0.31961, + "grad_norm": 0.8624175333503749, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 31961 + }, + { + "epoch": 0.31962, + "grad_norm": 1.0284770809161496, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 31962 + }, + { + "epoch": 0.31963, + "grad_norm": 1.2004378167584853, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 31963 + }, + { + "epoch": 0.31964, + "grad_norm": 1.2749034604360052, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 31964 + }, + { + "epoch": 0.31965, + "grad_norm": 1.1616313195808199, + "learning_rate": 0.003, + "loss": 4.0936, + "step": 31965 + }, + { + "epoch": 0.31966, + "grad_norm": 0.865392073854147, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 31966 + }, + { + "epoch": 0.31967, + "grad_norm": 0.9482044650409749, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 31967 + }, + { + "epoch": 0.31968, + "grad_norm": 1.0383239711100305, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 31968 + }, + { + "epoch": 0.31969, + "grad_norm": 1.0498143532727757, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 31969 + }, + { + "epoch": 0.3197, + "grad_norm": 1.0809690815571886, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 31970 + }, + { + "epoch": 0.31971, + "grad_norm": 1.244662207599401, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 31971 + }, + { + "epoch": 0.31972, + "grad_norm": 1.1060749623948476, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 31972 + }, + { + "epoch": 0.31973, + "grad_norm": 1.2693388659779634, + "learning_rate": 0.003, + "loss": 4.078, + "step": 31973 + }, + { + "epoch": 0.31974, + "grad_norm": 0.9295830412385528, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 31974 + }, + { + "epoch": 0.31975, + "grad_norm": 0.8835644882934729, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 31975 + }, + { + "epoch": 0.31976, + "grad_norm": 0.9648393491815364, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 31976 + }, + { + "epoch": 0.31977, + "grad_norm": 0.9918387370197791, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 31977 + }, + { + "epoch": 0.31978, + "grad_norm": 1.1269779982823505, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 31978 + }, + { + "epoch": 0.31979, + "grad_norm": 0.9918305896713778, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 31979 + }, + { + "epoch": 0.3198, + "grad_norm": 0.9249372323757936, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 31980 + }, + { + "epoch": 0.31981, + "grad_norm": 1.0306827823574232, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 31981 + }, + { + "epoch": 0.31982, + "grad_norm": 1.3237780333536606, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 31982 + }, + { + "epoch": 0.31983, + "grad_norm": 1.216066670193554, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 31983 + }, + { + "epoch": 0.31984, + "grad_norm": 1.2588686992339484, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 31984 + }, + { + "epoch": 0.31985, + "grad_norm": 1.14178632160689, + "learning_rate": 0.003, + "loss": 4.051, + "step": 31985 + }, + { + "epoch": 0.31986, + "grad_norm": 1.000153398161612, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 31986 + }, + { + "epoch": 0.31987, + "grad_norm": 1.048367303903448, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 31987 + }, + { + "epoch": 0.31988, + "grad_norm": 1.1407210969020283, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 31988 + }, + { + "epoch": 0.31989, + "grad_norm": 1.017026642045237, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 31989 + }, + { + "epoch": 0.3199, + "grad_norm": 1.2898601074491352, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 31990 + }, + { + "epoch": 0.31991, + "grad_norm": 1.004521823921086, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 31991 + }, + { + "epoch": 0.31992, + "grad_norm": 0.9554300751148529, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 31992 + }, + { + "epoch": 0.31993, + "grad_norm": 0.9792815771000235, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 31993 + }, + { + "epoch": 0.31994, + "grad_norm": 1.253497426521402, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 31994 + }, + { + "epoch": 0.31995, + "grad_norm": 1.0877583487850384, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 31995 + }, + { + "epoch": 0.31996, + "grad_norm": 1.0083689726323524, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 31996 + }, + { + "epoch": 0.31997, + "grad_norm": 1.0744184457381307, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 31997 + }, + { + "epoch": 0.31998, + "grad_norm": 0.9265565201987565, + "learning_rate": 0.003, + "loss": 4.058, + "step": 31998 + }, + { + "epoch": 0.31999, + "grad_norm": 1.0967356418030536, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 31999 + }, + { + "epoch": 0.32, + "grad_norm": 1.0910442245849326, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 32000 + }, + { + "epoch": 0.32001, + "grad_norm": 0.9724613349125483, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 32001 + }, + { + "epoch": 0.32002, + "grad_norm": 1.055191531962308, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 32002 + }, + { + "epoch": 0.32003, + "grad_norm": 1.15131674610959, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 32003 + }, + { + "epoch": 0.32004, + "grad_norm": 1.1310667326716424, + "learning_rate": 0.003, + "loss": 4.053, + "step": 32004 + }, + { + "epoch": 0.32005, + "grad_norm": 1.2336877392323944, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 32005 + }, + { + "epoch": 0.32006, + "grad_norm": 0.9175532945743243, + "learning_rate": 0.003, + "loss": 4.067, + "step": 32006 + }, + { + "epoch": 0.32007, + "grad_norm": 1.1137757581897698, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 32007 + }, + { + "epoch": 0.32008, + "grad_norm": 1.3464149530838128, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 32008 + }, + { + "epoch": 0.32009, + "grad_norm": 1.1792366931920981, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 32009 + }, + { + "epoch": 0.3201, + "grad_norm": 1.422281017635926, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 32010 + }, + { + "epoch": 0.32011, + "grad_norm": 0.7204069716478683, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 32011 + }, + { + "epoch": 0.32012, + "grad_norm": 0.7384986805087486, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 32012 + }, + { + "epoch": 0.32013, + "grad_norm": 1.0103171265914392, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 32013 + }, + { + "epoch": 0.32014, + "grad_norm": 1.219576760918464, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 32014 + }, + { + "epoch": 0.32015, + "grad_norm": 0.9744046498634229, + "learning_rate": 0.003, + "loss": 4.019, + "step": 32015 + }, + { + "epoch": 0.32016, + "grad_norm": 1.1607237714631733, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 32016 + }, + { + "epoch": 0.32017, + "grad_norm": 1.0131868711162888, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 32017 + }, + { + "epoch": 0.32018, + "grad_norm": 1.293190331533858, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 32018 + }, + { + "epoch": 0.32019, + "grad_norm": 1.162798524737129, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 32019 + }, + { + "epoch": 0.3202, + "grad_norm": 1.248418318734186, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 32020 + }, + { + "epoch": 0.32021, + "grad_norm": 1.1350361145653514, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 32021 + }, + { + "epoch": 0.32022, + "grad_norm": 1.033485414234219, + "learning_rate": 0.003, + "loss": 4.042, + "step": 32022 + }, + { + "epoch": 0.32023, + "grad_norm": 1.1542333680145365, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 32023 + }, + { + "epoch": 0.32024, + "grad_norm": 1.0932610211873797, + "learning_rate": 0.003, + "loss": 4.022, + "step": 32024 + }, + { + "epoch": 0.32025, + "grad_norm": 1.1778125087940863, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 32025 + }, + { + "epoch": 0.32026, + "grad_norm": 0.8465245332072461, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 32026 + }, + { + "epoch": 0.32027, + "grad_norm": 0.8680298977901865, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 32027 + }, + { + "epoch": 0.32028, + "grad_norm": 0.9248850304385782, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 32028 + }, + { + "epoch": 0.32029, + "grad_norm": 0.9283235881629013, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 32029 + }, + { + "epoch": 0.3203, + "grad_norm": 0.9426998739371529, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 32030 + }, + { + "epoch": 0.32031, + "grad_norm": 1.0412357171591236, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 32031 + }, + { + "epoch": 0.32032, + "grad_norm": 1.1599563050065151, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 32032 + }, + { + "epoch": 0.32033, + "grad_norm": 1.1062941295684428, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 32033 + }, + { + "epoch": 0.32034, + "grad_norm": 1.060778792791246, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 32034 + }, + { + "epoch": 0.32035, + "grad_norm": 1.063025182220492, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 32035 + }, + { + "epoch": 0.32036, + "grad_norm": 1.2336308572327377, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 32036 + }, + { + "epoch": 0.32037, + "grad_norm": 1.0916534903019688, + "learning_rate": 0.003, + "loss": 4.011, + "step": 32037 + }, + { + "epoch": 0.32038, + "grad_norm": 1.1522259828341894, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 32038 + }, + { + "epoch": 0.32039, + "grad_norm": 1.0492251289975962, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 32039 + }, + { + "epoch": 0.3204, + "grad_norm": 1.0620495577424005, + "learning_rate": 0.003, + "loss": 4.025, + "step": 32040 + }, + { + "epoch": 0.32041, + "grad_norm": 1.1437933487664205, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 32041 + }, + { + "epoch": 0.32042, + "grad_norm": 1.013512762443919, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 32042 + }, + { + "epoch": 0.32043, + "grad_norm": 1.3017994061181004, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 32043 + }, + { + "epoch": 0.32044, + "grad_norm": 0.9196690846184193, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 32044 + }, + { + "epoch": 0.32045, + "grad_norm": 1.0513562659595486, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 32045 + }, + { + "epoch": 0.32046, + "grad_norm": 1.120832044440907, + "learning_rate": 0.003, + "loss": 4.039, + "step": 32046 + }, + { + "epoch": 0.32047, + "grad_norm": 0.9254448425855561, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 32047 + }, + { + "epoch": 0.32048, + "grad_norm": 0.9003990330429196, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 32048 + }, + { + "epoch": 0.32049, + "grad_norm": 1.1302803587783952, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 32049 + }, + { + "epoch": 0.3205, + "grad_norm": 1.00071079159606, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 32050 + }, + { + "epoch": 0.32051, + "grad_norm": 1.1353576414894626, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 32051 + }, + { + "epoch": 0.32052, + "grad_norm": 1.1328450501490555, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 32052 + }, + { + "epoch": 0.32053, + "grad_norm": 1.1012032413306687, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 32053 + }, + { + "epoch": 0.32054, + "grad_norm": 1.373604161945927, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 32054 + }, + { + "epoch": 0.32055, + "grad_norm": 0.9880548267548206, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 32055 + }, + { + "epoch": 0.32056, + "grad_norm": 1.067203499814224, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 32056 + }, + { + "epoch": 0.32057, + "grad_norm": 1.1014397208045834, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 32057 + }, + { + "epoch": 0.32058, + "grad_norm": 1.1491738285906303, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 32058 + }, + { + "epoch": 0.32059, + "grad_norm": 1.2334330219899428, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 32059 + }, + { + "epoch": 0.3206, + "grad_norm": 0.9741809665367496, + "learning_rate": 0.003, + "loss": 4.042, + "step": 32060 + }, + { + "epoch": 0.32061, + "grad_norm": 1.2591845106390929, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 32061 + }, + { + "epoch": 0.32062, + "grad_norm": 1.0233260266275557, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 32062 + }, + { + "epoch": 0.32063, + "grad_norm": 0.9723879018554382, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 32063 + }, + { + "epoch": 0.32064, + "grad_norm": 0.9724121656662035, + "learning_rate": 0.003, + "loss": 4.058, + "step": 32064 + }, + { + "epoch": 0.32065, + "grad_norm": 1.1793017896224738, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 32065 + }, + { + "epoch": 0.32066, + "grad_norm": 0.8990858344455834, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 32066 + }, + { + "epoch": 0.32067, + "grad_norm": 0.8591178382913587, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 32067 + }, + { + "epoch": 0.32068, + "grad_norm": 0.9301164634050418, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 32068 + }, + { + "epoch": 0.32069, + "grad_norm": 1.1670956567118367, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 32069 + }, + { + "epoch": 0.3207, + "grad_norm": 1.1090761747454878, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 32070 + }, + { + "epoch": 0.32071, + "grad_norm": 1.2255460553927355, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 32071 + }, + { + "epoch": 0.32072, + "grad_norm": 1.0232847189233787, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 32072 + }, + { + "epoch": 0.32073, + "grad_norm": 1.3202424516186912, + "learning_rate": 0.003, + "loss": 4.059, + "step": 32073 + }, + { + "epoch": 0.32074, + "grad_norm": 0.9562285755014477, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 32074 + }, + { + "epoch": 0.32075, + "grad_norm": 1.0495216425213085, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 32075 + }, + { + "epoch": 0.32076, + "grad_norm": 1.2086272165430652, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 32076 + }, + { + "epoch": 0.32077, + "grad_norm": 1.0530988169794673, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 32077 + }, + { + "epoch": 0.32078, + "grad_norm": 1.1552057995339864, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 32078 + }, + { + "epoch": 0.32079, + "grad_norm": 0.920733474162925, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 32079 + }, + { + "epoch": 0.3208, + "grad_norm": 1.1511897064293506, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 32080 + }, + { + "epoch": 0.32081, + "grad_norm": 1.3073349310074636, + "learning_rate": 0.003, + "loss": 4.055, + "step": 32081 + }, + { + "epoch": 0.32082, + "grad_norm": 0.8399405011677125, + "learning_rate": 0.003, + "loss": 4.037, + "step": 32082 + }, + { + "epoch": 0.32083, + "grad_norm": 1.0661374049987973, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 32083 + }, + { + "epoch": 0.32084, + "grad_norm": 1.3257121681588493, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 32084 + }, + { + "epoch": 0.32085, + "grad_norm": 1.155009333428461, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 32085 + }, + { + "epoch": 0.32086, + "grad_norm": 0.9445378412303226, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 32086 + }, + { + "epoch": 0.32087, + "grad_norm": 1.205947202886209, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 32087 + }, + { + "epoch": 0.32088, + "grad_norm": 1.1694303169142524, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 32088 + }, + { + "epoch": 0.32089, + "grad_norm": 0.840813264170393, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 32089 + }, + { + "epoch": 0.3209, + "grad_norm": 0.8134243024168345, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 32090 + }, + { + "epoch": 0.32091, + "grad_norm": 1.000717763785386, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 32091 + }, + { + "epoch": 0.32092, + "grad_norm": 1.4205855501576297, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 32092 + }, + { + "epoch": 0.32093, + "grad_norm": 0.7893881203038027, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 32093 + }, + { + "epoch": 0.32094, + "grad_norm": 0.908679480392228, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 32094 + }, + { + "epoch": 0.32095, + "grad_norm": 1.1386293594225534, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 32095 + }, + { + "epoch": 0.32096, + "grad_norm": 1.2359803351470366, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 32096 + }, + { + "epoch": 0.32097, + "grad_norm": 0.9025312137358116, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 32097 + }, + { + "epoch": 0.32098, + "grad_norm": 1.000996567499761, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 32098 + }, + { + "epoch": 0.32099, + "grad_norm": 1.4334953709709708, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 32099 + }, + { + "epoch": 0.321, + "grad_norm": 0.82348283229666, + "learning_rate": 0.003, + "loss": 4.043, + "step": 32100 + }, + { + "epoch": 0.32101, + "grad_norm": 0.967122066692614, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 32101 + }, + { + "epoch": 0.32102, + "grad_norm": 1.1240797358395063, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 32102 + }, + { + "epoch": 0.32103, + "grad_norm": 0.9653047473639215, + "learning_rate": 0.003, + "loss": 4.06, + "step": 32103 + }, + { + "epoch": 0.32104, + "grad_norm": 1.094456899502812, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 32104 + }, + { + "epoch": 0.32105, + "grad_norm": 1.1510449828212226, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 32105 + }, + { + "epoch": 0.32106, + "grad_norm": 1.3665991040762213, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 32106 + }, + { + "epoch": 0.32107, + "grad_norm": 0.8691271987125805, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 32107 + }, + { + "epoch": 0.32108, + "grad_norm": 0.9797647261117024, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 32108 + }, + { + "epoch": 0.32109, + "grad_norm": 1.4579065709164694, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 32109 + }, + { + "epoch": 0.3211, + "grad_norm": 0.9142550978412053, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 32110 + }, + { + "epoch": 0.32111, + "grad_norm": 1.0121486800281756, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 32111 + }, + { + "epoch": 0.32112, + "grad_norm": 1.506942886508187, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 32112 + }, + { + "epoch": 0.32113, + "grad_norm": 0.9426781818174497, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 32113 + }, + { + "epoch": 0.32114, + "grad_norm": 0.9684423958967764, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 32114 + }, + { + "epoch": 0.32115, + "grad_norm": 0.9629302089384705, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 32115 + }, + { + "epoch": 0.32116, + "grad_norm": 1.0949131338260398, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 32116 + }, + { + "epoch": 0.32117, + "grad_norm": 1.0810915326490547, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 32117 + }, + { + "epoch": 0.32118, + "grad_norm": 1.1395160387176977, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 32118 + }, + { + "epoch": 0.32119, + "grad_norm": 1.1468098694890605, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 32119 + }, + { + "epoch": 0.3212, + "grad_norm": 0.9946305886065189, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 32120 + }, + { + "epoch": 0.32121, + "grad_norm": 1.2920790979459822, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 32121 + }, + { + "epoch": 0.32122, + "grad_norm": 0.8821509232192283, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 32122 + }, + { + "epoch": 0.32123, + "grad_norm": 0.9658428011578131, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 32123 + }, + { + "epoch": 0.32124, + "grad_norm": 1.2555411283390072, + "learning_rate": 0.003, + "loss": 4.064, + "step": 32124 + }, + { + "epoch": 0.32125, + "grad_norm": 1.1180698738732906, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 32125 + }, + { + "epoch": 0.32126, + "grad_norm": 0.9052416057197187, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 32126 + }, + { + "epoch": 0.32127, + "grad_norm": 0.9119880315181756, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 32127 + }, + { + "epoch": 0.32128, + "grad_norm": 1.118236402658847, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 32128 + }, + { + "epoch": 0.32129, + "grad_norm": 1.2579663846362767, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 32129 + }, + { + "epoch": 0.3213, + "grad_norm": 0.9008865224374114, + "learning_rate": 0.003, + "loss": 3.998, + "step": 32130 + }, + { + "epoch": 0.32131, + "grad_norm": 0.9619224513929069, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 32131 + }, + { + "epoch": 0.32132, + "grad_norm": 1.3827485334726541, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 32132 + }, + { + "epoch": 0.32133, + "grad_norm": 0.967132744824875, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 32133 + }, + { + "epoch": 0.32134, + "grad_norm": 1.0346993450980617, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 32134 + }, + { + "epoch": 0.32135, + "grad_norm": 1.2509291966803588, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 32135 + }, + { + "epoch": 0.32136, + "grad_norm": 1.0349764842567284, + "learning_rate": 0.003, + "loss": 4.039, + "step": 32136 + }, + { + "epoch": 0.32137, + "grad_norm": 1.283736438392351, + "learning_rate": 0.003, + "loss": 4.04, + "step": 32137 + }, + { + "epoch": 0.32138, + "grad_norm": 1.0176224049231106, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 32138 + }, + { + "epoch": 0.32139, + "grad_norm": 1.1952091874465287, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 32139 + }, + { + "epoch": 0.3214, + "grad_norm": 1.0551263559010164, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 32140 + }, + { + "epoch": 0.32141, + "grad_norm": 1.2734158425632207, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 32141 + }, + { + "epoch": 0.32142, + "grad_norm": 0.9970543004381334, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 32142 + }, + { + "epoch": 0.32143, + "grad_norm": 1.1497586895656693, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 32143 + }, + { + "epoch": 0.32144, + "grad_norm": 1.067075980908234, + "learning_rate": 0.003, + "loss": 4.066, + "step": 32144 + }, + { + "epoch": 0.32145, + "grad_norm": 1.0251090927928501, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 32145 + }, + { + "epoch": 0.32146, + "grad_norm": 0.970316634947586, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 32146 + }, + { + "epoch": 0.32147, + "grad_norm": 0.9494553942168185, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 32147 + }, + { + "epoch": 0.32148, + "grad_norm": 1.1236950765724596, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 32148 + }, + { + "epoch": 0.32149, + "grad_norm": 1.068337919841664, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 32149 + }, + { + "epoch": 0.3215, + "grad_norm": 1.1962117135498203, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 32150 + }, + { + "epoch": 0.32151, + "grad_norm": 1.1389922733796294, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 32151 + }, + { + "epoch": 0.32152, + "grad_norm": 0.8413826179603455, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 32152 + }, + { + "epoch": 0.32153, + "grad_norm": 0.9785500695764021, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 32153 + }, + { + "epoch": 0.32154, + "grad_norm": 1.2769315924225872, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 32154 + }, + { + "epoch": 0.32155, + "grad_norm": 1.1711725188996696, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 32155 + }, + { + "epoch": 0.32156, + "grad_norm": 1.1500125475784697, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 32156 + }, + { + "epoch": 0.32157, + "grad_norm": 1.1364361222873167, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 32157 + }, + { + "epoch": 0.32158, + "grad_norm": 1.0279446464409732, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 32158 + }, + { + "epoch": 0.32159, + "grad_norm": 0.9976728394417755, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 32159 + }, + { + "epoch": 0.3216, + "grad_norm": 1.1796848450634358, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 32160 + }, + { + "epoch": 0.32161, + "grad_norm": 1.2383227771214624, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 32161 + }, + { + "epoch": 0.32162, + "grad_norm": 1.0933131948282828, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 32162 + }, + { + "epoch": 0.32163, + "grad_norm": 0.9944376522468465, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 32163 + }, + { + "epoch": 0.32164, + "grad_norm": 1.0235629872599834, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 32164 + }, + { + "epoch": 0.32165, + "grad_norm": 1.0671978757746359, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 32165 + }, + { + "epoch": 0.32166, + "grad_norm": 1.043959209658961, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 32166 + }, + { + "epoch": 0.32167, + "grad_norm": 1.107276536379961, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 32167 + }, + { + "epoch": 0.32168, + "grad_norm": 0.9897995429822203, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 32168 + }, + { + "epoch": 0.32169, + "grad_norm": 1.175217098469182, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 32169 + }, + { + "epoch": 0.3217, + "grad_norm": 1.0150244413471956, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 32170 + }, + { + "epoch": 0.32171, + "grad_norm": 1.0713190751245365, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 32171 + }, + { + "epoch": 0.32172, + "grad_norm": 0.9706908789997643, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 32172 + }, + { + "epoch": 0.32173, + "grad_norm": 1.0310298279084624, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 32173 + }, + { + "epoch": 0.32174, + "grad_norm": 1.010430943018412, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 32174 + }, + { + "epoch": 0.32175, + "grad_norm": 1.1361190350264134, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 32175 + }, + { + "epoch": 0.32176, + "grad_norm": 1.0088141944893447, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 32176 + }, + { + "epoch": 0.32177, + "grad_norm": 1.0763108531283847, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 32177 + }, + { + "epoch": 0.32178, + "grad_norm": 1.0676763879110376, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 32178 + }, + { + "epoch": 0.32179, + "grad_norm": 1.2669970922210922, + "learning_rate": 0.003, + "loss": 4.06, + "step": 32179 + }, + { + "epoch": 0.3218, + "grad_norm": 1.0227416119547434, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 32180 + }, + { + "epoch": 0.32181, + "grad_norm": 1.177303786187009, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 32181 + }, + { + "epoch": 0.32182, + "grad_norm": 1.0510423706742946, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 32182 + }, + { + "epoch": 0.32183, + "grad_norm": 1.1617527497986269, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 32183 + }, + { + "epoch": 0.32184, + "grad_norm": 1.2192378991031674, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 32184 + }, + { + "epoch": 0.32185, + "grad_norm": 1.048902804409032, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 32185 + }, + { + "epoch": 0.32186, + "grad_norm": 1.0916675041586414, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 32186 + }, + { + "epoch": 0.32187, + "grad_norm": 0.9783600624412457, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 32187 + }, + { + "epoch": 0.32188, + "grad_norm": 1.1510590213086955, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 32188 + }, + { + "epoch": 0.32189, + "grad_norm": 1.199726913024886, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 32189 + }, + { + "epoch": 0.3219, + "grad_norm": 1.0217638592369147, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 32190 + }, + { + "epoch": 0.32191, + "grad_norm": 1.1605132669388885, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 32191 + }, + { + "epoch": 0.32192, + "grad_norm": 1.2690851211553913, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 32192 + }, + { + "epoch": 0.32193, + "grad_norm": 1.0803398819681342, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 32193 + }, + { + "epoch": 0.32194, + "grad_norm": 0.8666602258282032, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 32194 + }, + { + "epoch": 0.32195, + "grad_norm": 0.9418132759790964, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 32195 + }, + { + "epoch": 0.32196, + "grad_norm": 1.2938590716766494, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 32196 + }, + { + "epoch": 0.32197, + "grad_norm": 0.854584893228695, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 32197 + }, + { + "epoch": 0.32198, + "grad_norm": 0.9497802660399589, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 32198 + }, + { + "epoch": 0.32199, + "grad_norm": 1.225328431390177, + "learning_rate": 0.003, + "loss": 4.057, + "step": 32199 + }, + { + "epoch": 0.322, + "grad_norm": 1.1315937025384777, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 32200 + }, + { + "epoch": 0.32201, + "grad_norm": 0.9948243471510648, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 32201 + }, + { + "epoch": 0.32202, + "grad_norm": 1.2546696649745819, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 32202 + }, + { + "epoch": 0.32203, + "grad_norm": 1.0476631310004367, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 32203 + }, + { + "epoch": 0.32204, + "grad_norm": 1.0216004855487566, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 32204 + }, + { + "epoch": 0.32205, + "grad_norm": 1.0428499466478114, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 32205 + }, + { + "epoch": 0.32206, + "grad_norm": 1.3544394786017568, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 32206 + }, + { + "epoch": 0.32207, + "grad_norm": 0.9357446563818532, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 32207 + }, + { + "epoch": 0.32208, + "grad_norm": 0.8659338282798246, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 32208 + }, + { + "epoch": 0.32209, + "grad_norm": 0.9050586751028818, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 32209 + }, + { + "epoch": 0.3221, + "grad_norm": 1.079273856309754, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 32210 + }, + { + "epoch": 0.32211, + "grad_norm": 1.1969079472952724, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 32211 + }, + { + "epoch": 0.32212, + "grad_norm": 0.8704467889339402, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 32212 + }, + { + "epoch": 0.32213, + "grad_norm": 0.8764952265279083, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 32213 + }, + { + "epoch": 0.32214, + "grad_norm": 1.0921043812729467, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 32214 + }, + { + "epoch": 0.32215, + "grad_norm": 1.1119633838671985, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 32215 + }, + { + "epoch": 0.32216, + "grad_norm": 1.0581761807141197, + "learning_rate": 0.003, + "loss": 4.056, + "step": 32216 + }, + { + "epoch": 0.32217, + "grad_norm": 1.1288922899185303, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 32217 + }, + { + "epoch": 0.32218, + "grad_norm": 0.909996805240633, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 32218 + }, + { + "epoch": 0.32219, + "grad_norm": 1.0294063563148268, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 32219 + }, + { + "epoch": 0.3222, + "grad_norm": 1.2429308848338878, + "learning_rate": 0.003, + "loss": 4.03, + "step": 32220 + }, + { + "epoch": 0.32221, + "grad_norm": 1.0416866059185599, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 32221 + }, + { + "epoch": 0.32222, + "grad_norm": 1.3339134302781388, + "learning_rate": 0.003, + "loss": 4.053, + "step": 32222 + }, + { + "epoch": 0.32223, + "grad_norm": 0.9871311970693925, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 32223 + }, + { + "epoch": 0.32224, + "grad_norm": 1.1205100012246367, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 32224 + }, + { + "epoch": 0.32225, + "grad_norm": 1.0861603773835988, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 32225 + }, + { + "epoch": 0.32226, + "grad_norm": 1.230645536710062, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 32226 + }, + { + "epoch": 0.32227, + "grad_norm": 1.2791470050391742, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 32227 + }, + { + "epoch": 0.32228, + "grad_norm": 1.0490585227600622, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 32228 + }, + { + "epoch": 0.32229, + "grad_norm": 0.8961364509753174, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 32229 + }, + { + "epoch": 0.3223, + "grad_norm": 0.9580170518034434, + "learning_rate": 0.003, + "loss": 4.03, + "step": 32230 + }, + { + "epoch": 0.32231, + "grad_norm": 1.1263250477697024, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 32231 + }, + { + "epoch": 0.32232, + "grad_norm": 0.9779761182216642, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 32232 + }, + { + "epoch": 0.32233, + "grad_norm": 1.124439748905529, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 32233 + }, + { + "epoch": 0.32234, + "grad_norm": 0.9790957558389489, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 32234 + }, + { + "epoch": 0.32235, + "grad_norm": 1.1004275720669312, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 32235 + }, + { + "epoch": 0.32236, + "grad_norm": 1.3493500362057982, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 32236 + }, + { + "epoch": 0.32237, + "grad_norm": 1.0756187173028156, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 32237 + }, + { + "epoch": 0.32238, + "grad_norm": 1.0015776269227454, + "learning_rate": 0.003, + "loss": 4.011, + "step": 32238 + }, + { + "epoch": 0.32239, + "grad_norm": 1.0114068293550154, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 32239 + }, + { + "epoch": 0.3224, + "grad_norm": 1.2268536339205496, + "learning_rate": 0.003, + "loss": 4.065, + "step": 32240 + }, + { + "epoch": 0.32241, + "grad_norm": 0.9706125713622027, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 32241 + }, + { + "epoch": 0.32242, + "grad_norm": 0.9618031989181872, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 32242 + }, + { + "epoch": 0.32243, + "grad_norm": 0.9571547170547305, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 32243 + }, + { + "epoch": 0.32244, + "grad_norm": 1.0725473871469264, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 32244 + }, + { + "epoch": 0.32245, + "grad_norm": 1.0663548262204237, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 32245 + }, + { + "epoch": 0.32246, + "grad_norm": 1.1882484492063963, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 32246 + }, + { + "epoch": 0.32247, + "grad_norm": 0.8760718759027971, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 32247 + }, + { + "epoch": 0.32248, + "grad_norm": 1.0405877498097251, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 32248 + }, + { + "epoch": 0.32249, + "grad_norm": 1.4458950571031168, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 32249 + }, + { + "epoch": 0.3225, + "grad_norm": 1.2609001324012266, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 32250 + }, + { + "epoch": 0.32251, + "grad_norm": 0.7714257232746313, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 32251 + }, + { + "epoch": 0.32252, + "grad_norm": 0.836301606165282, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 32252 + }, + { + "epoch": 0.32253, + "grad_norm": 0.9700669768267618, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 32253 + }, + { + "epoch": 0.32254, + "grad_norm": 1.2424851345459604, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 32254 + }, + { + "epoch": 0.32255, + "grad_norm": 1.0700288922581416, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 32255 + }, + { + "epoch": 0.32256, + "grad_norm": 1.5216721267479052, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 32256 + }, + { + "epoch": 0.32257, + "grad_norm": 0.8094281776092985, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 32257 + }, + { + "epoch": 0.32258, + "grad_norm": 0.8398735104712203, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 32258 + }, + { + "epoch": 0.32259, + "grad_norm": 1.0964291777473485, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 32259 + }, + { + "epoch": 0.3226, + "grad_norm": 1.1475181447948244, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 32260 + }, + { + "epoch": 0.32261, + "grad_norm": 1.1086045536684244, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 32261 + }, + { + "epoch": 0.32262, + "grad_norm": 1.30462387508675, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 32262 + }, + { + "epoch": 0.32263, + "grad_norm": 0.8331360455953755, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 32263 + }, + { + "epoch": 0.32264, + "grad_norm": 0.7688829805174585, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 32264 + }, + { + "epoch": 0.32265, + "grad_norm": 0.8539704665651591, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 32265 + }, + { + "epoch": 0.32266, + "grad_norm": 1.0015642605379138, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 32266 + }, + { + "epoch": 0.32267, + "grad_norm": 1.0293132768988127, + "learning_rate": 0.003, + "loss": 4.067, + "step": 32267 + }, + { + "epoch": 0.32268, + "grad_norm": 0.9722513769506097, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 32268 + }, + { + "epoch": 0.32269, + "grad_norm": 0.9625282353235104, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 32269 + }, + { + "epoch": 0.3227, + "grad_norm": 1.3490193754269888, + "learning_rate": 0.003, + "loss": 4.06, + "step": 32270 + }, + { + "epoch": 0.32271, + "grad_norm": 1.1600334418343334, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 32271 + }, + { + "epoch": 0.32272, + "grad_norm": 1.045983687979229, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 32272 + }, + { + "epoch": 0.32273, + "grad_norm": 1.2766554597609392, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 32273 + }, + { + "epoch": 0.32274, + "grad_norm": 1.0649870952873648, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 32274 + }, + { + "epoch": 0.32275, + "grad_norm": 1.0381285610731008, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 32275 + }, + { + "epoch": 0.32276, + "grad_norm": 1.311822679749782, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 32276 + }, + { + "epoch": 0.32277, + "grad_norm": 1.2573816828417095, + "learning_rate": 0.003, + "loss": 4.058, + "step": 32277 + }, + { + "epoch": 0.32278, + "grad_norm": 0.9932640254500386, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 32278 + }, + { + "epoch": 0.32279, + "grad_norm": 0.9549819630776405, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 32279 + }, + { + "epoch": 0.3228, + "grad_norm": 1.0000351912626888, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 32280 + }, + { + "epoch": 0.32281, + "grad_norm": 0.9320696405401002, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 32281 + }, + { + "epoch": 0.32282, + "grad_norm": 0.922694616315088, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 32282 + }, + { + "epoch": 0.32283, + "grad_norm": 1.10211682024759, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 32283 + }, + { + "epoch": 0.32284, + "grad_norm": 1.0001944957847295, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 32284 + }, + { + "epoch": 0.32285, + "grad_norm": 1.2612601559905965, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 32285 + }, + { + "epoch": 0.32286, + "grad_norm": 0.9211905078915822, + "learning_rate": 0.003, + "loss": 4.048, + "step": 32286 + }, + { + "epoch": 0.32287, + "grad_norm": 1.133265818814609, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 32287 + }, + { + "epoch": 0.32288, + "grad_norm": 0.9784981885025487, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 32288 + }, + { + "epoch": 0.32289, + "grad_norm": 1.0041849569583337, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 32289 + }, + { + "epoch": 0.3229, + "grad_norm": 1.262716349494285, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 32290 + }, + { + "epoch": 0.32291, + "grad_norm": 1.1494251768071944, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 32291 + }, + { + "epoch": 0.32292, + "grad_norm": 1.161436341918068, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 32292 + }, + { + "epoch": 0.32293, + "grad_norm": 1.1468416132544417, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 32293 + }, + { + "epoch": 0.32294, + "grad_norm": 1.0645945068150275, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 32294 + }, + { + "epoch": 0.32295, + "grad_norm": 1.2085801614164462, + "learning_rate": 0.003, + "loss": 4.052, + "step": 32295 + }, + { + "epoch": 0.32296, + "grad_norm": 1.0648594518183678, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 32296 + }, + { + "epoch": 0.32297, + "grad_norm": 1.0845011982698551, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 32297 + }, + { + "epoch": 0.32298, + "grad_norm": 1.0781954105740519, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 32298 + }, + { + "epoch": 0.32299, + "grad_norm": 1.100384720926805, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 32299 + }, + { + "epoch": 0.323, + "grad_norm": 1.3525597286250237, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 32300 + }, + { + "epoch": 0.32301, + "grad_norm": 0.8498101825909904, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 32301 + }, + { + "epoch": 0.32302, + "grad_norm": 0.8911611031387309, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 32302 + }, + { + "epoch": 0.32303, + "grad_norm": 1.2006420808763156, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 32303 + }, + { + "epoch": 0.32304, + "grad_norm": 1.180500056395411, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 32304 + }, + { + "epoch": 0.32305, + "grad_norm": 0.9368694586680094, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 32305 + }, + { + "epoch": 0.32306, + "grad_norm": 1.0318468690022469, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 32306 + }, + { + "epoch": 0.32307, + "grad_norm": 1.22452749072152, + "learning_rate": 0.003, + "loss": 4.045, + "step": 32307 + }, + { + "epoch": 0.32308, + "grad_norm": 0.9991353404582238, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 32308 + }, + { + "epoch": 0.32309, + "grad_norm": 1.0911335454296167, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 32309 + }, + { + "epoch": 0.3231, + "grad_norm": 1.0237298927961636, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 32310 + }, + { + "epoch": 0.32311, + "grad_norm": 1.236432142035171, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 32311 + }, + { + "epoch": 0.32312, + "grad_norm": 0.9096555924385036, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 32312 + }, + { + "epoch": 0.32313, + "grad_norm": 0.9192893174562199, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 32313 + }, + { + "epoch": 0.32314, + "grad_norm": 1.3024492557264837, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 32314 + }, + { + "epoch": 0.32315, + "grad_norm": 0.9693684708126917, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 32315 + }, + { + "epoch": 0.32316, + "grad_norm": 1.1482307862991719, + "learning_rate": 0.003, + "loss": 4.063, + "step": 32316 + }, + { + "epoch": 0.32317, + "grad_norm": 1.145330512272032, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 32317 + }, + { + "epoch": 0.32318, + "grad_norm": 0.9985838654390442, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 32318 + }, + { + "epoch": 0.32319, + "grad_norm": 1.1635413074012078, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 32319 + }, + { + "epoch": 0.3232, + "grad_norm": 1.0745967599890955, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 32320 + }, + { + "epoch": 0.32321, + "grad_norm": 0.9579876328019742, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 32321 + }, + { + "epoch": 0.32322, + "grad_norm": 1.1385315440468287, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 32322 + }, + { + "epoch": 0.32323, + "grad_norm": 1.105666182677298, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 32323 + }, + { + "epoch": 0.32324, + "grad_norm": 0.9310429144613603, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 32324 + }, + { + "epoch": 0.32325, + "grad_norm": 1.214760098180014, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 32325 + }, + { + "epoch": 0.32326, + "grad_norm": 0.9388193186875262, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 32326 + }, + { + "epoch": 0.32327, + "grad_norm": 1.3993830962260827, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 32327 + }, + { + "epoch": 0.32328, + "grad_norm": 0.9537232038420994, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 32328 + }, + { + "epoch": 0.32329, + "grad_norm": 1.0424988782178224, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 32329 + }, + { + "epoch": 0.3233, + "grad_norm": 1.3652842901520406, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 32330 + }, + { + "epoch": 0.32331, + "grad_norm": 0.8394651886885081, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 32331 + }, + { + "epoch": 0.32332, + "grad_norm": 1.098752952208296, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 32332 + }, + { + "epoch": 0.32333, + "grad_norm": 1.2543056055888564, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 32333 + }, + { + "epoch": 0.32334, + "grad_norm": 0.8090894217557097, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 32334 + }, + { + "epoch": 0.32335, + "grad_norm": 0.816727078259081, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 32335 + }, + { + "epoch": 0.32336, + "grad_norm": 0.9289074242961669, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 32336 + }, + { + "epoch": 0.32337, + "grad_norm": 1.1228244463469956, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 32337 + }, + { + "epoch": 0.32338, + "grad_norm": 0.8935434480292621, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 32338 + }, + { + "epoch": 0.32339, + "grad_norm": 1.0668194086236602, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 32339 + }, + { + "epoch": 0.3234, + "grad_norm": 1.17871680734743, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 32340 + }, + { + "epoch": 0.32341, + "grad_norm": 0.958535250312292, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 32341 + }, + { + "epoch": 0.32342, + "grad_norm": 0.9412907181214536, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 32342 + }, + { + "epoch": 0.32343, + "grad_norm": 0.849519108393143, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 32343 + }, + { + "epoch": 0.32344, + "grad_norm": 0.9141980699920238, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 32344 + }, + { + "epoch": 0.32345, + "grad_norm": 1.6207140627394143, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 32345 + }, + { + "epoch": 0.32346, + "grad_norm": 0.7494481190205089, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 32346 + }, + { + "epoch": 0.32347, + "grad_norm": 0.7395812434065252, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 32347 + }, + { + "epoch": 0.32348, + "grad_norm": 0.9390902486300334, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 32348 + }, + { + "epoch": 0.32349, + "grad_norm": 1.1512397824506315, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 32349 + }, + { + "epoch": 0.3235, + "grad_norm": 1.3569317552144047, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 32350 + }, + { + "epoch": 0.32351, + "grad_norm": 1.0123145986830786, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 32351 + }, + { + "epoch": 0.32352, + "grad_norm": 0.9298956714047747, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 32352 + }, + { + "epoch": 0.32353, + "grad_norm": 0.963487147429496, + "learning_rate": 0.003, + "loss": 4.034, + "step": 32353 + }, + { + "epoch": 0.32354, + "grad_norm": 1.328484702358541, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 32354 + }, + { + "epoch": 0.32355, + "grad_norm": 0.8367785580425461, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 32355 + }, + { + "epoch": 0.32356, + "grad_norm": 0.7739641142582966, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 32356 + }, + { + "epoch": 0.32357, + "grad_norm": 1.0753062692783741, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 32357 + }, + { + "epoch": 0.32358, + "grad_norm": 1.1337725399838123, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 32358 + }, + { + "epoch": 0.32359, + "grad_norm": 0.8459178316658426, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 32359 + }, + { + "epoch": 0.3236, + "grad_norm": 0.9090614949507353, + "learning_rate": 0.003, + "loss": 4.038, + "step": 32360 + }, + { + "epoch": 0.32361, + "grad_norm": 1.138186171421296, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 32361 + }, + { + "epoch": 0.32362, + "grad_norm": 1.3961413165878624, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 32362 + }, + { + "epoch": 0.32363, + "grad_norm": 0.8396118709642177, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 32363 + }, + { + "epoch": 0.32364, + "grad_norm": 1.189657281469211, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 32364 + }, + { + "epoch": 0.32365, + "grad_norm": 0.9202575041832537, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 32365 + }, + { + "epoch": 0.32366, + "grad_norm": 1.066914247979181, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 32366 + }, + { + "epoch": 0.32367, + "grad_norm": 1.4937356238987216, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 32367 + }, + { + "epoch": 0.32368, + "grad_norm": 1.0233100446473675, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 32368 + }, + { + "epoch": 0.32369, + "grad_norm": 1.2170347465415574, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 32369 + }, + { + "epoch": 0.3237, + "grad_norm": 1.1312872341890876, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 32370 + }, + { + "epoch": 0.32371, + "grad_norm": 0.9055526510406467, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 32371 + }, + { + "epoch": 0.32372, + "grad_norm": 0.8649291404064763, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 32372 + }, + { + "epoch": 0.32373, + "grad_norm": 1.0179917482561875, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 32373 + }, + { + "epoch": 0.32374, + "grad_norm": 1.1887216034640642, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 32374 + }, + { + "epoch": 0.32375, + "grad_norm": 1.0210131971368583, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 32375 + }, + { + "epoch": 0.32376, + "grad_norm": 1.087140320592426, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 32376 + }, + { + "epoch": 0.32377, + "grad_norm": 1.0192337539970597, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 32377 + }, + { + "epoch": 0.32378, + "grad_norm": 1.1807752507490423, + "learning_rate": 0.003, + "loss": 4.045, + "step": 32378 + }, + { + "epoch": 0.32379, + "grad_norm": 0.9348670908569028, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 32379 + }, + { + "epoch": 0.3238, + "grad_norm": 1.0723770959214967, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 32380 + }, + { + "epoch": 0.32381, + "grad_norm": 1.0568925694779865, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 32381 + }, + { + "epoch": 0.32382, + "grad_norm": 1.0929382148376876, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 32382 + }, + { + "epoch": 0.32383, + "grad_norm": 1.2055984672375812, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 32383 + }, + { + "epoch": 0.32384, + "grad_norm": 1.0627614943860897, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 32384 + }, + { + "epoch": 0.32385, + "grad_norm": 1.0323150412647741, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 32385 + }, + { + "epoch": 0.32386, + "grad_norm": 1.2861483444072561, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 32386 + }, + { + "epoch": 0.32387, + "grad_norm": 0.9511482763135065, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 32387 + }, + { + "epoch": 0.32388, + "grad_norm": 0.9411251513975777, + "learning_rate": 0.003, + "loss": 4.049, + "step": 32388 + }, + { + "epoch": 0.32389, + "grad_norm": 1.0225814700848364, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 32389 + }, + { + "epoch": 0.3239, + "grad_norm": 1.1871652981241967, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 32390 + }, + { + "epoch": 0.32391, + "grad_norm": 1.253866100038436, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 32391 + }, + { + "epoch": 0.32392, + "grad_norm": 1.0809450115739776, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 32392 + }, + { + "epoch": 0.32393, + "grad_norm": 1.3163047128878695, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 32393 + }, + { + "epoch": 0.32394, + "grad_norm": 0.7410139536186602, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 32394 + }, + { + "epoch": 0.32395, + "grad_norm": 0.8503169868395056, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 32395 + }, + { + "epoch": 0.32396, + "grad_norm": 1.061872264136207, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 32396 + }, + { + "epoch": 0.32397, + "grad_norm": 1.1832909834325376, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 32397 + }, + { + "epoch": 0.32398, + "grad_norm": 1.0110143019958928, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 32398 + }, + { + "epoch": 0.32399, + "grad_norm": 1.2184721566982992, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 32399 + }, + { + "epoch": 0.324, + "grad_norm": 1.1539764568534538, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 32400 + }, + { + "epoch": 0.32401, + "grad_norm": 1.10721616193958, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 32401 + }, + { + "epoch": 0.32402, + "grad_norm": 0.9767529917602094, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 32402 + }, + { + "epoch": 0.32403, + "grad_norm": 1.1775706086857154, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 32403 + }, + { + "epoch": 0.32404, + "grad_norm": 1.2290821306578605, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 32404 + }, + { + "epoch": 0.32405, + "grad_norm": 0.9718409123384276, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 32405 + }, + { + "epoch": 0.32406, + "grad_norm": 1.090206954773631, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 32406 + }, + { + "epoch": 0.32407, + "grad_norm": 1.0302142938362289, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 32407 + }, + { + "epoch": 0.32408, + "grad_norm": 1.0123752077099142, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 32408 + }, + { + "epoch": 0.32409, + "grad_norm": 1.202231841201695, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 32409 + }, + { + "epoch": 0.3241, + "grad_norm": 0.9787994012459231, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 32410 + }, + { + "epoch": 0.32411, + "grad_norm": 1.1480259834904158, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 32411 + }, + { + "epoch": 0.32412, + "grad_norm": 1.166113248114455, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 32412 + }, + { + "epoch": 0.32413, + "grad_norm": 1.050197398482545, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 32413 + }, + { + "epoch": 0.32414, + "grad_norm": 1.0088332571373109, + "learning_rate": 0.003, + "loss": 4.055, + "step": 32414 + }, + { + "epoch": 0.32415, + "grad_norm": 1.0510508810105135, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 32415 + }, + { + "epoch": 0.32416, + "grad_norm": 1.1992520352931004, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 32416 + }, + { + "epoch": 0.32417, + "grad_norm": 0.9398689249263076, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 32417 + }, + { + "epoch": 0.32418, + "grad_norm": 0.9829881586969853, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 32418 + }, + { + "epoch": 0.32419, + "grad_norm": 1.0474395346916794, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 32419 + }, + { + "epoch": 0.3242, + "grad_norm": 1.1128691154323342, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 32420 + }, + { + "epoch": 0.32421, + "grad_norm": 0.9452371063960797, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 32421 + }, + { + "epoch": 0.32422, + "grad_norm": 1.0753896299089793, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 32422 + }, + { + "epoch": 0.32423, + "grad_norm": 1.1203551987541267, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 32423 + }, + { + "epoch": 0.32424, + "grad_norm": 1.1878702842367637, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 32424 + }, + { + "epoch": 0.32425, + "grad_norm": 1.0673408326537488, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 32425 + }, + { + "epoch": 0.32426, + "grad_norm": 1.236378756881788, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 32426 + }, + { + "epoch": 0.32427, + "grad_norm": 1.1056409198974444, + "learning_rate": 0.003, + "loss": 4.04, + "step": 32427 + }, + { + "epoch": 0.32428, + "grad_norm": 1.1384977533535354, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 32428 + }, + { + "epoch": 0.32429, + "grad_norm": 1.052309154224397, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 32429 + }, + { + "epoch": 0.3243, + "grad_norm": 1.0965275782876396, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 32430 + }, + { + "epoch": 0.32431, + "grad_norm": 1.0701342513706522, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 32431 + }, + { + "epoch": 0.32432, + "grad_norm": 1.2041087224549893, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 32432 + }, + { + "epoch": 0.32433, + "grad_norm": 1.1970291600204024, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 32433 + }, + { + "epoch": 0.32434, + "grad_norm": 0.9557228332415438, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 32434 + }, + { + "epoch": 0.32435, + "grad_norm": 1.2209806116069448, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 32435 + }, + { + "epoch": 0.32436, + "grad_norm": 0.9497412455269136, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 32436 + }, + { + "epoch": 0.32437, + "grad_norm": 1.0515258794877511, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 32437 + }, + { + "epoch": 0.32438, + "grad_norm": 1.1588346076166083, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 32438 + }, + { + "epoch": 0.32439, + "grad_norm": 1.0743318154930335, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 32439 + }, + { + "epoch": 0.3244, + "grad_norm": 1.1237074814442927, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 32440 + }, + { + "epoch": 0.32441, + "grad_norm": 1.0710697298028045, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 32441 + }, + { + "epoch": 0.32442, + "grad_norm": 1.0181468060060825, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 32442 + }, + { + "epoch": 0.32443, + "grad_norm": 1.1344648108629438, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 32443 + }, + { + "epoch": 0.32444, + "grad_norm": 0.9530934525858707, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 32444 + }, + { + "epoch": 0.32445, + "grad_norm": 1.123047877380485, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 32445 + }, + { + "epoch": 0.32446, + "grad_norm": 1.0353386011265076, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 32446 + }, + { + "epoch": 0.32447, + "grad_norm": 0.9395051502397103, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 32447 + }, + { + "epoch": 0.32448, + "grad_norm": 1.0560812223156868, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 32448 + }, + { + "epoch": 0.32449, + "grad_norm": 1.3239362473336607, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 32449 + }, + { + "epoch": 0.3245, + "grad_norm": 1.0629066067653556, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 32450 + }, + { + "epoch": 0.32451, + "grad_norm": 1.2597656201634009, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 32451 + }, + { + "epoch": 0.32452, + "grad_norm": 0.9139632255681941, + "learning_rate": 0.003, + "loss": 4.059, + "step": 32452 + }, + { + "epoch": 0.32453, + "grad_norm": 1.2208498192102337, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 32453 + }, + { + "epoch": 0.32454, + "grad_norm": 1.2736528035726271, + "learning_rate": 0.003, + "loss": 4.016, + "step": 32454 + }, + { + "epoch": 0.32455, + "grad_norm": 1.0400496988041497, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 32455 + }, + { + "epoch": 0.32456, + "grad_norm": 1.0731379181778908, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 32456 + }, + { + "epoch": 0.32457, + "grad_norm": 1.087689354309093, + "learning_rate": 0.003, + "loss": 4.068, + "step": 32457 + }, + { + "epoch": 0.32458, + "grad_norm": 0.9645362416270481, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 32458 + }, + { + "epoch": 0.32459, + "grad_norm": 1.1640626107489165, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 32459 + }, + { + "epoch": 0.3246, + "grad_norm": 1.0716103491208444, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 32460 + }, + { + "epoch": 0.32461, + "grad_norm": 1.1883991046275189, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 32461 + }, + { + "epoch": 0.32462, + "grad_norm": 1.209247805469429, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 32462 + }, + { + "epoch": 0.32463, + "grad_norm": 0.873767389031544, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 32463 + }, + { + "epoch": 0.32464, + "grad_norm": 0.9353917172236381, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 32464 + }, + { + "epoch": 0.32465, + "grad_norm": 1.0274415449506926, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 32465 + }, + { + "epoch": 0.32466, + "grad_norm": 1.5182271020373224, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 32466 + }, + { + "epoch": 0.32467, + "grad_norm": 1.0076543342693385, + "learning_rate": 0.003, + "loss": 4.035, + "step": 32467 + }, + { + "epoch": 0.32468, + "grad_norm": 1.1526029707805596, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 32468 + }, + { + "epoch": 0.32469, + "grad_norm": 1.1400184890261476, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 32469 + }, + { + "epoch": 0.3247, + "grad_norm": 0.8618323055943568, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 32470 + }, + { + "epoch": 0.32471, + "grad_norm": 0.9952733121192249, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 32471 + }, + { + "epoch": 0.32472, + "grad_norm": 1.1718313999066068, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 32472 + }, + { + "epoch": 0.32473, + "grad_norm": 0.8547152374243784, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 32473 + }, + { + "epoch": 0.32474, + "grad_norm": 0.9154711288938518, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 32474 + }, + { + "epoch": 0.32475, + "grad_norm": 1.1454023660242774, + "learning_rate": 0.003, + "loss": 4.042, + "step": 32475 + }, + { + "epoch": 0.32476, + "grad_norm": 1.1890094587991387, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 32476 + }, + { + "epoch": 0.32477, + "grad_norm": 0.9435513928330196, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 32477 + }, + { + "epoch": 0.32478, + "grad_norm": 1.203833145089492, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 32478 + }, + { + "epoch": 0.32479, + "grad_norm": 1.1361436264721256, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 32479 + }, + { + "epoch": 0.3248, + "grad_norm": 1.271320912276781, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 32480 + }, + { + "epoch": 0.32481, + "grad_norm": 1.149459617627628, + "learning_rate": 0.003, + "loss": 4.042, + "step": 32481 + }, + { + "epoch": 0.32482, + "grad_norm": 1.149717943785927, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 32482 + }, + { + "epoch": 0.32483, + "grad_norm": 0.9412369696011353, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 32483 + }, + { + "epoch": 0.32484, + "grad_norm": 0.9017005936276337, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 32484 + }, + { + "epoch": 0.32485, + "grad_norm": 1.0659958440144026, + "learning_rate": 0.003, + "loss": 4.06, + "step": 32485 + }, + { + "epoch": 0.32486, + "grad_norm": 1.195180605259778, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 32486 + }, + { + "epoch": 0.32487, + "grad_norm": 1.1871513957767308, + "learning_rate": 0.003, + "loss": 4.034, + "step": 32487 + }, + { + "epoch": 0.32488, + "grad_norm": 0.9304554464366017, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 32488 + }, + { + "epoch": 0.32489, + "grad_norm": 0.9486305408259831, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 32489 + }, + { + "epoch": 0.3249, + "grad_norm": 1.2236769612454963, + "learning_rate": 0.003, + "loss": 4.06, + "step": 32490 + }, + { + "epoch": 0.32491, + "grad_norm": 0.9931478330515637, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 32491 + }, + { + "epoch": 0.32492, + "grad_norm": 1.352560910215808, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 32492 + }, + { + "epoch": 0.32493, + "grad_norm": 0.7585599738906713, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 32493 + }, + { + "epoch": 0.32494, + "grad_norm": 0.8678787203352133, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 32494 + }, + { + "epoch": 0.32495, + "grad_norm": 1.0181952425444232, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 32495 + }, + { + "epoch": 0.32496, + "grad_norm": 1.3092137093337681, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 32496 + }, + { + "epoch": 0.32497, + "grad_norm": 0.9546084439683095, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 32497 + }, + { + "epoch": 0.32498, + "grad_norm": 1.4011035724938563, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 32498 + }, + { + "epoch": 0.32499, + "grad_norm": 0.9349916222697623, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 32499 + }, + { + "epoch": 0.325, + "grad_norm": 1.0073924222322161, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 32500 + }, + { + "epoch": 0.32501, + "grad_norm": 1.4685843596576555, + "learning_rate": 0.003, + "loss": 4.053, + "step": 32501 + }, + { + "epoch": 0.32502, + "grad_norm": 1.147440887355232, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 32502 + }, + { + "epoch": 0.32503, + "grad_norm": 0.9933440971522545, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 32503 + }, + { + "epoch": 0.32504, + "grad_norm": 1.058203902293008, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 32504 + }, + { + "epoch": 0.32505, + "grad_norm": 1.0650920322087603, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 32505 + }, + { + "epoch": 0.32506, + "grad_norm": 1.2201100045829694, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 32506 + }, + { + "epoch": 0.32507, + "grad_norm": 1.1784167663036704, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 32507 + }, + { + "epoch": 0.32508, + "grad_norm": 1.089839973378581, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 32508 + }, + { + "epoch": 0.32509, + "grad_norm": 1.2193549243843054, + "learning_rate": 0.003, + "loss": 4.044, + "step": 32509 + }, + { + "epoch": 0.3251, + "grad_norm": 0.8144954109552364, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 32510 + }, + { + "epoch": 0.32511, + "grad_norm": 0.7572150387131418, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 32511 + }, + { + "epoch": 0.32512, + "grad_norm": 0.8774604589614188, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 32512 + }, + { + "epoch": 0.32513, + "grad_norm": 1.14827761013434, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 32513 + }, + { + "epoch": 0.32514, + "grad_norm": 1.0387678182360807, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 32514 + }, + { + "epoch": 0.32515, + "grad_norm": 1.255752876043969, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 32515 + }, + { + "epoch": 0.32516, + "grad_norm": 0.9822527265119309, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 32516 + }, + { + "epoch": 0.32517, + "grad_norm": 1.3265601465781782, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 32517 + }, + { + "epoch": 0.32518, + "grad_norm": 0.8175180857823106, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 32518 + }, + { + "epoch": 0.32519, + "grad_norm": 0.7846293544906638, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 32519 + }, + { + "epoch": 0.3252, + "grad_norm": 0.9519148647271236, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 32520 + }, + { + "epoch": 0.32521, + "grad_norm": 1.2065335300047244, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 32521 + }, + { + "epoch": 0.32522, + "grad_norm": 0.8957823933319111, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 32522 + }, + { + "epoch": 0.32523, + "grad_norm": 0.9563967668687223, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 32523 + }, + { + "epoch": 0.32524, + "grad_norm": 1.2442142250928148, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 32524 + }, + { + "epoch": 0.32525, + "grad_norm": 0.9086608060149213, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 32525 + }, + { + "epoch": 0.32526, + "grad_norm": 0.935472740853471, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 32526 + }, + { + "epoch": 0.32527, + "grad_norm": 1.0345370272993213, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 32527 + }, + { + "epoch": 0.32528, + "grad_norm": 1.0409327482915003, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 32528 + }, + { + "epoch": 0.32529, + "grad_norm": 1.540221245835266, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 32529 + }, + { + "epoch": 0.3253, + "grad_norm": 0.928679173933247, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 32530 + }, + { + "epoch": 0.32531, + "grad_norm": 1.106617720563528, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 32531 + }, + { + "epoch": 0.32532, + "grad_norm": 1.1866802235050984, + "learning_rate": 0.003, + "loss": 4.028, + "step": 32532 + }, + { + "epoch": 0.32533, + "grad_norm": 1.1304623541634313, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 32533 + }, + { + "epoch": 0.32534, + "grad_norm": 1.196403187607473, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 32534 + }, + { + "epoch": 0.32535, + "grad_norm": 1.0591131042839537, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 32535 + }, + { + "epoch": 0.32536, + "grad_norm": 1.0880180610616283, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 32536 + }, + { + "epoch": 0.32537, + "grad_norm": 0.8958566814453591, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 32537 + }, + { + "epoch": 0.32538, + "grad_norm": 1.1269272884112733, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 32538 + }, + { + "epoch": 0.32539, + "grad_norm": 1.042295909458301, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 32539 + }, + { + "epoch": 0.3254, + "grad_norm": 1.1013513668336457, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 32540 + }, + { + "epoch": 0.32541, + "grad_norm": 0.9513131976292009, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 32541 + }, + { + "epoch": 0.32542, + "grad_norm": 1.2357495259682816, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 32542 + }, + { + "epoch": 0.32543, + "grad_norm": 1.197916707279121, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 32543 + }, + { + "epoch": 0.32544, + "grad_norm": 1.0176628001128014, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 32544 + }, + { + "epoch": 0.32545, + "grad_norm": 1.3126906921257746, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 32545 + }, + { + "epoch": 0.32546, + "grad_norm": 0.9214565844494687, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 32546 + }, + { + "epoch": 0.32547, + "grad_norm": 0.9324672359830164, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 32547 + }, + { + "epoch": 0.32548, + "grad_norm": 1.077574839229732, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 32548 + }, + { + "epoch": 0.32549, + "grad_norm": 1.2230547372959224, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 32549 + }, + { + "epoch": 0.3255, + "grad_norm": 1.133728699754645, + "learning_rate": 0.003, + "loss": 4.03, + "step": 32550 + }, + { + "epoch": 0.32551, + "grad_norm": 1.0204233367634534, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 32551 + }, + { + "epoch": 0.32552, + "grad_norm": 1.048415045767757, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 32552 + }, + { + "epoch": 0.32553, + "grad_norm": 1.1648874289623894, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 32553 + }, + { + "epoch": 0.32554, + "grad_norm": 0.90327712204543, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 32554 + }, + { + "epoch": 0.32555, + "grad_norm": 0.9456488953439698, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 32555 + }, + { + "epoch": 0.32556, + "grad_norm": 1.053506156991521, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 32556 + }, + { + "epoch": 0.32557, + "grad_norm": 0.967515482032098, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 32557 + }, + { + "epoch": 0.32558, + "grad_norm": 1.1280653430861722, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 32558 + }, + { + "epoch": 0.32559, + "grad_norm": 1.0899580768584884, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 32559 + }, + { + "epoch": 0.3256, + "grad_norm": 1.3484826588558787, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 32560 + }, + { + "epoch": 0.32561, + "grad_norm": 1.129886383112505, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 32561 + }, + { + "epoch": 0.32562, + "grad_norm": 1.1376064310147578, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 32562 + }, + { + "epoch": 0.32563, + "grad_norm": 1.0653615199888729, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 32563 + }, + { + "epoch": 0.32564, + "grad_norm": 1.0353242908597462, + "learning_rate": 0.003, + "loss": 4.041, + "step": 32564 + }, + { + "epoch": 0.32565, + "grad_norm": 1.0570961499605311, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 32565 + }, + { + "epoch": 0.32566, + "grad_norm": 1.005442574445895, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 32566 + }, + { + "epoch": 0.32567, + "grad_norm": 0.9914838687629995, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 32567 + }, + { + "epoch": 0.32568, + "grad_norm": 1.1056946908192875, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 32568 + }, + { + "epoch": 0.32569, + "grad_norm": 1.1212359406676682, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 32569 + }, + { + "epoch": 0.3257, + "grad_norm": 0.9065341616026726, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 32570 + }, + { + "epoch": 0.32571, + "grad_norm": 0.8340532214705042, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 32571 + }, + { + "epoch": 0.32572, + "grad_norm": 0.9250201530709949, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 32572 + }, + { + "epoch": 0.32573, + "grad_norm": 1.1376631875049905, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 32573 + }, + { + "epoch": 0.32574, + "grad_norm": 1.2969169890376286, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 32574 + }, + { + "epoch": 0.32575, + "grad_norm": 1.0742356387078875, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 32575 + }, + { + "epoch": 0.32576, + "grad_norm": 1.17995627756027, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 32576 + }, + { + "epoch": 0.32577, + "grad_norm": 1.017927500432615, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 32577 + }, + { + "epoch": 0.32578, + "grad_norm": 1.0626906046312672, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 32578 + }, + { + "epoch": 0.32579, + "grad_norm": 1.2427319192939508, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 32579 + }, + { + "epoch": 0.3258, + "grad_norm": 1.1433454214377945, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 32580 + }, + { + "epoch": 0.32581, + "grad_norm": 1.1578374412694776, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 32581 + }, + { + "epoch": 0.32582, + "grad_norm": 1.0504036383625608, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 32582 + }, + { + "epoch": 0.32583, + "grad_norm": 1.1459402692986427, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 32583 + }, + { + "epoch": 0.32584, + "grad_norm": 0.9977186305646808, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 32584 + }, + { + "epoch": 0.32585, + "grad_norm": 1.2510686567677574, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 32585 + }, + { + "epoch": 0.32586, + "grad_norm": 0.7881311695108678, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 32586 + }, + { + "epoch": 0.32587, + "grad_norm": 1.0837246264072655, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 32587 + }, + { + "epoch": 0.32588, + "grad_norm": 1.2724660006872768, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 32588 + }, + { + "epoch": 0.32589, + "grad_norm": 0.9693205611724095, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 32589 + }, + { + "epoch": 0.3259, + "grad_norm": 1.0749165466169617, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 32590 + }, + { + "epoch": 0.32591, + "grad_norm": 1.0009289253951812, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 32591 + }, + { + "epoch": 0.32592, + "grad_norm": 1.1849271581897522, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 32592 + }, + { + "epoch": 0.32593, + "grad_norm": 1.0460180396694283, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 32593 + }, + { + "epoch": 0.32594, + "grad_norm": 1.0920216376793415, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 32594 + }, + { + "epoch": 0.32595, + "grad_norm": 1.0131201622344483, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 32595 + }, + { + "epoch": 0.32596, + "grad_norm": 1.199205372478249, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 32596 + }, + { + "epoch": 0.32597, + "grad_norm": 1.2329602988198824, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 32597 + }, + { + "epoch": 0.32598, + "grad_norm": 1.0866524579192003, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 32598 + }, + { + "epoch": 0.32599, + "grad_norm": 1.1712522586355132, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 32599 + }, + { + "epoch": 0.326, + "grad_norm": 1.1063386329614082, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 32600 + }, + { + "epoch": 0.32601, + "grad_norm": 0.8933435182858873, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 32601 + }, + { + "epoch": 0.32602, + "grad_norm": 1.017932883132579, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 32602 + }, + { + "epoch": 0.32603, + "grad_norm": 1.2683769823879676, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 32603 + }, + { + "epoch": 0.32604, + "grad_norm": 0.951743030982751, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 32604 + }, + { + "epoch": 0.32605, + "grad_norm": 1.105025940359306, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 32605 + }, + { + "epoch": 0.32606, + "grad_norm": 1.0966119244983548, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 32606 + }, + { + "epoch": 0.32607, + "grad_norm": 1.2525614488709589, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 32607 + }, + { + "epoch": 0.32608, + "grad_norm": 0.9966950656324479, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 32608 + }, + { + "epoch": 0.32609, + "grad_norm": 1.3717655822943993, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 32609 + }, + { + "epoch": 0.3261, + "grad_norm": 1.043327646035429, + "learning_rate": 0.003, + "loss": 4.064, + "step": 32610 + }, + { + "epoch": 0.32611, + "grad_norm": 1.0539079838115193, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 32611 + }, + { + "epoch": 0.32612, + "grad_norm": 1.003594183429228, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 32612 + }, + { + "epoch": 0.32613, + "grad_norm": 1.0777039658262455, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 32613 + }, + { + "epoch": 0.32614, + "grad_norm": 1.06913475785382, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 32614 + }, + { + "epoch": 0.32615, + "grad_norm": 1.2004857986926927, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 32615 + }, + { + "epoch": 0.32616, + "grad_norm": 1.0088192331196506, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 32616 + }, + { + "epoch": 0.32617, + "grad_norm": 1.1999116890647126, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 32617 + }, + { + "epoch": 0.32618, + "grad_norm": 0.930527425449492, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 32618 + }, + { + "epoch": 0.32619, + "grad_norm": 1.1036965758131903, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 32619 + }, + { + "epoch": 0.3262, + "grad_norm": 1.1607531580533648, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 32620 + }, + { + "epoch": 0.32621, + "grad_norm": 1.1215550454847285, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 32621 + }, + { + "epoch": 0.32622, + "grad_norm": 1.2078442403813925, + "learning_rate": 0.003, + "loss": 4.03, + "step": 32622 + }, + { + "epoch": 0.32623, + "grad_norm": 1.0653655108808098, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 32623 + }, + { + "epoch": 0.32624, + "grad_norm": 1.1183938328627328, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 32624 + }, + { + "epoch": 0.32625, + "grad_norm": 0.9941543682994044, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 32625 + }, + { + "epoch": 0.32626, + "grad_norm": 1.244404462696845, + "learning_rate": 0.003, + "loss": 4.038, + "step": 32626 + }, + { + "epoch": 0.32627, + "grad_norm": 1.0074617536041144, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 32627 + }, + { + "epoch": 0.32628, + "grad_norm": 1.267409805322788, + "learning_rate": 0.003, + "loss": 4.063, + "step": 32628 + }, + { + "epoch": 0.32629, + "grad_norm": 0.8699357030223396, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 32629 + }, + { + "epoch": 0.3263, + "grad_norm": 0.8102383030055186, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 32630 + }, + { + "epoch": 0.32631, + "grad_norm": 0.8920219861442784, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 32631 + }, + { + "epoch": 0.32632, + "grad_norm": 1.0405542717211187, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 32632 + }, + { + "epoch": 0.32633, + "grad_norm": 1.129639830388938, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 32633 + }, + { + "epoch": 0.32634, + "grad_norm": 0.94169427457491, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 32634 + }, + { + "epoch": 0.32635, + "grad_norm": 1.181813930350324, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 32635 + }, + { + "epoch": 0.32636, + "grad_norm": 0.921852300281146, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 32636 + }, + { + "epoch": 0.32637, + "grad_norm": 1.1987180906177448, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 32637 + }, + { + "epoch": 0.32638, + "grad_norm": 1.2520454169051236, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 32638 + }, + { + "epoch": 0.32639, + "grad_norm": 0.9816471908198209, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 32639 + }, + { + "epoch": 0.3264, + "grad_norm": 1.0914199400248916, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 32640 + }, + { + "epoch": 0.32641, + "grad_norm": 1.2421916854943331, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 32641 + }, + { + "epoch": 0.32642, + "grad_norm": 1.27503278481682, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 32642 + }, + { + "epoch": 0.32643, + "grad_norm": 1.0467335797818884, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 32643 + }, + { + "epoch": 0.32644, + "grad_norm": 1.250279493656514, + "learning_rate": 0.003, + "loss": 4.046, + "step": 32644 + }, + { + "epoch": 0.32645, + "grad_norm": 0.9163867569674198, + "learning_rate": 0.003, + "loss": 4.045, + "step": 32645 + }, + { + "epoch": 0.32646, + "grad_norm": 0.9100713107970934, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 32646 + }, + { + "epoch": 0.32647, + "grad_norm": 1.0829461020998674, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 32647 + }, + { + "epoch": 0.32648, + "grad_norm": 1.1897230768674194, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 32648 + }, + { + "epoch": 0.32649, + "grad_norm": 1.2535693667878323, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 32649 + }, + { + "epoch": 0.3265, + "grad_norm": 1.0550593418876595, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 32650 + }, + { + "epoch": 0.32651, + "grad_norm": 0.8881743534423914, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 32651 + }, + { + "epoch": 0.32652, + "grad_norm": 0.8813107086334819, + "learning_rate": 0.003, + "loss": 4.043, + "step": 32652 + }, + { + "epoch": 0.32653, + "grad_norm": 0.86955109685387, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 32653 + }, + { + "epoch": 0.32654, + "grad_norm": 1.077486620392478, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 32654 + }, + { + "epoch": 0.32655, + "grad_norm": 1.093025232076901, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 32655 + }, + { + "epoch": 0.32656, + "grad_norm": 1.0077623476892867, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 32656 + }, + { + "epoch": 0.32657, + "grad_norm": 1.1732114167628487, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 32657 + }, + { + "epoch": 0.32658, + "grad_norm": 0.9463768161002816, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 32658 + }, + { + "epoch": 0.32659, + "grad_norm": 0.9768808753521103, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 32659 + }, + { + "epoch": 0.3266, + "grad_norm": 1.0339231573243943, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 32660 + }, + { + "epoch": 0.32661, + "grad_norm": 1.0855322671055787, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 32661 + }, + { + "epoch": 0.32662, + "grad_norm": 1.102077343588366, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 32662 + }, + { + "epoch": 0.32663, + "grad_norm": 1.085238988302344, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 32663 + }, + { + "epoch": 0.32664, + "grad_norm": 1.186037858026576, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 32664 + }, + { + "epoch": 0.32665, + "grad_norm": 0.9546989948630725, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 32665 + }, + { + "epoch": 0.32666, + "grad_norm": 1.110174756311738, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 32666 + }, + { + "epoch": 0.32667, + "grad_norm": 1.2095579181575236, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 32667 + }, + { + "epoch": 0.32668, + "grad_norm": 1.1538554796739333, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 32668 + }, + { + "epoch": 0.32669, + "grad_norm": 0.9349447307805878, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 32669 + }, + { + "epoch": 0.3267, + "grad_norm": 1.038968897616925, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 32670 + }, + { + "epoch": 0.32671, + "grad_norm": 1.3416380594331805, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 32671 + }, + { + "epoch": 0.32672, + "grad_norm": 1.1881913942736673, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 32672 + }, + { + "epoch": 0.32673, + "grad_norm": 0.8778798141573366, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 32673 + }, + { + "epoch": 0.32674, + "grad_norm": 0.9644871300803451, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 32674 + }, + { + "epoch": 0.32675, + "grad_norm": 1.2837283734862397, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 32675 + }, + { + "epoch": 0.32676, + "grad_norm": 0.9391410928294471, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 32676 + }, + { + "epoch": 0.32677, + "grad_norm": 1.11787249590384, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 32677 + }, + { + "epoch": 0.32678, + "grad_norm": 1.0919942600342636, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 32678 + }, + { + "epoch": 0.32679, + "grad_norm": 1.0489018895689317, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 32679 + }, + { + "epoch": 0.3268, + "grad_norm": 1.1905448645202985, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 32680 + }, + { + "epoch": 0.32681, + "grad_norm": 1.252145475927943, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 32681 + }, + { + "epoch": 0.32682, + "grad_norm": 1.0992042173613328, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 32682 + }, + { + "epoch": 0.32683, + "grad_norm": 1.1490950484120062, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 32683 + }, + { + "epoch": 0.32684, + "grad_norm": 0.911946659557409, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 32684 + }, + { + "epoch": 0.32685, + "grad_norm": 0.9027094374998721, + "learning_rate": 0.003, + "loss": 4.068, + "step": 32685 + }, + { + "epoch": 0.32686, + "grad_norm": 1.1141263683111202, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 32686 + }, + { + "epoch": 0.32687, + "grad_norm": 1.1936077936505225, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 32687 + }, + { + "epoch": 0.32688, + "grad_norm": 0.9998295905074743, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 32688 + }, + { + "epoch": 0.32689, + "grad_norm": 1.059099294730739, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 32689 + }, + { + "epoch": 0.3269, + "grad_norm": 1.3001279833738477, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 32690 + }, + { + "epoch": 0.32691, + "grad_norm": 1.0312781162661542, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 32691 + }, + { + "epoch": 0.32692, + "grad_norm": 1.105243375401195, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 32692 + }, + { + "epoch": 0.32693, + "grad_norm": 1.1649473254049232, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 32693 + }, + { + "epoch": 0.32694, + "grad_norm": 0.9870099350079525, + "learning_rate": 0.003, + "loss": 4.063, + "step": 32694 + }, + { + "epoch": 0.32695, + "grad_norm": 1.1213715394056951, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 32695 + }, + { + "epoch": 0.32696, + "grad_norm": 1.0306604989066452, + "learning_rate": 0.003, + "loss": 4.047, + "step": 32696 + }, + { + "epoch": 0.32697, + "grad_norm": 1.046194257902389, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 32697 + }, + { + "epoch": 0.32698, + "grad_norm": 1.3355360405926389, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 32698 + }, + { + "epoch": 0.32699, + "grad_norm": 0.8946215722887539, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 32699 + }, + { + "epoch": 0.327, + "grad_norm": 1.0795956599415084, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 32700 + }, + { + "epoch": 0.32701, + "grad_norm": 1.1569096031346982, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 32701 + }, + { + "epoch": 0.32702, + "grad_norm": 1.0068581262541991, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 32702 + }, + { + "epoch": 0.32703, + "grad_norm": 0.9976032436285923, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 32703 + }, + { + "epoch": 0.32704, + "grad_norm": 1.0073802847406492, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 32704 + }, + { + "epoch": 0.32705, + "grad_norm": 1.0530203791125734, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 32705 + }, + { + "epoch": 0.32706, + "grad_norm": 1.0543339546914874, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 32706 + }, + { + "epoch": 0.32707, + "grad_norm": 1.1268572000509558, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 32707 + }, + { + "epoch": 0.32708, + "grad_norm": 0.9959131503244015, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 32708 + }, + { + "epoch": 0.32709, + "grad_norm": 1.0743147370206267, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 32709 + }, + { + "epoch": 0.3271, + "grad_norm": 1.1765877480042366, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 32710 + }, + { + "epoch": 0.32711, + "grad_norm": 0.8945754612028469, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 32711 + }, + { + "epoch": 0.32712, + "grad_norm": 0.9500838016849329, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 32712 + }, + { + "epoch": 0.32713, + "grad_norm": 1.2362721773559042, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 32713 + }, + { + "epoch": 0.32714, + "grad_norm": 0.9665035455348819, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 32714 + }, + { + "epoch": 0.32715, + "grad_norm": 1.3398576673545495, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 32715 + }, + { + "epoch": 0.32716, + "grad_norm": 0.7350785323495301, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 32716 + }, + { + "epoch": 0.32717, + "grad_norm": 0.8852039665761605, + "learning_rate": 0.003, + "loss": 4.055, + "step": 32717 + }, + { + "epoch": 0.32718, + "grad_norm": 1.2819802776475093, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 32718 + }, + { + "epoch": 0.32719, + "grad_norm": 1.0720068533717668, + "learning_rate": 0.003, + "loss": 4.024, + "step": 32719 + }, + { + "epoch": 0.3272, + "grad_norm": 1.3352154225795314, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 32720 + }, + { + "epoch": 0.32721, + "grad_norm": 1.0454590719682952, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 32721 + }, + { + "epoch": 0.32722, + "grad_norm": 0.9730784139583684, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 32722 + }, + { + "epoch": 0.32723, + "grad_norm": 1.0905748521205108, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 32723 + }, + { + "epoch": 0.32724, + "grad_norm": 1.0485452045693768, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 32724 + }, + { + "epoch": 0.32725, + "grad_norm": 1.1608894388341668, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 32725 + }, + { + "epoch": 0.32726, + "grad_norm": 1.151120693541435, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 32726 + }, + { + "epoch": 0.32727, + "grad_norm": 1.0270739153654145, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 32727 + }, + { + "epoch": 0.32728, + "grad_norm": 1.0742973171851518, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 32728 + }, + { + "epoch": 0.32729, + "grad_norm": 1.2098015862547478, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 32729 + }, + { + "epoch": 0.3273, + "grad_norm": 0.9055130878932395, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 32730 + }, + { + "epoch": 0.32731, + "grad_norm": 1.0103434191125908, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 32731 + }, + { + "epoch": 0.32732, + "grad_norm": 1.1248691351715816, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 32732 + }, + { + "epoch": 0.32733, + "grad_norm": 0.8731296458628333, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 32733 + }, + { + "epoch": 0.32734, + "grad_norm": 1.15387642486534, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 32734 + }, + { + "epoch": 0.32735, + "grad_norm": 1.0282976680484552, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 32735 + }, + { + "epoch": 0.32736, + "grad_norm": 1.272950698408999, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 32736 + }, + { + "epoch": 0.32737, + "grad_norm": 1.1683386085413126, + "learning_rate": 0.003, + "loss": 4.091, + "step": 32737 + }, + { + "epoch": 0.32738, + "grad_norm": 1.3234840537555028, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 32738 + }, + { + "epoch": 0.32739, + "grad_norm": 0.9244259125049237, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 32739 + }, + { + "epoch": 0.3274, + "grad_norm": 0.9278264967731583, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 32740 + }, + { + "epoch": 0.32741, + "grad_norm": 1.120624246825543, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 32741 + }, + { + "epoch": 0.32742, + "grad_norm": 1.0557891127240986, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 32742 + }, + { + "epoch": 0.32743, + "grad_norm": 0.9385548248871414, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 32743 + }, + { + "epoch": 0.32744, + "grad_norm": 0.8119159568641413, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 32744 + }, + { + "epoch": 0.32745, + "grad_norm": 0.9996516541732238, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 32745 + }, + { + "epoch": 0.32746, + "grad_norm": 1.2864480748499691, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 32746 + }, + { + "epoch": 0.32747, + "grad_norm": 1.13379488987333, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 32747 + }, + { + "epoch": 0.32748, + "grad_norm": 1.230425876969228, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 32748 + }, + { + "epoch": 0.32749, + "grad_norm": 0.9058998675009722, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 32749 + }, + { + "epoch": 0.3275, + "grad_norm": 1.0722728037932636, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 32750 + }, + { + "epoch": 0.32751, + "grad_norm": 1.2674635905966845, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 32751 + }, + { + "epoch": 0.32752, + "grad_norm": 0.9348914287066533, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 32752 + }, + { + "epoch": 0.32753, + "grad_norm": 1.228086209688431, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 32753 + }, + { + "epoch": 0.32754, + "grad_norm": 1.1565525228434734, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 32754 + }, + { + "epoch": 0.32755, + "grad_norm": 0.974649822948947, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 32755 + }, + { + "epoch": 0.32756, + "grad_norm": 1.1421899969978688, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 32756 + }, + { + "epoch": 0.32757, + "grad_norm": 1.2067528988330052, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 32757 + }, + { + "epoch": 0.32758, + "grad_norm": 1.0624913486887686, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 32758 + }, + { + "epoch": 0.32759, + "grad_norm": 1.033750206354991, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 32759 + }, + { + "epoch": 0.3276, + "grad_norm": 1.0814738152785024, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 32760 + }, + { + "epoch": 0.32761, + "grad_norm": 1.0499024048040593, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 32761 + }, + { + "epoch": 0.32762, + "grad_norm": 1.1095593114702897, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 32762 + }, + { + "epoch": 0.32763, + "grad_norm": 0.8836961454194503, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 32763 + }, + { + "epoch": 0.32764, + "grad_norm": 1.0959261171247463, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 32764 + }, + { + "epoch": 0.32765, + "grad_norm": 1.2296058631191846, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 32765 + }, + { + "epoch": 0.32766, + "grad_norm": 0.9799698302776657, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 32766 + }, + { + "epoch": 0.32767, + "grad_norm": 1.236681750332521, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 32767 + }, + { + "epoch": 0.32768, + "grad_norm": 1.1152032432112342, + "learning_rate": 0.003, + "loss": 4.043, + "step": 32768 + }, + { + "epoch": 0.32769, + "grad_norm": 1.0155593198701154, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 32769 + }, + { + "epoch": 0.3277, + "grad_norm": 1.4550569973222396, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 32770 + }, + { + "epoch": 0.32771, + "grad_norm": 1.026076565552671, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 32771 + }, + { + "epoch": 0.32772, + "grad_norm": 1.0821409633150898, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 32772 + }, + { + "epoch": 0.32773, + "grad_norm": 0.9002618720916135, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 32773 + }, + { + "epoch": 0.32774, + "grad_norm": 0.91022536105691, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 32774 + }, + { + "epoch": 0.32775, + "grad_norm": 1.1406778115652845, + "learning_rate": 0.003, + "loss": 4.014, + "step": 32775 + }, + { + "epoch": 0.32776, + "grad_norm": 1.1089602831264758, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 32776 + }, + { + "epoch": 0.32777, + "grad_norm": 1.0965183510688497, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 32777 + }, + { + "epoch": 0.32778, + "grad_norm": 1.0743859466428547, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 32778 + }, + { + "epoch": 0.32779, + "grad_norm": 1.2030995972615677, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 32779 + }, + { + "epoch": 0.3278, + "grad_norm": 1.028684455703738, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 32780 + }, + { + "epoch": 0.32781, + "grad_norm": 1.119340055451518, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 32781 + }, + { + "epoch": 0.32782, + "grad_norm": 1.0726963162985534, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 32782 + }, + { + "epoch": 0.32783, + "grad_norm": 1.004208526333181, + "learning_rate": 0.003, + "loss": 4.03, + "step": 32783 + }, + { + "epoch": 0.32784, + "grad_norm": 1.1756294365581403, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 32784 + }, + { + "epoch": 0.32785, + "grad_norm": 0.8633418029460062, + "learning_rate": 0.003, + "loss": 4.041, + "step": 32785 + }, + { + "epoch": 0.32786, + "grad_norm": 0.8734220537299441, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 32786 + }, + { + "epoch": 0.32787, + "grad_norm": 1.0436914520327472, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 32787 + }, + { + "epoch": 0.32788, + "grad_norm": 1.2437894079280203, + "learning_rate": 0.003, + "loss": 4.016, + "step": 32788 + }, + { + "epoch": 0.32789, + "grad_norm": 0.9226989424177049, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 32789 + }, + { + "epoch": 0.3279, + "grad_norm": 0.8862667984150538, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 32790 + }, + { + "epoch": 0.32791, + "grad_norm": 0.9243484654839489, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 32791 + }, + { + "epoch": 0.32792, + "grad_norm": 1.1665409651725445, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 32792 + }, + { + "epoch": 0.32793, + "grad_norm": 1.0546390378068022, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 32793 + }, + { + "epoch": 0.32794, + "grad_norm": 1.175189599681202, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 32794 + }, + { + "epoch": 0.32795, + "grad_norm": 1.0362484347250238, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 32795 + }, + { + "epoch": 0.32796, + "grad_norm": 1.0362861789046338, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 32796 + }, + { + "epoch": 0.32797, + "grad_norm": 0.9581814584941082, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 32797 + }, + { + "epoch": 0.32798, + "grad_norm": 1.2403531495180717, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 32798 + }, + { + "epoch": 0.32799, + "grad_norm": 0.8314066718066333, + "learning_rate": 0.003, + "loss": 4.036, + "step": 32799 + }, + { + "epoch": 0.328, + "grad_norm": 0.9141810507314327, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 32800 + }, + { + "epoch": 0.32801, + "grad_norm": 0.9691815065792353, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 32801 + }, + { + "epoch": 0.32802, + "grad_norm": 1.5784772320018488, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 32802 + }, + { + "epoch": 0.32803, + "grad_norm": 0.855185556579807, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 32803 + }, + { + "epoch": 0.32804, + "grad_norm": 0.8818287228667573, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 32804 + }, + { + "epoch": 0.32805, + "grad_norm": 1.1193751133842613, + "learning_rate": 0.003, + "loss": 4.042, + "step": 32805 + }, + { + "epoch": 0.32806, + "grad_norm": 1.0591024381798007, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 32806 + }, + { + "epoch": 0.32807, + "grad_norm": 1.1664045963242085, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 32807 + }, + { + "epoch": 0.32808, + "grad_norm": 0.9900934085672579, + "learning_rate": 0.003, + "loss": 4.1169, + "step": 32808 + }, + { + "epoch": 0.32809, + "grad_norm": 1.180838094844748, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 32809 + }, + { + "epoch": 0.3281, + "grad_norm": 1.0947134164815082, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 32810 + }, + { + "epoch": 0.32811, + "grad_norm": 1.1840969279379503, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 32811 + }, + { + "epoch": 0.32812, + "grad_norm": 1.0570246146756188, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 32812 + }, + { + "epoch": 0.32813, + "grad_norm": 1.0554874672586756, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 32813 + }, + { + "epoch": 0.32814, + "grad_norm": 1.491646664859719, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 32814 + }, + { + "epoch": 0.32815, + "grad_norm": 0.8890044350134411, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 32815 + }, + { + "epoch": 0.32816, + "grad_norm": 0.8944716083828899, + "learning_rate": 0.003, + "loss": 4.086, + "step": 32816 + }, + { + "epoch": 0.32817, + "grad_norm": 1.1085217681647597, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 32817 + }, + { + "epoch": 0.32818, + "grad_norm": 1.1230328954790243, + "learning_rate": 0.003, + "loss": 4.014, + "step": 32818 + }, + { + "epoch": 0.32819, + "grad_norm": 0.9675983474345159, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 32819 + }, + { + "epoch": 0.3282, + "grad_norm": 1.1262311656336752, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 32820 + }, + { + "epoch": 0.32821, + "grad_norm": 1.1801216940772865, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 32821 + }, + { + "epoch": 0.32822, + "grad_norm": 1.0396895257230472, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 32822 + }, + { + "epoch": 0.32823, + "grad_norm": 1.2213286613100893, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 32823 + }, + { + "epoch": 0.32824, + "grad_norm": 0.9726474543733651, + "learning_rate": 0.003, + "loss": 4.034, + "step": 32824 + }, + { + "epoch": 0.32825, + "grad_norm": 1.1648744463538816, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 32825 + }, + { + "epoch": 0.32826, + "grad_norm": 1.1583693938510027, + "learning_rate": 0.003, + "loss": 4.028, + "step": 32826 + }, + { + "epoch": 0.32827, + "grad_norm": 1.1262979007815903, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 32827 + }, + { + "epoch": 0.32828, + "grad_norm": 1.0721351548820965, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 32828 + }, + { + "epoch": 0.32829, + "grad_norm": 0.981274743416914, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 32829 + }, + { + "epoch": 0.3283, + "grad_norm": 1.022440934141349, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 32830 + }, + { + "epoch": 0.32831, + "grad_norm": 1.1873969162641806, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 32831 + }, + { + "epoch": 0.32832, + "grad_norm": 1.1132317642243197, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 32832 + }, + { + "epoch": 0.32833, + "grad_norm": 1.0260084397604687, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 32833 + }, + { + "epoch": 0.32834, + "grad_norm": 0.91647117079699, + "learning_rate": 0.003, + "loss": 4.039, + "step": 32834 + }, + { + "epoch": 0.32835, + "grad_norm": 1.0167089473693762, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 32835 + }, + { + "epoch": 0.32836, + "grad_norm": 1.0383438979336734, + "learning_rate": 0.003, + "loss": 4.044, + "step": 32836 + }, + { + "epoch": 0.32837, + "grad_norm": 1.0147080591761792, + "learning_rate": 0.003, + "loss": 4.055, + "step": 32837 + }, + { + "epoch": 0.32838, + "grad_norm": 1.1290206513413727, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 32838 + }, + { + "epoch": 0.32839, + "grad_norm": 1.04925012100973, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 32839 + }, + { + "epoch": 0.3284, + "grad_norm": 1.1580922695024463, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 32840 + }, + { + "epoch": 0.32841, + "grad_norm": 1.0320556535565115, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 32841 + }, + { + "epoch": 0.32842, + "grad_norm": 1.1192098410190932, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 32842 + }, + { + "epoch": 0.32843, + "grad_norm": 1.149195160985433, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 32843 + }, + { + "epoch": 0.32844, + "grad_norm": 1.0197539468795827, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 32844 + }, + { + "epoch": 0.32845, + "grad_norm": 1.3158712028801038, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 32845 + }, + { + "epoch": 0.32846, + "grad_norm": 0.8798074242410197, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 32846 + }, + { + "epoch": 0.32847, + "grad_norm": 0.9421607852422427, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 32847 + }, + { + "epoch": 0.32848, + "grad_norm": 1.370637702089417, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 32848 + }, + { + "epoch": 0.32849, + "grad_norm": 1.1366235979726986, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 32849 + }, + { + "epoch": 0.3285, + "grad_norm": 0.934328110117203, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 32850 + }, + { + "epoch": 0.32851, + "grad_norm": 0.8875888567335317, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 32851 + }, + { + "epoch": 0.32852, + "grad_norm": 1.0301076605271822, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 32852 + }, + { + "epoch": 0.32853, + "grad_norm": 1.1837586950257792, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 32853 + }, + { + "epoch": 0.32854, + "grad_norm": 0.8915644187548211, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 32854 + }, + { + "epoch": 0.32855, + "grad_norm": 1.1659299013820992, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 32855 + }, + { + "epoch": 0.32856, + "grad_norm": 1.007430711078981, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 32856 + }, + { + "epoch": 0.32857, + "grad_norm": 1.6523295708786083, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 32857 + }, + { + "epoch": 0.32858, + "grad_norm": 0.8064138986542201, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 32858 + }, + { + "epoch": 0.32859, + "grad_norm": 0.9050021756755514, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 32859 + }, + { + "epoch": 0.3286, + "grad_norm": 1.102180823576145, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 32860 + }, + { + "epoch": 0.32861, + "grad_norm": 1.0592818338655339, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 32861 + }, + { + "epoch": 0.32862, + "grad_norm": 1.0306645689267082, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 32862 + }, + { + "epoch": 0.32863, + "grad_norm": 1.1924139112979475, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 32863 + }, + { + "epoch": 0.32864, + "grad_norm": 0.9489170316545495, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 32864 + }, + { + "epoch": 0.32865, + "grad_norm": 1.1273692974280913, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 32865 + }, + { + "epoch": 0.32866, + "grad_norm": 1.300706720162915, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 32866 + }, + { + "epoch": 0.32867, + "grad_norm": 1.072262710822644, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 32867 + }, + { + "epoch": 0.32868, + "grad_norm": 1.0212393051950397, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 32868 + }, + { + "epoch": 0.32869, + "grad_norm": 1.1603594635740766, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 32869 + }, + { + "epoch": 0.3287, + "grad_norm": 0.9471889109386961, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 32870 + }, + { + "epoch": 0.32871, + "grad_norm": 1.1368942938078954, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 32871 + }, + { + "epoch": 0.32872, + "grad_norm": 1.189339133600247, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 32872 + }, + { + "epoch": 0.32873, + "grad_norm": 0.9443314920929559, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 32873 + }, + { + "epoch": 0.32874, + "grad_norm": 1.2114933136048176, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 32874 + }, + { + "epoch": 0.32875, + "grad_norm": 1.311887809688802, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 32875 + }, + { + "epoch": 0.32876, + "grad_norm": 0.8202246872643427, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 32876 + }, + { + "epoch": 0.32877, + "grad_norm": 0.8629943326718447, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 32877 + }, + { + "epoch": 0.32878, + "grad_norm": 0.9451573267823522, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 32878 + }, + { + "epoch": 0.32879, + "grad_norm": 1.1838151939169124, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 32879 + }, + { + "epoch": 0.3288, + "grad_norm": 1.2454994866778086, + "learning_rate": 0.003, + "loss": 4.061, + "step": 32880 + }, + { + "epoch": 0.32881, + "grad_norm": 1.1492796450316, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 32881 + }, + { + "epoch": 0.32882, + "grad_norm": 1.105880770703977, + "learning_rate": 0.003, + "loss": 4.054, + "step": 32882 + }, + { + "epoch": 0.32883, + "grad_norm": 1.0877699056989767, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 32883 + }, + { + "epoch": 0.32884, + "grad_norm": 1.0248362761035208, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 32884 + }, + { + "epoch": 0.32885, + "grad_norm": 0.9524646727755467, + "learning_rate": 0.003, + "loss": 4.027, + "step": 32885 + }, + { + "epoch": 0.32886, + "grad_norm": 1.0506400185464184, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 32886 + }, + { + "epoch": 0.32887, + "grad_norm": 1.023077599962643, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 32887 + }, + { + "epoch": 0.32888, + "grad_norm": 1.1402357089399275, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 32888 + }, + { + "epoch": 0.32889, + "grad_norm": 1.255815816968926, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 32889 + }, + { + "epoch": 0.3289, + "grad_norm": 0.7817024694043023, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 32890 + }, + { + "epoch": 0.32891, + "grad_norm": 1.0058632985856155, + "learning_rate": 0.003, + "loss": 4.035, + "step": 32891 + }, + { + "epoch": 0.32892, + "grad_norm": 1.3382152044993527, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 32892 + }, + { + "epoch": 0.32893, + "grad_norm": 1.1064197635110418, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 32893 + }, + { + "epoch": 0.32894, + "grad_norm": 1.1265729057156884, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 32894 + }, + { + "epoch": 0.32895, + "grad_norm": 1.1029533533395686, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 32895 + }, + { + "epoch": 0.32896, + "grad_norm": 1.2480346406702425, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 32896 + }, + { + "epoch": 0.32897, + "grad_norm": 0.9841560948716194, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 32897 + }, + { + "epoch": 0.32898, + "grad_norm": 0.9946509919386947, + "learning_rate": 0.003, + "loss": 4.028, + "step": 32898 + }, + { + "epoch": 0.32899, + "grad_norm": 1.1332491372102045, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 32899 + }, + { + "epoch": 0.329, + "grad_norm": 1.1523436342174072, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 32900 + }, + { + "epoch": 0.32901, + "grad_norm": 0.9532773812918214, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 32901 + }, + { + "epoch": 0.32902, + "grad_norm": 1.1570079012730594, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 32902 + }, + { + "epoch": 0.32903, + "grad_norm": 1.0365754080072775, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 32903 + }, + { + "epoch": 0.32904, + "grad_norm": 1.145271527960053, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 32904 + }, + { + "epoch": 0.32905, + "grad_norm": 1.1224328208687313, + "learning_rate": 0.003, + "loss": 4.045, + "step": 32905 + }, + { + "epoch": 0.32906, + "grad_norm": 1.208816221087569, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 32906 + }, + { + "epoch": 0.32907, + "grad_norm": 0.9684754528492481, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 32907 + }, + { + "epoch": 0.32908, + "grad_norm": 1.0353443389222892, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 32908 + }, + { + "epoch": 0.32909, + "grad_norm": 1.010778807849494, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 32909 + }, + { + "epoch": 0.3291, + "grad_norm": 1.183111714298224, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 32910 + }, + { + "epoch": 0.32911, + "grad_norm": 0.9907035893199502, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 32911 + }, + { + "epoch": 0.32912, + "grad_norm": 1.0803798559984676, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 32912 + }, + { + "epoch": 0.32913, + "grad_norm": 1.2954532729462294, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 32913 + }, + { + "epoch": 0.32914, + "grad_norm": 0.8293566958661698, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 32914 + }, + { + "epoch": 0.32915, + "grad_norm": 0.8952395952816325, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 32915 + }, + { + "epoch": 0.32916, + "grad_norm": 1.130695816494312, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 32916 + }, + { + "epoch": 0.32917, + "grad_norm": 1.2253996903302231, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 32917 + }, + { + "epoch": 0.32918, + "grad_norm": 0.7725889896210645, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 32918 + }, + { + "epoch": 0.32919, + "grad_norm": 0.8167484491429085, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 32919 + }, + { + "epoch": 0.3292, + "grad_norm": 0.9579043337424948, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 32920 + }, + { + "epoch": 0.32921, + "grad_norm": 1.218672796200067, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 32921 + }, + { + "epoch": 0.32922, + "grad_norm": 1.1574614423993979, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 32922 + }, + { + "epoch": 0.32923, + "grad_norm": 1.0941152286749714, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 32923 + }, + { + "epoch": 0.32924, + "grad_norm": 1.2393889363600534, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 32924 + }, + { + "epoch": 0.32925, + "grad_norm": 1.1479619353906638, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 32925 + }, + { + "epoch": 0.32926, + "grad_norm": 0.952432551175457, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 32926 + }, + { + "epoch": 0.32927, + "grad_norm": 0.911163402343291, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 32927 + }, + { + "epoch": 0.32928, + "grad_norm": 1.004075175875458, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 32928 + }, + { + "epoch": 0.32929, + "grad_norm": 1.1945293610684427, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 32929 + }, + { + "epoch": 0.3293, + "grad_norm": 1.0198820008466642, + "learning_rate": 0.003, + "loss": 4.043, + "step": 32930 + }, + { + "epoch": 0.32931, + "grad_norm": 1.1528449938766983, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 32931 + }, + { + "epoch": 0.32932, + "grad_norm": 1.1795399190267764, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 32932 + }, + { + "epoch": 0.32933, + "grad_norm": 1.0194790188218639, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 32933 + }, + { + "epoch": 0.32934, + "grad_norm": 1.1518456902505754, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 32934 + }, + { + "epoch": 0.32935, + "grad_norm": 1.0903669524770332, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 32935 + }, + { + "epoch": 0.32936, + "grad_norm": 1.1954928973739616, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 32936 + }, + { + "epoch": 0.32937, + "grad_norm": 1.0755674865851756, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 32937 + }, + { + "epoch": 0.32938, + "grad_norm": 1.319776185945577, + "learning_rate": 0.003, + "loss": 4.051, + "step": 32938 + }, + { + "epoch": 0.32939, + "grad_norm": 1.116021425432386, + "learning_rate": 0.003, + "loss": 4.086, + "step": 32939 + }, + { + "epoch": 0.3294, + "grad_norm": 1.0580009701479713, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 32940 + }, + { + "epoch": 0.32941, + "grad_norm": 1.1137922616293372, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 32941 + }, + { + "epoch": 0.32942, + "grad_norm": 1.1452015105627071, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 32942 + }, + { + "epoch": 0.32943, + "grad_norm": 1.144561065589319, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 32943 + }, + { + "epoch": 0.32944, + "grad_norm": 1.0906207247583548, + "learning_rate": 0.003, + "loss": 4.063, + "step": 32944 + }, + { + "epoch": 0.32945, + "grad_norm": 1.1389394357518747, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 32945 + }, + { + "epoch": 0.32946, + "grad_norm": 0.9515023705781903, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 32946 + }, + { + "epoch": 0.32947, + "grad_norm": 0.9664300662688718, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 32947 + }, + { + "epoch": 0.32948, + "grad_norm": 1.140669100213568, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 32948 + }, + { + "epoch": 0.32949, + "grad_norm": 1.1661864222179026, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 32949 + }, + { + "epoch": 0.3295, + "grad_norm": 0.9847615969081918, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 32950 + }, + { + "epoch": 0.32951, + "grad_norm": 1.2357486289527015, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 32951 + }, + { + "epoch": 0.32952, + "grad_norm": 1.1349921723212881, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 32952 + }, + { + "epoch": 0.32953, + "grad_norm": 1.1669159197908396, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 32953 + }, + { + "epoch": 0.32954, + "grad_norm": 1.0980671339181196, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 32954 + }, + { + "epoch": 0.32955, + "grad_norm": 1.2765540741919064, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 32955 + }, + { + "epoch": 0.32956, + "grad_norm": 0.8968485156520635, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 32956 + }, + { + "epoch": 0.32957, + "grad_norm": 0.8731108861507535, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 32957 + }, + { + "epoch": 0.32958, + "grad_norm": 1.163219999121643, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 32958 + }, + { + "epoch": 0.32959, + "grad_norm": 1.1959465968100442, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 32959 + }, + { + "epoch": 0.3296, + "grad_norm": 1.0114735728186572, + "learning_rate": 0.003, + "loss": 4.053, + "step": 32960 + }, + { + "epoch": 0.32961, + "grad_norm": 1.3881039464822014, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 32961 + }, + { + "epoch": 0.32962, + "grad_norm": 0.887738409880535, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 32962 + }, + { + "epoch": 0.32963, + "grad_norm": 0.8219970878861304, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 32963 + }, + { + "epoch": 0.32964, + "grad_norm": 1.2018954342342223, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 32964 + }, + { + "epoch": 0.32965, + "grad_norm": 1.1772261662295138, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 32965 + }, + { + "epoch": 0.32966, + "grad_norm": 1.2741734600522856, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 32966 + }, + { + "epoch": 0.32967, + "grad_norm": 0.9225616398890063, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 32967 + }, + { + "epoch": 0.32968, + "grad_norm": 0.760665558745161, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 32968 + }, + { + "epoch": 0.32969, + "grad_norm": 0.8181167956695393, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 32969 + }, + { + "epoch": 0.3297, + "grad_norm": 1.038996382778784, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 32970 + }, + { + "epoch": 0.32971, + "grad_norm": 1.1227093202299916, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 32971 + }, + { + "epoch": 0.32972, + "grad_norm": 1.3398477309533845, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 32972 + }, + { + "epoch": 0.32973, + "grad_norm": 0.8240256070189123, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 32973 + }, + { + "epoch": 0.32974, + "grad_norm": 0.7007749896815023, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 32974 + }, + { + "epoch": 0.32975, + "grad_norm": 0.7719779693286128, + "learning_rate": 0.003, + "loss": 4.038, + "step": 32975 + }, + { + "epoch": 0.32976, + "grad_norm": 1.1167847110717048, + "learning_rate": 0.003, + "loss": 4.046, + "step": 32976 + }, + { + "epoch": 0.32977, + "grad_norm": 1.1898889764728284, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 32977 + }, + { + "epoch": 0.32978, + "grad_norm": 0.8834644988458241, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 32978 + }, + { + "epoch": 0.32979, + "grad_norm": 0.957376223241626, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 32979 + }, + { + "epoch": 0.3298, + "grad_norm": 1.201211074196585, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 32980 + }, + { + "epoch": 0.32981, + "grad_norm": 0.9469161896467702, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 32981 + }, + { + "epoch": 0.32982, + "grad_norm": 1.3899114973302475, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 32982 + }, + { + "epoch": 0.32983, + "grad_norm": 1.0036349439390844, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 32983 + }, + { + "epoch": 0.32984, + "grad_norm": 1.0586695102688313, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 32984 + }, + { + "epoch": 0.32985, + "grad_norm": 1.1939607269430033, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 32985 + }, + { + "epoch": 0.32986, + "grad_norm": 1.108663950163169, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 32986 + }, + { + "epoch": 0.32987, + "grad_norm": 1.0876541065105494, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 32987 + }, + { + "epoch": 0.32988, + "grad_norm": 1.079554354611725, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 32988 + }, + { + "epoch": 0.32989, + "grad_norm": 1.2941288762588992, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 32989 + }, + { + "epoch": 0.3299, + "grad_norm": 0.9069371873107217, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 32990 + }, + { + "epoch": 0.32991, + "grad_norm": 1.0045125624470113, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 32991 + }, + { + "epoch": 0.32992, + "grad_norm": 1.2341509481298638, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 32992 + }, + { + "epoch": 0.32993, + "grad_norm": 1.0734369222154032, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 32993 + }, + { + "epoch": 0.32994, + "grad_norm": 1.3916651791482508, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 32994 + }, + { + "epoch": 0.32995, + "grad_norm": 0.9778811806118344, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 32995 + }, + { + "epoch": 0.32996, + "grad_norm": 0.9975454915739683, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 32996 + }, + { + "epoch": 0.32997, + "grad_norm": 1.1952058094353109, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 32997 + }, + { + "epoch": 0.32998, + "grad_norm": 1.2805562658302698, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 32998 + }, + { + "epoch": 0.32999, + "grad_norm": 1.3026531883063495, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 32999 + }, + { + "epoch": 0.33, + "grad_norm": 0.9235356928391467, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 33000 + }, + { + "epoch": 0.33001, + "grad_norm": 0.9555245729717873, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 33001 + }, + { + "epoch": 0.33002, + "grad_norm": 1.2315052128552126, + "learning_rate": 0.003, + "loss": 4.023, + "step": 33002 + }, + { + "epoch": 0.33003, + "grad_norm": 1.1153991866047148, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 33003 + }, + { + "epoch": 0.33004, + "grad_norm": 1.1589948404038821, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 33004 + }, + { + "epoch": 0.33005, + "grad_norm": 0.8487189568972537, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 33005 + }, + { + "epoch": 0.33006, + "grad_norm": 0.8397257140416016, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 33006 + }, + { + "epoch": 0.33007, + "grad_norm": 0.9346866972586558, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 33007 + }, + { + "epoch": 0.33008, + "grad_norm": 1.2668743943786236, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 33008 + }, + { + "epoch": 0.33009, + "grad_norm": 1.121187643470503, + "learning_rate": 0.003, + "loss": 4.031, + "step": 33009 + }, + { + "epoch": 0.3301, + "grad_norm": 1.01690378043657, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 33010 + }, + { + "epoch": 0.33011, + "grad_norm": 1.1709576164051256, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 33011 + }, + { + "epoch": 0.33012, + "grad_norm": 0.877439105653169, + "learning_rate": 0.003, + "loss": 4.043, + "step": 33012 + }, + { + "epoch": 0.33013, + "grad_norm": 0.9027321447402605, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 33013 + }, + { + "epoch": 0.33014, + "grad_norm": 1.0425830986962663, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 33014 + }, + { + "epoch": 0.33015, + "grad_norm": 1.1362921753481208, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 33015 + }, + { + "epoch": 0.33016, + "grad_norm": 1.132152027118666, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 33016 + }, + { + "epoch": 0.33017, + "grad_norm": 0.9766987103714856, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 33017 + }, + { + "epoch": 0.33018, + "grad_norm": 1.0646943881312132, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 33018 + }, + { + "epoch": 0.33019, + "grad_norm": 1.1091121684822307, + "learning_rate": 0.003, + "loss": 4.034, + "step": 33019 + }, + { + "epoch": 0.3302, + "grad_norm": 1.2850601591807553, + "learning_rate": 0.003, + "loss": 4.053, + "step": 33020 + }, + { + "epoch": 0.33021, + "grad_norm": 1.149783398935869, + "learning_rate": 0.003, + "loss": 4.051, + "step": 33021 + }, + { + "epoch": 0.33022, + "grad_norm": 0.8660663202953596, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 33022 + }, + { + "epoch": 0.33023, + "grad_norm": 0.7758663511147841, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 33023 + }, + { + "epoch": 0.33024, + "grad_norm": 0.7113131971167662, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 33024 + }, + { + "epoch": 0.33025, + "grad_norm": 0.9089201622662032, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 33025 + }, + { + "epoch": 0.33026, + "grad_norm": 1.217316690879419, + "learning_rate": 0.003, + "loss": 4.054, + "step": 33026 + }, + { + "epoch": 0.33027, + "grad_norm": 0.9176336302480094, + "learning_rate": 0.003, + "loss": 4.068, + "step": 33027 + }, + { + "epoch": 0.33028, + "grad_norm": 1.03189484044629, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 33028 + }, + { + "epoch": 0.33029, + "grad_norm": 1.0927667771422744, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 33029 + }, + { + "epoch": 0.3303, + "grad_norm": 1.0689159622149216, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 33030 + }, + { + "epoch": 0.33031, + "grad_norm": 1.1519158512550047, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 33031 + }, + { + "epoch": 0.33032, + "grad_norm": 1.0434389968631241, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 33032 + }, + { + "epoch": 0.33033, + "grad_norm": 1.0503488032856674, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 33033 + }, + { + "epoch": 0.33034, + "grad_norm": 1.0800819054022313, + "learning_rate": 0.003, + "loss": 4.056, + "step": 33034 + }, + { + "epoch": 0.33035, + "grad_norm": 1.3203598881885399, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 33035 + }, + { + "epoch": 0.33036, + "grad_norm": 1.1467398630436518, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 33036 + }, + { + "epoch": 0.33037, + "grad_norm": 1.3294480063709544, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 33037 + }, + { + "epoch": 0.33038, + "grad_norm": 0.9019924418162506, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 33038 + }, + { + "epoch": 0.33039, + "grad_norm": 0.8989733432254652, + "learning_rate": 0.003, + "loss": 4.054, + "step": 33039 + }, + { + "epoch": 0.3304, + "grad_norm": 1.2420668345536297, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 33040 + }, + { + "epoch": 0.33041, + "grad_norm": 1.0930097943891062, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 33041 + }, + { + "epoch": 0.33042, + "grad_norm": 1.0480352738927248, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 33042 + }, + { + "epoch": 0.33043, + "grad_norm": 1.2294357362275128, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 33043 + }, + { + "epoch": 0.33044, + "grad_norm": 1.171132426679432, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 33044 + }, + { + "epoch": 0.33045, + "grad_norm": 1.0982705151629168, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 33045 + }, + { + "epoch": 0.33046, + "grad_norm": 0.990840037411167, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 33046 + }, + { + "epoch": 0.33047, + "grad_norm": 1.229346888024568, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 33047 + }, + { + "epoch": 0.33048, + "grad_norm": 1.1874487878522066, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 33048 + }, + { + "epoch": 0.33049, + "grad_norm": 1.1443810487016073, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 33049 + }, + { + "epoch": 0.3305, + "grad_norm": 1.0652499375515043, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 33050 + }, + { + "epoch": 0.33051, + "grad_norm": 1.2079798166440847, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 33051 + }, + { + "epoch": 0.33052, + "grad_norm": 0.9086604760285475, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 33052 + }, + { + "epoch": 0.33053, + "grad_norm": 1.001346780758141, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 33053 + }, + { + "epoch": 0.33054, + "grad_norm": 1.2285153482206568, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 33054 + }, + { + "epoch": 0.33055, + "grad_norm": 0.8434761499137663, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 33055 + }, + { + "epoch": 0.33056, + "grad_norm": 0.8595853234420884, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 33056 + }, + { + "epoch": 0.33057, + "grad_norm": 1.0673898160871, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 33057 + }, + { + "epoch": 0.33058, + "grad_norm": 1.2962357599739935, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 33058 + }, + { + "epoch": 0.33059, + "grad_norm": 0.9026907952420866, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 33059 + }, + { + "epoch": 0.3306, + "grad_norm": 0.9299025836170267, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 33060 + }, + { + "epoch": 0.33061, + "grad_norm": 1.1158843252491213, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 33061 + }, + { + "epoch": 0.33062, + "grad_norm": 1.1593723646456886, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 33062 + }, + { + "epoch": 0.33063, + "grad_norm": 1.465661447067864, + "learning_rate": 0.003, + "loss": 4.057, + "step": 33063 + }, + { + "epoch": 0.33064, + "grad_norm": 0.9658012359701834, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 33064 + }, + { + "epoch": 0.33065, + "grad_norm": 1.1537889920267663, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 33065 + }, + { + "epoch": 0.33066, + "grad_norm": 1.1519311970406083, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 33066 + }, + { + "epoch": 0.33067, + "grad_norm": 1.1630938472363785, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 33067 + }, + { + "epoch": 0.33068, + "grad_norm": 1.0031777493201879, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 33068 + }, + { + "epoch": 0.33069, + "grad_norm": 1.2705122249965117, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 33069 + }, + { + "epoch": 0.3307, + "grad_norm": 1.0138926544943256, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 33070 + }, + { + "epoch": 0.33071, + "grad_norm": 1.16898798036113, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 33071 + }, + { + "epoch": 0.33072, + "grad_norm": 1.044043922166102, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 33072 + }, + { + "epoch": 0.33073, + "grad_norm": 1.1113144110849322, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 33073 + }, + { + "epoch": 0.33074, + "grad_norm": 1.1157199795689878, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 33074 + }, + { + "epoch": 0.33075, + "grad_norm": 1.0094605086283244, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 33075 + }, + { + "epoch": 0.33076, + "grad_norm": 1.0854252958208648, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 33076 + }, + { + "epoch": 0.33077, + "grad_norm": 1.0718714136152303, + "learning_rate": 0.003, + "loss": 4.033, + "step": 33077 + }, + { + "epoch": 0.33078, + "grad_norm": 1.0109716216581424, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 33078 + }, + { + "epoch": 0.33079, + "grad_norm": 1.0946015245595395, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 33079 + }, + { + "epoch": 0.3308, + "grad_norm": 0.9904097572434436, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 33080 + }, + { + "epoch": 0.33081, + "grad_norm": 1.1485852804942605, + "learning_rate": 0.003, + "loss": 4.035, + "step": 33081 + }, + { + "epoch": 0.33082, + "grad_norm": 1.1648160038464448, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 33082 + }, + { + "epoch": 0.33083, + "grad_norm": 0.9246028054636176, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 33083 + }, + { + "epoch": 0.33084, + "grad_norm": 1.180075203230791, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 33084 + }, + { + "epoch": 0.33085, + "grad_norm": 1.246943862794258, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 33085 + }, + { + "epoch": 0.33086, + "grad_norm": 1.0371018243586834, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 33086 + }, + { + "epoch": 0.33087, + "grad_norm": 1.089940248825774, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 33087 + }, + { + "epoch": 0.33088, + "grad_norm": 1.3252447187152157, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 33088 + }, + { + "epoch": 0.33089, + "grad_norm": 0.7836937117232808, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 33089 + }, + { + "epoch": 0.3309, + "grad_norm": 0.9326289527571072, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 33090 + }, + { + "epoch": 0.33091, + "grad_norm": 1.0419924410261496, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 33091 + }, + { + "epoch": 0.33092, + "grad_norm": 1.0991214285576576, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 33092 + }, + { + "epoch": 0.33093, + "grad_norm": 1.354864815236083, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 33093 + }, + { + "epoch": 0.33094, + "grad_norm": 0.7950068483415377, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 33094 + }, + { + "epoch": 0.33095, + "grad_norm": 0.9322225845085906, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 33095 + }, + { + "epoch": 0.33096, + "grad_norm": 1.059609839873906, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 33096 + }, + { + "epoch": 0.33097, + "grad_norm": 0.9265726116077287, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 33097 + }, + { + "epoch": 0.33098, + "grad_norm": 1.2144046231235777, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 33098 + }, + { + "epoch": 0.33099, + "grad_norm": 1.2017807876515618, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 33099 + }, + { + "epoch": 0.331, + "grad_norm": 1.310389259668618, + "learning_rate": 0.003, + "loss": 4.044, + "step": 33100 + }, + { + "epoch": 0.33101, + "grad_norm": 1.0024680666093722, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 33101 + }, + { + "epoch": 0.33102, + "grad_norm": 1.0855817523803153, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 33102 + }, + { + "epoch": 0.33103, + "grad_norm": 1.00707677929575, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 33103 + }, + { + "epoch": 0.33104, + "grad_norm": 0.9999744510100576, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 33104 + }, + { + "epoch": 0.33105, + "grad_norm": 1.1155790478073537, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 33105 + }, + { + "epoch": 0.33106, + "grad_norm": 1.0303771137551045, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 33106 + }, + { + "epoch": 0.33107, + "grad_norm": 1.0162713586924785, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 33107 + }, + { + "epoch": 0.33108, + "grad_norm": 1.32422627629117, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 33108 + }, + { + "epoch": 0.33109, + "grad_norm": 0.9341331626855786, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 33109 + }, + { + "epoch": 0.3311, + "grad_norm": 0.9323012927529603, + "learning_rate": 0.003, + "loss": 4.022, + "step": 33110 + }, + { + "epoch": 0.33111, + "grad_norm": 1.1237597095613447, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 33111 + }, + { + "epoch": 0.33112, + "grad_norm": 1.2925198094863934, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 33112 + }, + { + "epoch": 0.33113, + "grad_norm": 0.894843297578504, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 33113 + }, + { + "epoch": 0.33114, + "grad_norm": 0.9059563213422179, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 33114 + }, + { + "epoch": 0.33115, + "grad_norm": 1.0417776679700774, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 33115 + }, + { + "epoch": 0.33116, + "grad_norm": 1.335442122544576, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 33116 + }, + { + "epoch": 0.33117, + "grad_norm": 0.7517130865123741, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 33117 + }, + { + "epoch": 0.33118, + "grad_norm": 0.7701650353578516, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 33118 + }, + { + "epoch": 0.33119, + "grad_norm": 0.7323955605434908, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 33119 + }, + { + "epoch": 0.3312, + "grad_norm": 0.7467795432148829, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 33120 + }, + { + "epoch": 0.33121, + "grad_norm": 0.7792666208331651, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 33121 + }, + { + "epoch": 0.33122, + "grad_norm": 0.9641227511829373, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 33122 + }, + { + "epoch": 0.33123, + "grad_norm": 1.2549649436072279, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 33123 + }, + { + "epoch": 0.33124, + "grad_norm": 0.9485658344711339, + "learning_rate": 0.003, + "loss": 4.003, + "step": 33124 + }, + { + "epoch": 0.33125, + "grad_norm": 1.264830276306524, + "learning_rate": 0.003, + "loss": 4.038, + "step": 33125 + }, + { + "epoch": 0.33126, + "grad_norm": 1.036258492693461, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 33126 + }, + { + "epoch": 0.33127, + "grad_norm": 1.1080521894279063, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 33127 + }, + { + "epoch": 0.33128, + "grad_norm": 1.2220616540179812, + "learning_rate": 0.003, + "loss": 4.051, + "step": 33128 + }, + { + "epoch": 0.33129, + "grad_norm": 1.0198598150862883, + "learning_rate": 0.003, + "loss": 4.061, + "step": 33129 + }, + { + "epoch": 0.3313, + "grad_norm": 1.0094218971637283, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 33130 + }, + { + "epoch": 0.33131, + "grad_norm": 1.09241721845485, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 33131 + }, + { + "epoch": 0.33132, + "grad_norm": 1.3169547607732028, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 33132 + }, + { + "epoch": 0.33133, + "grad_norm": 1.0408144337965097, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 33133 + }, + { + "epoch": 0.33134, + "grad_norm": 1.2190930993874818, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 33134 + }, + { + "epoch": 0.33135, + "grad_norm": 1.0006600079930803, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 33135 + }, + { + "epoch": 0.33136, + "grad_norm": 1.3090823023182638, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 33136 + }, + { + "epoch": 0.33137, + "grad_norm": 1.3218657449694142, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 33137 + }, + { + "epoch": 0.33138, + "grad_norm": 1.0876376956334248, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 33138 + }, + { + "epoch": 0.33139, + "grad_norm": 1.1471068499044532, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 33139 + }, + { + "epoch": 0.3314, + "grad_norm": 1.2315328258922171, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 33140 + }, + { + "epoch": 0.33141, + "grad_norm": 0.9592692932618332, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 33141 + }, + { + "epoch": 0.33142, + "grad_norm": 1.084012376947994, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 33142 + }, + { + "epoch": 0.33143, + "grad_norm": 1.1272172122849646, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 33143 + }, + { + "epoch": 0.33144, + "grad_norm": 1.070105048135619, + "learning_rate": 0.003, + "loss": 4.055, + "step": 33144 + }, + { + "epoch": 0.33145, + "grad_norm": 1.108585993037001, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 33145 + }, + { + "epoch": 0.33146, + "grad_norm": 0.8732283532400141, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 33146 + }, + { + "epoch": 0.33147, + "grad_norm": 1.0169204901729048, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 33147 + }, + { + "epoch": 0.33148, + "grad_norm": 1.239878591726712, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 33148 + }, + { + "epoch": 0.33149, + "grad_norm": 1.0535142700409332, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 33149 + }, + { + "epoch": 0.3315, + "grad_norm": 1.3591477879808338, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 33150 + }, + { + "epoch": 0.33151, + "grad_norm": 0.7909573217943175, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 33151 + }, + { + "epoch": 0.33152, + "grad_norm": 0.7931778502613867, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 33152 + }, + { + "epoch": 0.33153, + "grad_norm": 1.1601604596178847, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 33153 + }, + { + "epoch": 0.33154, + "grad_norm": 1.5176569080892308, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 33154 + }, + { + "epoch": 0.33155, + "grad_norm": 0.8061668957547138, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 33155 + }, + { + "epoch": 0.33156, + "grad_norm": 0.9599647830579248, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 33156 + }, + { + "epoch": 0.33157, + "grad_norm": 1.2445039050655222, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 33157 + }, + { + "epoch": 0.33158, + "grad_norm": 0.9608686381957925, + "learning_rate": 0.003, + "loss": 4.025, + "step": 33158 + }, + { + "epoch": 0.33159, + "grad_norm": 1.0942870905798576, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 33159 + }, + { + "epoch": 0.3316, + "grad_norm": 0.9958674290527839, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 33160 + }, + { + "epoch": 0.33161, + "grad_norm": 1.0206328602906305, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 33161 + }, + { + "epoch": 0.33162, + "grad_norm": 0.978910226766089, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 33162 + }, + { + "epoch": 0.33163, + "grad_norm": 1.1155688519129432, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 33163 + }, + { + "epoch": 0.33164, + "grad_norm": 0.9763379607354878, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 33164 + }, + { + "epoch": 0.33165, + "grad_norm": 1.429124489282371, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 33165 + }, + { + "epoch": 0.33166, + "grad_norm": 0.9894428191188427, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 33166 + }, + { + "epoch": 0.33167, + "grad_norm": 0.9361011257155233, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 33167 + }, + { + "epoch": 0.33168, + "grad_norm": 1.1547224002274252, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 33168 + }, + { + "epoch": 0.33169, + "grad_norm": 1.3431315920833722, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 33169 + }, + { + "epoch": 0.3317, + "grad_norm": 1.2430078681867982, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 33170 + }, + { + "epoch": 0.33171, + "grad_norm": 0.8905417313316385, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 33171 + }, + { + "epoch": 0.33172, + "grad_norm": 0.9941983783472595, + "learning_rate": 0.003, + "loss": 4.027, + "step": 33172 + }, + { + "epoch": 0.33173, + "grad_norm": 1.3511758032512247, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 33173 + }, + { + "epoch": 0.33174, + "grad_norm": 1.0555559901989096, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 33174 + }, + { + "epoch": 0.33175, + "grad_norm": 1.1779853822815405, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 33175 + }, + { + "epoch": 0.33176, + "grad_norm": 1.1343963364192533, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 33176 + }, + { + "epoch": 0.33177, + "grad_norm": 0.9530496775146108, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 33177 + }, + { + "epoch": 0.33178, + "grad_norm": 1.0999974197905005, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 33178 + }, + { + "epoch": 0.33179, + "grad_norm": 1.1187126028540475, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 33179 + }, + { + "epoch": 0.3318, + "grad_norm": 1.2266823533116769, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 33180 + }, + { + "epoch": 0.33181, + "grad_norm": 1.0274930813909187, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 33181 + }, + { + "epoch": 0.33182, + "grad_norm": 1.0495858626925016, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 33182 + }, + { + "epoch": 0.33183, + "grad_norm": 1.252751194793797, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 33183 + }, + { + "epoch": 0.33184, + "grad_norm": 0.8468222484045087, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 33184 + }, + { + "epoch": 0.33185, + "grad_norm": 0.9089263326053736, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 33185 + }, + { + "epoch": 0.33186, + "grad_norm": 1.0864681165829408, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 33186 + }, + { + "epoch": 0.33187, + "grad_norm": 1.0642791237689513, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 33187 + }, + { + "epoch": 0.33188, + "grad_norm": 0.9218954983826967, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 33188 + }, + { + "epoch": 0.33189, + "grad_norm": 0.9807819299286287, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 33189 + }, + { + "epoch": 0.3319, + "grad_norm": 1.2529787753869819, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 33190 + }, + { + "epoch": 0.33191, + "grad_norm": 1.2553707200409114, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 33191 + }, + { + "epoch": 0.33192, + "grad_norm": 1.0257638515804892, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 33192 + }, + { + "epoch": 0.33193, + "grad_norm": 1.0140972086511348, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 33193 + }, + { + "epoch": 0.33194, + "grad_norm": 1.3061617073746357, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 33194 + }, + { + "epoch": 0.33195, + "grad_norm": 0.9712756314640071, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 33195 + }, + { + "epoch": 0.33196, + "grad_norm": 1.2517721835974727, + "learning_rate": 0.003, + "loss": 4.072, + "step": 33196 + }, + { + "epoch": 0.33197, + "grad_norm": 0.9650311743657992, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 33197 + }, + { + "epoch": 0.33198, + "grad_norm": 0.9999606060198339, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 33198 + }, + { + "epoch": 0.33199, + "grad_norm": 1.125274411564263, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 33199 + }, + { + "epoch": 0.332, + "grad_norm": 0.8747584875579906, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 33200 + }, + { + "epoch": 0.33201, + "grad_norm": 1.0103311328453743, + "learning_rate": 0.003, + "loss": 4.06, + "step": 33201 + }, + { + "epoch": 0.33202, + "grad_norm": 1.4986574125048466, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 33202 + }, + { + "epoch": 0.33203, + "grad_norm": 0.6525076065529439, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 33203 + }, + { + "epoch": 0.33204, + "grad_norm": 0.9076402083531954, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 33204 + }, + { + "epoch": 0.33205, + "grad_norm": 1.1643921756071869, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 33205 + }, + { + "epoch": 0.33206, + "grad_norm": 1.2899445231124504, + "learning_rate": 0.003, + "loss": 4.053, + "step": 33206 + }, + { + "epoch": 0.33207, + "grad_norm": 1.2239329959922123, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 33207 + }, + { + "epoch": 0.33208, + "grad_norm": 1.022437253770084, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 33208 + }, + { + "epoch": 0.33209, + "grad_norm": 0.9055323592494574, + "learning_rate": 0.003, + "loss": 4.072, + "step": 33209 + }, + { + "epoch": 0.3321, + "grad_norm": 0.9406643980341813, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 33210 + }, + { + "epoch": 0.33211, + "grad_norm": 0.9867120324339816, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 33211 + }, + { + "epoch": 0.33212, + "grad_norm": 1.182170548929911, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 33212 + }, + { + "epoch": 0.33213, + "grad_norm": 1.0248365335508212, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 33213 + }, + { + "epoch": 0.33214, + "grad_norm": 1.5306013708926434, + "learning_rate": 0.003, + "loss": 4.043, + "step": 33214 + }, + { + "epoch": 0.33215, + "grad_norm": 0.7312968016969003, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 33215 + }, + { + "epoch": 0.33216, + "grad_norm": 0.7993362722840395, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 33216 + }, + { + "epoch": 0.33217, + "grad_norm": 0.9292148727499144, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 33217 + }, + { + "epoch": 0.33218, + "grad_norm": 1.014552933119195, + "learning_rate": 0.003, + "loss": 4.052, + "step": 33218 + }, + { + "epoch": 0.33219, + "grad_norm": 1.2568234874255353, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 33219 + }, + { + "epoch": 0.3322, + "grad_norm": 0.9177566397518674, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 33220 + }, + { + "epoch": 0.33221, + "grad_norm": 1.2758254238718414, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 33221 + }, + { + "epoch": 0.33222, + "grad_norm": 0.9775135576673744, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 33222 + }, + { + "epoch": 0.33223, + "grad_norm": 1.0088718522974771, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 33223 + }, + { + "epoch": 0.33224, + "grad_norm": 1.0810779867205582, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 33224 + }, + { + "epoch": 0.33225, + "grad_norm": 1.13608540243067, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 33225 + }, + { + "epoch": 0.33226, + "grad_norm": 1.0446217770592099, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 33226 + }, + { + "epoch": 0.33227, + "grad_norm": 0.9649510320548248, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 33227 + }, + { + "epoch": 0.33228, + "grad_norm": 1.4374657516855698, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 33228 + }, + { + "epoch": 0.33229, + "grad_norm": 1.0709601914227962, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 33229 + }, + { + "epoch": 0.3323, + "grad_norm": 1.1192823110916215, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 33230 + }, + { + "epoch": 0.33231, + "grad_norm": 1.0582042186712388, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 33231 + }, + { + "epoch": 0.33232, + "grad_norm": 1.2605714660406155, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 33232 + }, + { + "epoch": 0.33233, + "grad_norm": 1.1921936011805438, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 33233 + }, + { + "epoch": 0.33234, + "grad_norm": 1.078884648232986, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 33234 + }, + { + "epoch": 0.33235, + "grad_norm": 1.0345015961568085, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 33235 + }, + { + "epoch": 0.33236, + "grad_norm": 1.2275713687043255, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 33236 + }, + { + "epoch": 0.33237, + "grad_norm": 0.7266703783773412, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 33237 + }, + { + "epoch": 0.33238, + "grad_norm": 0.786314810873333, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 33238 + }, + { + "epoch": 0.33239, + "grad_norm": 1.0016443850545484, + "learning_rate": 0.003, + "loss": 4.052, + "step": 33239 + }, + { + "epoch": 0.3324, + "grad_norm": 1.2118239513630895, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 33240 + }, + { + "epoch": 0.33241, + "grad_norm": 1.1903195340984176, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 33241 + }, + { + "epoch": 0.33242, + "grad_norm": 1.1723362431612034, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 33242 + }, + { + "epoch": 0.33243, + "grad_norm": 1.0635715713710663, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 33243 + }, + { + "epoch": 0.33244, + "grad_norm": 1.0616952927541154, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 33244 + }, + { + "epoch": 0.33245, + "grad_norm": 1.0565318485999213, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 33245 + }, + { + "epoch": 0.33246, + "grad_norm": 0.948377405222966, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 33246 + }, + { + "epoch": 0.33247, + "grad_norm": 1.019163550644036, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 33247 + }, + { + "epoch": 0.33248, + "grad_norm": 1.0253157836604772, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 33248 + }, + { + "epoch": 0.33249, + "grad_norm": 0.9611152329958741, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 33249 + }, + { + "epoch": 0.3325, + "grad_norm": 0.8814359804822163, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 33250 + }, + { + "epoch": 0.33251, + "grad_norm": 0.9347328717055793, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 33251 + }, + { + "epoch": 0.33252, + "grad_norm": 1.0864742095073787, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 33252 + }, + { + "epoch": 0.33253, + "grad_norm": 1.100558453523321, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 33253 + }, + { + "epoch": 0.33254, + "grad_norm": 0.9365888423313703, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 33254 + }, + { + "epoch": 0.33255, + "grad_norm": 1.1428423331148303, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 33255 + }, + { + "epoch": 0.33256, + "grad_norm": 1.6409977546319765, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 33256 + }, + { + "epoch": 0.33257, + "grad_norm": 0.8640866389845245, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 33257 + }, + { + "epoch": 0.33258, + "grad_norm": 0.9017632957426522, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 33258 + }, + { + "epoch": 0.33259, + "grad_norm": 1.1818215125289566, + "learning_rate": 0.003, + "loss": 4.05, + "step": 33259 + }, + { + "epoch": 0.3326, + "grad_norm": 1.3504458247043376, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 33260 + }, + { + "epoch": 0.33261, + "grad_norm": 0.9075674321618687, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 33261 + }, + { + "epoch": 0.33262, + "grad_norm": 1.0505923079917523, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 33262 + }, + { + "epoch": 0.33263, + "grad_norm": 1.2101809809255328, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 33263 + }, + { + "epoch": 0.33264, + "grad_norm": 0.8869815033140813, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 33264 + }, + { + "epoch": 0.33265, + "grad_norm": 1.0775857928876307, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 33265 + }, + { + "epoch": 0.33266, + "grad_norm": 1.356258960859761, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 33266 + }, + { + "epoch": 0.33267, + "grad_norm": 0.8795803338464434, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 33267 + }, + { + "epoch": 0.33268, + "grad_norm": 1.027033639280053, + "learning_rate": 0.003, + "loss": 4.029, + "step": 33268 + }, + { + "epoch": 0.33269, + "grad_norm": 1.1751036303870135, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 33269 + }, + { + "epoch": 0.3327, + "grad_norm": 0.9724096440529029, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 33270 + }, + { + "epoch": 0.33271, + "grad_norm": 1.0886210369162443, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 33271 + }, + { + "epoch": 0.33272, + "grad_norm": 0.8834465754684531, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 33272 + }, + { + "epoch": 0.33273, + "grad_norm": 0.9879801223375759, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 33273 + }, + { + "epoch": 0.33274, + "grad_norm": 1.2654244488173458, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 33274 + }, + { + "epoch": 0.33275, + "grad_norm": 1.0763893051644273, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 33275 + }, + { + "epoch": 0.33276, + "grad_norm": 1.0672472667361916, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 33276 + }, + { + "epoch": 0.33277, + "grad_norm": 1.020117567008848, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 33277 + }, + { + "epoch": 0.33278, + "grad_norm": 1.2640495724793188, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 33278 + }, + { + "epoch": 0.33279, + "grad_norm": 0.996843697816631, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 33279 + }, + { + "epoch": 0.3328, + "grad_norm": 1.2047906666419528, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 33280 + }, + { + "epoch": 0.33281, + "grad_norm": 0.966617995924135, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 33281 + }, + { + "epoch": 0.33282, + "grad_norm": 0.9872533244566479, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 33282 + }, + { + "epoch": 0.33283, + "grad_norm": 1.125066948607633, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 33283 + }, + { + "epoch": 0.33284, + "grad_norm": 0.9801677881758738, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 33284 + }, + { + "epoch": 0.33285, + "grad_norm": 0.8802646175979747, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 33285 + }, + { + "epoch": 0.33286, + "grad_norm": 1.0176188931977375, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 33286 + }, + { + "epoch": 0.33287, + "grad_norm": 1.2989192606758122, + "learning_rate": 0.003, + "loss": 4.036, + "step": 33287 + }, + { + "epoch": 0.33288, + "grad_norm": 0.8724503539425554, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 33288 + }, + { + "epoch": 0.33289, + "grad_norm": 0.9672400120284889, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 33289 + }, + { + "epoch": 0.3329, + "grad_norm": 1.1012968273465453, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 33290 + }, + { + "epoch": 0.33291, + "grad_norm": 1.0933726495649947, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 33291 + }, + { + "epoch": 0.33292, + "grad_norm": 0.9678372744780817, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 33292 + }, + { + "epoch": 0.33293, + "grad_norm": 1.149004661655332, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 33293 + }, + { + "epoch": 0.33294, + "grad_norm": 0.9874259718285584, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 33294 + }, + { + "epoch": 0.33295, + "grad_norm": 1.2688184818683126, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 33295 + }, + { + "epoch": 0.33296, + "grad_norm": 1.2501059985974852, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 33296 + }, + { + "epoch": 0.33297, + "grad_norm": 1.1672322407977167, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 33297 + }, + { + "epoch": 0.33298, + "grad_norm": 1.1345652381915254, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 33298 + }, + { + "epoch": 0.33299, + "grad_norm": 1.1770635360303745, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 33299 + }, + { + "epoch": 0.333, + "grad_norm": 0.9966780570377581, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 33300 + }, + { + "epoch": 0.33301, + "grad_norm": 1.0938492213096733, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 33301 + }, + { + "epoch": 0.33302, + "grad_norm": 1.121485308776742, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 33302 + }, + { + "epoch": 0.33303, + "grad_norm": 1.0297840417133255, + "learning_rate": 0.003, + "loss": 4.03, + "step": 33303 + }, + { + "epoch": 0.33304, + "grad_norm": 1.1500922000811162, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 33304 + }, + { + "epoch": 0.33305, + "grad_norm": 1.0795397673827745, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 33305 + }, + { + "epoch": 0.33306, + "grad_norm": 1.1318936344702302, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 33306 + }, + { + "epoch": 0.33307, + "grad_norm": 0.912572466235311, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 33307 + }, + { + "epoch": 0.33308, + "grad_norm": 1.0660500704955285, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 33308 + }, + { + "epoch": 0.33309, + "grad_norm": 1.17268663098436, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 33309 + }, + { + "epoch": 0.3331, + "grad_norm": 1.1248227294382886, + "learning_rate": 0.003, + "loss": 4.0922, + "step": 33310 + }, + { + "epoch": 0.33311, + "grad_norm": 1.2398333335031615, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 33311 + }, + { + "epoch": 0.33312, + "grad_norm": 0.9863330521790993, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 33312 + }, + { + "epoch": 0.33313, + "grad_norm": 1.1098649373082519, + "learning_rate": 0.003, + "loss": 4.033, + "step": 33313 + }, + { + "epoch": 0.33314, + "grad_norm": 1.0792458754391179, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 33314 + }, + { + "epoch": 0.33315, + "grad_norm": 1.1226166787205798, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 33315 + }, + { + "epoch": 0.33316, + "grad_norm": 1.1586378074508616, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 33316 + }, + { + "epoch": 0.33317, + "grad_norm": 1.058236810903858, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 33317 + }, + { + "epoch": 0.33318, + "grad_norm": 1.094525234338469, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 33318 + }, + { + "epoch": 0.33319, + "grad_norm": 1.0313985309047573, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 33319 + }, + { + "epoch": 0.3332, + "grad_norm": 0.9705779694539961, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 33320 + }, + { + "epoch": 0.33321, + "grad_norm": 1.1227499848497444, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 33321 + }, + { + "epoch": 0.33322, + "grad_norm": 1.115155694337554, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 33322 + }, + { + "epoch": 0.33323, + "grad_norm": 1.1954156114946353, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 33323 + }, + { + "epoch": 0.33324, + "grad_norm": 0.9238262465913571, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 33324 + }, + { + "epoch": 0.33325, + "grad_norm": 1.1619761873191745, + "learning_rate": 0.003, + "loss": 4.026, + "step": 33325 + }, + { + "epoch": 0.33326, + "grad_norm": 1.1097281400402477, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 33326 + }, + { + "epoch": 0.33327, + "grad_norm": 0.9782694074687519, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 33327 + }, + { + "epoch": 0.33328, + "grad_norm": 1.2367710991157015, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 33328 + }, + { + "epoch": 0.33329, + "grad_norm": 0.9232041286478272, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 33329 + }, + { + "epoch": 0.3333, + "grad_norm": 0.8808811438952645, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 33330 + }, + { + "epoch": 0.33331, + "grad_norm": 1.0267757268252682, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 33331 + }, + { + "epoch": 0.33332, + "grad_norm": 1.2175823193777786, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 33332 + }, + { + "epoch": 0.33333, + "grad_norm": 1.0237306006004487, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 33333 + }, + { + "epoch": 0.33334, + "grad_norm": 1.1698871195683778, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 33334 + }, + { + "epoch": 0.33335, + "grad_norm": 1.0337316212375058, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 33335 + }, + { + "epoch": 0.33336, + "grad_norm": 1.286275585073409, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 33336 + }, + { + "epoch": 0.33337, + "grad_norm": 1.0008372640854069, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 33337 + }, + { + "epoch": 0.33338, + "grad_norm": 1.3086844144747674, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 33338 + }, + { + "epoch": 0.33339, + "grad_norm": 1.155614192318637, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 33339 + }, + { + "epoch": 0.3334, + "grad_norm": 1.218649723245425, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 33340 + }, + { + "epoch": 0.33341, + "grad_norm": 0.9164382698938444, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 33341 + }, + { + "epoch": 0.33342, + "grad_norm": 1.0331132759943134, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 33342 + }, + { + "epoch": 0.33343, + "grad_norm": 1.1246980033177023, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 33343 + }, + { + "epoch": 0.33344, + "grad_norm": 1.0167498325820574, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 33344 + }, + { + "epoch": 0.33345, + "grad_norm": 1.1930498564013634, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 33345 + }, + { + "epoch": 0.33346, + "grad_norm": 0.8856677521453473, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 33346 + }, + { + "epoch": 0.33347, + "grad_norm": 0.9299230967919512, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 33347 + }, + { + "epoch": 0.33348, + "grad_norm": 1.0893618879602347, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 33348 + }, + { + "epoch": 0.33349, + "grad_norm": 1.2136638658250167, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 33349 + }, + { + "epoch": 0.3335, + "grad_norm": 0.9878595264450379, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 33350 + }, + { + "epoch": 0.33351, + "grad_norm": 1.172455210762899, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 33351 + }, + { + "epoch": 0.33352, + "grad_norm": 1.2248428867101768, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 33352 + }, + { + "epoch": 0.33353, + "grad_norm": 0.9961245980570909, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 33353 + }, + { + "epoch": 0.33354, + "grad_norm": 1.0559306578372865, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 33354 + }, + { + "epoch": 0.33355, + "grad_norm": 0.990937937125174, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 33355 + }, + { + "epoch": 0.33356, + "grad_norm": 1.1162299963080315, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 33356 + }, + { + "epoch": 0.33357, + "grad_norm": 1.0791262699520943, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 33357 + }, + { + "epoch": 0.33358, + "grad_norm": 1.1785929179564687, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 33358 + }, + { + "epoch": 0.33359, + "grad_norm": 0.9439931884961706, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 33359 + }, + { + "epoch": 0.3336, + "grad_norm": 1.0442669112589409, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 33360 + }, + { + "epoch": 0.33361, + "grad_norm": 1.1577727047183803, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 33361 + }, + { + "epoch": 0.33362, + "grad_norm": 1.3567056473238022, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 33362 + }, + { + "epoch": 0.33363, + "grad_norm": 1.1723590633133512, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 33363 + }, + { + "epoch": 0.33364, + "grad_norm": 0.8099414350238634, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 33364 + }, + { + "epoch": 0.33365, + "grad_norm": 0.9410303256127828, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 33365 + }, + { + "epoch": 0.33366, + "grad_norm": 1.2149384448203862, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 33366 + }, + { + "epoch": 0.33367, + "grad_norm": 0.927625339082501, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 33367 + }, + { + "epoch": 0.33368, + "grad_norm": 0.9201428825015607, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 33368 + }, + { + "epoch": 0.33369, + "grad_norm": 0.9360575509564406, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 33369 + }, + { + "epoch": 0.3337, + "grad_norm": 1.2383404031175334, + "learning_rate": 0.003, + "loss": 4.058, + "step": 33370 + }, + { + "epoch": 0.33371, + "grad_norm": 1.0768587699727705, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 33371 + }, + { + "epoch": 0.33372, + "grad_norm": 1.3966659760839257, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 33372 + }, + { + "epoch": 0.33373, + "grad_norm": 0.9997804694607474, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 33373 + }, + { + "epoch": 0.33374, + "grad_norm": 0.9269734135406177, + "learning_rate": 0.003, + "loss": 4.004, + "step": 33374 + }, + { + "epoch": 0.33375, + "grad_norm": 1.1312473287931886, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 33375 + }, + { + "epoch": 0.33376, + "grad_norm": 1.1617924806658877, + "learning_rate": 0.003, + "loss": 4.049, + "step": 33376 + }, + { + "epoch": 0.33377, + "grad_norm": 0.9674430710122011, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 33377 + }, + { + "epoch": 0.33378, + "grad_norm": 1.1348908955997443, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 33378 + }, + { + "epoch": 0.33379, + "grad_norm": 1.1450579903524387, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 33379 + }, + { + "epoch": 0.3338, + "grad_norm": 1.0428177329180535, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 33380 + }, + { + "epoch": 0.33381, + "grad_norm": 0.8980305085801041, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 33381 + }, + { + "epoch": 0.33382, + "grad_norm": 0.997075942202163, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 33382 + }, + { + "epoch": 0.33383, + "grad_norm": 1.36124654038813, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 33383 + }, + { + "epoch": 0.33384, + "grad_norm": 1.0143257775714227, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 33384 + }, + { + "epoch": 0.33385, + "grad_norm": 1.152459928930593, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 33385 + }, + { + "epoch": 0.33386, + "grad_norm": 0.9741990517788531, + "learning_rate": 0.003, + "loss": 4.017, + "step": 33386 + }, + { + "epoch": 0.33387, + "grad_norm": 1.1787522040376897, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 33387 + }, + { + "epoch": 0.33388, + "grad_norm": 1.033339710609771, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 33388 + }, + { + "epoch": 0.33389, + "grad_norm": 1.1950360997845146, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 33389 + }, + { + "epoch": 0.3339, + "grad_norm": 1.109587878909062, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 33390 + }, + { + "epoch": 0.33391, + "grad_norm": 1.2033013647073105, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 33391 + }, + { + "epoch": 0.33392, + "grad_norm": 0.9836273441574027, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 33392 + }, + { + "epoch": 0.33393, + "grad_norm": 1.2306251281274394, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 33393 + }, + { + "epoch": 0.33394, + "grad_norm": 0.9335008041322741, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 33394 + }, + { + "epoch": 0.33395, + "grad_norm": 0.9025092426019036, + "learning_rate": 0.003, + "loss": 4.049, + "step": 33395 + }, + { + "epoch": 0.33396, + "grad_norm": 1.1894627838587257, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 33396 + }, + { + "epoch": 0.33397, + "grad_norm": 1.2470921159783472, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 33397 + }, + { + "epoch": 0.33398, + "grad_norm": 0.9699083769501572, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 33398 + }, + { + "epoch": 0.33399, + "grad_norm": 1.2105344362315311, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 33399 + }, + { + "epoch": 0.334, + "grad_norm": 0.9990067227100452, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 33400 + }, + { + "epoch": 0.33401, + "grad_norm": 1.1287526897646218, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 33401 + }, + { + "epoch": 0.33402, + "grad_norm": 1.1165978918187682, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 33402 + }, + { + "epoch": 0.33403, + "grad_norm": 1.2456721830275117, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 33403 + }, + { + "epoch": 0.33404, + "grad_norm": 0.9512813313351012, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 33404 + }, + { + "epoch": 0.33405, + "grad_norm": 1.1888342101130318, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 33405 + }, + { + "epoch": 0.33406, + "grad_norm": 1.017856575998936, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 33406 + }, + { + "epoch": 0.33407, + "grad_norm": 1.290405644549135, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 33407 + }, + { + "epoch": 0.33408, + "grad_norm": 1.0464330795203391, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 33408 + }, + { + "epoch": 0.33409, + "grad_norm": 1.0833200956293354, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 33409 + }, + { + "epoch": 0.3341, + "grad_norm": 1.0806143051175532, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 33410 + }, + { + "epoch": 0.33411, + "grad_norm": 1.0821879783574089, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 33411 + }, + { + "epoch": 0.33412, + "grad_norm": 1.060566314071876, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 33412 + }, + { + "epoch": 0.33413, + "grad_norm": 1.244714993041856, + "learning_rate": 0.003, + "loss": 4.053, + "step": 33413 + }, + { + "epoch": 0.33414, + "grad_norm": 1.0526665300367803, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 33414 + }, + { + "epoch": 0.33415, + "grad_norm": 1.1405178252472692, + "learning_rate": 0.003, + "loss": 4.014, + "step": 33415 + }, + { + "epoch": 0.33416, + "grad_norm": 1.0478708502376324, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 33416 + }, + { + "epoch": 0.33417, + "grad_norm": 0.9936309632075992, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 33417 + }, + { + "epoch": 0.33418, + "grad_norm": 1.278595065126082, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 33418 + }, + { + "epoch": 0.33419, + "grad_norm": 0.9486826572973209, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 33419 + }, + { + "epoch": 0.3342, + "grad_norm": 1.118801975493145, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 33420 + }, + { + "epoch": 0.33421, + "grad_norm": 1.1182481551786019, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 33421 + }, + { + "epoch": 0.33422, + "grad_norm": 1.1444620516601456, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 33422 + }, + { + "epoch": 0.33423, + "grad_norm": 0.810501146968662, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 33423 + }, + { + "epoch": 0.33424, + "grad_norm": 0.8884686631095771, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 33424 + }, + { + "epoch": 0.33425, + "grad_norm": 1.0753417053687169, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 33425 + }, + { + "epoch": 0.33426, + "grad_norm": 1.1658460924544405, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 33426 + }, + { + "epoch": 0.33427, + "grad_norm": 1.0896348964270506, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 33427 + }, + { + "epoch": 0.33428, + "grad_norm": 0.9384181316241255, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 33428 + }, + { + "epoch": 0.33429, + "grad_norm": 0.9486408117987624, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 33429 + }, + { + "epoch": 0.3343, + "grad_norm": 1.137912016925253, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 33430 + }, + { + "epoch": 0.33431, + "grad_norm": 1.2159842174870479, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 33431 + }, + { + "epoch": 0.33432, + "grad_norm": 1.198001713348813, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 33432 + }, + { + "epoch": 0.33433, + "grad_norm": 0.9945340119630435, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 33433 + }, + { + "epoch": 0.33434, + "grad_norm": 1.130204757872619, + "learning_rate": 0.003, + "loss": 4.039, + "step": 33434 + }, + { + "epoch": 0.33435, + "grad_norm": 1.1374637733128379, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 33435 + }, + { + "epoch": 0.33436, + "grad_norm": 1.126422588259581, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 33436 + }, + { + "epoch": 0.33437, + "grad_norm": 1.1870512762145815, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 33437 + }, + { + "epoch": 0.33438, + "grad_norm": 0.8026597379204244, + "learning_rate": 0.003, + "loss": 4.02, + "step": 33438 + }, + { + "epoch": 0.33439, + "grad_norm": 0.8876881144154257, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 33439 + }, + { + "epoch": 0.3344, + "grad_norm": 1.037845650142723, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 33440 + }, + { + "epoch": 0.33441, + "grad_norm": 1.2695727530527143, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 33441 + }, + { + "epoch": 0.33442, + "grad_norm": 0.9409513446258215, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 33442 + }, + { + "epoch": 0.33443, + "grad_norm": 1.4053994403110979, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 33443 + }, + { + "epoch": 0.33444, + "grad_norm": 1.0396625394110912, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 33444 + }, + { + "epoch": 0.33445, + "grad_norm": 1.013914649716546, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 33445 + }, + { + "epoch": 0.33446, + "grad_norm": 1.0281860626709014, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 33446 + }, + { + "epoch": 0.33447, + "grad_norm": 1.046245663921306, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 33447 + }, + { + "epoch": 0.33448, + "grad_norm": 1.2971693896626464, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 33448 + }, + { + "epoch": 0.33449, + "grad_norm": 1.0506447845922824, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 33449 + }, + { + "epoch": 0.3345, + "grad_norm": 1.1203064239320233, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 33450 + }, + { + "epoch": 0.33451, + "grad_norm": 1.1449283441695104, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 33451 + }, + { + "epoch": 0.33452, + "grad_norm": 1.3249598109353178, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 33452 + }, + { + "epoch": 0.33453, + "grad_norm": 0.9405258793750317, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 33453 + }, + { + "epoch": 0.33454, + "grad_norm": 1.0604185953207104, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 33454 + }, + { + "epoch": 0.33455, + "grad_norm": 1.058251884490825, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 33455 + }, + { + "epoch": 0.33456, + "grad_norm": 1.2614731666840981, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 33456 + }, + { + "epoch": 0.33457, + "grad_norm": 0.9965033303339387, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 33457 + }, + { + "epoch": 0.33458, + "grad_norm": 1.279980230980588, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 33458 + }, + { + "epoch": 0.33459, + "grad_norm": 1.2266333327056629, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 33459 + }, + { + "epoch": 0.3346, + "grad_norm": 0.8925946768545643, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 33460 + }, + { + "epoch": 0.33461, + "grad_norm": 0.8604779561209858, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 33461 + }, + { + "epoch": 0.33462, + "grad_norm": 1.0113858806081624, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 33462 + }, + { + "epoch": 0.33463, + "grad_norm": 1.284000616526191, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 33463 + }, + { + "epoch": 0.33464, + "grad_norm": 0.8645877887836642, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 33464 + }, + { + "epoch": 0.33465, + "grad_norm": 1.0130843623986991, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 33465 + }, + { + "epoch": 0.33466, + "grad_norm": 1.3503005359584335, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 33466 + }, + { + "epoch": 0.33467, + "grad_norm": 0.9658765092454891, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 33467 + }, + { + "epoch": 0.33468, + "grad_norm": 1.0633983563118012, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 33468 + }, + { + "epoch": 0.33469, + "grad_norm": 1.076934209327917, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 33469 + }, + { + "epoch": 0.3347, + "grad_norm": 1.0836412908488067, + "learning_rate": 0.003, + "loss": 4.0974, + "step": 33470 + }, + { + "epoch": 0.33471, + "grad_norm": 1.062714197045719, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 33471 + }, + { + "epoch": 0.33472, + "grad_norm": 1.1244045090342323, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 33472 + }, + { + "epoch": 0.33473, + "grad_norm": 0.8795881061727154, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 33473 + }, + { + "epoch": 0.33474, + "grad_norm": 1.0840005346205708, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 33474 + }, + { + "epoch": 0.33475, + "grad_norm": 1.1155983806842342, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 33475 + }, + { + "epoch": 0.33476, + "grad_norm": 1.0802801558778725, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 33476 + }, + { + "epoch": 0.33477, + "grad_norm": 1.005037288175103, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 33477 + }, + { + "epoch": 0.33478, + "grad_norm": 1.2365840636550707, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 33478 + }, + { + "epoch": 0.33479, + "grad_norm": 0.9943988882515945, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 33479 + }, + { + "epoch": 0.3348, + "grad_norm": 1.2157810723686362, + "learning_rate": 0.003, + "loss": 4.066, + "step": 33480 + }, + { + "epoch": 0.33481, + "grad_norm": 1.0910907666992213, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 33481 + }, + { + "epoch": 0.33482, + "grad_norm": 0.9665758356223879, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 33482 + }, + { + "epoch": 0.33483, + "grad_norm": 1.1517076724849329, + "learning_rate": 0.003, + "loss": 4.023, + "step": 33483 + }, + { + "epoch": 0.33484, + "grad_norm": 1.3429790449568488, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 33484 + }, + { + "epoch": 0.33485, + "grad_norm": 0.9007177805883014, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 33485 + }, + { + "epoch": 0.33486, + "grad_norm": 0.971907905547948, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 33486 + }, + { + "epoch": 0.33487, + "grad_norm": 1.2562264271872305, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 33487 + }, + { + "epoch": 0.33488, + "grad_norm": 0.9148267218011045, + "learning_rate": 0.003, + "loss": 4.034, + "step": 33488 + }, + { + "epoch": 0.33489, + "grad_norm": 0.9923945481867834, + "learning_rate": 0.003, + "loss": 4.031, + "step": 33489 + }, + { + "epoch": 0.3349, + "grad_norm": 1.025128656889925, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 33490 + }, + { + "epoch": 0.33491, + "grad_norm": 0.9629350180978697, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 33491 + }, + { + "epoch": 0.33492, + "grad_norm": 1.2963785663403702, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 33492 + }, + { + "epoch": 0.33493, + "grad_norm": 0.8746061827618173, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 33493 + }, + { + "epoch": 0.33494, + "grad_norm": 1.0607192318173104, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 33494 + }, + { + "epoch": 0.33495, + "grad_norm": 1.402205460773169, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 33495 + }, + { + "epoch": 0.33496, + "grad_norm": 0.9167418467663975, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 33496 + }, + { + "epoch": 0.33497, + "grad_norm": 1.1265844083480614, + "learning_rate": 0.003, + "loss": 4.0906, + "step": 33497 + }, + { + "epoch": 0.33498, + "grad_norm": 1.1804176359285594, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 33498 + }, + { + "epoch": 0.33499, + "grad_norm": 1.2105581171485231, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 33499 + }, + { + "epoch": 0.335, + "grad_norm": 1.3105741028852989, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 33500 + }, + { + "epoch": 0.33501, + "grad_norm": 0.7937757990339293, + "learning_rate": 0.003, + "loss": 4.025, + "step": 33501 + }, + { + "epoch": 0.33502, + "grad_norm": 0.8209561320599486, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 33502 + }, + { + "epoch": 0.33503, + "grad_norm": 1.1525915706481435, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 33503 + }, + { + "epoch": 0.33504, + "grad_norm": 1.1334203590192018, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 33504 + }, + { + "epoch": 0.33505, + "grad_norm": 1.2925245256030422, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 33505 + }, + { + "epoch": 0.33506, + "grad_norm": 0.760100177677063, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 33506 + }, + { + "epoch": 0.33507, + "grad_norm": 0.7864657893621501, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 33507 + }, + { + "epoch": 0.33508, + "grad_norm": 0.9416397315014706, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 33508 + }, + { + "epoch": 0.33509, + "grad_norm": 1.113104153445045, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 33509 + }, + { + "epoch": 0.3351, + "grad_norm": 1.1114917570448548, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 33510 + }, + { + "epoch": 0.33511, + "grad_norm": 1.0721225041223077, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 33511 + }, + { + "epoch": 0.33512, + "grad_norm": 1.136335972436359, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 33512 + }, + { + "epoch": 0.33513, + "grad_norm": 1.1613044993448505, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 33513 + }, + { + "epoch": 0.33514, + "grad_norm": 1.1060166275160828, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 33514 + }, + { + "epoch": 0.33515, + "grad_norm": 1.001307886931013, + "learning_rate": 0.003, + "loss": 4.023, + "step": 33515 + }, + { + "epoch": 0.33516, + "grad_norm": 1.229736988546126, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 33516 + }, + { + "epoch": 0.33517, + "grad_norm": 0.9537298053598237, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 33517 + }, + { + "epoch": 0.33518, + "grad_norm": 1.1108377889537175, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 33518 + }, + { + "epoch": 0.33519, + "grad_norm": 1.0825279879478409, + "learning_rate": 0.003, + "loss": 4.012, + "step": 33519 + }, + { + "epoch": 0.3352, + "grad_norm": 1.0153540542324495, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 33520 + }, + { + "epoch": 0.33521, + "grad_norm": 1.0203539040641783, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 33521 + }, + { + "epoch": 0.33522, + "grad_norm": 1.0385457053350762, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 33522 + }, + { + "epoch": 0.33523, + "grad_norm": 0.9775341846609807, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 33523 + }, + { + "epoch": 0.33524, + "grad_norm": 1.0712082376612793, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 33524 + }, + { + "epoch": 0.33525, + "grad_norm": 1.1867023557104392, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 33525 + }, + { + "epoch": 0.33526, + "grad_norm": 0.9081344180629654, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 33526 + }, + { + "epoch": 0.33527, + "grad_norm": 1.0305914179335418, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 33527 + }, + { + "epoch": 0.33528, + "grad_norm": 1.3829222004073194, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 33528 + }, + { + "epoch": 0.33529, + "grad_norm": 0.873278919710927, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 33529 + }, + { + "epoch": 0.3353, + "grad_norm": 0.9204609345704192, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 33530 + }, + { + "epoch": 0.33531, + "grad_norm": 1.0283216599259157, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 33531 + }, + { + "epoch": 0.33532, + "grad_norm": 1.1622289449505827, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 33532 + }, + { + "epoch": 0.33533, + "grad_norm": 1.214873799402269, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 33533 + }, + { + "epoch": 0.33534, + "grad_norm": 1.2030738994052639, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 33534 + }, + { + "epoch": 0.33535, + "grad_norm": 1.1953457298898507, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 33535 + }, + { + "epoch": 0.33536, + "grad_norm": 1.2911951472609957, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 33536 + }, + { + "epoch": 0.33537, + "grad_norm": 0.9746561179064244, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 33537 + }, + { + "epoch": 0.33538, + "grad_norm": 1.0985117161064646, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 33538 + }, + { + "epoch": 0.33539, + "grad_norm": 1.1778806899071044, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 33539 + }, + { + "epoch": 0.3354, + "grad_norm": 0.9136070691770373, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 33540 + }, + { + "epoch": 0.33541, + "grad_norm": 1.1102777822601249, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 33541 + }, + { + "epoch": 0.33542, + "grad_norm": 1.1727956638539956, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 33542 + }, + { + "epoch": 0.33543, + "grad_norm": 0.8183502295098132, + "learning_rate": 0.003, + "loss": 4.023, + "step": 33543 + }, + { + "epoch": 0.33544, + "grad_norm": 0.9144197887030383, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 33544 + }, + { + "epoch": 0.33545, + "grad_norm": 1.1335536088061557, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 33545 + }, + { + "epoch": 0.33546, + "grad_norm": 1.1877916589745843, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 33546 + }, + { + "epoch": 0.33547, + "grad_norm": 1.3436761699984943, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 33547 + }, + { + "epoch": 0.33548, + "grad_norm": 1.04239463955051, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 33548 + }, + { + "epoch": 0.33549, + "grad_norm": 0.9940488823978096, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 33549 + }, + { + "epoch": 0.3355, + "grad_norm": 1.1172013799751983, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 33550 + }, + { + "epoch": 0.33551, + "grad_norm": 1.1165191784431423, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 33551 + }, + { + "epoch": 0.33552, + "grad_norm": 1.0294012057275401, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 33552 + }, + { + "epoch": 0.33553, + "grad_norm": 1.1417996560990198, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 33553 + }, + { + "epoch": 0.33554, + "grad_norm": 1.2999911668098485, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 33554 + }, + { + "epoch": 0.33555, + "grad_norm": 0.7608641443477125, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 33555 + }, + { + "epoch": 0.33556, + "grad_norm": 0.8909631128936206, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 33556 + }, + { + "epoch": 0.33557, + "grad_norm": 1.0698468338201912, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 33557 + }, + { + "epoch": 0.33558, + "grad_norm": 1.120843333044803, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 33558 + }, + { + "epoch": 0.33559, + "grad_norm": 1.0859582118464408, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 33559 + }, + { + "epoch": 0.3356, + "grad_norm": 1.157800488266656, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 33560 + }, + { + "epoch": 0.33561, + "grad_norm": 0.9646230520678868, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 33561 + }, + { + "epoch": 0.33562, + "grad_norm": 1.035365903527961, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 33562 + }, + { + "epoch": 0.33563, + "grad_norm": 1.0824426330698045, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 33563 + }, + { + "epoch": 0.33564, + "grad_norm": 1.1882812942378524, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 33564 + }, + { + "epoch": 0.33565, + "grad_norm": 1.5590797057476784, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 33565 + }, + { + "epoch": 0.33566, + "grad_norm": 1.0082362654350554, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 33566 + }, + { + "epoch": 0.33567, + "grad_norm": 0.8801923668272332, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 33567 + }, + { + "epoch": 0.33568, + "grad_norm": 0.9907863635825083, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 33568 + }, + { + "epoch": 0.33569, + "grad_norm": 1.3770473482572692, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 33569 + }, + { + "epoch": 0.3357, + "grad_norm": 1.0034027473190448, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 33570 + }, + { + "epoch": 0.33571, + "grad_norm": 1.3085263551607482, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 33571 + }, + { + "epoch": 0.33572, + "grad_norm": 0.8947551830668722, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 33572 + }, + { + "epoch": 0.33573, + "grad_norm": 0.873758438456406, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 33573 + }, + { + "epoch": 0.33574, + "grad_norm": 1.0532535473531375, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 33574 + }, + { + "epoch": 0.33575, + "grad_norm": 1.2132582932096383, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 33575 + }, + { + "epoch": 0.33576, + "grad_norm": 0.9971196894514947, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 33576 + }, + { + "epoch": 0.33577, + "grad_norm": 1.2478152593989134, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 33577 + }, + { + "epoch": 0.33578, + "grad_norm": 0.9989466182728001, + "learning_rate": 0.003, + "loss": 4.0959, + "step": 33578 + }, + { + "epoch": 0.33579, + "grad_norm": 1.1936723729361103, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 33579 + }, + { + "epoch": 0.3358, + "grad_norm": 0.985466496595951, + "learning_rate": 0.003, + "loss": 4.039, + "step": 33580 + }, + { + "epoch": 0.33581, + "grad_norm": 1.27939660235184, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 33581 + }, + { + "epoch": 0.33582, + "grad_norm": 1.052244411515386, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 33582 + }, + { + "epoch": 0.33583, + "grad_norm": 1.13822076419641, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 33583 + }, + { + "epoch": 0.33584, + "grad_norm": 1.0931684467672464, + "learning_rate": 0.003, + "loss": 4.04, + "step": 33584 + }, + { + "epoch": 0.33585, + "grad_norm": 1.278557889499896, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 33585 + }, + { + "epoch": 0.33586, + "grad_norm": 0.97525851268909, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 33586 + }, + { + "epoch": 0.33587, + "grad_norm": 1.166079247936948, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 33587 + }, + { + "epoch": 0.33588, + "grad_norm": 0.971905632400039, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 33588 + }, + { + "epoch": 0.33589, + "grad_norm": 1.0606398926496232, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 33589 + }, + { + "epoch": 0.3359, + "grad_norm": 1.1760376979046254, + "learning_rate": 0.003, + "loss": 4.0931, + "step": 33590 + }, + { + "epoch": 0.33591, + "grad_norm": 1.161436946647382, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 33591 + }, + { + "epoch": 0.33592, + "grad_norm": 0.87396238427341, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 33592 + }, + { + "epoch": 0.33593, + "grad_norm": 0.9133141787392055, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 33593 + }, + { + "epoch": 0.33594, + "grad_norm": 1.2369033057264032, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 33594 + }, + { + "epoch": 0.33595, + "grad_norm": 1.1033496289638687, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 33595 + }, + { + "epoch": 0.33596, + "grad_norm": 1.1647823245325648, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 33596 + }, + { + "epoch": 0.33597, + "grad_norm": 1.0405785372157172, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 33597 + }, + { + "epoch": 0.33598, + "grad_norm": 1.0059426073242446, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 33598 + }, + { + "epoch": 0.33599, + "grad_norm": 1.2337899629324025, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 33599 + }, + { + "epoch": 0.336, + "grad_norm": 0.9786383783547883, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 33600 + }, + { + "epoch": 0.33601, + "grad_norm": 0.9894917945380116, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 33601 + }, + { + "epoch": 0.33602, + "grad_norm": 1.0567436928226293, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 33602 + }, + { + "epoch": 0.33603, + "grad_norm": 1.1826941617486286, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 33603 + }, + { + "epoch": 0.33604, + "grad_norm": 0.8110645003195911, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 33604 + }, + { + "epoch": 0.33605, + "grad_norm": 0.9675438136674755, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 33605 + }, + { + "epoch": 0.33606, + "grad_norm": 1.4897966389942876, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 33606 + }, + { + "epoch": 0.33607, + "grad_norm": 1.0884322718363513, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 33607 + }, + { + "epoch": 0.33608, + "grad_norm": 1.2686134198043164, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 33608 + }, + { + "epoch": 0.33609, + "grad_norm": 0.7848441706017502, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 33609 + }, + { + "epoch": 0.3361, + "grad_norm": 0.9979903906537293, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 33610 + }, + { + "epoch": 0.33611, + "grad_norm": 1.248071694118968, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 33611 + }, + { + "epoch": 0.33612, + "grad_norm": 0.9524246948234667, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 33612 + }, + { + "epoch": 0.33613, + "grad_norm": 1.4865051591306235, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 33613 + }, + { + "epoch": 0.33614, + "grad_norm": 1.1314432718188248, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 33614 + }, + { + "epoch": 0.33615, + "grad_norm": 1.0886832013241994, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 33615 + }, + { + "epoch": 0.33616, + "grad_norm": 1.0826696849878208, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 33616 + }, + { + "epoch": 0.33617, + "grad_norm": 1.014840668970094, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 33617 + }, + { + "epoch": 0.33618, + "grad_norm": 1.1963419565769073, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 33618 + }, + { + "epoch": 0.33619, + "grad_norm": 0.8289002925880453, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 33619 + }, + { + "epoch": 0.3362, + "grad_norm": 0.895591811013665, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 33620 + }, + { + "epoch": 0.33621, + "grad_norm": 1.2228709606725294, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 33621 + }, + { + "epoch": 0.33622, + "grad_norm": 1.1672751276893718, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 33622 + }, + { + "epoch": 0.33623, + "grad_norm": 1.041231829110575, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 33623 + }, + { + "epoch": 0.33624, + "grad_norm": 1.1778332147031059, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 33624 + }, + { + "epoch": 0.33625, + "grad_norm": 1.0193384802829477, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 33625 + }, + { + "epoch": 0.33626, + "grad_norm": 1.1774069551267057, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 33626 + }, + { + "epoch": 0.33627, + "grad_norm": 1.0635652143417489, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 33627 + }, + { + "epoch": 0.33628, + "grad_norm": 1.0510091638839416, + "learning_rate": 0.003, + "loss": 4.065, + "step": 33628 + }, + { + "epoch": 0.33629, + "grad_norm": 1.169900213831366, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 33629 + }, + { + "epoch": 0.3363, + "grad_norm": 0.9420266635587355, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 33630 + }, + { + "epoch": 0.33631, + "grad_norm": 0.9061610489055771, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 33631 + }, + { + "epoch": 0.33632, + "grad_norm": 1.0098186181183781, + "learning_rate": 0.003, + "loss": 4.029, + "step": 33632 + }, + { + "epoch": 0.33633, + "grad_norm": 1.0549448418916845, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 33633 + }, + { + "epoch": 0.33634, + "grad_norm": 1.0332033654112902, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 33634 + }, + { + "epoch": 0.33635, + "grad_norm": 1.068182523036032, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 33635 + }, + { + "epoch": 0.33636, + "grad_norm": 0.9526520886659138, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 33636 + }, + { + "epoch": 0.33637, + "grad_norm": 1.2082040116317863, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 33637 + }, + { + "epoch": 0.33638, + "grad_norm": 1.1298436444820044, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 33638 + }, + { + "epoch": 0.33639, + "grad_norm": 1.2024070520189387, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 33639 + }, + { + "epoch": 0.3364, + "grad_norm": 0.847517306145917, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 33640 + }, + { + "epoch": 0.33641, + "grad_norm": 0.945174470031901, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 33641 + }, + { + "epoch": 0.33642, + "grad_norm": 1.2126261952991204, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 33642 + }, + { + "epoch": 0.33643, + "grad_norm": 1.06264889814489, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 33643 + }, + { + "epoch": 0.33644, + "grad_norm": 0.8401461716709637, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 33644 + }, + { + "epoch": 0.33645, + "grad_norm": 0.9872074619569438, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 33645 + }, + { + "epoch": 0.33646, + "grad_norm": 1.249594286990585, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 33646 + }, + { + "epoch": 0.33647, + "grad_norm": 1.064967151677333, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 33647 + }, + { + "epoch": 0.33648, + "grad_norm": 0.951878510977666, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 33648 + }, + { + "epoch": 0.33649, + "grad_norm": 1.1399586935073367, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 33649 + }, + { + "epoch": 0.3365, + "grad_norm": 1.195758672733086, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 33650 + }, + { + "epoch": 0.33651, + "grad_norm": 0.8523082024290569, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 33651 + }, + { + "epoch": 0.33652, + "grad_norm": 0.8308214059782925, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 33652 + }, + { + "epoch": 0.33653, + "grad_norm": 1.0325440600587497, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 33653 + }, + { + "epoch": 0.33654, + "grad_norm": 1.3986444375075955, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 33654 + }, + { + "epoch": 0.33655, + "grad_norm": 0.8057369465708372, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 33655 + }, + { + "epoch": 0.33656, + "grad_norm": 0.8881272733173603, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 33656 + }, + { + "epoch": 0.33657, + "grad_norm": 1.0242726215105937, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 33657 + }, + { + "epoch": 0.33658, + "grad_norm": 1.2194496310596612, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 33658 + }, + { + "epoch": 0.33659, + "grad_norm": 1.0302379871363807, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 33659 + }, + { + "epoch": 0.3366, + "grad_norm": 1.236466411836065, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 33660 + }, + { + "epoch": 0.33661, + "grad_norm": 1.0646602320170067, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 33661 + }, + { + "epoch": 0.33662, + "grad_norm": 1.149393491555629, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 33662 + }, + { + "epoch": 0.33663, + "grad_norm": 1.1192118783735459, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 33663 + }, + { + "epoch": 0.33664, + "grad_norm": 0.924851260370346, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 33664 + }, + { + "epoch": 0.33665, + "grad_norm": 1.1098370962357105, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 33665 + }, + { + "epoch": 0.33666, + "grad_norm": 1.2174543485209501, + "learning_rate": 0.003, + "loss": 4.043, + "step": 33666 + }, + { + "epoch": 0.33667, + "grad_norm": 1.1430482336355619, + "learning_rate": 0.003, + "loss": 4.037, + "step": 33667 + }, + { + "epoch": 0.33668, + "grad_norm": 0.9975072733856482, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 33668 + }, + { + "epoch": 0.33669, + "grad_norm": 1.1365035507616377, + "learning_rate": 0.003, + "loss": 4.04, + "step": 33669 + }, + { + "epoch": 0.3367, + "grad_norm": 0.960940165879838, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 33670 + }, + { + "epoch": 0.33671, + "grad_norm": 1.1549060763149708, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 33671 + }, + { + "epoch": 0.33672, + "grad_norm": 1.0833796239185247, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 33672 + }, + { + "epoch": 0.33673, + "grad_norm": 1.3945777080266457, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 33673 + }, + { + "epoch": 0.33674, + "grad_norm": 1.0430767156001355, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 33674 + }, + { + "epoch": 0.33675, + "grad_norm": 1.0611230857480136, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 33675 + }, + { + "epoch": 0.33676, + "grad_norm": 1.1784564652646612, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 33676 + }, + { + "epoch": 0.33677, + "grad_norm": 0.9760001435836277, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 33677 + }, + { + "epoch": 0.33678, + "grad_norm": 1.0197460289926707, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 33678 + }, + { + "epoch": 0.33679, + "grad_norm": 1.1777047552293929, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 33679 + }, + { + "epoch": 0.3368, + "grad_norm": 0.9654645824569923, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 33680 + }, + { + "epoch": 0.33681, + "grad_norm": 1.0791559858713273, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 33681 + }, + { + "epoch": 0.33682, + "grad_norm": 1.1128200633669225, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 33682 + }, + { + "epoch": 0.33683, + "grad_norm": 1.1715757273961696, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 33683 + }, + { + "epoch": 0.33684, + "grad_norm": 1.1042756674853236, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 33684 + }, + { + "epoch": 0.33685, + "grad_norm": 1.1837107335892871, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 33685 + }, + { + "epoch": 0.33686, + "grad_norm": 0.9134488657128694, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 33686 + }, + { + "epoch": 0.33687, + "grad_norm": 0.8941329340656302, + "learning_rate": 0.003, + "loss": 4.013, + "step": 33687 + }, + { + "epoch": 0.33688, + "grad_norm": 0.9674454429155855, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 33688 + }, + { + "epoch": 0.33689, + "grad_norm": 1.2091371527608246, + "learning_rate": 0.003, + "loss": 4.024, + "step": 33689 + }, + { + "epoch": 0.3369, + "grad_norm": 1.0371905359718734, + "learning_rate": 0.003, + "loss": 4.026, + "step": 33690 + }, + { + "epoch": 0.33691, + "grad_norm": 1.1781331408692868, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 33691 + }, + { + "epoch": 0.33692, + "grad_norm": 0.9893686697840157, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 33692 + }, + { + "epoch": 0.33693, + "grad_norm": 1.0241236437929035, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 33693 + }, + { + "epoch": 0.33694, + "grad_norm": 1.256680409865194, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 33694 + }, + { + "epoch": 0.33695, + "grad_norm": 0.8631586136724347, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 33695 + }, + { + "epoch": 0.33696, + "grad_norm": 0.7828961273036517, + "learning_rate": 0.003, + "loss": 4.02, + "step": 33696 + }, + { + "epoch": 0.33697, + "grad_norm": 0.8202960825438594, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 33697 + }, + { + "epoch": 0.33698, + "grad_norm": 0.9502008568566216, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 33698 + }, + { + "epoch": 0.33699, + "grad_norm": 1.1390417240288133, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 33699 + }, + { + "epoch": 0.337, + "grad_norm": 1.2772536186568968, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 33700 + }, + { + "epoch": 0.33701, + "grad_norm": 0.9676958379840487, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 33701 + }, + { + "epoch": 0.33702, + "grad_norm": 1.2339574677230054, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 33702 + }, + { + "epoch": 0.33703, + "grad_norm": 0.9784290015196682, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 33703 + }, + { + "epoch": 0.33704, + "grad_norm": 1.3380232533521212, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 33704 + }, + { + "epoch": 0.33705, + "grad_norm": 1.0695331153124459, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 33705 + }, + { + "epoch": 0.33706, + "grad_norm": 1.2660586195784376, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 33706 + }, + { + "epoch": 0.33707, + "grad_norm": 0.918830604415341, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 33707 + }, + { + "epoch": 0.33708, + "grad_norm": 1.1837620305127348, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 33708 + }, + { + "epoch": 0.33709, + "grad_norm": 1.0629085260363744, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 33709 + }, + { + "epoch": 0.3371, + "grad_norm": 1.2151056496270791, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 33710 + }, + { + "epoch": 0.33711, + "grad_norm": 1.0817773737706808, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 33711 + }, + { + "epoch": 0.33712, + "grad_norm": 1.25766577712652, + "learning_rate": 0.003, + "loss": 4.044, + "step": 33712 + }, + { + "epoch": 0.33713, + "grad_norm": 1.1541530684425658, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 33713 + }, + { + "epoch": 0.33714, + "grad_norm": 1.0815107448215984, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 33714 + }, + { + "epoch": 0.33715, + "grad_norm": 1.2329109359046315, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 33715 + }, + { + "epoch": 0.33716, + "grad_norm": 1.200798373227851, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 33716 + }, + { + "epoch": 0.33717, + "grad_norm": 1.0163201381815283, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 33717 + }, + { + "epoch": 0.33718, + "grad_norm": 1.154439431688693, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 33718 + }, + { + "epoch": 0.33719, + "grad_norm": 1.0471152254041145, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 33719 + }, + { + "epoch": 0.3372, + "grad_norm": 1.063803109569859, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 33720 + }, + { + "epoch": 0.33721, + "grad_norm": 0.9859643495105057, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 33721 + }, + { + "epoch": 0.33722, + "grad_norm": 1.0002055048648142, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 33722 + }, + { + "epoch": 0.33723, + "grad_norm": 1.2386022549061495, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 33723 + }, + { + "epoch": 0.33724, + "grad_norm": 0.9044599692042393, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 33724 + }, + { + "epoch": 0.33725, + "grad_norm": 0.8092984947515856, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 33725 + }, + { + "epoch": 0.33726, + "grad_norm": 0.8641016007771386, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 33726 + }, + { + "epoch": 0.33727, + "grad_norm": 1.0778295445551034, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 33727 + }, + { + "epoch": 0.33728, + "grad_norm": 1.1603883184604176, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 33728 + }, + { + "epoch": 0.33729, + "grad_norm": 0.9536483414437014, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 33729 + }, + { + "epoch": 0.3373, + "grad_norm": 1.2299894105656524, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 33730 + }, + { + "epoch": 0.33731, + "grad_norm": 1.3078655796150023, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 33731 + }, + { + "epoch": 0.33732, + "grad_norm": 1.066760685946242, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 33732 + }, + { + "epoch": 0.33733, + "grad_norm": 1.011571129694601, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 33733 + }, + { + "epoch": 0.33734, + "grad_norm": 1.1786998660923862, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 33734 + }, + { + "epoch": 0.33735, + "grad_norm": 1.0868641908860697, + "learning_rate": 0.003, + "loss": 4.033, + "step": 33735 + }, + { + "epoch": 0.33736, + "grad_norm": 1.0428296426545183, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 33736 + }, + { + "epoch": 0.33737, + "grad_norm": 1.1329004986499622, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 33737 + }, + { + "epoch": 0.33738, + "grad_norm": 0.933146158161402, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 33738 + }, + { + "epoch": 0.33739, + "grad_norm": 1.07360042486239, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 33739 + }, + { + "epoch": 0.3374, + "grad_norm": 1.0656828402653125, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 33740 + }, + { + "epoch": 0.33741, + "grad_norm": 1.047526351323221, + "learning_rate": 0.003, + "loss": 4.055, + "step": 33741 + }, + { + "epoch": 0.33742, + "grad_norm": 1.0745021637394767, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 33742 + }, + { + "epoch": 0.33743, + "grad_norm": 1.0023927629100366, + "learning_rate": 0.003, + "loss": 4.003, + "step": 33743 + }, + { + "epoch": 0.33744, + "grad_norm": 1.2030361091590092, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 33744 + }, + { + "epoch": 0.33745, + "grad_norm": 0.8916508943181377, + "learning_rate": 0.003, + "loss": 4.032, + "step": 33745 + }, + { + "epoch": 0.33746, + "grad_norm": 0.9261364969257578, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 33746 + }, + { + "epoch": 0.33747, + "grad_norm": 1.0362432044068348, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 33747 + }, + { + "epoch": 0.33748, + "grad_norm": 1.0096257458152176, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 33748 + }, + { + "epoch": 0.33749, + "grad_norm": 1.5981419952324576, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 33749 + }, + { + "epoch": 0.3375, + "grad_norm": 0.8098451491207329, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 33750 + }, + { + "epoch": 0.33751, + "grad_norm": 0.7985069618173487, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 33751 + }, + { + "epoch": 0.33752, + "grad_norm": 0.9970149616587736, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 33752 + }, + { + "epoch": 0.33753, + "grad_norm": 1.2550520022371214, + "learning_rate": 0.003, + "loss": 4.066, + "step": 33753 + }, + { + "epoch": 0.33754, + "grad_norm": 1.2534749922721797, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 33754 + }, + { + "epoch": 0.33755, + "grad_norm": 0.9182453341202866, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 33755 + }, + { + "epoch": 0.33756, + "grad_norm": 0.9775715639852036, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 33756 + }, + { + "epoch": 0.33757, + "grad_norm": 1.1131591818858033, + "learning_rate": 0.003, + "loss": 4.043, + "step": 33757 + }, + { + "epoch": 0.33758, + "grad_norm": 1.0368674546982244, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 33758 + }, + { + "epoch": 0.33759, + "grad_norm": 0.9986457383478954, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 33759 + }, + { + "epoch": 0.3376, + "grad_norm": 1.2381945797482552, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 33760 + }, + { + "epoch": 0.33761, + "grad_norm": 0.8378189197921854, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 33761 + }, + { + "epoch": 0.33762, + "grad_norm": 1.0193105772760143, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 33762 + }, + { + "epoch": 0.33763, + "grad_norm": 1.4168423488517314, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 33763 + }, + { + "epoch": 0.33764, + "grad_norm": 0.9837950603634447, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 33764 + }, + { + "epoch": 0.33765, + "grad_norm": 1.4391135547689407, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 33765 + }, + { + "epoch": 0.33766, + "grad_norm": 0.9552265433250858, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 33766 + }, + { + "epoch": 0.33767, + "grad_norm": 0.8465630388097051, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 33767 + }, + { + "epoch": 0.33768, + "grad_norm": 1.1973647204779256, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 33768 + }, + { + "epoch": 0.33769, + "grad_norm": 1.3098787067061002, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 33769 + }, + { + "epoch": 0.3377, + "grad_norm": 0.9508547848282117, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 33770 + }, + { + "epoch": 0.33771, + "grad_norm": 1.2504357957136896, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 33771 + }, + { + "epoch": 0.33772, + "grad_norm": 1.0542910576141586, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 33772 + }, + { + "epoch": 0.33773, + "grad_norm": 1.0683596417330634, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 33773 + }, + { + "epoch": 0.33774, + "grad_norm": 1.1589699350342995, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 33774 + }, + { + "epoch": 0.33775, + "grad_norm": 0.8705594484694119, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 33775 + }, + { + "epoch": 0.33776, + "grad_norm": 0.9377681227036608, + "learning_rate": 0.003, + "loss": 4.032, + "step": 33776 + }, + { + "epoch": 0.33777, + "grad_norm": 1.1208996644292735, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 33777 + }, + { + "epoch": 0.33778, + "grad_norm": 1.1742653574108874, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 33778 + }, + { + "epoch": 0.33779, + "grad_norm": 1.060658994663726, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 33779 + }, + { + "epoch": 0.3378, + "grad_norm": 1.1005870821512023, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 33780 + }, + { + "epoch": 0.33781, + "grad_norm": 0.9771271042266725, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 33781 + }, + { + "epoch": 0.33782, + "grad_norm": 0.9782185881761067, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 33782 + }, + { + "epoch": 0.33783, + "grad_norm": 1.0762002094694114, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 33783 + }, + { + "epoch": 0.33784, + "grad_norm": 0.883536597348243, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 33784 + }, + { + "epoch": 0.33785, + "grad_norm": 1.060553216395601, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 33785 + }, + { + "epoch": 0.33786, + "grad_norm": 1.0533707946566828, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 33786 + }, + { + "epoch": 0.33787, + "grad_norm": 1.053896412487799, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 33787 + }, + { + "epoch": 0.33788, + "grad_norm": 1.184910807568745, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 33788 + }, + { + "epoch": 0.33789, + "grad_norm": 1.1991052251173637, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 33789 + }, + { + "epoch": 0.3379, + "grad_norm": 1.124793978467813, + "learning_rate": 0.003, + "loss": 4.034, + "step": 33790 + }, + { + "epoch": 0.33791, + "grad_norm": 1.0383636479498926, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 33791 + }, + { + "epoch": 0.33792, + "grad_norm": 1.3167971289061238, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 33792 + }, + { + "epoch": 0.33793, + "grad_norm": 1.0403334607567483, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 33793 + }, + { + "epoch": 0.33794, + "grad_norm": 1.0823364812212881, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 33794 + }, + { + "epoch": 0.33795, + "grad_norm": 1.1348637896228166, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 33795 + }, + { + "epoch": 0.33796, + "grad_norm": 0.9683564548301012, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 33796 + }, + { + "epoch": 0.33797, + "grad_norm": 0.9848315534775902, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 33797 + }, + { + "epoch": 0.33798, + "grad_norm": 1.2424560273245038, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 33798 + }, + { + "epoch": 0.33799, + "grad_norm": 0.9699380979422515, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 33799 + }, + { + "epoch": 0.338, + "grad_norm": 1.19353243541847, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 33800 + }, + { + "epoch": 0.33801, + "grad_norm": 1.0039476105517822, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 33801 + }, + { + "epoch": 0.33802, + "grad_norm": 1.1561648653246313, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 33802 + }, + { + "epoch": 0.33803, + "grad_norm": 1.1599561242811236, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 33803 + }, + { + "epoch": 0.33804, + "grad_norm": 1.0184307182593049, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 33804 + }, + { + "epoch": 0.33805, + "grad_norm": 1.2340938013472955, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 33805 + }, + { + "epoch": 0.33806, + "grad_norm": 1.093026529167612, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 33806 + }, + { + "epoch": 0.33807, + "grad_norm": 1.1615496171236461, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 33807 + }, + { + "epoch": 0.33808, + "grad_norm": 0.9591407770591525, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 33808 + }, + { + "epoch": 0.33809, + "grad_norm": 1.1572452116267666, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 33809 + }, + { + "epoch": 0.3381, + "grad_norm": 1.1430164898283284, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 33810 + }, + { + "epoch": 0.33811, + "grad_norm": 0.9119961018434574, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 33811 + }, + { + "epoch": 0.33812, + "grad_norm": 0.9527110234938777, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 33812 + }, + { + "epoch": 0.33813, + "grad_norm": 0.9927339472876459, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 33813 + }, + { + "epoch": 0.33814, + "grad_norm": 1.016597399815991, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 33814 + }, + { + "epoch": 0.33815, + "grad_norm": 0.9998456790275937, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 33815 + }, + { + "epoch": 0.33816, + "grad_norm": 0.9950672581236902, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 33816 + }, + { + "epoch": 0.33817, + "grad_norm": 1.2189458497393184, + "learning_rate": 0.003, + "loss": 4.066, + "step": 33817 + }, + { + "epoch": 0.33818, + "grad_norm": 1.0368575377718752, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 33818 + }, + { + "epoch": 0.33819, + "grad_norm": 1.039363641143072, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 33819 + }, + { + "epoch": 0.3382, + "grad_norm": 1.1476998512137953, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 33820 + }, + { + "epoch": 0.33821, + "grad_norm": 0.9994323001167494, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 33821 + }, + { + "epoch": 0.33822, + "grad_norm": 1.2723994365031115, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 33822 + }, + { + "epoch": 0.33823, + "grad_norm": 1.1401010212750848, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 33823 + }, + { + "epoch": 0.33824, + "grad_norm": 1.2470867469874807, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 33824 + }, + { + "epoch": 0.33825, + "grad_norm": 0.9589836745631347, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 33825 + }, + { + "epoch": 0.33826, + "grad_norm": 1.0498449661616247, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 33826 + }, + { + "epoch": 0.33827, + "grad_norm": 1.2719359505239387, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 33827 + }, + { + "epoch": 0.33828, + "grad_norm": 0.9320727799624603, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 33828 + }, + { + "epoch": 0.33829, + "grad_norm": 1.0207247211035282, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 33829 + }, + { + "epoch": 0.3383, + "grad_norm": 1.2147590710379854, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 33830 + }, + { + "epoch": 0.33831, + "grad_norm": 1.133427459699124, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 33831 + }, + { + "epoch": 0.33832, + "grad_norm": 0.959333876341855, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 33832 + }, + { + "epoch": 0.33833, + "grad_norm": 1.0540825017633424, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 33833 + }, + { + "epoch": 0.33834, + "grad_norm": 1.1160884039902106, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 33834 + }, + { + "epoch": 0.33835, + "grad_norm": 0.9354069732224993, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 33835 + }, + { + "epoch": 0.33836, + "grad_norm": 1.0876207186568931, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 33836 + }, + { + "epoch": 0.33837, + "grad_norm": 1.2802116704184208, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 33837 + }, + { + "epoch": 0.33838, + "grad_norm": 0.933284577637626, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 33838 + }, + { + "epoch": 0.33839, + "grad_norm": 1.0238008634891582, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 33839 + }, + { + "epoch": 0.3384, + "grad_norm": 1.2745632447884, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 33840 + }, + { + "epoch": 0.33841, + "grad_norm": 1.1410973362881074, + "learning_rate": 0.003, + "loss": 4.053, + "step": 33841 + }, + { + "epoch": 0.33842, + "grad_norm": 1.2404406239808776, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 33842 + }, + { + "epoch": 0.33843, + "grad_norm": 1.174942240968663, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 33843 + }, + { + "epoch": 0.33844, + "grad_norm": 0.960815500549289, + "learning_rate": 0.003, + "loss": 4.039, + "step": 33844 + }, + { + "epoch": 0.33845, + "grad_norm": 1.2030758620019923, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 33845 + }, + { + "epoch": 0.33846, + "grad_norm": 1.057806719483501, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 33846 + }, + { + "epoch": 0.33847, + "grad_norm": 1.0939200334062422, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 33847 + }, + { + "epoch": 0.33848, + "grad_norm": 1.0283484475824236, + "learning_rate": 0.003, + "loss": 4.03, + "step": 33848 + }, + { + "epoch": 0.33849, + "grad_norm": 1.1374872541233352, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 33849 + }, + { + "epoch": 0.3385, + "grad_norm": 0.7570013343333227, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 33850 + }, + { + "epoch": 0.33851, + "grad_norm": 0.8355542988907526, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 33851 + }, + { + "epoch": 0.33852, + "grad_norm": 1.0076579492659787, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 33852 + }, + { + "epoch": 0.33853, + "grad_norm": 1.260594556208295, + "learning_rate": 0.003, + "loss": 4.049, + "step": 33853 + }, + { + "epoch": 0.33854, + "grad_norm": 1.005905070272558, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 33854 + }, + { + "epoch": 0.33855, + "grad_norm": 1.2503591453377731, + "learning_rate": 0.003, + "loss": 4.047, + "step": 33855 + }, + { + "epoch": 0.33856, + "grad_norm": 0.8439162631764893, + "learning_rate": 0.003, + "loss": 4.014, + "step": 33856 + }, + { + "epoch": 0.33857, + "grad_norm": 0.8349496835115554, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 33857 + }, + { + "epoch": 0.33858, + "grad_norm": 0.9150190717860883, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 33858 + }, + { + "epoch": 0.33859, + "grad_norm": 1.132744484266875, + "learning_rate": 0.003, + "loss": 4.046, + "step": 33859 + }, + { + "epoch": 0.3386, + "grad_norm": 1.3804028008993665, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 33860 + }, + { + "epoch": 0.33861, + "grad_norm": 0.8241634049411336, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 33861 + }, + { + "epoch": 0.33862, + "grad_norm": 0.9239672972771155, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 33862 + }, + { + "epoch": 0.33863, + "grad_norm": 1.2392392208160643, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 33863 + }, + { + "epoch": 0.33864, + "grad_norm": 1.023992047401061, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 33864 + }, + { + "epoch": 0.33865, + "grad_norm": 1.096354904410417, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 33865 + }, + { + "epoch": 0.33866, + "grad_norm": 1.1006235980223036, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 33866 + }, + { + "epoch": 0.33867, + "grad_norm": 1.0590945610519635, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 33867 + }, + { + "epoch": 0.33868, + "grad_norm": 1.176282562670036, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 33868 + }, + { + "epoch": 0.33869, + "grad_norm": 1.1081546088396645, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 33869 + }, + { + "epoch": 0.3387, + "grad_norm": 1.1736119151547857, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 33870 + }, + { + "epoch": 0.33871, + "grad_norm": 0.9072291728834118, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 33871 + }, + { + "epoch": 0.33872, + "grad_norm": 0.9759865289598662, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 33872 + }, + { + "epoch": 0.33873, + "grad_norm": 1.101505302727401, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 33873 + }, + { + "epoch": 0.33874, + "grad_norm": 1.2587007552562415, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 33874 + }, + { + "epoch": 0.33875, + "grad_norm": 0.8388390886933402, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 33875 + }, + { + "epoch": 0.33876, + "grad_norm": 0.9639603371040093, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 33876 + }, + { + "epoch": 0.33877, + "grad_norm": 1.5509525700191522, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 33877 + }, + { + "epoch": 0.33878, + "grad_norm": 1.162914499447136, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 33878 + }, + { + "epoch": 0.33879, + "grad_norm": 1.0598689796400187, + "learning_rate": 0.003, + "loss": 4.056, + "step": 33879 + }, + { + "epoch": 0.3388, + "grad_norm": 1.0226759702468058, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 33880 + }, + { + "epoch": 0.33881, + "grad_norm": 1.2535221540050787, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 33881 + }, + { + "epoch": 0.33882, + "grad_norm": 1.0614822068170675, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 33882 + }, + { + "epoch": 0.33883, + "grad_norm": 1.1363077015946659, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 33883 + }, + { + "epoch": 0.33884, + "grad_norm": 1.1287811743871272, + "learning_rate": 0.003, + "loss": 4.024, + "step": 33884 + }, + { + "epoch": 0.33885, + "grad_norm": 1.0487702521053468, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 33885 + }, + { + "epoch": 0.33886, + "grad_norm": 1.4570061900653255, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 33886 + }, + { + "epoch": 0.33887, + "grad_norm": 0.8157844343500497, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 33887 + }, + { + "epoch": 0.33888, + "grad_norm": 0.8112442358000183, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 33888 + }, + { + "epoch": 0.33889, + "grad_norm": 1.3484326714748338, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 33889 + }, + { + "epoch": 0.3389, + "grad_norm": 0.9977800874530913, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 33890 + }, + { + "epoch": 0.33891, + "grad_norm": 1.2744407308303578, + "learning_rate": 0.003, + "loss": 4.066, + "step": 33891 + }, + { + "epoch": 0.33892, + "grad_norm": 0.9515888032548256, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 33892 + }, + { + "epoch": 0.33893, + "grad_norm": 1.0083531808418447, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 33893 + }, + { + "epoch": 0.33894, + "grad_norm": 1.1943862523357467, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 33894 + }, + { + "epoch": 0.33895, + "grad_norm": 1.0115211633635037, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 33895 + }, + { + "epoch": 0.33896, + "grad_norm": 1.1893920873563635, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 33896 + }, + { + "epoch": 0.33897, + "grad_norm": 1.049313497051183, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 33897 + }, + { + "epoch": 0.33898, + "grad_norm": 1.0422430102512446, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 33898 + }, + { + "epoch": 0.33899, + "grad_norm": 1.0572126774378128, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 33899 + }, + { + "epoch": 0.339, + "grad_norm": 1.150277936769352, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 33900 + }, + { + "epoch": 0.33901, + "grad_norm": 0.8801611378673316, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 33901 + }, + { + "epoch": 0.33902, + "grad_norm": 1.032455142427172, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 33902 + }, + { + "epoch": 0.33903, + "grad_norm": 1.0957165426218027, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 33903 + }, + { + "epoch": 0.33904, + "grad_norm": 0.8642197663161446, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 33904 + }, + { + "epoch": 0.33905, + "grad_norm": 0.9455161391704371, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 33905 + }, + { + "epoch": 0.33906, + "grad_norm": 1.0670808956157565, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 33906 + }, + { + "epoch": 0.33907, + "grad_norm": 1.1518940542100564, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 33907 + }, + { + "epoch": 0.33908, + "grad_norm": 1.4723382205035958, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 33908 + }, + { + "epoch": 0.33909, + "grad_norm": 1.1929998354959843, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 33909 + }, + { + "epoch": 0.3391, + "grad_norm": 1.0775075489922734, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 33910 + }, + { + "epoch": 0.33911, + "grad_norm": 1.0459133067225506, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 33911 + }, + { + "epoch": 0.33912, + "grad_norm": 1.0058111988739544, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 33912 + }, + { + "epoch": 0.33913, + "grad_norm": 1.0290564668162137, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 33913 + }, + { + "epoch": 0.33914, + "grad_norm": 1.24159611915555, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 33914 + }, + { + "epoch": 0.33915, + "grad_norm": 0.8979278487089998, + "learning_rate": 0.003, + "loss": 4.029, + "step": 33915 + }, + { + "epoch": 0.33916, + "grad_norm": 1.130041402166701, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 33916 + }, + { + "epoch": 0.33917, + "grad_norm": 1.0414727667357289, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 33917 + }, + { + "epoch": 0.33918, + "grad_norm": 1.0876697028861324, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 33918 + }, + { + "epoch": 0.33919, + "grad_norm": 1.1969905688592655, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 33919 + }, + { + "epoch": 0.3392, + "grad_norm": 1.1431170961901795, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 33920 + }, + { + "epoch": 0.33921, + "grad_norm": 1.0204398557855967, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 33921 + }, + { + "epoch": 0.33922, + "grad_norm": 1.1288377823718485, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 33922 + }, + { + "epoch": 0.33923, + "grad_norm": 1.0431420071605866, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 33923 + }, + { + "epoch": 0.33924, + "grad_norm": 1.2572658800711352, + "learning_rate": 0.003, + "loss": 4.076, + "step": 33924 + }, + { + "epoch": 0.33925, + "grad_norm": 1.0043964127884177, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 33925 + }, + { + "epoch": 0.33926, + "grad_norm": 1.0684206142806913, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 33926 + }, + { + "epoch": 0.33927, + "grad_norm": 1.1896933551971234, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 33927 + }, + { + "epoch": 0.33928, + "grad_norm": 1.2666044428497678, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 33928 + }, + { + "epoch": 0.33929, + "grad_norm": 0.9628337013696862, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 33929 + }, + { + "epoch": 0.3393, + "grad_norm": 0.9896581856304671, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 33930 + }, + { + "epoch": 0.33931, + "grad_norm": 1.0921742533310208, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 33931 + }, + { + "epoch": 0.33932, + "grad_norm": 1.0862662824879834, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 33932 + }, + { + "epoch": 0.33933, + "grad_norm": 1.1969071387070314, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 33933 + }, + { + "epoch": 0.33934, + "grad_norm": 0.9384278085589371, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 33934 + }, + { + "epoch": 0.33935, + "grad_norm": 1.2323121803099149, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 33935 + }, + { + "epoch": 0.33936, + "grad_norm": 1.1082651671558201, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 33936 + }, + { + "epoch": 0.33937, + "grad_norm": 1.0037903425772772, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 33937 + }, + { + "epoch": 0.33938, + "grad_norm": 1.0943239721748919, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 33938 + }, + { + "epoch": 0.33939, + "grad_norm": 1.0213800261488908, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 33939 + }, + { + "epoch": 0.3394, + "grad_norm": 0.8587020017471257, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 33940 + }, + { + "epoch": 0.33941, + "grad_norm": 0.9724394622256191, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 33941 + }, + { + "epoch": 0.33942, + "grad_norm": 1.375521584073671, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 33942 + }, + { + "epoch": 0.33943, + "grad_norm": 1.0140043159882484, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 33943 + }, + { + "epoch": 0.33944, + "grad_norm": 1.2011662496726374, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 33944 + }, + { + "epoch": 0.33945, + "grad_norm": 0.8602925876157633, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 33945 + }, + { + "epoch": 0.33946, + "grad_norm": 0.9450900573897222, + "learning_rate": 0.003, + "loss": 4.058, + "step": 33946 + }, + { + "epoch": 0.33947, + "grad_norm": 1.383703642798844, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 33947 + }, + { + "epoch": 0.33948, + "grad_norm": 0.9346402540161012, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 33948 + }, + { + "epoch": 0.33949, + "grad_norm": 0.964967583574119, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 33949 + }, + { + "epoch": 0.3395, + "grad_norm": 1.07787982999581, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 33950 + }, + { + "epoch": 0.33951, + "grad_norm": 1.0516457575184268, + "learning_rate": 0.003, + "loss": 4.062, + "step": 33951 + }, + { + "epoch": 0.33952, + "grad_norm": 1.3195627579514406, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 33952 + }, + { + "epoch": 0.33953, + "grad_norm": 1.0183721790544757, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 33953 + }, + { + "epoch": 0.33954, + "grad_norm": 1.1272340913579353, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 33954 + }, + { + "epoch": 0.33955, + "grad_norm": 1.0923507805069361, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 33955 + }, + { + "epoch": 0.33956, + "grad_norm": 1.0893325268198393, + "learning_rate": 0.003, + "loss": 4.014, + "step": 33956 + }, + { + "epoch": 0.33957, + "grad_norm": 1.3244318296284476, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 33957 + }, + { + "epoch": 0.33958, + "grad_norm": 0.9961833237806981, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 33958 + }, + { + "epoch": 0.33959, + "grad_norm": 1.271078568299366, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 33959 + }, + { + "epoch": 0.3396, + "grad_norm": 0.9858704602079096, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 33960 + }, + { + "epoch": 0.33961, + "grad_norm": 0.9378907888453093, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 33961 + }, + { + "epoch": 0.33962, + "grad_norm": 0.9710358480040545, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 33962 + }, + { + "epoch": 0.33963, + "grad_norm": 1.1762995144729245, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 33963 + }, + { + "epoch": 0.33964, + "grad_norm": 1.1639307193797876, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 33964 + }, + { + "epoch": 0.33965, + "grad_norm": 1.1926375357343906, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 33965 + }, + { + "epoch": 0.33966, + "grad_norm": 1.0616328752537036, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 33966 + }, + { + "epoch": 0.33967, + "grad_norm": 0.9990015778533008, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 33967 + }, + { + "epoch": 0.33968, + "grad_norm": 1.222022891757408, + "learning_rate": 0.003, + "loss": 4.048, + "step": 33968 + }, + { + "epoch": 0.33969, + "grad_norm": 1.04740524655996, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 33969 + }, + { + "epoch": 0.3397, + "grad_norm": 1.3393587901687276, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 33970 + }, + { + "epoch": 0.33971, + "grad_norm": 0.8912367594809687, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 33971 + }, + { + "epoch": 0.33972, + "grad_norm": 0.9406432590488256, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 33972 + }, + { + "epoch": 0.33973, + "grad_norm": 1.2151707345074039, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 33973 + }, + { + "epoch": 0.33974, + "grad_norm": 1.0204880927566378, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 33974 + }, + { + "epoch": 0.33975, + "grad_norm": 1.008931004004869, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 33975 + }, + { + "epoch": 0.33976, + "grad_norm": 1.098209993633499, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 33976 + }, + { + "epoch": 0.33977, + "grad_norm": 0.9666419883426326, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 33977 + }, + { + "epoch": 0.33978, + "grad_norm": 1.2872418748889942, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 33978 + }, + { + "epoch": 0.33979, + "grad_norm": 1.0152967672955304, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 33979 + }, + { + "epoch": 0.3398, + "grad_norm": 0.9616403863963165, + "learning_rate": 0.003, + "loss": 4.025, + "step": 33980 + }, + { + "epoch": 0.33981, + "grad_norm": 0.8482729528191202, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 33981 + }, + { + "epoch": 0.33982, + "grad_norm": 0.8859565122120978, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 33982 + }, + { + "epoch": 0.33983, + "grad_norm": 0.8878861023401772, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 33983 + }, + { + "epoch": 0.33984, + "grad_norm": 1.0077371272248146, + "learning_rate": 0.003, + "loss": 4.036, + "step": 33984 + }, + { + "epoch": 0.33985, + "grad_norm": 1.1702622758320864, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 33985 + }, + { + "epoch": 0.33986, + "grad_norm": 1.1659476277129877, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 33986 + }, + { + "epoch": 0.33987, + "grad_norm": 1.0043658981623131, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 33987 + }, + { + "epoch": 0.33988, + "grad_norm": 1.02040941300848, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 33988 + }, + { + "epoch": 0.33989, + "grad_norm": 1.1276138130320321, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 33989 + }, + { + "epoch": 0.3399, + "grad_norm": 1.3254601195429105, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 33990 + }, + { + "epoch": 0.33991, + "grad_norm": 1.039844292606068, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 33991 + }, + { + "epoch": 0.33992, + "grad_norm": 1.378834730487504, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 33992 + }, + { + "epoch": 0.33993, + "grad_norm": 0.8479137597841141, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 33993 + }, + { + "epoch": 0.33994, + "grad_norm": 0.8511490794203215, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 33994 + }, + { + "epoch": 0.33995, + "grad_norm": 1.0371253283053181, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 33995 + }, + { + "epoch": 0.33996, + "grad_norm": 1.3744443027769924, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 33996 + }, + { + "epoch": 0.33997, + "grad_norm": 0.9075172101946021, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 33997 + }, + { + "epoch": 0.33998, + "grad_norm": 1.0496197713178212, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 33998 + }, + { + "epoch": 0.33999, + "grad_norm": 1.2417541325889798, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 33999 + }, + { + "epoch": 0.34, + "grad_norm": 1.1566722631661213, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 34000 + }, + { + "epoch": 0.34001, + "grad_norm": 1.0909780737119317, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 34001 + }, + { + "epoch": 0.34002, + "grad_norm": 1.0341627386989518, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 34002 + }, + { + "epoch": 0.34003, + "grad_norm": 1.1436342221042695, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 34003 + }, + { + "epoch": 0.34004, + "grad_norm": 1.0282097194814768, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 34004 + }, + { + "epoch": 0.34005, + "grad_norm": 1.2306460511235746, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 34005 + }, + { + "epoch": 0.34006, + "grad_norm": 0.9724826181188113, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 34006 + }, + { + "epoch": 0.34007, + "grad_norm": 1.1347512065171155, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 34007 + }, + { + "epoch": 0.34008, + "grad_norm": 1.1553862081336626, + "learning_rate": 0.003, + "loss": 4.022, + "step": 34008 + }, + { + "epoch": 0.34009, + "grad_norm": 1.00586441834169, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 34009 + }, + { + "epoch": 0.3401, + "grad_norm": 1.2394077260563274, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 34010 + }, + { + "epoch": 0.34011, + "grad_norm": 0.9633392187706981, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 34011 + }, + { + "epoch": 0.34012, + "grad_norm": 1.0412452775857235, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 34012 + }, + { + "epoch": 0.34013, + "grad_norm": 1.177690629830753, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 34013 + }, + { + "epoch": 0.34014, + "grad_norm": 0.8710123418591585, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 34014 + }, + { + "epoch": 0.34015, + "grad_norm": 0.9011591043219936, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 34015 + }, + { + "epoch": 0.34016, + "grad_norm": 1.1096588267264655, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 34016 + }, + { + "epoch": 0.34017, + "grad_norm": 1.1850123526833003, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 34017 + }, + { + "epoch": 0.34018, + "grad_norm": 1.1023187330505484, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 34018 + }, + { + "epoch": 0.34019, + "grad_norm": 1.2459696572650132, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 34019 + }, + { + "epoch": 0.3402, + "grad_norm": 1.2329780392680882, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 34020 + }, + { + "epoch": 0.34021, + "grad_norm": 1.0030811233568864, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 34021 + }, + { + "epoch": 0.34022, + "grad_norm": 1.0717637275243002, + "learning_rate": 0.003, + "loss": 4.0871, + "step": 34022 + }, + { + "epoch": 0.34023, + "grad_norm": 0.9495729174198266, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 34023 + }, + { + "epoch": 0.34024, + "grad_norm": 1.1702248861026225, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 34024 + }, + { + "epoch": 0.34025, + "grad_norm": 1.1231648587195069, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 34025 + }, + { + "epoch": 0.34026, + "grad_norm": 1.1899629223035162, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 34026 + }, + { + "epoch": 0.34027, + "grad_norm": 0.8560131419802525, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 34027 + }, + { + "epoch": 0.34028, + "grad_norm": 0.8400707209347874, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 34028 + }, + { + "epoch": 0.34029, + "grad_norm": 0.9241108745578459, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 34029 + }, + { + "epoch": 0.3403, + "grad_norm": 1.0399149490765047, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 34030 + }, + { + "epoch": 0.34031, + "grad_norm": 1.111755024017703, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 34031 + }, + { + "epoch": 0.34032, + "grad_norm": 1.0770695760130158, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 34032 + }, + { + "epoch": 0.34033, + "grad_norm": 1.0311936635690555, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 34033 + }, + { + "epoch": 0.34034, + "grad_norm": 0.8832134789285913, + "learning_rate": 0.003, + "loss": 4.047, + "step": 34034 + }, + { + "epoch": 0.34035, + "grad_norm": 0.9661783451105227, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 34035 + }, + { + "epoch": 0.34036, + "grad_norm": 1.3336297880548929, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 34036 + }, + { + "epoch": 0.34037, + "grad_norm": 1.2368512652818322, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 34037 + }, + { + "epoch": 0.34038, + "grad_norm": 1.0722061260603286, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 34038 + }, + { + "epoch": 0.34039, + "grad_norm": 0.9754841349935764, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 34039 + }, + { + "epoch": 0.3404, + "grad_norm": 1.0562331512527254, + "learning_rate": 0.003, + "loss": 4.033, + "step": 34040 + }, + { + "epoch": 0.34041, + "grad_norm": 1.1120734738495344, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 34041 + }, + { + "epoch": 0.34042, + "grad_norm": 1.1066687681906926, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 34042 + }, + { + "epoch": 0.34043, + "grad_norm": 1.3576977129370238, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 34043 + }, + { + "epoch": 0.34044, + "grad_norm": 0.9197380926505477, + "learning_rate": 0.003, + "loss": 4.071, + "step": 34044 + }, + { + "epoch": 0.34045, + "grad_norm": 1.1211813326770694, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 34045 + }, + { + "epoch": 0.34046, + "grad_norm": 1.2667397449275262, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 34046 + }, + { + "epoch": 0.34047, + "grad_norm": 1.18499247253768, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 34047 + }, + { + "epoch": 0.34048, + "grad_norm": 0.9761929636807514, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 34048 + }, + { + "epoch": 0.34049, + "grad_norm": 1.1152090613683956, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 34049 + }, + { + "epoch": 0.3405, + "grad_norm": 1.111877800925453, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 34050 + }, + { + "epoch": 0.34051, + "grad_norm": 1.1538464981540595, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 34051 + }, + { + "epoch": 0.34052, + "grad_norm": 1.1337428874937912, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 34052 + }, + { + "epoch": 0.34053, + "grad_norm": 1.343777845381307, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 34053 + }, + { + "epoch": 0.34054, + "grad_norm": 0.8892147934102594, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 34054 + }, + { + "epoch": 0.34055, + "grad_norm": 0.918925178778528, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 34055 + }, + { + "epoch": 0.34056, + "grad_norm": 1.1199838142395346, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 34056 + }, + { + "epoch": 0.34057, + "grad_norm": 1.030714686810909, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 34057 + }, + { + "epoch": 0.34058, + "grad_norm": 1.2999146805047217, + "learning_rate": 0.003, + "loss": 4.053, + "step": 34058 + }, + { + "epoch": 0.34059, + "grad_norm": 0.9464208174279344, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 34059 + }, + { + "epoch": 0.3406, + "grad_norm": 1.0532349136374193, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 34060 + }, + { + "epoch": 0.34061, + "grad_norm": 0.7872890319553634, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 34061 + }, + { + "epoch": 0.34062, + "grad_norm": 0.9977727055423192, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 34062 + }, + { + "epoch": 0.34063, + "grad_norm": 1.1118904075473492, + "learning_rate": 0.003, + "loss": 4.059, + "step": 34063 + }, + { + "epoch": 0.34064, + "grad_norm": 0.8597309602430572, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 34064 + }, + { + "epoch": 0.34065, + "grad_norm": 1.0662103848309998, + "learning_rate": 0.003, + "loss": 4.058, + "step": 34065 + }, + { + "epoch": 0.34066, + "grad_norm": 1.014579237166296, + "learning_rate": 0.003, + "loss": 4.036, + "step": 34066 + }, + { + "epoch": 0.34067, + "grad_norm": 1.2876346873019142, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 34067 + }, + { + "epoch": 0.34068, + "grad_norm": 1.116203485155106, + "learning_rate": 0.003, + "loss": 4.023, + "step": 34068 + }, + { + "epoch": 0.34069, + "grad_norm": 1.0984066375148451, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 34069 + }, + { + "epoch": 0.3407, + "grad_norm": 1.1854816574755596, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 34070 + }, + { + "epoch": 0.34071, + "grad_norm": 1.2322948395316555, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 34071 + }, + { + "epoch": 0.34072, + "grad_norm": 1.0319871086023091, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 34072 + }, + { + "epoch": 0.34073, + "grad_norm": 1.1959067142722446, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 34073 + }, + { + "epoch": 0.34074, + "grad_norm": 0.9131166930222002, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 34074 + }, + { + "epoch": 0.34075, + "grad_norm": 0.8946311427551017, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 34075 + }, + { + "epoch": 0.34076, + "grad_norm": 1.050041839589098, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 34076 + }, + { + "epoch": 0.34077, + "grad_norm": 1.08345724501189, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 34077 + }, + { + "epoch": 0.34078, + "grad_norm": 1.041775051186548, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 34078 + }, + { + "epoch": 0.34079, + "grad_norm": 1.0935617094391639, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 34079 + }, + { + "epoch": 0.3408, + "grad_norm": 1.0565981707680832, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 34080 + }, + { + "epoch": 0.34081, + "grad_norm": 1.074557029047314, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 34081 + }, + { + "epoch": 0.34082, + "grad_norm": 0.9954584221730496, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 34082 + }, + { + "epoch": 0.34083, + "grad_norm": 1.0521539974057887, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 34083 + }, + { + "epoch": 0.34084, + "grad_norm": 1.0656292702831702, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 34084 + }, + { + "epoch": 0.34085, + "grad_norm": 1.2379245659314742, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 34085 + }, + { + "epoch": 0.34086, + "grad_norm": 1.01564868614807, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 34086 + }, + { + "epoch": 0.34087, + "grad_norm": 1.1691385121283584, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 34087 + }, + { + "epoch": 0.34088, + "grad_norm": 0.9530075700788229, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 34088 + }, + { + "epoch": 0.34089, + "grad_norm": 1.2452851216365055, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 34089 + }, + { + "epoch": 0.3409, + "grad_norm": 1.1258594887840085, + "learning_rate": 0.003, + "loss": 4.032, + "step": 34090 + }, + { + "epoch": 0.34091, + "grad_norm": 1.3499375838616077, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 34091 + }, + { + "epoch": 0.34092, + "grad_norm": 1.1738124063649853, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 34092 + }, + { + "epoch": 0.34093, + "grad_norm": 0.9956630251627508, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 34093 + }, + { + "epoch": 0.34094, + "grad_norm": 0.9533129520440055, + "learning_rate": 0.003, + "loss": 4.052, + "step": 34094 + }, + { + "epoch": 0.34095, + "grad_norm": 1.069355999335036, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 34095 + }, + { + "epoch": 0.34096, + "grad_norm": 1.2305370943359526, + "learning_rate": 0.003, + "loss": 4.041, + "step": 34096 + }, + { + "epoch": 0.34097, + "grad_norm": 1.1104682325850497, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 34097 + }, + { + "epoch": 0.34098, + "grad_norm": 1.337395575875309, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 34098 + }, + { + "epoch": 0.34099, + "grad_norm": 0.9316902452191907, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 34099 + }, + { + "epoch": 0.341, + "grad_norm": 0.8326546547522677, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 34100 + }, + { + "epoch": 0.34101, + "grad_norm": 0.8237104791791329, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 34101 + }, + { + "epoch": 0.34102, + "grad_norm": 0.920477053457992, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 34102 + }, + { + "epoch": 0.34103, + "grad_norm": 1.3118449869539335, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 34103 + }, + { + "epoch": 0.34104, + "grad_norm": 1.1533501899993843, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 34104 + }, + { + "epoch": 0.34105, + "grad_norm": 1.1097068796470406, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 34105 + }, + { + "epoch": 0.34106, + "grad_norm": 1.1082220160778316, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 34106 + }, + { + "epoch": 0.34107, + "grad_norm": 1.337864014188281, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 34107 + }, + { + "epoch": 0.34108, + "grad_norm": 1.0600327742221443, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 34108 + }, + { + "epoch": 0.34109, + "grad_norm": 1.0451573580096818, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 34109 + }, + { + "epoch": 0.3411, + "grad_norm": 0.9774859221363222, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 34110 + }, + { + "epoch": 0.34111, + "grad_norm": 1.0211685546832092, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 34111 + }, + { + "epoch": 0.34112, + "grad_norm": 1.0301951333457466, + "learning_rate": 0.003, + "loss": 4.049, + "step": 34112 + }, + { + "epoch": 0.34113, + "grad_norm": 1.1876360915142836, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 34113 + }, + { + "epoch": 0.34114, + "grad_norm": 1.0340330509043147, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 34114 + }, + { + "epoch": 0.34115, + "grad_norm": 1.2290771140881707, + "learning_rate": 0.003, + "loss": 4.037, + "step": 34115 + }, + { + "epoch": 0.34116, + "grad_norm": 1.0062003296700477, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 34116 + }, + { + "epoch": 0.34117, + "grad_norm": 1.1373907540172246, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 34117 + }, + { + "epoch": 0.34118, + "grad_norm": 0.9455668112106291, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 34118 + }, + { + "epoch": 0.34119, + "grad_norm": 1.066041493554447, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 34119 + }, + { + "epoch": 0.3412, + "grad_norm": 1.0795027666098265, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 34120 + }, + { + "epoch": 0.34121, + "grad_norm": 1.2494776085096406, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 34121 + }, + { + "epoch": 0.34122, + "grad_norm": 0.9718752532377073, + "learning_rate": 0.003, + "loss": 4.038, + "step": 34122 + }, + { + "epoch": 0.34123, + "grad_norm": 0.9283029468877221, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 34123 + }, + { + "epoch": 0.34124, + "grad_norm": 0.9646069290684527, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 34124 + }, + { + "epoch": 0.34125, + "grad_norm": 1.2035718639662196, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 34125 + }, + { + "epoch": 0.34126, + "grad_norm": 1.1841785924894326, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 34126 + }, + { + "epoch": 0.34127, + "grad_norm": 1.1672946444878398, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 34127 + }, + { + "epoch": 0.34128, + "grad_norm": 0.97201426102269, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 34128 + }, + { + "epoch": 0.34129, + "grad_norm": 1.0601208789489875, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 34129 + }, + { + "epoch": 0.3413, + "grad_norm": 1.2350262240508816, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 34130 + }, + { + "epoch": 0.34131, + "grad_norm": 1.023419831055957, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 34131 + }, + { + "epoch": 0.34132, + "grad_norm": 0.9681836078193599, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 34132 + }, + { + "epoch": 0.34133, + "grad_norm": 1.1619885256362632, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 34133 + }, + { + "epoch": 0.34134, + "grad_norm": 1.0373668498837676, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 34134 + }, + { + "epoch": 0.34135, + "grad_norm": 1.2436692821380795, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 34135 + }, + { + "epoch": 0.34136, + "grad_norm": 1.039229984650176, + "learning_rate": 0.003, + "loss": 4.045, + "step": 34136 + }, + { + "epoch": 0.34137, + "grad_norm": 0.9491512137728171, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 34137 + }, + { + "epoch": 0.34138, + "grad_norm": 1.012852496777197, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 34138 + }, + { + "epoch": 0.34139, + "grad_norm": 1.2691953946257735, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 34139 + }, + { + "epoch": 0.3414, + "grad_norm": 1.1498790968697712, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 34140 + }, + { + "epoch": 0.34141, + "grad_norm": 0.9311236050379476, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 34141 + }, + { + "epoch": 0.34142, + "grad_norm": 1.0714471938406696, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 34142 + }, + { + "epoch": 0.34143, + "grad_norm": 1.279697818977371, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 34143 + }, + { + "epoch": 0.34144, + "grad_norm": 1.0633601844720806, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 34144 + }, + { + "epoch": 0.34145, + "grad_norm": 1.2351068361281028, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 34145 + }, + { + "epoch": 0.34146, + "grad_norm": 0.8282156885011412, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 34146 + }, + { + "epoch": 0.34147, + "grad_norm": 0.8521067676073478, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 34147 + }, + { + "epoch": 0.34148, + "grad_norm": 1.1048623872135186, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 34148 + }, + { + "epoch": 0.34149, + "grad_norm": 1.3241383966037163, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 34149 + }, + { + "epoch": 0.3415, + "grad_norm": 1.1720660083004197, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 34150 + }, + { + "epoch": 0.34151, + "grad_norm": 1.0414751459315863, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 34151 + }, + { + "epoch": 0.34152, + "grad_norm": 1.0403159668883895, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 34152 + }, + { + "epoch": 0.34153, + "grad_norm": 1.192986586790585, + "learning_rate": 0.003, + "loss": 4.017, + "step": 34153 + }, + { + "epoch": 0.34154, + "grad_norm": 1.1063530562522954, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 34154 + }, + { + "epoch": 0.34155, + "grad_norm": 1.084822971772218, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 34155 + }, + { + "epoch": 0.34156, + "grad_norm": 1.3306389865742856, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 34156 + }, + { + "epoch": 0.34157, + "grad_norm": 0.8096151269653733, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 34157 + }, + { + "epoch": 0.34158, + "grad_norm": 0.8411692419562977, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 34158 + }, + { + "epoch": 0.34159, + "grad_norm": 0.8629252008931224, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 34159 + }, + { + "epoch": 0.3416, + "grad_norm": 0.9781488397989982, + "learning_rate": 0.003, + "loss": 4.033, + "step": 34160 + }, + { + "epoch": 0.34161, + "grad_norm": 1.274840117424837, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 34161 + }, + { + "epoch": 0.34162, + "grad_norm": 1.0952363127696392, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 34162 + }, + { + "epoch": 0.34163, + "grad_norm": 0.996139838796033, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 34163 + }, + { + "epoch": 0.34164, + "grad_norm": 1.1958247257212145, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 34164 + }, + { + "epoch": 0.34165, + "grad_norm": 0.9988222816696795, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 34165 + }, + { + "epoch": 0.34166, + "grad_norm": 0.9926182854844572, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 34166 + }, + { + "epoch": 0.34167, + "grad_norm": 1.0265898472042059, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 34167 + }, + { + "epoch": 0.34168, + "grad_norm": 1.1698172302546825, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 34168 + }, + { + "epoch": 0.34169, + "grad_norm": 1.032876536231665, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 34169 + }, + { + "epoch": 0.3417, + "grad_norm": 1.1251649785342264, + "learning_rate": 0.003, + "loss": 4.063, + "step": 34170 + }, + { + "epoch": 0.34171, + "grad_norm": 0.9831917638357686, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 34171 + }, + { + "epoch": 0.34172, + "grad_norm": 1.1225527252163472, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 34172 + }, + { + "epoch": 0.34173, + "grad_norm": 1.1331846898857638, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 34173 + }, + { + "epoch": 0.34174, + "grad_norm": 1.0465721603253044, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 34174 + }, + { + "epoch": 0.34175, + "grad_norm": 1.1971662444535283, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 34175 + }, + { + "epoch": 0.34176, + "grad_norm": 0.955988427917266, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 34176 + }, + { + "epoch": 0.34177, + "grad_norm": 1.0178900379204814, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 34177 + }, + { + "epoch": 0.34178, + "grad_norm": 1.1810566094187465, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 34178 + }, + { + "epoch": 0.34179, + "grad_norm": 0.9881862038465359, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 34179 + }, + { + "epoch": 0.3418, + "grad_norm": 1.1414528510875124, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 34180 + }, + { + "epoch": 0.34181, + "grad_norm": 1.0418792912316448, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 34181 + }, + { + "epoch": 0.34182, + "grad_norm": 1.1026626571636617, + "learning_rate": 0.003, + "loss": 4.04, + "step": 34182 + }, + { + "epoch": 0.34183, + "grad_norm": 1.0219172921116204, + "learning_rate": 0.003, + "loss": 4.064, + "step": 34183 + }, + { + "epoch": 0.34184, + "grad_norm": 1.29984606959971, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 34184 + }, + { + "epoch": 0.34185, + "grad_norm": 0.9686203625466211, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 34185 + }, + { + "epoch": 0.34186, + "grad_norm": 1.292446687791519, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 34186 + }, + { + "epoch": 0.34187, + "grad_norm": 1.1888856949107318, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 34187 + }, + { + "epoch": 0.34188, + "grad_norm": 1.1066832884522786, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 34188 + }, + { + "epoch": 0.34189, + "grad_norm": 1.1090676979862222, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 34189 + }, + { + "epoch": 0.3419, + "grad_norm": 1.2717103817746185, + "learning_rate": 0.003, + "loss": 4.053, + "step": 34190 + }, + { + "epoch": 0.34191, + "grad_norm": 1.220865378499801, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 34191 + }, + { + "epoch": 0.34192, + "grad_norm": 1.1281406387224522, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 34192 + }, + { + "epoch": 0.34193, + "grad_norm": 1.270810120076017, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 34193 + }, + { + "epoch": 0.34194, + "grad_norm": 1.0622971888525377, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 34194 + }, + { + "epoch": 0.34195, + "grad_norm": 1.1569751618134034, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 34195 + }, + { + "epoch": 0.34196, + "grad_norm": 1.017982430466405, + "learning_rate": 0.003, + "loss": 4.026, + "step": 34196 + }, + { + "epoch": 0.34197, + "grad_norm": 1.349794231323246, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 34197 + }, + { + "epoch": 0.34198, + "grad_norm": 0.8979210414591274, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 34198 + }, + { + "epoch": 0.34199, + "grad_norm": 1.0210535939302223, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 34199 + }, + { + "epoch": 0.342, + "grad_norm": 1.2931158779825418, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 34200 + }, + { + "epoch": 0.34201, + "grad_norm": 0.9155120593018996, + "learning_rate": 0.003, + "loss": 4.053, + "step": 34201 + }, + { + "epoch": 0.34202, + "grad_norm": 1.2036164114423555, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 34202 + }, + { + "epoch": 0.34203, + "grad_norm": 1.208675830671688, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 34203 + }, + { + "epoch": 0.34204, + "grad_norm": 0.8703665595190949, + "learning_rate": 0.003, + "loss": 4.047, + "step": 34204 + }, + { + "epoch": 0.34205, + "grad_norm": 0.9275336035316256, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 34205 + }, + { + "epoch": 0.34206, + "grad_norm": 1.0808324043627635, + "learning_rate": 0.003, + "loss": 4.058, + "step": 34206 + }, + { + "epoch": 0.34207, + "grad_norm": 1.01272032434139, + "learning_rate": 0.003, + "loss": 4.035, + "step": 34207 + }, + { + "epoch": 0.34208, + "grad_norm": 1.0510696889662885, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 34208 + }, + { + "epoch": 0.34209, + "grad_norm": 1.0827528031823692, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 34209 + }, + { + "epoch": 0.3421, + "grad_norm": 1.022122084017462, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 34210 + }, + { + "epoch": 0.34211, + "grad_norm": 1.0681231453868893, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 34211 + }, + { + "epoch": 0.34212, + "grad_norm": 1.2358489143037077, + "learning_rate": 0.003, + "loss": 4.038, + "step": 34212 + }, + { + "epoch": 0.34213, + "grad_norm": 0.9530222492923599, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 34213 + }, + { + "epoch": 0.34214, + "grad_norm": 1.0265679361099778, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 34214 + }, + { + "epoch": 0.34215, + "grad_norm": 1.4455028430215777, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 34215 + }, + { + "epoch": 0.34216, + "grad_norm": 1.181876275740678, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 34216 + }, + { + "epoch": 0.34217, + "grad_norm": 1.0708966701155558, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 34217 + }, + { + "epoch": 0.34218, + "grad_norm": 0.991982138234559, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 34218 + }, + { + "epoch": 0.34219, + "grad_norm": 1.0695076469313551, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 34219 + }, + { + "epoch": 0.3422, + "grad_norm": 1.2146841395066033, + "learning_rate": 0.003, + "loss": 4.051, + "step": 34220 + }, + { + "epoch": 0.34221, + "grad_norm": 0.8940996703829103, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 34221 + }, + { + "epoch": 0.34222, + "grad_norm": 1.0249977881549597, + "learning_rate": 0.003, + "loss": 4.034, + "step": 34222 + }, + { + "epoch": 0.34223, + "grad_norm": 1.2645435615555072, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 34223 + }, + { + "epoch": 0.34224, + "grad_norm": 0.795724322557917, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 34224 + }, + { + "epoch": 0.34225, + "grad_norm": 0.9311719752162445, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 34225 + }, + { + "epoch": 0.34226, + "grad_norm": 0.9371670035852077, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 34226 + }, + { + "epoch": 0.34227, + "grad_norm": 0.8969544984571669, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 34227 + }, + { + "epoch": 0.34228, + "grad_norm": 1.1198163859898318, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 34228 + }, + { + "epoch": 0.34229, + "grad_norm": 1.1400181131142453, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 34229 + }, + { + "epoch": 0.3423, + "grad_norm": 1.1244240000381254, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 34230 + }, + { + "epoch": 0.34231, + "grad_norm": 0.9747710965413843, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 34231 + }, + { + "epoch": 0.34232, + "grad_norm": 1.1127570373426718, + "learning_rate": 0.003, + "loss": 4.049, + "step": 34232 + }, + { + "epoch": 0.34233, + "grad_norm": 1.007657596130511, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 34233 + }, + { + "epoch": 0.34234, + "grad_norm": 1.2850163901345744, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 34234 + }, + { + "epoch": 0.34235, + "grad_norm": 0.8446694295693635, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 34235 + }, + { + "epoch": 0.34236, + "grad_norm": 1.0867399354521954, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 34236 + }, + { + "epoch": 0.34237, + "grad_norm": 1.6505098547332764, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 34237 + }, + { + "epoch": 0.34238, + "grad_norm": 0.9877916296149962, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 34238 + }, + { + "epoch": 0.34239, + "grad_norm": 1.078106776575685, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 34239 + }, + { + "epoch": 0.3424, + "grad_norm": 1.0020921703073142, + "learning_rate": 0.003, + "loss": 4.033, + "step": 34240 + }, + { + "epoch": 0.34241, + "grad_norm": 1.2766857254986077, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 34241 + }, + { + "epoch": 0.34242, + "grad_norm": 1.0611707713649894, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 34242 + }, + { + "epoch": 0.34243, + "grad_norm": 1.2657852396873808, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 34243 + }, + { + "epoch": 0.34244, + "grad_norm": 1.0652640137071607, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 34244 + }, + { + "epoch": 0.34245, + "grad_norm": 0.8996606717194064, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 34245 + }, + { + "epoch": 0.34246, + "grad_norm": 0.9557119946234675, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 34246 + }, + { + "epoch": 0.34247, + "grad_norm": 1.102943617878365, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 34247 + }, + { + "epoch": 0.34248, + "grad_norm": 1.1652219413588591, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 34248 + }, + { + "epoch": 0.34249, + "grad_norm": 1.147102140384689, + "learning_rate": 0.003, + "loss": 4.051, + "step": 34249 + }, + { + "epoch": 0.3425, + "grad_norm": 0.9404631886712789, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 34250 + }, + { + "epoch": 0.34251, + "grad_norm": 0.8666095806954309, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 34251 + }, + { + "epoch": 0.34252, + "grad_norm": 0.9584479022577769, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 34252 + }, + { + "epoch": 0.34253, + "grad_norm": 1.2471769115682108, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 34253 + }, + { + "epoch": 0.34254, + "grad_norm": 0.8777778846068609, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 34254 + }, + { + "epoch": 0.34255, + "grad_norm": 0.7862813858560832, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 34255 + }, + { + "epoch": 0.34256, + "grad_norm": 0.9759223850783387, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 34256 + }, + { + "epoch": 0.34257, + "grad_norm": 1.3671335350921274, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 34257 + }, + { + "epoch": 0.34258, + "grad_norm": 1.123100998058137, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 34258 + }, + { + "epoch": 0.34259, + "grad_norm": 0.8602578846160382, + "learning_rate": 0.003, + "loss": 4.05, + "step": 34259 + }, + { + "epoch": 0.3426, + "grad_norm": 1.0163578858293154, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 34260 + }, + { + "epoch": 0.34261, + "grad_norm": 1.1780123547253492, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 34261 + }, + { + "epoch": 0.34262, + "grad_norm": 0.9970575845823726, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 34262 + }, + { + "epoch": 0.34263, + "grad_norm": 1.6150069228098378, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 34263 + }, + { + "epoch": 0.34264, + "grad_norm": 0.9621408198279282, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 34264 + }, + { + "epoch": 0.34265, + "grad_norm": 1.030259678090019, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 34265 + }, + { + "epoch": 0.34266, + "grad_norm": 1.0643344782524198, + "learning_rate": 0.003, + "loss": 4.049, + "step": 34266 + }, + { + "epoch": 0.34267, + "grad_norm": 0.9758631702254871, + "learning_rate": 0.003, + "loss": 4.031, + "step": 34267 + }, + { + "epoch": 0.34268, + "grad_norm": 1.086420393909518, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 34268 + }, + { + "epoch": 0.34269, + "grad_norm": 1.1160588374999825, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 34269 + }, + { + "epoch": 0.3427, + "grad_norm": 1.1957327969031535, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 34270 + }, + { + "epoch": 0.34271, + "grad_norm": 0.9092922438493196, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 34271 + }, + { + "epoch": 0.34272, + "grad_norm": 1.0654567586475823, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 34272 + }, + { + "epoch": 0.34273, + "grad_norm": 1.3132797089494375, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 34273 + }, + { + "epoch": 0.34274, + "grad_norm": 0.9411643637465419, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 34274 + }, + { + "epoch": 0.34275, + "grad_norm": 1.159091252287261, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 34275 + }, + { + "epoch": 0.34276, + "grad_norm": 1.327603492677768, + "learning_rate": 0.003, + "loss": 4.054, + "step": 34276 + }, + { + "epoch": 0.34277, + "grad_norm": 0.7884233003668002, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 34277 + }, + { + "epoch": 0.34278, + "grad_norm": 0.7346994975604143, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 34278 + }, + { + "epoch": 0.34279, + "grad_norm": 0.8195888413776732, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 34279 + }, + { + "epoch": 0.3428, + "grad_norm": 0.9821590475602906, + "learning_rate": 0.003, + "loss": 4.04, + "step": 34280 + }, + { + "epoch": 0.34281, + "grad_norm": 1.1844000844237144, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 34281 + }, + { + "epoch": 0.34282, + "grad_norm": 1.0402345540051858, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 34282 + }, + { + "epoch": 0.34283, + "grad_norm": 1.0861838717244883, + "learning_rate": 0.003, + "loss": 4.018, + "step": 34283 + }, + { + "epoch": 0.34284, + "grad_norm": 1.205517967206391, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 34284 + }, + { + "epoch": 0.34285, + "grad_norm": 1.1773173554159013, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 34285 + }, + { + "epoch": 0.34286, + "grad_norm": 1.0756243535617398, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 34286 + }, + { + "epoch": 0.34287, + "grad_norm": 1.108506325659439, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 34287 + }, + { + "epoch": 0.34288, + "grad_norm": 1.0558263699232615, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 34288 + }, + { + "epoch": 0.34289, + "grad_norm": 1.021966748542985, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 34289 + }, + { + "epoch": 0.3429, + "grad_norm": 1.2931176109039435, + "learning_rate": 0.003, + "loss": 4.015, + "step": 34290 + }, + { + "epoch": 0.34291, + "grad_norm": 0.9116598720127139, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 34291 + }, + { + "epoch": 0.34292, + "grad_norm": 0.9614504388264141, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 34292 + }, + { + "epoch": 0.34293, + "grad_norm": 1.2658254442350898, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 34293 + }, + { + "epoch": 0.34294, + "grad_norm": 1.0706839996857025, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 34294 + }, + { + "epoch": 0.34295, + "grad_norm": 1.0578546939096125, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 34295 + }, + { + "epoch": 0.34296, + "grad_norm": 1.1864363435395868, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 34296 + }, + { + "epoch": 0.34297, + "grad_norm": 1.0693920747017205, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 34297 + }, + { + "epoch": 0.34298, + "grad_norm": 1.1412568330831354, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 34298 + }, + { + "epoch": 0.34299, + "grad_norm": 0.963784991969164, + "learning_rate": 0.003, + "loss": 4.041, + "step": 34299 + }, + { + "epoch": 0.343, + "grad_norm": 1.1063689589775916, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 34300 + }, + { + "epoch": 0.34301, + "grad_norm": 0.9769025848334794, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 34301 + }, + { + "epoch": 0.34302, + "grad_norm": 1.0409675947949169, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 34302 + }, + { + "epoch": 0.34303, + "grad_norm": 1.0893790052877528, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 34303 + }, + { + "epoch": 0.34304, + "grad_norm": 0.855237430630082, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 34304 + }, + { + "epoch": 0.34305, + "grad_norm": 1.2475432775675868, + "learning_rate": 0.003, + "loss": 4.047, + "step": 34305 + }, + { + "epoch": 0.34306, + "grad_norm": 0.9856024134044679, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 34306 + }, + { + "epoch": 0.34307, + "grad_norm": 1.2494279090224942, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 34307 + }, + { + "epoch": 0.34308, + "grad_norm": 1.0683045734816532, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 34308 + }, + { + "epoch": 0.34309, + "grad_norm": 1.3205890037865755, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 34309 + }, + { + "epoch": 0.3431, + "grad_norm": 0.9573023147083429, + "learning_rate": 0.003, + "loss": 4.045, + "step": 34310 + }, + { + "epoch": 0.34311, + "grad_norm": 1.0956857682188417, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 34311 + }, + { + "epoch": 0.34312, + "grad_norm": 1.0011790776082563, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 34312 + }, + { + "epoch": 0.34313, + "grad_norm": 1.2075859726858982, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 34313 + }, + { + "epoch": 0.34314, + "grad_norm": 1.0568116159754664, + "learning_rate": 0.003, + "loss": 4.054, + "step": 34314 + }, + { + "epoch": 0.34315, + "grad_norm": 0.9720268111899834, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 34315 + }, + { + "epoch": 0.34316, + "grad_norm": 0.9927488677941723, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 34316 + }, + { + "epoch": 0.34317, + "grad_norm": 1.3992018478054833, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 34317 + }, + { + "epoch": 0.34318, + "grad_norm": 1.2000299673006731, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 34318 + }, + { + "epoch": 0.34319, + "grad_norm": 1.1206509508184526, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 34319 + }, + { + "epoch": 0.3432, + "grad_norm": 1.066143514337155, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 34320 + }, + { + "epoch": 0.34321, + "grad_norm": 1.1381239568442965, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 34321 + }, + { + "epoch": 0.34322, + "grad_norm": 1.0387138729369216, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 34322 + }, + { + "epoch": 0.34323, + "grad_norm": 1.1305850807473248, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 34323 + }, + { + "epoch": 0.34324, + "grad_norm": 0.974671385631424, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 34324 + }, + { + "epoch": 0.34325, + "grad_norm": 1.0005539387896605, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 34325 + }, + { + "epoch": 0.34326, + "grad_norm": 1.1217618061385677, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 34326 + }, + { + "epoch": 0.34327, + "grad_norm": 1.1567440990222617, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 34327 + }, + { + "epoch": 0.34328, + "grad_norm": 0.9811604040684305, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 34328 + }, + { + "epoch": 0.34329, + "grad_norm": 1.1945939326989814, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 34329 + }, + { + "epoch": 0.3433, + "grad_norm": 1.075623148947667, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 34330 + }, + { + "epoch": 0.34331, + "grad_norm": 1.1768513871548358, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 34331 + }, + { + "epoch": 0.34332, + "grad_norm": 1.1288589835130138, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 34332 + }, + { + "epoch": 0.34333, + "grad_norm": 1.0931400223522045, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 34333 + }, + { + "epoch": 0.34334, + "grad_norm": 1.030659549347335, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 34334 + }, + { + "epoch": 0.34335, + "grad_norm": 1.032335011042104, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 34335 + }, + { + "epoch": 0.34336, + "grad_norm": 1.179603253956772, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 34336 + }, + { + "epoch": 0.34337, + "grad_norm": 0.8718311099289218, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 34337 + }, + { + "epoch": 0.34338, + "grad_norm": 0.8408333467270137, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 34338 + }, + { + "epoch": 0.34339, + "grad_norm": 1.079595719823177, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 34339 + }, + { + "epoch": 0.3434, + "grad_norm": 1.2251655901985743, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 34340 + }, + { + "epoch": 0.34341, + "grad_norm": 0.9303357749412673, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 34341 + }, + { + "epoch": 0.34342, + "grad_norm": 1.0258167096228872, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 34342 + }, + { + "epoch": 0.34343, + "grad_norm": 1.205035124454518, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 34343 + }, + { + "epoch": 0.34344, + "grad_norm": 1.2168358777197203, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 34344 + }, + { + "epoch": 0.34345, + "grad_norm": 1.187112579087249, + "learning_rate": 0.003, + "loss": 4.062, + "step": 34345 + }, + { + "epoch": 0.34346, + "grad_norm": 1.1368567631155344, + "learning_rate": 0.003, + "loss": 4.034, + "step": 34346 + }, + { + "epoch": 0.34347, + "grad_norm": 0.8858971192208428, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 34347 + }, + { + "epoch": 0.34348, + "grad_norm": 1.1223866341414548, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 34348 + }, + { + "epoch": 0.34349, + "grad_norm": 1.3928366669892136, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 34349 + }, + { + "epoch": 0.3435, + "grad_norm": 0.9437742596385027, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 34350 + }, + { + "epoch": 0.34351, + "grad_norm": 0.997851918604408, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 34351 + }, + { + "epoch": 0.34352, + "grad_norm": 1.4051396483019778, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 34352 + }, + { + "epoch": 0.34353, + "grad_norm": 1.1501439720448163, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 34353 + }, + { + "epoch": 0.34354, + "grad_norm": 0.8900396799417867, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 34354 + }, + { + "epoch": 0.34355, + "grad_norm": 0.8865411834393244, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 34355 + }, + { + "epoch": 0.34356, + "grad_norm": 1.1554596115438396, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 34356 + }, + { + "epoch": 0.34357, + "grad_norm": 1.1363084471454674, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 34357 + }, + { + "epoch": 0.34358, + "grad_norm": 1.0015246840357344, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 34358 + }, + { + "epoch": 0.34359, + "grad_norm": 1.1248374862206525, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 34359 + }, + { + "epoch": 0.3436, + "grad_norm": 1.059021878006882, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 34360 + }, + { + "epoch": 0.34361, + "grad_norm": 1.0710319914152844, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 34361 + }, + { + "epoch": 0.34362, + "grad_norm": 0.9608784426968547, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 34362 + }, + { + "epoch": 0.34363, + "grad_norm": 1.0662226166782893, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 34363 + }, + { + "epoch": 0.34364, + "grad_norm": 1.2417262668046072, + "learning_rate": 0.003, + "loss": 4.1006, + "step": 34364 + }, + { + "epoch": 0.34365, + "grad_norm": 1.1267184858506023, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 34365 + }, + { + "epoch": 0.34366, + "grad_norm": 1.1968370391208707, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 34366 + }, + { + "epoch": 0.34367, + "grad_norm": 0.8455921492335274, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 34367 + }, + { + "epoch": 0.34368, + "grad_norm": 0.7894854933221787, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 34368 + }, + { + "epoch": 0.34369, + "grad_norm": 0.9372161958306626, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 34369 + }, + { + "epoch": 0.3437, + "grad_norm": 1.435646271984004, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 34370 + }, + { + "epoch": 0.34371, + "grad_norm": 1.0249527844671584, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 34371 + }, + { + "epoch": 0.34372, + "grad_norm": 1.1591946601786929, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 34372 + }, + { + "epoch": 0.34373, + "grad_norm": 1.1335302112677104, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 34373 + }, + { + "epoch": 0.34374, + "grad_norm": 1.0800997991307064, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 34374 + }, + { + "epoch": 0.34375, + "grad_norm": 0.9326751218821264, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 34375 + }, + { + "epoch": 0.34376, + "grad_norm": 0.9225890831394509, + "learning_rate": 0.003, + "loss": 4.006, + "step": 34376 + }, + { + "epoch": 0.34377, + "grad_norm": 1.060448973020737, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 34377 + }, + { + "epoch": 0.34378, + "grad_norm": 1.2399634083054718, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 34378 + }, + { + "epoch": 0.34379, + "grad_norm": 0.996829824499122, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 34379 + }, + { + "epoch": 0.3438, + "grad_norm": 1.1577443705486932, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 34380 + }, + { + "epoch": 0.34381, + "grad_norm": 0.9683368435596569, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 34381 + }, + { + "epoch": 0.34382, + "grad_norm": 1.0350607496708961, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 34382 + }, + { + "epoch": 0.34383, + "grad_norm": 1.1561353447546552, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 34383 + }, + { + "epoch": 0.34384, + "grad_norm": 1.144320759608624, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 34384 + }, + { + "epoch": 0.34385, + "grad_norm": 1.145103606670467, + "learning_rate": 0.003, + "loss": 4.047, + "step": 34385 + }, + { + "epoch": 0.34386, + "grad_norm": 1.1045209990384688, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 34386 + }, + { + "epoch": 0.34387, + "grad_norm": 1.1945573978957928, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 34387 + }, + { + "epoch": 0.34388, + "grad_norm": 1.247488565372402, + "learning_rate": 0.003, + "loss": 4.059, + "step": 34388 + }, + { + "epoch": 0.34389, + "grad_norm": 0.9415907274917961, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 34389 + }, + { + "epoch": 0.3439, + "grad_norm": 1.1375468446483077, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 34390 + }, + { + "epoch": 0.34391, + "grad_norm": 1.2464770458617342, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 34391 + }, + { + "epoch": 0.34392, + "grad_norm": 1.0636118325544666, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 34392 + }, + { + "epoch": 0.34393, + "grad_norm": 1.0785198627172419, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 34393 + }, + { + "epoch": 0.34394, + "grad_norm": 1.0023080949803131, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 34394 + }, + { + "epoch": 0.34395, + "grad_norm": 1.1727033538491467, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 34395 + }, + { + "epoch": 0.34396, + "grad_norm": 1.0733405512742908, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 34396 + }, + { + "epoch": 0.34397, + "grad_norm": 0.9097377922619316, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 34397 + }, + { + "epoch": 0.34398, + "grad_norm": 1.1071129202129135, + "learning_rate": 0.003, + "loss": 4.063, + "step": 34398 + }, + { + "epoch": 0.34399, + "grad_norm": 1.2723793762894435, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 34399 + }, + { + "epoch": 0.344, + "grad_norm": 1.0089670283834078, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 34400 + }, + { + "epoch": 0.34401, + "grad_norm": 0.9618385021195759, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 34401 + }, + { + "epoch": 0.34402, + "grad_norm": 1.1178667567167735, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 34402 + }, + { + "epoch": 0.34403, + "grad_norm": 1.173853489221578, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 34403 + }, + { + "epoch": 0.34404, + "grad_norm": 1.0608531874353326, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 34404 + }, + { + "epoch": 0.34405, + "grad_norm": 1.0953577592136927, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 34405 + }, + { + "epoch": 0.34406, + "grad_norm": 0.9511926893115736, + "learning_rate": 0.003, + "loss": 4.032, + "step": 34406 + }, + { + "epoch": 0.34407, + "grad_norm": 0.9395167665396553, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 34407 + }, + { + "epoch": 0.34408, + "grad_norm": 1.1633435290289866, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 34408 + }, + { + "epoch": 0.34409, + "grad_norm": 1.154624228039658, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 34409 + }, + { + "epoch": 0.3441, + "grad_norm": 1.0380760755690346, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 34410 + }, + { + "epoch": 0.34411, + "grad_norm": 1.0494361048342669, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 34411 + }, + { + "epoch": 0.34412, + "grad_norm": 1.0273683758792918, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 34412 + }, + { + "epoch": 0.34413, + "grad_norm": 1.113449082354047, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 34413 + }, + { + "epoch": 0.34414, + "grad_norm": 1.2444609980360672, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 34414 + }, + { + "epoch": 0.34415, + "grad_norm": 1.090546661854788, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 34415 + }, + { + "epoch": 0.34416, + "grad_norm": 1.1840914165332221, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 34416 + }, + { + "epoch": 0.34417, + "grad_norm": 1.0304329339245737, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 34417 + }, + { + "epoch": 0.34418, + "grad_norm": 1.1425319848778333, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 34418 + }, + { + "epoch": 0.34419, + "grad_norm": 1.10536549440644, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 34419 + }, + { + "epoch": 0.3442, + "grad_norm": 0.9797331743674014, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 34420 + }, + { + "epoch": 0.34421, + "grad_norm": 1.0463816376646808, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 34421 + }, + { + "epoch": 0.34422, + "grad_norm": 1.1787297371854395, + "learning_rate": 0.003, + "loss": 4.049, + "step": 34422 + }, + { + "epoch": 0.34423, + "grad_norm": 1.0485501055093143, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 34423 + }, + { + "epoch": 0.34424, + "grad_norm": 1.065451029434722, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 34424 + }, + { + "epoch": 0.34425, + "grad_norm": 1.2181209460231668, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 34425 + }, + { + "epoch": 0.34426, + "grad_norm": 0.8739017139394357, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 34426 + }, + { + "epoch": 0.34427, + "grad_norm": 0.8976125926482074, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 34427 + }, + { + "epoch": 0.34428, + "grad_norm": 1.1672023793466535, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 34428 + }, + { + "epoch": 0.34429, + "grad_norm": 1.3311408768738586, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 34429 + }, + { + "epoch": 0.3443, + "grad_norm": 0.9479928022395492, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 34430 + }, + { + "epoch": 0.34431, + "grad_norm": 1.0678374707300557, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 34431 + }, + { + "epoch": 0.34432, + "grad_norm": 1.441122403033255, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 34432 + }, + { + "epoch": 0.34433, + "grad_norm": 0.9215498164522137, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 34433 + }, + { + "epoch": 0.34434, + "grad_norm": 1.0393663935589699, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 34434 + }, + { + "epoch": 0.34435, + "grad_norm": 1.343912227055471, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 34435 + }, + { + "epoch": 0.34436, + "grad_norm": 0.9629706784900298, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 34436 + }, + { + "epoch": 0.34437, + "grad_norm": 1.0103213230708155, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 34437 + }, + { + "epoch": 0.34438, + "grad_norm": 1.135629056045014, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 34438 + }, + { + "epoch": 0.34439, + "grad_norm": 1.0923898529025855, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 34439 + }, + { + "epoch": 0.3444, + "grad_norm": 0.9991303295013053, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 34440 + }, + { + "epoch": 0.34441, + "grad_norm": 1.2684757218821958, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 34441 + }, + { + "epoch": 0.34442, + "grad_norm": 1.0675723160943753, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 34442 + }, + { + "epoch": 0.34443, + "grad_norm": 1.0632550837806138, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 34443 + }, + { + "epoch": 0.34444, + "grad_norm": 1.2021121960017094, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 34444 + }, + { + "epoch": 0.34445, + "grad_norm": 1.0228918083351872, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 34445 + }, + { + "epoch": 0.34446, + "grad_norm": 1.3122064810554948, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 34446 + }, + { + "epoch": 0.34447, + "grad_norm": 1.0191545991293067, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 34447 + }, + { + "epoch": 0.34448, + "grad_norm": 1.1245897985869333, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 34448 + }, + { + "epoch": 0.34449, + "grad_norm": 1.1130977337295485, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 34449 + }, + { + "epoch": 0.3445, + "grad_norm": 1.2404314323560053, + "learning_rate": 0.003, + "loss": 4.042, + "step": 34450 + }, + { + "epoch": 0.34451, + "grad_norm": 1.0638726672986722, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 34451 + }, + { + "epoch": 0.34452, + "grad_norm": 1.0254934260873478, + "learning_rate": 0.003, + "loss": 4.026, + "step": 34452 + }, + { + "epoch": 0.34453, + "grad_norm": 1.2324374909648081, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 34453 + }, + { + "epoch": 0.34454, + "grad_norm": 1.0420012531402159, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 34454 + }, + { + "epoch": 0.34455, + "grad_norm": 1.1291993726081564, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 34455 + }, + { + "epoch": 0.34456, + "grad_norm": 0.9472107085523773, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 34456 + }, + { + "epoch": 0.34457, + "grad_norm": 0.9894540384375469, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 34457 + }, + { + "epoch": 0.34458, + "grad_norm": 1.047291677693332, + "learning_rate": 0.003, + "loss": 4.0898, + "step": 34458 + }, + { + "epoch": 0.34459, + "grad_norm": 0.9993270397097432, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 34459 + }, + { + "epoch": 0.3446, + "grad_norm": 1.2376400729084978, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 34460 + }, + { + "epoch": 0.34461, + "grad_norm": 0.9873479403383387, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 34461 + }, + { + "epoch": 0.34462, + "grad_norm": 1.1616177256151823, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 34462 + }, + { + "epoch": 0.34463, + "grad_norm": 1.0714019240921755, + "learning_rate": 0.003, + "loss": 4.053, + "step": 34463 + }, + { + "epoch": 0.34464, + "grad_norm": 1.1890356846047612, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 34464 + }, + { + "epoch": 0.34465, + "grad_norm": 1.0804038949425545, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 34465 + }, + { + "epoch": 0.34466, + "grad_norm": 0.9661959405648028, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 34466 + }, + { + "epoch": 0.34467, + "grad_norm": 0.9105210365324069, + "learning_rate": 0.003, + "loss": 4.037, + "step": 34467 + }, + { + "epoch": 0.34468, + "grad_norm": 1.0809668479952121, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 34468 + }, + { + "epoch": 0.34469, + "grad_norm": 1.153064076238308, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 34469 + }, + { + "epoch": 0.3447, + "grad_norm": 1.0154475509719951, + "learning_rate": 0.003, + "loss": 4.014, + "step": 34470 + }, + { + "epoch": 0.34471, + "grad_norm": 1.175084941530833, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 34471 + }, + { + "epoch": 0.34472, + "grad_norm": 0.8736257618957576, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 34472 + }, + { + "epoch": 0.34473, + "grad_norm": 1.1377209698626776, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 34473 + }, + { + "epoch": 0.34474, + "grad_norm": 1.312082228574056, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 34474 + }, + { + "epoch": 0.34475, + "grad_norm": 0.973187962768112, + "learning_rate": 0.003, + "loss": 4.029, + "step": 34475 + }, + { + "epoch": 0.34476, + "grad_norm": 1.159243410294867, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 34476 + }, + { + "epoch": 0.34477, + "grad_norm": 1.1059314371743734, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 34477 + }, + { + "epoch": 0.34478, + "grad_norm": 1.257403136485606, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 34478 + }, + { + "epoch": 0.34479, + "grad_norm": 0.7773451849498993, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 34479 + }, + { + "epoch": 0.3448, + "grad_norm": 0.8464536112094444, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 34480 + }, + { + "epoch": 0.34481, + "grad_norm": 1.2015042102264286, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 34481 + }, + { + "epoch": 0.34482, + "grad_norm": 1.3774947832765685, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 34482 + }, + { + "epoch": 0.34483, + "grad_norm": 0.9047716767833524, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 34483 + }, + { + "epoch": 0.34484, + "grad_norm": 0.9283898611016924, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 34484 + }, + { + "epoch": 0.34485, + "grad_norm": 1.1362894633173228, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 34485 + }, + { + "epoch": 0.34486, + "grad_norm": 1.121681898929463, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 34486 + }, + { + "epoch": 0.34487, + "grad_norm": 0.9631516380978984, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 34487 + }, + { + "epoch": 0.34488, + "grad_norm": 1.189838477692484, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 34488 + }, + { + "epoch": 0.34489, + "grad_norm": 1.0459899612352928, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 34489 + }, + { + "epoch": 0.3449, + "grad_norm": 1.0866760474213057, + "learning_rate": 0.003, + "loss": 4.038, + "step": 34490 + }, + { + "epoch": 0.34491, + "grad_norm": 1.1352364066020066, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 34491 + }, + { + "epoch": 0.34492, + "grad_norm": 1.1839784502723127, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 34492 + }, + { + "epoch": 0.34493, + "grad_norm": 1.0909190847348964, + "learning_rate": 0.003, + "loss": 4.047, + "step": 34493 + }, + { + "epoch": 0.34494, + "grad_norm": 1.0428403606237693, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 34494 + }, + { + "epoch": 0.34495, + "grad_norm": 1.10309752313844, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 34495 + }, + { + "epoch": 0.34496, + "grad_norm": 0.9492679384800012, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 34496 + }, + { + "epoch": 0.34497, + "grad_norm": 0.8776563978487001, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 34497 + }, + { + "epoch": 0.34498, + "grad_norm": 0.9382838941288548, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 34498 + }, + { + "epoch": 0.34499, + "grad_norm": 1.2627407300558977, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 34499 + }, + { + "epoch": 0.345, + "grad_norm": 0.8697403157545301, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 34500 + }, + { + "epoch": 0.34501, + "grad_norm": 0.9963726973242036, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 34501 + }, + { + "epoch": 0.34502, + "grad_norm": 1.2364602399606024, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 34502 + }, + { + "epoch": 0.34503, + "grad_norm": 0.8692341018680425, + "learning_rate": 0.003, + "loss": 4.033, + "step": 34503 + }, + { + "epoch": 0.34504, + "grad_norm": 0.9935214518619855, + "learning_rate": 0.003, + "loss": 4.041, + "step": 34504 + }, + { + "epoch": 0.34505, + "grad_norm": 1.2711180514204432, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 34505 + }, + { + "epoch": 0.34506, + "grad_norm": 0.882503312222552, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 34506 + }, + { + "epoch": 0.34507, + "grad_norm": 1.1124556982321392, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 34507 + }, + { + "epoch": 0.34508, + "grad_norm": 1.4934706010969578, + "learning_rate": 0.003, + "loss": 4.08, + "step": 34508 + }, + { + "epoch": 0.34509, + "grad_norm": 1.0029079913956445, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 34509 + }, + { + "epoch": 0.3451, + "grad_norm": 1.0836321461011313, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 34510 + }, + { + "epoch": 0.34511, + "grad_norm": 1.103694869167866, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 34511 + }, + { + "epoch": 0.34512, + "grad_norm": 1.1204065232792306, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 34512 + }, + { + "epoch": 0.34513, + "grad_norm": 1.0207349960225407, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 34513 + }, + { + "epoch": 0.34514, + "grad_norm": 1.1419505702362633, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 34514 + }, + { + "epoch": 0.34515, + "grad_norm": 1.0433438659980998, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 34515 + }, + { + "epoch": 0.34516, + "grad_norm": 1.181883672443725, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 34516 + }, + { + "epoch": 0.34517, + "grad_norm": 1.0979500101127118, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 34517 + }, + { + "epoch": 0.34518, + "grad_norm": 1.2039335674732161, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 34518 + }, + { + "epoch": 0.34519, + "grad_norm": 1.0017513408131442, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 34519 + }, + { + "epoch": 0.3452, + "grad_norm": 1.0747819219032564, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 34520 + }, + { + "epoch": 0.34521, + "grad_norm": 1.092785095012136, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 34521 + }, + { + "epoch": 0.34522, + "grad_norm": 1.0149636002242768, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 34522 + }, + { + "epoch": 0.34523, + "grad_norm": 1.2872141578290284, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 34523 + }, + { + "epoch": 0.34524, + "grad_norm": 0.9357118193938935, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 34524 + }, + { + "epoch": 0.34525, + "grad_norm": 1.0996461187875541, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 34525 + }, + { + "epoch": 0.34526, + "grad_norm": 1.1680488477722262, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 34526 + }, + { + "epoch": 0.34527, + "grad_norm": 0.9777798573538227, + "learning_rate": 0.003, + "loss": 4.047, + "step": 34527 + }, + { + "epoch": 0.34528, + "grad_norm": 1.3004326953875893, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 34528 + }, + { + "epoch": 0.34529, + "grad_norm": 1.101310913122926, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 34529 + }, + { + "epoch": 0.3453, + "grad_norm": 1.0285331665008326, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 34530 + }, + { + "epoch": 0.34531, + "grad_norm": 1.1099883713809686, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 34531 + }, + { + "epoch": 0.34532, + "grad_norm": 0.9582157203982202, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 34532 + }, + { + "epoch": 0.34533, + "grad_norm": 1.037631487835121, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 34533 + }, + { + "epoch": 0.34534, + "grad_norm": 1.1711698655152962, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 34534 + }, + { + "epoch": 0.34535, + "grad_norm": 1.0667394798770835, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 34535 + }, + { + "epoch": 0.34536, + "grad_norm": 1.3096288602962214, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 34536 + }, + { + "epoch": 0.34537, + "grad_norm": 0.9873886229564557, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 34537 + }, + { + "epoch": 0.34538, + "grad_norm": 1.0527250574420928, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 34538 + }, + { + "epoch": 0.34539, + "grad_norm": 1.1010920070442032, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 34539 + }, + { + "epoch": 0.3454, + "grad_norm": 1.133629306201421, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 34540 + }, + { + "epoch": 0.34541, + "grad_norm": 1.1604572630528438, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 34541 + }, + { + "epoch": 0.34542, + "grad_norm": 1.0450299478041332, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 34542 + }, + { + "epoch": 0.34543, + "grad_norm": 1.1344904844109196, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 34543 + }, + { + "epoch": 0.34544, + "grad_norm": 1.0014323141343593, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 34544 + }, + { + "epoch": 0.34545, + "grad_norm": 1.033750850464248, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 34545 + }, + { + "epoch": 0.34546, + "grad_norm": 1.0634533631953897, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 34546 + }, + { + "epoch": 0.34547, + "grad_norm": 1.3270515985963502, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 34547 + }, + { + "epoch": 0.34548, + "grad_norm": 0.9259613706410639, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 34548 + }, + { + "epoch": 0.34549, + "grad_norm": 1.003652796685915, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 34549 + }, + { + "epoch": 0.3455, + "grad_norm": 1.210051734704486, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 34550 + }, + { + "epoch": 0.34551, + "grad_norm": 0.9216014826336848, + "learning_rate": 0.003, + "loss": 4.051, + "step": 34551 + }, + { + "epoch": 0.34552, + "grad_norm": 0.9063475824033662, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 34552 + }, + { + "epoch": 0.34553, + "grad_norm": 1.0807350146329724, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 34553 + }, + { + "epoch": 0.34554, + "grad_norm": 1.2548195974166052, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 34554 + }, + { + "epoch": 0.34555, + "grad_norm": 1.0009322608869087, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 34555 + }, + { + "epoch": 0.34556, + "grad_norm": 1.367960408167423, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 34556 + }, + { + "epoch": 0.34557, + "grad_norm": 0.8988139525653036, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 34557 + }, + { + "epoch": 0.34558, + "grad_norm": 1.142544989691883, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 34558 + }, + { + "epoch": 0.34559, + "grad_norm": 1.1596754668064686, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 34559 + }, + { + "epoch": 0.3456, + "grad_norm": 1.2061860233067336, + "learning_rate": 0.003, + "loss": 4.031, + "step": 34560 + }, + { + "epoch": 0.34561, + "grad_norm": 0.9293105634359753, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 34561 + }, + { + "epoch": 0.34562, + "grad_norm": 1.1234999459348485, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 34562 + }, + { + "epoch": 0.34563, + "grad_norm": 1.2230951521557023, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 34563 + }, + { + "epoch": 0.34564, + "grad_norm": 0.9923948976408027, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 34564 + }, + { + "epoch": 0.34565, + "grad_norm": 1.175891239423349, + "learning_rate": 0.003, + "loss": 4.052, + "step": 34565 + }, + { + "epoch": 0.34566, + "grad_norm": 0.9987303442299005, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 34566 + }, + { + "epoch": 0.34567, + "grad_norm": 0.9988809879764203, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 34567 + }, + { + "epoch": 0.34568, + "grad_norm": 1.1785654366321827, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 34568 + }, + { + "epoch": 0.34569, + "grad_norm": 0.8391386664929004, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 34569 + }, + { + "epoch": 0.3457, + "grad_norm": 0.8147626775104346, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 34570 + }, + { + "epoch": 0.34571, + "grad_norm": 0.9665670260153781, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 34571 + }, + { + "epoch": 0.34572, + "grad_norm": 1.0363716413218877, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 34572 + }, + { + "epoch": 0.34573, + "grad_norm": 0.9372905169981248, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 34573 + }, + { + "epoch": 0.34574, + "grad_norm": 1.0591259209460098, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 34574 + }, + { + "epoch": 0.34575, + "grad_norm": 1.2444889736590088, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 34575 + }, + { + "epoch": 0.34576, + "grad_norm": 0.9975291091185307, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 34576 + }, + { + "epoch": 0.34577, + "grad_norm": 1.4454311972262395, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 34577 + }, + { + "epoch": 0.34578, + "grad_norm": 0.9487576567755055, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 34578 + }, + { + "epoch": 0.34579, + "grad_norm": 1.1792518363083155, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 34579 + }, + { + "epoch": 0.3458, + "grad_norm": 1.3390492288103146, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 34580 + }, + { + "epoch": 0.34581, + "grad_norm": 0.9029090758872884, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 34581 + }, + { + "epoch": 0.34582, + "grad_norm": 1.1076844631677525, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 34582 + }, + { + "epoch": 0.34583, + "grad_norm": 1.2215239127569717, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 34583 + }, + { + "epoch": 0.34584, + "grad_norm": 1.1355358759387406, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 34584 + }, + { + "epoch": 0.34585, + "grad_norm": 0.9325371680303802, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 34585 + }, + { + "epoch": 0.34586, + "grad_norm": 0.8987847500935281, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 34586 + }, + { + "epoch": 0.34587, + "grad_norm": 1.0721232156612748, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 34587 + }, + { + "epoch": 0.34588, + "grad_norm": 1.1961237203239603, + "learning_rate": 0.003, + "loss": 4.04, + "step": 34588 + }, + { + "epoch": 0.34589, + "grad_norm": 1.0607003231622938, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 34589 + }, + { + "epoch": 0.3459, + "grad_norm": 1.1055511983165889, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 34590 + }, + { + "epoch": 0.34591, + "grad_norm": 1.040094128790307, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 34591 + }, + { + "epoch": 0.34592, + "grad_norm": 0.992835866945854, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 34592 + }, + { + "epoch": 0.34593, + "grad_norm": 1.1587215137962856, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 34593 + }, + { + "epoch": 0.34594, + "grad_norm": 0.9312909477308969, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 34594 + }, + { + "epoch": 0.34595, + "grad_norm": 0.8867056181095341, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 34595 + }, + { + "epoch": 0.34596, + "grad_norm": 1.0112615448032973, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 34596 + }, + { + "epoch": 0.34597, + "grad_norm": 1.3013120231086546, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 34597 + }, + { + "epoch": 0.34598, + "grad_norm": 1.1410630125288672, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 34598 + }, + { + "epoch": 0.34599, + "grad_norm": 1.2426191646178313, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 34599 + }, + { + "epoch": 0.346, + "grad_norm": 0.9946927063597509, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 34600 + }, + { + "epoch": 0.34601, + "grad_norm": 1.2677060285699824, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 34601 + }, + { + "epoch": 0.34602, + "grad_norm": 1.004110533361268, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 34602 + }, + { + "epoch": 0.34603, + "grad_norm": 1.2004699004250199, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 34603 + }, + { + "epoch": 0.34604, + "grad_norm": 1.0548649958909995, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 34604 + }, + { + "epoch": 0.34605, + "grad_norm": 1.0305372417357612, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 34605 + }, + { + "epoch": 0.34606, + "grad_norm": 1.0990275655751056, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 34606 + }, + { + "epoch": 0.34607, + "grad_norm": 1.0206053637126749, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 34607 + }, + { + "epoch": 0.34608, + "grad_norm": 1.18266095750583, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 34608 + }, + { + "epoch": 0.34609, + "grad_norm": 0.9318701397500743, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 34609 + }, + { + "epoch": 0.3461, + "grad_norm": 1.1628386321367865, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 34610 + }, + { + "epoch": 0.34611, + "grad_norm": 1.0420927859358793, + "learning_rate": 0.003, + "loss": 4.053, + "step": 34611 + }, + { + "epoch": 0.34612, + "grad_norm": 1.2170877935838331, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 34612 + }, + { + "epoch": 0.34613, + "grad_norm": 1.0235048863101133, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 34613 + }, + { + "epoch": 0.34614, + "grad_norm": 1.27866812246563, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 34614 + }, + { + "epoch": 0.34615, + "grad_norm": 1.0092400497989662, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 34615 + }, + { + "epoch": 0.34616, + "grad_norm": 1.1184139556062176, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 34616 + }, + { + "epoch": 0.34617, + "grad_norm": 0.9507733863761583, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 34617 + }, + { + "epoch": 0.34618, + "grad_norm": 1.2503535564315258, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 34618 + }, + { + "epoch": 0.34619, + "grad_norm": 1.0665900660602115, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 34619 + }, + { + "epoch": 0.3462, + "grad_norm": 1.02323630132289, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 34620 + }, + { + "epoch": 0.34621, + "grad_norm": 1.113751994034353, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 34621 + }, + { + "epoch": 0.34622, + "grad_norm": 1.3513763550584867, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 34622 + }, + { + "epoch": 0.34623, + "grad_norm": 1.106656189759034, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 34623 + }, + { + "epoch": 0.34624, + "grad_norm": 1.3120536444185282, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 34624 + }, + { + "epoch": 0.34625, + "grad_norm": 0.8976893625191695, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 34625 + }, + { + "epoch": 0.34626, + "grad_norm": 0.940122226343621, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 34626 + }, + { + "epoch": 0.34627, + "grad_norm": 1.1485231412308123, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 34627 + }, + { + "epoch": 0.34628, + "grad_norm": 1.2559852677024843, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 34628 + }, + { + "epoch": 0.34629, + "grad_norm": 1.1314118227762928, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 34629 + }, + { + "epoch": 0.3463, + "grad_norm": 1.203638213860725, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 34630 + }, + { + "epoch": 0.34631, + "grad_norm": 0.8537946848034833, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 34631 + }, + { + "epoch": 0.34632, + "grad_norm": 0.814868180653638, + "learning_rate": 0.003, + "loss": 4.052, + "step": 34632 + }, + { + "epoch": 0.34633, + "grad_norm": 0.9586285716524866, + "learning_rate": 0.003, + "loss": 4.056, + "step": 34633 + }, + { + "epoch": 0.34634, + "grad_norm": 1.0939332750492439, + "learning_rate": 0.003, + "loss": 4.03, + "step": 34634 + }, + { + "epoch": 0.34635, + "grad_norm": 1.1094399009287668, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 34635 + }, + { + "epoch": 0.34636, + "grad_norm": 1.3458382399008637, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 34636 + }, + { + "epoch": 0.34637, + "grad_norm": 0.9519570822326945, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 34637 + }, + { + "epoch": 0.34638, + "grad_norm": 1.088427374550152, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 34638 + }, + { + "epoch": 0.34639, + "grad_norm": 1.0295009543175184, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 34639 + }, + { + "epoch": 0.3464, + "grad_norm": 1.0628203673085548, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 34640 + }, + { + "epoch": 0.34641, + "grad_norm": 1.0383772981271855, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 34641 + }, + { + "epoch": 0.34642, + "grad_norm": 1.0544789131288894, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 34642 + }, + { + "epoch": 0.34643, + "grad_norm": 1.2890104315378796, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 34643 + }, + { + "epoch": 0.34644, + "grad_norm": 1.024890368698296, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 34644 + }, + { + "epoch": 0.34645, + "grad_norm": 1.1700077749590638, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 34645 + }, + { + "epoch": 0.34646, + "grad_norm": 1.2358440507081674, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 34646 + }, + { + "epoch": 0.34647, + "grad_norm": 1.0959016037139462, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 34647 + }, + { + "epoch": 0.34648, + "grad_norm": 1.1674962924443633, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 34648 + }, + { + "epoch": 0.34649, + "grad_norm": 0.9715408499688554, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 34649 + }, + { + "epoch": 0.3465, + "grad_norm": 0.9511307156401978, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 34650 + }, + { + "epoch": 0.34651, + "grad_norm": 0.991402181869661, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 34651 + }, + { + "epoch": 0.34652, + "grad_norm": 1.0427256049521245, + "learning_rate": 0.003, + "loss": 4.07, + "step": 34652 + }, + { + "epoch": 0.34653, + "grad_norm": 1.172120961269069, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 34653 + }, + { + "epoch": 0.34654, + "grad_norm": 1.0779708576524651, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 34654 + }, + { + "epoch": 0.34655, + "grad_norm": 1.1791560334739766, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 34655 + }, + { + "epoch": 0.34656, + "grad_norm": 1.2757430567888697, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 34656 + }, + { + "epoch": 0.34657, + "grad_norm": 0.9139516725375675, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 34657 + }, + { + "epoch": 0.34658, + "grad_norm": 0.9879144938394865, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 34658 + }, + { + "epoch": 0.34659, + "grad_norm": 1.2133177801321358, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 34659 + }, + { + "epoch": 0.3466, + "grad_norm": 1.0464011059471858, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 34660 + }, + { + "epoch": 0.34661, + "grad_norm": 1.116803341306035, + "learning_rate": 0.003, + "loss": 4.034, + "step": 34661 + }, + { + "epoch": 0.34662, + "grad_norm": 1.1395737071629113, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 34662 + }, + { + "epoch": 0.34663, + "grad_norm": 0.9155381737961222, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 34663 + }, + { + "epoch": 0.34664, + "grad_norm": 1.071947110482163, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 34664 + }, + { + "epoch": 0.34665, + "grad_norm": 1.0177114799363616, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 34665 + }, + { + "epoch": 0.34666, + "grad_norm": 0.9660896719116978, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 34666 + }, + { + "epoch": 0.34667, + "grad_norm": 1.1090816171708484, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 34667 + }, + { + "epoch": 0.34668, + "grad_norm": 0.99905947279207, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 34668 + }, + { + "epoch": 0.34669, + "grad_norm": 1.2258229161258472, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 34669 + }, + { + "epoch": 0.3467, + "grad_norm": 0.9824136722649299, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 34670 + }, + { + "epoch": 0.34671, + "grad_norm": 1.1639281282530654, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 34671 + }, + { + "epoch": 0.34672, + "grad_norm": 1.2292305378229489, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 34672 + }, + { + "epoch": 0.34673, + "grad_norm": 0.949550035939364, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 34673 + }, + { + "epoch": 0.34674, + "grad_norm": 1.0298793266713946, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 34674 + }, + { + "epoch": 0.34675, + "grad_norm": 1.1875017003547486, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 34675 + }, + { + "epoch": 0.34676, + "grad_norm": 0.9906156972909791, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 34676 + }, + { + "epoch": 0.34677, + "grad_norm": 1.1187191550111535, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 34677 + }, + { + "epoch": 0.34678, + "grad_norm": 0.87553773749331, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 34678 + }, + { + "epoch": 0.34679, + "grad_norm": 1.1013025020282505, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 34679 + }, + { + "epoch": 0.3468, + "grad_norm": 1.2159153932814517, + "learning_rate": 0.003, + "loss": 4.03, + "step": 34680 + }, + { + "epoch": 0.34681, + "grad_norm": 0.8755684859113395, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 34681 + }, + { + "epoch": 0.34682, + "grad_norm": 0.9124875612768798, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 34682 + }, + { + "epoch": 0.34683, + "grad_norm": 1.1181718347091563, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 34683 + }, + { + "epoch": 0.34684, + "grad_norm": 1.2370409017989514, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 34684 + }, + { + "epoch": 0.34685, + "grad_norm": 0.7880443246555622, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 34685 + }, + { + "epoch": 0.34686, + "grad_norm": 0.7666665226338994, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 34686 + }, + { + "epoch": 0.34687, + "grad_norm": 0.92524749794835, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 34687 + }, + { + "epoch": 0.34688, + "grad_norm": 1.2351720575461556, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 34688 + }, + { + "epoch": 0.34689, + "grad_norm": 1.1420966055296984, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 34689 + }, + { + "epoch": 0.3469, + "grad_norm": 1.2296145541589567, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 34690 + }, + { + "epoch": 0.34691, + "grad_norm": 1.0950698652300888, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 34691 + }, + { + "epoch": 0.34692, + "grad_norm": 1.2185001248850789, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 34692 + }, + { + "epoch": 0.34693, + "grad_norm": 0.9832421885217651, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 34693 + }, + { + "epoch": 0.34694, + "grad_norm": 1.0214333370716875, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 34694 + }, + { + "epoch": 0.34695, + "grad_norm": 1.106939342947082, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 34695 + }, + { + "epoch": 0.34696, + "grad_norm": 1.2283735205153177, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 34696 + }, + { + "epoch": 0.34697, + "grad_norm": 1.0152124364062785, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 34697 + }, + { + "epoch": 0.34698, + "grad_norm": 1.3201615558201536, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 34698 + }, + { + "epoch": 0.34699, + "grad_norm": 0.980462786965123, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 34699 + }, + { + "epoch": 0.347, + "grad_norm": 1.0568177295396786, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 34700 + }, + { + "epoch": 0.34701, + "grad_norm": 1.2295233723164276, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 34701 + }, + { + "epoch": 0.34702, + "grad_norm": 1.0285674373504325, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 34702 + }, + { + "epoch": 0.34703, + "grad_norm": 1.207353337605062, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 34703 + }, + { + "epoch": 0.34704, + "grad_norm": 1.0273940717875862, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 34704 + }, + { + "epoch": 0.34705, + "grad_norm": 1.2964987977299955, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 34705 + }, + { + "epoch": 0.34706, + "grad_norm": 1.052878628951556, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 34706 + }, + { + "epoch": 0.34707, + "grad_norm": 1.1179127302944007, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 34707 + }, + { + "epoch": 0.34708, + "grad_norm": 1.0572626402027387, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 34708 + }, + { + "epoch": 0.34709, + "grad_norm": 0.9286015223277748, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 34709 + }, + { + "epoch": 0.3471, + "grad_norm": 0.9844417519997285, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 34710 + }, + { + "epoch": 0.34711, + "grad_norm": 1.178180664541173, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 34711 + }, + { + "epoch": 0.34712, + "grad_norm": 1.331370754092897, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 34712 + }, + { + "epoch": 0.34713, + "grad_norm": 1.1107291406020106, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 34713 + }, + { + "epoch": 0.34714, + "grad_norm": 0.9574497539726918, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 34714 + }, + { + "epoch": 0.34715, + "grad_norm": 1.0850886438130813, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 34715 + }, + { + "epoch": 0.34716, + "grad_norm": 1.1944449442197478, + "learning_rate": 0.003, + "loss": 4.036, + "step": 34716 + }, + { + "epoch": 0.34717, + "grad_norm": 1.0049633890534215, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 34717 + }, + { + "epoch": 0.34718, + "grad_norm": 1.1522963977220948, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 34718 + }, + { + "epoch": 0.34719, + "grad_norm": 0.9990124236689076, + "learning_rate": 0.003, + "loss": 4.02, + "step": 34719 + }, + { + "epoch": 0.3472, + "grad_norm": 1.219856785446624, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 34720 + }, + { + "epoch": 0.34721, + "grad_norm": 1.0981609947961362, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 34721 + }, + { + "epoch": 0.34722, + "grad_norm": 1.0652413073062037, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 34722 + }, + { + "epoch": 0.34723, + "grad_norm": 0.9699632941910848, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 34723 + }, + { + "epoch": 0.34724, + "grad_norm": 1.1293519200914568, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 34724 + }, + { + "epoch": 0.34725, + "grad_norm": 1.2697501308076298, + "learning_rate": 0.003, + "loss": 4.024, + "step": 34725 + }, + { + "epoch": 0.34726, + "grad_norm": 1.2784755616608057, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 34726 + }, + { + "epoch": 0.34727, + "grad_norm": 0.9441577237973676, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 34727 + }, + { + "epoch": 0.34728, + "grad_norm": 1.1064128218965363, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 34728 + }, + { + "epoch": 0.34729, + "grad_norm": 1.2802112121763698, + "learning_rate": 0.003, + "loss": 4.056, + "step": 34729 + }, + { + "epoch": 0.3473, + "grad_norm": 0.9554204268425464, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 34730 + }, + { + "epoch": 0.34731, + "grad_norm": 0.9991185805709686, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 34731 + }, + { + "epoch": 0.34732, + "grad_norm": 1.1896870077137653, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 34732 + }, + { + "epoch": 0.34733, + "grad_norm": 1.1055021961313989, + "learning_rate": 0.003, + "loss": 4.052, + "step": 34733 + }, + { + "epoch": 0.34734, + "grad_norm": 1.054490023720566, + "learning_rate": 0.003, + "loss": 4.028, + "step": 34734 + }, + { + "epoch": 0.34735, + "grad_norm": 1.1433198066236792, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 34735 + }, + { + "epoch": 0.34736, + "grad_norm": 1.1596022570672266, + "learning_rate": 0.003, + "loss": 4.035, + "step": 34736 + }, + { + "epoch": 0.34737, + "grad_norm": 0.9382925908759948, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 34737 + }, + { + "epoch": 0.34738, + "grad_norm": 1.0851488076621458, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 34738 + }, + { + "epoch": 0.34739, + "grad_norm": 1.2536975289627224, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 34739 + }, + { + "epoch": 0.3474, + "grad_norm": 0.9944587062978574, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 34740 + }, + { + "epoch": 0.34741, + "grad_norm": 1.090009736752007, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 34741 + }, + { + "epoch": 0.34742, + "grad_norm": 1.0841986226520197, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 34742 + }, + { + "epoch": 0.34743, + "grad_norm": 0.8700550033765636, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 34743 + }, + { + "epoch": 0.34744, + "grad_norm": 1.191535519094507, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 34744 + }, + { + "epoch": 0.34745, + "grad_norm": 1.046627402992397, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 34745 + }, + { + "epoch": 0.34746, + "grad_norm": 1.0787281440168552, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 34746 + }, + { + "epoch": 0.34747, + "grad_norm": 1.1725078061517111, + "learning_rate": 0.003, + "loss": 4.03, + "step": 34747 + }, + { + "epoch": 0.34748, + "grad_norm": 1.1667897598895256, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 34748 + }, + { + "epoch": 0.34749, + "grad_norm": 1.084945118254595, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 34749 + }, + { + "epoch": 0.3475, + "grad_norm": 1.2448993027642064, + "learning_rate": 0.003, + "loss": 4.045, + "step": 34750 + }, + { + "epoch": 0.34751, + "grad_norm": 1.0959685341454353, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 34751 + }, + { + "epoch": 0.34752, + "grad_norm": 1.2332519059413105, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 34752 + }, + { + "epoch": 0.34753, + "grad_norm": 1.0354024225991334, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 34753 + }, + { + "epoch": 0.34754, + "grad_norm": 1.0089292007110116, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 34754 + }, + { + "epoch": 0.34755, + "grad_norm": 1.360904063860321, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 34755 + }, + { + "epoch": 0.34756, + "grad_norm": 0.8546385209500669, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 34756 + }, + { + "epoch": 0.34757, + "grad_norm": 0.9181674390179626, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 34757 + }, + { + "epoch": 0.34758, + "grad_norm": 1.117931461897469, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 34758 + }, + { + "epoch": 0.34759, + "grad_norm": 0.9089324337125912, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 34759 + }, + { + "epoch": 0.3476, + "grad_norm": 1.0198892003911646, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 34760 + }, + { + "epoch": 0.34761, + "grad_norm": 1.223130537943272, + "learning_rate": 0.003, + "loss": 4.055, + "step": 34761 + }, + { + "epoch": 0.34762, + "grad_norm": 0.8666248017548354, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 34762 + }, + { + "epoch": 0.34763, + "grad_norm": 0.9851893452815864, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 34763 + }, + { + "epoch": 0.34764, + "grad_norm": 1.2180767968346908, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 34764 + }, + { + "epoch": 0.34765, + "grad_norm": 1.1366121806452132, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 34765 + }, + { + "epoch": 0.34766, + "grad_norm": 1.0346806828389314, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 34766 + }, + { + "epoch": 0.34767, + "grad_norm": 1.2481200617120554, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 34767 + }, + { + "epoch": 0.34768, + "grad_norm": 1.0941837600594535, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 34768 + }, + { + "epoch": 0.34769, + "grad_norm": 1.0911475598987603, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 34769 + }, + { + "epoch": 0.3477, + "grad_norm": 0.9921325362225656, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 34770 + }, + { + "epoch": 0.34771, + "grad_norm": 0.9215595238093056, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 34771 + }, + { + "epoch": 0.34772, + "grad_norm": 1.057387196850224, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 34772 + }, + { + "epoch": 0.34773, + "grad_norm": 1.305695311112206, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 34773 + }, + { + "epoch": 0.34774, + "grad_norm": 0.85730232527034, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 34774 + }, + { + "epoch": 0.34775, + "grad_norm": 0.9358441815587467, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 34775 + }, + { + "epoch": 0.34776, + "grad_norm": 0.978469134391879, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 34776 + }, + { + "epoch": 0.34777, + "grad_norm": 1.1746545836282127, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 34777 + }, + { + "epoch": 0.34778, + "grad_norm": 1.138517244888704, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 34778 + }, + { + "epoch": 0.34779, + "grad_norm": 1.1558183529729973, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 34779 + }, + { + "epoch": 0.3478, + "grad_norm": 0.9546644584192621, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 34780 + }, + { + "epoch": 0.34781, + "grad_norm": 1.250418333857631, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 34781 + }, + { + "epoch": 0.34782, + "grad_norm": 1.1291188929523608, + "learning_rate": 0.003, + "loss": 4.028, + "step": 34782 + }, + { + "epoch": 0.34783, + "grad_norm": 0.8840811213936768, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 34783 + }, + { + "epoch": 0.34784, + "grad_norm": 0.890721643290648, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 34784 + }, + { + "epoch": 0.34785, + "grad_norm": 1.0587912904982941, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 34785 + }, + { + "epoch": 0.34786, + "grad_norm": 0.987523775666879, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 34786 + }, + { + "epoch": 0.34787, + "grad_norm": 1.2459084051568226, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 34787 + }, + { + "epoch": 0.34788, + "grad_norm": 0.9049543723945214, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 34788 + }, + { + "epoch": 0.34789, + "grad_norm": 1.1671535926654635, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 34789 + }, + { + "epoch": 0.3479, + "grad_norm": 1.048914889242975, + "learning_rate": 0.003, + "loss": 4.046, + "step": 34790 + }, + { + "epoch": 0.34791, + "grad_norm": 1.4811299046989255, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 34791 + }, + { + "epoch": 0.34792, + "grad_norm": 1.0701551222433847, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 34792 + }, + { + "epoch": 0.34793, + "grad_norm": 1.034871069773209, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 34793 + }, + { + "epoch": 0.34794, + "grad_norm": 1.1434303029895934, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 34794 + }, + { + "epoch": 0.34795, + "grad_norm": 0.9414564136228174, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 34795 + }, + { + "epoch": 0.34796, + "grad_norm": 0.9757159076474735, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 34796 + }, + { + "epoch": 0.34797, + "grad_norm": 0.9857576572553661, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 34797 + }, + { + "epoch": 0.34798, + "grad_norm": 1.1862337669788918, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 34798 + }, + { + "epoch": 0.34799, + "grad_norm": 1.0699730181506006, + "learning_rate": 0.003, + "loss": 4.047, + "step": 34799 + }, + { + "epoch": 0.348, + "grad_norm": 1.0724879496273592, + "learning_rate": 0.003, + "loss": 4.045, + "step": 34800 + }, + { + "epoch": 0.34801, + "grad_norm": 1.1792456879300826, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 34801 + }, + { + "epoch": 0.34802, + "grad_norm": 1.0568341264080068, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 34802 + }, + { + "epoch": 0.34803, + "grad_norm": 1.281947538814292, + "learning_rate": 0.003, + "loss": 4.043, + "step": 34803 + }, + { + "epoch": 0.34804, + "grad_norm": 1.0155947818613684, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 34804 + }, + { + "epoch": 0.34805, + "grad_norm": 1.136167994830044, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 34805 + }, + { + "epoch": 0.34806, + "grad_norm": 1.0375287861965004, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 34806 + }, + { + "epoch": 0.34807, + "grad_norm": 1.1584603390448642, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 34807 + }, + { + "epoch": 0.34808, + "grad_norm": 1.2394464216112215, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 34808 + }, + { + "epoch": 0.34809, + "grad_norm": 1.0948558847674297, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 34809 + }, + { + "epoch": 0.3481, + "grad_norm": 1.025040378092018, + "learning_rate": 0.003, + "loss": 4.054, + "step": 34810 + }, + { + "epoch": 0.34811, + "grad_norm": 1.0791912048880596, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 34811 + }, + { + "epoch": 0.34812, + "grad_norm": 0.99959739365138, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 34812 + }, + { + "epoch": 0.34813, + "grad_norm": 1.1120046534566699, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 34813 + }, + { + "epoch": 0.34814, + "grad_norm": 1.1047660679860871, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 34814 + }, + { + "epoch": 0.34815, + "grad_norm": 1.0743191495600997, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 34815 + }, + { + "epoch": 0.34816, + "grad_norm": 0.9964843303689765, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 34816 + }, + { + "epoch": 0.34817, + "grad_norm": 1.19076264671751, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 34817 + }, + { + "epoch": 0.34818, + "grad_norm": 1.253220363495525, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 34818 + }, + { + "epoch": 0.34819, + "grad_norm": 1.0523583581116627, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 34819 + }, + { + "epoch": 0.3482, + "grad_norm": 1.0794298782900247, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 34820 + }, + { + "epoch": 0.34821, + "grad_norm": 1.0715578996472825, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 34821 + }, + { + "epoch": 0.34822, + "grad_norm": 1.1086243492325758, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 34822 + }, + { + "epoch": 0.34823, + "grad_norm": 1.0314212766069222, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 34823 + }, + { + "epoch": 0.34824, + "grad_norm": 1.024054410195294, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 34824 + }, + { + "epoch": 0.34825, + "grad_norm": 1.0255826271423676, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 34825 + }, + { + "epoch": 0.34826, + "grad_norm": 0.9989644943614341, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 34826 + }, + { + "epoch": 0.34827, + "grad_norm": 1.115376005795658, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 34827 + }, + { + "epoch": 0.34828, + "grad_norm": 1.1309838760394244, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 34828 + }, + { + "epoch": 0.34829, + "grad_norm": 1.2901261464982732, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 34829 + }, + { + "epoch": 0.3483, + "grad_norm": 0.9177278872557897, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 34830 + }, + { + "epoch": 0.34831, + "grad_norm": 1.005858287193552, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 34831 + }, + { + "epoch": 0.34832, + "grad_norm": 1.1803926501775641, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 34832 + }, + { + "epoch": 0.34833, + "grad_norm": 0.8282841605881793, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 34833 + }, + { + "epoch": 0.34834, + "grad_norm": 0.9799627152180356, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 34834 + }, + { + "epoch": 0.34835, + "grad_norm": 1.4732173223512166, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 34835 + }, + { + "epoch": 0.34836, + "grad_norm": 1.0291975997698808, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 34836 + }, + { + "epoch": 0.34837, + "grad_norm": 1.0031691465873664, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 34837 + }, + { + "epoch": 0.34838, + "grad_norm": 1.2584149275819498, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 34838 + }, + { + "epoch": 0.34839, + "grad_norm": 1.0534318797985682, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 34839 + }, + { + "epoch": 0.3484, + "grad_norm": 1.0073506391469176, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 34840 + }, + { + "epoch": 0.34841, + "grad_norm": 1.1230999120960334, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 34841 + }, + { + "epoch": 0.34842, + "grad_norm": 1.074384696873213, + "learning_rate": 0.003, + "loss": 4.043, + "step": 34842 + }, + { + "epoch": 0.34843, + "grad_norm": 1.005492547984754, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 34843 + }, + { + "epoch": 0.34844, + "grad_norm": 1.0986047466530935, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 34844 + }, + { + "epoch": 0.34845, + "grad_norm": 1.060897094364185, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 34845 + }, + { + "epoch": 0.34846, + "grad_norm": 1.1841995387926374, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 34846 + }, + { + "epoch": 0.34847, + "grad_norm": 0.954304422388616, + "learning_rate": 0.003, + "loss": 4.03, + "step": 34847 + }, + { + "epoch": 0.34848, + "grad_norm": 0.8391025862309616, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 34848 + }, + { + "epoch": 0.34849, + "grad_norm": 1.3342748058678475, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 34849 + }, + { + "epoch": 0.3485, + "grad_norm": 1.2002698955963316, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 34850 + }, + { + "epoch": 0.34851, + "grad_norm": 1.163090674866263, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 34851 + }, + { + "epoch": 0.34852, + "grad_norm": 0.9726968665418168, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 34852 + }, + { + "epoch": 0.34853, + "grad_norm": 1.0749624573284542, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 34853 + }, + { + "epoch": 0.34854, + "grad_norm": 1.0772331931045882, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 34854 + }, + { + "epoch": 0.34855, + "grad_norm": 1.084365804188474, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 34855 + }, + { + "epoch": 0.34856, + "grad_norm": 0.9035614418982033, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 34856 + }, + { + "epoch": 0.34857, + "grad_norm": 0.9547643965170278, + "learning_rate": 0.003, + "loss": 4.046, + "step": 34857 + }, + { + "epoch": 0.34858, + "grad_norm": 1.4368694761481966, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 34858 + }, + { + "epoch": 0.34859, + "grad_norm": 1.0020524971277178, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 34859 + }, + { + "epoch": 0.3486, + "grad_norm": 1.1894958224607834, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 34860 + }, + { + "epoch": 0.34861, + "grad_norm": 0.8661169151617757, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 34861 + }, + { + "epoch": 0.34862, + "grad_norm": 0.9110442948814188, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 34862 + }, + { + "epoch": 0.34863, + "grad_norm": 0.9925810430994272, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 34863 + }, + { + "epoch": 0.34864, + "grad_norm": 1.1768666171829685, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 34864 + }, + { + "epoch": 0.34865, + "grad_norm": 1.0937997653079694, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 34865 + }, + { + "epoch": 0.34866, + "grad_norm": 1.0562893391427288, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 34866 + }, + { + "epoch": 0.34867, + "grad_norm": 1.120941154863872, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 34867 + }, + { + "epoch": 0.34868, + "grad_norm": 1.1631145868672303, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 34868 + }, + { + "epoch": 0.34869, + "grad_norm": 0.9357059037969901, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 34869 + }, + { + "epoch": 0.3487, + "grad_norm": 1.162918734245065, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 34870 + }, + { + "epoch": 0.34871, + "grad_norm": 1.1758591120069095, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 34871 + }, + { + "epoch": 0.34872, + "grad_norm": 1.1639566418794975, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 34872 + }, + { + "epoch": 0.34873, + "grad_norm": 1.0691067552064641, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 34873 + }, + { + "epoch": 0.34874, + "grad_norm": 1.050369682140266, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 34874 + }, + { + "epoch": 0.34875, + "grad_norm": 1.0684484878189942, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 34875 + }, + { + "epoch": 0.34876, + "grad_norm": 0.9552029983519287, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 34876 + }, + { + "epoch": 0.34877, + "grad_norm": 1.0496326434682586, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 34877 + }, + { + "epoch": 0.34878, + "grad_norm": 0.9526344247279405, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 34878 + }, + { + "epoch": 0.34879, + "grad_norm": 1.27296950077819, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 34879 + }, + { + "epoch": 0.3488, + "grad_norm": 0.9329216647585216, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 34880 + }, + { + "epoch": 0.34881, + "grad_norm": 0.9371240931060444, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 34881 + }, + { + "epoch": 0.34882, + "grad_norm": 0.8942513430666733, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 34882 + }, + { + "epoch": 0.34883, + "grad_norm": 1.0598151899402144, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 34883 + }, + { + "epoch": 0.34884, + "grad_norm": 1.0911927228519844, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 34884 + }, + { + "epoch": 0.34885, + "grad_norm": 1.1125986716171357, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 34885 + }, + { + "epoch": 0.34886, + "grad_norm": 1.3130963950648513, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 34886 + }, + { + "epoch": 0.34887, + "grad_norm": 1.0186986438473258, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 34887 + }, + { + "epoch": 0.34888, + "grad_norm": 1.1788712854389718, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 34888 + }, + { + "epoch": 0.34889, + "grad_norm": 1.2027653559520366, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 34889 + }, + { + "epoch": 0.3489, + "grad_norm": 1.116648153187078, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 34890 + }, + { + "epoch": 0.34891, + "grad_norm": 0.9797575932452717, + "learning_rate": 0.003, + "loss": 4.045, + "step": 34891 + }, + { + "epoch": 0.34892, + "grad_norm": 1.2031380243012144, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 34892 + }, + { + "epoch": 0.34893, + "grad_norm": 1.1097482286105353, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 34893 + }, + { + "epoch": 0.34894, + "grad_norm": 1.1760475170938556, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 34894 + }, + { + "epoch": 0.34895, + "grad_norm": 0.9758649928261086, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 34895 + }, + { + "epoch": 0.34896, + "grad_norm": 1.0990275103257063, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 34896 + }, + { + "epoch": 0.34897, + "grad_norm": 1.0111818719044259, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 34897 + }, + { + "epoch": 0.34898, + "grad_norm": 0.9969985123973375, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 34898 + }, + { + "epoch": 0.34899, + "grad_norm": 1.0654311295825538, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 34899 + }, + { + "epoch": 0.349, + "grad_norm": 1.1656158199936033, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 34900 + }, + { + "epoch": 0.34901, + "grad_norm": 0.9953157108353846, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 34901 + }, + { + "epoch": 0.34902, + "grad_norm": 1.253262297040725, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 34902 + }, + { + "epoch": 0.34903, + "grad_norm": 1.0541828066702326, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 34903 + }, + { + "epoch": 0.34904, + "grad_norm": 1.1525139545467464, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 34904 + }, + { + "epoch": 0.34905, + "grad_norm": 0.9066906177954452, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 34905 + }, + { + "epoch": 0.34906, + "grad_norm": 0.9829334200609329, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 34906 + }, + { + "epoch": 0.34907, + "grad_norm": 1.1340179842645681, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 34907 + }, + { + "epoch": 0.34908, + "grad_norm": 0.9955664496999502, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 34908 + }, + { + "epoch": 0.34909, + "grad_norm": 1.2090763312825532, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 34909 + }, + { + "epoch": 0.3491, + "grad_norm": 1.1771426699028464, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 34910 + }, + { + "epoch": 0.34911, + "grad_norm": 1.1036470224455803, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 34911 + }, + { + "epoch": 0.34912, + "grad_norm": 0.9088700003263573, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 34912 + }, + { + "epoch": 0.34913, + "grad_norm": 0.7595297839220525, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 34913 + }, + { + "epoch": 0.34914, + "grad_norm": 0.9244154495064038, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 34914 + }, + { + "epoch": 0.34915, + "grad_norm": 1.2211746510109662, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 34915 + }, + { + "epoch": 0.34916, + "grad_norm": 0.945519785229097, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 34916 + }, + { + "epoch": 0.34917, + "grad_norm": 1.0958490868790054, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 34917 + }, + { + "epoch": 0.34918, + "grad_norm": 1.398428267990295, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 34918 + }, + { + "epoch": 0.34919, + "grad_norm": 1.0415485239613615, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 34919 + }, + { + "epoch": 0.3492, + "grad_norm": 1.0574679241718137, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 34920 + }, + { + "epoch": 0.34921, + "grad_norm": 1.1217960448262367, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 34921 + }, + { + "epoch": 0.34922, + "grad_norm": 0.9631356083567862, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 34922 + }, + { + "epoch": 0.34923, + "grad_norm": 1.3378312440762694, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 34923 + }, + { + "epoch": 0.34924, + "grad_norm": 1.0612973502679095, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 34924 + }, + { + "epoch": 0.34925, + "grad_norm": 1.0520884231872456, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 34925 + }, + { + "epoch": 0.34926, + "grad_norm": 1.1134346851173091, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 34926 + }, + { + "epoch": 0.34927, + "grad_norm": 1.1141031587118793, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 34927 + }, + { + "epoch": 0.34928, + "grad_norm": 0.9338482949239286, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 34928 + }, + { + "epoch": 0.34929, + "grad_norm": 1.0542834491318842, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 34929 + }, + { + "epoch": 0.3493, + "grad_norm": 1.2352887219424176, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 34930 + }, + { + "epoch": 0.34931, + "grad_norm": 1.0802885385718088, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 34931 + }, + { + "epoch": 0.34932, + "grad_norm": 1.024383002083958, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 34932 + }, + { + "epoch": 0.34933, + "grad_norm": 0.9169055304240463, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 34933 + }, + { + "epoch": 0.34934, + "grad_norm": 1.053168354498545, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 34934 + }, + { + "epoch": 0.34935, + "grad_norm": 1.1858556348986466, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 34935 + }, + { + "epoch": 0.34936, + "grad_norm": 1.0685373369818754, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 34936 + }, + { + "epoch": 0.34937, + "grad_norm": 1.2279693797326223, + "learning_rate": 0.003, + "loss": 4.04, + "step": 34937 + }, + { + "epoch": 0.34938, + "grad_norm": 1.165253302847199, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 34938 + }, + { + "epoch": 0.34939, + "grad_norm": 1.15030259782111, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 34939 + }, + { + "epoch": 0.3494, + "grad_norm": 1.0175447063789653, + "learning_rate": 0.003, + "loss": 4.039, + "step": 34940 + }, + { + "epoch": 0.34941, + "grad_norm": 1.2417201520616277, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 34941 + }, + { + "epoch": 0.34942, + "grad_norm": 0.9102940080600476, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 34942 + }, + { + "epoch": 0.34943, + "grad_norm": 1.076124747442075, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 34943 + }, + { + "epoch": 0.34944, + "grad_norm": 1.6291120910509882, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 34944 + }, + { + "epoch": 0.34945, + "grad_norm": 0.696772706655121, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 34945 + }, + { + "epoch": 0.34946, + "grad_norm": 0.9283705959877883, + "learning_rate": 0.003, + "loss": 4.048, + "step": 34946 + }, + { + "epoch": 0.34947, + "grad_norm": 1.3757302431910496, + "learning_rate": 0.003, + "loss": 4.031, + "step": 34947 + }, + { + "epoch": 0.34948, + "grad_norm": 1.1071872032940897, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 34948 + }, + { + "epoch": 0.34949, + "grad_norm": 1.1396051400140284, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 34949 + }, + { + "epoch": 0.3495, + "grad_norm": 1.0552441735401838, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 34950 + }, + { + "epoch": 0.34951, + "grad_norm": 1.0500327629853279, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 34951 + }, + { + "epoch": 0.34952, + "grad_norm": 1.2591470894484371, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 34952 + }, + { + "epoch": 0.34953, + "grad_norm": 1.0167863762638008, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 34953 + }, + { + "epoch": 0.34954, + "grad_norm": 1.2345131511599485, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 34954 + }, + { + "epoch": 0.34955, + "grad_norm": 0.8331934436996794, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 34955 + }, + { + "epoch": 0.34956, + "grad_norm": 0.9397647249075318, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 34956 + }, + { + "epoch": 0.34957, + "grad_norm": 0.8517362433522978, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 34957 + }, + { + "epoch": 0.34958, + "grad_norm": 1.0155690554647745, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 34958 + }, + { + "epoch": 0.34959, + "grad_norm": 1.1845533135158361, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 34959 + }, + { + "epoch": 0.3496, + "grad_norm": 1.0440790245243832, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 34960 + }, + { + "epoch": 0.34961, + "grad_norm": 1.1293938530599557, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 34961 + }, + { + "epoch": 0.34962, + "grad_norm": 0.9453417376935052, + "learning_rate": 0.003, + "loss": 4.042, + "step": 34962 + }, + { + "epoch": 0.34963, + "grad_norm": 1.0225542937029557, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 34963 + }, + { + "epoch": 0.34964, + "grad_norm": 1.2968633111303114, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 34964 + }, + { + "epoch": 0.34965, + "grad_norm": 0.9678719415533693, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 34965 + }, + { + "epoch": 0.34966, + "grad_norm": 1.1650386002117532, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 34966 + }, + { + "epoch": 0.34967, + "grad_norm": 1.0140083712252566, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 34967 + }, + { + "epoch": 0.34968, + "grad_norm": 1.261834411536581, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 34968 + }, + { + "epoch": 0.34969, + "grad_norm": 1.596290388134109, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 34969 + }, + { + "epoch": 0.3497, + "grad_norm": 0.8842224225388544, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 34970 + }, + { + "epoch": 0.34971, + "grad_norm": 0.9297658578321728, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 34971 + }, + { + "epoch": 0.34972, + "grad_norm": 1.0781962600123292, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 34972 + }, + { + "epoch": 0.34973, + "grad_norm": 1.2543989146401873, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 34973 + }, + { + "epoch": 0.34974, + "grad_norm": 1.2287599695081357, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 34974 + }, + { + "epoch": 0.34975, + "grad_norm": 0.8442278373363744, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 34975 + }, + { + "epoch": 0.34976, + "grad_norm": 0.788552060031113, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 34976 + }, + { + "epoch": 0.34977, + "grad_norm": 0.8503291335991351, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 34977 + }, + { + "epoch": 0.34978, + "grad_norm": 0.8837215660616383, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 34978 + }, + { + "epoch": 0.34979, + "grad_norm": 0.9951794557055391, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 34979 + }, + { + "epoch": 0.3498, + "grad_norm": 1.4362856163053561, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 34980 + }, + { + "epoch": 0.34981, + "grad_norm": 1.0126131419214355, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 34981 + }, + { + "epoch": 0.34982, + "grad_norm": 1.3368958730977556, + "learning_rate": 0.003, + "loss": 4.029, + "step": 34982 + }, + { + "epoch": 0.34983, + "grad_norm": 0.9999802400686244, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 34983 + }, + { + "epoch": 0.34984, + "grad_norm": 1.0219793086549693, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 34984 + }, + { + "epoch": 0.34985, + "grad_norm": 1.0475037582357292, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 34985 + }, + { + "epoch": 0.34986, + "grad_norm": 1.3158155333229404, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 34986 + }, + { + "epoch": 0.34987, + "grad_norm": 1.023528803737236, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 34987 + }, + { + "epoch": 0.34988, + "grad_norm": 1.233205372519872, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 34988 + }, + { + "epoch": 0.34989, + "grad_norm": 1.097926080681261, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 34989 + }, + { + "epoch": 0.3499, + "grad_norm": 1.099999301534502, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 34990 + }, + { + "epoch": 0.34991, + "grad_norm": 0.9957698291304612, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 34991 + }, + { + "epoch": 0.34992, + "grad_norm": 1.1789521258183906, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 34992 + }, + { + "epoch": 0.34993, + "grad_norm": 1.0457388390117954, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 34993 + }, + { + "epoch": 0.34994, + "grad_norm": 1.022404144514737, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 34994 + }, + { + "epoch": 0.34995, + "grad_norm": 1.032281012234823, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 34995 + }, + { + "epoch": 0.34996, + "grad_norm": 1.0749179288137634, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 34996 + }, + { + "epoch": 0.34997, + "grad_norm": 1.0107382921120285, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 34997 + }, + { + "epoch": 0.34998, + "grad_norm": 1.1563647916680568, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 34998 + }, + { + "epoch": 0.34999, + "grad_norm": 0.9738191232305603, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 34999 + }, + { + "epoch": 0.35, + "grad_norm": 1.1283647179198286, + "learning_rate": 0.003, + "loss": 4.054, + "step": 35000 + }, + { + "epoch": 0.35001, + "grad_norm": 1.271032180227936, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 35001 + }, + { + "epoch": 0.35002, + "grad_norm": 1.1166282373410503, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 35002 + }, + { + "epoch": 0.35003, + "grad_norm": 1.1419841035318852, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 35003 + }, + { + "epoch": 0.35004, + "grad_norm": 1.1757950485024518, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 35004 + }, + { + "epoch": 0.35005, + "grad_norm": 1.3581668744581463, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 35005 + }, + { + "epoch": 0.35006, + "grad_norm": 0.6883960846535727, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 35006 + }, + { + "epoch": 0.35007, + "grad_norm": 0.8431748345647658, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 35007 + }, + { + "epoch": 0.35008, + "grad_norm": 1.1424996261168223, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 35008 + }, + { + "epoch": 0.35009, + "grad_norm": 1.450697957538779, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 35009 + }, + { + "epoch": 0.3501, + "grad_norm": 0.8056549896793421, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 35010 + }, + { + "epoch": 0.35011, + "grad_norm": 0.7874777345805947, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 35011 + }, + { + "epoch": 0.35012, + "grad_norm": 1.001527698064804, + "learning_rate": 0.003, + "loss": 4.03, + "step": 35012 + }, + { + "epoch": 0.35013, + "grad_norm": 1.2457573366044499, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 35013 + }, + { + "epoch": 0.35014, + "grad_norm": 0.7063488098854339, + "learning_rate": 0.003, + "loss": 4.029, + "step": 35014 + }, + { + "epoch": 0.35015, + "grad_norm": 0.7663635285597418, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 35015 + }, + { + "epoch": 0.35016, + "grad_norm": 0.9790121142535203, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 35016 + }, + { + "epoch": 0.35017, + "grad_norm": 1.2614180683574672, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 35017 + }, + { + "epoch": 0.35018, + "grad_norm": 0.8483414823683842, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 35018 + }, + { + "epoch": 0.35019, + "grad_norm": 0.9319498756666376, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 35019 + }, + { + "epoch": 0.3502, + "grad_norm": 0.9961856876553982, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 35020 + }, + { + "epoch": 0.35021, + "grad_norm": 1.1582980816656359, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 35021 + }, + { + "epoch": 0.35022, + "grad_norm": 1.2043578103368353, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 35022 + }, + { + "epoch": 0.35023, + "grad_norm": 1.1271597150695927, + "learning_rate": 0.003, + "loss": 4.039, + "step": 35023 + }, + { + "epoch": 0.35024, + "grad_norm": 1.1778884950444255, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 35024 + }, + { + "epoch": 0.35025, + "grad_norm": 1.3246144392702102, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 35025 + }, + { + "epoch": 0.35026, + "grad_norm": 0.9241920313836272, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 35026 + }, + { + "epoch": 0.35027, + "grad_norm": 0.9471816550819787, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 35027 + }, + { + "epoch": 0.35028, + "grad_norm": 1.3932360755995046, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 35028 + }, + { + "epoch": 0.35029, + "grad_norm": 1.0213535625539256, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 35029 + }, + { + "epoch": 0.3503, + "grad_norm": 1.157617490162696, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 35030 + }, + { + "epoch": 0.35031, + "grad_norm": 1.0170977356157622, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 35031 + }, + { + "epoch": 0.35032, + "grad_norm": 1.1167325645619368, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 35032 + }, + { + "epoch": 0.35033, + "grad_norm": 0.9368800987827076, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 35033 + }, + { + "epoch": 0.35034, + "grad_norm": 1.0753796264407522, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 35034 + }, + { + "epoch": 0.35035, + "grad_norm": 1.1027229176085176, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 35035 + }, + { + "epoch": 0.35036, + "grad_norm": 1.0867131467930937, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 35036 + }, + { + "epoch": 0.35037, + "grad_norm": 1.204349151014532, + "learning_rate": 0.003, + "loss": 4.033, + "step": 35037 + }, + { + "epoch": 0.35038, + "grad_norm": 1.1461872431723628, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 35038 + }, + { + "epoch": 0.35039, + "grad_norm": 1.2818123340358631, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 35039 + }, + { + "epoch": 0.3504, + "grad_norm": 1.1438211136737084, + "learning_rate": 0.003, + "loss": 4.041, + "step": 35040 + }, + { + "epoch": 0.35041, + "grad_norm": 0.838022461377939, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 35041 + }, + { + "epoch": 0.35042, + "grad_norm": 0.8396253942192884, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 35042 + }, + { + "epoch": 0.35043, + "grad_norm": 1.0448695708750957, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 35043 + }, + { + "epoch": 0.35044, + "grad_norm": 1.1127133792302062, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 35044 + }, + { + "epoch": 0.35045, + "grad_norm": 1.2266018587587355, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 35045 + }, + { + "epoch": 0.35046, + "grad_norm": 1.097758722286596, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 35046 + }, + { + "epoch": 0.35047, + "grad_norm": 1.01870735145425, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 35047 + }, + { + "epoch": 0.35048, + "grad_norm": 1.258218815750722, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 35048 + }, + { + "epoch": 0.35049, + "grad_norm": 1.1975489191846764, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 35049 + }, + { + "epoch": 0.3505, + "grad_norm": 0.8943671184500396, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 35050 + }, + { + "epoch": 0.35051, + "grad_norm": 0.8815344890236573, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 35051 + }, + { + "epoch": 0.35052, + "grad_norm": 1.0833910658236008, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 35052 + }, + { + "epoch": 0.35053, + "grad_norm": 1.267651898559493, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 35053 + }, + { + "epoch": 0.35054, + "grad_norm": 1.0450239995962582, + "learning_rate": 0.003, + "loss": 4.021, + "step": 35054 + }, + { + "epoch": 0.35055, + "grad_norm": 1.2038361081450064, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 35055 + }, + { + "epoch": 0.35056, + "grad_norm": 1.0447346125699493, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 35056 + }, + { + "epoch": 0.35057, + "grad_norm": 0.9532045548670321, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 35057 + }, + { + "epoch": 0.35058, + "grad_norm": 1.016087577708776, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 35058 + }, + { + "epoch": 0.35059, + "grad_norm": 1.0717822126169374, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 35059 + }, + { + "epoch": 0.3506, + "grad_norm": 1.0467001640514257, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 35060 + }, + { + "epoch": 0.35061, + "grad_norm": 1.225434080736295, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 35061 + }, + { + "epoch": 0.35062, + "grad_norm": 0.9002183388410356, + "learning_rate": 0.003, + "loss": 4.047, + "step": 35062 + }, + { + "epoch": 0.35063, + "grad_norm": 1.0986528570760983, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 35063 + }, + { + "epoch": 0.35064, + "grad_norm": 1.3578088853619852, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 35064 + }, + { + "epoch": 0.35065, + "grad_norm": 1.0435388718872396, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 35065 + }, + { + "epoch": 0.35066, + "grad_norm": 1.186223721833528, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 35066 + }, + { + "epoch": 0.35067, + "grad_norm": 1.425613463485605, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 35067 + }, + { + "epoch": 0.35068, + "grad_norm": 1.0736272971866354, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 35068 + }, + { + "epoch": 0.35069, + "grad_norm": 1.4619377329331085, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 35069 + }, + { + "epoch": 0.3507, + "grad_norm": 0.7385082525636367, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 35070 + }, + { + "epoch": 0.35071, + "grad_norm": 0.7785258763740139, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 35071 + }, + { + "epoch": 0.35072, + "grad_norm": 1.1554810768320019, + "learning_rate": 0.003, + "loss": 4.024, + "step": 35072 + }, + { + "epoch": 0.35073, + "grad_norm": 1.2126828876950548, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 35073 + }, + { + "epoch": 0.35074, + "grad_norm": 1.2637651779961145, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 35074 + }, + { + "epoch": 0.35075, + "grad_norm": 0.8611486283719374, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 35075 + }, + { + "epoch": 0.35076, + "grad_norm": 0.8787512496475968, + "learning_rate": 0.003, + "loss": 4.023, + "step": 35076 + }, + { + "epoch": 0.35077, + "grad_norm": 1.176013726623187, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 35077 + }, + { + "epoch": 0.35078, + "grad_norm": 1.1618584949240907, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 35078 + }, + { + "epoch": 0.35079, + "grad_norm": 1.0451189825578677, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 35079 + }, + { + "epoch": 0.3508, + "grad_norm": 1.0178547347029245, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 35080 + }, + { + "epoch": 0.35081, + "grad_norm": 1.2016422839152607, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 35081 + }, + { + "epoch": 0.35082, + "grad_norm": 0.9708403641186586, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 35082 + }, + { + "epoch": 0.35083, + "grad_norm": 1.0827433601490186, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 35083 + }, + { + "epoch": 0.35084, + "grad_norm": 1.0176584027357694, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 35084 + }, + { + "epoch": 0.35085, + "grad_norm": 1.1802275777214946, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 35085 + }, + { + "epoch": 0.35086, + "grad_norm": 1.2024731185019668, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 35086 + }, + { + "epoch": 0.35087, + "grad_norm": 0.9966029546319225, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 35087 + }, + { + "epoch": 0.35088, + "grad_norm": 1.0601574061593348, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 35088 + }, + { + "epoch": 0.35089, + "grad_norm": 1.1599740242431245, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 35089 + }, + { + "epoch": 0.3509, + "grad_norm": 0.9208370531604236, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 35090 + }, + { + "epoch": 0.35091, + "grad_norm": 0.9832725411045101, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 35091 + }, + { + "epoch": 0.35092, + "grad_norm": 1.4921305358733432, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 35092 + }, + { + "epoch": 0.35093, + "grad_norm": 0.8308108182412794, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 35093 + }, + { + "epoch": 0.35094, + "grad_norm": 0.8723726774935318, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 35094 + }, + { + "epoch": 0.35095, + "grad_norm": 1.1120713585749182, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 35095 + }, + { + "epoch": 0.35096, + "grad_norm": 1.3144088467799537, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 35096 + }, + { + "epoch": 0.35097, + "grad_norm": 0.9147852210886204, + "learning_rate": 0.003, + "loss": 4.045, + "step": 35097 + }, + { + "epoch": 0.35098, + "grad_norm": 1.1207186935968663, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 35098 + }, + { + "epoch": 0.35099, + "grad_norm": 1.190160336990496, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 35099 + }, + { + "epoch": 0.351, + "grad_norm": 1.059609689007313, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 35100 + }, + { + "epoch": 0.35101, + "grad_norm": 0.9697766875775828, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 35101 + }, + { + "epoch": 0.35102, + "grad_norm": 0.8506823639044674, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 35102 + }, + { + "epoch": 0.35103, + "grad_norm": 0.8398752129288076, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 35103 + }, + { + "epoch": 0.35104, + "grad_norm": 1.3122592762190064, + "learning_rate": 0.003, + "loss": 4.056, + "step": 35104 + }, + { + "epoch": 0.35105, + "grad_norm": 1.1144661877381552, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 35105 + }, + { + "epoch": 0.35106, + "grad_norm": 1.0329149144863496, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 35106 + }, + { + "epoch": 0.35107, + "grad_norm": 1.139982925965349, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 35107 + }, + { + "epoch": 0.35108, + "grad_norm": 1.052185787040762, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 35108 + }, + { + "epoch": 0.35109, + "grad_norm": 1.2224842978241168, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 35109 + }, + { + "epoch": 0.3511, + "grad_norm": 1.167926720897326, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 35110 + }, + { + "epoch": 0.35111, + "grad_norm": 1.0382958489198475, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 35111 + }, + { + "epoch": 0.35112, + "grad_norm": 1.0728942945109674, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 35112 + }, + { + "epoch": 0.35113, + "grad_norm": 1.207541922546702, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 35113 + }, + { + "epoch": 0.35114, + "grad_norm": 0.9657811065071282, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 35114 + }, + { + "epoch": 0.35115, + "grad_norm": 1.16036676816592, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 35115 + }, + { + "epoch": 0.35116, + "grad_norm": 1.1087649439176714, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 35116 + }, + { + "epoch": 0.35117, + "grad_norm": 1.198055463074635, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 35117 + }, + { + "epoch": 0.35118, + "grad_norm": 0.9219352802601745, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 35118 + }, + { + "epoch": 0.35119, + "grad_norm": 1.1374163620923539, + "learning_rate": 0.003, + "loss": 4.028, + "step": 35119 + }, + { + "epoch": 0.3512, + "grad_norm": 1.159451524013875, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 35120 + }, + { + "epoch": 0.35121, + "grad_norm": 0.9915462434331676, + "learning_rate": 0.003, + "loss": 4.032, + "step": 35121 + }, + { + "epoch": 0.35122, + "grad_norm": 1.0319716351719304, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 35122 + }, + { + "epoch": 0.35123, + "grad_norm": 1.2800455390364356, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 35123 + }, + { + "epoch": 0.35124, + "grad_norm": 1.1029323121762102, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 35124 + }, + { + "epoch": 0.35125, + "grad_norm": 1.069046400129774, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 35125 + }, + { + "epoch": 0.35126, + "grad_norm": 1.0454256487675542, + "learning_rate": 0.003, + "loss": 4.048, + "step": 35126 + }, + { + "epoch": 0.35127, + "grad_norm": 1.2641936099713933, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 35127 + }, + { + "epoch": 0.35128, + "grad_norm": 1.014657878072598, + "learning_rate": 0.003, + "loss": 4.045, + "step": 35128 + }, + { + "epoch": 0.35129, + "grad_norm": 1.3289752830139816, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 35129 + }, + { + "epoch": 0.3513, + "grad_norm": 0.9969305027967421, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 35130 + }, + { + "epoch": 0.35131, + "grad_norm": 1.2082352174426854, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 35131 + }, + { + "epoch": 0.35132, + "grad_norm": 1.047681387689566, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 35132 + }, + { + "epoch": 0.35133, + "grad_norm": 0.994379475961578, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 35133 + }, + { + "epoch": 0.35134, + "grad_norm": 1.058842759409239, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 35134 + }, + { + "epoch": 0.35135, + "grad_norm": 1.3766218879394925, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 35135 + }, + { + "epoch": 0.35136, + "grad_norm": 0.7379673273748399, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 35136 + }, + { + "epoch": 0.35137, + "grad_norm": 0.8473878868787658, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 35137 + }, + { + "epoch": 0.35138, + "grad_norm": 1.0582085918261093, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 35138 + }, + { + "epoch": 0.35139, + "grad_norm": 1.204555999152817, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 35139 + }, + { + "epoch": 0.3514, + "grad_norm": 1.0011343148830634, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 35140 + }, + { + "epoch": 0.35141, + "grad_norm": 1.332015180428653, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 35141 + }, + { + "epoch": 0.35142, + "grad_norm": 0.9867380846703625, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 35142 + }, + { + "epoch": 0.35143, + "grad_norm": 1.0760310812789597, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 35143 + }, + { + "epoch": 0.35144, + "grad_norm": 1.1122305478793315, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 35144 + }, + { + "epoch": 0.35145, + "grad_norm": 1.2057348411596092, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 35145 + }, + { + "epoch": 0.35146, + "grad_norm": 1.1194037474568879, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 35146 + }, + { + "epoch": 0.35147, + "grad_norm": 1.151065767636874, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 35147 + }, + { + "epoch": 0.35148, + "grad_norm": 1.0833365882621588, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 35148 + }, + { + "epoch": 0.35149, + "grad_norm": 1.214056391679558, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 35149 + }, + { + "epoch": 0.3515, + "grad_norm": 1.134241831719395, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 35150 + }, + { + "epoch": 0.35151, + "grad_norm": 1.211333720094181, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 35151 + }, + { + "epoch": 0.35152, + "grad_norm": 1.0701992623544587, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 35152 + }, + { + "epoch": 0.35153, + "grad_norm": 1.1845046036166191, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 35153 + }, + { + "epoch": 0.35154, + "grad_norm": 0.8646310450883632, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 35154 + }, + { + "epoch": 0.35155, + "grad_norm": 0.772189836447939, + "learning_rate": 0.003, + "loss": 4.044, + "step": 35155 + }, + { + "epoch": 0.35156, + "grad_norm": 0.8554972515571767, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 35156 + }, + { + "epoch": 0.35157, + "grad_norm": 0.8853933269814118, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 35157 + }, + { + "epoch": 0.35158, + "grad_norm": 0.9348777542565745, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 35158 + }, + { + "epoch": 0.35159, + "grad_norm": 1.2988766108362497, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 35159 + }, + { + "epoch": 0.3516, + "grad_norm": 1.0820515330492195, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 35160 + }, + { + "epoch": 0.35161, + "grad_norm": 1.2822156459028669, + "learning_rate": 0.003, + "loss": 4.013, + "step": 35161 + }, + { + "epoch": 0.35162, + "grad_norm": 1.040116990837634, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 35162 + }, + { + "epoch": 0.35163, + "grad_norm": 1.164096208015248, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 35163 + }, + { + "epoch": 0.35164, + "grad_norm": 0.987741090917036, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 35164 + }, + { + "epoch": 0.35165, + "grad_norm": 1.5006088853013717, + "learning_rate": 0.003, + "loss": 4.025, + "step": 35165 + }, + { + "epoch": 0.35166, + "grad_norm": 0.9168479623543114, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 35166 + }, + { + "epoch": 0.35167, + "grad_norm": 1.0065017166197339, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 35167 + }, + { + "epoch": 0.35168, + "grad_norm": 1.2939808931296977, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 35168 + }, + { + "epoch": 0.35169, + "grad_norm": 0.9951749316020146, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 35169 + }, + { + "epoch": 0.3517, + "grad_norm": 1.3602642572921608, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 35170 + }, + { + "epoch": 0.35171, + "grad_norm": 0.8997517927960679, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 35171 + }, + { + "epoch": 0.35172, + "grad_norm": 1.0475646063260675, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 35172 + }, + { + "epoch": 0.35173, + "grad_norm": 1.212333852716913, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 35173 + }, + { + "epoch": 0.35174, + "grad_norm": 0.9951859741533411, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 35174 + }, + { + "epoch": 0.35175, + "grad_norm": 0.9990186719718984, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 35175 + }, + { + "epoch": 0.35176, + "grad_norm": 1.0097635981604949, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 35176 + }, + { + "epoch": 0.35177, + "grad_norm": 1.170251519958874, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 35177 + }, + { + "epoch": 0.35178, + "grad_norm": 1.1706873789566743, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 35178 + }, + { + "epoch": 0.35179, + "grad_norm": 1.1230458317423806, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 35179 + }, + { + "epoch": 0.3518, + "grad_norm": 1.0080730905747046, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 35180 + }, + { + "epoch": 0.35181, + "grad_norm": 1.0719307263701559, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 35181 + }, + { + "epoch": 0.35182, + "grad_norm": 1.256454197776436, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 35182 + }, + { + "epoch": 0.35183, + "grad_norm": 1.0379189159821491, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 35183 + }, + { + "epoch": 0.35184, + "grad_norm": 1.1254431057333498, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 35184 + }, + { + "epoch": 0.35185, + "grad_norm": 1.0319141235863258, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 35185 + }, + { + "epoch": 0.35186, + "grad_norm": 1.0840023786102184, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 35186 + }, + { + "epoch": 0.35187, + "grad_norm": 0.9887992263078056, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 35187 + }, + { + "epoch": 0.35188, + "grad_norm": 1.10527547526829, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 35188 + }, + { + "epoch": 0.35189, + "grad_norm": 1.0958906154868648, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 35189 + }, + { + "epoch": 0.3519, + "grad_norm": 1.0982169306289054, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 35190 + }, + { + "epoch": 0.35191, + "grad_norm": 1.205069469210958, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 35191 + }, + { + "epoch": 0.35192, + "grad_norm": 1.0101675760808577, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 35192 + }, + { + "epoch": 0.35193, + "grad_norm": 1.167632680655999, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 35193 + }, + { + "epoch": 0.35194, + "grad_norm": 1.0081597046770903, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 35194 + }, + { + "epoch": 0.35195, + "grad_norm": 1.256878290819856, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 35195 + }, + { + "epoch": 0.35196, + "grad_norm": 1.1044774749022568, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 35196 + }, + { + "epoch": 0.35197, + "grad_norm": 1.0515752715733577, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 35197 + }, + { + "epoch": 0.35198, + "grad_norm": 0.9722149698067419, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 35198 + }, + { + "epoch": 0.35199, + "grad_norm": 1.1441981389666784, + "learning_rate": 0.003, + "loss": 4.058, + "step": 35199 + }, + { + "epoch": 0.352, + "grad_norm": 0.9862806024877852, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 35200 + }, + { + "epoch": 0.35201, + "grad_norm": 1.5837295318653775, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 35201 + }, + { + "epoch": 0.35202, + "grad_norm": 1.0300592976938379, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 35202 + }, + { + "epoch": 0.35203, + "grad_norm": 1.3636037601645281, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 35203 + }, + { + "epoch": 0.35204, + "grad_norm": 1.0056549404728066, + "learning_rate": 0.003, + "loss": 4.0934, + "step": 35204 + }, + { + "epoch": 0.35205, + "grad_norm": 1.111592368875962, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 35205 + }, + { + "epoch": 0.35206, + "grad_norm": 1.0315021314695483, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 35206 + }, + { + "epoch": 0.35207, + "grad_norm": 1.3151876611600766, + "learning_rate": 0.003, + "loss": 4.0959, + "step": 35207 + }, + { + "epoch": 0.35208, + "grad_norm": 0.9492444372708225, + "learning_rate": 0.003, + "loss": 4.04, + "step": 35208 + }, + { + "epoch": 0.35209, + "grad_norm": 1.0108183455673778, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 35209 + }, + { + "epoch": 0.3521, + "grad_norm": 1.5379851319996933, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 35210 + }, + { + "epoch": 0.35211, + "grad_norm": 0.8530075660512072, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 35211 + }, + { + "epoch": 0.35212, + "grad_norm": 0.7006348736751017, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 35212 + }, + { + "epoch": 0.35213, + "grad_norm": 0.7299760551742707, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 35213 + }, + { + "epoch": 0.35214, + "grad_norm": 0.9685208780789366, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 35214 + }, + { + "epoch": 0.35215, + "grad_norm": 1.311698915261674, + "learning_rate": 0.003, + "loss": 4.0864, + "step": 35215 + }, + { + "epoch": 0.35216, + "grad_norm": 0.9773469390322586, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 35216 + }, + { + "epoch": 0.35217, + "grad_norm": 1.2622231557128458, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 35217 + }, + { + "epoch": 0.35218, + "grad_norm": 0.9031084415983675, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 35218 + }, + { + "epoch": 0.35219, + "grad_norm": 0.8747351862626436, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 35219 + }, + { + "epoch": 0.3522, + "grad_norm": 1.0959941025225437, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 35220 + }, + { + "epoch": 0.35221, + "grad_norm": 1.1666661181841975, + "learning_rate": 0.003, + "loss": 4.052, + "step": 35221 + }, + { + "epoch": 0.35222, + "grad_norm": 1.123602785142103, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 35222 + }, + { + "epoch": 0.35223, + "grad_norm": 1.0722078851218133, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 35223 + }, + { + "epoch": 0.35224, + "grad_norm": 0.9390960544433561, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 35224 + }, + { + "epoch": 0.35225, + "grad_norm": 1.062722459408616, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 35225 + }, + { + "epoch": 0.35226, + "grad_norm": 1.1833871854221512, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 35226 + }, + { + "epoch": 0.35227, + "grad_norm": 1.0237104131221555, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 35227 + }, + { + "epoch": 0.35228, + "grad_norm": 1.1389896562221802, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 35228 + }, + { + "epoch": 0.35229, + "grad_norm": 1.1106716598363129, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 35229 + }, + { + "epoch": 0.3523, + "grad_norm": 1.0274671361630667, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 35230 + }, + { + "epoch": 0.35231, + "grad_norm": 1.031899742079587, + "learning_rate": 0.003, + "loss": 4.022, + "step": 35231 + }, + { + "epoch": 0.35232, + "grad_norm": 1.0975747277864787, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 35232 + }, + { + "epoch": 0.35233, + "grad_norm": 1.15767239278544, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 35233 + }, + { + "epoch": 0.35234, + "grad_norm": 0.8804914068073882, + "learning_rate": 0.003, + "loss": 4.055, + "step": 35234 + }, + { + "epoch": 0.35235, + "grad_norm": 0.995863727000329, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 35235 + }, + { + "epoch": 0.35236, + "grad_norm": 1.284854656175508, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 35236 + }, + { + "epoch": 0.35237, + "grad_norm": 1.099148583114505, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 35237 + }, + { + "epoch": 0.35238, + "grad_norm": 1.1373278340751558, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 35238 + }, + { + "epoch": 0.35239, + "grad_norm": 0.9954833035395896, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 35239 + }, + { + "epoch": 0.3524, + "grad_norm": 1.1244632308945353, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 35240 + }, + { + "epoch": 0.35241, + "grad_norm": 1.22486487011655, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 35241 + }, + { + "epoch": 0.35242, + "grad_norm": 1.0085881271154025, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 35242 + }, + { + "epoch": 0.35243, + "grad_norm": 1.1496177679429929, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 35243 + }, + { + "epoch": 0.35244, + "grad_norm": 1.0726209296899567, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 35244 + }, + { + "epoch": 0.35245, + "grad_norm": 1.005972190531522, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 35245 + }, + { + "epoch": 0.35246, + "grad_norm": 1.1960767272489568, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 35246 + }, + { + "epoch": 0.35247, + "grad_norm": 1.0919258143900348, + "learning_rate": 0.003, + "loss": 4.044, + "step": 35247 + }, + { + "epoch": 0.35248, + "grad_norm": 1.142091409617738, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 35248 + }, + { + "epoch": 0.35249, + "grad_norm": 0.9940480904692286, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 35249 + }, + { + "epoch": 0.3525, + "grad_norm": 1.2876924192446928, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 35250 + }, + { + "epoch": 0.35251, + "grad_norm": 1.10379296181195, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 35251 + }, + { + "epoch": 0.35252, + "grad_norm": 1.195667354785666, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 35252 + }, + { + "epoch": 0.35253, + "grad_norm": 0.9289135695387426, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 35253 + }, + { + "epoch": 0.35254, + "grad_norm": 0.9645918211072203, + "learning_rate": 0.003, + "loss": 4.052, + "step": 35254 + }, + { + "epoch": 0.35255, + "grad_norm": 1.3272180125159412, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 35255 + }, + { + "epoch": 0.35256, + "grad_norm": 0.926594112199224, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 35256 + }, + { + "epoch": 0.35257, + "grad_norm": 1.0965787174577744, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 35257 + }, + { + "epoch": 0.35258, + "grad_norm": 1.1789629708706528, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 35258 + }, + { + "epoch": 0.35259, + "grad_norm": 0.8389963908081807, + "learning_rate": 0.003, + "loss": 4.045, + "step": 35259 + }, + { + "epoch": 0.3526, + "grad_norm": 0.9552962273505238, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 35260 + }, + { + "epoch": 0.35261, + "grad_norm": 1.1232938353599347, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 35261 + }, + { + "epoch": 0.35262, + "grad_norm": 1.056803385564251, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 35262 + }, + { + "epoch": 0.35263, + "grad_norm": 1.0260852092418427, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 35263 + }, + { + "epoch": 0.35264, + "grad_norm": 1.382141025170784, + "learning_rate": 0.003, + "loss": 4.032, + "step": 35264 + }, + { + "epoch": 0.35265, + "grad_norm": 0.9421773710989504, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 35265 + }, + { + "epoch": 0.35266, + "grad_norm": 0.9153767229348604, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 35266 + }, + { + "epoch": 0.35267, + "grad_norm": 1.0708265314319247, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 35267 + }, + { + "epoch": 0.35268, + "grad_norm": 1.1258051511100784, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 35268 + }, + { + "epoch": 0.35269, + "grad_norm": 1.099184238811196, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 35269 + }, + { + "epoch": 0.3527, + "grad_norm": 1.0638504468905337, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 35270 + }, + { + "epoch": 0.35271, + "grad_norm": 1.1278865544328964, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 35271 + }, + { + "epoch": 0.35272, + "grad_norm": 1.006698350810172, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 35272 + }, + { + "epoch": 0.35273, + "grad_norm": 1.2219167167071028, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 35273 + }, + { + "epoch": 0.35274, + "grad_norm": 1.0205457337680368, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 35274 + }, + { + "epoch": 0.35275, + "grad_norm": 1.163147168609544, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 35275 + }, + { + "epoch": 0.35276, + "grad_norm": 1.0554282253502485, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 35276 + }, + { + "epoch": 0.35277, + "grad_norm": 0.9065243441710843, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 35277 + }, + { + "epoch": 0.35278, + "grad_norm": 1.0521274432767065, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 35278 + }, + { + "epoch": 0.35279, + "grad_norm": 1.3256933466983354, + "learning_rate": 0.003, + "loss": 4.067, + "step": 35279 + }, + { + "epoch": 0.3528, + "grad_norm": 1.0557167811280406, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 35280 + }, + { + "epoch": 0.35281, + "grad_norm": 1.159284977604106, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 35281 + }, + { + "epoch": 0.35282, + "grad_norm": 1.1992046527115747, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 35282 + }, + { + "epoch": 0.35283, + "grad_norm": 0.9666734755485986, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 35283 + }, + { + "epoch": 0.35284, + "grad_norm": 1.0747811057323222, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 35284 + }, + { + "epoch": 0.35285, + "grad_norm": 1.10457994415811, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 35285 + }, + { + "epoch": 0.35286, + "grad_norm": 1.2436485104217443, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 35286 + }, + { + "epoch": 0.35287, + "grad_norm": 0.9556262530460833, + "learning_rate": 0.003, + "loss": 4.018, + "step": 35287 + }, + { + "epoch": 0.35288, + "grad_norm": 1.1156286271712068, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 35288 + }, + { + "epoch": 0.35289, + "grad_norm": 1.1260883222671354, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 35289 + }, + { + "epoch": 0.3529, + "grad_norm": 1.0937781613469362, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 35290 + }, + { + "epoch": 0.35291, + "grad_norm": 0.951092241166077, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 35291 + }, + { + "epoch": 0.35292, + "grad_norm": 1.1444513337194817, + "learning_rate": 0.003, + "loss": 4.054, + "step": 35292 + }, + { + "epoch": 0.35293, + "grad_norm": 0.8538087938871355, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 35293 + }, + { + "epoch": 0.35294, + "grad_norm": 0.9616771117921259, + "learning_rate": 0.003, + "loss": 4.044, + "step": 35294 + }, + { + "epoch": 0.35295, + "grad_norm": 1.107475274393163, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 35295 + }, + { + "epoch": 0.35296, + "grad_norm": 1.0601443155943484, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 35296 + }, + { + "epoch": 0.35297, + "grad_norm": 1.0716647393972993, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 35297 + }, + { + "epoch": 0.35298, + "grad_norm": 1.155661660185143, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 35298 + }, + { + "epoch": 0.35299, + "grad_norm": 1.2480127592371395, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 35299 + }, + { + "epoch": 0.353, + "grad_norm": 0.9822715383351442, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 35300 + }, + { + "epoch": 0.35301, + "grad_norm": 1.1283624888984902, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 35301 + }, + { + "epoch": 0.35302, + "grad_norm": 1.1805196673784144, + "learning_rate": 0.003, + "loss": 4.033, + "step": 35302 + }, + { + "epoch": 0.35303, + "grad_norm": 0.958506924967713, + "learning_rate": 0.003, + "loss": 4.056, + "step": 35303 + }, + { + "epoch": 0.35304, + "grad_norm": 1.1774074762935967, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 35304 + }, + { + "epoch": 0.35305, + "grad_norm": 1.301371340330293, + "learning_rate": 0.003, + "loss": 4.027, + "step": 35305 + }, + { + "epoch": 0.35306, + "grad_norm": 0.9076907547489832, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 35306 + }, + { + "epoch": 0.35307, + "grad_norm": 0.9899550950458699, + "learning_rate": 0.003, + "loss": 4.062, + "step": 35307 + }, + { + "epoch": 0.35308, + "grad_norm": 1.1736647347368012, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 35308 + }, + { + "epoch": 0.35309, + "grad_norm": 0.9996467409813728, + "learning_rate": 0.003, + "loss": 4.011, + "step": 35309 + }, + { + "epoch": 0.3531, + "grad_norm": 1.214256352536846, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 35310 + }, + { + "epoch": 0.35311, + "grad_norm": 1.016955179917925, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 35311 + }, + { + "epoch": 0.35312, + "grad_norm": 1.3435250359124917, + "learning_rate": 0.003, + "loss": 4.042, + "step": 35312 + }, + { + "epoch": 0.35313, + "grad_norm": 1.065047844300601, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 35313 + }, + { + "epoch": 0.35314, + "grad_norm": 1.0231874180408642, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 35314 + }, + { + "epoch": 0.35315, + "grad_norm": 0.9890648016251014, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 35315 + }, + { + "epoch": 0.35316, + "grad_norm": 1.0799763643420515, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 35316 + }, + { + "epoch": 0.35317, + "grad_norm": 0.864806581476304, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 35317 + }, + { + "epoch": 0.35318, + "grad_norm": 0.8855086547512037, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 35318 + }, + { + "epoch": 0.35319, + "grad_norm": 1.0283916426418367, + "learning_rate": 0.003, + "loss": 4.035, + "step": 35319 + }, + { + "epoch": 0.3532, + "grad_norm": 1.2793213175416869, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 35320 + }, + { + "epoch": 0.35321, + "grad_norm": 1.2883528898265129, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 35321 + }, + { + "epoch": 0.35322, + "grad_norm": 1.0013863947562838, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 35322 + }, + { + "epoch": 0.35323, + "grad_norm": 1.2295366316915022, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 35323 + }, + { + "epoch": 0.35324, + "grad_norm": 0.9426318507462338, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 35324 + }, + { + "epoch": 0.35325, + "grad_norm": 1.0847068973165692, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 35325 + }, + { + "epoch": 0.35326, + "grad_norm": 1.158775555717033, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 35326 + }, + { + "epoch": 0.35327, + "grad_norm": 1.009136103224398, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 35327 + }, + { + "epoch": 0.35328, + "grad_norm": 1.0899769101058032, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 35328 + }, + { + "epoch": 0.35329, + "grad_norm": 1.0641610760840472, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 35329 + }, + { + "epoch": 0.3533, + "grad_norm": 1.0651067046739202, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 35330 + }, + { + "epoch": 0.35331, + "grad_norm": 1.2004977029703423, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 35331 + }, + { + "epoch": 0.35332, + "grad_norm": 1.0334471373424745, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 35332 + }, + { + "epoch": 0.35333, + "grad_norm": 1.1379764226346742, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 35333 + }, + { + "epoch": 0.35334, + "grad_norm": 1.1874163312090167, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 35334 + }, + { + "epoch": 0.35335, + "grad_norm": 0.9673770424881707, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 35335 + }, + { + "epoch": 0.35336, + "grad_norm": 1.0293593854073018, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 35336 + }, + { + "epoch": 0.35337, + "grad_norm": 1.0634743291782043, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 35337 + }, + { + "epoch": 0.35338, + "grad_norm": 0.9305607421730626, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 35338 + }, + { + "epoch": 0.35339, + "grad_norm": 0.9375734939553004, + "learning_rate": 0.003, + "loss": 4.061, + "step": 35339 + }, + { + "epoch": 0.3534, + "grad_norm": 1.251496530842142, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 35340 + }, + { + "epoch": 0.35341, + "grad_norm": 1.076426693277424, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 35341 + }, + { + "epoch": 0.35342, + "grad_norm": 1.253030038786931, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 35342 + }, + { + "epoch": 0.35343, + "grad_norm": 1.103580369881204, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 35343 + }, + { + "epoch": 0.35344, + "grad_norm": 1.103656210587188, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 35344 + }, + { + "epoch": 0.35345, + "grad_norm": 1.197102889031292, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 35345 + }, + { + "epoch": 0.35346, + "grad_norm": 0.9282260342635278, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 35346 + }, + { + "epoch": 0.35347, + "grad_norm": 1.023534396551096, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 35347 + }, + { + "epoch": 0.35348, + "grad_norm": 1.0562577406990759, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 35348 + }, + { + "epoch": 0.35349, + "grad_norm": 1.0794927293144505, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 35349 + }, + { + "epoch": 0.3535, + "grad_norm": 1.0584206637368954, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 35350 + }, + { + "epoch": 0.35351, + "grad_norm": 1.0541119749622916, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 35351 + }, + { + "epoch": 0.35352, + "grad_norm": 1.4599093219855057, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 35352 + }, + { + "epoch": 0.35353, + "grad_norm": 0.8333825351170584, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 35353 + }, + { + "epoch": 0.35354, + "grad_norm": 1.0655570215051917, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 35354 + }, + { + "epoch": 0.35355, + "grad_norm": 1.0776531870393984, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 35355 + }, + { + "epoch": 0.35356, + "grad_norm": 1.3163880356811244, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 35356 + }, + { + "epoch": 0.35357, + "grad_norm": 1.157291885335012, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 35357 + }, + { + "epoch": 0.35358, + "grad_norm": 0.8980220177992865, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 35358 + }, + { + "epoch": 0.35359, + "grad_norm": 0.8590417289132891, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 35359 + }, + { + "epoch": 0.3536, + "grad_norm": 0.9645628308923946, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 35360 + }, + { + "epoch": 0.35361, + "grad_norm": 1.2804853461433288, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 35361 + }, + { + "epoch": 0.35362, + "grad_norm": 0.9411056395658333, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 35362 + }, + { + "epoch": 0.35363, + "grad_norm": 0.9663024314017004, + "learning_rate": 0.003, + "loss": 4.026, + "step": 35363 + }, + { + "epoch": 0.35364, + "grad_norm": 1.0745028981445741, + "learning_rate": 0.003, + "loss": 4.036, + "step": 35364 + }, + { + "epoch": 0.35365, + "grad_norm": 0.9762560778741938, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 35365 + }, + { + "epoch": 0.35366, + "grad_norm": 0.9683397319650876, + "learning_rate": 0.003, + "loss": 4.043, + "step": 35366 + }, + { + "epoch": 0.35367, + "grad_norm": 1.2561270676981289, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 35367 + }, + { + "epoch": 0.35368, + "grad_norm": 0.9681935071287254, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 35368 + }, + { + "epoch": 0.35369, + "grad_norm": 1.1924774697172227, + "learning_rate": 0.003, + "loss": 4.008, + "step": 35369 + }, + { + "epoch": 0.3537, + "grad_norm": 1.196995029503743, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 35370 + }, + { + "epoch": 0.35371, + "grad_norm": 1.5764464765314892, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 35371 + }, + { + "epoch": 0.35372, + "grad_norm": 0.820037819475443, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 35372 + }, + { + "epoch": 0.35373, + "grad_norm": 0.9087546743612851, + "learning_rate": 0.003, + "loss": 4.035, + "step": 35373 + }, + { + "epoch": 0.35374, + "grad_norm": 0.9925379756568413, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 35374 + }, + { + "epoch": 0.35375, + "grad_norm": 1.150955647482644, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 35375 + }, + { + "epoch": 0.35376, + "grad_norm": 1.4557153438551003, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 35376 + }, + { + "epoch": 0.35377, + "grad_norm": 0.981990574252401, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 35377 + }, + { + "epoch": 0.35378, + "grad_norm": 1.0592470945177204, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 35378 + }, + { + "epoch": 0.35379, + "grad_norm": 1.2419595184279002, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 35379 + }, + { + "epoch": 0.3538, + "grad_norm": 1.096184619881051, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 35380 + }, + { + "epoch": 0.35381, + "grad_norm": 0.8951863400439698, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 35381 + }, + { + "epoch": 0.35382, + "grad_norm": 0.874052107880109, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 35382 + }, + { + "epoch": 0.35383, + "grad_norm": 1.2483479351560232, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 35383 + }, + { + "epoch": 0.35384, + "grad_norm": 1.2591596582661575, + "learning_rate": 0.003, + "loss": 4.045, + "step": 35384 + }, + { + "epoch": 0.35385, + "grad_norm": 0.7559738697007036, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 35385 + }, + { + "epoch": 0.35386, + "grad_norm": 0.8417486430792628, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 35386 + }, + { + "epoch": 0.35387, + "grad_norm": 0.9008130729340037, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 35387 + }, + { + "epoch": 0.35388, + "grad_norm": 1.2011826518393656, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 35388 + }, + { + "epoch": 0.35389, + "grad_norm": 1.2207188528886315, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 35389 + }, + { + "epoch": 0.3539, + "grad_norm": 1.2383776217057678, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 35390 + }, + { + "epoch": 0.35391, + "grad_norm": 0.9074520191585378, + "learning_rate": 0.003, + "loss": 4.024, + "step": 35391 + }, + { + "epoch": 0.35392, + "grad_norm": 0.9715398198967014, + "learning_rate": 0.003, + "loss": 4.038, + "step": 35392 + }, + { + "epoch": 0.35393, + "grad_norm": 1.0638775438610082, + "learning_rate": 0.003, + "loss": 4.054, + "step": 35393 + }, + { + "epoch": 0.35394, + "grad_norm": 1.1080483075047836, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 35394 + }, + { + "epoch": 0.35395, + "grad_norm": 1.031600515527374, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 35395 + }, + { + "epoch": 0.35396, + "grad_norm": 1.1751999624928928, + "learning_rate": 0.003, + "loss": 4.048, + "step": 35396 + }, + { + "epoch": 0.35397, + "grad_norm": 1.2209622659733415, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 35397 + }, + { + "epoch": 0.35398, + "grad_norm": 1.092512025171629, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 35398 + }, + { + "epoch": 0.35399, + "grad_norm": 1.0206623246795608, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 35399 + }, + { + "epoch": 0.354, + "grad_norm": 1.4276088043961122, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 35400 + }, + { + "epoch": 0.35401, + "grad_norm": 1.018116892578526, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 35401 + }, + { + "epoch": 0.35402, + "grad_norm": 1.2220660538638197, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 35402 + }, + { + "epoch": 0.35403, + "grad_norm": 1.0330951464993041, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 35403 + }, + { + "epoch": 0.35404, + "grad_norm": 1.0694537947529226, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 35404 + }, + { + "epoch": 0.35405, + "grad_norm": 1.0420004858384586, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 35405 + }, + { + "epoch": 0.35406, + "grad_norm": 1.086104438283513, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 35406 + }, + { + "epoch": 0.35407, + "grad_norm": 1.048570410687985, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 35407 + }, + { + "epoch": 0.35408, + "grad_norm": 1.121616538817845, + "learning_rate": 0.003, + "loss": 4.049, + "step": 35408 + }, + { + "epoch": 0.35409, + "grad_norm": 1.021917225700126, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 35409 + }, + { + "epoch": 0.3541, + "grad_norm": 1.1056071602824393, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 35410 + }, + { + "epoch": 0.35411, + "grad_norm": 0.9756716611413416, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 35411 + }, + { + "epoch": 0.35412, + "grad_norm": 1.0223602633793387, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 35412 + }, + { + "epoch": 0.35413, + "grad_norm": 0.9818651543353925, + "learning_rate": 0.003, + "loss": 4.042, + "step": 35413 + }, + { + "epoch": 0.35414, + "grad_norm": 1.0077492942861168, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 35414 + }, + { + "epoch": 0.35415, + "grad_norm": 1.043672756487177, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 35415 + }, + { + "epoch": 0.35416, + "grad_norm": 0.917758218377558, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 35416 + }, + { + "epoch": 0.35417, + "grad_norm": 0.9944533659044833, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 35417 + }, + { + "epoch": 0.35418, + "grad_norm": 1.3413758918713656, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 35418 + }, + { + "epoch": 0.35419, + "grad_norm": 1.3151255864284936, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 35419 + }, + { + "epoch": 0.3542, + "grad_norm": 0.9184042063806114, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 35420 + }, + { + "epoch": 0.35421, + "grad_norm": 0.8722969228681837, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 35421 + }, + { + "epoch": 0.35422, + "grad_norm": 1.1148724660728424, + "learning_rate": 0.003, + "loss": 4.022, + "step": 35422 + }, + { + "epoch": 0.35423, + "grad_norm": 1.2086626273675987, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 35423 + }, + { + "epoch": 0.35424, + "grad_norm": 1.0045974187797384, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 35424 + }, + { + "epoch": 0.35425, + "grad_norm": 1.3971089541631159, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 35425 + }, + { + "epoch": 0.35426, + "grad_norm": 1.010950842167999, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 35426 + }, + { + "epoch": 0.35427, + "grad_norm": 0.9682365297863983, + "learning_rate": 0.003, + "loss": 4.04, + "step": 35427 + }, + { + "epoch": 0.35428, + "grad_norm": 1.1696140146609082, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 35428 + }, + { + "epoch": 0.35429, + "grad_norm": 1.0941327781652292, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 35429 + }, + { + "epoch": 0.3543, + "grad_norm": 1.3950930892918454, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 35430 + }, + { + "epoch": 0.35431, + "grad_norm": 0.9768059287848244, + "learning_rate": 0.003, + "loss": 4.057, + "step": 35431 + }, + { + "epoch": 0.35432, + "grad_norm": 1.0354614334380317, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 35432 + }, + { + "epoch": 0.35433, + "grad_norm": 1.0777978981155079, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 35433 + }, + { + "epoch": 0.35434, + "grad_norm": 1.0335445693833067, + "learning_rate": 0.003, + "loss": 4.023, + "step": 35434 + }, + { + "epoch": 0.35435, + "grad_norm": 1.1072312044052794, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 35435 + }, + { + "epoch": 0.35436, + "grad_norm": 1.1496168813270364, + "learning_rate": 0.003, + "loss": 4.053, + "step": 35436 + }, + { + "epoch": 0.35437, + "grad_norm": 0.9016490770867897, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 35437 + }, + { + "epoch": 0.35438, + "grad_norm": 1.0102628290090563, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 35438 + }, + { + "epoch": 0.35439, + "grad_norm": 1.1505229580660017, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 35439 + }, + { + "epoch": 0.3544, + "grad_norm": 1.247076418777222, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 35440 + }, + { + "epoch": 0.35441, + "grad_norm": 1.1869478116220513, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 35441 + }, + { + "epoch": 0.35442, + "grad_norm": 0.9482758325729708, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 35442 + }, + { + "epoch": 0.35443, + "grad_norm": 0.9732144686687805, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 35443 + }, + { + "epoch": 0.35444, + "grad_norm": 1.2250656944433578, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 35444 + }, + { + "epoch": 0.35445, + "grad_norm": 1.1015258806767705, + "learning_rate": 0.003, + "loss": 4.079, + "step": 35445 + }, + { + "epoch": 0.35446, + "grad_norm": 1.017739529240618, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 35446 + }, + { + "epoch": 0.35447, + "grad_norm": 1.1763051868327574, + "learning_rate": 0.003, + "loss": 4.038, + "step": 35447 + }, + { + "epoch": 0.35448, + "grad_norm": 1.1136757694235664, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 35448 + }, + { + "epoch": 0.35449, + "grad_norm": 1.0671644998451262, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 35449 + }, + { + "epoch": 0.3545, + "grad_norm": 0.9259742910309325, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 35450 + }, + { + "epoch": 0.35451, + "grad_norm": 0.949629371183701, + "learning_rate": 0.003, + "loss": 4.016, + "step": 35451 + }, + { + "epoch": 0.35452, + "grad_norm": 0.899348395761475, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 35452 + }, + { + "epoch": 0.35453, + "grad_norm": 0.9317125662485806, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 35453 + }, + { + "epoch": 0.35454, + "grad_norm": 1.1843287849376607, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 35454 + }, + { + "epoch": 0.35455, + "grad_norm": 1.135349772645484, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 35455 + }, + { + "epoch": 0.35456, + "grad_norm": 1.2054595240864145, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 35456 + }, + { + "epoch": 0.35457, + "grad_norm": 1.2770768703218176, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 35457 + }, + { + "epoch": 0.35458, + "grad_norm": 1.0026761802882234, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 35458 + }, + { + "epoch": 0.35459, + "grad_norm": 1.1631662938149292, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 35459 + }, + { + "epoch": 0.3546, + "grad_norm": 0.9683007587952852, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 35460 + }, + { + "epoch": 0.35461, + "grad_norm": 1.2533444154762754, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 35461 + }, + { + "epoch": 0.35462, + "grad_norm": 1.0083790351985482, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 35462 + }, + { + "epoch": 0.35463, + "grad_norm": 0.9611185255807982, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 35463 + }, + { + "epoch": 0.35464, + "grad_norm": 1.041278632076461, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 35464 + }, + { + "epoch": 0.35465, + "grad_norm": 1.2703928802069646, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 35465 + }, + { + "epoch": 0.35466, + "grad_norm": 0.846030196730307, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 35466 + }, + { + "epoch": 0.35467, + "grad_norm": 0.7243024486116023, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 35467 + }, + { + "epoch": 0.35468, + "grad_norm": 0.9662132580411676, + "learning_rate": 0.003, + "loss": 4.053, + "step": 35468 + }, + { + "epoch": 0.35469, + "grad_norm": 1.0215913426108978, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 35469 + }, + { + "epoch": 0.3547, + "grad_norm": 1.040188024637803, + "learning_rate": 0.003, + "loss": 4.071, + "step": 35470 + }, + { + "epoch": 0.35471, + "grad_norm": 1.1976480440661137, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 35471 + }, + { + "epoch": 0.35472, + "grad_norm": 1.0721751750149096, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 35472 + }, + { + "epoch": 0.35473, + "grad_norm": 1.0720482519221115, + "learning_rate": 0.003, + "loss": 4.028, + "step": 35473 + }, + { + "epoch": 0.35474, + "grad_norm": 1.0114367680331613, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 35474 + }, + { + "epoch": 0.35475, + "grad_norm": 0.9498823261221037, + "learning_rate": 0.003, + "loss": 4.05, + "step": 35475 + }, + { + "epoch": 0.35476, + "grad_norm": 0.9466495856763693, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 35476 + }, + { + "epoch": 0.35477, + "grad_norm": 1.0682172455609322, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 35477 + }, + { + "epoch": 0.35478, + "grad_norm": 1.1431949968097739, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 35478 + }, + { + "epoch": 0.35479, + "grad_norm": 1.0387926819534556, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 35479 + }, + { + "epoch": 0.3548, + "grad_norm": 1.072574280247357, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 35480 + }, + { + "epoch": 0.35481, + "grad_norm": 1.1752543031141822, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 35481 + }, + { + "epoch": 0.35482, + "grad_norm": 1.7547719163189284, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 35482 + }, + { + "epoch": 0.35483, + "grad_norm": 0.7040767001194593, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 35483 + }, + { + "epoch": 0.35484, + "grad_norm": 1.0247511747242917, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 35484 + }, + { + "epoch": 0.35485, + "grad_norm": 1.4520429094176173, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 35485 + }, + { + "epoch": 0.35486, + "grad_norm": 0.803057544818838, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 35486 + }, + { + "epoch": 0.35487, + "grad_norm": 1.050329723240252, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 35487 + }, + { + "epoch": 0.35488, + "grad_norm": 0.9109835458699538, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 35488 + }, + { + "epoch": 0.35489, + "grad_norm": 0.9711043315397743, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 35489 + }, + { + "epoch": 0.3549, + "grad_norm": 1.1978660432511392, + "learning_rate": 0.003, + "loss": 4.045, + "step": 35490 + }, + { + "epoch": 0.35491, + "grad_norm": 1.1690402184532522, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 35491 + }, + { + "epoch": 0.35492, + "grad_norm": 1.2300706136684034, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 35492 + }, + { + "epoch": 0.35493, + "grad_norm": 1.08959749699816, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 35493 + }, + { + "epoch": 0.35494, + "grad_norm": 0.9622005433585042, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 35494 + }, + { + "epoch": 0.35495, + "grad_norm": 1.121280555190489, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 35495 + }, + { + "epoch": 0.35496, + "grad_norm": 1.209595759297322, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 35496 + }, + { + "epoch": 0.35497, + "grad_norm": 0.9222064122436742, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 35497 + }, + { + "epoch": 0.35498, + "grad_norm": 1.2619790431707065, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 35498 + }, + { + "epoch": 0.35499, + "grad_norm": 0.9900298355950882, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 35499 + }, + { + "epoch": 0.355, + "grad_norm": 0.9692648342724822, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 35500 + }, + { + "epoch": 0.35501, + "grad_norm": 1.1307752518903318, + "learning_rate": 0.003, + "loss": 4.049, + "step": 35501 + }, + { + "epoch": 0.35502, + "grad_norm": 1.0714643229376195, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 35502 + }, + { + "epoch": 0.35503, + "grad_norm": 1.0228523565737473, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 35503 + }, + { + "epoch": 0.35504, + "grad_norm": 1.171318181652272, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 35504 + }, + { + "epoch": 0.35505, + "grad_norm": 1.213186903263662, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 35505 + }, + { + "epoch": 0.35506, + "grad_norm": 0.8843383715462079, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 35506 + }, + { + "epoch": 0.35507, + "grad_norm": 1.103762036649215, + "learning_rate": 0.003, + "loss": 4.029, + "step": 35507 + }, + { + "epoch": 0.35508, + "grad_norm": 1.379359398122215, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 35508 + }, + { + "epoch": 0.35509, + "grad_norm": 0.8005583894485412, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 35509 + }, + { + "epoch": 0.3551, + "grad_norm": 0.7627372882757958, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 35510 + }, + { + "epoch": 0.35511, + "grad_norm": 0.9660173443836291, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 35511 + }, + { + "epoch": 0.35512, + "grad_norm": 1.0394383678391905, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 35512 + }, + { + "epoch": 0.35513, + "grad_norm": 1.0356265239524383, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 35513 + }, + { + "epoch": 0.35514, + "grad_norm": 1.1619126616807423, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 35514 + }, + { + "epoch": 0.35515, + "grad_norm": 1.0993724524229471, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 35515 + }, + { + "epoch": 0.35516, + "grad_norm": 1.1022722268414322, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 35516 + }, + { + "epoch": 0.35517, + "grad_norm": 1.3327075134365989, + "learning_rate": 0.003, + "loss": 4.064, + "step": 35517 + }, + { + "epoch": 0.35518, + "grad_norm": 1.0317206531111245, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 35518 + }, + { + "epoch": 0.35519, + "grad_norm": 1.099481908284126, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 35519 + }, + { + "epoch": 0.3552, + "grad_norm": 1.023302006284792, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 35520 + }, + { + "epoch": 0.35521, + "grad_norm": 1.120636638963493, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 35521 + }, + { + "epoch": 0.35522, + "grad_norm": 1.0289998749564324, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 35522 + }, + { + "epoch": 0.35523, + "grad_norm": 1.2982009346429393, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 35523 + }, + { + "epoch": 0.35524, + "grad_norm": 0.9530303422252163, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 35524 + }, + { + "epoch": 0.35525, + "grad_norm": 1.2443185254839273, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 35525 + }, + { + "epoch": 0.35526, + "grad_norm": 1.1703824612712175, + "learning_rate": 0.003, + "loss": 4.07, + "step": 35526 + }, + { + "epoch": 0.35527, + "grad_norm": 1.281147721091665, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 35527 + }, + { + "epoch": 0.35528, + "grad_norm": 0.9187056394429827, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 35528 + }, + { + "epoch": 0.35529, + "grad_norm": 0.8684298077482888, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 35529 + }, + { + "epoch": 0.3553, + "grad_norm": 0.9499170527898176, + "learning_rate": 0.003, + "loss": 4.042, + "step": 35530 + }, + { + "epoch": 0.35531, + "grad_norm": 1.2772233945386782, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 35531 + }, + { + "epoch": 0.35532, + "grad_norm": 0.9497466093711919, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 35532 + }, + { + "epoch": 0.35533, + "grad_norm": 1.2576285268432368, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 35533 + }, + { + "epoch": 0.35534, + "grad_norm": 1.0338539445169013, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 35534 + }, + { + "epoch": 0.35535, + "grad_norm": 1.1068798429724107, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 35535 + }, + { + "epoch": 0.35536, + "grad_norm": 1.02765905333589, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 35536 + }, + { + "epoch": 0.35537, + "grad_norm": 1.1379066735153482, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 35537 + }, + { + "epoch": 0.35538, + "grad_norm": 1.1503617696727768, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 35538 + }, + { + "epoch": 0.35539, + "grad_norm": 1.1852294045106126, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 35539 + }, + { + "epoch": 0.3554, + "grad_norm": 1.0899097896396004, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 35540 + }, + { + "epoch": 0.35541, + "grad_norm": 1.157064561388665, + "learning_rate": 0.003, + "loss": 4.015, + "step": 35541 + }, + { + "epoch": 0.35542, + "grad_norm": 1.072034891474068, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 35542 + }, + { + "epoch": 0.35543, + "grad_norm": 1.1600076821003877, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 35543 + }, + { + "epoch": 0.35544, + "grad_norm": 1.1074986105190368, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 35544 + }, + { + "epoch": 0.35545, + "grad_norm": 1.061291895166256, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 35545 + }, + { + "epoch": 0.35546, + "grad_norm": 1.0808941383925763, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 35546 + }, + { + "epoch": 0.35547, + "grad_norm": 1.217273877464344, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 35547 + }, + { + "epoch": 0.35548, + "grad_norm": 1.0974141307387417, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 35548 + }, + { + "epoch": 0.35549, + "grad_norm": 1.0861309982954594, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 35549 + }, + { + "epoch": 0.3555, + "grad_norm": 1.0807477631937181, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 35550 + }, + { + "epoch": 0.35551, + "grad_norm": 0.9451022680774412, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 35551 + }, + { + "epoch": 0.35552, + "grad_norm": 1.1291641518453004, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 35552 + }, + { + "epoch": 0.35553, + "grad_norm": 1.0416127898066998, + "learning_rate": 0.003, + "loss": 4.038, + "step": 35553 + }, + { + "epoch": 0.35554, + "grad_norm": 1.2094879749558256, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 35554 + }, + { + "epoch": 0.35555, + "grad_norm": 1.048515702752777, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 35555 + }, + { + "epoch": 0.35556, + "grad_norm": 1.1306495932670504, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 35556 + }, + { + "epoch": 0.35557, + "grad_norm": 1.0846203527956944, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 35557 + }, + { + "epoch": 0.35558, + "grad_norm": 1.144112560442146, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 35558 + }, + { + "epoch": 0.35559, + "grad_norm": 1.2002342868542422, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 35559 + }, + { + "epoch": 0.3556, + "grad_norm": 0.9170331895871701, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 35560 + }, + { + "epoch": 0.35561, + "grad_norm": 0.8788359496861506, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 35561 + }, + { + "epoch": 0.35562, + "grad_norm": 0.9612976737907654, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 35562 + }, + { + "epoch": 0.35563, + "grad_norm": 1.0153572837765965, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 35563 + }, + { + "epoch": 0.35564, + "grad_norm": 1.2056614599637507, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 35564 + }, + { + "epoch": 0.35565, + "grad_norm": 1.1285978330400175, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 35565 + }, + { + "epoch": 0.35566, + "grad_norm": 1.192078429430611, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 35566 + }, + { + "epoch": 0.35567, + "grad_norm": 1.0057896680174405, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 35567 + }, + { + "epoch": 0.35568, + "grad_norm": 1.1528190943102277, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 35568 + }, + { + "epoch": 0.35569, + "grad_norm": 0.9499065887064395, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 35569 + }, + { + "epoch": 0.3557, + "grad_norm": 1.0404607458992228, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 35570 + }, + { + "epoch": 0.35571, + "grad_norm": 1.248166759463059, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 35571 + }, + { + "epoch": 0.35572, + "grad_norm": 0.9871744747566311, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 35572 + }, + { + "epoch": 0.35573, + "grad_norm": 1.1162878852223108, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 35573 + }, + { + "epoch": 0.35574, + "grad_norm": 1.1247638809037337, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 35574 + }, + { + "epoch": 0.35575, + "grad_norm": 1.2342170434598763, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 35575 + }, + { + "epoch": 0.35576, + "grad_norm": 1.1038108778282092, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 35576 + }, + { + "epoch": 0.35577, + "grad_norm": 0.9621952532767276, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 35577 + }, + { + "epoch": 0.35578, + "grad_norm": 0.9515007774253934, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 35578 + }, + { + "epoch": 0.35579, + "grad_norm": 1.2073325007778504, + "learning_rate": 0.003, + "loss": 4.056, + "step": 35579 + }, + { + "epoch": 0.3558, + "grad_norm": 0.9390075495334784, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 35580 + }, + { + "epoch": 0.35581, + "grad_norm": 1.1290033622733957, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 35581 + }, + { + "epoch": 0.35582, + "grad_norm": 1.1659643923191612, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 35582 + }, + { + "epoch": 0.35583, + "grad_norm": 1.0752717531836935, + "learning_rate": 0.003, + "loss": 4.038, + "step": 35583 + }, + { + "epoch": 0.35584, + "grad_norm": 0.9520510639835474, + "learning_rate": 0.003, + "loss": 4.034, + "step": 35584 + }, + { + "epoch": 0.35585, + "grad_norm": 0.8795528086439379, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 35585 + }, + { + "epoch": 0.35586, + "grad_norm": 0.9857920164132304, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 35586 + }, + { + "epoch": 0.35587, + "grad_norm": 1.554135607270914, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 35587 + }, + { + "epoch": 0.35588, + "grad_norm": 0.8218665269277823, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 35588 + }, + { + "epoch": 0.35589, + "grad_norm": 0.9560869138495331, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 35589 + }, + { + "epoch": 0.3559, + "grad_norm": 1.06381553617479, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 35590 + }, + { + "epoch": 0.35591, + "grad_norm": 1.1182446248132454, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 35591 + }, + { + "epoch": 0.35592, + "grad_norm": 0.9936814045115612, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 35592 + }, + { + "epoch": 0.35593, + "grad_norm": 1.223270840692896, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 35593 + }, + { + "epoch": 0.35594, + "grad_norm": 0.8843808929809455, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 35594 + }, + { + "epoch": 0.35595, + "grad_norm": 0.9457179014977011, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 35595 + }, + { + "epoch": 0.35596, + "grad_norm": 1.2950861760237842, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 35596 + }, + { + "epoch": 0.35597, + "grad_norm": 0.8695089236429102, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 35597 + }, + { + "epoch": 0.35598, + "grad_norm": 1.18957459844958, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 35598 + }, + { + "epoch": 0.35599, + "grad_norm": 1.2376018906772217, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 35599 + }, + { + "epoch": 0.356, + "grad_norm": 1.0345739485904186, + "learning_rate": 0.003, + "loss": 4.024, + "step": 35600 + }, + { + "epoch": 0.35601, + "grad_norm": 1.1291401280465876, + "learning_rate": 0.003, + "loss": 4.057, + "step": 35601 + }, + { + "epoch": 0.35602, + "grad_norm": 1.071531933161162, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 35602 + }, + { + "epoch": 0.35603, + "grad_norm": 1.2532265552332174, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 35603 + }, + { + "epoch": 0.35604, + "grad_norm": 1.1137862732022037, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 35604 + }, + { + "epoch": 0.35605, + "grad_norm": 1.0338071656189556, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 35605 + }, + { + "epoch": 0.35606, + "grad_norm": 1.1544370585804822, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 35606 + }, + { + "epoch": 0.35607, + "grad_norm": 1.0529615481071284, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 35607 + }, + { + "epoch": 0.35608, + "grad_norm": 1.094221967480766, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 35608 + }, + { + "epoch": 0.35609, + "grad_norm": 1.0277142428536508, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 35609 + }, + { + "epoch": 0.3561, + "grad_norm": 1.153013020597003, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 35610 + }, + { + "epoch": 0.35611, + "grad_norm": 0.8326564644317113, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 35611 + }, + { + "epoch": 0.35612, + "grad_norm": 0.9616953029561579, + "learning_rate": 0.003, + "loss": 4.041, + "step": 35612 + }, + { + "epoch": 0.35613, + "grad_norm": 1.212215254503468, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 35613 + }, + { + "epoch": 0.35614, + "grad_norm": 0.9951646610851534, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 35614 + }, + { + "epoch": 0.35615, + "grad_norm": 1.0295344663769714, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 35615 + }, + { + "epoch": 0.35616, + "grad_norm": 1.2823577374834008, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 35616 + }, + { + "epoch": 0.35617, + "grad_norm": 1.0795118590623776, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 35617 + }, + { + "epoch": 0.35618, + "grad_norm": 0.946835720617722, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 35618 + }, + { + "epoch": 0.35619, + "grad_norm": 0.937414585715965, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 35619 + }, + { + "epoch": 0.3562, + "grad_norm": 1.1194985446223085, + "learning_rate": 0.003, + "loss": 4.048, + "step": 35620 + }, + { + "epoch": 0.35621, + "grad_norm": 1.4081971405437532, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 35621 + }, + { + "epoch": 0.35622, + "grad_norm": 0.8424140160728647, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 35622 + }, + { + "epoch": 0.35623, + "grad_norm": 0.8452117877480922, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 35623 + }, + { + "epoch": 0.35624, + "grad_norm": 0.9015090521415571, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 35624 + }, + { + "epoch": 0.35625, + "grad_norm": 1.1797173644066858, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 35625 + }, + { + "epoch": 0.35626, + "grad_norm": 1.1794111485421754, + "learning_rate": 0.003, + "loss": 4.031, + "step": 35626 + }, + { + "epoch": 0.35627, + "grad_norm": 1.1782919883554077, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 35627 + }, + { + "epoch": 0.35628, + "grad_norm": 0.9927518390969776, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 35628 + }, + { + "epoch": 0.35629, + "grad_norm": 1.2007975186886282, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 35629 + }, + { + "epoch": 0.3563, + "grad_norm": 1.121296285187051, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 35630 + }, + { + "epoch": 0.35631, + "grad_norm": 1.0557190073621407, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 35631 + }, + { + "epoch": 0.35632, + "grad_norm": 1.2265821282552627, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 35632 + }, + { + "epoch": 0.35633, + "grad_norm": 0.9352107688465416, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 35633 + }, + { + "epoch": 0.35634, + "grad_norm": 1.0634653972568626, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 35634 + }, + { + "epoch": 0.35635, + "grad_norm": 1.074326032419467, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 35635 + }, + { + "epoch": 0.35636, + "grad_norm": 1.207737159304361, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 35636 + }, + { + "epoch": 0.35637, + "grad_norm": 1.221455551727294, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 35637 + }, + { + "epoch": 0.35638, + "grad_norm": 1.0080007003107792, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 35638 + }, + { + "epoch": 0.35639, + "grad_norm": 1.2790903809227745, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 35639 + }, + { + "epoch": 0.3564, + "grad_norm": 0.9250136766212208, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 35640 + }, + { + "epoch": 0.35641, + "grad_norm": 1.1563031102760521, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 35641 + }, + { + "epoch": 0.35642, + "grad_norm": 1.0819343136105706, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 35642 + }, + { + "epoch": 0.35643, + "grad_norm": 0.964486867644658, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 35643 + }, + { + "epoch": 0.35644, + "grad_norm": 1.15690347791061, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 35644 + }, + { + "epoch": 0.35645, + "grad_norm": 1.1117884992858538, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 35645 + }, + { + "epoch": 0.35646, + "grad_norm": 0.916545916093589, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 35646 + }, + { + "epoch": 0.35647, + "grad_norm": 1.051881925979099, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 35647 + }, + { + "epoch": 0.35648, + "grad_norm": 1.414801961075907, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 35648 + }, + { + "epoch": 0.35649, + "grad_norm": 0.9740581011408684, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 35649 + }, + { + "epoch": 0.3565, + "grad_norm": 1.040774847968717, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 35650 + }, + { + "epoch": 0.35651, + "grad_norm": 1.015862363538203, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 35651 + }, + { + "epoch": 0.35652, + "grad_norm": 1.2327602601149894, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 35652 + }, + { + "epoch": 0.35653, + "grad_norm": 1.0486345406038884, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 35653 + }, + { + "epoch": 0.35654, + "grad_norm": 0.9862457702207021, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 35654 + }, + { + "epoch": 0.35655, + "grad_norm": 1.1492272886270423, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 35655 + }, + { + "epoch": 0.35656, + "grad_norm": 0.9264287653502635, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 35656 + }, + { + "epoch": 0.35657, + "grad_norm": 1.149538444399977, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 35657 + }, + { + "epoch": 0.35658, + "grad_norm": 1.323578457145479, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 35658 + }, + { + "epoch": 0.35659, + "grad_norm": 0.8482956074012272, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 35659 + }, + { + "epoch": 0.3566, + "grad_norm": 0.9487989597777825, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 35660 + }, + { + "epoch": 0.35661, + "grad_norm": 1.2026475451847, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 35661 + }, + { + "epoch": 0.35662, + "grad_norm": 0.8986013808173199, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 35662 + }, + { + "epoch": 0.35663, + "grad_norm": 0.975855960148968, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 35663 + }, + { + "epoch": 0.35664, + "grad_norm": 1.1228405313200955, + "learning_rate": 0.003, + "loss": 4.039, + "step": 35664 + }, + { + "epoch": 0.35665, + "grad_norm": 1.1353852559698012, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 35665 + }, + { + "epoch": 0.35666, + "grad_norm": 1.0859077882750467, + "learning_rate": 0.003, + "loss": 4.017, + "step": 35666 + }, + { + "epoch": 0.35667, + "grad_norm": 1.0379438262751384, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 35667 + }, + { + "epoch": 0.35668, + "grad_norm": 1.2458276484932571, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 35668 + }, + { + "epoch": 0.35669, + "grad_norm": 1.2045893870510713, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 35669 + }, + { + "epoch": 0.3567, + "grad_norm": 0.7748144023106596, + "learning_rate": 0.003, + "loss": 4.045, + "step": 35670 + }, + { + "epoch": 0.35671, + "grad_norm": 0.8733028966951678, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 35671 + }, + { + "epoch": 0.35672, + "grad_norm": 1.046365096976066, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 35672 + }, + { + "epoch": 0.35673, + "grad_norm": 1.2570450287774066, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 35673 + }, + { + "epoch": 0.35674, + "grad_norm": 0.9704277569975305, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 35674 + }, + { + "epoch": 0.35675, + "grad_norm": 1.3550503115837986, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 35675 + }, + { + "epoch": 0.35676, + "grad_norm": 0.9632208908319092, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 35676 + }, + { + "epoch": 0.35677, + "grad_norm": 1.088864780526171, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 35677 + }, + { + "epoch": 0.35678, + "grad_norm": 1.1400624081281516, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 35678 + }, + { + "epoch": 0.35679, + "grad_norm": 1.1644919255870736, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 35679 + }, + { + "epoch": 0.3568, + "grad_norm": 1.4634163947114482, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 35680 + }, + { + "epoch": 0.35681, + "grad_norm": 0.8869969313878641, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 35681 + }, + { + "epoch": 0.35682, + "grad_norm": 0.8553524680512856, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 35682 + }, + { + "epoch": 0.35683, + "grad_norm": 0.8151499497556257, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 35683 + }, + { + "epoch": 0.35684, + "grad_norm": 1.0309591602525081, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 35684 + }, + { + "epoch": 0.35685, + "grad_norm": 1.0674102997582078, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 35685 + }, + { + "epoch": 0.35686, + "grad_norm": 0.9389063728036273, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 35686 + }, + { + "epoch": 0.35687, + "grad_norm": 0.9969005498880887, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 35687 + }, + { + "epoch": 0.35688, + "grad_norm": 1.1050640860137273, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 35688 + }, + { + "epoch": 0.35689, + "grad_norm": 1.2242331936242459, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 35689 + }, + { + "epoch": 0.3569, + "grad_norm": 1.0901308367646683, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 35690 + }, + { + "epoch": 0.35691, + "grad_norm": 1.3667194942814762, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 35691 + }, + { + "epoch": 0.35692, + "grad_norm": 1.2057899788822077, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 35692 + }, + { + "epoch": 0.35693, + "grad_norm": 1.06972140536982, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 35693 + }, + { + "epoch": 0.35694, + "grad_norm": 1.0909220799638217, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 35694 + }, + { + "epoch": 0.35695, + "grad_norm": 0.8644177125814803, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 35695 + }, + { + "epoch": 0.35696, + "grad_norm": 0.8863821476629986, + "learning_rate": 0.003, + "loss": 4.027, + "step": 35696 + }, + { + "epoch": 0.35697, + "grad_norm": 0.9757141104018375, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 35697 + }, + { + "epoch": 0.35698, + "grad_norm": 1.0228614587370006, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 35698 + }, + { + "epoch": 0.35699, + "grad_norm": 1.2263332822644764, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 35699 + }, + { + "epoch": 0.357, + "grad_norm": 1.0935253348376728, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 35700 + }, + { + "epoch": 0.35701, + "grad_norm": 1.1875192499926304, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 35701 + }, + { + "epoch": 0.35702, + "grad_norm": 1.244305614573257, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 35702 + }, + { + "epoch": 0.35703, + "grad_norm": 0.9840029557123735, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 35703 + }, + { + "epoch": 0.35704, + "grad_norm": 1.107010047053122, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 35704 + }, + { + "epoch": 0.35705, + "grad_norm": 1.260891099865273, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 35705 + }, + { + "epoch": 0.35706, + "grad_norm": 1.0555431855995006, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 35706 + }, + { + "epoch": 0.35707, + "grad_norm": 1.0694923739483264, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 35707 + }, + { + "epoch": 0.35708, + "grad_norm": 1.0594112564084948, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 35708 + }, + { + "epoch": 0.35709, + "grad_norm": 1.3203379465939142, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 35709 + }, + { + "epoch": 0.3571, + "grad_norm": 0.7756857281155664, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 35710 + }, + { + "epoch": 0.35711, + "grad_norm": 0.7877786431557353, + "learning_rate": 0.003, + "loss": 4.029, + "step": 35711 + }, + { + "epoch": 0.35712, + "grad_norm": 0.8328349186321663, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 35712 + }, + { + "epoch": 0.35713, + "grad_norm": 0.9927966515878314, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 35713 + }, + { + "epoch": 0.35714, + "grad_norm": 1.243786961452899, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 35714 + }, + { + "epoch": 0.35715, + "grad_norm": 1.0340576362494363, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 35715 + }, + { + "epoch": 0.35716, + "grad_norm": 1.066511815386739, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 35716 + }, + { + "epoch": 0.35717, + "grad_norm": 1.04937902439261, + "learning_rate": 0.003, + "loss": 4.022, + "step": 35717 + }, + { + "epoch": 0.35718, + "grad_norm": 1.0202250139664322, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 35718 + }, + { + "epoch": 0.35719, + "grad_norm": 1.2218370139406889, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 35719 + }, + { + "epoch": 0.3572, + "grad_norm": 1.0879964803264262, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 35720 + }, + { + "epoch": 0.35721, + "grad_norm": 1.5413922654624748, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 35721 + }, + { + "epoch": 0.35722, + "grad_norm": 0.8690920218071385, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 35722 + }, + { + "epoch": 0.35723, + "grad_norm": 0.7982946175290851, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 35723 + }, + { + "epoch": 0.35724, + "grad_norm": 0.8457158322811773, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 35724 + }, + { + "epoch": 0.35725, + "grad_norm": 1.0324202708828307, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 35725 + }, + { + "epoch": 0.35726, + "grad_norm": 1.1429607893672968, + "learning_rate": 0.003, + "loss": 4.044, + "step": 35726 + }, + { + "epoch": 0.35727, + "grad_norm": 1.0453527847616744, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 35727 + }, + { + "epoch": 0.35728, + "grad_norm": 1.4108924636322524, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 35728 + }, + { + "epoch": 0.35729, + "grad_norm": 1.211359741259417, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 35729 + }, + { + "epoch": 0.3573, + "grad_norm": 0.999315639574949, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 35730 + }, + { + "epoch": 0.35731, + "grad_norm": 1.3153601306371958, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 35731 + }, + { + "epoch": 0.35732, + "grad_norm": 1.0346041532222303, + "learning_rate": 0.003, + "loss": 4.075, + "step": 35732 + }, + { + "epoch": 0.35733, + "grad_norm": 0.9909810315040625, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 35733 + }, + { + "epoch": 0.35734, + "grad_norm": 1.1597675509932412, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 35734 + }, + { + "epoch": 0.35735, + "grad_norm": 1.060834485082214, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 35735 + }, + { + "epoch": 0.35736, + "grad_norm": 1.023467349165769, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 35736 + }, + { + "epoch": 0.35737, + "grad_norm": 1.0518921431944321, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 35737 + }, + { + "epoch": 0.35738, + "grad_norm": 1.170067715978499, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 35738 + }, + { + "epoch": 0.35739, + "grad_norm": 1.1408298020791605, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 35739 + }, + { + "epoch": 0.3574, + "grad_norm": 1.2299211233879768, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 35740 + }, + { + "epoch": 0.35741, + "grad_norm": 1.1339918782526537, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 35741 + }, + { + "epoch": 0.35742, + "grad_norm": 1.323602065143769, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 35742 + }, + { + "epoch": 0.35743, + "grad_norm": 0.9965993515454062, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 35743 + }, + { + "epoch": 0.35744, + "grad_norm": 0.9755165292316584, + "learning_rate": 0.003, + "loss": 4.032, + "step": 35744 + }, + { + "epoch": 0.35745, + "grad_norm": 1.0605848807607334, + "learning_rate": 0.003, + "loss": 4.003, + "step": 35745 + }, + { + "epoch": 0.35746, + "grad_norm": 1.0422461645862144, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 35746 + }, + { + "epoch": 0.35747, + "grad_norm": 1.1282984440127601, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 35747 + }, + { + "epoch": 0.35748, + "grad_norm": 0.98194228005656, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 35748 + }, + { + "epoch": 0.35749, + "grad_norm": 1.1987361728625345, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 35749 + }, + { + "epoch": 0.3575, + "grad_norm": 0.8753301145347189, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 35750 + }, + { + "epoch": 0.35751, + "grad_norm": 0.8236747546646713, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 35751 + }, + { + "epoch": 0.35752, + "grad_norm": 1.2298668409764055, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 35752 + }, + { + "epoch": 0.35753, + "grad_norm": 0.9455400051654079, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 35753 + }, + { + "epoch": 0.35754, + "grad_norm": 0.9010852793385364, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 35754 + }, + { + "epoch": 0.35755, + "grad_norm": 1.0005285333890195, + "learning_rate": 0.003, + "loss": 4.051, + "step": 35755 + }, + { + "epoch": 0.35756, + "grad_norm": 1.0874045105633858, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 35756 + }, + { + "epoch": 0.35757, + "grad_norm": 1.0067177449916684, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 35757 + }, + { + "epoch": 0.35758, + "grad_norm": 1.1897712894202035, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 35758 + }, + { + "epoch": 0.35759, + "grad_norm": 1.095613057979795, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 35759 + }, + { + "epoch": 0.3576, + "grad_norm": 0.9968464228293243, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 35760 + }, + { + "epoch": 0.35761, + "grad_norm": 1.349757927527724, + "learning_rate": 0.003, + "loss": 4.051, + "step": 35761 + }, + { + "epoch": 0.35762, + "grad_norm": 1.3152335508583783, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 35762 + }, + { + "epoch": 0.35763, + "grad_norm": 1.00074396874111, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 35763 + }, + { + "epoch": 0.35764, + "grad_norm": 0.9324179306428958, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 35764 + }, + { + "epoch": 0.35765, + "grad_norm": 1.0913285289031274, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 35765 + }, + { + "epoch": 0.35766, + "grad_norm": 1.101808688964492, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 35766 + }, + { + "epoch": 0.35767, + "grad_norm": 1.4532430820547446, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 35767 + }, + { + "epoch": 0.35768, + "grad_norm": 0.9497352597629249, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 35768 + }, + { + "epoch": 0.35769, + "grad_norm": 0.8926493836360513, + "learning_rate": 0.003, + "loss": 4.013, + "step": 35769 + }, + { + "epoch": 0.3577, + "grad_norm": 0.8304895120993552, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 35770 + }, + { + "epoch": 0.35771, + "grad_norm": 0.9619171720751024, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 35771 + }, + { + "epoch": 0.35772, + "grad_norm": 1.2345605951835634, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 35772 + }, + { + "epoch": 0.35773, + "grad_norm": 1.0194109379250427, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 35773 + }, + { + "epoch": 0.35774, + "grad_norm": 0.9553705337896224, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 35774 + }, + { + "epoch": 0.35775, + "grad_norm": 1.1317281986965422, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 35775 + }, + { + "epoch": 0.35776, + "grad_norm": 1.053860645295171, + "learning_rate": 0.003, + "loss": 4.056, + "step": 35776 + }, + { + "epoch": 0.35777, + "grad_norm": 1.3876847396893086, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 35777 + }, + { + "epoch": 0.35778, + "grad_norm": 1.106003179401289, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 35778 + }, + { + "epoch": 0.35779, + "grad_norm": 1.008394350001395, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 35779 + }, + { + "epoch": 0.3578, + "grad_norm": 1.1707808933822899, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 35780 + }, + { + "epoch": 0.35781, + "grad_norm": 1.0756963779232809, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 35781 + }, + { + "epoch": 0.35782, + "grad_norm": 1.2437568836985613, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 35782 + }, + { + "epoch": 0.35783, + "grad_norm": 0.966795930430896, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 35783 + }, + { + "epoch": 0.35784, + "grad_norm": 1.2722536729171117, + "learning_rate": 0.003, + "loss": 4.023, + "step": 35784 + }, + { + "epoch": 0.35785, + "grad_norm": 1.0237766191405684, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 35785 + }, + { + "epoch": 0.35786, + "grad_norm": 0.9890188196264915, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 35786 + }, + { + "epoch": 0.35787, + "grad_norm": 1.0778845997855517, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 35787 + }, + { + "epoch": 0.35788, + "grad_norm": 1.1630532380052598, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 35788 + }, + { + "epoch": 0.35789, + "grad_norm": 1.037308715832548, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 35789 + }, + { + "epoch": 0.3579, + "grad_norm": 1.1257523072042281, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 35790 + }, + { + "epoch": 0.35791, + "grad_norm": 0.8661226849717869, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 35791 + }, + { + "epoch": 0.35792, + "grad_norm": 0.996302428277914, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 35792 + }, + { + "epoch": 0.35793, + "grad_norm": 1.1661768823600627, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 35793 + }, + { + "epoch": 0.35794, + "grad_norm": 1.1651865334769493, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 35794 + }, + { + "epoch": 0.35795, + "grad_norm": 1.1411011399742548, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 35795 + }, + { + "epoch": 0.35796, + "grad_norm": 1.026177328121155, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 35796 + }, + { + "epoch": 0.35797, + "grad_norm": 1.53268015811194, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 35797 + }, + { + "epoch": 0.35798, + "grad_norm": 0.8563165288774551, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 35798 + }, + { + "epoch": 0.35799, + "grad_norm": 0.8418764446633559, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 35799 + }, + { + "epoch": 0.358, + "grad_norm": 1.0579538069059304, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 35800 + }, + { + "epoch": 0.35801, + "grad_norm": 1.1219600250803239, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 35801 + }, + { + "epoch": 0.35802, + "grad_norm": 1.12612404902505, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 35802 + }, + { + "epoch": 0.35803, + "grad_norm": 1.5601071195529528, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 35803 + }, + { + "epoch": 0.35804, + "grad_norm": 0.8040117828538547, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 35804 + }, + { + "epoch": 0.35805, + "grad_norm": 1.015965770261241, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 35805 + }, + { + "epoch": 0.35806, + "grad_norm": 1.435378709086305, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 35806 + }, + { + "epoch": 0.35807, + "grad_norm": 0.9274960035812129, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 35807 + }, + { + "epoch": 0.35808, + "grad_norm": 1.217517452695287, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 35808 + }, + { + "epoch": 0.35809, + "grad_norm": 0.9800283408695148, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 35809 + }, + { + "epoch": 0.3581, + "grad_norm": 0.9150424524829164, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 35810 + }, + { + "epoch": 0.35811, + "grad_norm": 1.1744580408140428, + "learning_rate": 0.003, + "loss": 4.066, + "step": 35811 + }, + { + "epoch": 0.35812, + "grad_norm": 1.1864878757685586, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 35812 + }, + { + "epoch": 0.35813, + "grad_norm": 0.8992374523457576, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 35813 + }, + { + "epoch": 0.35814, + "grad_norm": 0.9409376785277731, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 35814 + }, + { + "epoch": 0.35815, + "grad_norm": 1.274115148965271, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 35815 + }, + { + "epoch": 0.35816, + "grad_norm": 1.1038500960700188, + "learning_rate": 0.003, + "loss": 4.019, + "step": 35816 + }, + { + "epoch": 0.35817, + "grad_norm": 1.139905548062641, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 35817 + }, + { + "epoch": 0.35818, + "grad_norm": 0.7848474610621521, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 35818 + }, + { + "epoch": 0.35819, + "grad_norm": 0.7205018541493902, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 35819 + }, + { + "epoch": 0.3582, + "grad_norm": 0.7344011333332022, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 35820 + }, + { + "epoch": 0.35821, + "grad_norm": 0.8077792908300779, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 35821 + }, + { + "epoch": 0.35822, + "grad_norm": 1.0053449113978792, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 35822 + }, + { + "epoch": 0.35823, + "grad_norm": 1.1709297702252548, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 35823 + }, + { + "epoch": 0.35824, + "grad_norm": 0.9618195980987533, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 35824 + }, + { + "epoch": 0.35825, + "grad_norm": 1.2746675834597778, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 35825 + }, + { + "epoch": 0.35826, + "grad_norm": 1.329588526875158, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 35826 + }, + { + "epoch": 0.35827, + "grad_norm": 0.9632936713172245, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 35827 + }, + { + "epoch": 0.35828, + "grad_norm": 1.0470876935104123, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 35828 + }, + { + "epoch": 0.35829, + "grad_norm": 1.0996550932674982, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 35829 + }, + { + "epoch": 0.3583, + "grad_norm": 1.187766321885879, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 35830 + }, + { + "epoch": 0.35831, + "grad_norm": 0.9971388001112075, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 35831 + }, + { + "epoch": 0.35832, + "grad_norm": 1.3337422308684386, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 35832 + }, + { + "epoch": 0.35833, + "grad_norm": 0.8691540396057957, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 35833 + }, + { + "epoch": 0.35834, + "grad_norm": 1.016323924615636, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 35834 + }, + { + "epoch": 0.35835, + "grad_norm": 1.3689648839772788, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 35835 + }, + { + "epoch": 0.35836, + "grad_norm": 0.9383861540302391, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 35836 + }, + { + "epoch": 0.35837, + "grad_norm": 1.0656230698481688, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 35837 + }, + { + "epoch": 0.35838, + "grad_norm": 1.4057551016334011, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 35838 + }, + { + "epoch": 0.35839, + "grad_norm": 0.9732374556965492, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 35839 + }, + { + "epoch": 0.3584, + "grad_norm": 1.1129779721948714, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 35840 + }, + { + "epoch": 0.35841, + "grad_norm": 0.941265196254004, + "learning_rate": 0.003, + "loss": 4.043, + "step": 35841 + }, + { + "epoch": 0.35842, + "grad_norm": 1.0029577424939347, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 35842 + }, + { + "epoch": 0.35843, + "grad_norm": 1.1582231752035446, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 35843 + }, + { + "epoch": 0.35844, + "grad_norm": 1.0131160496211942, + "learning_rate": 0.003, + "loss": 4.049, + "step": 35844 + }, + { + "epoch": 0.35845, + "grad_norm": 0.9849939637346845, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 35845 + }, + { + "epoch": 0.35846, + "grad_norm": 1.0887974084254228, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 35846 + }, + { + "epoch": 0.35847, + "grad_norm": 1.1973213934325593, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 35847 + }, + { + "epoch": 0.35848, + "grad_norm": 1.2824071109099018, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 35848 + }, + { + "epoch": 0.35849, + "grad_norm": 1.036013948155982, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 35849 + }, + { + "epoch": 0.3585, + "grad_norm": 1.3448685995474186, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 35850 + }, + { + "epoch": 0.35851, + "grad_norm": 0.982094405340596, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 35851 + }, + { + "epoch": 0.35852, + "grad_norm": 1.1815977572631438, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 35852 + }, + { + "epoch": 0.35853, + "grad_norm": 1.184019007999987, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 35853 + }, + { + "epoch": 0.35854, + "grad_norm": 0.918260000046395, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 35854 + }, + { + "epoch": 0.35855, + "grad_norm": 1.135312016901875, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 35855 + }, + { + "epoch": 0.35856, + "grad_norm": 1.045492017844325, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 35856 + }, + { + "epoch": 0.35857, + "grad_norm": 1.1158533291684332, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 35857 + }, + { + "epoch": 0.35858, + "grad_norm": 1.1878868738355786, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 35858 + }, + { + "epoch": 0.35859, + "grad_norm": 1.0881820866670344, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 35859 + }, + { + "epoch": 0.3586, + "grad_norm": 0.9164900102854271, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 35860 + }, + { + "epoch": 0.35861, + "grad_norm": 1.2863704353816432, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 35861 + }, + { + "epoch": 0.35862, + "grad_norm": 1.160654212207456, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 35862 + }, + { + "epoch": 0.35863, + "grad_norm": 1.0746772210342965, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 35863 + }, + { + "epoch": 0.35864, + "grad_norm": 1.2690060777667265, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 35864 + }, + { + "epoch": 0.35865, + "grad_norm": 0.9119801752590876, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 35865 + }, + { + "epoch": 0.35866, + "grad_norm": 0.8984319625600701, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 35866 + }, + { + "epoch": 0.35867, + "grad_norm": 1.1693055531858187, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 35867 + }, + { + "epoch": 0.35868, + "grad_norm": 0.9628607455590213, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 35868 + }, + { + "epoch": 0.35869, + "grad_norm": 0.9664347001503688, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 35869 + }, + { + "epoch": 0.3587, + "grad_norm": 1.093366952235885, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 35870 + }, + { + "epoch": 0.35871, + "grad_norm": 1.1705982508039643, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 35871 + }, + { + "epoch": 0.35872, + "grad_norm": 1.169935855012751, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 35872 + }, + { + "epoch": 0.35873, + "grad_norm": 1.1479453369611845, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 35873 + }, + { + "epoch": 0.35874, + "grad_norm": 1.0747576322116403, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 35874 + }, + { + "epoch": 0.35875, + "grad_norm": 0.915739671011638, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 35875 + }, + { + "epoch": 0.35876, + "grad_norm": 1.3060806092312003, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 35876 + }, + { + "epoch": 0.35877, + "grad_norm": 1.1089798413360283, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 35877 + }, + { + "epoch": 0.35878, + "grad_norm": 1.0623311328006633, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 35878 + }, + { + "epoch": 0.35879, + "grad_norm": 1.1731485015600371, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 35879 + }, + { + "epoch": 0.3588, + "grad_norm": 0.9292100412231236, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 35880 + }, + { + "epoch": 0.35881, + "grad_norm": 0.9634777670124951, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 35881 + }, + { + "epoch": 0.35882, + "grad_norm": 1.0946760771244628, + "learning_rate": 0.003, + "loss": 4.076, + "step": 35882 + }, + { + "epoch": 0.35883, + "grad_norm": 1.2557321730083848, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 35883 + }, + { + "epoch": 0.35884, + "grad_norm": 1.0987508902420002, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 35884 + }, + { + "epoch": 0.35885, + "grad_norm": 1.2085307761786628, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 35885 + }, + { + "epoch": 0.35886, + "grad_norm": 0.995665406102532, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 35886 + }, + { + "epoch": 0.35887, + "grad_norm": 1.1375603531353304, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 35887 + }, + { + "epoch": 0.35888, + "grad_norm": 0.9649100520338744, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 35888 + }, + { + "epoch": 0.35889, + "grad_norm": 0.9147231718300748, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 35889 + }, + { + "epoch": 0.3589, + "grad_norm": 0.9361977479580739, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 35890 + }, + { + "epoch": 0.35891, + "grad_norm": 1.087286565291011, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 35891 + }, + { + "epoch": 0.35892, + "grad_norm": 1.0885246395018298, + "learning_rate": 0.003, + "loss": 4.016, + "step": 35892 + }, + { + "epoch": 0.35893, + "grad_norm": 1.1433879934293973, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 35893 + }, + { + "epoch": 0.35894, + "grad_norm": 1.4493390237722907, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 35894 + }, + { + "epoch": 0.35895, + "grad_norm": 0.7933377882196838, + "learning_rate": 0.003, + "loss": 4.02, + "step": 35895 + }, + { + "epoch": 0.35896, + "grad_norm": 0.888916527760576, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 35896 + }, + { + "epoch": 0.35897, + "grad_norm": 1.145151612950037, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 35897 + }, + { + "epoch": 0.35898, + "grad_norm": 1.159993956123528, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 35898 + }, + { + "epoch": 0.35899, + "grad_norm": 0.9365132396451717, + "learning_rate": 0.003, + "loss": 4.034, + "step": 35899 + }, + { + "epoch": 0.359, + "grad_norm": 1.0232640508846012, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 35900 + }, + { + "epoch": 0.35901, + "grad_norm": 1.3120445073824958, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 35901 + }, + { + "epoch": 0.35902, + "grad_norm": 0.9373423755114666, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 35902 + }, + { + "epoch": 0.35903, + "grad_norm": 0.9905011838264411, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 35903 + }, + { + "epoch": 0.35904, + "grad_norm": 1.244919798412617, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 35904 + }, + { + "epoch": 0.35905, + "grad_norm": 0.8793693832226369, + "learning_rate": 0.003, + "loss": 4.026, + "step": 35905 + }, + { + "epoch": 0.35906, + "grad_norm": 1.0061036368456302, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 35906 + }, + { + "epoch": 0.35907, + "grad_norm": 1.143937881425009, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 35907 + }, + { + "epoch": 0.35908, + "grad_norm": 1.1599418003621311, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 35908 + }, + { + "epoch": 0.35909, + "grad_norm": 0.9953138874788132, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 35909 + }, + { + "epoch": 0.3591, + "grad_norm": 1.3505185466698963, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 35910 + }, + { + "epoch": 0.35911, + "grad_norm": 0.920329352237062, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 35911 + }, + { + "epoch": 0.35912, + "grad_norm": 1.2526203436698395, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 35912 + }, + { + "epoch": 0.35913, + "grad_norm": 1.0915535059250803, + "learning_rate": 0.003, + "loss": 4.044, + "step": 35913 + }, + { + "epoch": 0.35914, + "grad_norm": 1.0228985903726004, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 35914 + }, + { + "epoch": 0.35915, + "grad_norm": 1.1515370204238762, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 35915 + }, + { + "epoch": 0.35916, + "grad_norm": 1.2321297936227962, + "learning_rate": 0.003, + "loss": 4.038, + "step": 35916 + }, + { + "epoch": 0.35917, + "grad_norm": 1.11247258148998, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 35917 + }, + { + "epoch": 0.35918, + "grad_norm": 1.0347587130234692, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 35918 + }, + { + "epoch": 0.35919, + "grad_norm": 1.1502973908195075, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 35919 + }, + { + "epoch": 0.3592, + "grad_norm": 0.9669336426756877, + "learning_rate": 0.003, + "loss": 4.021, + "step": 35920 + }, + { + "epoch": 0.35921, + "grad_norm": 0.9884267923130708, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 35921 + }, + { + "epoch": 0.35922, + "grad_norm": 0.9359977727619547, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 35922 + }, + { + "epoch": 0.35923, + "grad_norm": 0.8627473503644087, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 35923 + }, + { + "epoch": 0.35924, + "grad_norm": 0.9877955179872208, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 35924 + }, + { + "epoch": 0.35925, + "grad_norm": 0.9120516245639521, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 35925 + }, + { + "epoch": 0.35926, + "grad_norm": 1.0994219351827794, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 35926 + }, + { + "epoch": 0.35927, + "grad_norm": 1.2441822004182244, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 35927 + }, + { + "epoch": 0.35928, + "grad_norm": 1.096426772918347, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 35928 + }, + { + "epoch": 0.35929, + "grad_norm": 1.1655576820177629, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 35929 + }, + { + "epoch": 0.3593, + "grad_norm": 1.2475895915354829, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 35930 + }, + { + "epoch": 0.35931, + "grad_norm": 1.24548922706177, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 35931 + }, + { + "epoch": 0.35932, + "grad_norm": 1.2280697538389749, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 35932 + }, + { + "epoch": 0.35933, + "grad_norm": 1.034241638223239, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 35933 + }, + { + "epoch": 0.35934, + "grad_norm": 1.2311927532984794, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 35934 + }, + { + "epoch": 0.35935, + "grad_norm": 1.2567773344327269, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 35935 + }, + { + "epoch": 0.35936, + "grad_norm": 0.9378995645713172, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 35936 + }, + { + "epoch": 0.35937, + "grad_norm": 1.0049452702175616, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 35937 + }, + { + "epoch": 0.35938, + "grad_norm": 1.3491478131636632, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 35938 + }, + { + "epoch": 0.35939, + "grad_norm": 1.0179861945004784, + "learning_rate": 0.003, + "loss": 4.049, + "step": 35939 + }, + { + "epoch": 0.3594, + "grad_norm": 1.334914711397424, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 35940 + }, + { + "epoch": 0.35941, + "grad_norm": 0.8744574160766733, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 35941 + }, + { + "epoch": 0.35942, + "grad_norm": 0.8826818806545126, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 35942 + }, + { + "epoch": 0.35943, + "grad_norm": 1.0856531153919524, + "learning_rate": 0.003, + "loss": 4.037, + "step": 35943 + }, + { + "epoch": 0.35944, + "grad_norm": 1.2555403883375476, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 35944 + }, + { + "epoch": 0.35945, + "grad_norm": 1.1054270375417703, + "learning_rate": 0.003, + "loss": 4.026, + "step": 35945 + }, + { + "epoch": 0.35946, + "grad_norm": 0.9827347723116815, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 35946 + }, + { + "epoch": 0.35947, + "grad_norm": 0.9144172215447615, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 35947 + }, + { + "epoch": 0.35948, + "grad_norm": 1.1585568761176017, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 35948 + }, + { + "epoch": 0.35949, + "grad_norm": 1.4312178535170008, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 35949 + }, + { + "epoch": 0.3595, + "grad_norm": 0.774285620174308, + "learning_rate": 0.003, + "loss": 4.004, + "step": 35950 + }, + { + "epoch": 0.35951, + "grad_norm": 0.8479833578100805, + "learning_rate": 0.003, + "loss": 4.052, + "step": 35951 + }, + { + "epoch": 0.35952, + "grad_norm": 1.0987144011474286, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 35952 + }, + { + "epoch": 0.35953, + "grad_norm": 1.4226312996567116, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 35953 + }, + { + "epoch": 0.35954, + "grad_norm": 0.7858516052888822, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 35954 + }, + { + "epoch": 0.35955, + "grad_norm": 0.6843235512822887, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 35955 + }, + { + "epoch": 0.35956, + "grad_norm": 0.7986196150473206, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 35956 + }, + { + "epoch": 0.35957, + "grad_norm": 1.0060260102692893, + "learning_rate": 0.003, + "loss": 4.062, + "step": 35957 + }, + { + "epoch": 0.35958, + "grad_norm": 1.2680758329040083, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 35958 + }, + { + "epoch": 0.35959, + "grad_norm": 1.0850639749877213, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 35959 + }, + { + "epoch": 0.3596, + "grad_norm": 1.3050759521892579, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 35960 + }, + { + "epoch": 0.35961, + "grad_norm": 0.9498943978762849, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 35961 + }, + { + "epoch": 0.35962, + "grad_norm": 1.0623266027992977, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 35962 + }, + { + "epoch": 0.35963, + "grad_norm": 1.1239782531866933, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 35963 + }, + { + "epoch": 0.35964, + "grad_norm": 1.152748403196938, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 35964 + }, + { + "epoch": 0.35965, + "grad_norm": 1.0031654739074363, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 35965 + }, + { + "epoch": 0.35966, + "grad_norm": 1.0472105683467807, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 35966 + }, + { + "epoch": 0.35967, + "grad_norm": 0.9532980556241696, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 35967 + }, + { + "epoch": 0.35968, + "grad_norm": 0.8973505786296024, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 35968 + }, + { + "epoch": 0.35969, + "grad_norm": 1.0646733788844887, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 35969 + }, + { + "epoch": 0.3597, + "grad_norm": 1.2815575073839818, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 35970 + }, + { + "epoch": 0.35971, + "grad_norm": 0.9235678254461337, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 35971 + }, + { + "epoch": 0.35972, + "grad_norm": 1.0997666308840912, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 35972 + }, + { + "epoch": 0.35973, + "grad_norm": 1.0631960925747952, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 35973 + }, + { + "epoch": 0.35974, + "grad_norm": 1.0261863030874936, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 35974 + }, + { + "epoch": 0.35975, + "grad_norm": 1.233418943090699, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 35975 + }, + { + "epoch": 0.35976, + "grad_norm": 1.2251523030599218, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 35976 + }, + { + "epoch": 0.35977, + "grad_norm": 1.0414697880102275, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 35977 + }, + { + "epoch": 0.35978, + "grad_norm": 1.150766311351624, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 35978 + }, + { + "epoch": 0.35979, + "grad_norm": 1.2509408258561237, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 35979 + }, + { + "epoch": 0.3598, + "grad_norm": 0.888213228038585, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 35980 + }, + { + "epoch": 0.35981, + "grad_norm": 0.9134756293056818, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 35981 + }, + { + "epoch": 0.35982, + "grad_norm": 1.0787958195698077, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 35982 + }, + { + "epoch": 0.35983, + "grad_norm": 1.106573022404464, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 35983 + }, + { + "epoch": 0.35984, + "grad_norm": 0.8846442535957574, + "learning_rate": 0.003, + "loss": 4.076, + "step": 35984 + }, + { + "epoch": 0.35985, + "grad_norm": 0.9687367849372949, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 35985 + }, + { + "epoch": 0.35986, + "grad_norm": 1.1950470524670445, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 35986 + }, + { + "epoch": 0.35987, + "grad_norm": 1.1137827303838606, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 35987 + }, + { + "epoch": 0.35988, + "grad_norm": 1.1878607737486422, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 35988 + }, + { + "epoch": 0.35989, + "grad_norm": 1.1586918775453106, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 35989 + }, + { + "epoch": 0.3599, + "grad_norm": 0.9747793526660342, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 35990 + }, + { + "epoch": 0.35991, + "grad_norm": 1.165412478583046, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 35991 + }, + { + "epoch": 0.35992, + "grad_norm": 1.012515820765459, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 35992 + }, + { + "epoch": 0.35993, + "grad_norm": 1.298795013374699, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 35993 + }, + { + "epoch": 0.35994, + "grad_norm": 0.9510973704396044, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 35994 + }, + { + "epoch": 0.35995, + "grad_norm": 1.309378130007583, + "learning_rate": 0.003, + "loss": 4.046, + "step": 35995 + }, + { + "epoch": 0.35996, + "grad_norm": 1.0758631891835555, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 35996 + }, + { + "epoch": 0.35997, + "grad_norm": 1.1562808350183247, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 35997 + }, + { + "epoch": 0.35998, + "grad_norm": 1.1551447835902418, + "learning_rate": 0.003, + "loss": 4.025, + "step": 35998 + }, + { + "epoch": 0.35999, + "grad_norm": 0.9556011699293208, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 35999 + }, + { + "epoch": 0.36, + "grad_norm": 1.205737302291432, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 36000 + }, + { + "epoch": 0.36001, + "grad_norm": 0.9064928872950471, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 36001 + }, + { + "epoch": 0.36002, + "grad_norm": 0.9310026858269284, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 36002 + }, + { + "epoch": 0.36003, + "grad_norm": 1.1792014848521502, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 36003 + }, + { + "epoch": 0.36004, + "grad_norm": 1.1408276843198146, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 36004 + }, + { + "epoch": 0.36005, + "grad_norm": 1.0923475366718343, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 36005 + }, + { + "epoch": 0.36006, + "grad_norm": 1.1060345940049727, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 36006 + }, + { + "epoch": 0.36007, + "grad_norm": 0.9749511769602551, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 36007 + }, + { + "epoch": 0.36008, + "grad_norm": 1.1182596210212394, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 36008 + }, + { + "epoch": 0.36009, + "grad_norm": 1.142872810213152, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 36009 + }, + { + "epoch": 0.3601, + "grad_norm": 1.0996389503404096, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 36010 + }, + { + "epoch": 0.36011, + "grad_norm": 1.152052615509361, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 36011 + }, + { + "epoch": 0.36012, + "grad_norm": 1.0749742208997408, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 36012 + }, + { + "epoch": 0.36013, + "grad_norm": 1.057428007917624, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 36013 + }, + { + "epoch": 0.36014, + "grad_norm": 1.1608667399832897, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 36014 + }, + { + "epoch": 0.36015, + "grad_norm": 1.1220464278618827, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 36015 + }, + { + "epoch": 0.36016, + "grad_norm": 1.0060240795021715, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 36016 + }, + { + "epoch": 0.36017, + "grad_norm": 1.3125613543079664, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 36017 + }, + { + "epoch": 0.36018, + "grad_norm": 0.971419422563042, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 36018 + }, + { + "epoch": 0.36019, + "grad_norm": 1.198073520925754, + "learning_rate": 0.003, + "loss": 4.026, + "step": 36019 + }, + { + "epoch": 0.3602, + "grad_norm": 1.0081078860474633, + "learning_rate": 0.003, + "loss": 4.007, + "step": 36020 + }, + { + "epoch": 0.36021, + "grad_norm": 1.2750021520569337, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 36021 + }, + { + "epoch": 0.36022, + "grad_norm": 0.918417669048894, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 36022 + }, + { + "epoch": 0.36023, + "grad_norm": 0.9538622602075877, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 36023 + }, + { + "epoch": 0.36024, + "grad_norm": 1.0420606286393583, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 36024 + }, + { + "epoch": 0.36025, + "grad_norm": 1.3661306372058493, + "learning_rate": 0.003, + "loss": 4.06, + "step": 36025 + }, + { + "epoch": 0.36026, + "grad_norm": 0.9765400742290132, + "learning_rate": 0.003, + "loss": 4.061, + "step": 36026 + }, + { + "epoch": 0.36027, + "grad_norm": 1.1784919882969422, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 36027 + }, + { + "epoch": 0.36028, + "grad_norm": 1.0908782420413945, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 36028 + }, + { + "epoch": 0.36029, + "grad_norm": 1.006210356361578, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 36029 + }, + { + "epoch": 0.3603, + "grad_norm": 0.904364562288725, + "learning_rate": 0.003, + "loss": 4.056, + "step": 36030 + }, + { + "epoch": 0.36031, + "grad_norm": 0.9964145699634117, + "learning_rate": 0.003, + "loss": 4.066, + "step": 36031 + }, + { + "epoch": 0.36032, + "grad_norm": 1.183675352670878, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 36032 + }, + { + "epoch": 0.36033, + "grad_norm": 1.002985787635039, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 36033 + }, + { + "epoch": 0.36034, + "grad_norm": 1.1614044180872607, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 36034 + }, + { + "epoch": 0.36035, + "grad_norm": 0.970647838745693, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 36035 + }, + { + "epoch": 0.36036, + "grad_norm": 1.057217555158728, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 36036 + }, + { + "epoch": 0.36037, + "grad_norm": 1.054804695300783, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 36037 + }, + { + "epoch": 0.36038, + "grad_norm": 1.0630749844798695, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 36038 + }, + { + "epoch": 0.36039, + "grad_norm": 0.9986156062360545, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 36039 + }, + { + "epoch": 0.3604, + "grad_norm": 1.1888824734351455, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 36040 + }, + { + "epoch": 0.36041, + "grad_norm": 1.0802074646279043, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 36041 + }, + { + "epoch": 0.36042, + "grad_norm": 1.0849073660714688, + "learning_rate": 0.003, + "loss": 4.052, + "step": 36042 + }, + { + "epoch": 0.36043, + "grad_norm": 1.1002329780094693, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 36043 + }, + { + "epoch": 0.36044, + "grad_norm": 1.2217473838042294, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 36044 + }, + { + "epoch": 0.36045, + "grad_norm": 1.1989257908921855, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 36045 + }, + { + "epoch": 0.36046, + "grad_norm": 1.1758973226872143, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 36046 + }, + { + "epoch": 0.36047, + "grad_norm": 1.115378716734179, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 36047 + }, + { + "epoch": 0.36048, + "grad_norm": 1.1486077294408341, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 36048 + }, + { + "epoch": 0.36049, + "grad_norm": 1.0881129135924625, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 36049 + }, + { + "epoch": 0.3605, + "grad_norm": 1.0920368456053995, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 36050 + }, + { + "epoch": 0.36051, + "grad_norm": 1.040484739707713, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 36051 + }, + { + "epoch": 0.36052, + "grad_norm": 1.041472865115166, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 36052 + }, + { + "epoch": 0.36053, + "grad_norm": 1.0662792617868555, + "learning_rate": 0.003, + "loss": 4.023, + "step": 36053 + }, + { + "epoch": 0.36054, + "grad_norm": 1.2381343926511674, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 36054 + }, + { + "epoch": 0.36055, + "grad_norm": 0.9548361125873576, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 36055 + }, + { + "epoch": 0.36056, + "grad_norm": 1.0205992313195607, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 36056 + }, + { + "epoch": 0.36057, + "grad_norm": 0.9738095977022263, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 36057 + }, + { + "epoch": 0.36058, + "grad_norm": 1.1226250101993387, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 36058 + }, + { + "epoch": 0.36059, + "grad_norm": 1.0419718803016147, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 36059 + }, + { + "epoch": 0.3606, + "grad_norm": 1.152353376689778, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 36060 + }, + { + "epoch": 0.36061, + "grad_norm": 1.1723745203078892, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 36061 + }, + { + "epoch": 0.36062, + "grad_norm": 1.072524402233613, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 36062 + }, + { + "epoch": 0.36063, + "grad_norm": 1.0084044268906842, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 36063 + }, + { + "epoch": 0.36064, + "grad_norm": 1.0247367898076876, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 36064 + }, + { + "epoch": 0.36065, + "grad_norm": 1.133918058226698, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 36065 + }, + { + "epoch": 0.36066, + "grad_norm": 1.220840481353774, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 36066 + }, + { + "epoch": 0.36067, + "grad_norm": 1.2598353222817844, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 36067 + }, + { + "epoch": 0.36068, + "grad_norm": 1.1911913655834523, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 36068 + }, + { + "epoch": 0.36069, + "grad_norm": 1.0350504042371345, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 36069 + }, + { + "epoch": 0.3607, + "grad_norm": 1.083366712005143, + "learning_rate": 0.003, + "loss": 4.042, + "step": 36070 + }, + { + "epoch": 0.36071, + "grad_norm": 1.0640185481277686, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 36071 + }, + { + "epoch": 0.36072, + "grad_norm": 1.1280322068194044, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 36072 + }, + { + "epoch": 0.36073, + "grad_norm": 1.013127173024313, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 36073 + }, + { + "epoch": 0.36074, + "grad_norm": 1.1545159753213776, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 36074 + }, + { + "epoch": 0.36075, + "grad_norm": 1.23454496487031, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 36075 + }, + { + "epoch": 0.36076, + "grad_norm": 0.9923642761905479, + "learning_rate": 0.003, + "loss": 4.029, + "step": 36076 + }, + { + "epoch": 0.36077, + "grad_norm": 1.3286902951135684, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 36077 + }, + { + "epoch": 0.36078, + "grad_norm": 0.9591528360268986, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 36078 + }, + { + "epoch": 0.36079, + "grad_norm": 1.1481235681044162, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 36079 + }, + { + "epoch": 0.3608, + "grad_norm": 1.1951656750650466, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 36080 + }, + { + "epoch": 0.36081, + "grad_norm": 0.9400907967658543, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 36081 + }, + { + "epoch": 0.36082, + "grad_norm": 0.9785516671166529, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 36082 + }, + { + "epoch": 0.36083, + "grad_norm": 1.3050273290808523, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 36083 + }, + { + "epoch": 0.36084, + "grad_norm": 0.9903167392161238, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 36084 + }, + { + "epoch": 0.36085, + "grad_norm": 1.3278092772464738, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 36085 + }, + { + "epoch": 0.36086, + "grad_norm": 1.0657251497683644, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 36086 + }, + { + "epoch": 0.36087, + "grad_norm": 1.0525323956558388, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 36087 + }, + { + "epoch": 0.36088, + "grad_norm": 1.0193769514067281, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 36088 + }, + { + "epoch": 0.36089, + "grad_norm": 1.22680389043169, + "learning_rate": 0.003, + "loss": 4.05, + "step": 36089 + }, + { + "epoch": 0.3609, + "grad_norm": 1.0064435303320565, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 36090 + }, + { + "epoch": 0.36091, + "grad_norm": 1.1598727636057211, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 36091 + }, + { + "epoch": 0.36092, + "grad_norm": 0.9436294089714403, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 36092 + }, + { + "epoch": 0.36093, + "grad_norm": 1.0823838860357657, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 36093 + }, + { + "epoch": 0.36094, + "grad_norm": 0.9874442252198946, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 36094 + }, + { + "epoch": 0.36095, + "grad_norm": 1.103591361021858, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 36095 + }, + { + "epoch": 0.36096, + "grad_norm": 0.9768168601115986, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 36096 + }, + { + "epoch": 0.36097, + "grad_norm": 1.2258515821449312, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 36097 + }, + { + "epoch": 0.36098, + "grad_norm": 1.1335741857542714, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 36098 + }, + { + "epoch": 0.36099, + "grad_norm": 1.064085300769959, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 36099 + }, + { + "epoch": 0.361, + "grad_norm": 1.063802369740798, + "learning_rate": 0.003, + "loss": 4.03, + "step": 36100 + }, + { + "epoch": 0.36101, + "grad_norm": 1.225138180581416, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 36101 + }, + { + "epoch": 0.36102, + "grad_norm": 1.1262833863599662, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 36102 + }, + { + "epoch": 0.36103, + "grad_norm": 1.0609378242089078, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 36103 + }, + { + "epoch": 0.36104, + "grad_norm": 1.052192286303146, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 36104 + }, + { + "epoch": 0.36105, + "grad_norm": 1.119725018218206, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 36105 + }, + { + "epoch": 0.36106, + "grad_norm": 1.0288739888704688, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 36106 + }, + { + "epoch": 0.36107, + "grad_norm": 1.1160845585983583, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 36107 + }, + { + "epoch": 0.36108, + "grad_norm": 1.1645727766129348, + "learning_rate": 0.003, + "loss": 4.062, + "step": 36108 + }, + { + "epoch": 0.36109, + "grad_norm": 1.0625428888451898, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 36109 + }, + { + "epoch": 0.3611, + "grad_norm": 1.0718141736757738, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 36110 + }, + { + "epoch": 0.36111, + "grad_norm": 1.2119618806312777, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 36111 + }, + { + "epoch": 0.36112, + "grad_norm": 1.0283250482946646, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 36112 + }, + { + "epoch": 0.36113, + "grad_norm": 1.0535495162756578, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 36113 + }, + { + "epoch": 0.36114, + "grad_norm": 1.086611851082065, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 36114 + }, + { + "epoch": 0.36115, + "grad_norm": 1.03003953638163, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 36115 + }, + { + "epoch": 0.36116, + "grad_norm": 1.2145610039045882, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 36116 + }, + { + "epoch": 0.36117, + "grad_norm": 1.140300137015742, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 36117 + }, + { + "epoch": 0.36118, + "grad_norm": 1.037395599328845, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 36118 + }, + { + "epoch": 0.36119, + "grad_norm": 1.1193877470465985, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 36119 + }, + { + "epoch": 0.3612, + "grad_norm": 1.0743975840357936, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 36120 + }, + { + "epoch": 0.36121, + "grad_norm": 1.1425446932020296, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 36121 + }, + { + "epoch": 0.36122, + "grad_norm": 1.1387817165764114, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 36122 + }, + { + "epoch": 0.36123, + "grad_norm": 1.216495665393574, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 36123 + }, + { + "epoch": 0.36124, + "grad_norm": 1.0403597046413224, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 36124 + }, + { + "epoch": 0.36125, + "grad_norm": 1.2598860692814597, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 36125 + }, + { + "epoch": 0.36126, + "grad_norm": 0.925879869434711, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 36126 + }, + { + "epoch": 0.36127, + "grad_norm": 0.9935457059729857, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 36127 + }, + { + "epoch": 0.36128, + "grad_norm": 1.1326569248166172, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 36128 + }, + { + "epoch": 0.36129, + "grad_norm": 1.0328463769143132, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 36129 + }, + { + "epoch": 0.3613, + "grad_norm": 1.1478846440065855, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 36130 + }, + { + "epoch": 0.36131, + "grad_norm": 1.017297071219917, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 36131 + }, + { + "epoch": 0.36132, + "grad_norm": 1.3424194117615016, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 36132 + }, + { + "epoch": 0.36133, + "grad_norm": 0.862576591090026, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 36133 + }, + { + "epoch": 0.36134, + "grad_norm": 0.8475911291710518, + "learning_rate": 0.003, + "loss": 4.027, + "step": 36134 + }, + { + "epoch": 0.36135, + "grad_norm": 1.0001132960020522, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 36135 + }, + { + "epoch": 0.36136, + "grad_norm": 1.4462165869222143, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 36136 + }, + { + "epoch": 0.36137, + "grad_norm": 1.0401822570205472, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 36137 + }, + { + "epoch": 0.36138, + "grad_norm": 1.0417017437018938, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 36138 + }, + { + "epoch": 0.36139, + "grad_norm": 1.194252208594879, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 36139 + }, + { + "epoch": 0.3614, + "grad_norm": 0.9733370675691718, + "learning_rate": 0.003, + "loss": 4.051, + "step": 36140 + }, + { + "epoch": 0.36141, + "grad_norm": 1.0893260801906754, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 36141 + }, + { + "epoch": 0.36142, + "grad_norm": 1.0950003807511592, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 36142 + }, + { + "epoch": 0.36143, + "grad_norm": 1.116410945404137, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 36143 + }, + { + "epoch": 0.36144, + "grad_norm": 0.9399934262815682, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 36144 + }, + { + "epoch": 0.36145, + "grad_norm": 1.1243752269652059, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 36145 + }, + { + "epoch": 0.36146, + "grad_norm": 1.2317357932501414, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 36146 + }, + { + "epoch": 0.36147, + "grad_norm": 1.199287485477577, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 36147 + }, + { + "epoch": 0.36148, + "grad_norm": 0.9110399599955669, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 36148 + }, + { + "epoch": 0.36149, + "grad_norm": 0.9437892007554418, + "learning_rate": 0.003, + "loss": 4.067, + "step": 36149 + }, + { + "epoch": 0.3615, + "grad_norm": 1.3403936909572773, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 36150 + }, + { + "epoch": 0.36151, + "grad_norm": 1.092202165445448, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 36151 + }, + { + "epoch": 0.36152, + "grad_norm": 0.9224298099309391, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 36152 + }, + { + "epoch": 0.36153, + "grad_norm": 0.8949769753672273, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 36153 + }, + { + "epoch": 0.36154, + "grad_norm": 0.9852320076256462, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 36154 + }, + { + "epoch": 0.36155, + "grad_norm": 1.103196385970357, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 36155 + }, + { + "epoch": 0.36156, + "grad_norm": 1.2383059328607968, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 36156 + }, + { + "epoch": 0.36157, + "grad_norm": 1.0462252849448275, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 36157 + }, + { + "epoch": 0.36158, + "grad_norm": 1.082117491226415, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 36158 + }, + { + "epoch": 0.36159, + "grad_norm": 1.1739830857443685, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 36159 + }, + { + "epoch": 0.3616, + "grad_norm": 1.0657438919854334, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 36160 + }, + { + "epoch": 0.36161, + "grad_norm": 1.0264670722364355, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 36161 + }, + { + "epoch": 0.36162, + "grad_norm": 0.9770940925057766, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 36162 + }, + { + "epoch": 0.36163, + "grad_norm": 1.0818042201037876, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 36163 + }, + { + "epoch": 0.36164, + "grad_norm": 1.1917453520291006, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 36164 + }, + { + "epoch": 0.36165, + "grad_norm": 1.1617450945515055, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 36165 + }, + { + "epoch": 0.36166, + "grad_norm": 1.0626650164536022, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 36166 + }, + { + "epoch": 0.36167, + "grad_norm": 1.179393482965644, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 36167 + }, + { + "epoch": 0.36168, + "grad_norm": 1.1774216814201952, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 36168 + }, + { + "epoch": 0.36169, + "grad_norm": 1.1030407479838424, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 36169 + }, + { + "epoch": 0.3617, + "grad_norm": 1.0684567100004145, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 36170 + }, + { + "epoch": 0.36171, + "grad_norm": 1.100466559461608, + "learning_rate": 0.003, + "loss": 4.061, + "step": 36171 + }, + { + "epoch": 0.36172, + "grad_norm": 1.0839860285042417, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 36172 + }, + { + "epoch": 0.36173, + "grad_norm": 1.1275022888917625, + "learning_rate": 0.003, + "loss": 4.019, + "step": 36173 + }, + { + "epoch": 0.36174, + "grad_norm": 1.2418984903985288, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 36174 + }, + { + "epoch": 0.36175, + "grad_norm": 1.2167149850342367, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 36175 + }, + { + "epoch": 0.36176, + "grad_norm": 1.4029320187631118, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 36176 + }, + { + "epoch": 0.36177, + "grad_norm": 0.8863573008946374, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 36177 + }, + { + "epoch": 0.36178, + "grad_norm": 1.0186274472852355, + "learning_rate": 0.003, + "loss": 4.051, + "step": 36178 + }, + { + "epoch": 0.36179, + "grad_norm": 1.33761600733048, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 36179 + }, + { + "epoch": 0.3618, + "grad_norm": 0.8865274258570781, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 36180 + }, + { + "epoch": 0.36181, + "grad_norm": 0.927838655599057, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 36181 + }, + { + "epoch": 0.36182, + "grad_norm": 1.0525915917192579, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 36182 + }, + { + "epoch": 0.36183, + "grad_norm": 1.346465248981819, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 36183 + }, + { + "epoch": 0.36184, + "grad_norm": 1.1235531350079166, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 36184 + }, + { + "epoch": 0.36185, + "grad_norm": 1.0931056475659084, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 36185 + }, + { + "epoch": 0.36186, + "grad_norm": 1.111415512458663, + "learning_rate": 0.003, + "loss": 3.997, + "step": 36186 + }, + { + "epoch": 0.36187, + "grad_norm": 0.8584534514965892, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 36187 + }, + { + "epoch": 0.36188, + "grad_norm": 0.8130085353231821, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 36188 + }, + { + "epoch": 0.36189, + "grad_norm": 0.9612543850399455, + "learning_rate": 0.003, + "loss": 4.0856, + "step": 36189 + }, + { + "epoch": 0.3619, + "grad_norm": 1.0553331107612993, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 36190 + }, + { + "epoch": 0.36191, + "grad_norm": 0.9782667240263083, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 36191 + }, + { + "epoch": 0.36192, + "grad_norm": 1.0535556274589992, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 36192 + }, + { + "epoch": 0.36193, + "grad_norm": 0.9531856813641226, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 36193 + }, + { + "epoch": 0.36194, + "grad_norm": 1.2774775059472434, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 36194 + }, + { + "epoch": 0.36195, + "grad_norm": 1.1910394170455436, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 36195 + }, + { + "epoch": 0.36196, + "grad_norm": 1.2758888870178848, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 36196 + }, + { + "epoch": 0.36197, + "grad_norm": 1.1309695315997266, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 36197 + }, + { + "epoch": 0.36198, + "grad_norm": 0.9379941430882806, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 36198 + }, + { + "epoch": 0.36199, + "grad_norm": 0.9123444493843925, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 36199 + }, + { + "epoch": 0.362, + "grad_norm": 1.0902589103926128, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 36200 + }, + { + "epoch": 0.36201, + "grad_norm": 1.2546130439407126, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 36201 + }, + { + "epoch": 0.36202, + "grad_norm": 1.1151355703849408, + "learning_rate": 0.003, + "loss": 4.028, + "step": 36202 + }, + { + "epoch": 0.36203, + "grad_norm": 1.1795235952524952, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 36203 + }, + { + "epoch": 0.36204, + "grad_norm": 0.9889830948782139, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 36204 + }, + { + "epoch": 0.36205, + "grad_norm": 1.0331217765383447, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 36205 + }, + { + "epoch": 0.36206, + "grad_norm": 1.192346675806331, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 36206 + }, + { + "epoch": 0.36207, + "grad_norm": 0.9734742711383406, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 36207 + }, + { + "epoch": 0.36208, + "grad_norm": 1.0645399907156068, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 36208 + }, + { + "epoch": 0.36209, + "grad_norm": 1.0684383802893498, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 36209 + }, + { + "epoch": 0.3621, + "grad_norm": 1.0269144884493508, + "learning_rate": 0.003, + "loss": 4.032, + "step": 36210 + }, + { + "epoch": 0.36211, + "grad_norm": 1.3260274300827106, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 36211 + }, + { + "epoch": 0.36212, + "grad_norm": 1.0660995915678415, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 36212 + }, + { + "epoch": 0.36213, + "grad_norm": 1.3130831548628665, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 36213 + }, + { + "epoch": 0.36214, + "grad_norm": 0.9837420105536792, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 36214 + }, + { + "epoch": 0.36215, + "grad_norm": 0.9900759241045025, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 36215 + }, + { + "epoch": 0.36216, + "grad_norm": 1.103400558187684, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 36216 + }, + { + "epoch": 0.36217, + "grad_norm": 1.14304713420627, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 36217 + }, + { + "epoch": 0.36218, + "grad_norm": 0.8571399316866893, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 36218 + }, + { + "epoch": 0.36219, + "grad_norm": 0.9202622183192712, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 36219 + }, + { + "epoch": 0.3622, + "grad_norm": 1.0825469509396828, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 36220 + }, + { + "epoch": 0.36221, + "grad_norm": 1.2203515985711952, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 36221 + }, + { + "epoch": 0.36222, + "grad_norm": 1.1124636796933642, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 36222 + }, + { + "epoch": 0.36223, + "grad_norm": 1.1318830776303124, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 36223 + }, + { + "epoch": 0.36224, + "grad_norm": 1.1806962294391459, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 36224 + }, + { + "epoch": 0.36225, + "grad_norm": 1.0438489461465, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 36225 + }, + { + "epoch": 0.36226, + "grad_norm": 1.1323805247903218, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 36226 + }, + { + "epoch": 0.36227, + "grad_norm": 1.099092779056212, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 36227 + }, + { + "epoch": 0.36228, + "grad_norm": 1.090743719271549, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 36228 + }, + { + "epoch": 0.36229, + "grad_norm": 0.9755069134831077, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 36229 + }, + { + "epoch": 0.3623, + "grad_norm": 1.1291414157698616, + "learning_rate": 0.003, + "loss": 4.061, + "step": 36230 + }, + { + "epoch": 0.36231, + "grad_norm": 0.9764142316536714, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 36231 + }, + { + "epoch": 0.36232, + "grad_norm": 1.221581132133646, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 36232 + }, + { + "epoch": 0.36233, + "grad_norm": 0.828794031034768, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 36233 + }, + { + "epoch": 0.36234, + "grad_norm": 0.8362883953835856, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 36234 + }, + { + "epoch": 0.36235, + "grad_norm": 0.9519780795495802, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 36235 + }, + { + "epoch": 0.36236, + "grad_norm": 1.1434222930741333, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 36236 + }, + { + "epoch": 0.36237, + "grad_norm": 1.0692773328196277, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 36237 + }, + { + "epoch": 0.36238, + "grad_norm": 1.2712563982596312, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 36238 + }, + { + "epoch": 0.36239, + "grad_norm": 0.9707978658413767, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 36239 + }, + { + "epoch": 0.3624, + "grad_norm": 1.1863467953623847, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 36240 + }, + { + "epoch": 0.36241, + "grad_norm": 1.1173791098778783, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 36241 + }, + { + "epoch": 0.36242, + "grad_norm": 1.0588957857019663, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 36242 + }, + { + "epoch": 0.36243, + "grad_norm": 1.0162545495662683, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 36243 + }, + { + "epoch": 0.36244, + "grad_norm": 1.1893642663324406, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 36244 + }, + { + "epoch": 0.36245, + "grad_norm": 1.00272048165899, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 36245 + }, + { + "epoch": 0.36246, + "grad_norm": 1.1454771386295823, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 36246 + }, + { + "epoch": 0.36247, + "grad_norm": 0.9450506349412096, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 36247 + }, + { + "epoch": 0.36248, + "grad_norm": 1.2922914269263812, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 36248 + }, + { + "epoch": 0.36249, + "grad_norm": 1.1708390018126795, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 36249 + }, + { + "epoch": 0.3625, + "grad_norm": 1.2025462149986712, + "learning_rate": 0.003, + "loss": 4.016, + "step": 36250 + }, + { + "epoch": 0.36251, + "grad_norm": 1.18223916764514, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 36251 + }, + { + "epoch": 0.36252, + "grad_norm": 1.2201303000818011, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 36252 + }, + { + "epoch": 0.36253, + "grad_norm": 1.044709046787619, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 36253 + }, + { + "epoch": 0.36254, + "grad_norm": 1.0467358514356622, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 36254 + }, + { + "epoch": 0.36255, + "grad_norm": 1.0459370246435948, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 36255 + }, + { + "epoch": 0.36256, + "grad_norm": 1.0226208068614075, + "learning_rate": 0.003, + "loss": 4.04, + "step": 36256 + }, + { + "epoch": 0.36257, + "grad_norm": 1.1168892150830492, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 36257 + }, + { + "epoch": 0.36258, + "grad_norm": 1.0413464686425693, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 36258 + }, + { + "epoch": 0.36259, + "grad_norm": 0.9759423687965816, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 36259 + }, + { + "epoch": 0.3626, + "grad_norm": 1.1653136658088652, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 36260 + }, + { + "epoch": 0.36261, + "grad_norm": 0.8864930498445526, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 36261 + }, + { + "epoch": 0.36262, + "grad_norm": 0.8558746039600269, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 36262 + }, + { + "epoch": 0.36263, + "grad_norm": 0.9865102504606796, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 36263 + }, + { + "epoch": 0.36264, + "grad_norm": 1.0833647767175871, + "learning_rate": 0.003, + "loss": 4.034, + "step": 36264 + }, + { + "epoch": 0.36265, + "grad_norm": 1.2196372721032875, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 36265 + }, + { + "epoch": 0.36266, + "grad_norm": 1.0681141611989153, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 36266 + }, + { + "epoch": 0.36267, + "grad_norm": 1.3589577818045722, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 36267 + }, + { + "epoch": 0.36268, + "grad_norm": 0.7963356507456792, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 36268 + }, + { + "epoch": 0.36269, + "grad_norm": 0.9299294219191527, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 36269 + }, + { + "epoch": 0.3627, + "grad_norm": 1.4810438096118386, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 36270 + }, + { + "epoch": 0.36271, + "grad_norm": 0.8937055913995273, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 36271 + }, + { + "epoch": 0.36272, + "grad_norm": 1.134505569844896, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 36272 + }, + { + "epoch": 0.36273, + "grad_norm": 1.3289402307313258, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 36273 + }, + { + "epoch": 0.36274, + "grad_norm": 0.9421539442095292, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 36274 + }, + { + "epoch": 0.36275, + "grad_norm": 1.0043867957337578, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 36275 + }, + { + "epoch": 0.36276, + "grad_norm": 1.2144368527482308, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 36276 + }, + { + "epoch": 0.36277, + "grad_norm": 0.9320844571705629, + "learning_rate": 0.003, + "loss": 4.042, + "step": 36277 + }, + { + "epoch": 0.36278, + "grad_norm": 0.9385490539911125, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 36278 + }, + { + "epoch": 0.36279, + "grad_norm": 1.0781957108722002, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 36279 + }, + { + "epoch": 0.3628, + "grad_norm": 1.1218389943267006, + "learning_rate": 0.003, + "loss": 4.07, + "step": 36280 + }, + { + "epoch": 0.36281, + "grad_norm": 1.00501548909617, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 36281 + }, + { + "epoch": 0.36282, + "grad_norm": 1.0319862771717208, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 36282 + }, + { + "epoch": 0.36283, + "grad_norm": 1.109722556227984, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 36283 + }, + { + "epoch": 0.36284, + "grad_norm": 1.2731881208166465, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 36284 + }, + { + "epoch": 0.36285, + "grad_norm": 0.973776702537539, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 36285 + }, + { + "epoch": 0.36286, + "grad_norm": 1.302381701007981, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 36286 + }, + { + "epoch": 0.36287, + "grad_norm": 1.1029401550922908, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 36287 + }, + { + "epoch": 0.36288, + "grad_norm": 1.1685043693795607, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 36288 + }, + { + "epoch": 0.36289, + "grad_norm": 0.9332226402978115, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 36289 + }, + { + "epoch": 0.3629, + "grad_norm": 1.1540638140707542, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 36290 + }, + { + "epoch": 0.36291, + "grad_norm": 0.9579914931734302, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 36291 + }, + { + "epoch": 0.36292, + "grad_norm": 1.0178708354106674, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 36292 + }, + { + "epoch": 0.36293, + "grad_norm": 0.9651234561095976, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 36293 + }, + { + "epoch": 0.36294, + "grad_norm": 0.9538090971346304, + "learning_rate": 0.003, + "loss": 4.025, + "step": 36294 + }, + { + "epoch": 0.36295, + "grad_norm": 0.86844258006929, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 36295 + }, + { + "epoch": 0.36296, + "grad_norm": 1.0286191263913012, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 36296 + }, + { + "epoch": 0.36297, + "grad_norm": 1.6234568880139661, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 36297 + }, + { + "epoch": 0.36298, + "grad_norm": 1.0289223045473055, + "learning_rate": 0.003, + "loss": 4.005, + "step": 36298 + }, + { + "epoch": 0.36299, + "grad_norm": 1.2682430266018316, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 36299 + }, + { + "epoch": 0.363, + "grad_norm": 0.930992969705686, + "learning_rate": 0.003, + "loss": 4.044, + "step": 36300 + }, + { + "epoch": 0.36301, + "grad_norm": 0.9057861539163327, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 36301 + }, + { + "epoch": 0.36302, + "grad_norm": 1.0647139024701522, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 36302 + }, + { + "epoch": 0.36303, + "grad_norm": 1.3115970162929071, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 36303 + }, + { + "epoch": 0.36304, + "grad_norm": 1.0768261923381943, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 36304 + }, + { + "epoch": 0.36305, + "grad_norm": 1.1713846956618046, + "learning_rate": 0.003, + "loss": 4.022, + "step": 36305 + }, + { + "epoch": 0.36306, + "grad_norm": 1.1034492668535942, + "learning_rate": 0.003, + "loss": 4.05, + "step": 36306 + }, + { + "epoch": 0.36307, + "grad_norm": 1.011502744847271, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 36307 + }, + { + "epoch": 0.36308, + "grad_norm": 1.1630652476375225, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 36308 + }, + { + "epoch": 0.36309, + "grad_norm": 1.0220981435049308, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 36309 + }, + { + "epoch": 0.3631, + "grad_norm": 0.9758942563086818, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 36310 + }, + { + "epoch": 0.36311, + "grad_norm": 0.966914150993004, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 36311 + }, + { + "epoch": 0.36312, + "grad_norm": 1.003853715135205, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 36312 + }, + { + "epoch": 0.36313, + "grad_norm": 1.1910253528022772, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 36313 + }, + { + "epoch": 0.36314, + "grad_norm": 0.9357848931069456, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 36314 + }, + { + "epoch": 0.36315, + "grad_norm": 1.109407874254288, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 36315 + }, + { + "epoch": 0.36316, + "grad_norm": 1.2257473832796075, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 36316 + }, + { + "epoch": 0.36317, + "grad_norm": 1.0180114193710834, + "learning_rate": 0.003, + "loss": 4.03, + "step": 36317 + }, + { + "epoch": 0.36318, + "grad_norm": 1.1247529634852662, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 36318 + }, + { + "epoch": 0.36319, + "grad_norm": 1.0827748999434916, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 36319 + }, + { + "epoch": 0.3632, + "grad_norm": 1.2723887901415993, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 36320 + }, + { + "epoch": 0.36321, + "grad_norm": 0.9492558932439303, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 36321 + }, + { + "epoch": 0.36322, + "grad_norm": 1.0684748294074213, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 36322 + }, + { + "epoch": 0.36323, + "grad_norm": 1.1670312439825463, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 36323 + }, + { + "epoch": 0.36324, + "grad_norm": 1.129426501401701, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 36324 + }, + { + "epoch": 0.36325, + "grad_norm": 1.1926154987366369, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 36325 + }, + { + "epoch": 0.36326, + "grad_norm": 0.975710794956171, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 36326 + }, + { + "epoch": 0.36327, + "grad_norm": 1.0545931720008015, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 36327 + }, + { + "epoch": 0.36328, + "grad_norm": 1.1286974921878328, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 36328 + }, + { + "epoch": 0.36329, + "grad_norm": 1.0521679024383381, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 36329 + }, + { + "epoch": 0.3633, + "grad_norm": 1.0878042636053704, + "learning_rate": 0.003, + "loss": 4.005, + "step": 36330 + }, + { + "epoch": 0.36331, + "grad_norm": 1.0010045238585077, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 36331 + }, + { + "epoch": 0.36332, + "grad_norm": 1.0570195701047376, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 36332 + }, + { + "epoch": 0.36333, + "grad_norm": 0.9522824105908478, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 36333 + }, + { + "epoch": 0.36334, + "grad_norm": 1.149062214273742, + "learning_rate": 0.003, + "loss": 4.078, + "step": 36334 + }, + { + "epoch": 0.36335, + "grad_norm": 1.1407554696068594, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 36335 + }, + { + "epoch": 0.36336, + "grad_norm": 1.3293202467407808, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 36336 + }, + { + "epoch": 0.36337, + "grad_norm": 0.9383284460969379, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 36337 + }, + { + "epoch": 0.36338, + "grad_norm": 1.1515542856678898, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 36338 + }, + { + "epoch": 0.36339, + "grad_norm": 1.2235014719304962, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 36339 + }, + { + "epoch": 0.3634, + "grad_norm": 0.9219959439065155, + "learning_rate": 0.003, + "loss": 4.033, + "step": 36340 + }, + { + "epoch": 0.36341, + "grad_norm": 0.9092181414969958, + "learning_rate": 0.003, + "loss": 4.031, + "step": 36341 + }, + { + "epoch": 0.36342, + "grad_norm": 0.9870663257056472, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 36342 + }, + { + "epoch": 0.36343, + "grad_norm": 1.326975082369914, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 36343 + }, + { + "epoch": 0.36344, + "grad_norm": 0.9907974539566656, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 36344 + }, + { + "epoch": 0.36345, + "grad_norm": 1.0511487443505798, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 36345 + }, + { + "epoch": 0.36346, + "grad_norm": 1.0024786093916336, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 36346 + }, + { + "epoch": 0.36347, + "grad_norm": 1.1545853575130114, + "learning_rate": 0.003, + "loss": 4.082, + "step": 36347 + }, + { + "epoch": 0.36348, + "grad_norm": 0.9827852808035185, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 36348 + }, + { + "epoch": 0.36349, + "grad_norm": 1.0470314867176909, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 36349 + }, + { + "epoch": 0.3635, + "grad_norm": 1.1326051475208114, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 36350 + }, + { + "epoch": 0.36351, + "grad_norm": 1.0133532075900602, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 36351 + }, + { + "epoch": 0.36352, + "grad_norm": 1.0733263946626843, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 36352 + }, + { + "epoch": 0.36353, + "grad_norm": 1.1650876031908635, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 36353 + }, + { + "epoch": 0.36354, + "grad_norm": 1.1171929502904236, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 36354 + }, + { + "epoch": 0.36355, + "grad_norm": 1.2988638643279278, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 36355 + }, + { + "epoch": 0.36356, + "grad_norm": 1.2317266035413947, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 36356 + }, + { + "epoch": 0.36357, + "grad_norm": 1.0298081350169612, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 36357 + }, + { + "epoch": 0.36358, + "grad_norm": 1.11456939812675, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 36358 + }, + { + "epoch": 0.36359, + "grad_norm": 1.1702771308326934, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 36359 + }, + { + "epoch": 0.3636, + "grad_norm": 1.1666678834147501, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 36360 + }, + { + "epoch": 0.36361, + "grad_norm": 0.9316656140999264, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 36361 + }, + { + "epoch": 0.36362, + "grad_norm": 1.0278167067512884, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 36362 + }, + { + "epoch": 0.36363, + "grad_norm": 1.3168526070745055, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 36363 + }, + { + "epoch": 0.36364, + "grad_norm": 0.8584685009672232, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 36364 + }, + { + "epoch": 0.36365, + "grad_norm": 0.9231118568441412, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 36365 + }, + { + "epoch": 0.36366, + "grad_norm": 1.082131649388544, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 36366 + }, + { + "epoch": 0.36367, + "grad_norm": 1.1837749371417832, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 36367 + }, + { + "epoch": 0.36368, + "grad_norm": 0.917076726394652, + "learning_rate": 0.003, + "loss": 4.041, + "step": 36368 + }, + { + "epoch": 0.36369, + "grad_norm": 0.8998654029408915, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 36369 + }, + { + "epoch": 0.3637, + "grad_norm": 0.9410460642144105, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 36370 + }, + { + "epoch": 0.36371, + "grad_norm": 1.0516794992014151, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 36371 + }, + { + "epoch": 0.36372, + "grad_norm": 1.1784616548445417, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 36372 + }, + { + "epoch": 0.36373, + "grad_norm": 0.9678282938676069, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 36373 + }, + { + "epoch": 0.36374, + "grad_norm": 1.0499447867923353, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 36374 + }, + { + "epoch": 0.36375, + "grad_norm": 1.0906208932245507, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 36375 + }, + { + "epoch": 0.36376, + "grad_norm": 1.216109460654913, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 36376 + }, + { + "epoch": 0.36377, + "grad_norm": 0.8664479909167958, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 36377 + }, + { + "epoch": 0.36378, + "grad_norm": 1.1230163749251727, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 36378 + }, + { + "epoch": 0.36379, + "grad_norm": 1.2673632959742513, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 36379 + }, + { + "epoch": 0.3638, + "grad_norm": 0.9143540506585363, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 36380 + }, + { + "epoch": 0.36381, + "grad_norm": 0.938817446833839, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 36381 + }, + { + "epoch": 0.36382, + "grad_norm": 1.2311821069423314, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 36382 + }, + { + "epoch": 0.36383, + "grad_norm": 1.3837755416970263, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 36383 + }, + { + "epoch": 0.36384, + "grad_norm": 0.8131665517213336, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 36384 + }, + { + "epoch": 0.36385, + "grad_norm": 0.8915105138910483, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 36385 + }, + { + "epoch": 0.36386, + "grad_norm": 1.1298485860983292, + "learning_rate": 0.003, + "loss": 4.027, + "step": 36386 + }, + { + "epoch": 0.36387, + "grad_norm": 1.4595340027772563, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 36387 + }, + { + "epoch": 0.36388, + "grad_norm": 0.8490265922711295, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 36388 + }, + { + "epoch": 0.36389, + "grad_norm": 1.0007601508817314, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 36389 + }, + { + "epoch": 0.3639, + "grad_norm": 1.66181590280009, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 36390 + }, + { + "epoch": 0.36391, + "grad_norm": 0.9361147203995758, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 36391 + }, + { + "epoch": 0.36392, + "grad_norm": 1.0484790203876408, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 36392 + }, + { + "epoch": 0.36393, + "grad_norm": 1.1770145406489094, + "learning_rate": 0.003, + "loss": 4.049, + "step": 36393 + }, + { + "epoch": 0.36394, + "grad_norm": 1.0161384376594482, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 36394 + }, + { + "epoch": 0.36395, + "grad_norm": 1.135692814137398, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 36395 + }, + { + "epoch": 0.36396, + "grad_norm": 1.0692753554073668, + "learning_rate": 0.003, + "loss": 4.047, + "step": 36396 + }, + { + "epoch": 0.36397, + "grad_norm": 1.2733260508301687, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 36397 + }, + { + "epoch": 0.36398, + "grad_norm": 0.919341872471184, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 36398 + }, + { + "epoch": 0.36399, + "grad_norm": 0.8622015575757598, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 36399 + }, + { + "epoch": 0.364, + "grad_norm": 0.9739952893560696, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 36400 + }, + { + "epoch": 0.36401, + "grad_norm": 1.2024431298757867, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 36401 + }, + { + "epoch": 0.36402, + "grad_norm": 0.8174585819379212, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 36402 + }, + { + "epoch": 0.36403, + "grad_norm": 0.8636389689892126, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 36403 + }, + { + "epoch": 0.36404, + "grad_norm": 1.109246740875807, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 36404 + }, + { + "epoch": 0.36405, + "grad_norm": 1.2099569569875996, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 36405 + }, + { + "epoch": 0.36406, + "grad_norm": 1.1401396603123977, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 36406 + }, + { + "epoch": 0.36407, + "grad_norm": 1.0840063951977876, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 36407 + }, + { + "epoch": 0.36408, + "grad_norm": 1.0968101277698536, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 36408 + }, + { + "epoch": 0.36409, + "grad_norm": 1.0176265987114599, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 36409 + }, + { + "epoch": 0.3641, + "grad_norm": 1.142605823656213, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 36410 + }, + { + "epoch": 0.36411, + "grad_norm": 1.0144740630766154, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 36411 + }, + { + "epoch": 0.36412, + "grad_norm": 1.2928519701989514, + "learning_rate": 0.003, + "loss": 4.061, + "step": 36412 + }, + { + "epoch": 0.36413, + "grad_norm": 1.0062088194264627, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 36413 + }, + { + "epoch": 0.36414, + "grad_norm": 1.3228115470042197, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 36414 + }, + { + "epoch": 0.36415, + "grad_norm": 0.9230966697375987, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 36415 + }, + { + "epoch": 0.36416, + "grad_norm": 1.1126359249862559, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 36416 + }, + { + "epoch": 0.36417, + "grad_norm": 1.3480164936171393, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 36417 + }, + { + "epoch": 0.36418, + "grad_norm": 1.088522512465166, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 36418 + }, + { + "epoch": 0.36419, + "grad_norm": 1.2093766857131716, + "learning_rate": 0.003, + "loss": 4.052, + "step": 36419 + }, + { + "epoch": 0.3642, + "grad_norm": 1.1448505867684595, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 36420 + }, + { + "epoch": 0.36421, + "grad_norm": 0.8849106476982478, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 36421 + }, + { + "epoch": 0.36422, + "grad_norm": 1.0157078365141385, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 36422 + }, + { + "epoch": 0.36423, + "grad_norm": 1.4332151930776662, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 36423 + }, + { + "epoch": 0.36424, + "grad_norm": 1.0214380462057073, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 36424 + }, + { + "epoch": 0.36425, + "grad_norm": 1.013054281189697, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 36425 + }, + { + "epoch": 0.36426, + "grad_norm": 1.1527628637232563, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 36426 + }, + { + "epoch": 0.36427, + "grad_norm": 0.8999303609421675, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 36427 + }, + { + "epoch": 0.36428, + "grad_norm": 1.1591278967145016, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 36428 + }, + { + "epoch": 0.36429, + "grad_norm": 0.9729556815157927, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 36429 + }, + { + "epoch": 0.3643, + "grad_norm": 1.0216332063532876, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 36430 + }, + { + "epoch": 0.36431, + "grad_norm": 1.1132344942318235, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 36431 + }, + { + "epoch": 0.36432, + "grad_norm": 0.9998559970782971, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 36432 + }, + { + "epoch": 0.36433, + "grad_norm": 1.2298241390848872, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 36433 + }, + { + "epoch": 0.36434, + "grad_norm": 0.8904644918861668, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 36434 + }, + { + "epoch": 0.36435, + "grad_norm": 0.7985827480808176, + "learning_rate": 0.003, + "loss": 4.028, + "step": 36435 + }, + { + "epoch": 0.36436, + "grad_norm": 0.9814270324898015, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 36436 + }, + { + "epoch": 0.36437, + "grad_norm": 1.3505613746556024, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 36437 + }, + { + "epoch": 0.36438, + "grad_norm": 0.7445790201499954, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 36438 + }, + { + "epoch": 0.36439, + "grad_norm": 0.7717814416348626, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 36439 + }, + { + "epoch": 0.3644, + "grad_norm": 0.8043118142493898, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 36440 + }, + { + "epoch": 0.36441, + "grad_norm": 0.9757464830921303, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 36441 + }, + { + "epoch": 0.36442, + "grad_norm": 1.1903282855380868, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 36442 + }, + { + "epoch": 0.36443, + "grad_norm": 1.3661649291113465, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 36443 + }, + { + "epoch": 0.36444, + "grad_norm": 0.8657616167035024, + "learning_rate": 0.003, + "loss": 4.014, + "step": 36444 + }, + { + "epoch": 0.36445, + "grad_norm": 0.848312500551311, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 36445 + }, + { + "epoch": 0.36446, + "grad_norm": 0.9912049183537102, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 36446 + }, + { + "epoch": 0.36447, + "grad_norm": 1.2218367798639493, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 36447 + }, + { + "epoch": 0.36448, + "grad_norm": 0.9976614231794451, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 36448 + }, + { + "epoch": 0.36449, + "grad_norm": 1.4018066600879078, + "learning_rate": 0.003, + "loss": 4.045, + "step": 36449 + }, + { + "epoch": 0.3645, + "grad_norm": 0.8651462602098132, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 36450 + }, + { + "epoch": 0.36451, + "grad_norm": 1.056102937659938, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 36451 + }, + { + "epoch": 0.36452, + "grad_norm": 1.4722202569809117, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 36452 + }, + { + "epoch": 0.36453, + "grad_norm": 0.9550861496688409, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 36453 + }, + { + "epoch": 0.36454, + "grad_norm": 1.134852040494466, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 36454 + }, + { + "epoch": 0.36455, + "grad_norm": 1.1873464243319682, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 36455 + }, + { + "epoch": 0.36456, + "grad_norm": 1.1016351511984137, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 36456 + }, + { + "epoch": 0.36457, + "grad_norm": 1.3654624427465405, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 36457 + }, + { + "epoch": 0.36458, + "grad_norm": 1.0276275004782742, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 36458 + }, + { + "epoch": 0.36459, + "grad_norm": 0.9715242012272751, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 36459 + }, + { + "epoch": 0.3646, + "grad_norm": 1.0519204085706084, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 36460 + }, + { + "epoch": 0.36461, + "grad_norm": 1.0276417769367603, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 36461 + }, + { + "epoch": 0.36462, + "grad_norm": 1.2280466504664294, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 36462 + }, + { + "epoch": 0.36463, + "grad_norm": 0.9646779225409777, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 36463 + }, + { + "epoch": 0.36464, + "grad_norm": 0.9067914913893769, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 36464 + }, + { + "epoch": 0.36465, + "grad_norm": 1.0243941383942041, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 36465 + }, + { + "epoch": 0.36466, + "grad_norm": 1.2795671708280962, + "learning_rate": 0.003, + "loss": 4.042, + "step": 36466 + }, + { + "epoch": 0.36467, + "grad_norm": 0.957911682883715, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 36467 + }, + { + "epoch": 0.36468, + "grad_norm": 1.1127150565568893, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 36468 + }, + { + "epoch": 0.36469, + "grad_norm": 1.0667446716048656, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 36469 + }, + { + "epoch": 0.3647, + "grad_norm": 0.9306772544678422, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 36470 + }, + { + "epoch": 0.36471, + "grad_norm": 0.8879211988814945, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 36471 + }, + { + "epoch": 0.36472, + "grad_norm": 1.0337075995588154, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 36472 + }, + { + "epoch": 0.36473, + "grad_norm": 1.094588150294246, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 36473 + }, + { + "epoch": 0.36474, + "grad_norm": 1.3088674943178409, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 36474 + }, + { + "epoch": 0.36475, + "grad_norm": 1.015532074964339, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 36475 + }, + { + "epoch": 0.36476, + "grad_norm": 1.3564515226091336, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 36476 + }, + { + "epoch": 0.36477, + "grad_norm": 0.9842167466924912, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 36477 + }, + { + "epoch": 0.36478, + "grad_norm": 1.0800328097197887, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 36478 + }, + { + "epoch": 0.36479, + "grad_norm": 1.1112811532901923, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 36479 + }, + { + "epoch": 0.3648, + "grad_norm": 0.9493625923304377, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 36480 + }, + { + "epoch": 0.36481, + "grad_norm": 1.2071113286294852, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 36481 + }, + { + "epoch": 0.36482, + "grad_norm": 0.9845189018418498, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 36482 + }, + { + "epoch": 0.36483, + "grad_norm": 1.212681268361876, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 36483 + }, + { + "epoch": 0.36484, + "grad_norm": 0.979869448687421, + "learning_rate": 0.003, + "loss": 4.04, + "step": 36484 + }, + { + "epoch": 0.36485, + "grad_norm": 1.3342405180978325, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 36485 + }, + { + "epoch": 0.36486, + "grad_norm": 1.0116804096569163, + "learning_rate": 0.003, + "loss": 4.022, + "step": 36486 + }, + { + "epoch": 0.36487, + "grad_norm": 1.1592341176923335, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 36487 + }, + { + "epoch": 0.36488, + "grad_norm": 0.9996537631692719, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 36488 + }, + { + "epoch": 0.36489, + "grad_norm": 1.1667930774789426, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 36489 + }, + { + "epoch": 0.3649, + "grad_norm": 0.9650615099868576, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 36490 + }, + { + "epoch": 0.36491, + "grad_norm": 1.2850557868655523, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 36491 + }, + { + "epoch": 0.36492, + "grad_norm": 0.9445744440422246, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 36492 + }, + { + "epoch": 0.36493, + "grad_norm": 1.1967132601118753, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 36493 + }, + { + "epoch": 0.36494, + "grad_norm": 1.0868700240185245, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 36494 + }, + { + "epoch": 0.36495, + "grad_norm": 1.0003160241064486, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 36495 + }, + { + "epoch": 0.36496, + "grad_norm": 1.2347098302204231, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 36496 + }, + { + "epoch": 0.36497, + "grad_norm": 0.8862851284375163, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 36497 + }, + { + "epoch": 0.36498, + "grad_norm": 0.8934106292465013, + "learning_rate": 0.003, + "loss": 4.054, + "step": 36498 + }, + { + "epoch": 0.36499, + "grad_norm": 1.114893520855251, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 36499 + }, + { + "epoch": 0.365, + "grad_norm": 1.1345307672363065, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 36500 + }, + { + "epoch": 0.36501, + "grad_norm": 1.0209341145035111, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 36501 + }, + { + "epoch": 0.36502, + "grad_norm": 1.068614670317011, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 36502 + }, + { + "epoch": 0.36503, + "grad_norm": 1.0940227405645455, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 36503 + }, + { + "epoch": 0.36504, + "grad_norm": 1.1672224151501105, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 36504 + }, + { + "epoch": 0.36505, + "grad_norm": 1.0887029537593427, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 36505 + }, + { + "epoch": 0.36506, + "grad_norm": 1.0019180793265716, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 36506 + }, + { + "epoch": 0.36507, + "grad_norm": 0.9620212785957406, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 36507 + }, + { + "epoch": 0.36508, + "grad_norm": 1.0020832116024871, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 36508 + }, + { + "epoch": 0.36509, + "grad_norm": 1.0749341773481929, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 36509 + }, + { + "epoch": 0.3651, + "grad_norm": 0.8939576237467676, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 36510 + }, + { + "epoch": 0.36511, + "grad_norm": 1.1575267949090464, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 36511 + }, + { + "epoch": 0.36512, + "grad_norm": 1.127670856813205, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 36512 + }, + { + "epoch": 0.36513, + "grad_norm": 1.1479186943026467, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 36513 + }, + { + "epoch": 0.36514, + "grad_norm": 1.1224461469448899, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 36514 + }, + { + "epoch": 0.36515, + "grad_norm": 1.041486016924068, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 36515 + }, + { + "epoch": 0.36516, + "grad_norm": 1.088335232286472, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 36516 + }, + { + "epoch": 0.36517, + "grad_norm": 1.1522551588491652, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 36517 + }, + { + "epoch": 0.36518, + "grad_norm": 0.934278683569707, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 36518 + }, + { + "epoch": 0.36519, + "grad_norm": 1.059612468113911, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 36519 + }, + { + "epoch": 0.3652, + "grad_norm": 1.0156713423871964, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 36520 + }, + { + "epoch": 0.36521, + "grad_norm": 1.2351778313820572, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 36521 + }, + { + "epoch": 0.36522, + "grad_norm": 1.0913565626080373, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 36522 + }, + { + "epoch": 0.36523, + "grad_norm": 1.048678034044099, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 36523 + }, + { + "epoch": 0.36524, + "grad_norm": 1.0124891842673138, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 36524 + }, + { + "epoch": 0.36525, + "grad_norm": 1.1139961961065257, + "learning_rate": 0.003, + "loss": 4.044, + "step": 36525 + }, + { + "epoch": 0.36526, + "grad_norm": 1.0002754300850272, + "learning_rate": 0.003, + "loss": 4.044, + "step": 36526 + }, + { + "epoch": 0.36527, + "grad_norm": 1.1661603520333528, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 36527 + }, + { + "epoch": 0.36528, + "grad_norm": 0.9300971080253277, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 36528 + }, + { + "epoch": 0.36529, + "grad_norm": 1.0367445057408555, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 36529 + }, + { + "epoch": 0.3653, + "grad_norm": 1.4232184955887546, + "learning_rate": 0.003, + "loss": 4.026, + "step": 36530 + }, + { + "epoch": 0.36531, + "grad_norm": 0.9647940274048792, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 36531 + }, + { + "epoch": 0.36532, + "grad_norm": 1.1210382013953621, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 36532 + }, + { + "epoch": 0.36533, + "grad_norm": 1.110515570821808, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 36533 + }, + { + "epoch": 0.36534, + "grad_norm": 0.9248366967960251, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 36534 + }, + { + "epoch": 0.36535, + "grad_norm": 0.97526608240351, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 36535 + }, + { + "epoch": 0.36536, + "grad_norm": 1.1777975243821555, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 36536 + }, + { + "epoch": 0.36537, + "grad_norm": 1.1391929322261147, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 36537 + }, + { + "epoch": 0.36538, + "grad_norm": 1.2415605579051054, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 36538 + }, + { + "epoch": 0.36539, + "grad_norm": 1.1694833075639297, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 36539 + }, + { + "epoch": 0.3654, + "grad_norm": 1.2570341998067192, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 36540 + }, + { + "epoch": 0.36541, + "grad_norm": 0.8706940961091333, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 36541 + }, + { + "epoch": 0.36542, + "grad_norm": 0.7359177809328198, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 36542 + }, + { + "epoch": 0.36543, + "grad_norm": 0.8365411111404649, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 36543 + }, + { + "epoch": 0.36544, + "grad_norm": 1.1310432314856484, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 36544 + }, + { + "epoch": 0.36545, + "grad_norm": 1.1527964308960807, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 36545 + }, + { + "epoch": 0.36546, + "grad_norm": 1.0971378726196912, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 36546 + }, + { + "epoch": 0.36547, + "grad_norm": 1.2576412217488429, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 36547 + }, + { + "epoch": 0.36548, + "grad_norm": 0.8538736220113133, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 36548 + }, + { + "epoch": 0.36549, + "grad_norm": 0.7780610773738508, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 36549 + }, + { + "epoch": 0.3655, + "grad_norm": 0.7541409344521617, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 36550 + }, + { + "epoch": 0.36551, + "grad_norm": 0.7905016996563268, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 36551 + }, + { + "epoch": 0.36552, + "grad_norm": 0.9259778622989399, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 36552 + }, + { + "epoch": 0.36553, + "grad_norm": 1.10431978267401, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 36553 + }, + { + "epoch": 0.36554, + "grad_norm": 1.0466810998634934, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 36554 + }, + { + "epoch": 0.36555, + "grad_norm": 1.1194352895178228, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 36555 + }, + { + "epoch": 0.36556, + "grad_norm": 1.016691161367263, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 36556 + }, + { + "epoch": 0.36557, + "grad_norm": 0.9785283995586823, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 36557 + }, + { + "epoch": 0.36558, + "grad_norm": 1.0106204913824717, + "learning_rate": 0.003, + "loss": 4.065, + "step": 36558 + }, + { + "epoch": 0.36559, + "grad_norm": 1.278577330727775, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 36559 + }, + { + "epoch": 0.3656, + "grad_norm": 1.0551323206010432, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 36560 + }, + { + "epoch": 0.36561, + "grad_norm": 1.5655459515195322, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 36561 + }, + { + "epoch": 0.36562, + "grad_norm": 0.8547248087827107, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 36562 + }, + { + "epoch": 0.36563, + "grad_norm": 0.8361426407745726, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 36563 + }, + { + "epoch": 0.36564, + "grad_norm": 1.0021803902365642, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 36564 + }, + { + "epoch": 0.36565, + "grad_norm": 1.492680343585003, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 36565 + }, + { + "epoch": 0.36566, + "grad_norm": 0.9829652584711334, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 36566 + }, + { + "epoch": 0.36567, + "grad_norm": 1.2106143313411715, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 36567 + }, + { + "epoch": 0.36568, + "grad_norm": 0.9400462458214952, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 36568 + }, + { + "epoch": 0.36569, + "grad_norm": 1.0378720008182307, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 36569 + }, + { + "epoch": 0.3657, + "grad_norm": 1.1081142530885117, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 36570 + }, + { + "epoch": 0.36571, + "grad_norm": 1.0940215524002594, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 36571 + }, + { + "epoch": 0.36572, + "grad_norm": 1.3556815775144888, + "learning_rate": 0.003, + "loss": 4.056, + "step": 36572 + }, + { + "epoch": 0.36573, + "grad_norm": 1.0415765852940193, + "learning_rate": 0.003, + "loss": 4.033, + "step": 36573 + }, + { + "epoch": 0.36574, + "grad_norm": 1.3592048605581075, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 36574 + }, + { + "epoch": 0.36575, + "grad_norm": 1.0251390283730206, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 36575 + }, + { + "epoch": 0.36576, + "grad_norm": 0.9478019822068926, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 36576 + }, + { + "epoch": 0.36577, + "grad_norm": 1.0663768423541131, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 36577 + }, + { + "epoch": 0.36578, + "grad_norm": 1.2031136144809542, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 36578 + }, + { + "epoch": 0.36579, + "grad_norm": 0.913935716516725, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 36579 + }, + { + "epoch": 0.3658, + "grad_norm": 1.1038810607408482, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 36580 + }, + { + "epoch": 0.36581, + "grad_norm": 1.4967036786917622, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 36581 + }, + { + "epoch": 0.36582, + "grad_norm": 0.8440387724936942, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 36582 + }, + { + "epoch": 0.36583, + "grad_norm": 0.7548228404484802, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 36583 + }, + { + "epoch": 0.36584, + "grad_norm": 1.0035353419837196, + "learning_rate": 0.003, + "loss": 4.027, + "step": 36584 + }, + { + "epoch": 0.36585, + "grad_norm": 0.9991679147728755, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 36585 + }, + { + "epoch": 0.36586, + "grad_norm": 1.00554504883429, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 36586 + }, + { + "epoch": 0.36587, + "grad_norm": 0.9655078172788837, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 36587 + }, + { + "epoch": 0.36588, + "grad_norm": 1.133863882149194, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 36588 + }, + { + "epoch": 0.36589, + "grad_norm": 1.149724835837993, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 36589 + }, + { + "epoch": 0.3659, + "grad_norm": 0.9606386086394575, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 36590 + }, + { + "epoch": 0.36591, + "grad_norm": 1.3252001583860247, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 36591 + }, + { + "epoch": 0.36592, + "grad_norm": 1.1500998517865035, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 36592 + }, + { + "epoch": 0.36593, + "grad_norm": 1.0003688869742189, + "learning_rate": 0.003, + "loss": 4.062, + "step": 36593 + }, + { + "epoch": 0.36594, + "grad_norm": 1.149606418692574, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 36594 + }, + { + "epoch": 0.36595, + "grad_norm": 1.1397935312851266, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 36595 + }, + { + "epoch": 0.36596, + "grad_norm": 1.0898526617887283, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 36596 + }, + { + "epoch": 0.36597, + "grad_norm": 1.1456652066606632, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 36597 + }, + { + "epoch": 0.36598, + "grad_norm": 1.0392036569200744, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 36598 + }, + { + "epoch": 0.36599, + "grad_norm": 1.1618483546017893, + "learning_rate": 0.003, + "loss": 4.031, + "step": 36599 + }, + { + "epoch": 0.366, + "grad_norm": 0.9563680187396261, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 36600 + }, + { + "epoch": 0.36601, + "grad_norm": 1.2997075491462724, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 36601 + }, + { + "epoch": 0.36602, + "grad_norm": 1.0706219253039448, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 36602 + }, + { + "epoch": 0.36603, + "grad_norm": 1.3157791882194338, + "learning_rate": 0.003, + "loss": 4.031, + "step": 36603 + }, + { + "epoch": 0.36604, + "grad_norm": 0.7552526235486002, + "learning_rate": 0.003, + "loss": 4.045, + "step": 36604 + }, + { + "epoch": 0.36605, + "grad_norm": 0.7526296172155325, + "learning_rate": 0.003, + "loss": 4.052, + "step": 36605 + }, + { + "epoch": 0.36606, + "grad_norm": 0.9187914284455685, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 36606 + }, + { + "epoch": 0.36607, + "grad_norm": 1.0606630879596537, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 36607 + }, + { + "epoch": 0.36608, + "grad_norm": 1.2500694183201446, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 36608 + }, + { + "epoch": 0.36609, + "grad_norm": 0.8456743347710562, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 36609 + }, + { + "epoch": 0.3661, + "grad_norm": 0.8633738536475027, + "learning_rate": 0.003, + "loss": 4.028, + "step": 36610 + }, + { + "epoch": 0.36611, + "grad_norm": 0.915890152820535, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 36611 + }, + { + "epoch": 0.36612, + "grad_norm": 1.193808533118332, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 36612 + }, + { + "epoch": 0.36613, + "grad_norm": 0.9429302923899262, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 36613 + }, + { + "epoch": 0.36614, + "grad_norm": 1.31236455216912, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 36614 + }, + { + "epoch": 0.36615, + "grad_norm": 0.9384052244818855, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 36615 + }, + { + "epoch": 0.36616, + "grad_norm": 0.9244113302075305, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 36616 + }, + { + "epoch": 0.36617, + "grad_norm": 1.038985627580323, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 36617 + }, + { + "epoch": 0.36618, + "grad_norm": 0.987008602227122, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 36618 + }, + { + "epoch": 0.36619, + "grad_norm": 1.265851296741423, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 36619 + }, + { + "epoch": 0.3662, + "grad_norm": 1.457054670557986, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 36620 + }, + { + "epoch": 0.36621, + "grad_norm": 0.8815046423692197, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 36621 + }, + { + "epoch": 0.36622, + "grad_norm": 1.1433893829533797, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 36622 + }, + { + "epoch": 0.36623, + "grad_norm": 1.564871504169373, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 36623 + }, + { + "epoch": 0.36624, + "grad_norm": 0.924794282069868, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 36624 + }, + { + "epoch": 0.36625, + "grad_norm": 1.0268358717519712, + "learning_rate": 0.003, + "loss": 4.025, + "step": 36625 + }, + { + "epoch": 0.36626, + "grad_norm": 1.285328325679809, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 36626 + }, + { + "epoch": 0.36627, + "grad_norm": 1.077703075459853, + "learning_rate": 0.003, + "loss": 4.041, + "step": 36627 + }, + { + "epoch": 0.36628, + "grad_norm": 1.3527653477255717, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 36628 + }, + { + "epoch": 0.36629, + "grad_norm": 1.0645398042533567, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 36629 + }, + { + "epoch": 0.3663, + "grad_norm": 1.050037383230738, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 36630 + }, + { + "epoch": 0.36631, + "grad_norm": 1.1780221635685166, + "learning_rate": 0.003, + "loss": 4.032, + "step": 36631 + }, + { + "epoch": 0.36632, + "grad_norm": 0.8713301996736001, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 36632 + }, + { + "epoch": 0.36633, + "grad_norm": 0.8166478870477796, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 36633 + }, + { + "epoch": 0.36634, + "grad_norm": 0.9101064839072048, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 36634 + }, + { + "epoch": 0.36635, + "grad_norm": 1.259455641188934, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 36635 + }, + { + "epoch": 0.36636, + "grad_norm": 0.9232468671354163, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 36636 + }, + { + "epoch": 0.36637, + "grad_norm": 0.9300883758189907, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 36637 + }, + { + "epoch": 0.36638, + "grad_norm": 1.1248516700338465, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 36638 + }, + { + "epoch": 0.36639, + "grad_norm": 1.2043611760843687, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 36639 + }, + { + "epoch": 0.3664, + "grad_norm": 0.9637544081547083, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 36640 + }, + { + "epoch": 0.36641, + "grad_norm": 0.9058350391736724, + "learning_rate": 0.003, + "loss": 4.043, + "step": 36641 + }, + { + "epoch": 0.36642, + "grad_norm": 0.7378996528312682, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 36642 + }, + { + "epoch": 0.36643, + "grad_norm": 0.9591999593759513, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 36643 + }, + { + "epoch": 0.36644, + "grad_norm": 1.7867765265964548, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 36644 + }, + { + "epoch": 0.36645, + "grad_norm": 0.7344062930132821, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 36645 + }, + { + "epoch": 0.36646, + "grad_norm": 0.9966828039842984, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 36646 + }, + { + "epoch": 0.36647, + "grad_norm": 1.209829058796796, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 36647 + }, + { + "epoch": 0.36648, + "grad_norm": 0.9600621935359032, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 36648 + }, + { + "epoch": 0.36649, + "grad_norm": 0.9496065865790352, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 36649 + }, + { + "epoch": 0.3665, + "grad_norm": 1.0596680161677248, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 36650 + }, + { + "epoch": 0.36651, + "grad_norm": 0.8599837031129639, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 36651 + }, + { + "epoch": 0.36652, + "grad_norm": 0.9855141261036481, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 36652 + }, + { + "epoch": 0.36653, + "grad_norm": 1.3536633801799094, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 36653 + }, + { + "epoch": 0.36654, + "grad_norm": 1.001631887964961, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 36654 + }, + { + "epoch": 0.36655, + "grad_norm": 1.3659862701411893, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 36655 + }, + { + "epoch": 0.36656, + "grad_norm": 1.1682371172604085, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 36656 + }, + { + "epoch": 0.36657, + "grad_norm": 1.0617884295128808, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 36657 + }, + { + "epoch": 0.36658, + "grad_norm": 1.2936512149987218, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 36658 + }, + { + "epoch": 0.36659, + "grad_norm": 1.0732094181152783, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 36659 + }, + { + "epoch": 0.3666, + "grad_norm": 1.0942868066736606, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 36660 + }, + { + "epoch": 0.36661, + "grad_norm": 1.129582102238108, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 36661 + }, + { + "epoch": 0.36662, + "grad_norm": 1.0529455137639283, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 36662 + }, + { + "epoch": 0.36663, + "grad_norm": 1.0765207282958489, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 36663 + }, + { + "epoch": 0.36664, + "grad_norm": 1.0212061215887798, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 36664 + }, + { + "epoch": 0.36665, + "grad_norm": 1.225288103270454, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 36665 + }, + { + "epoch": 0.36666, + "grad_norm": 0.8936292455441225, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 36666 + }, + { + "epoch": 0.36667, + "grad_norm": 1.0359735890396933, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 36667 + }, + { + "epoch": 0.36668, + "grad_norm": 1.1624382056839075, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 36668 + }, + { + "epoch": 0.36669, + "grad_norm": 0.9209172416340784, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 36669 + }, + { + "epoch": 0.3667, + "grad_norm": 1.1225851056542506, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 36670 + }, + { + "epoch": 0.36671, + "grad_norm": 1.0008073591412907, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 36671 + }, + { + "epoch": 0.36672, + "grad_norm": 1.108557953006431, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 36672 + }, + { + "epoch": 0.36673, + "grad_norm": 1.3306270611021644, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 36673 + }, + { + "epoch": 0.36674, + "grad_norm": 1.032929321405938, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 36674 + }, + { + "epoch": 0.36675, + "grad_norm": 0.9923875693573929, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 36675 + }, + { + "epoch": 0.36676, + "grad_norm": 1.1745945997507456, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 36676 + }, + { + "epoch": 0.36677, + "grad_norm": 1.3312811374340814, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 36677 + }, + { + "epoch": 0.36678, + "grad_norm": 1.0358273262002886, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 36678 + }, + { + "epoch": 0.36679, + "grad_norm": 1.014313038073604, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 36679 + }, + { + "epoch": 0.3668, + "grad_norm": 1.3348540222846816, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 36680 + }, + { + "epoch": 0.36681, + "grad_norm": 1.0288314890525077, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 36681 + }, + { + "epoch": 0.36682, + "grad_norm": 1.2802376258825057, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 36682 + }, + { + "epoch": 0.36683, + "grad_norm": 1.0541809415569254, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 36683 + }, + { + "epoch": 0.36684, + "grad_norm": 1.015914668363804, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 36684 + }, + { + "epoch": 0.36685, + "grad_norm": 0.9458966496774639, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 36685 + }, + { + "epoch": 0.36686, + "grad_norm": 1.0311208701236954, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 36686 + }, + { + "epoch": 0.36687, + "grad_norm": 1.2716920613787408, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 36687 + }, + { + "epoch": 0.36688, + "grad_norm": 0.8434647358958413, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 36688 + }, + { + "epoch": 0.36689, + "grad_norm": 0.9187259940218374, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 36689 + }, + { + "epoch": 0.3669, + "grad_norm": 1.2706687916152084, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 36690 + }, + { + "epoch": 0.36691, + "grad_norm": 1.0310539338067068, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 36691 + }, + { + "epoch": 0.36692, + "grad_norm": 1.5675493691652396, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 36692 + }, + { + "epoch": 0.36693, + "grad_norm": 0.6849030432487769, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 36693 + }, + { + "epoch": 0.36694, + "grad_norm": 0.9415149967644655, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 36694 + }, + { + "epoch": 0.36695, + "grad_norm": 1.742608534796348, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 36695 + }, + { + "epoch": 0.36696, + "grad_norm": 0.6767158342948962, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 36696 + }, + { + "epoch": 0.36697, + "grad_norm": 1.374379029613088, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 36697 + }, + { + "epoch": 0.36698, + "grad_norm": 1.2966652938312064, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 36698 + }, + { + "epoch": 0.36699, + "grad_norm": 1.139345494157859, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 36699 + }, + { + "epoch": 0.367, + "grad_norm": 1.1351536645920552, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 36700 + }, + { + "epoch": 0.36701, + "grad_norm": 1.0700187110414061, + "learning_rate": 0.003, + "loss": 4.055, + "step": 36701 + }, + { + "epoch": 0.36702, + "grad_norm": 0.9730907884168878, + "learning_rate": 0.003, + "loss": 4.056, + "step": 36702 + }, + { + "epoch": 0.36703, + "grad_norm": 1.1365384872156639, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 36703 + }, + { + "epoch": 0.36704, + "grad_norm": 0.9174664074626036, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 36704 + }, + { + "epoch": 0.36705, + "grad_norm": 0.9932640637570507, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 36705 + }, + { + "epoch": 0.36706, + "grad_norm": 1.2158607195515079, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 36706 + }, + { + "epoch": 0.36707, + "grad_norm": 1.309848487527252, + "learning_rate": 0.003, + "loss": 4.058, + "step": 36707 + }, + { + "epoch": 0.36708, + "grad_norm": 0.8736793884933508, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 36708 + }, + { + "epoch": 0.36709, + "grad_norm": 0.9573199252944322, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 36709 + }, + { + "epoch": 0.3671, + "grad_norm": 1.1332200621512942, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 36710 + }, + { + "epoch": 0.36711, + "grad_norm": 1.2812939898381115, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 36711 + }, + { + "epoch": 0.36712, + "grad_norm": 0.9969087403506102, + "learning_rate": 0.003, + "loss": 4.077, + "step": 36712 + }, + { + "epoch": 0.36713, + "grad_norm": 1.2175406802816289, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 36713 + }, + { + "epoch": 0.36714, + "grad_norm": 0.9008886329842216, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 36714 + }, + { + "epoch": 0.36715, + "grad_norm": 0.8785780509807796, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 36715 + }, + { + "epoch": 0.36716, + "grad_norm": 0.963023463043161, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 36716 + }, + { + "epoch": 0.36717, + "grad_norm": 1.3004584787598514, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 36717 + }, + { + "epoch": 0.36718, + "grad_norm": 1.0492649392448308, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 36718 + }, + { + "epoch": 0.36719, + "grad_norm": 1.137928776748521, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 36719 + }, + { + "epoch": 0.3672, + "grad_norm": 1.1538613051063928, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 36720 + }, + { + "epoch": 0.36721, + "grad_norm": 1.1592562939423672, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 36721 + }, + { + "epoch": 0.36722, + "grad_norm": 0.9578077287979162, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 36722 + }, + { + "epoch": 0.36723, + "grad_norm": 1.0169779516835546, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 36723 + }, + { + "epoch": 0.36724, + "grad_norm": 1.1942991780444228, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 36724 + }, + { + "epoch": 0.36725, + "grad_norm": 1.157529652129079, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 36725 + }, + { + "epoch": 0.36726, + "grad_norm": 1.0590589532194683, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 36726 + }, + { + "epoch": 0.36727, + "grad_norm": 1.1188861528588578, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 36727 + }, + { + "epoch": 0.36728, + "grad_norm": 1.0412627549327063, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 36728 + }, + { + "epoch": 0.36729, + "grad_norm": 1.2060227522239972, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 36729 + }, + { + "epoch": 0.3673, + "grad_norm": 0.9522545361804458, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 36730 + }, + { + "epoch": 0.36731, + "grad_norm": 1.145549887975287, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 36731 + }, + { + "epoch": 0.36732, + "grad_norm": 1.1972900455010156, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 36732 + }, + { + "epoch": 0.36733, + "grad_norm": 1.0557959347939956, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 36733 + }, + { + "epoch": 0.36734, + "grad_norm": 1.1039432289077018, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 36734 + }, + { + "epoch": 0.36735, + "grad_norm": 1.0917210191811262, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 36735 + }, + { + "epoch": 0.36736, + "grad_norm": 1.1039496292297304, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 36736 + }, + { + "epoch": 0.36737, + "grad_norm": 1.3651661800591235, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 36737 + }, + { + "epoch": 0.36738, + "grad_norm": 0.9049340933329851, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 36738 + }, + { + "epoch": 0.36739, + "grad_norm": 1.0089515851622026, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 36739 + }, + { + "epoch": 0.3674, + "grad_norm": 1.2593948200019387, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 36740 + }, + { + "epoch": 0.36741, + "grad_norm": 0.9706243828032232, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 36741 + }, + { + "epoch": 0.36742, + "grad_norm": 1.1159101397515938, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 36742 + }, + { + "epoch": 0.36743, + "grad_norm": 0.8780169817905793, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 36743 + }, + { + "epoch": 0.36744, + "grad_norm": 0.9391598858563764, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 36744 + }, + { + "epoch": 0.36745, + "grad_norm": 1.1797043780065108, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 36745 + }, + { + "epoch": 0.36746, + "grad_norm": 0.8883656585131373, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 36746 + }, + { + "epoch": 0.36747, + "grad_norm": 1.0778372283973536, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 36747 + }, + { + "epoch": 0.36748, + "grad_norm": 1.258418378409869, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 36748 + }, + { + "epoch": 0.36749, + "grad_norm": 1.0094829375944143, + "learning_rate": 0.003, + "loss": 4.039, + "step": 36749 + }, + { + "epoch": 0.3675, + "grad_norm": 1.184589464081366, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 36750 + }, + { + "epoch": 0.36751, + "grad_norm": 0.9860909597113886, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 36751 + }, + { + "epoch": 0.36752, + "grad_norm": 1.2544962184636783, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 36752 + }, + { + "epoch": 0.36753, + "grad_norm": 1.1152005033069332, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 36753 + }, + { + "epoch": 0.36754, + "grad_norm": 0.9425752942717838, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 36754 + }, + { + "epoch": 0.36755, + "grad_norm": 1.1019841784539128, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 36755 + }, + { + "epoch": 0.36756, + "grad_norm": 1.0620534448214458, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 36756 + }, + { + "epoch": 0.36757, + "grad_norm": 1.0288985730767333, + "learning_rate": 0.003, + "loss": 4.053, + "step": 36757 + }, + { + "epoch": 0.36758, + "grad_norm": 1.0136375790744812, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 36758 + }, + { + "epoch": 0.36759, + "grad_norm": 1.1806653677042926, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 36759 + }, + { + "epoch": 0.3676, + "grad_norm": 1.2009612562162393, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 36760 + }, + { + "epoch": 0.36761, + "grad_norm": 1.071463386082698, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 36761 + }, + { + "epoch": 0.36762, + "grad_norm": 1.1723802426589567, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 36762 + }, + { + "epoch": 0.36763, + "grad_norm": 0.9716618624809251, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 36763 + }, + { + "epoch": 0.36764, + "grad_norm": 1.0751402980776052, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 36764 + }, + { + "epoch": 0.36765, + "grad_norm": 1.0568339248082685, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 36765 + }, + { + "epoch": 0.36766, + "grad_norm": 1.1538258520284372, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 36766 + }, + { + "epoch": 0.36767, + "grad_norm": 1.2272422685159348, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 36767 + }, + { + "epoch": 0.36768, + "grad_norm": 1.0439447534678081, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 36768 + }, + { + "epoch": 0.36769, + "grad_norm": 0.9823430379072373, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 36769 + }, + { + "epoch": 0.3677, + "grad_norm": 1.1035599449447657, + "learning_rate": 0.003, + "loss": 4.033, + "step": 36770 + }, + { + "epoch": 0.36771, + "grad_norm": 0.9871874544835367, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 36771 + }, + { + "epoch": 0.36772, + "grad_norm": 1.1822101443410746, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 36772 + }, + { + "epoch": 0.36773, + "grad_norm": 1.0071878285829137, + "learning_rate": 0.003, + "loss": 4.024, + "step": 36773 + }, + { + "epoch": 0.36774, + "grad_norm": 1.3184413986020673, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 36774 + }, + { + "epoch": 0.36775, + "grad_norm": 0.8079959926288589, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 36775 + }, + { + "epoch": 0.36776, + "grad_norm": 0.7919320278734041, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 36776 + }, + { + "epoch": 0.36777, + "grad_norm": 1.0900857060898959, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 36777 + }, + { + "epoch": 0.36778, + "grad_norm": 1.162294936035993, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 36778 + }, + { + "epoch": 0.36779, + "grad_norm": 1.0818874773516982, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 36779 + }, + { + "epoch": 0.3678, + "grad_norm": 1.162344014728076, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 36780 + }, + { + "epoch": 0.36781, + "grad_norm": 1.115430284825253, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 36781 + }, + { + "epoch": 0.36782, + "grad_norm": 1.0793749122828793, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 36782 + }, + { + "epoch": 0.36783, + "grad_norm": 1.0903159937639797, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 36783 + }, + { + "epoch": 0.36784, + "grad_norm": 0.979676672893971, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 36784 + }, + { + "epoch": 0.36785, + "grad_norm": 1.237463412605686, + "learning_rate": 0.003, + "loss": 4.02, + "step": 36785 + }, + { + "epoch": 0.36786, + "grad_norm": 1.1209401330938669, + "learning_rate": 0.003, + "loss": 4.052, + "step": 36786 + }, + { + "epoch": 0.36787, + "grad_norm": 1.2879136277615275, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 36787 + }, + { + "epoch": 0.36788, + "grad_norm": 1.1202631488326948, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 36788 + }, + { + "epoch": 0.36789, + "grad_norm": 0.9135730002238083, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 36789 + }, + { + "epoch": 0.3679, + "grad_norm": 0.9544467219021396, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 36790 + }, + { + "epoch": 0.36791, + "grad_norm": 1.156019075063474, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 36791 + }, + { + "epoch": 0.36792, + "grad_norm": 1.2020076542656577, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 36792 + }, + { + "epoch": 0.36793, + "grad_norm": 0.9982666888651991, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 36793 + }, + { + "epoch": 0.36794, + "grad_norm": 1.2311279220634512, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 36794 + }, + { + "epoch": 0.36795, + "grad_norm": 1.1300606236285844, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 36795 + }, + { + "epoch": 0.36796, + "grad_norm": 0.9415868768832414, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 36796 + }, + { + "epoch": 0.36797, + "grad_norm": 1.136926416038257, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 36797 + }, + { + "epoch": 0.36798, + "grad_norm": 1.2730126272231532, + "learning_rate": 0.003, + "loss": 4.03, + "step": 36798 + }, + { + "epoch": 0.36799, + "grad_norm": 0.8976533978754061, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 36799 + }, + { + "epoch": 0.368, + "grad_norm": 0.9184057728192875, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 36800 + }, + { + "epoch": 0.36801, + "grad_norm": 0.9738983349463793, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 36801 + }, + { + "epoch": 0.36802, + "grad_norm": 1.0491302928328454, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 36802 + }, + { + "epoch": 0.36803, + "grad_norm": 1.2393683906484203, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 36803 + }, + { + "epoch": 0.36804, + "grad_norm": 0.8600670437034211, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 36804 + }, + { + "epoch": 0.36805, + "grad_norm": 0.8230744858360624, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 36805 + }, + { + "epoch": 0.36806, + "grad_norm": 0.9429179578677309, + "learning_rate": 0.003, + "loss": 4.019, + "step": 36806 + }, + { + "epoch": 0.36807, + "grad_norm": 0.9816576553371672, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 36807 + }, + { + "epoch": 0.36808, + "grad_norm": 1.3806159121647612, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 36808 + }, + { + "epoch": 0.36809, + "grad_norm": 1.030816629905426, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 36809 + }, + { + "epoch": 0.3681, + "grad_norm": 1.1145269957891504, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 36810 + }, + { + "epoch": 0.36811, + "grad_norm": 1.0395495935518235, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 36811 + }, + { + "epoch": 0.36812, + "grad_norm": 1.0726318603360405, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 36812 + }, + { + "epoch": 0.36813, + "grad_norm": 1.1162550213399283, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 36813 + }, + { + "epoch": 0.36814, + "grad_norm": 1.3311388966510278, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 36814 + }, + { + "epoch": 0.36815, + "grad_norm": 1.0423452457440578, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 36815 + }, + { + "epoch": 0.36816, + "grad_norm": 1.1978775383903506, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 36816 + }, + { + "epoch": 0.36817, + "grad_norm": 1.2555957294498163, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 36817 + }, + { + "epoch": 0.36818, + "grad_norm": 0.991484403446095, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 36818 + }, + { + "epoch": 0.36819, + "grad_norm": 1.2079112295218704, + "learning_rate": 0.003, + "loss": 4.041, + "step": 36819 + }, + { + "epoch": 0.3682, + "grad_norm": 0.8831663756258646, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 36820 + }, + { + "epoch": 0.36821, + "grad_norm": 0.8812970217750674, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 36821 + }, + { + "epoch": 0.36822, + "grad_norm": 1.1598843000942587, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 36822 + }, + { + "epoch": 0.36823, + "grad_norm": 1.316404071388688, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 36823 + }, + { + "epoch": 0.36824, + "grad_norm": 0.8491312327258894, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 36824 + }, + { + "epoch": 0.36825, + "grad_norm": 0.8263630416156109, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 36825 + }, + { + "epoch": 0.36826, + "grad_norm": 0.9918837042742982, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 36826 + }, + { + "epoch": 0.36827, + "grad_norm": 1.1301494737773772, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 36827 + }, + { + "epoch": 0.36828, + "grad_norm": 0.9119003336024476, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 36828 + }, + { + "epoch": 0.36829, + "grad_norm": 1.1045648225384548, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 36829 + }, + { + "epoch": 0.3683, + "grad_norm": 1.1467045758182597, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 36830 + }, + { + "epoch": 0.36831, + "grad_norm": 1.214214126431885, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 36831 + }, + { + "epoch": 0.36832, + "grad_norm": 0.9110105350685518, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 36832 + }, + { + "epoch": 0.36833, + "grad_norm": 0.9579243356333408, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 36833 + }, + { + "epoch": 0.36834, + "grad_norm": 1.3455197636278275, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 36834 + }, + { + "epoch": 0.36835, + "grad_norm": 1.0873598366300596, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 36835 + }, + { + "epoch": 0.36836, + "grad_norm": 1.29222567947839, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 36836 + }, + { + "epoch": 0.36837, + "grad_norm": 0.9684451992247683, + "learning_rate": 0.003, + "loss": 4.05, + "step": 36837 + }, + { + "epoch": 0.36838, + "grad_norm": 0.9807242033941272, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 36838 + }, + { + "epoch": 0.36839, + "grad_norm": 1.2098257705061106, + "learning_rate": 0.003, + "loss": 4.05, + "step": 36839 + }, + { + "epoch": 0.3684, + "grad_norm": 1.1774938551141587, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 36840 + }, + { + "epoch": 0.36841, + "grad_norm": 1.1045698817984688, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 36841 + }, + { + "epoch": 0.36842, + "grad_norm": 1.0612958725134574, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 36842 + }, + { + "epoch": 0.36843, + "grad_norm": 1.1841277178369853, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 36843 + }, + { + "epoch": 0.36844, + "grad_norm": 0.9914428531684562, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 36844 + }, + { + "epoch": 0.36845, + "grad_norm": 1.0908397463348416, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 36845 + }, + { + "epoch": 0.36846, + "grad_norm": 0.9194648796078642, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 36846 + }, + { + "epoch": 0.36847, + "grad_norm": 1.1258836892750659, + "learning_rate": 0.003, + "loss": 4.059, + "step": 36847 + }, + { + "epoch": 0.36848, + "grad_norm": 0.9990940504154364, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 36848 + }, + { + "epoch": 0.36849, + "grad_norm": 1.2036169692588825, + "learning_rate": 0.003, + "loss": 4.025, + "step": 36849 + }, + { + "epoch": 0.3685, + "grad_norm": 1.065715066663558, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 36850 + }, + { + "epoch": 0.36851, + "grad_norm": 1.117103815473708, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 36851 + }, + { + "epoch": 0.36852, + "grad_norm": 1.0753354279582257, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 36852 + }, + { + "epoch": 0.36853, + "grad_norm": 0.9088959540051702, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 36853 + }, + { + "epoch": 0.36854, + "grad_norm": 1.0405248564406466, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 36854 + }, + { + "epoch": 0.36855, + "grad_norm": 1.4633430038974256, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 36855 + }, + { + "epoch": 0.36856, + "grad_norm": 0.9452603725917076, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 36856 + }, + { + "epoch": 0.36857, + "grad_norm": 1.041224826332132, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 36857 + }, + { + "epoch": 0.36858, + "grad_norm": 1.0290862908606873, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 36858 + }, + { + "epoch": 0.36859, + "grad_norm": 1.2483401654909971, + "learning_rate": 0.003, + "loss": 4.039, + "step": 36859 + }, + { + "epoch": 0.3686, + "grad_norm": 0.9428967357648123, + "learning_rate": 0.003, + "loss": 4.015, + "step": 36860 + }, + { + "epoch": 0.36861, + "grad_norm": 1.0356106114073311, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 36861 + }, + { + "epoch": 0.36862, + "grad_norm": 1.3380584025882774, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 36862 + }, + { + "epoch": 0.36863, + "grad_norm": 0.9902782875592723, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 36863 + }, + { + "epoch": 0.36864, + "grad_norm": 1.0211459303779915, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 36864 + }, + { + "epoch": 0.36865, + "grad_norm": 1.0290491610511499, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 36865 + }, + { + "epoch": 0.36866, + "grad_norm": 0.9774167294721918, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 36866 + }, + { + "epoch": 0.36867, + "grad_norm": 1.4649884719705781, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 36867 + }, + { + "epoch": 0.36868, + "grad_norm": 0.9626934789079992, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 36868 + }, + { + "epoch": 0.36869, + "grad_norm": 1.0714382202771984, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 36869 + }, + { + "epoch": 0.3687, + "grad_norm": 1.0811877649964088, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 36870 + }, + { + "epoch": 0.36871, + "grad_norm": 1.0876942824039477, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 36871 + }, + { + "epoch": 0.36872, + "grad_norm": 1.141885359664663, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 36872 + }, + { + "epoch": 0.36873, + "grad_norm": 0.9557214028374863, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 36873 + }, + { + "epoch": 0.36874, + "grad_norm": 1.0697290576677052, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 36874 + }, + { + "epoch": 0.36875, + "grad_norm": 1.02273088313861, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 36875 + }, + { + "epoch": 0.36876, + "grad_norm": 1.1985504414319297, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 36876 + }, + { + "epoch": 0.36877, + "grad_norm": 1.0739194716020999, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 36877 + }, + { + "epoch": 0.36878, + "grad_norm": 1.4078830002462401, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 36878 + }, + { + "epoch": 0.36879, + "grad_norm": 0.8363379937668536, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 36879 + }, + { + "epoch": 0.3688, + "grad_norm": 0.9208248641019772, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 36880 + }, + { + "epoch": 0.36881, + "grad_norm": 1.3157737291270606, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 36881 + }, + { + "epoch": 0.36882, + "grad_norm": 0.952418828290737, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 36882 + }, + { + "epoch": 0.36883, + "grad_norm": 1.0068175298188362, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 36883 + }, + { + "epoch": 0.36884, + "grad_norm": 1.3476383732988695, + "learning_rate": 0.003, + "loss": 4.068, + "step": 36884 + }, + { + "epoch": 0.36885, + "grad_norm": 1.0211984000625696, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 36885 + }, + { + "epoch": 0.36886, + "grad_norm": 1.3147284575339786, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 36886 + }, + { + "epoch": 0.36887, + "grad_norm": 0.9300886532867488, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 36887 + }, + { + "epoch": 0.36888, + "grad_norm": 1.063920337959678, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 36888 + }, + { + "epoch": 0.36889, + "grad_norm": 1.1462844532165206, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 36889 + }, + { + "epoch": 0.3689, + "grad_norm": 1.177064499530388, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 36890 + }, + { + "epoch": 0.36891, + "grad_norm": 1.138523902091858, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 36891 + }, + { + "epoch": 0.36892, + "grad_norm": 0.9743779341943675, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 36892 + }, + { + "epoch": 0.36893, + "grad_norm": 1.0591317165292768, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 36893 + }, + { + "epoch": 0.36894, + "grad_norm": 1.0411352650738364, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 36894 + }, + { + "epoch": 0.36895, + "grad_norm": 1.0927061859501181, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 36895 + }, + { + "epoch": 0.36896, + "grad_norm": 1.0068639784767648, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 36896 + }, + { + "epoch": 0.36897, + "grad_norm": 1.2565699437067663, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 36897 + }, + { + "epoch": 0.36898, + "grad_norm": 1.048936416186365, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 36898 + }, + { + "epoch": 0.36899, + "grad_norm": 1.0724456629436478, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 36899 + }, + { + "epoch": 0.369, + "grad_norm": 1.2946496090659878, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 36900 + }, + { + "epoch": 0.36901, + "grad_norm": 1.0228702431999794, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 36901 + }, + { + "epoch": 0.36902, + "grad_norm": 1.0734008812983213, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 36902 + }, + { + "epoch": 0.36903, + "grad_norm": 1.2959550315652766, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 36903 + }, + { + "epoch": 0.36904, + "grad_norm": 0.7504156820537512, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 36904 + }, + { + "epoch": 0.36905, + "grad_norm": 0.8955780169809412, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 36905 + }, + { + "epoch": 0.36906, + "grad_norm": 1.020090463808491, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 36906 + }, + { + "epoch": 0.36907, + "grad_norm": 1.1756789769232836, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 36907 + }, + { + "epoch": 0.36908, + "grad_norm": 1.2080219689353557, + "learning_rate": 0.003, + "loss": 4.048, + "step": 36908 + }, + { + "epoch": 0.36909, + "grad_norm": 1.022606681228568, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 36909 + }, + { + "epoch": 0.3691, + "grad_norm": 1.3264101822629075, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 36910 + }, + { + "epoch": 0.36911, + "grad_norm": 1.1654791191110194, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 36911 + }, + { + "epoch": 0.36912, + "grad_norm": 1.124516257649447, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 36912 + }, + { + "epoch": 0.36913, + "grad_norm": 1.017849961373814, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 36913 + }, + { + "epoch": 0.36914, + "grad_norm": 1.2165521358856883, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 36914 + }, + { + "epoch": 0.36915, + "grad_norm": 1.0290434173927372, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 36915 + }, + { + "epoch": 0.36916, + "grad_norm": 1.2334425973529206, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 36916 + }, + { + "epoch": 0.36917, + "grad_norm": 0.934905535699199, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 36917 + }, + { + "epoch": 0.36918, + "grad_norm": 0.9755792838165008, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 36918 + }, + { + "epoch": 0.36919, + "grad_norm": 1.0046452916898796, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 36919 + }, + { + "epoch": 0.3692, + "grad_norm": 1.1624665291970189, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 36920 + }, + { + "epoch": 0.36921, + "grad_norm": 1.014485857955837, + "learning_rate": 0.003, + "loss": 4.039, + "step": 36921 + }, + { + "epoch": 0.36922, + "grad_norm": 1.2118135157073417, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 36922 + }, + { + "epoch": 0.36923, + "grad_norm": 1.094247048695912, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 36923 + }, + { + "epoch": 0.36924, + "grad_norm": 1.0073193685187052, + "learning_rate": 0.003, + "loss": 4.028, + "step": 36924 + }, + { + "epoch": 0.36925, + "grad_norm": 1.0002710562711832, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 36925 + }, + { + "epoch": 0.36926, + "grad_norm": 1.0360728188085002, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 36926 + }, + { + "epoch": 0.36927, + "grad_norm": 1.150471405538932, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 36927 + }, + { + "epoch": 0.36928, + "grad_norm": 1.1488667647496584, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 36928 + }, + { + "epoch": 0.36929, + "grad_norm": 1.1797848639666648, + "learning_rate": 0.003, + "loss": 4.06, + "step": 36929 + }, + { + "epoch": 0.3693, + "grad_norm": 1.0782036742858423, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 36930 + }, + { + "epoch": 0.36931, + "grad_norm": 1.1103652934240373, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 36931 + }, + { + "epoch": 0.36932, + "grad_norm": 1.14314947904196, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 36932 + }, + { + "epoch": 0.36933, + "grad_norm": 0.8581284504130965, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 36933 + }, + { + "epoch": 0.36934, + "grad_norm": 0.937657502166885, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 36934 + }, + { + "epoch": 0.36935, + "grad_norm": 1.1826651651256377, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 36935 + }, + { + "epoch": 0.36936, + "grad_norm": 1.0761288066293166, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 36936 + }, + { + "epoch": 0.36937, + "grad_norm": 1.2311404623982678, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 36937 + }, + { + "epoch": 0.36938, + "grad_norm": 0.9670598741010287, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 36938 + }, + { + "epoch": 0.36939, + "grad_norm": 1.0933156042671575, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 36939 + }, + { + "epoch": 0.3694, + "grad_norm": 0.9015330099672482, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 36940 + }, + { + "epoch": 0.36941, + "grad_norm": 0.9103628659660244, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 36941 + }, + { + "epoch": 0.36942, + "grad_norm": 0.9381013104418867, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 36942 + }, + { + "epoch": 0.36943, + "grad_norm": 1.049747619834424, + "learning_rate": 0.003, + "loss": 4.046, + "step": 36943 + }, + { + "epoch": 0.36944, + "grad_norm": 1.1853249264023682, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 36944 + }, + { + "epoch": 0.36945, + "grad_norm": 1.22717203732281, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 36945 + }, + { + "epoch": 0.36946, + "grad_norm": 0.8933089761891145, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 36946 + }, + { + "epoch": 0.36947, + "grad_norm": 0.8231798886678069, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 36947 + }, + { + "epoch": 0.36948, + "grad_norm": 0.9241248612304654, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 36948 + }, + { + "epoch": 0.36949, + "grad_norm": 1.1721897227240206, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 36949 + }, + { + "epoch": 0.3695, + "grad_norm": 1.095683207191754, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 36950 + }, + { + "epoch": 0.36951, + "grad_norm": 1.4090259287616493, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 36951 + }, + { + "epoch": 0.36952, + "grad_norm": 0.8935562159653101, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 36952 + }, + { + "epoch": 0.36953, + "grad_norm": 1.0316220885504888, + "learning_rate": 0.003, + "loss": 4.037, + "step": 36953 + }, + { + "epoch": 0.36954, + "grad_norm": 1.207837771450202, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 36954 + }, + { + "epoch": 0.36955, + "grad_norm": 0.9194690041869815, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 36955 + }, + { + "epoch": 0.36956, + "grad_norm": 1.0884812221275302, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 36956 + }, + { + "epoch": 0.36957, + "grad_norm": 1.1902478365169329, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 36957 + }, + { + "epoch": 0.36958, + "grad_norm": 1.178551194588627, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 36958 + }, + { + "epoch": 0.36959, + "grad_norm": 0.9916716874794539, + "learning_rate": 0.003, + "loss": 4.02, + "step": 36959 + }, + { + "epoch": 0.3696, + "grad_norm": 0.999861124664258, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 36960 + }, + { + "epoch": 0.36961, + "grad_norm": 1.1654578947411602, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 36961 + }, + { + "epoch": 0.36962, + "grad_norm": 1.1234630096639329, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 36962 + }, + { + "epoch": 0.36963, + "grad_norm": 0.878672283220401, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 36963 + }, + { + "epoch": 0.36964, + "grad_norm": 0.9704659016045442, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 36964 + }, + { + "epoch": 0.36965, + "grad_norm": 1.5004283542347405, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 36965 + }, + { + "epoch": 0.36966, + "grad_norm": 1.1249801297320703, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 36966 + }, + { + "epoch": 0.36967, + "grad_norm": 1.0635306338423665, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 36967 + }, + { + "epoch": 0.36968, + "grad_norm": 1.0596907347301117, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 36968 + }, + { + "epoch": 0.36969, + "grad_norm": 1.1568659643059684, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 36969 + }, + { + "epoch": 0.3697, + "grad_norm": 1.4108868045125014, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 36970 + }, + { + "epoch": 0.36971, + "grad_norm": 0.8863032943623759, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 36971 + }, + { + "epoch": 0.36972, + "grad_norm": 0.8599623639314211, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 36972 + }, + { + "epoch": 0.36973, + "grad_norm": 0.9298570865725653, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 36973 + }, + { + "epoch": 0.36974, + "grad_norm": 1.0935576926951076, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 36974 + }, + { + "epoch": 0.36975, + "grad_norm": 1.2398236103735163, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 36975 + }, + { + "epoch": 0.36976, + "grad_norm": 1.0208029874003899, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 36976 + }, + { + "epoch": 0.36977, + "grad_norm": 1.2989441091165457, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 36977 + }, + { + "epoch": 0.36978, + "grad_norm": 0.8843421511176169, + "learning_rate": 0.003, + "loss": 4.038, + "step": 36978 + }, + { + "epoch": 0.36979, + "grad_norm": 0.9486668730599822, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 36979 + }, + { + "epoch": 0.3698, + "grad_norm": 1.2605020913138694, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 36980 + }, + { + "epoch": 0.36981, + "grad_norm": 1.1772917007079207, + "learning_rate": 0.003, + "loss": 4.029, + "step": 36981 + }, + { + "epoch": 0.36982, + "grad_norm": 1.0673741891725559, + "learning_rate": 0.003, + "loss": 4.039, + "step": 36982 + }, + { + "epoch": 0.36983, + "grad_norm": 1.0917774487470617, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 36983 + }, + { + "epoch": 0.36984, + "grad_norm": 1.0223222535744914, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 36984 + }, + { + "epoch": 0.36985, + "grad_norm": 0.9838612511661106, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 36985 + }, + { + "epoch": 0.36986, + "grad_norm": 1.1691134220437283, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 36986 + }, + { + "epoch": 0.36987, + "grad_norm": 1.209945599154538, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 36987 + }, + { + "epoch": 0.36988, + "grad_norm": 1.173796527158436, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 36988 + }, + { + "epoch": 0.36989, + "grad_norm": 1.0463921538376828, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 36989 + }, + { + "epoch": 0.3699, + "grad_norm": 1.119909207853039, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 36990 + }, + { + "epoch": 0.36991, + "grad_norm": 1.0629472171920558, + "learning_rate": 0.003, + "loss": 4.018, + "step": 36991 + }, + { + "epoch": 0.36992, + "grad_norm": 1.0991720353161851, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 36992 + }, + { + "epoch": 0.36993, + "grad_norm": 1.1515128132902783, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 36993 + }, + { + "epoch": 0.36994, + "grad_norm": 1.0641580209552102, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 36994 + }, + { + "epoch": 0.36995, + "grad_norm": 1.1231493850631167, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 36995 + }, + { + "epoch": 0.36996, + "grad_norm": 0.857304922075148, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 36996 + }, + { + "epoch": 0.36997, + "grad_norm": 0.8582631657127212, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 36997 + }, + { + "epoch": 0.36998, + "grad_norm": 1.0663322051992756, + "learning_rate": 0.003, + "loss": 4.018, + "step": 36998 + }, + { + "epoch": 0.36999, + "grad_norm": 1.3155167775445304, + "learning_rate": 0.003, + "loss": 4.044, + "step": 36999 + }, + { + "epoch": 0.37, + "grad_norm": 1.1426554634015604, + "learning_rate": 0.003, + "loss": 4.013, + "step": 37000 + }, + { + "epoch": 0.37001, + "grad_norm": 1.264721309685217, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 37001 + }, + { + "epoch": 0.37002, + "grad_norm": 0.9728098608184237, + "learning_rate": 0.003, + "loss": 4.008, + "step": 37002 + }, + { + "epoch": 0.37003, + "grad_norm": 1.1218469380739515, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 37003 + }, + { + "epoch": 0.37004, + "grad_norm": 1.2636479021001585, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 37004 + }, + { + "epoch": 0.37005, + "grad_norm": 1.102240126934674, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 37005 + }, + { + "epoch": 0.37006, + "grad_norm": 1.1804601675826987, + "learning_rate": 0.003, + "loss": 4.046, + "step": 37006 + }, + { + "epoch": 0.37007, + "grad_norm": 0.9743923505996531, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 37007 + }, + { + "epoch": 0.37008, + "grad_norm": 1.0946212840263196, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 37008 + }, + { + "epoch": 0.37009, + "grad_norm": 0.9084149233253436, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 37009 + }, + { + "epoch": 0.3701, + "grad_norm": 0.9889752484158121, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 37010 + }, + { + "epoch": 0.37011, + "grad_norm": 1.136153892238816, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 37011 + }, + { + "epoch": 0.37012, + "grad_norm": 1.1692959495305257, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 37012 + }, + { + "epoch": 0.37013, + "grad_norm": 1.0187920669158892, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 37013 + }, + { + "epoch": 0.37014, + "grad_norm": 1.1372431031070718, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 37014 + }, + { + "epoch": 0.37015, + "grad_norm": 0.9956397539442069, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 37015 + }, + { + "epoch": 0.37016, + "grad_norm": 1.3418075679270536, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 37016 + }, + { + "epoch": 0.37017, + "grad_norm": 1.2975586845726375, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 37017 + }, + { + "epoch": 0.37018, + "grad_norm": 1.0088062522593608, + "learning_rate": 0.003, + "loss": 4.021, + "step": 37018 + }, + { + "epoch": 0.37019, + "grad_norm": 1.1434716923256814, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 37019 + }, + { + "epoch": 0.3702, + "grad_norm": 1.2053805064227714, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 37020 + }, + { + "epoch": 0.37021, + "grad_norm": 0.8789220023246899, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 37021 + }, + { + "epoch": 0.37022, + "grad_norm": 0.9010464439601598, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 37022 + }, + { + "epoch": 0.37023, + "grad_norm": 1.028739332954875, + "learning_rate": 0.003, + "loss": 4.038, + "step": 37023 + }, + { + "epoch": 0.37024, + "grad_norm": 1.2005897642822247, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 37024 + }, + { + "epoch": 0.37025, + "grad_norm": 1.033953828602899, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 37025 + }, + { + "epoch": 0.37026, + "grad_norm": 1.124655540411057, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 37026 + }, + { + "epoch": 0.37027, + "grad_norm": 1.4160540794373777, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 37027 + }, + { + "epoch": 0.37028, + "grad_norm": 0.9693055479096144, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 37028 + }, + { + "epoch": 0.37029, + "grad_norm": 1.3511328199791666, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 37029 + }, + { + "epoch": 0.3703, + "grad_norm": 0.9098087471164509, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 37030 + }, + { + "epoch": 0.37031, + "grad_norm": 1.0434113205636941, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 37031 + }, + { + "epoch": 0.37032, + "grad_norm": 1.4349967706474869, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 37032 + }, + { + "epoch": 0.37033, + "grad_norm": 0.9425477357980881, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 37033 + }, + { + "epoch": 0.37034, + "grad_norm": 0.9582444169581359, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 37034 + }, + { + "epoch": 0.37035, + "grad_norm": 1.2875908165954866, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 37035 + }, + { + "epoch": 0.37036, + "grad_norm": 1.300380170605558, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 37036 + }, + { + "epoch": 0.37037, + "grad_norm": 0.8262713863184468, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 37037 + }, + { + "epoch": 0.37038, + "grad_norm": 0.7613528836081436, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 37038 + }, + { + "epoch": 0.37039, + "grad_norm": 0.947511938898671, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 37039 + }, + { + "epoch": 0.3704, + "grad_norm": 1.1088364796805366, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 37040 + }, + { + "epoch": 0.37041, + "grad_norm": 0.8690991714990448, + "learning_rate": 0.003, + "loss": 4.02, + "step": 37041 + }, + { + "epoch": 0.37042, + "grad_norm": 1.0897528092975926, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 37042 + }, + { + "epoch": 0.37043, + "grad_norm": 1.0927829142934415, + "learning_rate": 0.003, + "loss": 4.035, + "step": 37043 + }, + { + "epoch": 0.37044, + "grad_norm": 1.0432705504315607, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 37044 + }, + { + "epoch": 0.37045, + "grad_norm": 1.415574486440808, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 37045 + }, + { + "epoch": 0.37046, + "grad_norm": 0.7542582511502509, + "learning_rate": 0.003, + "loss": 4.024, + "step": 37046 + }, + { + "epoch": 0.37047, + "grad_norm": 0.749448281719521, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 37047 + }, + { + "epoch": 0.37048, + "grad_norm": 1.0079407645556617, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 37048 + }, + { + "epoch": 0.37049, + "grad_norm": 1.4307869883828295, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 37049 + }, + { + "epoch": 0.3705, + "grad_norm": 1.0564415091739836, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 37050 + }, + { + "epoch": 0.37051, + "grad_norm": 1.1683390487716603, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 37051 + }, + { + "epoch": 0.37052, + "grad_norm": 1.0351116614415015, + "learning_rate": 0.003, + "loss": 4.035, + "step": 37052 + }, + { + "epoch": 0.37053, + "grad_norm": 1.3668187540584809, + "learning_rate": 0.003, + "loss": 4.04, + "step": 37053 + }, + { + "epoch": 0.37054, + "grad_norm": 0.8326662870599263, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 37054 + }, + { + "epoch": 0.37055, + "grad_norm": 0.972283608392925, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 37055 + }, + { + "epoch": 0.37056, + "grad_norm": 1.3346403562855125, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 37056 + }, + { + "epoch": 0.37057, + "grad_norm": 0.9959758056833174, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 37057 + }, + { + "epoch": 0.37058, + "grad_norm": 1.17510636473766, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 37058 + }, + { + "epoch": 0.37059, + "grad_norm": 1.2058659145611537, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 37059 + }, + { + "epoch": 0.3706, + "grad_norm": 1.0086537606086703, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 37060 + }, + { + "epoch": 0.37061, + "grad_norm": 1.221665713564495, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 37061 + }, + { + "epoch": 0.37062, + "grad_norm": 1.117542233394171, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 37062 + }, + { + "epoch": 0.37063, + "grad_norm": 0.8853850422030948, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 37063 + }, + { + "epoch": 0.37064, + "grad_norm": 0.9233949573132626, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 37064 + }, + { + "epoch": 0.37065, + "grad_norm": 1.1885995219235765, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 37065 + }, + { + "epoch": 0.37066, + "grad_norm": 1.3549663378332562, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 37066 + }, + { + "epoch": 0.37067, + "grad_norm": 0.9358474424791327, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 37067 + }, + { + "epoch": 0.37068, + "grad_norm": 1.0067225357424496, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 37068 + }, + { + "epoch": 0.37069, + "grad_norm": 1.2023506704292273, + "learning_rate": 0.003, + "loss": 4.04, + "step": 37069 + }, + { + "epoch": 0.3707, + "grad_norm": 1.0481216896052423, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 37070 + }, + { + "epoch": 0.37071, + "grad_norm": 1.2371678759761084, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 37071 + }, + { + "epoch": 0.37072, + "grad_norm": 0.8189078317732481, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 37072 + }, + { + "epoch": 0.37073, + "grad_norm": 0.8060115784903223, + "learning_rate": 0.003, + "loss": 4.043, + "step": 37073 + }, + { + "epoch": 0.37074, + "grad_norm": 0.8532097829770753, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 37074 + }, + { + "epoch": 0.37075, + "grad_norm": 0.991575337999935, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 37075 + }, + { + "epoch": 0.37076, + "grad_norm": 1.3850024921575828, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 37076 + }, + { + "epoch": 0.37077, + "grad_norm": 1.0869791408249974, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 37077 + }, + { + "epoch": 0.37078, + "grad_norm": 1.1795451862175463, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 37078 + }, + { + "epoch": 0.37079, + "grad_norm": 0.9140202236588096, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 37079 + }, + { + "epoch": 0.3708, + "grad_norm": 1.015436404591647, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 37080 + }, + { + "epoch": 0.37081, + "grad_norm": 1.0951555474719061, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 37081 + }, + { + "epoch": 0.37082, + "grad_norm": 0.9636788145488028, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 37082 + }, + { + "epoch": 0.37083, + "grad_norm": 0.9604923863130896, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 37083 + }, + { + "epoch": 0.37084, + "grad_norm": 1.1596767372642278, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 37084 + }, + { + "epoch": 0.37085, + "grad_norm": 1.0897373965187624, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 37085 + }, + { + "epoch": 0.37086, + "grad_norm": 1.417586323927644, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 37086 + }, + { + "epoch": 0.37087, + "grad_norm": 0.9880549434986805, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 37087 + }, + { + "epoch": 0.37088, + "grad_norm": 1.2167472817095553, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 37088 + }, + { + "epoch": 0.37089, + "grad_norm": 1.2281841545983438, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 37089 + }, + { + "epoch": 0.3709, + "grad_norm": 1.165679081924488, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 37090 + }, + { + "epoch": 0.37091, + "grad_norm": 0.9464035702592727, + "learning_rate": 0.003, + "loss": 4.032, + "step": 37091 + }, + { + "epoch": 0.37092, + "grad_norm": 1.1230951505586422, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 37092 + }, + { + "epoch": 0.37093, + "grad_norm": 1.194174029304608, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 37093 + }, + { + "epoch": 0.37094, + "grad_norm": 0.9150640066893434, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 37094 + }, + { + "epoch": 0.37095, + "grad_norm": 1.0672776457525963, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 37095 + }, + { + "epoch": 0.37096, + "grad_norm": 1.12041378183744, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 37096 + }, + { + "epoch": 0.37097, + "grad_norm": 1.002518170963861, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 37097 + }, + { + "epoch": 0.37098, + "grad_norm": 1.1421123265791793, + "learning_rate": 0.003, + "loss": 4.0845, + "step": 37098 + }, + { + "epoch": 0.37099, + "grad_norm": 0.9038906969548965, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 37099 + }, + { + "epoch": 0.371, + "grad_norm": 0.9855785327044623, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 37100 + }, + { + "epoch": 0.37101, + "grad_norm": 1.1677585236737937, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 37101 + }, + { + "epoch": 0.37102, + "grad_norm": 0.908484401625296, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 37102 + }, + { + "epoch": 0.37103, + "grad_norm": 0.8826441294858622, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 37103 + }, + { + "epoch": 0.37104, + "grad_norm": 0.8833314779325786, + "learning_rate": 0.003, + "loss": 4.054, + "step": 37104 + }, + { + "epoch": 0.37105, + "grad_norm": 0.8894792636139226, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 37105 + }, + { + "epoch": 0.37106, + "grad_norm": 1.0437205554329478, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 37106 + }, + { + "epoch": 0.37107, + "grad_norm": 1.0142371772614305, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 37107 + }, + { + "epoch": 0.37108, + "grad_norm": 1.0455311899923316, + "learning_rate": 0.003, + "loss": 4.042, + "step": 37108 + }, + { + "epoch": 0.37109, + "grad_norm": 1.0670232173303167, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 37109 + }, + { + "epoch": 0.3711, + "grad_norm": 1.0810177017871088, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 37110 + }, + { + "epoch": 0.37111, + "grad_norm": 1.1772854165143691, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 37111 + }, + { + "epoch": 0.37112, + "grad_norm": 1.3970241944314632, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 37112 + }, + { + "epoch": 0.37113, + "grad_norm": 0.868799557962799, + "learning_rate": 0.003, + "loss": 4.066, + "step": 37113 + }, + { + "epoch": 0.37114, + "grad_norm": 0.8544017707936274, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 37114 + }, + { + "epoch": 0.37115, + "grad_norm": 0.851889730429501, + "learning_rate": 0.003, + "loss": 4.039, + "step": 37115 + }, + { + "epoch": 0.37116, + "grad_norm": 1.0019591666840186, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 37116 + }, + { + "epoch": 0.37117, + "grad_norm": 1.320429039905215, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 37117 + }, + { + "epoch": 0.37118, + "grad_norm": 0.8801187694084649, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 37118 + }, + { + "epoch": 0.37119, + "grad_norm": 1.1847363299036315, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 37119 + }, + { + "epoch": 0.3712, + "grad_norm": 0.9757434508455685, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 37120 + }, + { + "epoch": 0.37121, + "grad_norm": 1.3521027930117349, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 37121 + }, + { + "epoch": 0.37122, + "grad_norm": 1.229689818956991, + "learning_rate": 0.003, + "loss": 4.04, + "step": 37122 + }, + { + "epoch": 0.37123, + "grad_norm": 1.0281584805761583, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 37123 + }, + { + "epoch": 0.37124, + "grad_norm": 1.4805420719978093, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 37124 + }, + { + "epoch": 0.37125, + "grad_norm": 1.0199914927873912, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 37125 + }, + { + "epoch": 0.37126, + "grad_norm": 1.2849844862615125, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 37126 + }, + { + "epoch": 0.37127, + "grad_norm": 0.8255079855691503, + "learning_rate": 0.003, + "loss": 4.028, + "step": 37127 + }, + { + "epoch": 0.37128, + "grad_norm": 0.9720988217919602, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 37128 + }, + { + "epoch": 0.37129, + "grad_norm": 1.170200570473301, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 37129 + }, + { + "epoch": 0.3713, + "grad_norm": 1.0905386953787766, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 37130 + }, + { + "epoch": 0.37131, + "grad_norm": 1.2782121681530592, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 37131 + }, + { + "epoch": 0.37132, + "grad_norm": 1.005775701084189, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 37132 + }, + { + "epoch": 0.37133, + "grad_norm": 1.2134207161961006, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 37133 + }, + { + "epoch": 0.37134, + "grad_norm": 1.072126214010269, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 37134 + }, + { + "epoch": 0.37135, + "grad_norm": 0.9943999798263725, + "learning_rate": 0.003, + "loss": 4.038, + "step": 37135 + }, + { + "epoch": 0.37136, + "grad_norm": 1.1348483539230176, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 37136 + }, + { + "epoch": 0.37137, + "grad_norm": 0.8893704898847425, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 37137 + }, + { + "epoch": 0.37138, + "grad_norm": 1.139169464124516, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 37138 + }, + { + "epoch": 0.37139, + "grad_norm": 1.1831029419889205, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 37139 + }, + { + "epoch": 0.3714, + "grad_norm": 0.9584842642975469, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 37140 + }, + { + "epoch": 0.37141, + "grad_norm": 1.0750807450813291, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 37141 + }, + { + "epoch": 0.37142, + "grad_norm": 1.1827800408055547, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 37142 + }, + { + "epoch": 0.37143, + "grad_norm": 0.8680969877170456, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 37143 + }, + { + "epoch": 0.37144, + "grad_norm": 0.83372003207101, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 37144 + }, + { + "epoch": 0.37145, + "grad_norm": 1.0278815044275451, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 37145 + }, + { + "epoch": 0.37146, + "grad_norm": 1.3509878537896742, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 37146 + }, + { + "epoch": 0.37147, + "grad_norm": 0.936691314196718, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 37147 + }, + { + "epoch": 0.37148, + "grad_norm": 1.0965468866348338, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 37148 + }, + { + "epoch": 0.37149, + "grad_norm": 1.1068324858671328, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 37149 + }, + { + "epoch": 0.3715, + "grad_norm": 0.8424301042997818, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 37150 + }, + { + "epoch": 0.37151, + "grad_norm": 1.092991925528588, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 37151 + }, + { + "epoch": 0.37152, + "grad_norm": 1.059699230756336, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 37152 + }, + { + "epoch": 0.37153, + "grad_norm": 1.0796771605985553, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 37153 + }, + { + "epoch": 0.37154, + "grad_norm": 1.054560888573632, + "learning_rate": 0.003, + "loss": 4.039, + "step": 37154 + }, + { + "epoch": 0.37155, + "grad_norm": 1.1142065549468803, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 37155 + }, + { + "epoch": 0.37156, + "grad_norm": 0.9346066056374402, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 37156 + }, + { + "epoch": 0.37157, + "grad_norm": 1.329991531717086, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 37157 + }, + { + "epoch": 0.37158, + "grad_norm": 0.9432340066028426, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 37158 + }, + { + "epoch": 0.37159, + "grad_norm": 0.9696955432022503, + "learning_rate": 0.003, + "loss": 4.03, + "step": 37159 + }, + { + "epoch": 0.3716, + "grad_norm": 1.4181803074551371, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 37160 + }, + { + "epoch": 0.37161, + "grad_norm": 1.0953173484745318, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 37161 + }, + { + "epoch": 0.37162, + "grad_norm": 1.2341735847315494, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 37162 + }, + { + "epoch": 0.37163, + "grad_norm": 1.062896338712121, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 37163 + }, + { + "epoch": 0.37164, + "grad_norm": 1.2142129070040066, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 37164 + }, + { + "epoch": 0.37165, + "grad_norm": 1.024883972891802, + "learning_rate": 0.003, + "loss": 4.061, + "step": 37165 + }, + { + "epoch": 0.37166, + "grad_norm": 1.117693011737729, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 37166 + }, + { + "epoch": 0.37167, + "grad_norm": 1.0716390059893832, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 37167 + }, + { + "epoch": 0.37168, + "grad_norm": 1.0685148631332335, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 37168 + }, + { + "epoch": 0.37169, + "grad_norm": 1.1539267376412956, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 37169 + }, + { + "epoch": 0.3717, + "grad_norm": 1.3021248012042115, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 37170 + }, + { + "epoch": 0.37171, + "grad_norm": 1.1676633228479099, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 37171 + }, + { + "epoch": 0.37172, + "grad_norm": 1.174428090176443, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 37172 + }, + { + "epoch": 0.37173, + "grad_norm": 0.9303712092082048, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 37173 + }, + { + "epoch": 0.37174, + "grad_norm": 1.0478801675959861, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 37174 + }, + { + "epoch": 0.37175, + "grad_norm": 1.0870751364387785, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 37175 + }, + { + "epoch": 0.37176, + "grad_norm": 1.1156382974730559, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 37176 + }, + { + "epoch": 0.37177, + "grad_norm": 1.3466072991482718, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 37177 + }, + { + "epoch": 0.37178, + "grad_norm": 1.0404381794721658, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 37178 + }, + { + "epoch": 0.37179, + "grad_norm": 1.0355460327404642, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 37179 + }, + { + "epoch": 0.3718, + "grad_norm": 1.2343598115866328, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 37180 + }, + { + "epoch": 0.37181, + "grad_norm": 1.1246001072849645, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 37181 + }, + { + "epoch": 0.37182, + "grad_norm": 0.8919319699578974, + "learning_rate": 0.003, + "loss": 4.023, + "step": 37182 + }, + { + "epoch": 0.37183, + "grad_norm": 0.8156721201947141, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 37183 + }, + { + "epoch": 0.37184, + "grad_norm": 0.838080175566527, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 37184 + }, + { + "epoch": 0.37185, + "grad_norm": 0.8993822103708191, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 37185 + }, + { + "epoch": 0.37186, + "grad_norm": 0.9879013158375289, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 37186 + }, + { + "epoch": 0.37187, + "grad_norm": 1.1856557649500212, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 37187 + }, + { + "epoch": 0.37188, + "grad_norm": 1.019285791869121, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 37188 + }, + { + "epoch": 0.37189, + "grad_norm": 1.1295053267758182, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 37189 + }, + { + "epoch": 0.3719, + "grad_norm": 1.139247059450531, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 37190 + }, + { + "epoch": 0.37191, + "grad_norm": 1.2587120004962782, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 37191 + }, + { + "epoch": 0.37192, + "grad_norm": 0.9700462961227266, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 37192 + }, + { + "epoch": 0.37193, + "grad_norm": 1.0697238447538875, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 37193 + }, + { + "epoch": 0.37194, + "grad_norm": 1.2135813749741984, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 37194 + }, + { + "epoch": 0.37195, + "grad_norm": 0.994953985433186, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 37195 + }, + { + "epoch": 0.37196, + "grad_norm": 1.2545456417925873, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 37196 + }, + { + "epoch": 0.37197, + "grad_norm": 0.938545596555212, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 37197 + }, + { + "epoch": 0.37198, + "grad_norm": 1.0704756545742227, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 37198 + }, + { + "epoch": 0.37199, + "grad_norm": 1.1188990979246944, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 37199 + }, + { + "epoch": 0.372, + "grad_norm": 1.0717090562923643, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 37200 + }, + { + "epoch": 0.37201, + "grad_norm": 1.0400460350161904, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 37201 + }, + { + "epoch": 0.37202, + "grad_norm": 1.1942996226339904, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 37202 + }, + { + "epoch": 0.37203, + "grad_norm": 1.1049765453049085, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 37203 + }, + { + "epoch": 0.37204, + "grad_norm": 0.9176995107887879, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 37204 + }, + { + "epoch": 0.37205, + "grad_norm": 0.9506549288292506, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 37205 + }, + { + "epoch": 0.37206, + "grad_norm": 1.2227893433163959, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 37206 + }, + { + "epoch": 0.37207, + "grad_norm": 0.8659906698157105, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 37207 + }, + { + "epoch": 0.37208, + "grad_norm": 0.8036808020592672, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 37208 + }, + { + "epoch": 0.37209, + "grad_norm": 0.7930559417622611, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 37209 + }, + { + "epoch": 0.3721, + "grad_norm": 1.1213956570628516, + "learning_rate": 0.003, + "loss": 4.061, + "step": 37210 + }, + { + "epoch": 0.37211, + "grad_norm": 1.3036670807356994, + "learning_rate": 0.003, + "loss": 4.036, + "step": 37211 + }, + { + "epoch": 0.37212, + "grad_norm": 1.0153883956284089, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 37212 + }, + { + "epoch": 0.37213, + "grad_norm": 1.361619928144273, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 37213 + }, + { + "epoch": 0.37214, + "grad_norm": 0.9922304790905283, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 37214 + }, + { + "epoch": 0.37215, + "grad_norm": 0.9644676403961456, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 37215 + }, + { + "epoch": 0.37216, + "grad_norm": 0.9902177433576633, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 37216 + }, + { + "epoch": 0.37217, + "grad_norm": 1.3957651932270714, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 37217 + }, + { + "epoch": 0.37218, + "grad_norm": 1.2186832948214197, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 37218 + }, + { + "epoch": 0.37219, + "grad_norm": 0.9516261287432092, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 37219 + }, + { + "epoch": 0.3722, + "grad_norm": 1.0939614298746576, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 37220 + }, + { + "epoch": 0.37221, + "grad_norm": 1.1382132016730946, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 37221 + }, + { + "epoch": 0.37222, + "grad_norm": 1.057932062891032, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 37222 + }, + { + "epoch": 0.37223, + "grad_norm": 1.0719149682197855, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 37223 + }, + { + "epoch": 0.37224, + "grad_norm": 1.1608490137564582, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 37224 + }, + { + "epoch": 0.37225, + "grad_norm": 1.172927860148032, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 37225 + }, + { + "epoch": 0.37226, + "grad_norm": 0.968070059772194, + "learning_rate": 0.003, + "loss": 4.028, + "step": 37226 + }, + { + "epoch": 0.37227, + "grad_norm": 1.1281618170499363, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 37227 + }, + { + "epoch": 0.37228, + "grad_norm": 1.0794277567769772, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 37228 + }, + { + "epoch": 0.37229, + "grad_norm": 1.154037429440169, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 37229 + }, + { + "epoch": 0.3723, + "grad_norm": 1.0613217747788937, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 37230 + }, + { + "epoch": 0.37231, + "grad_norm": 1.3387009037997433, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 37231 + }, + { + "epoch": 0.37232, + "grad_norm": 0.9756410014863596, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 37232 + }, + { + "epoch": 0.37233, + "grad_norm": 0.8795042553772309, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 37233 + }, + { + "epoch": 0.37234, + "grad_norm": 0.9728144745201945, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 37234 + }, + { + "epoch": 0.37235, + "grad_norm": 1.380962451987385, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 37235 + }, + { + "epoch": 0.37236, + "grad_norm": 1.092982135862332, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 37236 + }, + { + "epoch": 0.37237, + "grad_norm": 1.0938272851940316, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 37237 + }, + { + "epoch": 0.37238, + "grad_norm": 1.0914358548374927, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 37238 + }, + { + "epoch": 0.37239, + "grad_norm": 1.00863095739212, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 37239 + }, + { + "epoch": 0.3724, + "grad_norm": 1.0321918427082646, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 37240 + }, + { + "epoch": 0.37241, + "grad_norm": 1.1110194213662576, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 37241 + }, + { + "epoch": 0.37242, + "grad_norm": 1.0976704988264838, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 37242 + }, + { + "epoch": 0.37243, + "grad_norm": 1.029546477511332, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 37243 + }, + { + "epoch": 0.37244, + "grad_norm": 0.9931622160918218, + "learning_rate": 0.003, + "loss": 4.036, + "step": 37244 + }, + { + "epoch": 0.37245, + "grad_norm": 1.1389888122826202, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 37245 + }, + { + "epoch": 0.37246, + "grad_norm": 1.0085065402127955, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 37246 + }, + { + "epoch": 0.37247, + "grad_norm": 1.1806466446638424, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 37247 + }, + { + "epoch": 0.37248, + "grad_norm": 1.0502741414471684, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 37248 + }, + { + "epoch": 0.37249, + "grad_norm": 0.9771159337190254, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 37249 + }, + { + "epoch": 0.3725, + "grad_norm": 1.2341116606488316, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 37250 + }, + { + "epoch": 0.37251, + "grad_norm": 1.0254762538873026, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 37251 + }, + { + "epoch": 0.37252, + "grad_norm": 1.1517595131188423, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 37252 + }, + { + "epoch": 0.37253, + "grad_norm": 1.393640665916975, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 37253 + }, + { + "epoch": 0.37254, + "grad_norm": 0.8366726341754787, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 37254 + }, + { + "epoch": 0.37255, + "grad_norm": 0.8977156327010546, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 37255 + }, + { + "epoch": 0.37256, + "grad_norm": 1.2613245168178393, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 37256 + }, + { + "epoch": 0.37257, + "grad_norm": 1.2590296765877458, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 37257 + }, + { + "epoch": 0.37258, + "grad_norm": 0.9715731691973885, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 37258 + }, + { + "epoch": 0.37259, + "grad_norm": 1.1844535526001083, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 37259 + }, + { + "epoch": 0.3726, + "grad_norm": 1.0110434372597905, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 37260 + }, + { + "epoch": 0.37261, + "grad_norm": 1.0216107153312328, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 37261 + }, + { + "epoch": 0.37262, + "grad_norm": 1.0690209602648653, + "learning_rate": 0.003, + "loss": 4.044, + "step": 37262 + }, + { + "epoch": 0.37263, + "grad_norm": 1.3712848560373367, + "learning_rate": 0.003, + "loss": 4.06, + "step": 37263 + }, + { + "epoch": 0.37264, + "grad_norm": 0.9355237620568017, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 37264 + }, + { + "epoch": 0.37265, + "grad_norm": 0.973834991536279, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 37265 + }, + { + "epoch": 0.37266, + "grad_norm": 1.111979527665987, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 37266 + }, + { + "epoch": 0.37267, + "grad_norm": 0.9577731849281126, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 37267 + }, + { + "epoch": 0.37268, + "grad_norm": 1.1486103159303696, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 37268 + }, + { + "epoch": 0.37269, + "grad_norm": 0.9283847735854027, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 37269 + }, + { + "epoch": 0.3727, + "grad_norm": 1.0294962547694688, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 37270 + }, + { + "epoch": 0.37271, + "grad_norm": 1.3503198714084323, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 37271 + }, + { + "epoch": 0.37272, + "grad_norm": 1.1193941957723679, + "learning_rate": 0.003, + "loss": 4.016, + "step": 37272 + }, + { + "epoch": 0.37273, + "grad_norm": 0.8988030107457338, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 37273 + }, + { + "epoch": 0.37274, + "grad_norm": 0.9190597358937553, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 37274 + }, + { + "epoch": 0.37275, + "grad_norm": 0.8935240385053601, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 37275 + }, + { + "epoch": 0.37276, + "grad_norm": 1.132299849811687, + "learning_rate": 0.003, + "loss": 4.062, + "step": 37276 + }, + { + "epoch": 0.37277, + "grad_norm": 1.01173972538598, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 37277 + }, + { + "epoch": 0.37278, + "grad_norm": 1.0563887287721827, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 37278 + }, + { + "epoch": 0.37279, + "grad_norm": 1.0989654492424707, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 37279 + }, + { + "epoch": 0.3728, + "grad_norm": 1.10072905525729, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 37280 + }, + { + "epoch": 0.37281, + "grad_norm": 1.070449161561095, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 37281 + }, + { + "epoch": 0.37282, + "grad_norm": 1.1502277493181412, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 37282 + }, + { + "epoch": 0.37283, + "grad_norm": 1.0643707879189568, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 37283 + }, + { + "epoch": 0.37284, + "grad_norm": 0.9653951596246818, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 37284 + }, + { + "epoch": 0.37285, + "grad_norm": 0.8673005668224193, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 37285 + }, + { + "epoch": 0.37286, + "grad_norm": 1.1523961236715752, + "learning_rate": 0.003, + "loss": 4.037, + "step": 37286 + }, + { + "epoch": 0.37287, + "grad_norm": 1.282522912072688, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 37287 + }, + { + "epoch": 0.37288, + "grad_norm": 1.2206808659069663, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 37288 + }, + { + "epoch": 0.37289, + "grad_norm": 1.0143426060270393, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 37289 + }, + { + "epoch": 0.3729, + "grad_norm": 1.1905080309947813, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 37290 + }, + { + "epoch": 0.37291, + "grad_norm": 1.0221670062173356, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 37291 + }, + { + "epoch": 0.37292, + "grad_norm": 1.20033261389089, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 37292 + }, + { + "epoch": 0.37293, + "grad_norm": 1.1353918485352716, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 37293 + }, + { + "epoch": 0.37294, + "grad_norm": 0.9935704412302636, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 37294 + }, + { + "epoch": 0.37295, + "grad_norm": 1.214712780838503, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 37295 + }, + { + "epoch": 0.37296, + "grad_norm": 0.907465414286833, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 37296 + }, + { + "epoch": 0.37297, + "grad_norm": 1.041977146957818, + "learning_rate": 0.003, + "loss": 4.013, + "step": 37297 + }, + { + "epoch": 0.37298, + "grad_norm": 1.066787751118034, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 37298 + }, + { + "epoch": 0.37299, + "grad_norm": 0.9740228900059867, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 37299 + }, + { + "epoch": 0.373, + "grad_norm": 1.0425308499867918, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 37300 + }, + { + "epoch": 0.37301, + "grad_norm": 1.1785500648237122, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 37301 + }, + { + "epoch": 0.37302, + "grad_norm": 1.0773943582715428, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 37302 + }, + { + "epoch": 0.37303, + "grad_norm": 1.1972383099160637, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 37303 + }, + { + "epoch": 0.37304, + "grad_norm": 1.1292024704025194, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 37304 + }, + { + "epoch": 0.37305, + "grad_norm": 1.1005395095375083, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 37305 + }, + { + "epoch": 0.37306, + "grad_norm": 1.1851914015076739, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 37306 + }, + { + "epoch": 0.37307, + "grad_norm": 0.8584657589807831, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 37307 + }, + { + "epoch": 0.37308, + "grad_norm": 0.8843009572229519, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 37308 + }, + { + "epoch": 0.37309, + "grad_norm": 1.13374971795011, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 37309 + }, + { + "epoch": 0.3731, + "grad_norm": 1.3257693853141181, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 37310 + }, + { + "epoch": 0.37311, + "grad_norm": 0.9988413832527319, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 37311 + }, + { + "epoch": 0.37312, + "grad_norm": 1.149368375587091, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 37312 + }, + { + "epoch": 0.37313, + "grad_norm": 0.9083378029602222, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 37313 + }, + { + "epoch": 0.37314, + "grad_norm": 1.0617375564116869, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 37314 + }, + { + "epoch": 0.37315, + "grad_norm": 1.312117663476405, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 37315 + }, + { + "epoch": 0.37316, + "grad_norm": 0.8724891518404734, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 37316 + }, + { + "epoch": 0.37317, + "grad_norm": 0.8825642375161291, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 37317 + }, + { + "epoch": 0.37318, + "grad_norm": 0.9929496134206991, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 37318 + }, + { + "epoch": 0.37319, + "grad_norm": 1.2092061658430004, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 37319 + }, + { + "epoch": 0.3732, + "grad_norm": 0.9996941375337175, + "learning_rate": 0.003, + "loss": 4.033, + "step": 37320 + }, + { + "epoch": 0.37321, + "grad_norm": 1.1917618901251037, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 37321 + }, + { + "epoch": 0.37322, + "grad_norm": 1.0757656076438336, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 37322 + }, + { + "epoch": 0.37323, + "grad_norm": 1.085733927539735, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 37323 + }, + { + "epoch": 0.37324, + "grad_norm": 1.2116123504742642, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 37324 + }, + { + "epoch": 0.37325, + "grad_norm": 1.0443451361060747, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 37325 + }, + { + "epoch": 0.37326, + "grad_norm": 1.2004745048328425, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 37326 + }, + { + "epoch": 0.37327, + "grad_norm": 1.0429296127426464, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 37327 + }, + { + "epoch": 0.37328, + "grad_norm": 1.2533320091191917, + "learning_rate": 0.003, + "loss": 4.036, + "step": 37328 + }, + { + "epoch": 0.37329, + "grad_norm": 0.7823929570630502, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 37329 + }, + { + "epoch": 0.3733, + "grad_norm": 0.8345319951725235, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 37330 + }, + { + "epoch": 0.37331, + "grad_norm": 1.1115850856428933, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 37331 + }, + { + "epoch": 0.37332, + "grad_norm": 1.0370760405924055, + "learning_rate": 0.003, + "loss": 4.051, + "step": 37332 + }, + { + "epoch": 0.37333, + "grad_norm": 1.2470730780548966, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 37333 + }, + { + "epoch": 0.37334, + "grad_norm": 1.1837355012498643, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 37334 + }, + { + "epoch": 0.37335, + "grad_norm": 0.9526557824139935, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 37335 + }, + { + "epoch": 0.37336, + "grad_norm": 1.0657917734515165, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 37336 + }, + { + "epoch": 0.37337, + "grad_norm": 1.0981560329159723, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 37337 + }, + { + "epoch": 0.37338, + "grad_norm": 1.0337515812142029, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 37338 + }, + { + "epoch": 0.37339, + "grad_norm": 1.0229172652208052, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 37339 + }, + { + "epoch": 0.3734, + "grad_norm": 1.0018944842516244, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 37340 + }, + { + "epoch": 0.37341, + "grad_norm": 1.15207062333167, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 37341 + }, + { + "epoch": 0.37342, + "grad_norm": 1.0822702155588577, + "learning_rate": 0.003, + "loss": 4.022, + "step": 37342 + }, + { + "epoch": 0.37343, + "grad_norm": 1.24519268029663, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 37343 + }, + { + "epoch": 0.37344, + "grad_norm": 0.8872273942030457, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 37344 + }, + { + "epoch": 0.37345, + "grad_norm": 1.0056229522183573, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 37345 + }, + { + "epoch": 0.37346, + "grad_norm": 1.110385493237873, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 37346 + }, + { + "epoch": 0.37347, + "grad_norm": 0.8416245862871874, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 37347 + }, + { + "epoch": 0.37348, + "grad_norm": 0.840519993063102, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 37348 + }, + { + "epoch": 0.37349, + "grad_norm": 0.9729884414582329, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 37349 + }, + { + "epoch": 0.3735, + "grad_norm": 1.3626012393210107, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 37350 + }, + { + "epoch": 0.37351, + "grad_norm": 1.1737266633312273, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 37351 + }, + { + "epoch": 0.37352, + "grad_norm": 1.2137756463035705, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 37352 + }, + { + "epoch": 0.37353, + "grad_norm": 0.973471472101587, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 37353 + }, + { + "epoch": 0.37354, + "grad_norm": 1.197439203907825, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 37354 + }, + { + "epoch": 0.37355, + "grad_norm": 0.984056488368164, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 37355 + }, + { + "epoch": 0.37356, + "grad_norm": 0.9681561923356437, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 37356 + }, + { + "epoch": 0.37357, + "grad_norm": 1.0767597641148499, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 37357 + }, + { + "epoch": 0.37358, + "grad_norm": 1.2056572634528202, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 37358 + }, + { + "epoch": 0.37359, + "grad_norm": 1.251049181363899, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 37359 + }, + { + "epoch": 0.3736, + "grad_norm": 0.7585358760189505, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 37360 + }, + { + "epoch": 0.37361, + "grad_norm": 0.8345861809187424, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 37361 + }, + { + "epoch": 0.37362, + "grad_norm": 0.9072747293794283, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 37362 + }, + { + "epoch": 0.37363, + "grad_norm": 1.4137323483178577, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 37363 + }, + { + "epoch": 0.37364, + "grad_norm": 1.0041274077133, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 37364 + }, + { + "epoch": 0.37365, + "grad_norm": 1.1831626491533336, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 37365 + }, + { + "epoch": 0.37366, + "grad_norm": 1.0825388225693593, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 37366 + }, + { + "epoch": 0.37367, + "grad_norm": 0.944359619587197, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 37367 + }, + { + "epoch": 0.37368, + "grad_norm": 1.1033768374694217, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 37368 + }, + { + "epoch": 0.37369, + "grad_norm": 0.9379304346809966, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 37369 + }, + { + "epoch": 0.3737, + "grad_norm": 1.1928543000267477, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 37370 + }, + { + "epoch": 0.37371, + "grad_norm": 1.2275727256664213, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 37371 + }, + { + "epoch": 0.37372, + "grad_norm": 1.3491957125052594, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 37372 + }, + { + "epoch": 0.37373, + "grad_norm": 1.1133709091380317, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 37373 + }, + { + "epoch": 0.37374, + "grad_norm": 1.0299799691309026, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 37374 + }, + { + "epoch": 0.37375, + "grad_norm": 1.1894865421453418, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 37375 + }, + { + "epoch": 0.37376, + "grad_norm": 0.9646879282204593, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 37376 + }, + { + "epoch": 0.37377, + "grad_norm": 1.2414847420178914, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 37377 + }, + { + "epoch": 0.37378, + "grad_norm": 0.830219428826606, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 37378 + }, + { + "epoch": 0.37379, + "grad_norm": 0.8887265165413628, + "learning_rate": 0.003, + "loss": 4.019, + "step": 37379 + }, + { + "epoch": 0.3738, + "grad_norm": 1.0665647791562145, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 37380 + }, + { + "epoch": 0.37381, + "grad_norm": 1.1181425648616325, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 37381 + }, + { + "epoch": 0.37382, + "grad_norm": 1.4068621284244334, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 37382 + }, + { + "epoch": 0.37383, + "grad_norm": 0.9519255460604087, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 37383 + }, + { + "epoch": 0.37384, + "grad_norm": 0.9190025420478093, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 37384 + }, + { + "epoch": 0.37385, + "grad_norm": 1.0072844766517572, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 37385 + }, + { + "epoch": 0.37386, + "grad_norm": 1.2773972959976148, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 37386 + }, + { + "epoch": 0.37387, + "grad_norm": 0.8831267727815932, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 37387 + }, + { + "epoch": 0.37388, + "grad_norm": 0.8925776984547468, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 37388 + }, + { + "epoch": 0.37389, + "grad_norm": 1.1227892571692075, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 37389 + }, + { + "epoch": 0.3739, + "grad_norm": 1.5038524192689962, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 37390 + }, + { + "epoch": 0.37391, + "grad_norm": 1.0286766924374922, + "learning_rate": 0.003, + "loss": 4.038, + "step": 37391 + }, + { + "epoch": 0.37392, + "grad_norm": 0.9123146009032342, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 37392 + }, + { + "epoch": 0.37393, + "grad_norm": 1.0849522519069283, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 37393 + }, + { + "epoch": 0.37394, + "grad_norm": 1.1420604419057143, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 37394 + }, + { + "epoch": 0.37395, + "grad_norm": 1.3443468768874225, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 37395 + }, + { + "epoch": 0.37396, + "grad_norm": 1.0274200434608665, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 37396 + }, + { + "epoch": 0.37397, + "grad_norm": 1.6402538968903224, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 37397 + }, + { + "epoch": 0.37398, + "grad_norm": 0.9289239947933058, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 37398 + }, + { + "epoch": 0.37399, + "grad_norm": 0.8568455300867168, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 37399 + }, + { + "epoch": 0.374, + "grad_norm": 1.16939290186867, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 37400 + }, + { + "epoch": 0.37401, + "grad_norm": 1.0900188675490987, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 37401 + }, + { + "epoch": 0.37402, + "grad_norm": 1.0909061232245756, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 37402 + }, + { + "epoch": 0.37403, + "grad_norm": 1.2104918543572663, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 37403 + }, + { + "epoch": 0.37404, + "grad_norm": 1.0897293084951507, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 37404 + }, + { + "epoch": 0.37405, + "grad_norm": 1.0980871963967536, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 37405 + }, + { + "epoch": 0.37406, + "grad_norm": 1.1049977927811543, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 37406 + }, + { + "epoch": 0.37407, + "grad_norm": 1.0528613459703249, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 37407 + }, + { + "epoch": 0.37408, + "grad_norm": 1.1275559202911125, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 37408 + }, + { + "epoch": 0.37409, + "grad_norm": 0.8569663344240197, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 37409 + }, + { + "epoch": 0.3741, + "grad_norm": 0.9884985462399584, + "learning_rate": 0.003, + "loss": 4.038, + "step": 37410 + }, + { + "epoch": 0.37411, + "grad_norm": 1.2786919838447184, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 37411 + }, + { + "epoch": 0.37412, + "grad_norm": 1.1702246208901075, + "learning_rate": 0.003, + "loss": 4.024, + "step": 37412 + }, + { + "epoch": 0.37413, + "grad_norm": 1.0523378824194476, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 37413 + }, + { + "epoch": 0.37414, + "grad_norm": 1.1283898512916177, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 37414 + }, + { + "epoch": 0.37415, + "grad_norm": 0.8829210054168882, + "learning_rate": 0.003, + "loss": 4.039, + "step": 37415 + }, + { + "epoch": 0.37416, + "grad_norm": 0.9512914154921638, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 37416 + }, + { + "epoch": 0.37417, + "grad_norm": 1.167835558218107, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 37417 + }, + { + "epoch": 0.37418, + "grad_norm": 1.1126013980796352, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 37418 + }, + { + "epoch": 0.37419, + "grad_norm": 1.0876817771741314, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 37419 + }, + { + "epoch": 0.3742, + "grad_norm": 1.1013556004273675, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 37420 + }, + { + "epoch": 0.37421, + "grad_norm": 1.13591110102086, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 37421 + }, + { + "epoch": 0.37422, + "grad_norm": 1.197406323783045, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 37422 + }, + { + "epoch": 0.37423, + "grad_norm": 0.7923610416405894, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 37423 + }, + { + "epoch": 0.37424, + "grad_norm": 0.7084831790316342, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 37424 + }, + { + "epoch": 0.37425, + "grad_norm": 0.8500591525519157, + "learning_rate": 0.003, + "loss": 4.048, + "step": 37425 + }, + { + "epoch": 0.37426, + "grad_norm": 1.0829835314867426, + "learning_rate": 0.003, + "loss": 4.056, + "step": 37426 + }, + { + "epoch": 0.37427, + "grad_norm": 1.1310672336139476, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 37427 + }, + { + "epoch": 0.37428, + "grad_norm": 1.0864075965236297, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 37428 + }, + { + "epoch": 0.37429, + "grad_norm": 1.068184461343839, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 37429 + }, + { + "epoch": 0.3743, + "grad_norm": 1.0002077090244517, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 37430 + }, + { + "epoch": 0.37431, + "grad_norm": 1.2313902850283238, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 37431 + }, + { + "epoch": 0.37432, + "grad_norm": 1.1338217358741847, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 37432 + }, + { + "epoch": 0.37433, + "grad_norm": 1.4954963977842795, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 37433 + }, + { + "epoch": 0.37434, + "grad_norm": 1.0148952591327733, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 37434 + }, + { + "epoch": 0.37435, + "grad_norm": 1.1715690036038995, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 37435 + }, + { + "epoch": 0.37436, + "grad_norm": 0.9375063708350094, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 37436 + }, + { + "epoch": 0.37437, + "grad_norm": 0.9962600789962562, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 37437 + }, + { + "epoch": 0.37438, + "grad_norm": 1.199071710380091, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 37438 + }, + { + "epoch": 0.37439, + "grad_norm": 0.9604135308739917, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 37439 + }, + { + "epoch": 0.3744, + "grad_norm": 1.0066048573804078, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 37440 + }, + { + "epoch": 0.37441, + "grad_norm": 1.0359545645736723, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 37441 + }, + { + "epoch": 0.37442, + "grad_norm": 0.9990109809891465, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 37442 + }, + { + "epoch": 0.37443, + "grad_norm": 1.1458783092031948, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 37443 + }, + { + "epoch": 0.37444, + "grad_norm": 1.2724674806818328, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 37444 + }, + { + "epoch": 0.37445, + "grad_norm": 1.2507978741799126, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 37445 + }, + { + "epoch": 0.37446, + "grad_norm": 1.0858444674167975, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 37446 + }, + { + "epoch": 0.37447, + "grad_norm": 1.2094177820810115, + "learning_rate": 0.003, + "loss": 4.036, + "step": 37447 + }, + { + "epoch": 0.37448, + "grad_norm": 0.8918273938071531, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 37448 + }, + { + "epoch": 0.37449, + "grad_norm": 0.8470943904965522, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 37449 + }, + { + "epoch": 0.3745, + "grad_norm": 1.0480382922606106, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 37450 + }, + { + "epoch": 0.37451, + "grad_norm": 1.4503861190813057, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 37451 + }, + { + "epoch": 0.37452, + "grad_norm": 0.821701925153018, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 37452 + }, + { + "epoch": 0.37453, + "grad_norm": 0.8839868932729024, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 37453 + }, + { + "epoch": 0.37454, + "grad_norm": 1.0863834019080247, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 37454 + }, + { + "epoch": 0.37455, + "grad_norm": 1.3251949016791997, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 37455 + }, + { + "epoch": 0.37456, + "grad_norm": 0.727046810360347, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 37456 + }, + { + "epoch": 0.37457, + "grad_norm": 0.8412246177249084, + "learning_rate": 0.003, + "loss": 4.024, + "step": 37457 + }, + { + "epoch": 0.37458, + "grad_norm": 1.1118887428673325, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 37458 + }, + { + "epoch": 0.37459, + "grad_norm": 1.1935992284698689, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 37459 + }, + { + "epoch": 0.3746, + "grad_norm": 1.0464912683254182, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 37460 + }, + { + "epoch": 0.37461, + "grad_norm": 1.0246376598552427, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 37461 + }, + { + "epoch": 0.37462, + "grad_norm": 1.0238031436137973, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 37462 + }, + { + "epoch": 0.37463, + "grad_norm": 1.298460343562505, + "learning_rate": 0.003, + "loss": 4.038, + "step": 37463 + }, + { + "epoch": 0.37464, + "grad_norm": 1.0338288896345076, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 37464 + }, + { + "epoch": 0.37465, + "grad_norm": 1.1728229855733356, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 37465 + }, + { + "epoch": 0.37466, + "grad_norm": 1.143682895651975, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 37466 + }, + { + "epoch": 0.37467, + "grad_norm": 1.1360040569012584, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 37467 + }, + { + "epoch": 0.37468, + "grad_norm": 1.0400398966865139, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 37468 + }, + { + "epoch": 0.37469, + "grad_norm": 1.1123723089071094, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 37469 + }, + { + "epoch": 0.3747, + "grad_norm": 1.0624326145138003, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 37470 + }, + { + "epoch": 0.37471, + "grad_norm": 1.0836596004653907, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 37471 + }, + { + "epoch": 0.37472, + "grad_norm": 1.0345380305386718, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 37472 + }, + { + "epoch": 0.37473, + "grad_norm": 1.0369441317734214, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 37473 + }, + { + "epoch": 0.37474, + "grad_norm": 1.1051842692151423, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 37474 + }, + { + "epoch": 0.37475, + "grad_norm": 1.1529850597687235, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 37475 + }, + { + "epoch": 0.37476, + "grad_norm": 1.1172750688689834, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 37476 + }, + { + "epoch": 0.37477, + "grad_norm": 1.1113405183837815, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 37477 + }, + { + "epoch": 0.37478, + "grad_norm": 1.226137895369373, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 37478 + }, + { + "epoch": 0.37479, + "grad_norm": 0.9301991691219836, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 37479 + }, + { + "epoch": 0.3748, + "grad_norm": 1.068132269353788, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 37480 + }, + { + "epoch": 0.37481, + "grad_norm": 1.0805708795188897, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 37481 + }, + { + "epoch": 0.37482, + "grad_norm": 0.937885954057119, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 37482 + }, + { + "epoch": 0.37483, + "grad_norm": 1.1048462736649387, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 37483 + }, + { + "epoch": 0.37484, + "grad_norm": 1.08238939422432, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 37484 + }, + { + "epoch": 0.37485, + "grad_norm": 1.2571043132370363, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 37485 + }, + { + "epoch": 0.37486, + "grad_norm": 0.8580891816473895, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 37486 + }, + { + "epoch": 0.37487, + "grad_norm": 0.9387373730195551, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 37487 + }, + { + "epoch": 0.37488, + "grad_norm": 1.3720492074523332, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 37488 + }, + { + "epoch": 0.37489, + "grad_norm": 0.914863247987598, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 37489 + }, + { + "epoch": 0.3749, + "grad_norm": 0.9966056233009836, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 37490 + }, + { + "epoch": 0.37491, + "grad_norm": 0.9704475042883107, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 37491 + }, + { + "epoch": 0.37492, + "grad_norm": 0.9156200199842454, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 37492 + }, + { + "epoch": 0.37493, + "grad_norm": 1.1240776424901908, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 37493 + }, + { + "epoch": 0.37494, + "grad_norm": 1.0895395571361943, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 37494 + }, + { + "epoch": 0.37495, + "grad_norm": 1.3132046646865427, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 37495 + }, + { + "epoch": 0.37496, + "grad_norm": 1.0713373071942265, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 37496 + }, + { + "epoch": 0.37497, + "grad_norm": 0.9749550330462541, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 37497 + }, + { + "epoch": 0.37498, + "grad_norm": 1.059202306986532, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 37498 + }, + { + "epoch": 0.37499, + "grad_norm": 1.103308376621372, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 37499 + }, + { + "epoch": 0.375, + "grad_norm": 0.9926582751461216, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 37500 + }, + { + "epoch": 0.37501, + "grad_norm": 1.0133293782727988, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 37501 + }, + { + "epoch": 0.37502, + "grad_norm": 1.455263448162951, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 37502 + }, + { + "epoch": 0.37503, + "grad_norm": 0.9930918597730285, + "learning_rate": 0.003, + "loss": 4.0819, + "step": 37503 + }, + { + "epoch": 0.37504, + "grad_norm": 1.2332394600046197, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 37504 + }, + { + "epoch": 0.37505, + "grad_norm": 0.9814411330826899, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 37505 + }, + { + "epoch": 0.37506, + "grad_norm": 1.113396119883311, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 37506 + }, + { + "epoch": 0.37507, + "grad_norm": 1.1445117009513934, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 37507 + }, + { + "epoch": 0.37508, + "grad_norm": 1.2745920953050018, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 37508 + }, + { + "epoch": 0.37509, + "grad_norm": 0.9307630542934887, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 37509 + }, + { + "epoch": 0.3751, + "grad_norm": 0.8412549953207533, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 37510 + }, + { + "epoch": 0.37511, + "grad_norm": 1.0176363987182122, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 37511 + }, + { + "epoch": 0.37512, + "grad_norm": 1.253085361262804, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 37512 + }, + { + "epoch": 0.37513, + "grad_norm": 1.0284824799534757, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 37513 + }, + { + "epoch": 0.37514, + "grad_norm": 1.0201018947013154, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 37514 + }, + { + "epoch": 0.37515, + "grad_norm": 1.226554155326389, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 37515 + }, + { + "epoch": 0.37516, + "grad_norm": 1.0915565058518362, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 37516 + }, + { + "epoch": 0.37517, + "grad_norm": 0.9663107180427447, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 37517 + }, + { + "epoch": 0.37518, + "grad_norm": 1.1634122782372018, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 37518 + }, + { + "epoch": 0.37519, + "grad_norm": 1.253707000342034, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 37519 + }, + { + "epoch": 0.3752, + "grad_norm": 0.8467993482332051, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 37520 + }, + { + "epoch": 0.37521, + "grad_norm": 0.9031402976826207, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 37521 + }, + { + "epoch": 0.37522, + "grad_norm": 1.0764451246282671, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 37522 + }, + { + "epoch": 0.37523, + "grad_norm": 1.3096910725806212, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 37523 + }, + { + "epoch": 0.37524, + "grad_norm": 0.991528386567545, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 37524 + }, + { + "epoch": 0.37525, + "grad_norm": 1.0757597823044651, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 37525 + }, + { + "epoch": 0.37526, + "grad_norm": 0.9727175796818078, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 37526 + }, + { + "epoch": 0.37527, + "grad_norm": 1.0028156933472163, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 37527 + }, + { + "epoch": 0.37528, + "grad_norm": 1.1022837297307941, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 37528 + }, + { + "epoch": 0.37529, + "grad_norm": 0.9372778002248798, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 37529 + }, + { + "epoch": 0.3753, + "grad_norm": 1.1383795306353406, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 37530 + }, + { + "epoch": 0.37531, + "grad_norm": 1.0919333181960686, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 37531 + }, + { + "epoch": 0.37532, + "grad_norm": 0.9981802171015356, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 37532 + }, + { + "epoch": 0.37533, + "grad_norm": 1.1642215476932656, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 37533 + }, + { + "epoch": 0.37534, + "grad_norm": 1.1478025677683288, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 37534 + }, + { + "epoch": 0.37535, + "grad_norm": 1.2605327048281239, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 37535 + }, + { + "epoch": 0.37536, + "grad_norm": 1.0689374104684075, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 37536 + }, + { + "epoch": 0.37537, + "grad_norm": 1.3430899194319446, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 37537 + }, + { + "epoch": 0.37538, + "grad_norm": 0.8570200283733946, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 37538 + }, + { + "epoch": 0.37539, + "grad_norm": 0.6966895545660587, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 37539 + }, + { + "epoch": 0.3754, + "grad_norm": 0.782964418974471, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 37540 + }, + { + "epoch": 0.37541, + "grad_norm": 0.9620565459588075, + "learning_rate": 0.003, + "loss": 4.051, + "step": 37541 + }, + { + "epoch": 0.37542, + "grad_norm": 1.1447906010797204, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 37542 + }, + { + "epoch": 0.37543, + "grad_norm": 0.917691786172769, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 37543 + }, + { + "epoch": 0.37544, + "grad_norm": 1.2008772190392638, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 37544 + }, + { + "epoch": 0.37545, + "grad_norm": 1.1005274927788633, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 37545 + }, + { + "epoch": 0.37546, + "grad_norm": 1.2539403162047396, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 37546 + }, + { + "epoch": 0.37547, + "grad_norm": 1.1555551474543806, + "learning_rate": 0.003, + "loss": 4.055, + "step": 37547 + }, + { + "epoch": 0.37548, + "grad_norm": 1.0740833718634843, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 37548 + }, + { + "epoch": 0.37549, + "grad_norm": 0.9475263533299161, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 37549 + }, + { + "epoch": 0.3755, + "grad_norm": 0.8864232067856073, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 37550 + }, + { + "epoch": 0.37551, + "grad_norm": 1.0374949653756367, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 37551 + }, + { + "epoch": 0.37552, + "grad_norm": 1.3395055865468373, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 37552 + }, + { + "epoch": 0.37553, + "grad_norm": 0.9179347137786718, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 37553 + }, + { + "epoch": 0.37554, + "grad_norm": 1.1055552794756953, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 37554 + }, + { + "epoch": 0.37555, + "grad_norm": 0.8890702865050827, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 37555 + }, + { + "epoch": 0.37556, + "grad_norm": 1.0684543038803338, + "learning_rate": 0.003, + "loss": 4.04, + "step": 37556 + }, + { + "epoch": 0.37557, + "grad_norm": 1.0920157623142508, + "learning_rate": 0.003, + "loss": 4.026, + "step": 37557 + }, + { + "epoch": 0.37558, + "grad_norm": 1.0682656412255815, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 37558 + }, + { + "epoch": 0.37559, + "grad_norm": 1.2210478107496958, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 37559 + }, + { + "epoch": 0.3756, + "grad_norm": 1.1193846767389979, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 37560 + }, + { + "epoch": 0.37561, + "grad_norm": 1.1039811036953435, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 37561 + }, + { + "epoch": 0.37562, + "grad_norm": 0.9836638165903714, + "learning_rate": 0.003, + "loss": 4.037, + "step": 37562 + }, + { + "epoch": 0.37563, + "grad_norm": 1.247002073588649, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 37563 + }, + { + "epoch": 0.37564, + "grad_norm": 1.09134671510424, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 37564 + }, + { + "epoch": 0.37565, + "grad_norm": 1.2006241468481997, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 37565 + }, + { + "epoch": 0.37566, + "grad_norm": 1.1959303045910918, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 37566 + }, + { + "epoch": 0.37567, + "grad_norm": 0.9899332057197936, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 37567 + }, + { + "epoch": 0.37568, + "grad_norm": 0.8977245967539504, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 37568 + }, + { + "epoch": 0.37569, + "grad_norm": 0.8793968923271226, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 37569 + }, + { + "epoch": 0.3757, + "grad_norm": 0.9862923053902573, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 37570 + }, + { + "epoch": 0.37571, + "grad_norm": 1.263845571446357, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 37571 + }, + { + "epoch": 0.37572, + "grad_norm": 1.0275207405609805, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 37572 + }, + { + "epoch": 0.37573, + "grad_norm": 1.070110439695937, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 37573 + }, + { + "epoch": 0.37574, + "grad_norm": 1.1549984949854564, + "learning_rate": 0.003, + "loss": 4.036, + "step": 37574 + }, + { + "epoch": 0.37575, + "grad_norm": 1.21690549669249, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 37575 + }, + { + "epoch": 0.37576, + "grad_norm": 1.1519320750021598, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 37576 + }, + { + "epoch": 0.37577, + "grad_norm": 0.9560612200423896, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 37577 + }, + { + "epoch": 0.37578, + "grad_norm": 1.1798126198463519, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 37578 + }, + { + "epoch": 0.37579, + "grad_norm": 1.1775832127603418, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 37579 + }, + { + "epoch": 0.3758, + "grad_norm": 1.0798630919802343, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 37580 + }, + { + "epoch": 0.37581, + "grad_norm": 1.2187063345512348, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 37581 + }, + { + "epoch": 0.37582, + "grad_norm": 0.9107466737172845, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 37582 + }, + { + "epoch": 0.37583, + "grad_norm": 0.893391570556649, + "learning_rate": 0.003, + "loss": 4.033, + "step": 37583 + }, + { + "epoch": 0.37584, + "grad_norm": 1.1535668476380192, + "learning_rate": 0.003, + "loss": 4.008, + "step": 37584 + }, + { + "epoch": 0.37585, + "grad_norm": 1.0773833819089644, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 37585 + }, + { + "epoch": 0.37586, + "grad_norm": 1.0874449511399893, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 37586 + }, + { + "epoch": 0.37587, + "grad_norm": 1.1288250625004002, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 37587 + }, + { + "epoch": 0.37588, + "grad_norm": 0.9774962085351876, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 37588 + }, + { + "epoch": 0.37589, + "grad_norm": 1.0826844018577813, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 37589 + }, + { + "epoch": 0.3759, + "grad_norm": 1.2790278668739004, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 37590 + }, + { + "epoch": 0.37591, + "grad_norm": 1.0409570769789909, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 37591 + }, + { + "epoch": 0.37592, + "grad_norm": 1.0794463372433345, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 37592 + }, + { + "epoch": 0.37593, + "grad_norm": 1.2153439139706341, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 37593 + }, + { + "epoch": 0.37594, + "grad_norm": 1.0116400530762197, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 37594 + }, + { + "epoch": 0.37595, + "grad_norm": 1.2309240611688983, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 37595 + }, + { + "epoch": 0.37596, + "grad_norm": 0.8858350586979835, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 37596 + }, + { + "epoch": 0.37597, + "grad_norm": 0.9739850171453904, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 37597 + }, + { + "epoch": 0.37598, + "grad_norm": 1.2007054179295555, + "learning_rate": 0.003, + "loss": 4.025, + "step": 37598 + }, + { + "epoch": 0.37599, + "grad_norm": 1.250937183804164, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 37599 + }, + { + "epoch": 0.376, + "grad_norm": 0.8695223005270967, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 37600 + }, + { + "epoch": 0.37601, + "grad_norm": 0.8296184164382523, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 37601 + }, + { + "epoch": 0.37602, + "grad_norm": 1.0713494637129224, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 37602 + }, + { + "epoch": 0.37603, + "grad_norm": 1.2580670492597303, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 37603 + }, + { + "epoch": 0.37604, + "grad_norm": 0.9846972020879547, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 37604 + }, + { + "epoch": 0.37605, + "grad_norm": 1.049117435519296, + "learning_rate": 0.003, + "loss": 4.055, + "step": 37605 + }, + { + "epoch": 0.37606, + "grad_norm": 1.0547038796811499, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 37606 + }, + { + "epoch": 0.37607, + "grad_norm": 1.1034016922668433, + "learning_rate": 0.003, + "loss": 4.051, + "step": 37607 + }, + { + "epoch": 0.37608, + "grad_norm": 1.090812219904647, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 37608 + }, + { + "epoch": 0.37609, + "grad_norm": 1.1591598611394516, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 37609 + }, + { + "epoch": 0.3761, + "grad_norm": 1.1158548891578546, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 37610 + }, + { + "epoch": 0.37611, + "grad_norm": 1.0914521255021299, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 37611 + }, + { + "epoch": 0.37612, + "grad_norm": 1.0391019910924697, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 37612 + }, + { + "epoch": 0.37613, + "grad_norm": 1.2147945557502058, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 37613 + }, + { + "epoch": 0.37614, + "grad_norm": 1.0275799271508934, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 37614 + }, + { + "epoch": 0.37615, + "grad_norm": 1.1257413608646976, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 37615 + }, + { + "epoch": 0.37616, + "grad_norm": 1.0212736281724435, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 37616 + }, + { + "epoch": 0.37617, + "grad_norm": 1.0679108202258338, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 37617 + }, + { + "epoch": 0.37618, + "grad_norm": 1.213024557130562, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 37618 + }, + { + "epoch": 0.37619, + "grad_norm": 0.9190861225378176, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 37619 + }, + { + "epoch": 0.3762, + "grad_norm": 1.2289780791621414, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 37620 + }, + { + "epoch": 0.37621, + "grad_norm": 1.1040311781666068, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 37621 + }, + { + "epoch": 0.37622, + "grad_norm": 1.071816972780309, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 37622 + }, + { + "epoch": 0.37623, + "grad_norm": 1.0400012740072497, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 37623 + }, + { + "epoch": 0.37624, + "grad_norm": 0.9788979073580956, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 37624 + }, + { + "epoch": 0.37625, + "grad_norm": 1.21062936096592, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 37625 + }, + { + "epoch": 0.37626, + "grad_norm": 1.1886075864777836, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 37626 + }, + { + "epoch": 0.37627, + "grad_norm": 1.1827226208663615, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 37627 + }, + { + "epoch": 0.37628, + "grad_norm": 1.2851760125123928, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 37628 + }, + { + "epoch": 0.37629, + "grad_norm": 0.9296054010327869, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 37629 + }, + { + "epoch": 0.3763, + "grad_norm": 0.9968984463860135, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 37630 + }, + { + "epoch": 0.37631, + "grad_norm": 1.15386283693688, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 37631 + }, + { + "epoch": 0.37632, + "grad_norm": 1.3212884170834311, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 37632 + }, + { + "epoch": 0.37633, + "grad_norm": 0.8345468507513016, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 37633 + }, + { + "epoch": 0.37634, + "grad_norm": 0.8653666321686873, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 37634 + }, + { + "epoch": 0.37635, + "grad_norm": 1.037076519096909, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 37635 + }, + { + "epoch": 0.37636, + "grad_norm": 1.169017483362535, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 37636 + }, + { + "epoch": 0.37637, + "grad_norm": 1.0578048545808452, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 37637 + }, + { + "epoch": 0.37638, + "grad_norm": 1.1441200129165718, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 37638 + }, + { + "epoch": 0.37639, + "grad_norm": 1.0884900155615294, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 37639 + }, + { + "epoch": 0.3764, + "grad_norm": 1.2596818060295307, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 37640 + }, + { + "epoch": 0.37641, + "grad_norm": 1.1980454926504374, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 37641 + }, + { + "epoch": 0.37642, + "grad_norm": 0.9825588157523991, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 37642 + }, + { + "epoch": 0.37643, + "grad_norm": 1.1127127691597432, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 37643 + }, + { + "epoch": 0.37644, + "grad_norm": 1.2528330531936713, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 37644 + }, + { + "epoch": 0.37645, + "grad_norm": 0.9243518321250597, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 37645 + }, + { + "epoch": 0.37646, + "grad_norm": 0.9482897956449794, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 37646 + }, + { + "epoch": 0.37647, + "grad_norm": 1.1556268041377273, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 37647 + }, + { + "epoch": 0.37648, + "grad_norm": 1.1202309840967317, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 37648 + }, + { + "epoch": 0.37649, + "grad_norm": 1.0150216211232281, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 37649 + }, + { + "epoch": 0.3765, + "grad_norm": 0.9893894437394458, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 37650 + }, + { + "epoch": 0.37651, + "grad_norm": 1.0221207518687916, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 37651 + }, + { + "epoch": 0.37652, + "grad_norm": 1.058332033034951, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 37652 + }, + { + "epoch": 0.37653, + "grad_norm": 1.0672157394203792, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 37653 + }, + { + "epoch": 0.37654, + "grad_norm": 1.302546766059152, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 37654 + }, + { + "epoch": 0.37655, + "grad_norm": 1.0241286690524873, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 37655 + }, + { + "epoch": 0.37656, + "grad_norm": 1.11486042586206, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 37656 + }, + { + "epoch": 0.37657, + "grad_norm": 1.0246878388277092, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 37657 + }, + { + "epoch": 0.37658, + "grad_norm": 1.0669496587565659, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 37658 + }, + { + "epoch": 0.37659, + "grad_norm": 1.1114022908654766, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 37659 + }, + { + "epoch": 0.3766, + "grad_norm": 0.9879789963214866, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 37660 + }, + { + "epoch": 0.37661, + "grad_norm": 0.9481582073550999, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 37661 + }, + { + "epoch": 0.37662, + "grad_norm": 0.9792117725207146, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 37662 + }, + { + "epoch": 0.37663, + "grad_norm": 1.0744095758173688, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 37663 + }, + { + "epoch": 0.37664, + "grad_norm": 1.0408050203595587, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 37664 + }, + { + "epoch": 0.37665, + "grad_norm": 1.2393113108665759, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 37665 + }, + { + "epoch": 0.37666, + "grad_norm": 1.012817753561915, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 37666 + }, + { + "epoch": 0.37667, + "grad_norm": 1.3503511537118955, + "learning_rate": 0.003, + "loss": 4.035, + "step": 37667 + }, + { + "epoch": 0.37668, + "grad_norm": 1.0154232740716267, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 37668 + }, + { + "epoch": 0.37669, + "grad_norm": 1.4093768625030603, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 37669 + }, + { + "epoch": 0.3767, + "grad_norm": 1.117403015529841, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 37670 + }, + { + "epoch": 0.37671, + "grad_norm": 0.8972848014040183, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 37671 + }, + { + "epoch": 0.37672, + "grad_norm": 0.9887172998170809, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 37672 + }, + { + "epoch": 0.37673, + "grad_norm": 1.47352655737785, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 37673 + }, + { + "epoch": 0.37674, + "grad_norm": 0.8986419929377656, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 37674 + }, + { + "epoch": 0.37675, + "grad_norm": 0.9143962523994213, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 37675 + }, + { + "epoch": 0.37676, + "grad_norm": 1.0108943132658088, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 37676 + }, + { + "epoch": 0.37677, + "grad_norm": 1.11856254364856, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 37677 + }, + { + "epoch": 0.37678, + "grad_norm": 1.0645504842320337, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 37678 + }, + { + "epoch": 0.37679, + "grad_norm": 1.4277359045235378, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 37679 + }, + { + "epoch": 0.3768, + "grad_norm": 1.0212052814121841, + "learning_rate": 0.003, + "loss": 4.053, + "step": 37680 + }, + { + "epoch": 0.37681, + "grad_norm": 1.1673110935919506, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 37681 + }, + { + "epoch": 0.37682, + "grad_norm": 0.9492727282793234, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 37682 + }, + { + "epoch": 0.37683, + "grad_norm": 1.0191217197331, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 37683 + }, + { + "epoch": 0.37684, + "grad_norm": 1.01906712243396, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 37684 + }, + { + "epoch": 0.37685, + "grad_norm": 0.9898594624966938, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 37685 + }, + { + "epoch": 0.37686, + "grad_norm": 1.1384126779145645, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 37686 + }, + { + "epoch": 0.37687, + "grad_norm": 1.1289287677915403, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 37687 + }, + { + "epoch": 0.37688, + "grad_norm": 1.171130422738786, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 37688 + }, + { + "epoch": 0.37689, + "grad_norm": 1.09007220488996, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 37689 + }, + { + "epoch": 0.3769, + "grad_norm": 0.994587022088737, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 37690 + }, + { + "epoch": 0.37691, + "grad_norm": 1.0345764089427019, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 37691 + }, + { + "epoch": 0.37692, + "grad_norm": 1.1629083638134166, + "learning_rate": 0.003, + "loss": 4.038, + "step": 37692 + }, + { + "epoch": 0.37693, + "grad_norm": 1.3197211488711464, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 37693 + }, + { + "epoch": 0.37694, + "grad_norm": 0.7750178313151677, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 37694 + }, + { + "epoch": 0.37695, + "grad_norm": 0.8231864043266106, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 37695 + }, + { + "epoch": 0.37696, + "grad_norm": 0.9097072334317058, + "learning_rate": 0.003, + "loss": 4.056, + "step": 37696 + }, + { + "epoch": 0.37697, + "grad_norm": 1.0005337787969464, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 37697 + }, + { + "epoch": 0.37698, + "grad_norm": 1.1835701417111804, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 37698 + }, + { + "epoch": 0.37699, + "grad_norm": 0.9425979356470253, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 37699 + }, + { + "epoch": 0.377, + "grad_norm": 1.0203021128635907, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 37700 + }, + { + "epoch": 0.37701, + "grad_norm": 1.1331686390643418, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 37701 + }, + { + "epoch": 0.37702, + "grad_norm": 1.1867565086254581, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 37702 + }, + { + "epoch": 0.37703, + "grad_norm": 1.0526898655869685, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 37703 + }, + { + "epoch": 0.37704, + "grad_norm": 1.1983110358464648, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 37704 + }, + { + "epoch": 0.37705, + "grad_norm": 1.311536639255419, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 37705 + }, + { + "epoch": 0.37706, + "grad_norm": 1.0495136590636411, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 37706 + }, + { + "epoch": 0.37707, + "grad_norm": 1.1398318703755992, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 37707 + }, + { + "epoch": 0.37708, + "grad_norm": 1.1308491203738478, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 37708 + }, + { + "epoch": 0.37709, + "grad_norm": 0.8735292111894486, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 37709 + }, + { + "epoch": 0.3771, + "grad_norm": 0.9239526953302684, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 37710 + }, + { + "epoch": 0.37711, + "grad_norm": 1.1896737785379405, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 37711 + }, + { + "epoch": 0.37712, + "grad_norm": 1.2011945252671115, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 37712 + }, + { + "epoch": 0.37713, + "grad_norm": 1.1121124339013444, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 37713 + }, + { + "epoch": 0.37714, + "grad_norm": 1.054211884700278, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 37714 + }, + { + "epoch": 0.37715, + "grad_norm": 1.4191814768245663, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 37715 + }, + { + "epoch": 0.37716, + "grad_norm": 0.8826713705722323, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 37716 + }, + { + "epoch": 0.37717, + "grad_norm": 0.8087995978171435, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 37717 + }, + { + "epoch": 0.37718, + "grad_norm": 0.942874670606866, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 37718 + }, + { + "epoch": 0.37719, + "grad_norm": 1.2343811419476256, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 37719 + }, + { + "epoch": 0.3772, + "grad_norm": 0.9747322504360919, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 37720 + }, + { + "epoch": 0.37721, + "grad_norm": 1.2529430434220707, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 37721 + }, + { + "epoch": 0.37722, + "grad_norm": 1.2271223765586208, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 37722 + }, + { + "epoch": 0.37723, + "grad_norm": 1.0590255047884976, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 37723 + }, + { + "epoch": 0.37724, + "grad_norm": 1.2835693108555803, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 37724 + }, + { + "epoch": 0.37725, + "grad_norm": 1.0856184155535074, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 37725 + }, + { + "epoch": 0.37726, + "grad_norm": 1.1353902716779742, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 37726 + }, + { + "epoch": 0.37727, + "grad_norm": 0.9354123032279335, + "learning_rate": 0.003, + "loss": 4.026, + "step": 37727 + }, + { + "epoch": 0.37728, + "grad_norm": 1.1909007819472137, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 37728 + }, + { + "epoch": 0.37729, + "grad_norm": 1.2825834834840948, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 37729 + }, + { + "epoch": 0.3773, + "grad_norm": 0.9541046330506003, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 37730 + }, + { + "epoch": 0.37731, + "grad_norm": 0.956952385588665, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 37731 + }, + { + "epoch": 0.37732, + "grad_norm": 1.1531840542920535, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 37732 + }, + { + "epoch": 0.37733, + "grad_norm": 1.1532699551165047, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 37733 + }, + { + "epoch": 0.37734, + "grad_norm": 1.1858772024094018, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 37734 + }, + { + "epoch": 0.37735, + "grad_norm": 0.9863817724953883, + "learning_rate": 0.003, + "loss": 4.036, + "step": 37735 + }, + { + "epoch": 0.37736, + "grad_norm": 0.9352841790406119, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 37736 + }, + { + "epoch": 0.37737, + "grad_norm": 1.0300373968549386, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 37737 + }, + { + "epoch": 0.37738, + "grad_norm": 1.138616895415841, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 37738 + }, + { + "epoch": 0.37739, + "grad_norm": 0.7627273866706218, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 37739 + }, + { + "epoch": 0.3774, + "grad_norm": 0.9151656757915707, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 37740 + }, + { + "epoch": 0.37741, + "grad_norm": 1.022571800931771, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 37741 + }, + { + "epoch": 0.37742, + "grad_norm": 1.2045474811133414, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 37742 + }, + { + "epoch": 0.37743, + "grad_norm": 1.0422630660479555, + "learning_rate": 0.003, + "loss": 4.059, + "step": 37743 + }, + { + "epoch": 0.37744, + "grad_norm": 1.1987769016938012, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 37744 + }, + { + "epoch": 0.37745, + "grad_norm": 0.9599457400374766, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 37745 + }, + { + "epoch": 0.37746, + "grad_norm": 1.1766094141023313, + "learning_rate": 0.003, + "loss": 4.0878, + "step": 37746 + }, + { + "epoch": 0.37747, + "grad_norm": 1.246121197383843, + "learning_rate": 0.003, + "loss": 4.045, + "step": 37747 + }, + { + "epoch": 0.37748, + "grad_norm": 1.0083145508941687, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 37748 + }, + { + "epoch": 0.37749, + "grad_norm": 1.1596831090333333, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 37749 + }, + { + "epoch": 0.3775, + "grad_norm": 1.147905805136515, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 37750 + }, + { + "epoch": 0.37751, + "grad_norm": 1.2084313727208056, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 37751 + }, + { + "epoch": 0.37752, + "grad_norm": 1.4614372743124202, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 37752 + }, + { + "epoch": 0.37753, + "grad_norm": 0.9590153182907339, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 37753 + }, + { + "epoch": 0.37754, + "grad_norm": 1.2260697380799606, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 37754 + }, + { + "epoch": 0.37755, + "grad_norm": 1.024020388341691, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 37755 + }, + { + "epoch": 0.37756, + "grad_norm": 1.1030722357909588, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 37756 + }, + { + "epoch": 0.37757, + "grad_norm": 1.1874272421672551, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 37757 + }, + { + "epoch": 0.37758, + "grad_norm": 0.9879959029984982, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 37758 + }, + { + "epoch": 0.37759, + "grad_norm": 1.0011149449659482, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 37759 + }, + { + "epoch": 0.3776, + "grad_norm": 1.1256745554904992, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 37760 + }, + { + "epoch": 0.37761, + "grad_norm": 0.9794072211389486, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 37761 + }, + { + "epoch": 0.37762, + "grad_norm": 1.0236972693610922, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 37762 + }, + { + "epoch": 0.37763, + "grad_norm": 1.0933845883107378, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 37763 + }, + { + "epoch": 0.37764, + "grad_norm": 1.2984863176827444, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 37764 + }, + { + "epoch": 0.37765, + "grad_norm": 1.0305337090289788, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 37765 + }, + { + "epoch": 0.37766, + "grad_norm": 1.0444202012378696, + "learning_rate": 0.003, + "loss": 4.06, + "step": 37766 + }, + { + "epoch": 0.37767, + "grad_norm": 1.0681564840970055, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 37767 + }, + { + "epoch": 0.37768, + "grad_norm": 1.2245903499604116, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 37768 + }, + { + "epoch": 0.37769, + "grad_norm": 1.0017570081740588, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 37769 + }, + { + "epoch": 0.3777, + "grad_norm": 1.0114448725789302, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 37770 + }, + { + "epoch": 0.37771, + "grad_norm": 1.1862417521189392, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 37771 + }, + { + "epoch": 0.37772, + "grad_norm": 0.9746573752183902, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 37772 + }, + { + "epoch": 0.37773, + "grad_norm": 1.0017718476216375, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 37773 + }, + { + "epoch": 0.37774, + "grad_norm": 1.1783767693423166, + "learning_rate": 0.003, + "loss": 4.033, + "step": 37774 + }, + { + "epoch": 0.37775, + "grad_norm": 0.9929901953061399, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 37775 + }, + { + "epoch": 0.37776, + "grad_norm": 1.4188120566220672, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 37776 + }, + { + "epoch": 0.37777, + "grad_norm": 0.8349150535024532, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 37777 + }, + { + "epoch": 0.37778, + "grad_norm": 0.8895964288702152, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 37778 + }, + { + "epoch": 0.37779, + "grad_norm": 1.044250432616606, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 37779 + }, + { + "epoch": 0.3778, + "grad_norm": 1.4420759476831904, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 37780 + }, + { + "epoch": 0.37781, + "grad_norm": 1.017082788316604, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 37781 + }, + { + "epoch": 0.37782, + "grad_norm": 1.3029655991730054, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 37782 + }, + { + "epoch": 0.37783, + "grad_norm": 0.8856919456585735, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 37783 + }, + { + "epoch": 0.37784, + "grad_norm": 0.950740699600656, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 37784 + }, + { + "epoch": 0.37785, + "grad_norm": 1.0994624217836533, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 37785 + }, + { + "epoch": 0.37786, + "grad_norm": 0.9394245836615953, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 37786 + }, + { + "epoch": 0.37787, + "grad_norm": 1.2015837964226772, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 37787 + }, + { + "epoch": 0.37788, + "grad_norm": 1.1103966896435444, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 37788 + }, + { + "epoch": 0.37789, + "grad_norm": 0.9696852115425195, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 37789 + }, + { + "epoch": 0.3779, + "grad_norm": 1.109807116996365, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 37790 + }, + { + "epoch": 0.37791, + "grad_norm": 1.3735356639897807, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 37791 + }, + { + "epoch": 0.37792, + "grad_norm": 1.0216227449179334, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 37792 + }, + { + "epoch": 0.37793, + "grad_norm": 1.0756858075159639, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 37793 + }, + { + "epoch": 0.37794, + "grad_norm": 0.8126586563269045, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 37794 + }, + { + "epoch": 0.37795, + "grad_norm": 0.9438767789756963, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 37795 + }, + { + "epoch": 0.37796, + "grad_norm": 1.4928451100311788, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 37796 + }, + { + "epoch": 0.37797, + "grad_norm": 0.701556516110975, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 37797 + }, + { + "epoch": 0.37798, + "grad_norm": 0.7751405145000948, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 37798 + }, + { + "epoch": 0.37799, + "grad_norm": 1.3219541887945614, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 37799 + }, + { + "epoch": 0.378, + "grad_norm": 0.9766014337965431, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 37800 + }, + { + "epoch": 0.37801, + "grad_norm": 1.1754288653859946, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 37801 + }, + { + "epoch": 0.37802, + "grad_norm": 0.9867195194833475, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 37802 + }, + { + "epoch": 0.37803, + "grad_norm": 0.9634320643228529, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 37803 + }, + { + "epoch": 0.37804, + "grad_norm": 1.1324629668653594, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 37804 + }, + { + "epoch": 0.37805, + "grad_norm": 0.8742069162482239, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 37805 + }, + { + "epoch": 0.37806, + "grad_norm": 1.108248358818601, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 37806 + }, + { + "epoch": 0.37807, + "grad_norm": 1.1989057204585163, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 37807 + }, + { + "epoch": 0.37808, + "grad_norm": 0.9063128762043744, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 37808 + }, + { + "epoch": 0.37809, + "grad_norm": 0.9872532114759037, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 37809 + }, + { + "epoch": 0.3781, + "grad_norm": 1.3302560558216965, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 37810 + }, + { + "epoch": 0.37811, + "grad_norm": 1.2929634549575562, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 37811 + }, + { + "epoch": 0.37812, + "grad_norm": 0.9537777436823922, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 37812 + }, + { + "epoch": 0.37813, + "grad_norm": 1.0515330346901077, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 37813 + }, + { + "epoch": 0.37814, + "grad_norm": 1.1486932102996044, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 37814 + }, + { + "epoch": 0.37815, + "grad_norm": 0.9924949369861672, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 37815 + }, + { + "epoch": 0.37816, + "grad_norm": 1.1902367071888271, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 37816 + }, + { + "epoch": 0.37817, + "grad_norm": 1.0785854298740958, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 37817 + }, + { + "epoch": 0.37818, + "grad_norm": 1.3047789720369891, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 37818 + }, + { + "epoch": 0.37819, + "grad_norm": 1.0146686157640254, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 37819 + }, + { + "epoch": 0.3782, + "grad_norm": 1.2467724095259711, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 37820 + }, + { + "epoch": 0.37821, + "grad_norm": 1.0044732828112006, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 37821 + }, + { + "epoch": 0.37822, + "grad_norm": 1.4285187280945744, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 37822 + }, + { + "epoch": 0.37823, + "grad_norm": 0.908855061946722, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 37823 + }, + { + "epoch": 0.37824, + "grad_norm": 1.0462584337488252, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 37824 + }, + { + "epoch": 0.37825, + "grad_norm": 1.176086756939959, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 37825 + }, + { + "epoch": 0.37826, + "grad_norm": 0.9269573279104107, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 37826 + }, + { + "epoch": 0.37827, + "grad_norm": 1.02166943051148, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 37827 + }, + { + "epoch": 0.37828, + "grad_norm": 1.0951778716524112, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 37828 + }, + { + "epoch": 0.37829, + "grad_norm": 0.9214481389319976, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 37829 + }, + { + "epoch": 0.3783, + "grad_norm": 1.0102026955390813, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 37830 + }, + { + "epoch": 0.37831, + "grad_norm": 1.047621726843969, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 37831 + }, + { + "epoch": 0.37832, + "grad_norm": 1.07672251472238, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 37832 + }, + { + "epoch": 0.37833, + "grad_norm": 1.0992282932347517, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 37833 + }, + { + "epoch": 0.37834, + "grad_norm": 1.3153295711315212, + "learning_rate": 0.003, + "loss": 4.025, + "step": 37834 + }, + { + "epoch": 0.37835, + "grad_norm": 0.9004348464068251, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 37835 + }, + { + "epoch": 0.37836, + "grad_norm": 0.9952646897981177, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 37836 + }, + { + "epoch": 0.37837, + "grad_norm": 1.460977975927896, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 37837 + }, + { + "epoch": 0.37838, + "grad_norm": 1.0709108517482695, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 37838 + }, + { + "epoch": 0.37839, + "grad_norm": 1.0535964143453664, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 37839 + }, + { + "epoch": 0.3784, + "grad_norm": 1.2646665929412142, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 37840 + }, + { + "epoch": 0.37841, + "grad_norm": 1.027295029428548, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 37841 + }, + { + "epoch": 0.37842, + "grad_norm": 0.9444647364854375, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 37842 + }, + { + "epoch": 0.37843, + "grad_norm": 1.0092701399778583, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 37843 + }, + { + "epoch": 0.37844, + "grad_norm": 1.177437473453843, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 37844 + }, + { + "epoch": 0.37845, + "grad_norm": 1.5090797509068108, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 37845 + }, + { + "epoch": 0.37846, + "grad_norm": 1.0351773227510708, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 37846 + }, + { + "epoch": 0.37847, + "grad_norm": 1.110421646487538, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 37847 + }, + { + "epoch": 0.37848, + "grad_norm": 0.9524989011952723, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 37848 + }, + { + "epoch": 0.37849, + "grad_norm": 1.1637778598874744, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 37849 + }, + { + "epoch": 0.3785, + "grad_norm": 1.118235326383712, + "learning_rate": 0.003, + "loss": 4.04, + "step": 37850 + }, + { + "epoch": 0.37851, + "grad_norm": 0.9371552812427654, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 37851 + }, + { + "epoch": 0.37852, + "grad_norm": 1.0271588539954482, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 37852 + }, + { + "epoch": 0.37853, + "grad_norm": 1.4312172633584996, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 37853 + }, + { + "epoch": 0.37854, + "grad_norm": 1.0357858461445872, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 37854 + }, + { + "epoch": 0.37855, + "grad_norm": 0.9727170326719637, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 37855 + }, + { + "epoch": 0.37856, + "grad_norm": 1.1530257747419654, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 37856 + }, + { + "epoch": 0.37857, + "grad_norm": 1.0965731251069062, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 37857 + }, + { + "epoch": 0.37858, + "grad_norm": 1.1400944801590052, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 37858 + }, + { + "epoch": 0.37859, + "grad_norm": 0.9472886907018038, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 37859 + }, + { + "epoch": 0.3786, + "grad_norm": 1.201725007213808, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 37860 + }, + { + "epoch": 0.37861, + "grad_norm": 1.0730790467070461, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 37861 + }, + { + "epoch": 0.37862, + "grad_norm": 1.1523362616484676, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 37862 + }, + { + "epoch": 0.37863, + "grad_norm": 1.0101408610287297, + "learning_rate": 0.003, + "loss": 4.011, + "step": 37863 + }, + { + "epoch": 0.37864, + "grad_norm": 0.9016825010956985, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 37864 + }, + { + "epoch": 0.37865, + "grad_norm": 0.9585401225671542, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 37865 + }, + { + "epoch": 0.37866, + "grad_norm": 1.109086612073418, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 37866 + }, + { + "epoch": 0.37867, + "grad_norm": 1.1235728238091303, + "learning_rate": 0.003, + "loss": 4.056, + "step": 37867 + }, + { + "epoch": 0.37868, + "grad_norm": 1.0702017119938123, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 37868 + }, + { + "epoch": 0.37869, + "grad_norm": 0.9653597642296273, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 37869 + }, + { + "epoch": 0.3787, + "grad_norm": 1.1071245645304744, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 37870 + }, + { + "epoch": 0.37871, + "grad_norm": 1.2125623949114528, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 37871 + }, + { + "epoch": 0.37872, + "grad_norm": 0.9536391851562978, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 37872 + }, + { + "epoch": 0.37873, + "grad_norm": 1.0498381756151383, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 37873 + }, + { + "epoch": 0.37874, + "grad_norm": 1.3624393965063888, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 37874 + }, + { + "epoch": 0.37875, + "grad_norm": 0.9303470940139932, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 37875 + }, + { + "epoch": 0.37876, + "grad_norm": 0.9108552657256191, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 37876 + }, + { + "epoch": 0.37877, + "grad_norm": 1.0793048073124214, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 37877 + }, + { + "epoch": 0.37878, + "grad_norm": 1.1379084659688143, + "learning_rate": 0.003, + "loss": 4.034, + "step": 37878 + }, + { + "epoch": 0.37879, + "grad_norm": 0.894798416612511, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 37879 + }, + { + "epoch": 0.3788, + "grad_norm": 1.0464646153365338, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 37880 + }, + { + "epoch": 0.37881, + "grad_norm": 1.68934832614725, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 37881 + }, + { + "epoch": 0.37882, + "grad_norm": 0.9306538964599012, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 37882 + }, + { + "epoch": 0.37883, + "grad_norm": 1.0446339610302084, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 37883 + }, + { + "epoch": 0.37884, + "grad_norm": 1.1523376095715128, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 37884 + }, + { + "epoch": 0.37885, + "grad_norm": 1.15362830967809, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 37885 + }, + { + "epoch": 0.37886, + "grad_norm": 1.308191582983755, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 37886 + }, + { + "epoch": 0.37887, + "grad_norm": 0.865446815566557, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 37887 + }, + { + "epoch": 0.37888, + "grad_norm": 1.06434975003464, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 37888 + }, + { + "epoch": 0.37889, + "grad_norm": 1.1355205590840822, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 37889 + }, + { + "epoch": 0.3789, + "grad_norm": 0.9008729326242946, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 37890 + }, + { + "epoch": 0.37891, + "grad_norm": 1.012996219258367, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 37891 + }, + { + "epoch": 0.37892, + "grad_norm": 1.3128662440439713, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 37892 + }, + { + "epoch": 0.37893, + "grad_norm": 0.8035653958857822, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 37893 + }, + { + "epoch": 0.37894, + "grad_norm": 0.9027987861818721, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 37894 + }, + { + "epoch": 0.37895, + "grad_norm": 1.3238164353399167, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 37895 + }, + { + "epoch": 0.37896, + "grad_norm": 1.077546369539363, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 37896 + }, + { + "epoch": 0.37897, + "grad_norm": 1.0274971139730582, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 37897 + }, + { + "epoch": 0.37898, + "grad_norm": 1.1042391851065632, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 37898 + }, + { + "epoch": 0.37899, + "grad_norm": 1.045827951585061, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 37899 + }, + { + "epoch": 0.379, + "grad_norm": 1.0008033430337908, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 37900 + }, + { + "epoch": 0.37901, + "grad_norm": 1.0894250158281669, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 37901 + }, + { + "epoch": 0.37902, + "grad_norm": 1.0298298449256855, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 37902 + }, + { + "epoch": 0.37903, + "grad_norm": 1.0307111954073522, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 37903 + }, + { + "epoch": 0.37904, + "grad_norm": 1.1965120887976515, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 37904 + }, + { + "epoch": 0.37905, + "grad_norm": 0.8264080516778497, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 37905 + }, + { + "epoch": 0.37906, + "grad_norm": 0.6891928721775216, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 37906 + }, + { + "epoch": 0.37907, + "grad_norm": 0.8961678854121592, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 37907 + }, + { + "epoch": 0.37908, + "grad_norm": 1.4550377304754734, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 37908 + }, + { + "epoch": 0.37909, + "grad_norm": 1.235198690993513, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 37909 + }, + { + "epoch": 0.3791, + "grad_norm": 0.9469296599697627, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 37910 + }, + { + "epoch": 0.37911, + "grad_norm": 0.9609301890124005, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 37911 + }, + { + "epoch": 0.37912, + "grad_norm": 1.1953202633093176, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 37912 + }, + { + "epoch": 0.37913, + "grad_norm": 1.1352351637989155, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 37913 + }, + { + "epoch": 0.37914, + "grad_norm": 1.01369826769462, + "learning_rate": 0.003, + "loss": 4.024, + "step": 37914 + }, + { + "epoch": 0.37915, + "grad_norm": 0.9791357522577887, + "learning_rate": 0.003, + "loss": 4.027, + "step": 37915 + }, + { + "epoch": 0.37916, + "grad_norm": 1.0940802878751692, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 37916 + }, + { + "epoch": 0.37917, + "grad_norm": 1.176920868573392, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 37917 + }, + { + "epoch": 0.37918, + "grad_norm": 1.0008955613429795, + "learning_rate": 0.003, + "loss": 4.054, + "step": 37918 + }, + { + "epoch": 0.37919, + "grad_norm": 0.9966810378865361, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 37919 + }, + { + "epoch": 0.3792, + "grad_norm": 1.1710607070619734, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 37920 + }, + { + "epoch": 0.37921, + "grad_norm": 1.1232588434851425, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 37921 + }, + { + "epoch": 0.37922, + "grad_norm": 1.1140954195410813, + "learning_rate": 0.003, + "loss": 4.044, + "step": 37922 + }, + { + "epoch": 0.37923, + "grad_norm": 1.3484028087795614, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 37923 + }, + { + "epoch": 0.37924, + "grad_norm": 0.9057310620875999, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 37924 + }, + { + "epoch": 0.37925, + "grad_norm": 1.0394571947262026, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 37925 + }, + { + "epoch": 0.37926, + "grad_norm": 1.2698294817904174, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 37926 + }, + { + "epoch": 0.37927, + "grad_norm": 0.773284350488124, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 37927 + }, + { + "epoch": 0.37928, + "grad_norm": 0.9589202494191116, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 37928 + }, + { + "epoch": 0.37929, + "grad_norm": 1.1155587126597715, + "learning_rate": 0.003, + "loss": 3.991, + "step": 37929 + }, + { + "epoch": 0.3793, + "grad_norm": 0.9234726633994004, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 37930 + }, + { + "epoch": 0.37931, + "grad_norm": 1.0075300652959545, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 37931 + }, + { + "epoch": 0.37932, + "grad_norm": 1.0882957237764166, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 37932 + }, + { + "epoch": 0.37933, + "grad_norm": 1.0890374846231177, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 37933 + }, + { + "epoch": 0.37934, + "grad_norm": 1.124340477889388, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 37934 + }, + { + "epoch": 0.37935, + "grad_norm": 1.0382463180874872, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 37935 + }, + { + "epoch": 0.37936, + "grad_norm": 1.2395184313752332, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 37936 + }, + { + "epoch": 0.37937, + "grad_norm": 1.2064525374335966, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 37937 + }, + { + "epoch": 0.37938, + "grad_norm": 1.3057851417630306, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 37938 + }, + { + "epoch": 0.37939, + "grad_norm": 1.0633248177305248, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 37939 + }, + { + "epoch": 0.3794, + "grad_norm": 1.0242151667976127, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 37940 + }, + { + "epoch": 0.37941, + "grad_norm": 1.1141799942759731, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 37941 + }, + { + "epoch": 0.37942, + "grad_norm": 1.1883027217247475, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 37942 + }, + { + "epoch": 0.37943, + "grad_norm": 0.9589588548085056, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 37943 + }, + { + "epoch": 0.37944, + "grad_norm": 1.1029744070071914, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 37944 + }, + { + "epoch": 0.37945, + "grad_norm": 1.2020893136946247, + "learning_rate": 0.003, + "loss": 4.028, + "step": 37945 + }, + { + "epoch": 0.37946, + "grad_norm": 1.0021705268299277, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 37946 + }, + { + "epoch": 0.37947, + "grad_norm": 1.0292925580924548, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 37947 + }, + { + "epoch": 0.37948, + "grad_norm": 1.1954906763193394, + "learning_rate": 0.003, + "loss": 4.013, + "step": 37948 + }, + { + "epoch": 0.37949, + "grad_norm": 1.2311331933947676, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 37949 + }, + { + "epoch": 0.3795, + "grad_norm": 1.1694350611177282, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 37950 + }, + { + "epoch": 0.37951, + "grad_norm": 0.8744356676472446, + "learning_rate": 0.003, + "loss": 4.023, + "step": 37951 + }, + { + "epoch": 0.37952, + "grad_norm": 1.047789943526462, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 37952 + }, + { + "epoch": 0.37953, + "grad_norm": 1.1047679829239252, + "learning_rate": 0.003, + "loss": 4.029, + "step": 37953 + }, + { + "epoch": 0.37954, + "grad_norm": 0.9674410733345532, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 37954 + }, + { + "epoch": 0.37955, + "grad_norm": 1.0517383220558019, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 37955 + }, + { + "epoch": 0.37956, + "grad_norm": 1.1526197288444722, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 37956 + }, + { + "epoch": 0.37957, + "grad_norm": 1.1425085059429696, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 37957 + }, + { + "epoch": 0.37958, + "grad_norm": 1.1117485680100572, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 37958 + }, + { + "epoch": 0.37959, + "grad_norm": 0.9131429434071086, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 37959 + }, + { + "epoch": 0.3796, + "grad_norm": 1.1030216307775522, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 37960 + }, + { + "epoch": 0.37961, + "grad_norm": 1.1229306825645933, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 37961 + }, + { + "epoch": 0.37962, + "grad_norm": 1.0209820795753992, + "learning_rate": 0.003, + "loss": 4.026, + "step": 37962 + }, + { + "epoch": 0.37963, + "grad_norm": 1.0921564547676104, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 37963 + }, + { + "epoch": 0.37964, + "grad_norm": 1.2036739194654988, + "learning_rate": 0.003, + "loss": 4.04, + "step": 37964 + }, + { + "epoch": 0.37965, + "grad_norm": 0.9712556451890662, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 37965 + }, + { + "epoch": 0.37966, + "grad_norm": 0.9131478021037087, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 37966 + }, + { + "epoch": 0.37967, + "grad_norm": 1.0311178370970853, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 37967 + }, + { + "epoch": 0.37968, + "grad_norm": 1.3329351263633609, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 37968 + }, + { + "epoch": 0.37969, + "grad_norm": 0.9186732235258371, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 37969 + }, + { + "epoch": 0.3797, + "grad_norm": 0.923031729220218, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 37970 + }, + { + "epoch": 0.37971, + "grad_norm": 1.0694496463844732, + "learning_rate": 0.003, + "loss": 4.038, + "step": 37971 + }, + { + "epoch": 0.37972, + "grad_norm": 1.1308063840264684, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 37972 + }, + { + "epoch": 0.37973, + "grad_norm": 1.017892630593658, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 37973 + }, + { + "epoch": 0.37974, + "grad_norm": 0.9572482335748121, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 37974 + }, + { + "epoch": 0.37975, + "grad_norm": 1.1856453708376964, + "learning_rate": 0.003, + "loss": 4.059, + "step": 37975 + }, + { + "epoch": 0.37976, + "grad_norm": 1.2346935152339666, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 37976 + }, + { + "epoch": 0.37977, + "grad_norm": 1.1456030560714072, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 37977 + }, + { + "epoch": 0.37978, + "grad_norm": 1.1329048705990452, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 37978 + }, + { + "epoch": 0.37979, + "grad_norm": 1.1360832341426397, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 37979 + }, + { + "epoch": 0.3798, + "grad_norm": 1.2682562313533263, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 37980 + }, + { + "epoch": 0.37981, + "grad_norm": 1.0138263483820293, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 37981 + }, + { + "epoch": 0.37982, + "grad_norm": 1.0888092155078686, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 37982 + }, + { + "epoch": 0.37983, + "grad_norm": 1.2802495661782158, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 37983 + }, + { + "epoch": 0.37984, + "grad_norm": 0.9305542730345041, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 37984 + }, + { + "epoch": 0.37985, + "grad_norm": 1.0699243137109917, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 37985 + }, + { + "epoch": 0.37986, + "grad_norm": 0.9775759925068196, + "learning_rate": 0.003, + "loss": 4.035, + "step": 37986 + }, + { + "epoch": 0.37987, + "grad_norm": 1.0669256935331641, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 37987 + }, + { + "epoch": 0.37988, + "grad_norm": 0.9723375790697429, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 37988 + }, + { + "epoch": 0.37989, + "grad_norm": 0.9154480911152824, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 37989 + }, + { + "epoch": 0.3799, + "grad_norm": 1.02847103708076, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 37990 + }, + { + "epoch": 0.37991, + "grad_norm": 1.2840465223564337, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 37991 + }, + { + "epoch": 0.37992, + "grad_norm": 1.2903285338248376, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 37992 + }, + { + "epoch": 0.37993, + "grad_norm": 1.1056920232048613, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 37993 + }, + { + "epoch": 0.37994, + "grad_norm": 0.7721024241561648, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 37994 + }, + { + "epoch": 0.37995, + "grad_norm": 0.7100954912874247, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 37995 + }, + { + "epoch": 0.37996, + "grad_norm": 0.7160978296513162, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 37996 + }, + { + "epoch": 0.37997, + "grad_norm": 0.8472201652636024, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 37997 + }, + { + "epoch": 0.37998, + "grad_norm": 0.9839315426453641, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 37998 + }, + { + "epoch": 0.37999, + "grad_norm": 1.2328256532748738, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 37999 + }, + { + "epoch": 0.38, + "grad_norm": 1.1868032299939297, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 38000 + }, + { + "epoch": 0.38001, + "grad_norm": 1.0486977221896547, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 38001 + }, + { + "epoch": 0.38002, + "grad_norm": 1.2120754547625558, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 38002 + }, + { + "epoch": 0.38003, + "grad_norm": 0.8641034549631681, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 38003 + }, + { + "epoch": 0.38004, + "grad_norm": 0.829523045099462, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 38004 + }, + { + "epoch": 0.38005, + "grad_norm": 1.0486704959769084, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 38005 + }, + { + "epoch": 0.38006, + "grad_norm": 1.1473075166962319, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 38006 + }, + { + "epoch": 0.38007, + "grad_norm": 0.9582548689924827, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 38007 + }, + { + "epoch": 0.38008, + "grad_norm": 0.9101281614429825, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 38008 + }, + { + "epoch": 0.38009, + "grad_norm": 1.0165752063483287, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 38009 + }, + { + "epoch": 0.3801, + "grad_norm": 1.1207932433388204, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 38010 + }, + { + "epoch": 0.38011, + "grad_norm": 1.0922966429632304, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 38011 + }, + { + "epoch": 0.38012, + "grad_norm": 1.1527737433839915, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 38012 + }, + { + "epoch": 0.38013, + "grad_norm": 1.1030870646992232, + "learning_rate": 0.003, + "loss": 4.037, + "step": 38013 + }, + { + "epoch": 0.38014, + "grad_norm": 1.0359431188281405, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 38014 + }, + { + "epoch": 0.38015, + "grad_norm": 1.3171721135059056, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 38015 + }, + { + "epoch": 0.38016, + "grad_norm": 1.110741214226886, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 38016 + }, + { + "epoch": 0.38017, + "grad_norm": 1.15792953131861, + "learning_rate": 0.003, + "loss": 4.04, + "step": 38017 + }, + { + "epoch": 0.38018, + "grad_norm": 0.9120599447950377, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 38018 + }, + { + "epoch": 0.38019, + "grad_norm": 1.0837651361809646, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 38019 + }, + { + "epoch": 0.3802, + "grad_norm": 1.2007284118154067, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 38020 + }, + { + "epoch": 0.38021, + "grad_norm": 0.9985930625836433, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 38021 + }, + { + "epoch": 0.38022, + "grad_norm": 1.0602318416554106, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 38022 + }, + { + "epoch": 0.38023, + "grad_norm": 1.0652374957464377, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 38023 + }, + { + "epoch": 0.38024, + "grad_norm": 1.2010108960551749, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 38024 + }, + { + "epoch": 0.38025, + "grad_norm": 0.9285208151161407, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 38025 + }, + { + "epoch": 0.38026, + "grad_norm": 1.1317687018075302, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 38026 + }, + { + "epoch": 0.38027, + "grad_norm": 1.181453712301815, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 38027 + }, + { + "epoch": 0.38028, + "grad_norm": 1.1120879611158019, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 38028 + }, + { + "epoch": 0.38029, + "grad_norm": 1.0773024524492996, + "learning_rate": 0.003, + "loss": 4.044, + "step": 38029 + }, + { + "epoch": 0.3803, + "grad_norm": 1.0993373429392423, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 38030 + }, + { + "epoch": 0.38031, + "grad_norm": 1.144588121818685, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 38031 + }, + { + "epoch": 0.38032, + "grad_norm": 1.2171584606533372, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 38032 + }, + { + "epoch": 0.38033, + "grad_norm": 1.001491666161331, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 38033 + }, + { + "epoch": 0.38034, + "grad_norm": 1.3435411540577546, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 38034 + }, + { + "epoch": 0.38035, + "grad_norm": 0.9010841019594298, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 38035 + }, + { + "epoch": 0.38036, + "grad_norm": 1.0579983032499536, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 38036 + }, + { + "epoch": 0.38037, + "grad_norm": 1.0932807906707827, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 38037 + }, + { + "epoch": 0.38038, + "grad_norm": 1.0789498917108633, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 38038 + }, + { + "epoch": 0.38039, + "grad_norm": 1.0967378231987592, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 38039 + }, + { + "epoch": 0.3804, + "grad_norm": 1.3100735620870652, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 38040 + }, + { + "epoch": 0.38041, + "grad_norm": 1.0633750124559862, + "learning_rate": 0.003, + "loss": 4.034, + "step": 38041 + }, + { + "epoch": 0.38042, + "grad_norm": 0.9438880544271857, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 38042 + }, + { + "epoch": 0.38043, + "grad_norm": 0.9646462411602048, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 38043 + }, + { + "epoch": 0.38044, + "grad_norm": 1.0571923197090989, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 38044 + }, + { + "epoch": 0.38045, + "grad_norm": 1.1436387093853675, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 38045 + }, + { + "epoch": 0.38046, + "grad_norm": 0.9462539147760534, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 38046 + }, + { + "epoch": 0.38047, + "grad_norm": 1.1497286424932618, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 38047 + }, + { + "epoch": 0.38048, + "grad_norm": 0.977580745575867, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 38048 + }, + { + "epoch": 0.38049, + "grad_norm": 1.1119599240625302, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 38049 + }, + { + "epoch": 0.3805, + "grad_norm": 1.1931727030388501, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 38050 + }, + { + "epoch": 0.38051, + "grad_norm": 0.8748249609595738, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 38051 + }, + { + "epoch": 0.38052, + "grad_norm": 0.9740483393676522, + "learning_rate": 0.003, + "loss": 4.032, + "step": 38052 + }, + { + "epoch": 0.38053, + "grad_norm": 1.2434377902638354, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 38053 + }, + { + "epoch": 0.38054, + "grad_norm": 1.1076107984694998, + "learning_rate": 0.003, + "loss": 4.035, + "step": 38054 + }, + { + "epoch": 0.38055, + "grad_norm": 1.361243672594244, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 38055 + }, + { + "epoch": 0.38056, + "grad_norm": 1.1870012396347664, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 38056 + }, + { + "epoch": 0.38057, + "grad_norm": 0.9831219838044603, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 38057 + }, + { + "epoch": 0.38058, + "grad_norm": 1.0294426022913994, + "learning_rate": 0.003, + "loss": 4.054, + "step": 38058 + }, + { + "epoch": 0.38059, + "grad_norm": 1.0978091419091836, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 38059 + }, + { + "epoch": 0.3806, + "grad_norm": 1.1438019406844617, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 38060 + }, + { + "epoch": 0.38061, + "grad_norm": 0.9148065127681451, + "learning_rate": 0.003, + "loss": 4.061, + "step": 38061 + }, + { + "epoch": 0.38062, + "grad_norm": 1.2057647229810957, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 38062 + }, + { + "epoch": 0.38063, + "grad_norm": 1.311516017959224, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 38063 + }, + { + "epoch": 0.38064, + "grad_norm": 0.8233959832371883, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 38064 + }, + { + "epoch": 0.38065, + "grad_norm": 0.9214151607219639, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 38065 + }, + { + "epoch": 0.38066, + "grad_norm": 1.0324431053229295, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 38066 + }, + { + "epoch": 0.38067, + "grad_norm": 1.306456502832609, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 38067 + }, + { + "epoch": 0.38068, + "grad_norm": 0.8303795550795626, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 38068 + }, + { + "epoch": 0.38069, + "grad_norm": 0.8158841471151264, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 38069 + }, + { + "epoch": 0.3807, + "grad_norm": 0.9607962634500584, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 38070 + }, + { + "epoch": 0.38071, + "grad_norm": 1.3173215046453615, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 38071 + }, + { + "epoch": 0.38072, + "grad_norm": 0.7603996342015612, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 38072 + }, + { + "epoch": 0.38073, + "grad_norm": 0.7505862063298655, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 38073 + }, + { + "epoch": 0.38074, + "grad_norm": 0.9451139708858803, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 38074 + }, + { + "epoch": 0.38075, + "grad_norm": 1.2504479259285344, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 38075 + }, + { + "epoch": 0.38076, + "grad_norm": 0.9175464667952272, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 38076 + }, + { + "epoch": 0.38077, + "grad_norm": 1.1607138300954762, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 38077 + }, + { + "epoch": 0.38078, + "grad_norm": 1.2029621963597525, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 38078 + }, + { + "epoch": 0.38079, + "grad_norm": 0.9974290899545118, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 38079 + }, + { + "epoch": 0.3808, + "grad_norm": 1.1637941085100265, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 38080 + }, + { + "epoch": 0.38081, + "grad_norm": 1.0088134526158616, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 38081 + }, + { + "epoch": 0.38082, + "grad_norm": 1.07439197359407, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 38082 + }, + { + "epoch": 0.38083, + "grad_norm": 1.159060181223573, + "learning_rate": 0.003, + "loss": 4.042, + "step": 38083 + }, + { + "epoch": 0.38084, + "grad_norm": 1.0278740709619956, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 38084 + }, + { + "epoch": 0.38085, + "grad_norm": 1.0275127449732926, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 38085 + }, + { + "epoch": 0.38086, + "grad_norm": 0.9885449344460161, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 38086 + }, + { + "epoch": 0.38087, + "grad_norm": 1.139770669388103, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 38087 + }, + { + "epoch": 0.38088, + "grad_norm": 0.9714445842502204, + "learning_rate": 0.003, + "loss": 4.042, + "step": 38088 + }, + { + "epoch": 0.38089, + "grad_norm": 1.2895250660620912, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 38089 + }, + { + "epoch": 0.3809, + "grad_norm": 0.9758062884239043, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 38090 + }, + { + "epoch": 0.38091, + "grad_norm": 1.3841009006254852, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 38091 + }, + { + "epoch": 0.38092, + "grad_norm": 0.9677282275956353, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 38092 + }, + { + "epoch": 0.38093, + "grad_norm": 1.2631056455948908, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 38093 + }, + { + "epoch": 0.38094, + "grad_norm": 1.0507265355788775, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 38094 + }, + { + "epoch": 0.38095, + "grad_norm": 1.2223064410905073, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 38095 + }, + { + "epoch": 0.38096, + "grad_norm": 1.2333254108997038, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 38096 + }, + { + "epoch": 0.38097, + "grad_norm": 1.0019507172150448, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 38097 + }, + { + "epoch": 0.38098, + "grad_norm": 1.0316248087950615, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 38098 + }, + { + "epoch": 0.38099, + "grad_norm": 1.2030398068737762, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 38099 + }, + { + "epoch": 0.381, + "grad_norm": 1.3680514598756357, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 38100 + }, + { + "epoch": 0.38101, + "grad_norm": 0.8407307215116205, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 38101 + }, + { + "epoch": 0.38102, + "grad_norm": 0.7398398933891888, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 38102 + }, + { + "epoch": 0.38103, + "grad_norm": 0.7651534064899338, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 38103 + }, + { + "epoch": 0.38104, + "grad_norm": 0.9349617715150088, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 38104 + }, + { + "epoch": 0.38105, + "grad_norm": 1.1819526341038278, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 38105 + }, + { + "epoch": 0.38106, + "grad_norm": 1.079443437125494, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 38106 + }, + { + "epoch": 0.38107, + "grad_norm": 1.1488362052728918, + "learning_rate": 0.003, + "loss": 4.05, + "step": 38107 + }, + { + "epoch": 0.38108, + "grad_norm": 1.0001484273613086, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 38108 + }, + { + "epoch": 0.38109, + "grad_norm": 1.2770347544988068, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 38109 + }, + { + "epoch": 0.3811, + "grad_norm": 1.1300660677339842, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 38110 + }, + { + "epoch": 0.38111, + "grad_norm": 1.0480232061213162, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 38111 + }, + { + "epoch": 0.38112, + "grad_norm": 1.1943987405578949, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 38112 + }, + { + "epoch": 0.38113, + "grad_norm": 0.9889147108175474, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 38113 + }, + { + "epoch": 0.38114, + "grad_norm": 1.123540231417467, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 38114 + }, + { + "epoch": 0.38115, + "grad_norm": 1.0837447294545575, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 38115 + }, + { + "epoch": 0.38116, + "grad_norm": 1.0334065615123502, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 38116 + }, + { + "epoch": 0.38117, + "grad_norm": 1.0470786777663017, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 38117 + }, + { + "epoch": 0.38118, + "grad_norm": 1.1208590473861935, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 38118 + }, + { + "epoch": 0.38119, + "grad_norm": 1.0203713563455794, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 38119 + }, + { + "epoch": 0.3812, + "grad_norm": 1.1525904668639801, + "learning_rate": 0.003, + "loss": 4.046, + "step": 38120 + }, + { + "epoch": 0.38121, + "grad_norm": 0.887182611737285, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 38121 + }, + { + "epoch": 0.38122, + "grad_norm": 1.0465015574458456, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 38122 + }, + { + "epoch": 0.38123, + "grad_norm": 0.9982209313411047, + "learning_rate": 0.003, + "loss": 4.035, + "step": 38123 + }, + { + "epoch": 0.38124, + "grad_norm": 0.9954627160417705, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 38124 + }, + { + "epoch": 0.38125, + "grad_norm": 1.1432169186871066, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 38125 + }, + { + "epoch": 0.38126, + "grad_norm": 0.9208807548742902, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 38126 + }, + { + "epoch": 0.38127, + "grad_norm": 1.0949043836459122, + "learning_rate": 0.003, + "loss": 4.04, + "step": 38127 + }, + { + "epoch": 0.38128, + "grad_norm": 1.1986765767137533, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 38128 + }, + { + "epoch": 0.38129, + "grad_norm": 0.8685267371299874, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 38129 + }, + { + "epoch": 0.3813, + "grad_norm": 1.0073920462033101, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 38130 + }, + { + "epoch": 0.38131, + "grad_norm": 1.2564615706713, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 38131 + }, + { + "epoch": 0.38132, + "grad_norm": 0.9913359675455141, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 38132 + }, + { + "epoch": 0.38133, + "grad_norm": 1.4365424180757023, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 38133 + }, + { + "epoch": 0.38134, + "grad_norm": 0.9129636329219143, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 38134 + }, + { + "epoch": 0.38135, + "grad_norm": 1.181602692163288, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 38135 + }, + { + "epoch": 0.38136, + "grad_norm": 1.4861755115615505, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 38136 + }, + { + "epoch": 0.38137, + "grad_norm": 0.6870431872021237, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 38137 + }, + { + "epoch": 0.38138, + "grad_norm": 0.7095926375243152, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 38138 + }, + { + "epoch": 0.38139, + "grad_norm": 0.8751538354591519, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 38139 + }, + { + "epoch": 0.3814, + "grad_norm": 1.151436334406677, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 38140 + }, + { + "epoch": 0.38141, + "grad_norm": 1.4620299090575846, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 38141 + }, + { + "epoch": 0.38142, + "grad_norm": 1.0539433577590023, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 38142 + }, + { + "epoch": 0.38143, + "grad_norm": 1.1158806283348568, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 38143 + }, + { + "epoch": 0.38144, + "grad_norm": 1.1334787695099406, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 38144 + }, + { + "epoch": 0.38145, + "grad_norm": 1.1930481902651862, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 38145 + }, + { + "epoch": 0.38146, + "grad_norm": 0.8783058630554252, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 38146 + }, + { + "epoch": 0.38147, + "grad_norm": 0.9894014764511156, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 38147 + }, + { + "epoch": 0.38148, + "grad_norm": 1.292233504578954, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 38148 + }, + { + "epoch": 0.38149, + "grad_norm": 0.9729189318510605, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 38149 + }, + { + "epoch": 0.3815, + "grad_norm": 1.1443658817609155, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 38150 + }, + { + "epoch": 0.38151, + "grad_norm": 1.197455063777796, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 38151 + }, + { + "epoch": 0.38152, + "grad_norm": 0.9901992754377585, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 38152 + }, + { + "epoch": 0.38153, + "grad_norm": 1.0058408076001493, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 38153 + }, + { + "epoch": 0.38154, + "grad_norm": 0.9624527162536458, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 38154 + }, + { + "epoch": 0.38155, + "grad_norm": 1.1775473224659236, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 38155 + }, + { + "epoch": 0.38156, + "grad_norm": 1.0178242748140456, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 38156 + }, + { + "epoch": 0.38157, + "grad_norm": 1.1710669734860149, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 38157 + }, + { + "epoch": 0.38158, + "grad_norm": 1.075709192013079, + "learning_rate": 0.003, + "loss": 4.051, + "step": 38158 + }, + { + "epoch": 0.38159, + "grad_norm": 1.1936083401007578, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 38159 + }, + { + "epoch": 0.3816, + "grad_norm": 1.1289407327022771, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 38160 + }, + { + "epoch": 0.38161, + "grad_norm": 0.980447493841913, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 38161 + }, + { + "epoch": 0.38162, + "grad_norm": 1.1069439074704004, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 38162 + }, + { + "epoch": 0.38163, + "grad_norm": 1.1685424936387725, + "learning_rate": 0.003, + "loss": 4.034, + "step": 38163 + }, + { + "epoch": 0.38164, + "grad_norm": 1.0438404944324402, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 38164 + }, + { + "epoch": 0.38165, + "grad_norm": 1.2816292871584285, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 38165 + }, + { + "epoch": 0.38166, + "grad_norm": 1.0760510775037702, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 38166 + }, + { + "epoch": 0.38167, + "grad_norm": 1.039742682346043, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 38167 + }, + { + "epoch": 0.38168, + "grad_norm": 1.239367642355639, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 38168 + }, + { + "epoch": 0.38169, + "grad_norm": 0.9704319312460977, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 38169 + }, + { + "epoch": 0.3817, + "grad_norm": 1.1813145514113237, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 38170 + }, + { + "epoch": 0.38171, + "grad_norm": 1.1192364239695984, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 38171 + }, + { + "epoch": 0.38172, + "grad_norm": 1.0085663856814326, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 38172 + }, + { + "epoch": 0.38173, + "grad_norm": 0.994761987822227, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 38173 + }, + { + "epoch": 0.38174, + "grad_norm": 1.195832496513494, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 38174 + }, + { + "epoch": 0.38175, + "grad_norm": 1.0361304400826317, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 38175 + }, + { + "epoch": 0.38176, + "grad_norm": 0.8288150978112586, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 38176 + }, + { + "epoch": 0.38177, + "grad_norm": 0.9754593154057256, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 38177 + }, + { + "epoch": 0.38178, + "grad_norm": 1.0090238912349252, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 38178 + }, + { + "epoch": 0.38179, + "grad_norm": 1.0611433874150953, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 38179 + }, + { + "epoch": 0.3818, + "grad_norm": 1.0453533853488104, + "learning_rate": 0.003, + "loss": 4.065, + "step": 38180 + }, + { + "epoch": 0.38181, + "grad_norm": 1.2056102592262175, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 38181 + }, + { + "epoch": 0.38182, + "grad_norm": 1.1276573701417132, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 38182 + }, + { + "epoch": 0.38183, + "grad_norm": 1.066443594046972, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 38183 + }, + { + "epoch": 0.38184, + "grad_norm": 1.068607282576246, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 38184 + }, + { + "epoch": 0.38185, + "grad_norm": 1.3288050921788739, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 38185 + }, + { + "epoch": 0.38186, + "grad_norm": 1.187846170759396, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 38186 + }, + { + "epoch": 0.38187, + "grad_norm": 0.9237315276223316, + "learning_rate": 0.003, + "loss": 4.063, + "step": 38187 + }, + { + "epoch": 0.38188, + "grad_norm": 1.018055145052292, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 38188 + }, + { + "epoch": 0.38189, + "grad_norm": 1.2857836831892377, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 38189 + }, + { + "epoch": 0.3819, + "grad_norm": 1.1286221407263974, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 38190 + }, + { + "epoch": 0.38191, + "grad_norm": 1.2099180298854577, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 38191 + }, + { + "epoch": 0.38192, + "grad_norm": 0.8698331516637408, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 38192 + }, + { + "epoch": 0.38193, + "grad_norm": 0.8198734580080109, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 38193 + }, + { + "epoch": 0.38194, + "grad_norm": 0.9972846039415544, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 38194 + }, + { + "epoch": 0.38195, + "grad_norm": 1.2698549084889457, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 38195 + }, + { + "epoch": 0.38196, + "grad_norm": 0.9187689352935985, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 38196 + }, + { + "epoch": 0.38197, + "grad_norm": 1.0455680304917532, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 38197 + }, + { + "epoch": 0.38198, + "grad_norm": 0.9326639600769181, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 38198 + }, + { + "epoch": 0.38199, + "grad_norm": 1.0795102652723656, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 38199 + }, + { + "epoch": 0.382, + "grad_norm": 1.0367840551963958, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 38200 + }, + { + "epoch": 0.38201, + "grad_norm": 1.0892562214644357, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 38201 + }, + { + "epoch": 0.38202, + "grad_norm": 0.999543862189938, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 38202 + }, + { + "epoch": 0.38203, + "grad_norm": 1.2771047946346712, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 38203 + }, + { + "epoch": 0.38204, + "grad_norm": 1.146627815744344, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 38204 + }, + { + "epoch": 0.38205, + "grad_norm": 1.4322075166203554, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 38205 + }, + { + "epoch": 0.38206, + "grad_norm": 1.0518849096322291, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 38206 + }, + { + "epoch": 0.38207, + "grad_norm": 1.0107409517843093, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 38207 + }, + { + "epoch": 0.38208, + "grad_norm": 1.178210456962775, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 38208 + }, + { + "epoch": 0.38209, + "grad_norm": 0.9636836211591852, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 38209 + }, + { + "epoch": 0.3821, + "grad_norm": 1.1442074662813688, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 38210 + }, + { + "epoch": 0.38211, + "grad_norm": 1.0627281710829144, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 38211 + }, + { + "epoch": 0.38212, + "grad_norm": 1.1073834177552144, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 38212 + }, + { + "epoch": 0.38213, + "grad_norm": 1.2118915581085417, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 38213 + }, + { + "epoch": 0.38214, + "grad_norm": 1.1152403555218395, + "learning_rate": 0.003, + "loss": 4.052, + "step": 38214 + }, + { + "epoch": 0.38215, + "grad_norm": 1.0518496755689035, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 38215 + }, + { + "epoch": 0.38216, + "grad_norm": 1.238609426445459, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 38216 + }, + { + "epoch": 0.38217, + "grad_norm": 0.9736515255070838, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 38217 + }, + { + "epoch": 0.38218, + "grad_norm": 0.9721070499378875, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 38218 + }, + { + "epoch": 0.38219, + "grad_norm": 0.9595761399038462, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 38219 + }, + { + "epoch": 0.3822, + "grad_norm": 0.9571491840565509, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 38220 + }, + { + "epoch": 0.38221, + "grad_norm": 1.041399835290153, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 38221 + }, + { + "epoch": 0.38222, + "grad_norm": 1.083938504824791, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 38222 + }, + { + "epoch": 0.38223, + "grad_norm": 1.009899987866786, + "learning_rate": 0.003, + "loss": 4.067, + "step": 38223 + }, + { + "epoch": 0.38224, + "grad_norm": 1.4710189197875059, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 38224 + }, + { + "epoch": 0.38225, + "grad_norm": 1.0716078973613594, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 38225 + }, + { + "epoch": 0.38226, + "grad_norm": 1.0226910532303615, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 38226 + }, + { + "epoch": 0.38227, + "grad_norm": 1.0860491977313946, + "learning_rate": 0.003, + "loss": 4.067, + "step": 38227 + }, + { + "epoch": 0.38228, + "grad_norm": 1.0953084371655528, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 38228 + }, + { + "epoch": 0.38229, + "grad_norm": 1.010026139164369, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 38229 + }, + { + "epoch": 0.3823, + "grad_norm": 0.9693023248392855, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 38230 + }, + { + "epoch": 0.38231, + "grad_norm": 1.2569643386075113, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 38231 + }, + { + "epoch": 0.38232, + "grad_norm": 0.9556865857507574, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 38232 + }, + { + "epoch": 0.38233, + "grad_norm": 1.3163225655673017, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 38233 + }, + { + "epoch": 0.38234, + "grad_norm": 1.240075518212444, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 38234 + }, + { + "epoch": 0.38235, + "grad_norm": 1.040792078446498, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 38235 + }, + { + "epoch": 0.38236, + "grad_norm": 0.9686623097951925, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 38236 + }, + { + "epoch": 0.38237, + "grad_norm": 1.0756872175247667, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 38237 + }, + { + "epoch": 0.38238, + "grad_norm": 1.3911681133249538, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 38238 + }, + { + "epoch": 0.38239, + "grad_norm": 0.9909400085548209, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 38239 + }, + { + "epoch": 0.3824, + "grad_norm": 0.951225167525819, + "learning_rate": 0.003, + "loss": 3.985, + "step": 38240 + }, + { + "epoch": 0.38241, + "grad_norm": 0.9338048963838911, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 38241 + }, + { + "epoch": 0.38242, + "grad_norm": 1.0465784749641796, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 38242 + }, + { + "epoch": 0.38243, + "grad_norm": 1.1536422145910563, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 38243 + }, + { + "epoch": 0.38244, + "grad_norm": 1.0276333552832693, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 38244 + }, + { + "epoch": 0.38245, + "grad_norm": 0.9069943131344743, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 38245 + }, + { + "epoch": 0.38246, + "grad_norm": 0.9023834757099167, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 38246 + }, + { + "epoch": 0.38247, + "grad_norm": 1.0790961969873276, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 38247 + }, + { + "epoch": 0.38248, + "grad_norm": 1.2074353841789482, + "learning_rate": 0.003, + "loss": 4.03, + "step": 38248 + }, + { + "epoch": 0.38249, + "grad_norm": 1.101037289349962, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 38249 + }, + { + "epoch": 0.3825, + "grad_norm": 0.9346430250998501, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 38250 + }, + { + "epoch": 0.38251, + "grad_norm": 0.9350105848040702, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 38251 + }, + { + "epoch": 0.38252, + "grad_norm": 0.9119033600568999, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 38252 + }, + { + "epoch": 0.38253, + "grad_norm": 1.053342613702654, + "learning_rate": 0.003, + "loss": 4.03, + "step": 38253 + }, + { + "epoch": 0.38254, + "grad_norm": 1.1334965747531824, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 38254 + }, + { + "epoch": 0.38255, + "grad_norm": 1.0709131961895435, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 38255 + }, + { + "epoch": 0.38256, + "grad_norm": 1.1174188843011303, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 38256 + }, + { + "epoch": 0.38257, + "grad_norm": 1.2369603585208393, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 38257 + }, + { + "epoch": 0.38258, + "grad_norm": 0.9405450458791387, + "learning_rate": 0.003, + "loss": 4.034, + "step": 38258 + }, + { + "epoch": 0.38259, + "grad_norm": 1.2087512743452273, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 38259 + }, + { + "epoch": 0.3826, + "grad_norm": 1.255075850032323, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 38260 + }, + { + "epoch": 0.38261, + "grad_norm": 1.1654320184880986, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 38261 + }, + { + "epoch": 0.38262, + "grad_norm": 1.1236686347361637, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 38262 + }, + { + "epoch": 0.38263, + "grad_norm": 1.1254369278285665, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 38263 + }, + { + "epoch": 0.38264, + "grad_norm": 1.1197666242767006, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 38264 + }, + { + "epoch": 0.38265, + "grad_norm": 1.1536810641829753, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 38265 + }, + { + "epoch": 0.38266, + "grad_norm": 0.9202392046294435, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 38266 + }, + { + "epoch": 0.38267, + "grad_norm": 0.8513927373609639, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 38267 + }, + { + "epoch": 0.38268, + "grad_norm": 0.9254074298616374, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 38268 + }, + { + "epoch": 0.38269, + "grad_norm": 1.2128877551160975, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 38269 + }, + { + "epoch": 0.3827, + "grad_norm": 1.039802342993711, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 38270 + }, + { + "epoch": 0.38271, + "grad_norm": 0.8571840938628142, + "learning_rate": 0.003, + "loss": 4.049, + "step": 38271 + }, + { + "epoch": 0.38272, + "grad_norm": 0.9472508754813342, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 38272 + }, + { + "epoch": 0.38273, + "grad_norm": 1.1971871698134027, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 38273 + }, + { + "epoch": 0.38274, + "grad_norm": 0.9243849592213093, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 38274 + }, + { + "epoch": 0.38275, + "grad_norm": 1.089248491568835, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 38275 + }, + { + "epoch": 0.38276, + "grad_norm": 1.485866938318443, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 38276 + }, + { + "epoch": 0.38277, + "grad_norm": 0.8056854697264497, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 38277 + }, + { + "epoch": 0.38278, + "grad_norm": 0.8640986620124285, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 38278 + }, + { + "epoch": 0.38279, + "grad_norm": 0.8977372641787023, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 38279 + }, + { + "epoch": 0.3828, + "grad_norm": 1.073141331006545, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 38280 + }, + { + "epoch": 0.38281, + "grad_norm": 1.3082039915183166, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 38281 + }, + { + "epoch": 0.38282, + "grad_norm": 1.0230472285080032, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 38282 + }, + { + "epoch": 0.38283, + "grad_norm": 1.184090439373477, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 38283 + }, + { + "epoch": 0.38284, + "grad_norm": 0.9983016937014179, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 38284 + }, + { + "epoch": 0.38285, + "grad_norm": 1.3096679878281834, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 38285 + }, + { + "epoch": 0.38286, + "grad_norm": 1.2066263756690325, + "learning_rate": 0.003, + "loss": 4.032, + "step": 38286 + }, + { + "epoch": 0.38287, + "grad_norm": 0.8561974226101544, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 38287 + }, + { + "epoch": 0.38288, + "grad_norm": 0.887315082301393, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 38288 + }, + { + "epoch": 0.38289, + "grad_norm": 1.1962577010304893, + "learning_rate": 0.003, + "loss": 4.037, + "step": 38289 + }, + { + "epoch": 0.3829, + "grad_norm": 1.3016689262288694, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 38290 + }, + { + "epoch": 0.38291, + "grad_norm": 0.8073096731739383, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 38291 + }, + { + "epoch": 0.38292, + "grad_norm": 0.8907918696025293, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 38292 + }, + { + "epoch": 0.38293, + "grad_norm": 1.2358659581425921, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 38293 + }, + { + "epoch": 0.38294, + "grad_norm": 1.159431203815093, + "learning_rate": 0.003, + "loss": 4.029, + "step": 38294 + }, + { + "epoch": 0.38295, + "grad_norm": 1.003415026947427, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 38295 + }, + { + "epoch": 0.38296, + "grad_norm": 1.1504823091952014, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 38296 + }, + { + "epoch": 0.38297, + "grad_norm": 1.1295694932864808, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 38297 + }, + { + "epoch": 0.38298, + "grad_norm": 1.0765505812472735, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 38298 + }, + { + "epoch": 0.38299, + "grad_norm": 1.1205174067897832, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 38299 + }, + { + "epoch": 0.383, + "grad_norm": 0.9246685399950589, + "learning_rate": 0.003, + "loss": 4.012, + "step": 38300 + }, + { + "epoch": 0.38301, + "grad_norm": 1.0867496584474685, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 38301 + }, + { + "epoch": 0.38302, + "grad_norm": 1.239791712955602, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 38302 + }, + { + "epoch": 0.38303, + "grad_norm": 1.1667900176217076, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 38303 + }, + { + "epoch": 0.38304, + "grad_norm": 1.0683109598519875, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 38304 + }, + { + "epoch": 0.38305, + "grad_norm": 0.8141033059567465, + "learning_rate": 0.003, + "loss": 3.998, + "step": 38305 + }, + { + "epoch": 0.38306, + "grad_norm": 0.9007414979518908, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 38306 + }, + { + "epoch": 0.38307, + "grad_norm": 1.3289357625805913, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 38307 + }, + { + "epoch": 0.38308, + "grad_norm": 1.0627992653553433, + "learning_rate": 0.003, + "loss": 4.039, + "step": 38308 + }, + { + "epoch": 0.38309, + "grad_norm": 1.3469326374670447, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 38309 + }, + { + "epoch": 0.3831, + "grad_norm": 1.1026683950645835, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 38310 + }, + { + "epoch": 0.38311, + "grad_norm": 0.9757384059484905, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 38311 + }, + { + "epoch": 0.38312, + "grad_norm": 1.0764256138010968, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 38312 + }, + { + "epoch": 0.38313, + "grad_norm": 0.8855787480868601, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 38313 + }, + { + "epoch": 0.38314, + "grad_norm": 0.8876316720302632, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 38314 + }, + { + "epoch": 0.38315, + "grad_norm": 1.0003195711642012, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 38315 + }, + { + "epoch": 0.38316, + "grad_norm": 1.2428042904866554, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 38316 + }, + { + "epoch": 0.38317, + "grad_norm": 1.0613422043564518, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 38317 + }, + { + "epoch": 0.38318, + "grad_norm": 1.0425103852516937, + "learning_rate": 0.003, + "loss": 4.028, + "step": 38318 + }, + { + "epoch": 0.38319, + "grad_norm": 1.1952922937131059, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 38319 + }, + { + "epoch": 0.3832, + "grad_norm": 0.9785456880143363, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 38320 + }, + { + "epoch": 0.38321, + "grad_norm": 1.1925889047051197, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 38321 + }, + { + "epoch": 0.38322, + "grad_norm": 1.0390050154837276, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 38322 + }, + { + "epoch": 0.38323, + "grad_norm": 1.2874238940128355, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 38323 + }, + { + "epoch": 0.38324, + "grad_norm": 0.8103850938261442, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 38324 + }, + { + "epoch": 0.38325, + "grad_norm": 0.8215947350229436, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 38325 + }, + { + "epoch": 0.38326, + "grad_norm": 1.1073107478056814, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 38326 + }, + { + "epoch": 0.38327, + "grad_norm": 1.2416648981240654, + "learning_rate": 0.003, + "loss": 4.041, + "step": 38327 + }, + { + "epoch": 0.38328, + "grad_norm": 1.2696015541682042, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 38328 + }, + { + "epoch": 0.38329, + "grad_norm": 1.141776106318225, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 38329 + }, + { + "epoch": 0.3833, + "grad_norm": 1.091085810645877, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 38330 + }, + { + "epoch": 0.38331, + "grad_norm": 1.1598489337237003, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 38331 + }, + { + "epoch": 0.38332, + "grad_norm": 0.9935702873532778, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 38332 + }, + { + "epoch": 0.38333, + "grad_norm": 1.121105766277725, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 38333 + }, + { + "epoch": 0.38334, + "grad_norm": 0.92699418755818, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 38334 + }, + { + "epoch": 0.38335, + "grad_norm": 1.0551218439268046, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 38335 + }, + { + "epoch": 0.38336, + "grad_norm": 1.0251975616821396, + "learning_rate": 0.003, + "loss": 4.007, + "step": 38336 + }, + { + "epoch": 0.38337, + "grad_norm": 1.049760431752997, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 38337 + }, + { + "epoch": 0.38338, + "grad_norm": 1.0573116656066728, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 38338 + }, + { + "epoch": 0.38339, + "grad_norm": 1.1527462925425702, + "learning_rate": 0.003, + "loss": 4.067, + "step": 38339 + }, + { + "epoch": 0.3834, + "grad_norm": 0.9859804582549887, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 38340 + }, + { + "epoch": 0.38341, + "grad_norm": 1.0076770737871004, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 38341 + }, + { + "epoch": 0.38342, + "grad_norm": 1.124091662227899, + "learning_rate": 0.003, + "loss": 4.019, + "step": 38342 + }, + { + "epoch": 0.38343, + "grad_norm": 1.07736316534735, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 38343 + }, + { + "epoch": 0.38344, + "grad_norm": 1.0193011165972643, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 38344 + }, + { + "epoch": 0.38345, + "grad_norm": 1.2294252092989597, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 38345 + }, + { + "epoch": 0.38346, + "grad_norm": 1.0157919267831481, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 38346 + }, + { + "epoch": 0.38347, + "grad_norm": 1.2231026941989056, + "learning_rate": 0.003, + "loss": 4.076, + "step": 38347 + }, + { + "epoch": 0.38348, + "grad_norm": 0.9865463102472756, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 38348 + }, + { + "epoch": 0.38349, + "grad_norm": 1.0928359372888636, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 38349 + }, + { + "epoch": 0.3835, + "grad_norm": 1.1290034161375726, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 38350 + }, + { + "epoch": 0.38351, + "grad_norm": 1.028528217218629, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 38351 + }, + { + "epoch": 0.38352, + "grad_norm": 1.6099280922433394, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 38352 + }, + { + "epoch": 0.38353, + "grad_norm": 1.2489807247712557, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 38353 + }, + { + "epoch": 0.38354, + "grad_norm": 1.1334407562981925, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 38354 + }, + { + "epoch": 0.38355, + "grad_norm": 1.0256744462026781, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 38355 + }, + { + "epoch": 0.38356, + "grad_norm": 1.2551440361947284, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 38356 + }, + { + "epoch": 0.38357, + "grad_norm": 1.0431302891721814, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 38357 + }, + { + "epoch": 0.38358, + "grad_norm": 0.9915563731970428, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 38358 + }, + { + "epoch": 0.38359, + "grad_norm": 1.174665159245864, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 38359 + }, + { + "epoch": 0.3836, + "grad_norm": 1.216089759827949, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 38360 + }, + { + "epoch": 0.38361, + "grad_norm": 0.8803321131259567, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 38361 + }, + { + "epoch": 0.38362, + "grad_norm": 0.8657324102247316, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 38362 + }, + { + "epoch": 0.38363, + "grad_norm": 0.9888644668050385, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 38363 + }, + { + "epoch": 0.38364, + "grad_norm": 1.1355489127168386, + "learning_rate": 0.003, + "loss": 4.0874, + "step": 38364 + }, + { + "epoch": 0.38365, + "grad_norm": 1.1709742655070425, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 38365 + }, + { + "epoch": 0.38366, + "grad_norm": 0.9462672883915179, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 38366 + }, + { + "epoch": 0.38367, + "grad_norm": 1.0421654904412718, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 38367 + }, + { + "epoch": 0.38368, + "grad_norm": 1.1048846803381416, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 38368 + }, + { + "epoch": 0.38369, + "grad_norm": 0.8704281721416365, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 38369 + }, + { + "epoch": 0.3837, + "grad_norm": 0.9588843122513211, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 38370 + }, + { + "epoch": 0.38371, + "grad_norm": 1.0437968443660515, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 38371 + }, + { + "epoch": 0.38372, + "grad_norm": 1.1122980712903023, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 38372 + }, + { + "epoch": 0.38373, + "grad_norm": 0.8681251269785177, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 38373 + }, + { + "epoch": 0.38374, + "grad_norm": 0.9759883000686815, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 38374 + }, + { + "epoch": 0.38375, + "grad_norm": 1.3566397057133461, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 38375 + }, + { + "epoch": 0.38376, + "grad_norm": 0.8665258803637487, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 38376 + }, + { + "epoch": 0.38377, + "grad_norm": 0.9941497781395191, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 38377 + }, + { + "epoch": 0.38378, + "grad_norm": 1.363944356391025, + "learning_rate": 0.003, + "loss": 4.0881, + "step": 38378 + }, + { + "epoch": 0.38379, + "grad_norm": 0.8746591979499077, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 38379 + }, + { + "epoch": 0.3838, + "grad_norm": 1.125926994128703, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 38380 + }, + { + "epoch": 0.38381, + "grad_norm": 1.2600438038778494, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 38381 + }, + { + "epoch": 0.38382, + "grad_norm": 0.9616468128201542, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 38382 + }, + { + "epoch": 0.38383, + "grad_norm": 1.06600489243647, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 38383 + }, + { + "epoch": 0.38384, + "grad_norm": 0.9597986466787282, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 38384 + }, + { + "epoch": 0.38385, + "grad_norm": 0.9363059386477561, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 38385 + }, + { + "epoch": 0.38386, + "grad_norm": 1.026122420552899, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 38386 + }, + { + "epoch": 0.38387, + "grad_norm": 1.2462297651570333, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 38387 + }, + { + "epoch": 0.38388, + "grad_norm": 0.8927321528740982, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 38388 + }, + { + "epoch": 0.38389, + "grad_norm": 1.0986773933529523, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 38389 + }, + { + "epoch": 0.3839, + "grad_norm": 1.3014879457479716, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 38390 + }, + { + "epoch": 0.38391, + "grad_norm": 1.081637131045942, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 38391 + }, + { + "epoch": 0.38392, + "grad_norm": 1.1359404362481533, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 38392 + }, + { + "epoch": 0.38393, + "grad_norm": 0.8754789517010867, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 38393 + }, + { + "epoch": 0.38394, + "grad_norm": 0.9677873224916202, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 38394 + }, + { + "epoch": 0.38395, + "grad_norm": 1.2191553598663603, + "learning_rate": 0.003, + "loss": 4.069, + "step": 38395 + }, + { + "epoch": 0.38396, + "grad_norm": 1.096059317262968, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 38396 + }, + { + "epoch": 0.38397, + "grad_norm": 1.0860460588619645, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 38397 + }, + { + "epoch": 0.38398, + "grad_norm": 1.0357167864621373, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 38398 + }, + { + "epoch": 0.38399, + "grad_norm": 1.1905982014340188, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 38399 + }, + { + "epoch": 0.384, + "grad_norm": 1.0701748027068727, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 38400 + }, + { + "epoch": 0.38401, + "grad_norm": 1.0369892932148856, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 38401 + }, + { + "epoch": 0.38402, + "grad_norm": 1.182515513718645, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 38402 + }, + { + "epoch": 0.38403, + "grad_norm": 0.8710375223099341, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 38403 + }, + { + "epoch": 0.38404, + "grad_norm": 0.8420064742962039, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 38404 + }, + { + "epoch": 0.38405, + "grad_norm": 0.8454642816318583, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 38405 + }, + { + "epoch": 0.38406, + "grad_norm": 0.9276564331202878, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 38406 + }, + { + "epoch": 0.38407, + "grad_norm": 1.3325337344378756, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 38407 + }, + { + "epoch": 0.38408, + "grad_norm": 1.2867136453947083, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 38408 + }, + { + "epoch": 0.38409, + "grad_norm": 1.1459223562222982, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 38409 + }, + { + "epoch": 0.3841, + "grad_norm": 1.0359753305630988, + "learning_rate": 0.003, + "loss": 4.013, + "step": 38410 + }, + { + "epoch": 0.38411, + "grad_norm": 1.0967922858301073, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 38411 + }, + { + "epoch": 0.38412, + "grad_norm": 1.1365512521392709, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 38412 + }, + { + "epoch": 0.38413, + "grad_norm": 1.3697424790080948, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 38413 + }, + { + "epoch": 0.38414, + "grad_norm": 0.910955196090727, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 38414 + }, + { + "epoch": 0.38415, + "grad_norm": 1.0221449475472475, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 38415 + }, + { + "epoch": 0.38416, + "grad_norm": 1.064530154839081, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 38416 + }, + { + "epoch": 0.38417, + "grad_norm": 0.9325303892825407, + "learning_rate": 0.003, + "loss": 4.042, + "step": 38417 + }, + { + "epoch": 0.38418, + "grad_norm": 0.8316118761701172, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 38418 + }, + { + "epoch": 0.38419, + "grad_norm": 0.949520578026658, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 38419 + }, + { + "epoch": 0.3842, + "grad_norm": 1.2143953164107595, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 38420 + }, + { + "epoch": 0.38421, + "grad_norm": 1.3324863792768749, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 38421 + }, + { + "epoch": 0.38422, + "grad_norm": 0.6888520422209793, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 38422 + }, + { + "epoch": 0.38423, + "grad_norm": 0.8151645111139775, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 38423 + }, + { + "epoch": 0.38424, + "grad_norm": 1.0878578910337036, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 38424 + }, + { + "epoch": 0.38425, + "grad_norm": 1.3030000044870425, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 38425 + }, + { + "epoch": 0.38426, + "grad_norm": 1.1196571857004272, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 38426 + }, + { + "epoch": 0.38427, + "grad_norm": 1.304043306349534, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 38427 + }, + { + "epoch": 0.38428, + "grad_norm": 0.9344203400908362, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 38428 + }, + { + "epoch": 0.38429, + "grad_norm": 0.9783233383842233, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 38429 + }, + { + "epoch": 0.3843, + "grad_norm": 0.9729307456863787, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 38430 + }, + { + "epoch": 0.38431, + "grad_norm": 0.920556866448468, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 38431 + }, + { + "epoch": 0.38432, + "grad_norm": 1.0948333194704862, + "learning_rate": 0.003, + "loss": 4.044, + "step": 38432 + }, + { + "epoch": 0.38433, + "grad_norm": 1.0366293278082228, + "learning_rate": 0.003, + "loss": 4.039, + "step": 38433 + }, + { + "epoch": 0.38434, + "grad_norm": 1.0649447796153264, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 38434 + }, + { + "epoch": 0.38435, + "grad_norm": 1.0322376772179462, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 38435 + }, + { + "epoch": 0.38436, + "grad_norm": 1.3309678484005039, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 38436 + }, + { + "epoch": 0.38437, + "grad_norm": 0.9819417775659708, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 38437 + }, + { + "epoch": 0.38438, + "grad_norm": 1.1450981167239571, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 38438 + }, + { + "epoch": 0.38439, + "grad_norm": 1.1876172309317232, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 38439 + }, + { + "epoch": 0.3844, + "grad_norm": 0.927226982295003, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 38440 + }, + { + "epoch": 0.38441, + "grad_norm": 0.9677749126497, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 38441 + }, + { + "epoch": 0.38442, + "grad_norm": 1.0824304118409478, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 38442 + }, + { + "epoch": 0.38443, + "grad_norm": 1.107138483619919, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 38443 + }, + { + "epoch": 0.38444, + "grad_norm": 1.0266480255612351, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 38444 + }, + { + "epoch": 0.38445, + "grad_norm": 1.1588439782460698, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 38445 + }, + { + "epoch": 0.38446, + "grad_norm": 1.1843307912496053, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 38446 + }, + { + "epoch": 0.38447, + "grad_norm": 1.213297299350173, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 38447 + }, + { + "epoch": 0.38448, + "grad_norm": 1.13706507206682, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 38448 + }, + { + "epoch": 0.38449, + "grad_norm": 1.1047176762306385, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 38449 + }, + { + "epoch": 0.3845, + "grad_norm": 0.9023105774383312, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 38450 + }, + { + "epoch": 0.38451, + "grad_norm": 1.0891649498501699, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 38451 + }, + { + "epoch": 0.38452, + "grad_norm": 1.2049171962810226, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 38452 + }, + { + "epoch": 0.38453, + "grad_norm": 0.897685996877255, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 38453 + }, + { + "epoch": 0.38454, + "grad_norm": 1.0148543526652167, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 38454 + }, + { + "epoch": 0.38455, + "grad_norm": 1.3019915079800075, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 38455 + }, + { + "epoch": 0.38456, + "grad_norm": 1.0161989387573553, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 38456 + }, + { + "epoch": 0.38457, + "grad_norm": 1.3808334353261449, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 38457 + }, + { + "epoch": 0.38458, + "grad_norm": 1.241056400699966, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 38458 + }, + { + "epoch": 0.38459, + "grad_norm": 1.152393175990933, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 38459 + }, + { + "epoch": 0.3846, + "grad_norm": 0.924226527588201, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 38460 + }, + { + "epoch": 0.38461, + "grad_norm": 0.9296022193185773, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 38461 + }, + { + "epoch": 0.38462, + "grad_norm": 1.082492272794741, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 38462 + }, + { + "epoch": 0.38463, + "grad_norm": 1.0500390793592722, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 38463 + }, + { + "epoch": 0.38464, + "grad_norm": 1.0934224022872407, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 38464 + }, + { + "epoch": 0.38465, + "grad_norm": 1.2889082849879834, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 38465 + }, + { + "epoch": 0.38466, + "grad_norm": 0.8872457848371775, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 38466 + }, + { + "epoch": 0.38467, + "grad_norm": 0.912644225965169, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 38467 + }, + { + "epoch": 0.38468, + "grad_norm": 1.1372744796847518, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 38468 + }, + { + "epoch": 0.38469, + "grad_norm": 0.9010659885151047, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 38469 + }, + { + "epoch": 0.3847, + "grad_norm": 1.0330232651622437, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 38470 + }, + { + "epoch": 0.38471, + "grad_norm": 1.2496589670666778, + "learning_rate": 0.003, + "loss": 4.056, + "step": 38471 + }, + { + "epoch": 0.38472, + "grad_norm": 1.1339788838038953, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 38472 + }, + { + "epoch": 0.38473, + "grad_norm": 1.1932585340725812, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 38473 + }, + { + "epoch": 0.38474, + "grad_norm": 1.0016021191471887, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 38474 + }, + { + "epoch": 0.38475, + "grad_norm": 1.0703295546585285, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 38475 + }, + { + "epoch": 0.38476, + "grad_norm": 1.091001128439685, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 38476 + }, + { + "epoch": 0.38477, + "grad_norm": 1.2754890484510522, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 38477 + }, + { + "epoch": 0.38478, + "grad_norm": 0.794969405869243, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 38478 + }, + { + "epoch": 0.38479, + "grad_norm": 0.8546540181581078, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 38479 + }, + { + "epoch": 0.3848, + "grad_norm": 1.1063119963171624, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 38480 + }, + { + "epoch": 0.38481, + "grad_norm": 1.1382590728024706, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 38481 + }, + { + "epoch": 0.38482, + "grad_norm": 1.0475098106515353, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 38482 + }, + { + "epoch": 0.38483, + "grad_norm": 0.9489448341362337, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 38483 + }, + { + "epoch": 0.38484, + "grad_norm": 0.9843194796344075, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 38484 + }, + { + "epoch": 0.38485, + "grad_norm": 1.1852489055102229, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 38485 + }, + { + "epoch": 0.38486, + "grad_norm": 1.2470536719139387, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 38486 + }, + { + "epoch": 0.38487, + "grad_norm": 1.1397784524000802, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 38487 + }, + { + "epoch": 0.38488, + "grad_norm": 1.089632863526774, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 38488 + }, + { + "epoch": 0.38489, + "grad_norm": 1.0487334481657722, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 38489 + }, + { + "epoch": 0.3849, + "grad_norm": 1.3870291079510315, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 38490 + }, + { + "epoch": 0.38491, + "grad_norm": 0.9906065179763611, + "learning_rate": 0.003, + "loss": 4.044, + "step": 38491 + }, + { + "epoch": 0.38492, + "grad_norm": 1.04242578440936, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 38492 + }, + { + "epoch": 0.38493, + "grad_norm": 1.156463086609641, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 38493 + }, + { + "epoch": 0.38494, + "grad_norm": 1.1715411907518098, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 38494 + }, + { + "epoch": 0.38495, + "grad_norm": 1.1499414800244283, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 38495 + }, + { + "epoch": 0.38496, + "grad_norm": 0.8765306295108535, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 38496 + }, + { + "epoch": 0.38497, + "grad_norm": 1.0168543375881538, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 38497 + }, + { + "epoch": 0.38498, + "grad_norm": 1.1786741008663626, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 38498 + }, + { + "epoch": 0.38499, + "grad_norm": 0.9905087437769013, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 38499 + }, + { + "epoch": 0.385, + "grad_norm": 1.1460813398513068, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 38500 + }, + { + "epoch": 0.38501, + "grad_norm": 1.1364612760760038, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 38501 + }, + { + "epoch": 0.38502, + "grad_norm": 0.9754580869559506, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 38502 + }, + { + "epoch": 0.38503, + "grad_norm": 1.1645203010865834, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 38503 + }, + { + "epoch": 0.38504, + "grad_norm": 1.0130485278431864, + "learning_rate": 0.003, + "loss": 4.005, + "step": 38504 + }, + { + "epoch": 0.38505, + "grad_norm": 1.0903397570637712, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 38505 + }, + { + "epoch": 0.38506, + "grad_norm": 1.0161161530765903, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 38506 + }, + { + "epoch": 0.38507, + "grad_norm": 1.2472327557393565, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 38507 + }, + { + "epoch": 0.38508, + "grad_norm": 1.0034863758640544, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 38508 + }, + { + "epoch": 0.38509, + "grad_norm": 1.1057386292900542, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 38509 + }, + { + "epoch": 0.3851, + "grad_norm": 0.9167242074402249, + "learning_rate": 0.003, + "loss": 4.037, + "step": 38510 + }, + { + "epoch": 0.38511, + "grad_norm": 0.8198361256717345, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 38511 + }, + { + "epoch": 0.38512, + "grad_norm": 0.9454188995610213, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 38512 + }, + { + "epoch": 0.38513, + "grad_norm": 1.0748987020573155, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 38513 + }, + { + "epoch": 0.38514, + "grad_norm": 1.172137934824207, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 38514 + }, + { + "epoch": 0.38515, + "grad_norm": 1.2026877635342117, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 38515 + }, + { + "epoch": 0.38516, + "grad_norm": 0.9603957249877642, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 38516 + }, + { + "epoch": 0.38517, + "grad_norm": 0.8868818018215552, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 38517 + }, + { + "epoch": 0.38518, + "grad_norm": 1.2188707278924331, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 38518 + }, + { + "epoch": 0.38519, + "grad_norm": 1.2770652349627984, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 38519 + }, + { + "epoch": 0.3852, + "grad_norm": 1.0279724138678568, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 38520 + }, + { + "epoch": 0.38521, + "grad_norm": 1.2584335091691126, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 38521 + }, + { + "epoch": 0.38522, + "grad_norm": 1.109089144555001, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 38522 + }, + { + "epoch": 0.38523, + "grad_norm": 1.017643239491343, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 38523 + }, + { + "epoch": 0.38524, + "grad_norm": 0.985400819645524, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 38524 + }, + { + "epoch": 0.38525, + "grad_norm": 1.1298430776842558, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 38525 + }, + { + "epoch": 0.38526, + "grad_norm": 1.0915243703989388, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 38526 + }, + { + "epoch": 0.38527, + "grad_norm": 0.9662216213463002, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 38527 + }, + { + "epoch": 0.38528, + "grad_norm": 1.1423033485601073, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 38528 + }, + { + "epoch": 0.38529, + "grad_norm": 1.1135422308254377, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 38529 + }, + { + "epoch": 0.3853, + "grad_norm": 0.9557390544773309, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 38530 + }, + { + "epoch": 0.38531, + "grad_norm": 0.990658540538504, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 38531 + }, + { + "epoch": 0.38532, + "grad_norm": 1.125396008342692, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 38532 + }, + { + "epoch": 0.38533, + "grad_norm": 0.98503647384524, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 38533 + }, + { + "epoch": 0.38534, + "grad_norm": 1.1575362203082986, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 38534 + }, + { + "epoch": 0.38535, + "grad_norm": 1.3465819643852155, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 38535 + }, + { + "epoch": 0.38536, + "grad_norm": 1.179267053411992, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 38536 + }, + { + "epoch": 0.38537, + "grad_norm": 1.1268308147539694, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 38537 + }, + { + "epoch": 0.38538, + "grad_norm": 1.0525739446733067, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 38538 + }, + { + "epoch": 0.38539, + "grad_norm": 1.05591515781724, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 38539 + }, + { + "epoch": 0.3854, + "grad_norm": 1.1287557488424291, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 38540 + }, + { + "epoch": 0.38541, + "grad_norm": 1.0124681371558735, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 38541 + }, + { + "epoch": 0.38542, + "grad_norm": 1.1090111579111037, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 38542 + }, + { + "epoch": 0.38543, + "grad_norm": 1.1769297918895798, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 38543 + }, + { + "epoch": 0.38544, + "grad_norm": 0.946704093189971, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 38544 + }, + { + "epoch": 0.38545, + "grad_norm": 1.1052663976026582, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 38545 + }, + { + "epoch": 0.38546, + "grad_norm": 1.1656273297912334, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 38546 + }, + { + "epoch": 0.38547, + "grad_norm": 1.1914617427614036, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 38547 + }, + { + "epoch": 0.38548, + "grad_norm": 1.0230663993593667, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 38548 + }, + { + "epoch": 0.38549, + "grad_norm": 1.1711236877992193, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 38549 + }, + { + "epoch": 0.3855, + "grad_norm": 1.0021512007307565, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 38550 + }, + { + "epoch": 0.38551, + "grad_norm": 1.119001386936165, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 38551 + }, + { + "epoch": 0.38552, + "grad_norm": 1.0183168257134587, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 38552 + }, + { + "epoch": 0.38553, + "grad_norm": 1.206372565135626, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 38553 + }, + { + "epoch": 0.38554, + "grad_norm": 1.118734773771679, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 38554 + }, + { + "epoch": 0.38555, + "grad_norm": 1.2804008156857984, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 38555 + }, + { + "epoch": 0.38556, + "grad_norm": 0.896039737431737, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 38556 + }, + { + "epoch": 0.38557, + "grad_norm": 0.8658360892207, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 38557 + }, + { + "epoch": 0.38558, + "grad_norm": 0.9852529779574895, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 38558 + }, + { + "epoch": 0.38559, + "grad_norm": 1.2102128653315407, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 38559 + }, + { + "epoch": 0.3856, + "grad_norm": 1.0784168921566273, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 38560 + }, + { + "epoch": 0.38561, + "grad_norm": 1.112477293903535, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 38561 + }, + { + "epoch": 0.38562, + "grad_norm": 1.1082333817889225, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 38562 + }, + { + "epoch": 0.38563, + "grad_norm": 1.1247431831459196, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 38563 + }, + { + "epoch": 0.38564, + "grad_norm": 1.2008810405979293, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 38564 + }, + { + "epoch": 0.38565, + "grad_norm": 1.0676469840284446, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 38565 + }, + { + "epoch": 0.38566, + "grad_norm": 0.9772534226294478, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 38566 + }, + { + "epoch": 0.38567, + "grad_norm": 1.0662455711016525, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 38567 + }, + { + "epoch": 0.38568, + "grad_norm": 1.2210026645623533, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 38568 + }, + { + "epoch": 0.38569, + "grad_norm": 0.9595484800533735, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 38569 + }, + { + "epoch": 0.3857, + "grad_norm": 1.3415198494018794, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 38570 + }, + { + "epoch": 0.38571, + "grad_norm": 0.9404045779239868, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 38571 + }, + { + "epoch": 0.38572, + "grad_norm": 0.9631733539651095, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 38572 + }, + { + "epoch": 0.38573, + "grad_norm": 0.9452802541352445, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 38573 + }, + { + "epoch": 0.38574, + "grad_norm": 1.2304154512364682, + "learning_rate": 0.003, + "loss": 4.046, + "step": 38574 + }, + { + "epoch": 0.38575, + "grad_norm": 1.172355905492865, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 38575 + }, + { + "epoch": 0.38576, + "grad_norm": 1.1637220766922467, + "learning_rate": 0.003, + "loss": 3.995, + "step": 38576 + }, + { + "epoch": 0.38577, + "grad_norm": 0.9023350425637743, + "learning_rate": 0.003, + "loss": 4.043, + "step": 38577 + }, + { + "epoch": 0.38578, + "grad_norm": 0.9877313333552988, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 38578 + }, + { + "epoch": 0.38579, + "grad_norm": 1.2089421666487523, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 38579 + }, + { + "epoch": 0.3858, + "grad_norm": 0.8871504460904813, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 38580 + }, + { + "epoch": 0.38581, + "grad_norm": 1.1374791745613673, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 38581 + }, + { + "epoch": 0.38582, + "grad_norm": 1.239684724825323, + "learning_rate": 0.003, + "loss": 4.063, + "step": 38582 + }, + { + "epoch": 0.38583, + "grad_norm": 0.8834173856474604, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 38583 + }, + { + "epoch": 0.38584, + "grad_norm": 0.9525297161729239, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 38584 + }, + { + "epoch": 0.38585, + "grad_norm": 1.0921508330528402, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 38585 + }, + { + "epoch": 0.38586, + "grad_norm": 1.0298628103615524, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 38586 + }, + { + "epoch": 0.38587, + "grad_norm": 1.0671803029135865, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 38587 + }, + { + "epoch": 0.38588, + "grad_norm": 1.1645689698122972, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 38588 + }, + { + "epoch": 0.38589, + "grad_norm": 1.0989342784836929, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 38589 + }, + { + "epoch": 0.3859, + "grad_norm": 1.287074548703676, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 38590 + }, + { + "epoch": 0.38591, + "grad_norm": 1.1460006667860043, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 38591 + }, + { + "epoch": 0.38592, + "grad_norm": 1.1667428088007092, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 38592 + }, + { + "epoch": 0.38593, + "grad_norm": 1.167023554476506, + "learning_rate": 0.003, + "loss": 4.029, + "step": 38593 + }, + { + "epoch": 0.38594, + "grad_norm": 0.9015779693447001, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 38594 + }, + { + "epoch": 0.38595, + "grad_norm": 0.9682403703604915, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 38595 + }, + { + "epoch": 0.38596, + "grad_norm": 1.1656999492219717, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 38596 + }, + { + "epoch": 0.38597, + "grad_norm": 1.0450928030137605, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 38597 + }, + { + "epoch": 0.38598, + "grad_norm": 1.1604904546637664, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 38598 + }, + { + "epoch": 0.38599, + "grad_norm": 1.1797791531650696, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 38599 + }, + { + "epoch": 0.386, + "grad_norm": 1.049701410760732, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 38600 + }, + { + "epoch": 0.38601, + "grad_norm": 1.113437972999597, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 38601 + }, + { + "epoch": 0.38602, + "grad_norm": 1.0610077035996417, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 38602 + }, + { + "epoch": 0.38603, + "grad_norm": 1.0294031672670065, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 38603 + }, + { + "epoch": 0.38604, + "grad_norm": 0.9427131258637567, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 38604 + }, + { + "epoch": 0.38605, + "grad_norm": 0.84368232597669, + "learning_rate": 0.003, + "loss": 4.036, + "step": 38605 + }, + { + "epoch": 0.38606, + "grad_norm": 1.037333817436421, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 38606 + }, + { + "epoch": 0.38607, + "grad_norm": 1.1623133124185918, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 38607 + }, + { + "epoch": 0.38608, + "grad_norm": 0.9930649006100121, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 38608 + }, + { + "epoch": 0.38609, + "grad_norm": 1.0851388492018252, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 38609 + }, + { + "epoch": 0.3861, + "grad_norm": 0.9736941902118167, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 38610 + }, + { + "epoch": 0.38611, + "grad_norm": 1.011954788912737, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 38611 + }, + { + "epoch": 0.38612, + "grad_norm": 1.3714402200122096, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 38612 + }, + { + "epoch": 0.38613, + "grad_norm": 0.9568557341252752, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 38613 + }, + { + "epoch": 0.38614, + "grad_norm": 1.1334535138109203, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 38614 + }, + { + "epoch": 0.38615, + "grad_norm": 1.0161517016396362, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 38615 + }, + { + "epoch": 0.38616, + "grad_norm": 1.4151878309676398, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 38616 + }, + { + "epoch": 0.38617, + "grad_norm": 1.0197732208869952, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 38617 + }, + { + "epoch": 0.38618, + "grad_norm": 0.9000786107918584, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 38618 + }, + { + "epoch": 0.38619, + "grad_norm": 0.8368686150021876, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 38619 + }, + { + "epoch": 0.3862, + "grad_norm": 1.1379413671919667, + "learning_rate": 0.003, + "loss": 4.068, + "step": 38620 + }, + { + "epoch": 0.38621, + "grad_norm": 1.3073706029667689, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 38621 + }, + { + "epoch": 0.38622, + "grad_norm": 0.9400163980989189, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 38622 + }, + { + "epoch": 0.38623, + "grad_norm": 1.026958866549631, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 38623 + }, + { + "epoch": 0.38624, + "grad_norm": 1.1175720468765513, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 38624 + }, + { + "epoch": 0.38625, + "grad_norm": 1.0731788233774502, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 38625 + }, + { + "epoch": 0.38626, + "grad_norm": 1.3060494521691601, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 38626 + }, + { + "epoch": 0.38627, + "grad_norm": 1.1339297504675971, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 38627 + }, + { + "epoch": 0.38628, + "grad_norm": 0.9955928954310289, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 38628 + }, + { + "epoch": 0.38629, + "grad_norm": 1.0858626171595391, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 38629 + }, + { + "epoch": 0.3863, + "grad_norm": 1.1454515265185998, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 38630 + }, + { + "epoch": 0.38631, + "grad_norm": 1.0702760383094416, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 38631 + }, + { + "epoch": 0.38632, + "grad_norm": 1.149633442851239, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 38632 + }, + { + "epoch": 0.38633, + "grad_norm": 1.1962206832741142, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 38633 + }, + { + "epoch": 0.38634, + "grad_norm": 1.0408828975726516, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 38634 + }, + { + "epoch": 0.38635, + "grad_norm": 1.1390716880865814, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 38635 + }, + { + "epoch": 0.38636, + "grad_norm": 1.011812382020061, + "learning_rate": 0.003, + "loss": 4.051, + "step": 38636 + }, + { + "epoch": 0.38637, + "grad_norm": 1.0157210172476108, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 38637 + }, + { + "epoch": 0.38638, + "grad_norm": 0.9114574787839199, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 38638 + }, + { + "epoch": 0.38639, + "grad_norm": 0.8999533312192864, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 38639 + }, + { + "epoch": 0.3864, + "grad_norm": 0.9561198964916565, + "learning_rate": 0.003, + "loss": 4.033, + "step": 38640 + }, + { + "epoch": 0.38641, + "grad_norm": 1.0467711821866814, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 38641 + }, + { + "epoch": 0.38642, + "grad_norm": 1.1022411864146302, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 38642 + }, + { + "epoch": 0.38643, + "grad_norm": 0.8992600134510487, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 38643 + }, + { + "epoch": 0.38644, + "grad_norm": 1.0319942090203893, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 38644 + }, + { + "epoch": 0.38645, + "grad_norm": 1.523567949279088, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 38645 + }, + { + "epoch": 0.38646, + "grad_norm": 0.6472082742546329, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 38646 + }, + { + "epoch": 0.38647, + "grad_norm": 0.8567120039948847, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 38647 + }, + { + "epoch": 0.38648, + "grad_norm": 1.2932626889435985, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 38648 + }, + { + "epoch": 0.38649, + "grad_norm": 0.838077848798536, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 38649 + }, + { + "epoch": 0.3865, + "grad_norm": 0.7556089129529504, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 38650 + }, + { + "epoch": 0.38651, + "grad_norm": 0.9636934485762529, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 38651 + }, + { + "epoch": 0.38652, + "grad_norm": 1.3681481374373061, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 38652 + }, + { + "epoch": 0.38653, + "grad_norm": 0.8139163699654792, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 38653 + }, + { + "epoch": 0.38654, + "grad_norm": 1.056095784984025, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 38654 + }, + { + "epoch": 0.38655, + "grad_norm": 1.2434302974938993, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 38655 + }, + { + "epoch": 0.38656, + "grad_norm": 1.0317051942466744, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 38656 + }, + { + "epoch": 0.38657, + "grad_norm": 1.1854183549513169, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 38657 + }, + { + "epoch": 0.38658, + "grad_norm": 1.2514180432253401, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 38658 + }, + { + "epoch": 0.38659, + "grad_norm": 0.901224191420163, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 38659 + }, + { + "epoch": 0.3866, + "grad_norm": 1.1904191954833765, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 38660 + }, + { + "epoch": 0.38661, + "grad_norm": 1.0775966257028038, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 38661 + }, + { + "epoch": 0.38662, + "grad_norm": 1.0932932312701653, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 38662 + }, + { + "epoch": 0.38663, + "grad_norm": 1.0603194937849028, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 38663 + }, + { + "epoch": 0.38664, + "grad_norm": 1.0931535624927589, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 38664 + }, + { + "epoch": 0.38665, + "grad_norm": 1.2131642247423284, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 38665 + }, + { + "epoch": 0.38666, + "grad_norm": 1.0820808645734667, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 38666 + }, + { + "epoch": 0.38667, + "grad_norm": 0.9878391989398602, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 38667 + }, + { + "epoch": 0.38668, + "grad_norm": 0.9276111816225423, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 38668 + }, + { + "epoch": 0.38669, + "grad_norm": 1.284371121533289, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 38669 + }, + { + "epoch": 0.3867, + "grad_norm": 1.1302794540235008, + "learning_rate": 0.003, + "loss": 4.065, + "step": 38670 + }, + { + "epoch": 0.38671, + "grad_norm": 1.3275352032567287, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 38671 + }, + { + "epoch": 0.38672, + "grad_norm": 1.1474237880837697, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 38672 + }, + { + "epoch": 0.38673, + "grad_norm": 1.09124309181528, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 38673 + }, + { + "epoch": 0.38674, + "grad_norm": 1.0357596762142829, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 38674 + }, + { + "epoch": 0.38675, + "grad_norm": 1.0267625123306563, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 38675 + }, + { + "epoch": 0.38676, + "grad_norm": 1.18326656490713, + "learning_rate": 0.003, + "loss": 4.036, + "step": 38676 + }, + { + "epoch": 0.38677, + "grad_norm": 1.0353630611795415, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 38677 + }, + { + "epoch": 0.38678, + "grad_norm": 1.369766471825508, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 38678 + }, + { + "epoch": 0.38679, + "grad_norm": 0.9814119422442764, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 38679 + }, + { + "epoch": 0.3868, + "grad_norm": 1.1055905535066297, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 38680 + }, + { + "epoch": 0.38681, + "grad_norm": 1.0239433825659006, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 38681 + }, + { + "epoch": 0.38682, + "grad_norm": 1.154545940768406, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 38682 + }, + { + "epoch": 0.38683, + "grad_norm": 1.0336783938566396, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 38683 + }, + { + "epoch": 0.38684, + "grad_norm": 1.150988002463637, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 38684 + }, + { + "epoch": 0.38685, + "grad_norm": 1.1061204238509976, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 38685 + }, + { + "epoch": 0.38686, + "grad_norm": 0.9255391790288957, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 38686 + }, + { + "epoch": 0.38687, + "grad_norm": 0.8833421325792721, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 38687 + }, + { + "epoch": 0.38688, + "grad_norm": 0.9181565658909621, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 38688 + }, + { + "epoch": 0.38689, + "grad_norm": 1.2338220585620667, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 38689 + }, + { + "epoch": 0.3869, + "grad_norm": 1.1939373547692709, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 38690 + }, + { + "epoch": 0.38691, + "grad_norm": 1.0429665866646862, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 38691 + }, + { + "epoch": 0.38692, + "grad_norm": 1.0942624142907846, + "learning_rate": 0.003, + "loss": 4.026, + "step": 38692 + }, + { + "epoch": 0.38693, + "grad_norm": 1.1831857815923932, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 38693 + }, + { + "epoch": 0.38694, + "grad_norm": 1.0739059691915287, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 38694 + }, + { + "epoch": 0.38695, + "grad_norm": 0.7723711115044983, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 38695 + }, + { + "epoch": 0.38696, + "grad_norm": 0.6990114575513825, + "learning_rate": 0.003, + "loss": 4.044, + "step": 38696 + }, + { + "epoch": 0.38697, + "grad_norm": 0.7748565287472476, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 38697 + }, + { + "epoch": 0.38698, + "grad_norm": 0.9400095133928473, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 38698 + }, + { + "epoch": 0.38699, + "grad_norm": 1.2616258441910904, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 38699 + }, + { + "epoch": 0.387, + "grad_norm": 0.7796983208217593, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 38700 + }, + { + "epoch": 0.38701, + "grad_norm": 0.8391657358498028, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 38701 + }, + { + "epoch": 0.38702, + "grad_norm": 0.8967544492419596, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 38702 + }, + { + "epoch": 0.38703, + "grad_norm": 1.1458591151004707, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 38703 + }, + { + "epoch": 0.38704, + "grad_norm": 1.2818277655693022, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 38704 + }, + { + "epoch": 0.38705, + "grad_norm": 0.8184337912800108, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 38705 + }, + { + "epoch": 0.38706, + "grad_norm": 0.8716469002563729, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 38706 + }, + { + "epoch": 0.38707, + "grad_norm": 1.29001199196172, + "learning_rate": 0.003, + "loss": 4.039, + "step": 38707 + }, + { + "epoch": 0.38708, + "grad_norm": 1.1039515386984653, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 38708 + }, + { + "epoch": 0.38709, + "grad_norm": 1.0303424062483242, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 38709 + }, + { + "epoch": 0.3871, + "grad_norm": 1.1156743970355094, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 38710 + }, + { + "epoch": 0.38711, + "grad_norm": 1.284025610695132, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 38711 + }, + { + "epoch": 0.38712, + "grad_norm": 0.979892666605223, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 38712 + }, + { + "epoch": 0.38713, + "grad_norm": 1.073308691651438, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 38713 + }, + { + "epoch": 0.38714, + "grad_norm": 1.27612490688341, + "learning_rate": 0.003, + "loss": 4.049, + "step": 38714 + }, + { + "epoch": 0.38715, + "grad_norm": 0.9201610450588059, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 38715 + }, + { + "epoch": 0.38716, + "grad_norm": 0.9277196929281021, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 38716 + }, + { + "epoch": 0.38717, + "grad_norm": 1.1350859974461174, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 38717 + }, + { + "epoch": 0.38718, + "grad_norm": 1.2933331579227656, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 38718 + }, + { + "epoch": 0.38719, + "grad_norm": 0.8354614925541244, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 38719 + }, + { + "epoch": 0.3872, + "grad_norm": 0.8848336693487878, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 38720 + }, + { + "epoch": 0.38721, + "grad_norm": 1.0417387567315226, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 38721 + }, + { + "epoch": 0.38722, + "grad_norm": 1.1559457311449357, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 38722 + }, + { + "epoch": 0.38723, + "grad_norm": 0.9914669222878395, + "learning_rate": 0.003, + "loss": 4.052, + "step": 38723 + }, + { + "epoch": 0.38724, + "grad_norm": 1.136578558799472, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 38724 + }, + { + "epoch": 0.38725, + "grad_norm": 0.965687658690405, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 38725 + }, + { + "epoch": 0.38726, + "grad_norm": 1.208590721150987, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 38726 + }, + { + "epoch": 0.38727, + "grad_norm": 1.0960035783161939, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 38727 + }, + { + "epoch": 0.38728, + "grad_norm": 1.1677710726342647, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 38728 + }, + { + "epoch": 0.38729, + "grad_norm": 1.2921904894011382, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 38729 + }, + { + "epoch": 0.3873, + "grad_norm": 1.0474620686132126, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 38730 + }, + { + "epoch": 0.38731, + "grad_norm": 0.9742991563718784, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 38731 + }, + { + "epoch": 0.38732, + "grad_norm": 1.256197882810664, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 38732 + }, + { + "epoch": 0.38733, + "grad_norm": 1.0204037593122357, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 38733 + }, + { + "epoch": 0.38734, + "grad_norm": 1.186869810772536, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 38734 + }, + { + "epoch": 0.38735, + "grad_norm": 1.227165917344482, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 38735 + }, + { + "epoch": 0.38736, + "grad_norm": 0.8030547729015161, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 38736 + }, + { + "epoch": 0.38737, + "grad_norm": 0.8536773704221247, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 38737 + }, + { + "epoch": 0.38738, + "grad_norm": 1.151899451546315, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 38738 + }, + { + "epoch": 0.38739, + "grad_norm": 1.2876784663612406, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 38739 + }, + { + "epoch": 0.3874, + "grad_norm": 1.1592912282036971, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 38740 + }, + { + "epoch": 0.38741, + "grad_norm": 1.042622430575193, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 38741 + }, + { + "epoch": 0.38742, + "grad_norm": 0.9459398940785201, + "learning_rate": 0.003, + "loss": 4.057, + "step": 38742 + }, + { + "epoch": 0.38743, + "grad_norm": 1.0880374080910482, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 38743 + }, + { + "epoch": 0.38744, + "grad_norm": 1.2014432229929883, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 38744 + }, + { + "epoch": 0.38745, + "grad_norm": 1.0242176797605185, + "learning_rate": 0.003, + "loss": 4.057, + "step": 38745 + }, + { + "epoch": 0.38746, + "grad_norm": 1.034379453027707, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 38746 + }, + { + "epoch": 0.38747, + "grad_norm": 1.0967809500052432, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 38747 + }, + { + "epoch": 0.38748, + "grad_norm": 1.3547590868850907, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 38748 + }, + { + "epoch": 0.38749, + "grad_norm": 1.0669618845043376, + "learning_rate": 0.003, + "loss": 4.005, + "step": 38749 + }, + { + "epoch": 0.3875, + "grad_norm": 1.1167104582757075, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 38750 + }, + { + "epoch": 0.38751, + "grad_norm": 1.0487797531212788, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 38751 + }, + { + "epoch": 0.38752, + "grad_norm": 0.9938981181283089, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 38752 + }, + { + "epoch": 0.38753, + "grad_norm": 1.3486702132608688, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 38753 + }, + { + "epoch": 0.38754, + "grad_norm": 0.85172587463024, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 38754 + }, + { + "epoch": 0.38755, + "grad_norm": 0.9756378349675168, + "learning_rate": 0.003, + "loss": 4.021, + "step": 38755 + }, + { + "epoch": 0.38756, + "grad_norm": 1.1862638768949993, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 38756 + }, + { + "epoch": 0.38757, + "grad_norm": 0.9423412265172554, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 38757 + }, + { + "epoch": 0.38758, + "grad_norm": 1.3264890254499606, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 38758 + }, + { + "epoch": 0.38759, + "grad_norm": 1.1153862716054443, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 38759 + }, + { + "epoch": 0.3876, + "grad_norm": 0.993894357494993, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 38760 + }, + { + "epoch": 0.38761, + "grad_norm": 1.1659584132499903, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 38761 + }, + { + "epoch": 0.38762, + "grad_norm": 0.9363831977241169, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 38762 + }, + { + "epoch": 0.38763, + "grad_norm": 1.1183800433684588, + "learning_rate": 0.003, + "loss": 4.013, + "step": 38763 + }, + { + "epoch": 0.38764, + "grad_norm": 1.1038223742946323, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 38764 + }, + { + "epoch": 0.38765, + "grad_norm": 1.069575972562539, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 38765 + }, + { + "epoch": 0.38766, + "grad_norm": 1.0651525010670866, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 38766 + }, + { + "epoch": 0.38767, + "grad_norm": 1.125127828499341, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 38767 + }, + { + "epoch": 0.38768, + "grad_norm": 1.0661582463303583, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 38768 + }, + { + "epoch": 0.38769, + "grad_norm": 1.0265376836008986, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 38769 + }, + { + "epoch": 0.3877, + "grad_norm": 1.1756034049355635, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 38770 + }, + { + "epoch": 0.38771, + "grad_norm": 1.0972309963423412, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 38771 + }, + { + "epoch": 0.38772, + "grad_norm": 1.1290716556622409, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 38772 + }, + { + "epoch": 0.38773, + "grad_norm": 0.935646326795091, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 38773 + }, + { + "epoch": 0.38774, + "grad_norm": 1.1768598324537205, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 38774 + }, + { + "epoch": 0.38775, + "grad_norm": 1.2809170221930508, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 38775 + }, + { + "epoch": 0.38776, + "grad_norm": 1.0847618004181345, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 38776 + }, + { + "epoch": 0.38777, + "grad_norm": 1.047417720776213, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 38777 + }, + { + "epoch": 0.38778, + "grad_norm": 1.0935659375115998, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 38778 + }, + { + "epoch": 0.38779, + "grad_norm": 0.933699241705118, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 38779 + }, + { + "epoch": 0.3878, + "grad_norm": 0.8726127641718461, + "learning_rate": 0.003, + "loss": 4.064, + "step": 38780 + }, + { + "epoch": 0.38781, + "grad_norm": 0.8930535366707121, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 38781 + }, + { + "epoch": 0.38782, + "grad_norm": 1.2249321591266373, + "learning_rate": 0.003, + "loss": 4.023, + "step": 38782 + }, + { + "epoch": 0.38783, + "grad_norm": 1.0940570634068456, + "learning_rate": 0.003, + "loss": 4.046, + "step": 38783 + }, + { + "epoch": 0.38784, + "grad_norm": 1.2045332742363846, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 38784 + }, + { + "epoch": 0.38785, + "grad_norm": 0.929277060599042, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 38785 + }, + { + "epoch": 0.38786, + "grad_norm": 0.9752485632955171, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 38786 + }, + { + "epoch": 0.38787, + "grad_norm": 1.0610409678453434, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 38787 + }, + { + "epoch": 0.38788, + "grad_norm": 1.3054162186358473, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 38788 + }, + { + "epoch": 0.38789, + "grad_norm": 1.101371750378466, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 38789 + }, + { + "epoch": 0.3879, + "grad_norm": 0.9612967270625817, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 38790 + }, + { + "epoch": 0.38791, + "grad_norm": 1.1593980385636253, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 38791 + }, + { + "epoch": 0.38792, + "grad_norm": 1.1125954457730223, + "learning_rate": 0.003, + "loss": 4.059, + "step": 38792 + }, + { + "epoch": 0.38793, + "grad_norm": 1.0174730826376985, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 38793 + }, + { + "epoch": 0.38794, + "grad_norm": 0.9453837715621227, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 38794 + }, + { + "epoch": 0.38795, + "grad_norm": 1.0509635866177562, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 38795 + }, + { + "epoch": 0.38796, + "grad_norm": 1.0608629798047533, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 38796 + }, + { + "epoch": 0.38797, + "grad_norm": 1.2010368021214153, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 38797 + }, + { + "epoch": 0.38798, + "grad_norm": 1.1620972741469853, + "learning_rate": 0.003, + "loss": 4.042, + "step": 38798 + }, + { + "epoch": 0.38799, + "grad_norm": 1.0612585648351993, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 38799 + }, + { + "epoch": 0.388, + "grad_norm": 0.8857254780081482, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 38800 + }, + { + "epoch": 0.38801, + "grad_norm": 0.8348044092705929, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 38801 + }, + { + "epoch": 0.38802, + "grad_norm": 0.8336024509789529, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 38802 + }, + { + "epoch": 0.38803, + "grad_norm": 0.9034316447687456, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 38803 + }, + { + "epoch": 0.38804, + "grad_norm": 0.9884858114052156, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 38804 + }, + { + "epoch": 0.38805, + "grad_norm": 1.39478176983987, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 38805 + }, + { + "epoch": 0.38806, + "grad_norm": 0.9184829642696667, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 38806 + }, + { + "epoch": 0.38807, + "grad_norm": 1.1894184074654905, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 38807 + }, + { + "epoch": 0.38808, + "grad_norm": 1.0208280348082515, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 38808 + }, + { + "epoch": 0.38809, + "grad_norm": 1.1088085452996315, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 38809 + }, + { + "epoch": 0.3881, + "grad_norm": 1.2610297931698324, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 38810 + }, + { + "epoch": 0.38811, + "grad_norm": 0.9033492713931802, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 38811 + }, + { + "epoch": 0.38812, + "grad_norm": 1.0177830326717912, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 38812 + }, + { + "epoch": 0.38813, + "grad_norm": 1.2688907032211711, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 38813 + }, + { + "epoch": 0.38814, + "grad_norm": 1.0307130726377638, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 38814 + }, + { + "epoch": 0.38815, + "grad_norm": 1.1876558289145067, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 38815 + }, + { + "epoch": 0.38816, + "grad_norm": 1.0168416697383458, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 38816 + }, + { + "epoch": 0.38817, + "grad_norm": 1.3285344195470474, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 38817 + }, + { + "epoch": 0.38818, + "grad_norm": 0.7715294094854522, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 38818 + }, + { + "epoch": 0.38819, + "grad_norm": 0.7775520593579708, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 38819 + }, + { + "epoch": 0.3882, + "grad_norm": 1.0785951106214962, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 38820 + }, + { + "epoch": 0.38821, + "grad_norm": 1.221580074615974, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 38821 + }, + { + "epoch": 0.38822, + "grad_norm": 0.8259684102669669, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 38822 + }, + { + "epoch": 0.38823, + "grad_norm": 1.3645608755182637, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 38823 + }, + { + "epoch": 0.38824, + "grad_norm": 1.236514540325314, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 38824 + }, + { + "epoch": 0.38825, + "grad_norm": 0.9827411695998356, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 38825 + }, + { + "epoch": 0.38826, + "grad_norm": 1.0009462664488713, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 38826 + }, + { + "epoch": 0.38827, + "grad_norm": 1.281197644724663, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 38827 + }, + { + "epoch": 0.38828, + "grad_norm": 1.0235402174337898, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 38828 + }, + { + "epoch": 0.38829, + "grad_norm": 1.0884123551726483, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 38829 + }, + { + "epoch": 0.3883, + "grad_norm": 0.9729227877026173, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 38830 + }, + { + "epoch": 0.38831, + "grad_norm": 1.1139931232881903, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 38831 + }, + { + "epoch": 0.38832, + "grad_norm": 1.1670702599074085, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 38832 + }, + { + "epoch": 0.38833, + "grad_norm": 0.9982774880299176, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 38833 + }, + { + "epoch": 0.38834, + "grad_norm": 1.0154352672154314, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 38834 + }, + { + "epoch": 0.38835, + "grad_norm": 1.0977145618139565, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 38835 + }, + { + "epoch": 0.38836, + "grad_norm": 1.4364055202752466, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 38836 + }, + { + "epoch": 0.38837, + "grad_norm": 0.926761559280869, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 38837 + }, + { + "epoch": 0.38838, + "grad_norm": 1.0514611916110508, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 38838 + }, + { + "epoch": 0.38839, + "grad_norm": 1.112375993534838, + "learning_rate": 0.003, + "loss": 4.05, + "step": 38839 + }, + { + "epoch": 0.3884, + "grad_norm": 1.223228758991962, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 38840 + }, + { + "epoch": 0.38841, + "grad_norm": 0.9087641795313198, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 38841 + }, + { + "epoch": 0.38842, + "grad_norm": 1.071530319262491, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 38842 + }, + { + "epoch": 0.38843, + "grad_norm": 1.2565802518381486, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 38843 + }, + { + "epoch": 0.38844, + "grad_norm": 1.1023277663225977, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 38844 + }, + { + "epoch": 0.38845, + "grad_norm": 1.0065305132915436, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 38845 + }, + { + "epoch": 0.38846, + "grad_norm": 0.9788847943276997, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 38846 + }, + { + "epoch": 0.38847, + "grad_norm": 1.1460525356589286, + "learning_rate": 0.003, + "loss": 4.042, + "step": 38847 + }, + { + "epoch": 0.38848, + "grad_norm": 1.2897153427888934, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 38848 + }, + { + "epoch": 0.38849, + "grad_norm": 0.795980336352499, + "learning_rate": 0.003, + "loss": 4.023, + "step": 38849 + }, + { + "epoch": 0.3885, + "grad_norm": 0.8043535409612209, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 38850 + }, + { + "epoch": 0.38851, + "grad_norm": 0.9963171230591236, + "learning_rate": 0.003, + "loss": 4.024, + "step": 38851 + }, + { + "epoch": 0.38852, + "grad_norm": 1.2184079977746212, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 38852 + }, + { + "epoch": 0.38853, + "grad_norm": 1.082414156136951, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 38853 + }, + { + "epoch": 0.38854, + "grad_norm": 1.3392210982712176, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 38854 + }, + { + "epoch": 0.38855, + "grad_norm": 0.7825508793906654, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 38855 + }, + { + "epoch": 0.38856, + "grad_norm": 0.9271443874866526, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 38856 + }, + { + "epoch": 0.38857, + "grad_norm": 1.1700628016269237, + "learning_rate": 0.003, + "loss": 4.039, + "step": 38857 + }, + { + "epoch": 0.38858, + "grad_norm": 0.9473373026959495, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 38858 + }, + { + "epoch": 0.38859, + "grad_norm": 1.1560288744648337, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 38859 + }, + { + "epoch": 0.3886, + "grad_norm": 1.0107366369155353, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 38860 + }, + { + "epoch": 0.38861, + "grad_norm": 1.0740415176269595, + "learning_rate": 0.003, + "loss": 4.025, + "step": 38861 + }, + { + "epoch": 0.38862, + "grad_norm": 1.1641170863139065, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 38862 + }, + { + "epoch": 0.38863, + "grad_norm": 1.0993180912407636, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 38863 + }, + { + "epoch": 0.38864, + "grad_norm": 1.0465356938128327, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 38864 + }, + { + "epoch": 0.38865, + "grad_norm": 1.1390063550053815, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 38865 + }, + { + "epoch": 0.38866, + "grad_norm": 1.078689906899154, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 38866 + }, + { + "epoch": 0.38867, + "grad_norm": 1.16991635212761, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 38867 + }, + { + "epoch": 0.38868, + "grad_norm": 0.990161606121034, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 38868 + }, + { + "epoch": 0.38869, + "grad_norm": 0.9602020285362679, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 38869 + }, + { + "epoch": 0.3887, + "grad_norm": 0.9992024450866367, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 38870 + }, + { + "epoch": 0.38871, + "grad_norm": 1.115026199887834, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 38871 + }, + { + "epoch": 0.38872, + "grad_norm": 1.2203869858800591, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 38872 + }, + { + "epoch": 0.38873, + "grad_norm": 1.2349588298914482, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 38873 + }, + { + "epoch": 0.38874, + "grad_norm": 1.2684064412087914, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 38874 + }, + { + "epoch": 0.38875, + "grad_norm": 1.0179956384876505, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 38875 + }, + { + "epoch": 0.38876, + "grad_norm": 1.1615384572868612, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 38876 + }, + { + "epoch": 0.38877, + "grad_norm": 1.161932143867455, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 38877 + }, + { + "epoch": 0.38878, + "grad_norm": 0.9180215395678407, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 38878 + }, + { + "epoch": 0.38879, + "grad_norm": 1.016274541331993, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 38879 + }, + { + "epoch": 0.3888, + "grad_norm": 1.3889336587219752, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 38880 + }, + { + "epoch": 0.38881, + "grad_norm": 0.8631364109522247, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 38881 + }, + { + "epoch": 0.38882, + "grad_norm": 0.8293563063599639, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 38882 + }, + { + "epoch": 0.38883, + "grad_norm": 1.0688652337310407, + "learning_rate": 0.003, + "loss": 4.052, + "step": 38883 + }, + { + "epoch": 0.38884, + "grad_norm": 1.1831489942977205, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 38884 + }, + { + "epoch": 0.38885, + "grad_norm": 0.9864816623886878, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 38885 + }, + { + "epoch": 0.38886, + "grad_norm": 1.009157574721923, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 38886 + }, + { + "epoch": 0.38887, + "grad_norm": 1.0388088750602704, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 38887 + }, + { + "epoch": 0.38888, + "grad_norm": 1.0032312888756956, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 38888 + }, + { + "epoch": 0.38889, + "grad_norm": 1.0411525643263402, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 38889 + }, + { + "epoch": 0.3889, + "grad_norm": 1.0791492403480447, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 38890 + }, + { + "epoch": 0.38891, + "grad_norm": 1.1026849967720302, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 38891 + }, + { + "epoch": 0.38892, + "grad_norm": 0.9424735698088533, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 38892 + }, + { + "epoch": 0.38893, + "grad_norm": 1.0288705633047082, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 38893 + }, + { + "epoch": 0.38894, + "grad_norm": 1.0680955878933356, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 38894 + }, + { + "epoch": 0.38895, + "grad_norm": 0.8325737044479854, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 38895 + }, + { + "epoch": 0.38896, + "grad_norm": 0.8403926332463153, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 38896 + }, + { + "epoch": 0.38897, + "grad_norm": 1.0164753995809803, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 38897 + }, + { + "epoch": 0.38898, + "grad_norm": 0.9609879761036147, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 38898 + }, + { + "epoch": 0.38899, + "grad_norm": 1.2649494612434926, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 38899 + }, + { + "epoch": 0.389, + "grad_norm": 0.9431840411855794, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 38900 + }, + { + "epoch": 0.38901, + "grad_norm": 1.1820109698558596, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 38901 + }, + { + "epoch": 0.38902, + "grad_norm": 1.2379304786934058, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 38902 + }, + { + "epoch": 0.38903, + "grad_norm": 1.2328124719990075, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 38903 + }, + { + "epoch": 0.38904, + "grad_norm": 1.0874381592731244, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 38904 + }, + { + "epoch": 0.38905, + "grad_norm": 1.1224586060537964, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 38905 + }, + { + "epoch": 0.38906, + "grad_norm": 1.1105154849330146, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 38906 + }, + { + "epoch": 0.38907, + "grad_norm": 1.0701322961177253, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 38907 + }, + { + "epoch": 0.38908, + "grad_norm": 1.0679862316565336, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 38908 + }, + { + "epoch": 0.38909, + "grad_norm": 0.9158693537410505, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 38909 + }, + { + "epoch": 0.3891, + "grad_norm": 1.0560484818156008, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 38910 + }, + { + "epoch": 0.38911, + "grad_norm": 1.070213684656132, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 38911 + }, + { + "epoch": 0.38912, + "grad_norm": 1.3444785765722977, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 38912 + }, + { + "epoch": 0.38913, + "grad_norm": 1.0342036203248448, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 38913 + }, + { + "epoch": 0.38914, + "grad_norm": 1.027592545539338, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 38914 + }, + { + "epoch": 0.38915, + "grad_norm": 0.9505284050789481, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 38915 + }, + { + "epoch": 0.38916, + "grad_norm": 1.1683284709287245, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 38916 + }, + { + "epoch": 0.38917, + "grad_norm": 0.902214409483891, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 38917 + }, + { + "epoch": 0.38918, + "grad_norm": 1.1244870333838395, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 38918 + }, + { + "epoch": 0.38919, + "grad_norm": 1.0664229353097605, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 38919 + }, + { + "epoch": 0.3892, + "grad_norm": 1.0667861949256001, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 38920 + }, + { + "epoch": 0.38921, + "grad_norm": 1.0486903960081715, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 38921 + }, + { + "epoch": 0.38922, + "grad_norm": 1.0976361719123986, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 38922 + }, + { + "epoch": 0.38923, + "grad_norm": 1.3235248073904384, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 38923 + }, + { + "epoch": 0.38924, + "grad_norm": 0.9272502871237421, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 38924 + }, + { + "epoch": 0.38925, + "grad_norm": 1.057073030178168, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 38925 + }, + { + "epoch": 0.38926, + "grad_norm": 1.1050130033361456, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 38926 + }, + { + "epoch": 0.38927, + "grad_norm": 1.0594480322958038, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 38927 + }, + { + "epoch": 0.38928, + "grad_norm": 1.6950772002178431, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 38928 + }, + { + "epoch": 0.38929, + "grad_norm": 0.6966845215265857, + "learning_rate": 0.003, + "loss": 4.053, + "step": 38929 + }, + { + "epoch": 0.3893, + "grad_norm": 0.9525315147091613, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 38930 + }, + { + "epoch": 0.38931, + "grad_norm": 1.1531034678674106, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 38931 + }, + { + "epoch": 0.38932, + "grad_norm": 0.9564917569398045, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 38932 + }, + { + "epoch": 0.38933, + "grad_norm": 1.0721274620856773, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 38933 + }, + { + "epoch": 0.38934, + "grad_norm": 1.2582799192166116, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 38934 + }, + { + "epoch": 0.38935, + "grad_norm": 0.911441083250517, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 38935 + }, + { + "epoch": 0.38936, + "grad_norm": 0.8330655689314437, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 38936 + }, + { + "epoch": 0.38937, + "grad_norm": 0.9237855193741675, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 38937 + }, + { + "epoch": 0.38938, + "grad_norm": 1.059330719471224, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 38938 + }, + { + "epoch": 0.38939, + "grad_norm": 1.0096003919359517, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 38939 + }, + { + "epoch": 0.3894, + "grad_norm": 1.1755446782059458, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 38940 + }, + { + "epoch": 0.38941, + "grad_norm": 0.9902379031166141, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 38941 + }, + { + "epoch": 0.38942, + "grad_norm": 1.1165397746584511, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 38942 + }, + { + "epoch": 0.38943, + "grad_norm": 1.1132779174766743, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 38943 + }, + { + "epoch": 0.38944, + "grad_norm": 0.9762936296670968, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 38944 + }, + { + "epoch": 0.38945, + "grad_norm": 1.2329993929691785, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 38945 + }, + { + "epoch": 0.38946, + "grad_norm": 1.2692280720515448, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 38946 + }, + { + "epoch": 0.38947, + "grad_norm": 1.1633889270816204, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 38947 + }, + { + "epoch": 0.38948, + "grad_norm": 1.2777153535533712, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 38948 + }, + { + "epoch": 0.38949, + "grad_norm": 1.005299869673439, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 38949 + }, + { + "epoch": 0.3895, + "grad_norm": 1.302731594870699, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 38950 + }, + { + "epoch": 0.38951, + "grad_norm": 1.010776075306977, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 38951 + }, + { + "epoch": 0.38952, + "grad_norm": 1.0732440559207295, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 38952 + }, + { + "epoch": 0.38953, + "grad_norm": 1.005935201954132, + "learning_rate": 0.003, + "loss": 4.004, + "step": 38953 + }, + { + "epoch": 0.38954, + "grad_norm": 1.0339675019679444, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 38954 + }, + { + "epoch": 0.38955, + "grad_norm": 0.9548439817932837, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 38955 + }, + { + "epoch": 0.38956, + "grad_norm": 1.0231295759399852, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 38956 + }, + { + "epoch": 0.38957, + "grad_norm": 0.9684159390316631, + "learning_rate": 0.003, + "loss": 4.063, + "step": 38957 + }, + { + "epoch": 0.38958, + "grad_norm": 1.415696376136498, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 38958 + }, + { + "epoch": 0.38959, + "grad_norm": 1.003428410022663, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 38959 + }, + { + "epoch": 0.3896, + "grad_norm": 1.2039666003996858, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 38960 + }, + { + "epoch": 0.38961, + "grad_norm": 0.9171808395503714, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 38961 + }, + { + "epoch": 0.38962, + "grad_norm": 1.1345209655548887, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 38962 + }, + { + "epoch": 0.38963, + "grad_norm": 1.4019554128553047, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 38963 + }, + { + "epoch": 0.38964, + "grad_norm": 0.815466989413752, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 38964 + }, + { + "epoch": 0.38965, + "grad_norm": 0.8471409084678595, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 38965 + }, + { + "epoch": 0.38966, + "grad_norm": 1.250276034216458, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 38966 + }, + { + "epoch": 0.38967, + "grad_norm": 1.4655443928148115, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 38967 + }, + { + "epoch": 0.38968, + "grad_norm": 0.7522735522103795, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 38968 + }, + { + "epoch": 0.38969, + "grad_norm": 0.7091080022178023, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 38969 + }, + { + "epoch": 0.3897, + "grad_norm": 1.0157137369519706, + "learning_rate": 0.003, + "loss": 4.037, + "step": 38970 + }, + { + "epoch": 0.38971, + "grad_norm": 1.3941524456872934, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 38971 + }, + { + "epoch": 0.38972, + "grad_norm": 1.2826357784433076, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 38972 + }, + { + "epoch": 0.38973, + "grad_norm": 0.8496387181281295, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 38973 + }, + { + "epoch": 0.38974, + "grad_norm": 0.8698105654243733, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 38974 + }, + { + "epoch": 0.38975, + "grad_norm": 1.173978521257272, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 38975 + }, + { + "epoch": 0.38976, + "grad_norm": 0.8896763724769233, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 38976 + }, + { + "epoch": 0.38977, + "grad_norm": 1.0292822398882242, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 38977 + }, + { + "epoch": 0.38978, + "grad_norm": 0.9706842685710119, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 38978 + }, + { + "epoch": 0.38979, + "grad_norm": 0.935531104261579, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 38979 + }, + { + "epoch": 0.3898, + "grad_norm": 0.9561193729912046, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 38980 + }, + { + "epoch": 0.38981, + "grad_norm": 1.0606587756921988, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 38981 + }, + { + "epoch": 0.38982, + "grad_norm": 1.0168004341871384, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 38982 + }, + { + "epoch": 0.38983, + "grad_norm": 1.5724280700583901, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 38983 + }, + { + "epoch": 0.38984, + "grad_norm": 0.9041557183251935, + "learning_rate": 0.003, + "loss": 4.032, + "step": 38984 + }, + { + "epoch": 0.38985, + "grad_norm": 1.0159860493312518, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 38985 + }, + { + "epoch": 0.38986, + "grad_norm": 1.0896717964824478, + "learning_rate": 0.003, + "loss": 4.048, + "step": 38986 + }, + { + "epoch": 0.38987, + "grad_norm": 1.376131375809585, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 38987 + }, + { + "epoch": 0.38988, + "grad_norm": 1.310852409930925, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 38988 + }, + { + "epoch": 0.38989, + "grad_norm": 0.7953711653133146, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 38989 + }, + { + "epoch": 0.3899, + "grad_norm": 0.6082871969734983, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 38990 + }, + { + "epoch": 0.38991, + "grad_norm": 0.7860625970164894, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 38991 + }, + { + "epoch": 0.38992, + "grad_norm": 1.10970365059547, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 38992 + }, + { + "epoch": 0.38993, + "grad_norm": 1.4394162651942792, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 38993 + }, + { + "epoch": 0.38994, + "grad_norm": 0.706865330463353, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 38994 + }, + { + "epoch": 0.38995, + "grad_norm": 0.7139068703964091, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 38995 + }, + { + "epoch": 0.38996, + "grad_norm": 0.7545220255320295, + "learning_rate": 0.003, + "loss": 4.036, + "step": 38996 + }, + { + "epoch": 0.38997, + "grad_norm": 0.8928269541654051, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 38997 + }, + { + "epoch": 0.38998, + "grad_norm": 1.0668302371183984, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 38998 + }, + { + "epoch": 0.38999, + "grad_norm": 1.3878522772455066, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 38999 + }, + { + "epoch": 0.39, + "grad_norm": 0.9559998381738452, + "learning_rate": 0.003, + "loss": 4.021, + "step": 39000 + }, + { + "epoch": 0.39001, + "grad_norm": 1.0237940094593925, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 39001 + }, + { + "epoch": 0.39002, + "grad_norm": 1.1734862919759799, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 39002 + }, + { + "epoch": 0.39003, + "grad_norm": 1.0196225772987817, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 39003 + }, + { + "epoch": 0.39004, + "grad_norm": 1.3943314089069807, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 39004 + }, + { + "epoch": 0.39005, + "grad_norm": 0.9668618851731021, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 39005 + }, + { + "epoch": 0.39006, + "grad_norm": 1.0279763406186873, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 39006 + }, + { + "epoch": 0.39007, + "grad_norm": 1.1356152858876212, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 39007 + }, + { + "epoch": 0.39008, + "grad_norm": 1.1173572697545755, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 39008 + }, + { + "epoch": 0.39009, + "grad_norm": 1.1369206027208243, + "learning_rate": 0.003, + "loss": 4.026, + "step": 39009 + }, + { + "epoch": 0.3901, + "grad_norm": 0.9447665502068827, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 39010 + }, + { + "epoch": 0.39011, + "grad_norm": 0.985358306539251, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 39011 + }, + { + "epoch": 0.39012, + "grad_norm": 0.9616684508317537, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 39012 + }, + { + "epoch": 0.39013, + "grad_norm": 1.151227905994873, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 39013 + }, + { + "epoch": 0.39014, + "grad_norm": 1.1284970355241268, + "learning_rate": 0.003, + "loss": 4.052, + "step": 39014 + }, + { + "epoch": 0.39015, + "grad_norm": 1.0075560726047192, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 39015 + }, + { + "epoch": 0.39016, + "grad_norm": 1.068586777116123, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 39016 + }, + { + "epoch": 0.39017, + "grad_norm": 1.0954500406472316, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 39017 + }, + { + "epoch": 0.39018, + "grad_norm": 1.118562846006535, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 39018 + }, + { + "epoch": 0.39019, + "grad_norm": 1.1529225631539333, + "learning_rate": 0.003, + "loss": 4.022, + "step": 39019 + }, + { + "epoch": 0.3902, + "grad_norm": 1.152779112477152, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 39020 + }, + { + "epoch": 0.39021, + "grad_norm": 1.1526785461356697, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 39021 + }, + { + "epoch": 0.39022, + "grad_norm": 1.049803317431784, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 39022 + }, + { + "epoch": 0.39023, + "grad_norm": 0.9635867216662075, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 39023 + }, + { + "epoch": 0.39024, + "grad_norm": 0.9133349327974815, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 39024 + }, + { + "epoch": 0.39025, + "grad_norm": 0.9670236596418559, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 39025 + }, + { + "epoch": 0.39026, + "grad_norm": 1.052359912517174, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 39026 + }, + { + "epoch": 0.39027, + "grad_norm": 1.0877204875349218, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 39027 + }, + { + "epoch": 0.39028, + "grad_norm": 1.2117001806865344, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 39028 + }, + { + "epoch": 0.39029, + "grad_norm": 1.0368822027736888, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 39029 + }, + { + "epoch": 0.3903, + "grad_norm": 1.1122576488196303, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 39030 + }, + { + "epoch": 0.39031, + "grad_norm": 1.22659811990041, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 39031 + }, + { + "epoch": 0.39032, + "grad_norm": 0.9672116286986376, + "learning_rate": 0.003, + "loss": 4.039, + "step": 39032 + }, + { + "epoch": 0.39033, + "grad_norm": 1.200494719824246, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 39033 + }, + { + "epoch": 0.39034, + "grad_norm": 1.0014479259520384, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 39034 + }, + { + "epoch": 0.39035, + "grad_norm": 1.430229478181485, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 39035 + }, + { + "epoch": 0.39036, + "grad_norm": 0.8187482510341242, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 39036 + }, + { + "epoch": 0.39037, + "grad_norm": 0.7693559973053548, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 39037 + }, + { + "epoch": 0.39038, + "grad_norm": 0.8942893136106522, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 39038 + }, + { + "epoch": 0.39039, + "grad_norm": 1.3965063029840703, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 39039 + }, + { + "epoch": 0.3904, + "grad_norm": 1.227210970122378, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 39040 + }, + { + "epoch": 0.39041, + "grad_norm": 0.8027894102537566, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 39041 + }, + { + "epoch": 0.39042, + "grad_norm": 0.8205787566734248, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 39042 + }, + { + "epoch": 0.39043, + "grad_norm": 0.9252974490983829, + "learning_rate": 0.003, + "loss": 4.042, + "step": 39043 + }, + { + "epoch": 0.39044, + "grad_norm": 1.291851104328029, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 39044 + }, + { + "epoch": 0.39045, + "grad_norm": 0.9220378443702995, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 39045 + }, + { + "epoch": 0.39046, + "grad_norm": 1.12144007550576, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 39046 + }, + { + "epoch": 0.39047, + "grad_norm": 1.297055916176243, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 39047 + }, + { + "epoch": 0.39048, + "grad_norm": 1.0450459009370061, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 39048 + }, + { + "epoch": 0.39049, + "grad_norm": 0.937215653320154, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 39049 + }, + { + "epoch": 0.3905, + "grad_norm": 0.999268548743829, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 39050 + }, + { + "epoch": 0.39051, + "grad_norm": 1.2742548588429143, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 39051 + }, + { + "epoch": 0.39052, + "grad_norm": 0.94720952769202, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 39052 + }, + { + "epoch": 0.39053, + "grad_norm": 1.114962008012982, + "learning_rate": 0.003, + "loss": 4.038, + "step": 39053 + }, + { + "epoch": 0.39054, + "grad_norm": 1.0584020804811245, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 39054 + }, + { + "epoch": 0.39055, + "grad_norm": 0.9887583510129117, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 39055 + }, + { + "epoch": 0.39056, + "grad_norm": 1.1509694130052637, + "learning_rate": 0.003, + "loss": 4.011, + "step": 39056 + }, + { + "epoch": 0.39057, + "grad_norm": 0.9534319128033002, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 39057 + }, + { + "epoch": 0.39058, + "grad_norm": 0.9377809398244366, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 39058 + }, + { + "epoch": 0.39059, + "grad_norm": 1.130019170067274, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 39059 + }, + { + "epoch": 0.3906, + "grad_norm": 1.043065794870807, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 39060 + }, + { + "epoch": 0.39061, + "grad_norm": 1.2718124769401187, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 39061 + }, + { + "epoch": 0.39062, + "grad_norm": 0.9226363459547848, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 39062 + }, + { + "epoch": 0.39063, + "grad_norm": 1.0582683113097842, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 39063 + }, + { + "epoch": 0.39064, + "grad_norm": 1.4078628099850754, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 39064 + }, + { + "epoch": 0.39065, + "grad_norm": 0.7086195066194975, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 39065 + }, + { + "epoch": 0.39066, + "grad_norm": 0.7641165447584206, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 39066 + }, + { + "epoch": 0.39067, + "grad_norm": 0.7695756813314002, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 39067 + }, + { + "epoch": 0.39068, + "grad_norm": 0.9864044224385264, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 39068 + }, + { + "epoch": 0.39069, + "grad_norm": 1.3492057438441072, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 39069 + }, + { + "epoch": 0.3907, + "grad_norm": 0.7931373984748378, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 39070 + }, + { + "epoch": 0.39071, + "grad_norm": 1.0549904277617648, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 39071 + }, + { + "epoch": 0.39072, + "grad_norm": 1.3673791573674146, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 39072 + }, + { + "epoch": 0.39073, + "grad_norm": 0.9113329730469768, + "learning_rate": 0.003, + "loss": 3.99, + "step": 39073 + }, + { + "epoch": 0.39074, + "grad_norm": 1.164095070962161, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 39074 + }, + { + "epoch": 0.39075, + "grad_norm": 1.3972571400046692, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 39075 + }, + { + "epoch": 0.39076, + "grad_norm": 0.987602508168385, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 39076 + }, + { + "epoch": 0.39077, + "grad_norm": 1.083070439909679, + "learning_rate": 0.003, + "loss": 4.049, + "step": 39077 + }, + { + "epoch": 0.39078, + "grad_norm": 1.2067147211387375, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 39078 + }, + { + "epoch": 0.39079, + "grad_norm": 1.1826923236051887, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 39079 + }, + { + "epoch": 0.3908, + "grad_norm": 1.1301822376096518, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 39080 + }, + { + "epoch": 0.39081, + "grad_norm": 1.1709630083637856, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 39081 + }, + { + "epoch": 0.39082, + "grad_norm": 1.118356685608785, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 39082 + }, + { + "epoch": 0.39083, + "grad_norm": 1.009735201006852, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 39083 + }, + { + "epoch": 0.39084, + "grad_norm": 1.0620947627033681, + "learning_rate": 0.003, + "loss": 4.054, + "step": 39084 + }, + { + "epoch": 0.39085, + "grad_norm": 1.006773926945385, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 39085 + }, + { + "epoch": 0.39086, + "grad_norm": 1.0672024864482843, + "learning_rate": 0.003, + "loss": 4.065, + "step": 39086 + }, + { + "epoch": 0.39087, + "grad_norm": 1.0997558520713413, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 39087 + }, + { + "epoch": 0.39088, + "grad_norm": 1.119742797701056, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 39088 + }, + { + "epoch": 0.39089, + "grad_norm": 0.9980480034411493, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 39089 + }, + { + "epoch": 0.3909, + "grad_norm": 1.1653829906803255, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 39090 + }, + { + "epoch": 0.39091, + "grad_norm": 1.1593604752882793, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 39091 + }, + { + "epoch": 0.39092, + "grad_norm": 1.0841579589826476, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 39092 + }, + { + "epoch": 0.39093, + "grad_norm": 0.940599281506768, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 39093 + }, + { + "epoch": 0.39094, + "grad_norm": 0.9650852255225177, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 39094 + }, + { + "epoch": 0.39095, + "grad_norm": 1.0445174602751073, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 39095 + }, + { + "epoch": 0.39096, + "grad_norm": 1.1206010752124524, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 39096 + }, + { + "epoch": 0.39097, + "grad_norm": 0.9926201035410968, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 39097 + }, + { + "epoch": 0.39098, + "grad_norm": 1.0003624677372582, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 39098 + }, + { + "epoch": 0.39099, + "grad_norm": 0.9977615532032472, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 39099 + }, + { + "epoch": 0.391, + "grad_norm": 0.918054139951152, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 39100 + }, + { + "epoch": 0.39101, + "grad_norm": 1.1431718688286114, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 39101 + }, + { + "epoch": 0.39102, + "grad_norm": 1.11445232837181, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 39102 + }, + { + "epoch": 0.39103, + "grad_norm": 1.0383531150303102, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 39103 + }, + { + "epoch": 0.39104, + "grad_norm": 1.025421106038541, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 39104 + }, + { + "epoch": 0.39105, + "grad_norm": 1.592306643252765, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 39105 + }, + { + "epoch": 0.39106, + "grad_norm": 0.9712585206767673, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 39106 + }, + { + "epoch": 0.39107, + "grad_norm": 1.0481461801350078, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 39107 + }, + { + "epoch": 0.39108, + "grad_norm": 1.088048356160132, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 39108 + }, + { + "epoch": 0.39109, + "grad_norm": 1.4528103743133605, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 39109 + }, + { + "epoch": 0.3911, + "grad_norm": 1.1183520793048738, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 39110 + }, + { + "epoch": 0.39111, + "grad_norm": 0.9312242475811744, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 39111 + }, + { + "epoch": 0.39112, + "grad_norm": 1.022575286718037, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 39112 + }, + { + "epoch": 0.39113, + "grad_norm": 1.2586720324000817, + "learning_rate": 0.003, + "loss": 4.038, + "step": 39113 + }, + { + "epoch": 0.39114, + "grad_norm": 0.8892515956272936, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 39114 + }, + { + "epoch": 0.39115, + "grad_norm": 1.0955491631637626, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 39115 + }, + { + "epoch": 0.39116, + "grad_norm": 1.36699679052604, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 39116 + }, + { + "epoch": 0.39117, + "grad_norm": 0.8707382189381672, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 39117 + }, + { + "epoch": 0.39118, + "grad_norm": 0.784867099987934, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 39118 + }, + { + "epoch": 0.39119, + "grad_norm": 0.9030832633281719, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 39119 + }, + { + "epoch": 0.3912, + "grad_norm": 1.2490830469750347, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 39120 + }, + { + "epoch": 0.39121, + "grad_norm": 0.9615758353292322, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 39121 + }, + { + "epoch": 0.39122, + "grad_norm": 1.3690481469042743, + "learning_rate": 0.003, + "loss": 4.037, + "step": 39122 + }, + { + "epoch": 0.39123, + "grad_norm": 0.8893633483603675, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 39123 + }, + { + "epoch": 0.39124, + "grad_norm": 1.0297771096231654, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 39124 + }, + { + "epoch": 0.39125, + "grad_norm": 1.151786533497205, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 39125 + }, + { + "epoch": 0.39126, + "grad_norm": 1.168894047027127, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 39126 + }, + { + "epoch": 0.39127, + "grad_norm": 1.3040748978457608, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 39127 + }, + { + "epoch": 0.39128, + "grad_norm": 1.0182632299438024, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 39128 + }, + { + "epoch": 0.39129, + "grad_norm": 1.1266186659730204, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 39129 + }, + { + "epoch": 0.3913, + "grad_norm": 1.1568725166508498, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 39130 + }, + { + "epoch": 0.39131, + "grad_norm": 1.075245331418054, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 39131 + }, + { + "epoch": 0.39132, + "grad_norm": 1.0284931123503234, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 39132 + }, + { + "epoch": 0.39133, + "grad_norm": 1.277507362654957, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 39133 + }, + { + "epoch": 0.39134, + "grad_norm": 1.01753637845435, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 39134 + }, + { + "epoch": 0.39135, + "grad_norm": 0.9827403902440973, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 39135 + }, + { + "epoch": 0.39136, + "grad_norm": 0.9388075195477151, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 39136 + }, + { + "epoch": 0.39137, + "grad_norm": 0.8485924109014807, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 39137 + }, + { + "epoch": 0.39138, + "grad_norm": 1.0156990519206717, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 39138 + }, + { + "epoch": 0.39139, + "grad_norm": 1.3942790380972008, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 39139 + }, + { + "epoch": 0.3914, + "grad_norm": 0.8444009651116594, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 39140 + }, + { + "epoch": 0.39141, + "grad_norm": 0.8500929135624559, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 39141 + }, + { + "epoch": 0.39142, + "grad_norm": 0.8005737017103461, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 39142 + }, + { + "epoch": 0.39143, + "grad_norm": 0.875570421803425, + "learning_rate": 0.003, + "loss": 4.029, + "step": 39143 + }, + { + "epoch": 0.39144, + "grad_norm": 1.0352905816261406, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 39144 + }, + { + "epoch": 0.39145, + "grad_norm": 1.3374702505329086, + "learning_rate": 0.003, + "loss": 4.06, + "step": 39145 + }, + { + "epoch": 0.39146, + "grad_norm": 1.0566516065384464, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 39146 + }, + { + "epoch": 0.39147, + "grad_norm": 1.1009297696747333, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 39147 + }, + { + "epoch": 0.39148, + "grad_norm": 1.1132172664313777, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 39148 + }, + { + "epoch": 0.39149, + "grad_norm": 1.1948319870868982, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 39149 + }, + { + "epoch": 0.3915, + "grad_norm": 0.9500892922963923, + "learning_rate": 0.003, + "loss": 4.053, + "step": 39150 + }, + { + "epoch": 0.39151, + "grad_norm": 0.8302019704858965, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 39151 + }, + { + "epoch": 0.39152, + "grad_norm": 1.1681213494596452, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 39152 + }, + { + "epoch": 0.39153, + "grad_norm": 1.2608035964790894, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 39153 + }, + { + "epoch": 0.39154, + "grad_norm": 0.8920496821755293, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 39154 + }, + { + "epoch": 0.39155, + "grad_norm": 1.0793769854326394, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 39155 + }, + { + "epoch": 0.39156, + "grad_norm": 1.2889882480552226, + "learning_rate": 0.003, + "loss": 4.013, + "step": 39156 + }, + { + "epoch": 0.39157, + "grad_norm": 0.9332724155541422, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 39157 + }, + { + "epoch": 0.39158, + "grad_norm": 0.9723922645147337, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 39158 + }, + { + "epoch": 0.39159, + "grad_norm": 1.17648879227164, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 39159 + }, + { + "epoch": 0.3916, + "grad_norm": 1.0429906288535404, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 39160 + }, + { + "epoch": 0.39161, + "grad_norm": 1.140151272116252, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 39161 + }, + { + "epoch": 0.39162, + "grad_norm": 1.1124373308528415, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 39162 + }, + { + "epoch": 0.39163, + "grad_norm": 1.0572086840157555, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 39163 + }, + { + "epoch": 0.39164, + "grad_norm": 1.0466274941580678, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 39164 + }, + { + "epoch": 0.39165, + "grad_norm": 1.2000973794274916, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 39165 + }, + { + "epoch": 0.39166, + "grad_norm": 0.9797551086621175, + "learning_rate": 0.003, + "loss": 4.026, + "step": 39166 + }, + { + "epoch": 0.39167, + "grad_norm": 1.012824109084087, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 39167 + }, + { + "epoch": 0.39168, + "grad_norm": 1.0700502091387463, + "learning_rate": 0.003, + "loss": 4.088, + "step": 39168 + }, + { + "epoch": 0.39169, + "grad_norm": 1.2277382136667845, + "learning_rate": 0.003, + "loss": 4.044, + "step": 39169 + }, + { + "epoch": 0.3917, + "grad_norm": 0.9422746713031581, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 39170 + }, + { + "epoch": 0.39171, + "grad_norm": 1.1467244881133645, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 39171 + }, + { + "epoch": 0.39172, + "grad_norm": 1.1201171941079244, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 39172 + }, + { + "epoch": 0.39173, + "grad_norm": 1.4848237651363456, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 39173 + }, + { + "epoch": 0.39174, + "grad_norm": 1.035574259371527, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 39174 + }, + { + "epoch": 0.39175, + "grad_norm": 1.1314634001082042, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 39175 + }, + { + "epoch": 0.39176, + "grad_norm": 1.0917761439319174, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 39176 + }, + { + "epoch": 0.39177, + "grad_norm": 1.2678131264573282, + "learning_rate": 0.003, + "loss": 4.032, + "step": 39177 + }, + { + "epoch": 0.39178, + "grad_norm": 0.9572818406610094, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 39178 + }, + { + "epoch": 0.39179, + "grad_norm": 1.060241769578332, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 39179 + }, + { + "epoch": 0.3918, + "grad_norm": 1.0968383501363632, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 39180 + }, + { + "epoch": 0.39181, + "grad_norm": 0.9979699672841684, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 39181 + }, + { + "epoch": 0.39182, + "grad_norm": 0.9691559063319037, + "learning_rate": 0.003, + "loss": 4.026, + "step": 39182 + }, + { + "epoch": 0.39183, + "grad_norm": 1.0280379538279245, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 39183 + }, + { + "epoch": 0.39184, + "grad_norm": 1.5576483513261952, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 39184 + }, + { + "epoch": 0.39185, + "grad_norm": 0.9246186090233488, + "learning_rate": 0.003, + "loss": 4.067, + "step": 39185 + }, + { + "epoch": 0.39186, + "grad_norm": 0.8410811961900365, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 39186 + }, + { + "epoch": 0.39187, + "grad_norm": 0.9448946603131936, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 39187 + }, + { + "epoch": 0.39188, + "grad_norm": 1.2751700240273405, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 39188 + }, + { + "epoch": 0.39189, + "grad_norm": 0.8328027892026245, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 39189 + }, + { + "epoch": 0.3919, + "grad_norm": 1.1221830083526703, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 39190 + }, + { + "epoch": 0.39191, + "grad_norm": 1.1100022653215162, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 39191 + }, + { + "epoch": 0.39192, + "grad_norm": 0.9564379737676854, + "learning_rate": 0.003, + "loss": 4.049, + "step": 39192 + }, + { + "epoch": 0.39193, + "grad_norm": 1.110516777224414, + "learning_rate": 0.003, + "loss": 4.029, + "step": 39193 + }, + { + "epoch": 0.39194, + "grad_norm": 1.370548526391566, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 39194 + }, + { + "epoch": 0.39195, + "grad_norm": 0.9734158126910007, + "learning_rate": 0.003, + "loss": 4.052, + "step": 39195 + }, + { + "epoch": 0.39196, + "grad_norm": 0.8884190871945947, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 39196 + }, + { + "epoch": 0.39197, + "grad_norm": 1.0793522248776528, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 39197 + }, + { + "epoch": 0.39198, + "grad_norm": 1.2569811361667178, + "learning_rate": 0.003, + "loss": 4.038, + "step": 39198 + }, + { + "epoch": 0.39199, + "grad_norm": 0.9098914331995281, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 39199 + }, + { + "epoch": 0.392, + "grad_norm": 1.0654173732525372, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 39200 + }, + { + "epoch": 0.39201, + "grad_norm": 1.2778696562146425, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 39201 + }, + { + "epoch": 0.39202, + "grad_norm": 0.916875293969278, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 39202 + }, + { + "epoch": 0.39203, + "grad_norm": 0.9134721957827362, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 39203 + }, + { + "epoch": 0.39204, + "grad_norm": 1.0942741460168546, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 39204 + }, + { + "epoch": 0.39205, + "grad_norm": 1.1295593607744054, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 39205 + }, + { + "epoch": 0.39206, + "grad_norm": 0.7276266430319634, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 39206 + }, + { + "epoch": 0.39207, + "grad_norm": 0.6861617886222239, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 39207 + }, + { + "epoch": 0.39208, + "grad_norm": 1.015570345461892, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 39208 + }, + { + "epoch": 0.39209, + "grad_norm": 1.2195894586290466, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 39209 + }, + { + "epoch": 0.3921, + "grad_norm": 0.8325784761662173, + "learning_rate": 0.003, + "loss": 4.033, + "step": 39210 + }, + { + "epoch": 0.39211, + "grad_norm": 0.8858404115965747, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 39211 + }, + { + "epoch": 0.39212, + "grad_norm": 1.0585778398172487, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 39212 + }, + { + "epoch": 0.39213, + "grad_norm": 1.2891899959938946, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 39213 + }, + { + "epoch": 0.39214, + "grad_norm": 1.0152425275166257, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 39214 + }, + { + "epoch": 0.39215, + "grad_norm": 1.2164472301398852, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 39215 + }, + { + "epoch": 0.39216, + "grad_norm": 1.0287075098372698, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 39216 + }, + { + "epoch": 0.39217, + "grad_norm": 1.0808467176634173, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 39217 + }, + { + "epoch": 0.39218, + "grad_norm": 1.0085037147241538, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 39218 + }, + { + "epoch": 0.39219, + "grad_norm": 1.3494803260724613, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 39219 + }, + { + "epoch": 0.3922, + "grad_norm": 1.3690462789787576, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 39220 + }, + { + "epoch": 0.39221, + "grad_norm": 1.0187976752514114, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 39221 + }, + { + "epoch": 0.39222, + "grad_norm": 0.9059059723277072, + "learning_rate": 0.003, + "loss": 4.046, + "step": 39222 + }, + { + "epoch": 0.39223, + "grad_norm": 1.1483264582126227, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 39223 + }, + { + "epoch": 0.39224, + "grad_norm": 1.0866005543825608, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 39224 + }, + { + "epoch": 0.39225, + "grad_norm": 1.2888914499478734, + "learning_rate": 0.003, + "loss": 4.065, + "step": 39225 + }, + { + "epoch": 0.39226, + "grad_norm": 1.0947796744447924, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 39226 + }, + { + "epoch": 0.39227, + "grad_norm": 0.995045929777054, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 39227 + }, + { + "epoch": 0.39228, + "grad_norm": 0.9916219406088008, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 39228 + }, + { + "epoch": 0.39229, + "grad_norm": 1.1989785913409223, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 39229 + }, + { + "epoch": 0.3923, + "grad_norm": 1.0830586014542125, + "learning_rate": 0.003, + "loss": 3.985, + "step": 39230 + }, + { + "epoch": 0.39231, + "grad_norm": 1.178393109815903, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 39231 + }, + { + "epoch": 0.39232, + "grad_norm": 1.254308724125386, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 39232 + }, + { + "epoch": 0.39233, + "grad_norm": 0.9595889251706782, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 39233 + }, + { + "epoch": 0.39234, + "grad_norm": 1.0077072421720543, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 39234 + }, + { + "epoch": 0.39235, + "grad_norm": 1.038173703706851, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 39235 + }, + { + "epoch": 0.39236, + "grad_norm": 1.169876610747858, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 39236 + }, + { + "epoch": 0.39237, + "grad_norm": 1.1525390254444523, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 39237 + }, + { + "epoch": 0.39238, + "grad_norm": 0.9575660388569116, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 39238 + }, + { + "epoch": 0.39239, + "grad_norm": 1.0066257701545003, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 39239 + }, + { + "epoch": 0.3924, + "grad_norm": 1.2700470673677433, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 39240 + }, + { + "epoch": 0.39241, + "grad_norm": 1.0121311688964096, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 39241 + }, + { + "epoch": 0.39242, + "grad_norm": 1.1038332961875654, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 39242 + }, + { + "epoch": 0.39243, + "grad_norm": 0.8591831924167226, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 39243 + }, + { + "epoch": 0.39244, + "grad_norm": 0.9192186754944294, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 39244 + }, + { + "epoch": 0.39245, + "grad_norm": 1.3836036729489771, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 39245 + }, + { + "epoch": 0.39246, + "grad_norm": 0.8721240433392972, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 39246 + }, + { + "epoch": 0.39247, + "grad_norm": 0.9442328111272911, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 39247 + }, + { + "epoch": 0.39248, + "grad_norm": 1.114705229683394, + "learning_rate": 0.003, + "loss": 4.062, + "step": 39248 + }, + { + "epoch": 0.39249, + "grad_norm": 1.197611745748418, + "learning_rate": 0.003, + "loss": 4.028, + "step": 39249 + }, + { + "epoch": 0.3925, + "grad_norm": 1.2054824119366043, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 39250 + }, + { + "epoch": 0.39251, + "grad_norm": 1.1499177023577936, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 39251 + }, + { + "epoch": 0.39252, + "grad_norm": 1.0954055192879757, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 39252 + }, + { + "epoch": 0.39253, + "grad_norm": 1.1577168814379992, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 39253 + }, + { + "epoch": 0.39254, + "grad_norm": 1.1106969440640504, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 39254 + }, + { + "epoch": 0.39255, + "grad_norm": 1.2296427779012549, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 39255 + }, + { + "epoch": 0.39256, + "grad_norm": 1.0949295476680665, + "learning_rate": 0.003, + "loss": 4.044, + "step": 39256 + }, + { + "epoch": 0.39257, + "grad_norm": 1.1043037949996037, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 39257 + }, + { + "epoch": 0.39258, + "grad_norm": 0.9024656652530272, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 39258 + }, + { + "epoch": 0.39259, + "grad_norm": 0.9708671441968447, + "learning_rate": 0.003, + "loss": 4.062, + "step": 39259 + }, + { + "epoch": 0.3926, + "grad_norm": 1.0721133522788364, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 39260 + }, + { + "epoch": 0.39261, + "grad_norm": 1.0291246026650955, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 39261 + }, + { + "epoch": 0.39262, + "grad_norm": 1.1755345547703224, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 39262 + }, + { + "epoch": 0.39263, + "grad_norm": 1.0801317328968076, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 39263 + }, + { + "epoch": 0.39264, + "grad_norm": 1.3433990365060076, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 39264 + }, + { + "epoch": 0.39265, + "grad_norm": 1.0188309026480125, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 39265 + }, + { + "epoch": 0.39266, + "grad_norm": 1.232104005308785, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 39266 + }, + { + "epoch": 0.39267, + "grad_norm": 0.918190335652169, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 39267 + }, + { + "epoch": 0.39268, + "grad_norm": 0.9245555713549231, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 39268 + }, + { + "epoch": 0.39269, + "grad_norm": 0.962360452864554, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 39269 + }, + { + "epoch": 0.3927, + "grad_norm": 1.2032402564519182, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 39270 + }, + { + "epoch": 0.39271, + "grad_norm": 1.327385857762658, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 39271 + }, + { + "epoch": 0.39272, + "grad_norm": 1.0128712552943127, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 39272 + }, + { + "epoch": 0.39273, + "grad_norm": 0.9707915180934074, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 39273 + }, + { + "epoch": 0.39274, + "grad_norm": 0.9803965696389646, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 39274 + }, + { + "epoch": 0.39275, + "grad_norm": 1.2726277891453128, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 39275 + }, + { + "epoch": 0.39276, + "grad_norm": 0.9201497850534409, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 39276 + }, + { + "epoch": 0.39277, + "grad_norm": 1.0852252874889392, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 39277 + }, + { + "epoch": 0.39278, + "grad_norm": 1.1039218230914778, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 39278 + }, + { + "epoch": 0.39279, + "grad_norm": 1.0690516454634407, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 39279 + }, + { + "epoch": 0.3928, + "grad_norm": 1.0576390863118508, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 39280 + }, + { + "epoch": 0.39281, + "grad_norm": 1.1597979174961948, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 39281 + }, + { + "epoch": 0.39282, + "grad_norm": 0.9440272244078362, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 39282 + }, + { + "epoch": 0.39283, + "grad_norm": 0.9922673417045462, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 39283 + }, + { + "epoch": 0.39284, + "grad_norm": 1.0988575042537663, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 39284 + }, + { + "epoch": 0.39285, + "grad_norm": 1.0093928434551296, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 39285 + }, + { + "epoch": 0.39286, + "grad_norm": 1.0869207816594388, + "learning_rate": 0.003, + "loss": 4.026, + "step": 39286 + }, + { + "epoch": 0.39287, + "grad_norm": 0.9926332078474238, + "learning_rate": 0.003, + "loss": 4.05, + "step": 39287 + }, + { + "epoch": 0.39288, + "grad_norm": 1.10018795146728, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 39288 + }, + { + "epoch": 0.39289, + "grad_norm": 1.0511548092908827, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 39289 + }, + { + "epoch": 0.3929, + "grad_norm": 1.0672530339467476, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 39290 + }, + { + "epoch": 0.39291, + "grad_norm": 1.1472412606792841, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 39291 + }, + { + "epoch": 0.39292, + "grad_norm": 1.1160501813297425, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 39292 + }, + { + "epoch": 0.39293, + "grad_norm": 1.1056492322510794, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 39293 + }, + { + "epoch": 0.39294, + "grad_norm": 1.2796268301747638, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 39294 + }, + { + "epoch": 0.39295, + "grad_norm": 1.0599846447296328, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 39295 + }, + { + "epoch": 0.39296, + "grad_norm": 1.0017511101263703, + "learning_rate": 0.003, + "loss": 4.042, + "step": 39296 + }, + { + "epoch": 0.39297, + "grad_norm": 1.0960428320886841, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 39297 + }, + { + "epoch": 0.39298, + "grad_norm": 1.087452831043349, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 39298 + }, + { + "epoch": 0.39299, + "grad_norm": 1.051145639882061, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 39299 + }, + { + "epoch": 0.393, + "grad_norm": 1.21614180723357, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 39300 + }, + { + "epoch": 0.39301, + "grad_norm": 1.0614872478342894, + "learning_rate": 0.003, + "loss": 4.022, + "step": 39301 + }, + { + "epoch": 0.39302, + "grad_norm": 1.126955292244651, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 39302 + }, + { + "epoch": 0.39303, + "grad_norm": 0.9470371499885408, + "learning_rate": 0.003, + "loss": 4.061, + "step": 39303 + }, + { + "epoch": 0.39304, + "grad_norm": 0.988636657996819, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 39304 + }, + { + "epoch": 0.39305, + "grad_norm": 1.0946954271460172, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 39305 + }, + { + "epoch": 0.39306, + "grad_norm": 1.1434559064647, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 39306 + }, + { + "epoch": 0.39307, + "grad_norm": 1.1282849878149432, + "learning_rate": 0.003, + "loss": 4.028, + "step": 39307 + }, + { + "epoch": 0.39308, + "grad_norm": 1.1946716597567346, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 39308 + }, + { + "epoch": 0.39309, + "grad_norm": 1.0017384652143533, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 39309 + }, + { + "epoch": 0.3931, + "grad_norm": 1.143243284577109, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 39310 + }, + { + "epoch": 0.39311, + "grad_norm": 1.022525516385117, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 39311 + }, + { + "epoch": 0.39312, + "grad_norm": 1.1917018741703485, + "learning_rate": 0.003, + "loss": 4.076, + "step": 39312 + }, + { + "epoch": 0.39313, + "grad_norm": 1.0013800703738187, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 39313 + }, + { + "epoch": 0.39314, + "grad_norm": 1.0747002990792767, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 39314 + }, + { + "epoch": 0.39315, + "grad_norm": 1.572683593315662, + "learning_rate": 0.003, + "loss": 4.049, + "step": 39315 + }, + { + "epoch": 0.39316, + "grad_norm": 0.9343776872693214, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 39316 + }, + { + "epoch": 0.39317, + "grad_norm": 0.8967199924780311, + "learning_rate": 0.003, + "loss": 4.035, + "step": 39317 + }, + { + "epoch": 0.39318, + "grad_norm": 1.0567723658370456, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 39318 + }, + { + "epoch": 0.39319, + "grad_norm": 0.9787959415827127, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 39319 + }, + { + "epoch": 0.3932, + "grad_norm": 1.2956746877791068, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 39320 + }, + { + "epoch": 0.39321, + "grad_norm": 1.0663976068945664, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 39321 + }, + { + "epoch": 0.39322, + "grad_norm": 1.5953036044994748, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 39322 + }, + { + "epoch": 0.39323, + "grad_norm": 0.727648445851257, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 39323 + }, + { + "epoch": 0.39324, + "grad_norm": 0.7896136293380706, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 39324 + }, + { + "epoch": 0.39325, + "grad_norm": 1.0564577949208567, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 39325 + }, + { + "epoch": 0.39326, + "grad_norm": 1.2644943107130824, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 39326 + }, + { + "epoch": 0.39327, + "grad_norm": 1.090882343773557, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 39327 + }, + { + "epoch": 0.39328, + "grad_norm": 1.211375452517233, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 39328 + }, + { + "epoch": 0.39329, + "grad_norm": 0.9929175023831115, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 39329 + }, + { + "epoch": 0.3933, + "grad_norm": 1.1761622207368285, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 39330 + }, + { + "epoch": 0.39331, + "grad_norm": 0.8628663961815884, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 39331 + }, + { + "epoch": 0.39332, + "grad_norm": 1.0658168742392196, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 39332 + }, + { + "epoch": 0.39333, + "grad_norm": 0.9969904992966786, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 39333 + }, + { + "epoch": 0.39334, + "grad_norm": 1.1274828691387242, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 39334 + }, + { + "epoch": 0.39335, + "grad_norm": 1.3625644192296709, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 39335 + }, + { + "epoch": 0.39336, + "grad_norm": 0.9079294170414495, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 39336 + }, + { + "epoch": 0.39337, + "grad_norm": 0.8711274531703131, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 39337 + }, + { + "epoch": 0.39338, + "grad_norm": 1.0610035126150759, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 39338 + }, + { + "epoch": 0.39339, + "grad_norm": 1.1284729870554373, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 39339 + }, + { + "epoch": 0.3934, + "grad_norm": 1.0494927540274424, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 39340 + }, + { + "epoch": 0.39341, + "grad_norm": 1.0400321282322966, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 39341 + }, + { + "epoch": 0.39342, + "grad_norm": 0.9715954566250217, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 39342 + }, + { + "epoch": 0.39343, + "grad_norm": 1.0499337076508874, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 39343 + }, + { + "epoch": 0.39344, + "grad_norm": 1.0940624247948, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 39344 + }, + { + "epoch": 0.39345, + "grad_norm": 1.0296193999692433, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 39345 + }, + { + "epoch": 0.39346, + "grad_norm": 1.0892123437584713, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 39346 + }, + { + "epoch": 0.39347, + "grad_norm": 1.0528320492496153, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 39347 + }, + { + "epoch": 0.39348, + "grad_norm": 1.0650833102880122, + "learning_rate": 0.003, + "loss": 4.039, + "step": 39348 + }, + { + "epoch": 0.39349, + "grad_norm": 1.0761319032858103, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 39349 + }, + { + "epoch": 0.3935, + "grad_norm": 1.080435738561484, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 39350 + }, + { + "epoch": 0.39351, + "grad_norm": 1.0571319943003437, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 39351 + }, + { + "epoch": 0.39352, + "grad_norm": 1.3236215586177131, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 39352 + }, + { + "epoch": 0.39353, + "grad_norm": 1.039190464025223, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 39353 + }, + { + "epoch": 0.39354, + "grad_norm": 1.05076077771878, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 39354 + }, + { + "epoch": 0.39355, + "grad_norm": 0.9356988213787176, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 39355 + }, + { + "epoch": 0.39356, + "grad_norm": 1.0971719910930495, + "learning_rate": 0.003, + "loss": 4.056, + "step": 39356 + }, + { + "epoch": 0.39357, + "grad_norm": 1.2855061231981337, + "learning_rate": 0.003, + "loss": 4.037, + "step": 39357 + }, + { + "epoch": 0.39358, + "grad_norm": 0.9875341766195443, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 39358 + }, + { + "epoch": 0.39359, + "grad_norm": 1.0815807922339071, + "learning_rate": 0.003, + "loss": 4.039, + "step": 39359 + }, + { + "epoch": 0.3936, + "grad_norm": 1.1592348146648477, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 39360 + }, + { + "epoch": 0.39361, + "grad_norm": 1.0816514245415483, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 39361 + }, + { + "epoch": 0.39362, + "grad_norm": 1.0037647089014048, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 39362 + }, + { + "epoch": 0.39363, + "grad_norm": 1.1550276160893689, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 39363 + }, + { + "epoch": 0.39364, + "grad_norm": 1.0210065630262248, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 39364 + }, + { + "epoch": 0.39365, + "grad_norm": 0.9708654558064884, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 39365 + }, + { + "epoch": 0.39366, + "grad_norm": 1.0826402621327533, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 39366 + }, + { + "epoch": 0.39367, + "grad_norm": 1.0001813941563424, + "learning_rate": 0.003, + "loss": 4.032, + "step": 39367 + }, + { + "epoch": 0.39368, + "grad_norm": 1.0401236737475685, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 39368 + }, + { + "epoch": 0.39369, + "grad_norm": 1.0252693607364058, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 39369 + }, + { + "epoch": 0.3937, + "grad_norm": 1.1008567988258862, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 39370 + }, + { + "epoch": 0.39371, + "grad_norm": 1.3870361803727007, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 39371 + }, + { + "epoch": 0.39372, + "grad_norm": 1.0954300576377216, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 39372 + }, + { + "epoch": 0.39373, + "grad_norm": 1.1718489595354618, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 39373 + }, + { + "epoch": 0.39374, + "grad_norm": 0.952126711192377, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 39374 + }, + { + "epoch": 0.39375, + "grad_norm": 1.2179825416747, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 39375 + }, + { + "epoch": 0.39376, + "grad_norm": 1.1251711005070435, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 39376 + }, + { + "epoch": 0.39377, + "grad_norm": 1.1172974119832353, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 39377 + }, + { + "epoch": 0.39378, + "grad_norm": 1.0975014501264284, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 39378 + }, + { + "epoch": 0.39379, + "grad_norm": 1.069200550111206, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 39379 + }, + { + "epoch": 0.3938, + "grad_norm": 0.9437426807580425, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 39380 + }, + { + "epoch": 0.39381, + "grad_norm": 1.1137606794714374, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 39381 + }, + { + "epoch": 0.39382, + "grad_norm": 1.1526597396251055, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 39382 + }, + { + "epoch": 0.39383, + "grad_norm": 1.1335017771566787, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 39383 + }, + { + "epoch": 0.39384, + "grad_norm": 1.0303493173551732, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 39384 + }, + { + "epoch": 0.39385, + "grad_norm": 1.186638756590633, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 39385 + }, + { + "epoch": 0.39386, + "grad_norm": 0.8957295254118088, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 39386 + }, + { + "epoch": 0.39387, + "grad_norm": 1.0686735574580077, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 39387 + }, + { + "epoch": 0.39388, + "grad_norm": 1.3188484069063737, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 39388 + }, + { + "epoch": 0.39389, + "grad_norm": 0.9108732568990766, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 39389 + }, + { + "epoch": 0.3939, + "grad_norm": 0.8721106079956069, + "learning_rate": 0.003, + "loss": 4.014, + "step": 39390 + }, + { + "epoch": 0.39391, + "grad_norm": 1.015369650407001, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 39391 + }, + { + "epoch": 0.39392, + "grad_norm": 1.3821467843778497, + "learning_rate": 0.003, + "loss": 4.005, + "step": 39392 + }, + { + "epoch": 0.39393, + "grad_norm": 0.7980436215323826, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 39393 + }, + { + "epoch": 0.39394, + "grad_norm": 0.7631042202087195, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 39394 + }, + { + "epoch": 0.39395, + "grad_norm": 0.9960483117906124, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 39395 + }, + { + "epoch": 0.39396, + "grad_norm": 1.3334147186667262, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 39396 + }, + { + "epoch": 0.39397, + "grad_norm": 0.8598818761151378, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 39397 + }, + { + "epoch": 0.39398, + "grad_norm": 1.056615182462742, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 39398 + }, + { + "epoch": 0.39399, + "grad_norm": 1.1639112148863768, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 39399 + }, + { + "epoch": 0.394, + "grad_norm": 0.9962110367265676, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 39400 + }, + { + "epoch": 0.39401, + "grad_norm": 0.9570505310746916, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 39401 + }, + { + "epoch": 0.39402, + "grad_norm": 1.0252076646069788, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 39402 + }, + { + "epoch": 0.39403, + "grad_norm": 1.1628083392056534, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 39403 + }, + { + "epoch": 0.39404, + "grad_norm": 0.9488926963047604, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 39404 + }, + { + "epoch": 0.39405, + "grad_norm": 1.2916104339430223, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 39405 + }, + { + "epoch": 0.39406, + "grad_norm": 1.0770523842462847, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 39406 + }, + { + "epoch": 0.39407, + "grad_norm": 1.0098687632854009, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 39407 + }, + { + "epoch": 0.39408, + "grad_norm": 1.0284094533631247, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 39408 + }, + { + "epoch": 0.39409, + "grad_norm": 1.155584058691031, + "learning_rate": 0.003, + "loss": 4.026, + "step": 39409 + }, + { + "epoch": 0.3941, + "grad_norm": 1.1403204273276273, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 39410 + }, + { + "epoch": 0.39411, + "grad_norm": 1.185096630732479, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 39411 + }, + { + "epoch": 0.39412, + "grad_norm": 1.011998073483414, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 39412 + }, + { + "epoch": 0.39413, + "grad_norm": 1.2832880103557067, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 39413 + }, + { + "epoch": 0.39414, + "grad_norm": 0.9245522369292437, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 39414 + }, + { + "epoch": 0.39415, + "grad_norm": 1.0245895004821532, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 39415 + }, + { + "epoch": 0.39416, + "grad_norm": 1.2685064635136585, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 39416 + }, + { + "epoch": 0.39417, + "grad_norm": 0.9580927533657115, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 39417 + }, + { + "epoch": 0.39418, + "grad_norm": 0.9162561430924698, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 39418 + }, + { + "epoch": 0.39419, + "grad_norm": 0.9195280282234141, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 39419 + }, + { + "epoch": 0.3942, + "grad_norm": 1.0034859429578713, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 39420 + }, + { + "epoch": 0.39421, + "grad_norm": 1.3558206848784542, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 39421 + }, + { + "epoch": 0.39422, + "grad_norm": 0.8807054447094301, + "learning_rate": 0.003, + "loss": 4.026, + "step": 39422 + }, + { + "epoch": 0.39423, + "grad_norm": 1.0143011483655644, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 39423 + }, + { + "epoch": 0.39424, + "grad_norm": 1.1158163387264677, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 39424 + }, + { + "epoch": 0.39425, + "grad_norm": 1.3372757360140077, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 39425 + }, + { + "epoch": 0.39426, + "grad_norm": 0.9640183948110603, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 39426 + }, + { + "epoch": 0.39427, + "grad_norm": 1.0307199722718665, + "learning_rate": 0.003, + "loss": 4.035, + "step": 39427 + }, + { + "epoch": 0.39428, + "grad_norm": 1.1621225508463269, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 39428 + }, + { + "epoch": 0.39429, + "grad_norm": 0.948244436551281, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 39429 + }, + { + "epoch": 0.3943, + "grad_norm": 1.0262686863820178, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 39430 + }, + { + "epoch": 0.39431, + "grad_norm": 1.0996958146236264, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 39431 + }, + { + "epoch": 0.39432, + "grad_norm": 0.9958158584913694, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 39432 + }, + { + "epoch": 0.39433, + "grad_norm": 1.0804047373322663, + "learning_rate": 0.003, + "loss": 4.036, + "step": 39433 + }, + { + "epoch": 0.39434, + "grad_norm": 1.077615895015348, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 39434 + }, + { + "epoch": 0.39435, + "grad_norm": 1.044729105345678, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 39435 + }, + { + "epoch": 0.39436, + "grad_norm": 1.0826323753443607, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 39436 + }, + { + "epoch": 0.39437, + "grad_norm": 0.9947396888452112, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 39437 + }, + { + "epoch": 0.39438, + "grad_norm": 1.1135507883921163, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 39438 + }, + { + "epoch": 0.39439, + "grad_norm": 0.9853377214443749, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 39439 + }, + { + "epoch": 0.3944, + "grad_norm": 1.0608679246472552, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 39440 + }, + { + "epoch": 0.39441, + "grad_norm": 1.3103183997237762, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 39441 + }, + { + "epoch": 0.39442, + "grad_norm": 0.6878694008357252, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 39442 + }, + { + "epoch": 0.39443, + "grad_norm": 0.7138113299921613, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 39443 + }, + { + "epoch": 0.39444, + "grad_norm": 1.0553142540651588, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 39444 + }, + { + "epoch": 0.39445, + "grad_norm": 1.7112829069409006, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 39445 + }, + { + "epoch": 0.39446, + "grad_norm": 0.855852330949626, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 39446 + }, + { + "epoch": 0.39447, + "grad_norm": 0.9644100764704854, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 39447 + }, + { + "epoch": 0.39448, + "grad_norm": 1.0830398597998214, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 39448 + }, + { + "epoch": 0.39449, + "grad_norm": 1.099099218025758, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 39449 + }, + { + "epoch": 0.3945, + "grad_norm": 1.144456208560266, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 39450 + }, + { + "epoch": 0.39451, + "grad_norm": 1.1570160643016834, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 39451 + }, + { + "epoch": 0.39452, + "grad_norm": 0.9623771723045639, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 39452 + }, + { + "epoch": 0.39453, + "grad_norm": 1.2627236986940353, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 39453 + }, + { + "epoch": 0.39454, + "grad_norm": 1.134391021860648, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 39454 + }, + { + "epoch": 0.39455, + "grad_norm": 0.9598095413149141, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 39455 + }, + { + "epoch": 0.39456, + "grad_norm": 1.1428370542977413, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 39456 + }, + { + "epoch": 0.39457, + "grad_norm": 1.1384137960438498, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 39457 + }, + { + "epoch": 0.39458, + "grad_norm": 0.9998065643348286, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 39458 + }, + { + "epoch": 0.39459, + "grad_norm": 1.0151209764793667, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 39459 + }, + { + "epoch": 0.3946, + "grad_norm": 0.9830997484555533, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 39460 + }, + { + "epoch": 0.39461, + "grad_norm": 0.8708301590387783, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 39461 + }, + { + "epoch": 0.39462, + "grad_norm": 1.0484145085969336, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 39462 + }, + { + "epoch": 0.39463, + "grad_norm": 1.4604623769693106, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 39463 + }, + { + "epoch": 0.39464, + "grad_norm": 0.6819013589894226, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 39464 + }, + { + "epoch": 0.39465, + "grad_norm": 0.8299763550680186, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 39465 + }, + { + "epoch": 0.39466, + "grad_norm": 0.9534334326709714, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 39466 + }, + { + "epoch": 0.39467, + "grad_norm": 1.0591055758744352, + "learning_rate": 0.003, + "loss": 4.042, + "step": 39467 + }, + { + "epoch": 0.39468, + "grad_norm": 0.9068914012708646, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 39468 + }, + { + "epoch": 0.39469, + "grad_norm": 1.0839393135452198, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 39469 + }, + { + "epoch": 0.3947, + "grad_norm": 1.3181846993040218, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 39470 + }, + { + "epoch": 0.39471, + "grad_norm": 0.9925163823515465, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 39471 + }, + { + "epoch": 0.39472, + "grad_norm": 1.2013969143488614, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 39472 + }, + { + "epoch": 0.39473, + "grad_norm": 1.0197169360414955, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 39473 + }, + { + "epoch": 0.39474, + "grad_norm": 1.0605264899920224, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 39474 + }, + { + "epoch": 0.39475, + "grad_norm": 1.1247843270751852, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 39475 + }, + { + "epoch": 0.39476, + "grad_norm": 1.3616997947764213, + "learning_rate": 0.003, + "loss": 4.05, + "step": 39476 + }, + { + "epoch": 0.39477, + "grad_norm": 0.9313043262611064, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 39477 + }, + { + "epoch": 0.39478, + "grad_norm": 1.0835171173644307, + "learning_rate": 0.003, + "loss": 4.049, + "step": 39478 + }, + { + "epoch": 0.39479, + "grad_norm": 1.3625335630250033, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 39479 + }, + { + "epoch": 0.3948, + "grad_norm": 0.8405060386415867, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 39480 + }, + { + "epoch": 0.39481, + "grad_norm": 0.9018380401594864, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 39481 + }, + { + "epoch": 0.39482, + "grad_norm": 1.1553307962724688, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 39482 + }, + { + "epoch": 0.39483, + "grad_norm": 1.2882763247573106, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 39483 + }, + { + "epoch": 0.39484, + "grad_norm": 0.9530597201570714, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 39484 + }, + { + "epoch": 0.39485, + "grad_norm": 0.945483546638423, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 39485 + }, + { + "epoch": 0.39486, + "grad_norm": 0.9486334806346157, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 39486 + }, + { + "epoch": 0.39487, + "grad_norm": 0.9971067964113752, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 39487 + }, + { + "epoch": 0.39488, + "grad_norm": 0.9825975043385191, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 39488 + }, + { + "epoch": 0.39489, + "grad_norm": 1.0206195245473935, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 39489 + }, + { + "epoch": 0.3949, + "grad_norm": 1.1741966106777937, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 39490 + }, + { + "epoch": 0.39491, + "grad_norm": 0.9092972807587759, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 39491 + }, + { + "epoch": 0.39492, + "grad_norm": 1.0511166969727126, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 39492 + }, + { + "epoch": 0.39493, + "grad_norm": 1.205557645677107, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 39493 + }, + { + "epoch": 0.39494, + "grad_norm": 1.0753421430624301, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 39494 + }, + { + "epoch": 0.39495, + "grad_norm": 1.2798046194579884, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 39495 + }, + { + "epoch": 0.39496, + "grad_norm": 0.9184086737948406, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 39496 + }, + { + "epoch": 0.39497, + "grad_norm": 1.1947568515152234, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 39497 + }, + { + "epoch": 0.39498, + "grad_norm": 1.179298641297218, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 39498 + }, + { + "epoch": 0.39499, + "grad_norm": 0.9374924897373572, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 39499 + }, + { + "epoch": 0.395, + "grad_norm": 1.0815480378473226, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 39500 + }, + { + "epoch": 0.39501, + "grad_norm": 1.1936578113725356, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 39501 + }, + { + "epoch": 0.39502, + "grad_norm": 1.0935488835648635, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 39502 + }, + { + "epoch": 0.39503, + "grad_norm": 1.0892176370815314, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 39503 + }, + { + "epoch": 0.39504, + "grad_norm": 1.2131625308119016, + "learning_rate": 0.003, + "loss": 4.034, + "step": 39504 + }, + { + "epoch": 0.39505, + "grad_norm": 0.9940863495769352, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 39505 + }, + { + "epoch": 0.39506, + "grad_norm": 1.0198340446498322, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 39506 + }, + { + "epoch": 0.39507, + "grad_norm": 1.0612069144239318, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 39507 + }, + { + "epoch": 0.39508, + "grad_norm": 1.2700366565266068, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 39508 + }, + { + "epoch": 0.39509, + "grad_norm": 0.9200006365948861, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 39509 + }, + { + "epoch": 0.3951, + "grad_norm": 0.8474797536761176, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 39510 + }, + { + "epoch": 0.39511, + "grad_norm": 0.9583842390954985, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 39511 + }, + { + "epoch": 0.39512, + "grad_norm": 1.3630206213272487, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 39512 + }, + { + "epoch": 0.39513, + "grad_norm": 0.9538488826740441, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 39513 + }, + { + "epoch": 0.39514, + "grad_norm": 0.9574777129797098, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 39514 + }, + { + "epoch": 0.39515, + "grad_norm": 1.0629755983032025, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 39515 + }, + { + "epoch": 0.39516, + "grad_norm": 1.3019499726177632, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 39516 + }, + { + "epoch": 0.39517, + "grad_norm": 0.8056252250664313, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 39517 + }, + { + "epoch": 0.39518, + "grad_norm": 0.71066574841106, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 39518 + }, + { + "epoch": 0.39519, + "grad_norm": 0.8386638577529708, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 39519 + }, + { + "epoch": 0.3952, + "grad_norm": 1.0505647683957164, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 39520 + }, + { + "epoch": 0.39521, + "grad_norm": 1.1495562372998451, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 39521 + }, + { + "epoch": 0.39522, + "grad_norm": 1.2726762824475186, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 39522 + }, + { + "epoch": 0.39523, + "grad_norm": 0.904408522381446, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 39523 + }, + { + "epoch": 0.39524, + "grad_norm": 1.149396090310492, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 39524 + }, + { + "epoch": 0.39525, + "grad_norm": 1.0916206383289868, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 39525 + }, + { + "epoch": 0.39526, + "grad_norm": 1.247512660298248, + "learning_rate": 0.003, + "loss": 4.051, + "step": 39526 + }, + { + "epoch": 0.39527, + "grad_norm": 0.904832523893185, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 39527 + }, + { + "epoch": 0.39528, + "grad_norm": 0.8260938587589753, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 39528 + }, + { + "epoch": 0.39529, + "grad_norm": 0.8868866834331444, + "learning_rate": 0.003, + "loss": 4.049, + "step": 39529 + }, + { + "epoch": 0.3953, + "grad_norm": 1.1667874922282384, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 39530 + }, + { + "epoch": 0.39531, + "grad_norm": 1.3060845599778466, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 39531 + }, + { + "epoch": 0.39532, + "grad_norm": 0.9212071277583199, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 39532 + }, + { + "epoch": 0.39533, + "grad_norm": 1.2229515305201168, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 39533 + }, + { + "epoch": 0.39534, + "grad_norm": 0.9676265109964612, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 39534 + }, + { + "epoch": 0.39535, + "grad_norm": 1.22442822203354, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 39535 + }, + { + "epoch": 0.39536, + "grad_norm": 1.2013217824763192, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 39536 + }, + { + "epoch": 0.39537, + "grad_norm": 1.021182717195546, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 39537 + }, + { + "epoch": 0.39538, + "grad_norm": 0.8482170213669108, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 39538 + }, + { + "epoch": 0.39539, + "grad_norm": 0.8729996564951161, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 39539 + }, + { + "epoch": 0.3954, + "grad_norm": 1.0413981070522056, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 39540 + }, + { + "epoch": 0.39541, + "grad_norm": 1.188204465527186, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 39541 + }, + { + "epoch": 0.39542, + "grad_norm": 1.1789785581354102, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 39542 + }, + { + "epoch": 0.39543, + "grad_norm": 1.1749084801032201, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 39543 + }, + { + "epoch": 0.39544, + "grad_norm": 1.1506752723673588, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 39544 + }, + { + "epoch": 0.39545, + "grad_norm": 1.0862033555192152, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 39545 + }, + { + "epoch": 0.39546, + "grad_norm": 0.9253785736340504, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 39546 + }, + { + "epoch": 0.39547, + "grad_norm": 1.1144409377215845, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 39547 + }, + { + "epoch": 0.39548, + "grad_norm": 1.0800409603323102, + "learning_rate": 0.003, + "loss": 4.004, + "step": 39548 + }, + { + "epoch": 0.39549, + "grad_norm": 1.147538294812319, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 39549 + }, + { + "epoch": 0.3955, + "grad_norm": 0.943848927395629, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 39550 + }, + { + "epoch": 0.39551, + "grad_norm": 1.2890250867138262, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 39551 + }, + { + "epoch": 0.39552, + "grad_norm": 1.263312387845453, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 39552 + }, + { + "epoch": 0.39553, + "grad_norm": 1.1703361476358873, + "learning_rate": 0.003, + "loss": 4.028, + "step": 39553 + }, + { + "epoch": 0.39554, + "grad_norm": 0.9719050017279088, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 39554 + }, + { + "epoch": 0.39555, + "grad_norm": 1.0072837449085463, + "learning_rate": 0.003, + "loss": 4.005, + "step": 39555 + }, + { + "epoch": 0.39556, + "grad_norm": 1.171638239900461, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 39556 + }, + { + "epoch": 0.39557, + "grad_norm": 1.110196706008205, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 39557 + }, + { + "epoch": 0.39558, + "grad_norm": 1.0047235232650267, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 39558 + }, + { + "epoch": 0.39559, + "grad_norm": 0.9746610636613785, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 39559 + }, + { + "epoch": 0.3956, + "grad_norm": 1.1701562929628422, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 39560 + }, + { + "epoch": 0.39561, + "grad_norm": 0.9359516714666237, + "learning_rate": 0.003, + "loss": 4.019, + "step": 39561 + }, + { + "epoch": 0.39562, + "grad_norm": 1.136137889079287, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 39562 + }, + { + "epoch": 0.39563, + "grad_norm": 1.005467284026473, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 39563 + }, + { + "epoch": 0.39564, + "grad_norm": 1.088684582516083, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 39564 + }, + { + "epoch": 0.39565, + "grad_norm": 1.088606859539093, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 39565 + }, + { + "epoch": 0.39566, + "grad_norm": 1.002621784598056, + "learning_rate": 0.003, + "loss": 4.049, + "step": 39566 + }, + { + "epoch": 0.39567, + "grad_norm": 1.1298089599352827, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 39567 + }, + { + "epoch": 0.39568, + "grad_norm": 0.9633001214151208, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 39568 + }, + { + "epoch": 0.39569, + "grad_norm": 1.0629654378616886, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 39569 + }, + { + "epoch": 0.3957, + "grad_norm": 1.1257244781550604, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 39570 + }, + { + "epoch": 0.39571, + "grad_norm": 1.0047218738637633, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 39571 + }, + { + "epoch": 0.39572, + "grad_norm": 1.5033037429991567, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 39572 + }, + { + "epoch": 0.39573, + "grad_norm": 1.013649080015332, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 39573 + }, + { + "epoch": 0.39574, + "grad_norm": 1.1322934087062244, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 39574 + }, + { + "epoch": 0.39575, + "grad_norm": 1.077744908531263, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 39575 + }, + { + "epoch": 0.39576, + "grad_norm": 1.1418930573102304, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 39576 + }, + { + "epoch": 0.39577, + "grad_norm": 0.9918860108535532, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 39577 + }, + { + "epoch": 0.39578, + "grad_norm": 1.1358105216681935, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 39578 + }, + { + "epoch": 0.39579, + "grad_norm": 0.9857081315528162, + "learning_rate": 0.003, + "loss": 4.046, + "step": 39579 + }, + { + "epoch": 0.3958, + "grad_norm": 1.1149233215729075, + "learning_rate": 0.003, + "loss": 4.094, + "step": 39580 + }, + { + "epoch": 0.39581, + "grad_norm": 0.9435295445542715, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 39581 + }, + { + "epoch": 0.39582, + "grad_norm": 1.0270234693958937, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 39582 + }, + { + "epoch": 0.39583, + "grad_norm": 1.2174102654515027, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 39583 + }, + { + "epoch": 0.39584, + "grad_norm": 1.1811277652799557, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 39584 + }, + { + "epoch": 0.39585, + "grad_norm": 0.8632595689868544, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 39585 + }, + { + "epoch": 0.39586, + "grad_norm": 0.9176810942672841, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 39586 + }, + { + "epoch": 0.39587, + "grad_norm": 1.1994936324102108, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 39587 + }, + { + "epoch": 0.39588, + "grad_norm": 1.113085275583254, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 39588 + }, + { + "epoch": 0.39589, + "grad_norm": 1.0932650534171227, + "learning_rate": 0.003, + "loss": 4.037, + "step": 39589 + }, + { + "epoch": 0.3959, + "grad_norm": 1.28126736803947, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 39590 + }, + { + "epoch": 0.39591, + "grad_norm": 0.936457742385687, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 39591 + }, + { + "epoch": 0.39592, + "grad_norm": 0.9239041422874463, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 39592 + }, + { + "epoch": 0.39593, + "grad_norm": 1.2326679844601272, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 39593 + }, + { + "epoch": 0.39594, + "grad_norm": 0.9227426343719258, + "learning_rate": 0.003, + "loss": 4.035, + "step": 39594 + }, + { + "epoch": 0.39595, + "grad_norm": 0.9540256370722633, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 39595 + }, + { + "epoch": 0.39596, + "grad_norm": 0.8508369701259573, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 39596 + }, + { + "epoch": 0.39597, + "grad_norm": 0.9378428626645219, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 39597 + }, + { + "epoch": 0.39598, + "grad_norm": 0.9007879455305282, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 39598 + }, + { + "epoch": 0.39599, + "grad_norm": 0.9533351111822762, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 39599 + }, + { + "epoch": 0.396, + "grad_norm": 1.235221674479344, + "learning_rate": 0.003, + "loss": 4.018, + "step": 39600 + }, + { + "epoch": 0.39601, + "grad_norm": 1.0448250508694144, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 39601 + }, + { + "epoch": 0.39602, + "grad_norm": 1.3794587213262042, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 39602 + }, + { + "epoch": 0.39603, + "grad_norm": 1.0152716462482994, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 39603 + }, + { + "epoch": 0.39604, + "grad_norm": 1.1223273210162252, + "learning_rate": 0.003, + "loss": 4.029, + "step": 39604 + }, + { + "epoch": 0.39605, + "grad_norm": 1.0110756981125435, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 39605 + }, + { + "epoch": 0.39606, + "grad_norm": 1.1952430847336142, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 39606 + }, + { + "epoch": 0.39607, + "grad_norm": 1.0464444123727785, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 39607 + }, + { + "epoch": 0.39608, + "grad_norm": 1.2073322444621708, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 39608 + }, + { + "epoch": 0.39609, + "grad_norm": 1.1040404712486618, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 39609 + }, + { + "epoch": 0.3961, + "grad_norm": 0.9694829953951618, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 39610 + }, + { + "epoch": 0.39611, + "grad_norm": 0.931653397538893, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 39611 + }, + { + "epoch": 0.39612, + "grad_norm": 0.9953907512020367, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 39612 + }, + { + "epoch": 0.39613, + "grad_norm": 1.079940613618274, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 39613 + }, + { + "epoch": 0.39614, + "grad_norm": 0.9799786311535778, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 39614 + }, + { + "epoch": 0.39615, + "grad_norm": 1.0330604593455235, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 39615 + }, + { + "epoch": 0.39616, + "grad_norm": 1.419718861956667, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 39616 + }, + { + "epoch": 0.39617, + "grad_norm": 0.8725095562711166, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 39617 + }, + { + "epoch": 0.39618, + "grad_norm": 1.2154728109360067, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 39618 + }, + { + "epoch": 0.39619, + "grad_norm": 0.9822563400377341, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 39619 + }, + { + "epoch": 0.3962, + "grad_norm": 1.1382832197857873, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 39620 + }, + { + "epoch": 0.39621, + "grad_norm": 0.9145306968686651, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 39621 + }, + { + "epoch": 0.39622, + "grad_norm": 0.9783364976295305, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 39622 + }, + { + "epoch": 0.39623, + "grad_norm": 1.1403674225461773, + "learning_rate": 0.003, + "loss": 4.031, + "step": 39623 + }, + { + "epoch": 0.39624, + "grad_norm": 1.0255821636955715, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 39624 + }, + { + "epoch": 0.39625, + "grad_norm": 1.1322959565387505, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 39625 + }, + { + "epoch": 0.39626, + "grad_norm": 1.0205097126155587, + "learning_rate": 0.003, + "loss": 4.032, + "step": 39626 + }, + { + "epoch": 0.39627, + "grad_norm": 1.450579516115374, + "learning_rate": 0.003, + "loss": 4.022, + "step": 39627 + }, + { + "epoch": 0.39628, + "grad_norm": 1.0277962692335643, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 39628 + }, + { + "epoch": 0.39629, + "grad_norm": 1.1664855364509987, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 39629 + }, + { + "epoch": 0.3963, + "grad_norm": 1.1377426016157626, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 39630 + }, + { + "epoch": 0.39631, + "grad_norm": 1.3585887144840734, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 39631 + }, + { + "epoch": 0.39632, + "grad_norm": 0.9688054509884139, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 39632 + }, + { + "epoch": 0.39633, + "grad_norm": 0.9958797941728861, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 39633 + }, + { + "epoch": 0.39634, + "grad_norm": 1.172883089098004, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 39634 + }, + { + "epoch": 0.39635, + "grad_norm": 1.1954600923006697, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 39635 + }, + { + "epoch": 0.39636, + "grad_norm": 0.9554781208001855, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 39636 + }, + { + "epoch": 0.39637, + "grad_norm": 1.039121907986708, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 39637 + }, + { + "epoch": 0.39638, + "grad_norm": 1.1938234745766647, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 39638 + }, + { + "epoch": 0.39639, + "grad_norm": 1.010944667233106, + "learning_rate": 0.003, + "loss": 4.045, + "step": 39639 + }, + { + "epoch": 0.3964, + "grad_norm": 1.2515873424431712, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 39640 + }, + { + "epoch": 0.39641, + "grad_norm": 1.0251260646182958, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 39641 + }, + { + "epoch": 0.39642, + "grad_norm": 1.1748862093535466, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 39642 + }, + { + "epoch": 0.39643, + "grad_norm": 0.9621832506327568, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 39643 + }, + { + "epoch": 0.39644, + "grad_norm": 1.0698691012351569, + "learning_rate": 0.003, + "loss": 4.058, + "step": 39644 + }, + { + "epoch": 0.39645, + "grad_norm": 1.1908846613445443, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 39645 + }, + { + "epoch": 0.39646, + "grad_norm": 1.0576864903783232, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 39646 + }, + { + "epoch": 0.39647, + "grad_norm": 1.1167341345485364, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 39647 + }, + { + "epoch": 0.39648, + "grad_norm": 0.9609702802695123, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 39648 + }, + { + "epoch": 0.39649, + "grad_norm": 1.1058448785125987, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 39649 + }, + { + "epoch": 0.3965, + "grad_norm": 1.5444658822246737, + "learning_rate": 0.003, + "loss": 4.056, + "step": 39650 + }, + { + "epoch": 0.39651, + "grad_norm": 0.8958162448946002, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 39651 + }, + { + "epoch": 0.39652, + "grad_norm": 0.8929274279277567, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 39652 + }, + { + "epoch": 0.39653, + "grad_norm": 1.0316462803308786, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 39653 + }, + { + "epoch": 0.39654, + "grad_norm": 1.4963910196996624, + "learning_rate": 0.003, + "loss": 4.029, + "step": 39654 + }, + { + "epoch": 0.39655, + "grad_norm": 0.8983194797100988, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 39655 + }, + { + "epoch": 0.39656, + "grad_norm": 0.8240426512599162, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 39656 + }, + { + "epoch": 0.39657, + "grad_norm": 1.1123464358874233, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 39657 + }, + { + "epoch": 0.39658, + "grad_norm": 1.4857318119280318, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 39658 + }, + { + "epoch": 0.39659, + "grad_norm": 0.664598512460935, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 39659 + }, + { + "epoch": 0.3966, + "grad_norm": 0.884707508499613, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 39660 + }, + { + "epoch": 0.39661, + "grad_norm": 1.0437558193868746, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 39661 + }, + { + "epoch": 0.39662, + "grad_norm": 1.0865577290301793, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 39662 + }, + { + "epoch": 0.39663, + "grad_norm": 1.0601898116853805, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 39663 + }, + { + "epoch": 0.39664, + "grad_norm": 1.1535324525770987, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 39664 + }, + { + "epoch": 0.39665, + "grad_norm": 0.89815934057396, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 39665 + }, + { + "epoch": 0.39666, + "grad_norm": 1.0155283535138466, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 39666 + }, + { + "epoch": 0.39667, + "grad_norm": 1.1546533939496493, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 39667 + }, + { + "epoch": 0.39668, + "grad_norm": 0.8060179954378502, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 39668 + }, + { + "epoch": 0.39669, + "grad_norm": 0.8211368284434252, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 39669 + }, + { + "epoch": 0.3967, + "grad_norm": 0.767433520542167, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 39670 + }, + { + "epoch": 0.39671, + "grad_norm": 0.8968512567314546, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 39671 + }, + { + "epoch": 0.39672, + "grad_norm": 1.3805098560785154, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 39672 + }, + { + "epoch": 0.39673, + "grad_norm": 1.3398579710783791, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 39673 + }, + { + "epoch": 0.39674, + "grad_norm": 1.0887424672485506, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 39674 + }, + { + "epoch": 0.39675, + "grad_norm": 1.0938625308762637, + "learning_rate": 0.003, + "loss": 4.03, + "step": 39675 + }, + { + "epoch": 0.39676, + "grad_norm": 1.2443984981528258, + "learning_rate": 0.003, + "loss": 4.036, + "step": 39676 + }, + { + "epoch": 0.39677, + "grad_norm": 1.1693938967703952, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 39677 + }, + { + "epoch": 0.39678, + "grad_norm": 0.9154052349789596, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 39678 + }, + { + "epoch": 0.39679, + "grad_norm": 0.9397850771497883, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 39679 + }, + { + "epoch": 0.3968, + "grad_norm": 1.1547885618921456, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 39680 + }, + { + "epoch": 0.39681, + "grad_norm": 1.106452179445062, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 39681 + }, + { + "epoch": 0.39682, + "grad_norm": 1.1827372708307577, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 39682 + }, + { + "epoch": 0.39683, + "grad_norm": 0.9778192559432551, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 39683 + }, + { + "epoch": 0.39684, + "grad_norm": 1.0248409656107547, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 39684 + }, + { + "epoch": 0.39685, + "grad_norm": 1.0620222190936262, + "learning_rate": 0.003, + "loss": 4.061, + "step": 39685 + }, + { + "epoch": 0.39686, + "grad_norm": 1.0694756903955096, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 39686 + }, + { + "epoch": 0.39687, + "grad_norm": 1.1832408120113547, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 39687 + }, + { + "epoch": 0.39688, + "grad_norm": 0.9787551756477284, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 39688 + }, + { + "epoch": 0.39689, + "grad_norm": 1.2430782077822031, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 39689 + }, + { + "epoch": 0.3969, + "grad_norm": 1.1086700134013627, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 39690 + }, + { + "epoch": 0.39691, + "grad_norm": 1.2823925531817084, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 39691 + }, + { + "epoch": 0.39692, + "grad_norm": 1.04396445847031, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 39692 + }, + { + "epoch": 0.39693, + "grad_norm": 1.0196153242566592, + "learning_rate": 0.003, + "loss": 3.975, + "step": 39693 + }, + { + "epoch": 0.39694, + "grad_norm": 0.998536658171138, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 39694 + }, + { + "epoch": 0.39695, + "grad_norm": 1.2418029141353175, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 39695 + }, + { + "epoch": 0.39696, + "grad_norm": 1.0805769714888493, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 39696 + }, + { + "epoch": 0.39697, + "grad_norm": 1.2546062859257825, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 39697 + }, + { + "epoch": 0.39698, + "grad_norm": 0.8804456142275197, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 39698 + }, + { + "epoch": 0.39699, + "grad_norm": 1.0054850637460406, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 39699 + }, + { + "epoch": 0.397, + "grad_norm": 1.2362253123023574, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 39700 + }, + { + "epoch": 0.39701, + "grad_norm": 1.1417422185367567, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 39701 + }, + { + "epoch": 0.39702, + "grad_norm": 0.9429420615179118, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 39702 + }, + { + "epoch": 0.39703, + "grad_norm": 0.9159439623344459, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 39703 + }, + { + "epoch": 0.39704, + "grad_norm": 1.0970204765073528, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 39704 + }, + { + "epoch": 0.39705, + "grad_norm": 0.9338922690294119, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 39705 + }, + { + "epoch": 0.39706, + "grad_norm": 1.0863669116277563, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 39706 + }, + { + "epoch": 0.39707, + "grad_norm": 1.376743884478655, + "learning_rate": 0.003, + "loss": 4.058, + "step": 39707 + }, + { + "epoch": 0.39708, + "grad_norm": 0.879590033546894, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 39708 + }, + { + "epoch": 0.39709, + "grad_norm": 0.9219673603388536, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 39709 + }, + { + "epoch": 0.3971, + "grad_norm": 1.176243451375422, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 39710 + }, + { + "epoch": 0.39711, + "grad_norm": 1.1613120480033745, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 39711 + }, + { + "epoch": 0.39712, + "grad_norm": 0.9259725974022404, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 39712 + }, + { + "epoch": 0.39713, + "grad_norm": 0.9900279003060364, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 39713 + }, + { + "epoch": 0.39714, + "grad_norm": 1.3225392860361591, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 39714 + }, + { + "epoch": 0.39715, + "grad_norm": 0.8476905970668128, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 39715 + }, + { + "epoch": 0.39716, + "grad_norm": 0.8678163505367655, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 39716 + }, + { + "epoch": 0.39717, + "grad_norm": 1.085616157458319, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 39717 + }, + { + "epoch": 0.39718, + "grad_norm": 1.1177670390872299, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 39718 + }, + { + "epoch": 0.39719, + "grad_norm": 0.9187943676766401, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 39719 + }, + { + "epoch": 0.3972, + "grad_norm": 0.90956137202729, + "learning_rate": 0.003, + "loss": 4.055, + "step": 39720 + }, + { + "epoch": 0.39721, + "grad_norm": 1.040571209104186, + "learning_rate": 0.003, + "loss": 4.018, + "step": 39721 + }, + { + "epoch": 0.39722, + "grad_norm": 1.1949670207668721, + "learning_rate": 0.003, + "loss": 4.051, + "step": 39722 + }, + { + "epoch": 0.39723, + "grad_norm": 1.1772912558290063, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 39723 + }, + { + "epoch": 0.39724, + "grad_norm": 1.1003554755757916, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 39724 + }, + { + "epoch": 0.39725, + "grad_norm": 1.1535387779761028, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 39725 + }, + { + "epoch": 0.39726, + "grad_norm": 1.0152084939295647, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 39726 + }, + { + "epoch": 0.39727, + "grad_norm": 1.1594581011075822, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 39727 + }, + { + "epoch": 0.39728, + "grad_norm": 0.9529334241858239, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 39728 + }, + { + "epoch": 0.39729, + "grad_norm": 1.0967384260713915, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 39729 + }, + { + "epoch": 0.3973, + "grad_norm": 1.2311519847631716, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 39730 + }, + { + "epoch": 0.39731, + "grad_norm": 1.0068838367075692, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 39731 + }, + { + "epoch": 0.39732, + "grad_norm": 1.1971870759211252, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 39732 + }, + { + "epoch": 0.39733, + "grad_norm": 1.0129425215017311, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 39733 + }, + { + "epoch": 0.39734, + "grad_norm": 1.1052406590646224, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 39734 + }, + { + "epoch": 0.39735, + "grad_norm": 1.103262018191955, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 39735 + }, + { + "epoch": 0.39736, + "grad_norm": 0.9310832873225584, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 39736 + }, + { + "epoch": 0.39737, + "grad_norm": 1.0433478401433838, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 39737 + }, + { + "epoch": 0.39738, + "grad_norm": 1.192606512203675, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 39738 + }, + { + "epoch": 0.39739, + "grad_norm": 1.0221150719394643, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 39739 + }, + { + "epoch": 0.3974, + "grad_norm": 1.1529209685740767, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 39740 + }, + { + "epoch": 0.39741, + "grad_norm": 1.1067917039599113, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 39741 + }, + { + "epoch": 0.39742, + "grad_norm": 1.002409073823219, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 39742 + }, + { + "epoch": 0.39743, + "grad_norm": 1.3054344045044763, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 39743 + }, + { + "epoch": 0.39744, + "grad_norm": 1.0587730053575748, + "learning_rate": 0.003, + "loss": 4.041, + "step": 39744 + }, + { + "epoch": 0.39745, + "grad_norm": 1.0016531323920153, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 39745 + }, + { + "epoch": 0.39746, + "grad_norm": 1.1082221950904894, + "learning_rate": 0.003, + "loss": 4.019, + "step": 39746 + }, + { + "epoch": 0.39747, + "grad_norm": 1.5340259205890485, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 39747 + }, + { + "epoch": 0.39748, + "grad_norm": 0.7990119182493673, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 39748 + }, + { + "epoch": 0.39749, + "grad_norm": 0.8075425200031595, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 39749 + }, + { + "epoch": 0.3975, + "grad_norm": 0.8402894901715633, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 39750 + }, + { + "epoch": 0.39751, + "grad_norm": 0.8834198503084629, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 39751 + }, + { + "epoch": 0.39752, + "grad_norm": 0.8968693454601242, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 39752 + }, + { + "epoch": 0.39753, + "grad_norm": 0.9896411566230473, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 39753 + }, + { + "epoch": 0.39754, + "grad_norm": 1.3404875272090375, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 39754 + }, + { + "epoch": 0.39755, + "grad_norm": 0.9147481647214385, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 39755 + }, + { + "epoch": 0.39756, + "grad_norm": 0.9801890004572759, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 39756 + }, + { + "epoch": 0.39757, + "grad_norm": 1.0826617131283038, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 39757 + }, + { + "epoch": 0.39758, + "grad_norm": 1.1634508557301768, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 39758 + }, + { + "epoch": 0.39759, + "grad_norm": 1.0729035647607956, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 39759 + }, + { + "epoch": 0.3976, + "grad_norm": 1.266363562832313, + "learning_rate": 0.003, + "loss": 4.068, + "step": 39760 + }, + { + "epoch": 0.39761, + "grad_norm": 1.094484392381517, + "learning_rate": 0.003, + "loss": 4.034, + "step": 39761 + }, + { + "epoch": 0.39762, + "grad_norm": 1.022869546470203, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 39762 + }, + { + "epoch": 0.39763, + "grad_norm": 0.9696334289733403, + "learning_rate": 0.003, + "loss": 4.046, + "step": 39763 + }, + { + "epoch": 0.39764, + "grad_norm": 1.1981274660463794, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 39764 + }, + { + "epoch": 0.39765, + "grad_norm": 1.1798574925619145, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 39765 + }, + { + "epoch": 0.39766, + "grad_norm": 1.0527343961476368, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 39766 + }, + { + "epoch": 0.39767, + "grad_norm": 1.075970363161624, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 39767 + }, + { + "epoch": 0.39768, + "grad_norm": 1.2179199534826468, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 39768 + }, + { + "epoch": 0.39769, + "grad_norm": 0.942575276403349, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 39769 + }, + { + "epoch": 0.3977, + "grad_norm": 1.066498469157128, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 39770 + }, + { + "epoch": 0.39771, + "grad_norm": 0.9155704720022676, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 39771 + }, + { + "epoch": 0.39772, + "grad_norm": 1.0265688000049973, + "learning_rate": 0.003, + "loss": 4.026, + "step": 39772 + }, + { + "epoch": 0.39773, + "grad_norm": 1.194779599310907, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 39773 + }, + { + "epoch": 0.39774, + "grad_norm": 1.1411833424526507, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 39774 + }, + { + "epoch": 0.39775, + "grad_norm": 1.2086415461810673, + "learning_rate": 0.003, + "loss": 4.057, + "step": 39775 + }, + { + "epoch": 0.39776, + "grad_norm": 0.9010032585516328, + "learning_rate": 0.003, + "loss": 4.036, + "step": 39776 + }, + { + "epoch": 0.39777, + "grad_norm": 1.0109551566049295, + "learning_rate": 0.003, + "loss": 4.032, + "step": 39777 + }, + { + "epoch": 0.39778, + "grad_norm": 1.310373912326102, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 39778 + }, + { + "epoch": 0.39779, + "grad_norm": 1.2213691438198429, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 39779 + }, + { + "epoch": 0.3978, + "grad_norm": 1.025568654424296, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 39780 + }, + { + "epoch": 0.39781, + "grad_norm": 0.9321442772777632, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 39781 + }, + { + "epoch": 0.39782, + "grad_norm": 0.9586437555130412, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 39782 + }, + { + "epoch": 0.39783, + "grad_norm": 1.0628080065307792, + "learning_rate": 0.003, + "loss": 4.06, + "step": 39783 + }, + { + "epoch": 0.39784, + "grad_norm": 0.9669524046235881, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 39784 + }, + { + "epoch": 0.39785, + "grad_norm": 1.0612588707538253, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 39785 + }, + { + "epoch": 0.39786, + "grad_norm": 1.2479952198658137, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 39786 + }, + { + "epoch": 0.39787, + "grad_norm": 1.0271780674992066, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 39787 + }, + { + "epoch": 0.39788, + "grad_norm": 1.1912435196388318, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 39788 + }, + { + "epoch": 0.39789, + "grad_norm": 0.9786123067058274, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 39789 + }, + { + "epoch": 0.3979, + "grad_norm": 1.0811937722848681, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 39790 + }, + { + "epoch": 0.39791, + "grad_norm": 1.2059817619775104, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 39791 + }, + { + "epoch": 0.39792, + "grad_norm": 0.7379483220865667, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 39792 + }, + { + "epoch": 0.39793, + "grad_norm": 0.747095284747321, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 39793 + }, + { + "epoch": 0.39794, + "grad_norm": 0.8518706924952681, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 39794 + }, + { + "epoch": 0.39795, + "grad_norm": 1.2281346756315397, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 39795 + }, + { + "epoch": 0.39796, + "grad_norm": 1.217327646627782, + "learning_rate": 0.003, + "loss": 4.022, + "step": 39796 + }, + { + "epoch": 0.39797, + "grad_norm": 1.1540579152519954, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 39797 + }, + { + "epoch": 0.39798, + "grad_norm": 1.27537650395604, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 39798 + }, + { + "epoch": 0.39799, + "grad_norm": 0.8230159383519529, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 39799 + }, + { + "epoch": 0.398, + "grad_norm": 0.7357573912580084, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 39800 + }, + { + "epoch": 0.39801, + "grad_norm": 0.8108335684749521, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 39801 + }, + { + "epoch": 0.39802, + "grad_norm": 0.9210208763259632, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 39802 + }, + { + "epoch": 0.39803, + "grad_norm": 1.2975152762175262, + "learning_rate": 0.003, + "loss": 4.03, + "step": 39803 + }, + { + "epoch": 0.39804, + "grad_norm": 0.937728742687371, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 39804 + }, + { + "epoch": 0.39805, + "grad_norm": 1.0094284972804124, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 39805 + }, + { + "epoch": 0.39806, + "grad_norm": 1.5256907586189337, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 39806 + }, + { + "epoch": 0.39807, + "grad_norm": 1.0015277494047392, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 39807 + }, + { + "epoch": 0.39808, + "grad_norm": 1.082866955913923, + "learning_rate": 0.003, + "loss": 4.046, + "step": 39808 + }, + { + "epoch": 0.39809, + "grad_norm": 1.09319640131449, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 39809 + }, + { + "epoch": 0.3981, + "grad_norm": 1.0405657265572836, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 39810 + }, + { + "epoch": 0.39811, + "grad_norm": 1.11521276632414, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 39811 + }, + { + "epoch": 0.39812, + "grad_norm": 1.103620694120051, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 39812 + }, + { + "epoch": 0.39813, + "grad_norm": 1.1147270945876462, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 39813 + }, + { + "epoch": 0.39814, + "grad_norm": 1.0012326798406865, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 39814 + }, + { + "epoch": 0.39815, + "grad_norm": 1.097650967452605, + "learning_rate": 0.003, + "loss": 4.048, + "step": 39815 + }, + { + "epoch": 0.39816, + "grad_norm": 1.1768707053750767, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 39816 + }, + { + "epoch": 0.39817, + "grad_norm": 1.0736546642297473, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 39817 + }, + { + "epoch": 0.39818, + "grad_norm": 0.9981344131940335, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 39818 + }, + { + "epoch": 0.39819, + "grad_norm": 0.95023419379907, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 39819 + }, + { + "epoch": 0.3982, + "grad_norm": 1.1526960956525822, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 39820 + }, + { + "epoch": 0.39821, + "grad_norm": 1.1680273172472173, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 39821 + }, + { + "epoch": 0.39822, + "grad_norm": 1.022921374564099, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 39822 + }, + { + "epoch": 0.39823, + "grad_norm": 1.2155741920180905, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 39823 + }, + { + "epoch": 0.39824, + "grad_norm": 0.9293112617544728, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 39824 + }, + { + "epoch": 0.39825, + "grad_norm": 0.9703752146494233, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 39825 + }, + { + "epoch": 0.39826, + "grad_norm": 1.2154265652910723, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 39826 + }, + { + "epoch": 0.39827, + "grad_norm": 0.8951844646207766, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 39827 + }, + { + "epoch": 0.39828, + "grad_norm": 1.1136975383512193, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 39828 + }, + { + "epoch": 0.39829, + "grad_norm": 1.307325212642836, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 39829 + }, + { + "epoch": 0.3983, + "grad_norm": 1.0938683580994049, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 39830 + }, + { + "epoch": 0.39831, + "grad_norm": 0.9765392724653817, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 39831 + }, + { + "epoch": 0.39832, + "grad_norm": 1.0595165261648503, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 39832 + }, + { + "epoch": 0.39833, + "grad_norm": 1.2932334713977096, + "learning_rate": 0.003, + "loss": 4.067, + "step": 39833 + }, + { + "epoch": 0.39834, + "grad_norm": 1.1090736401488122, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 39834 + }, + { + "epoch": 0.39835, + "grad_norm": 1.3615919041376436, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 39835 + }, + { + "epoch": 0.39836, + "grad_norm": 0.8789781419466771, + "learning_rate": 0.003, + "loss": 4.044, + "step": 39836 + }, + { + "epoch": 0.39837, + "grad_norm": 0.8255879205176726, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 39837 + }, + { + "epoch": 0.39838, + "grad_norm": 0.9348806894202902, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 39838 + }, + { + "epoch": 0.39839, + "grad_norm": 1.1702383380813521, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 39839 + }, + { + "epoch": 0.3984, + "grad_norm": 1.0786041131170678, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 39840 + }, + { + "epoch": 0.39841, + "grad_norm": 1.3691727853892741, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 39841 + }, + { + "epoch": 0.39842, + "grad_norm": 0.9772139387325627, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 39842 + }, + { + "epoch": 0.39843, + "grad_norm": 1.0638998076190183, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 39843 + }, + { + "epoch": 0.39844, + "grad_norm": 0.9025875067283418, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 39844 + }, + { + "epoch": 0.39845, + "grad_norm": 1.0556948940602324, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 39845 + }, + { + "epoch": 0.39846, + "grad_norm": 1.1621377462068303, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 39846 + }, + { + "epoch": 0.39847, + "grad_norm": 0.9711036268473122, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 39847 + }, + { + "epoch": 0.39848, + "grad_norm": 1.3258937660393912, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 39848 + }, + { + "epoch": 0.39849, + "grad_norm": 0.8882936179504, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 39849 + }, + { + "epoch": 0.3985, + "grad_norm": 1.1117153428776987, + "learning_rate": 0.003, + "loss": 4.044, + "step": 39850 + }, + { + "epoch": 0.39851, + "grad_norm": 1.0624654446943613, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 39851 + }, + { + "epoch": 0.39852, + "grad_norm": 1.2224734462744622, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 39852 + }, + { + "epoch": 0.39853, + "grad_norm": 1.1550790600700798, + "learning_rate": 0.003, + "loss": 4.035, + "step": 39853 + }, + { + "epoch": 0.39854, + "grad_norm": 1.1542763497764803, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 39854 + }, + { + "epoch": 0.39855, + "grad_norm": 1.2011467280025987, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 39855 + }, + { + "epoch": 0.39856, + "grad_norm": 1.079631142156973, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 39856 + }, + { + "epoch": 0.39857, + "grad_norm": 1.0149442272648228, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 39857 + }, + { + "epoch": 0.39858, + "grad_norm": 1.1316322201590492, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 39858 + }, + { + "epoch": 0.39859, + "grad_norm": 0.9701373105396969, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 39859 + }, + { + "epoch": 0.3986, + "grad_norm": 1.0322686288908034, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 39860 + }, + { + "epoch": 0.39861, + "grad_norm": 1.0858062955830148, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 39861 + }, + { + "epoch": 0.39862, + "grad_norm": 0.9573948131455293, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 39862 + }, + { + "epoch": 0.39863, + "grad_norm": 1.1131255714349013, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 39863 + }, + { + "epoch": 0.39864, + "grad_norm": 1.0751081276060404, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 39864 + }, + { + "epoch": 0.39865, + "grad_norm": 1.0658847052625682, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 39865 + }, + { + "epoch": 0.39866, + "grad_norm": 1.016170452912719, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 39866 + }, + { + "epoch": 0.39867, + "grad_norm": 0.9291457293158364, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 39867 + }, + { + "epoch": 0.39868, + "grad_norm": 1.0708944299978815, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 39868 + }, + { + "epoch": 0.39869, + "grad_norm": 1.2687698537759957, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 39869 + }, + { + "epoch": 0.3987, + "grad_norm": 0.9962810236436851, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 39870 + }, + { + "epoch": 0.39871, + "grad_norm": 1.1755571105153704, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 39871 + }, + { + "epoch": 0.39872, + "grad_norm": 1.0991105758188746, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 39872 + }, + { + "epoch": 0.39873, + "grad_norm": 1.0584952348798762, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 39873 + }, + { + "epoch": 0.39874, + "grad_norm": 1.2069362734882474, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 39874 + }, + { + "epoch": 0.39875, + "grad_norm": 1.2198218736025326, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 39875 + }, + { + "epoch": 0.39876, + "grad_norm": 0.907217583666471, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 39876 + }, + { + "epoch": 0.39877, + "grad_norm": 0.9179253258543691, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 39877 + }, + { + "epoch": 0.39878, + "grad_norm": 1.1088438198307267, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 39878 + }, + { + "epoch": 0.39879, + "grad_norm": 1.3410345081771493, + "learning_rate": 0.003, + "loss": 4.058, + "step": 39879 + }, + { + "epoch": 0.3988, + "grad_norm": 1.0609155122426948, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 39880 + }, + { + "epoch": 0.39881, + "grad_norm": 1.0356053850247982, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 39881 + }, + { + "epoch": 0.39882, + "grad_norm": 1.1681980196944262, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 39882 + }, + { + "epoch": 0.39883, + "grad_norm": 1.0331840760228985, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 39883 + }, + { + "epoch": 0.39884, + "grad_norm": 1.0523757447800526, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 39884 + }, + { + "epoch": 0.39885, + "grad_norm": 0.9928004428404028, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 39885 + }, + { + "epoch": 0.39886, + "grad_norm": 1.1788695864737557, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 39886 + }, + { + "epoch": 0.39887, + "grad_norm": 0.9753767420021379, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 39887 + }, + { + "epoch": 0.39888, + "grad_norm": 1.0221690889359691, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 39888 + }, + { + "epoch": 0.39889, + "grad_norm": 1.3202306706600784, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 39889 + }, + { + "epoch": 0.3989, + "grad_norm": 0.8811730812258731, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 39890 + }, + { + "epoch": 0.39891, + "grad_norm": 0.6952206678279327, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 39891 + }, + { + "epoch": 0.39892, + "grad_norm": 0.7981372270711562, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 39892 + }, + { + "epoch": 0.39893, + "grad_norm": 0.9353998877671835, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 39893 + }, + { + "epoch": 0.39894, + "grad_norm": 1.306182035653656, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 39894 + }, + { + "epoch": 0.39895, + "grad_norm": 1.066865547330856, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 39895 + }, + { + "epoch": 0.39896, + "grad_norm": 1.0089998574792345, + "learning_rate": 0.003, + "loss": 4.064, + "step": 39896 + }, + { + "epoch": 0.39897, + "grad_norm": 1.1131882796274393, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 39897 + }, + { + "epoch": 0.39898, + "grad_norm": 1.0175172816674845, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 39898 + }, + { + "epoch": 0.39899, + "grad_norm": 0.8893695744273166, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 39899 + }, + { + "epoch": 0.399, + "grad_norm": 1.0257444569328713, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 39900 + }, + { + "epoch": 0.39901, + "grad_norm": 1.0964033550909693, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 39901 + }, + { + "epoch": 0.39902, + "grad_norm": 0.9739958652833106, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 39902 + }, + { + "epoch": 0.39903, + "grad_norm": 1.0668728175700155, + "learning_rate": 0.003, + "loss": 4.042, + "step": 39903 + }, + { + "epoch": 0.39904, + "grad_norm": 1.0966664827508952, + "learning_rate": 0.003, + "loss": 4.067, + "step": 39904 + }, + { + "epoch": 0.39905, + "grad_norm": 1.1562291291551743, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 39905 + }, + { + "epoch": 0.39906, + "grad_norm": 0.9364615135101625, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 39906 + }, + { + "epoch": 0.39907, + "grad_norm": 1.167633505521684, + "learning_rate": 0.003, + "loss": 4.018, + "step": 39907 + }, + { + "epoch": 0.39908, + "grad_norm": 0.9169974183738576, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 39908 + }, + { + "epoch": 0.39909, + "grad_norm": 0.9195181883547996, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 39909 + }, + { + "epoch": 0.3991, + "grad_norm": 1.0550664345916685, + "learning_rate": 0.003, + "loss": 4.025, + "step": 39910 + }, + { + "epoch": 0.39911, + "grad_norm": 0.8806217166954874, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 39911 + }, + { + "epoch": 0.39912, + "grad_norm": 0.9564525304655477, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 39912 + }, + { + "epoch": 0.39913, + "grad_norm": 1.0584143869015363, + "learning_rate": 0.003, + "loss": 4.051, + "step": 39913 + }, + { + "epoch": 0.39914, + "grad_norm": 1.0887001793161208, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 39914 + }, + { + "epoch": 0.39915, + "grad_norm": 1.2534037392756947, + "learning_rate": 0.003, + "loss": 4.045, + "step": 39915 + }, + { + "epoch": 0.39916, + "grad_norm": 1.1545120105687088, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 39916 + }, + { + "epoch": 0.39917, + "grad_norm": 1.1415269654736049, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 39917 + }, + { + "epoch": 0.39918, + "grad_norm": 0.9586458439007999, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 39918 + }, + { + "epoch": 0.39919, + "grad_norm": 1.1794838488111203, + "learning_rate": 0.003, + "loss": 4.032, + "step": 39919 + }, + { + "epoch": 0.3992, + "grad_norm": 1.3558792180928096, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 39920 + }, + { + "epoch": 0.39921, + "grad_norm": 1.016948967167559, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 39921 + }, + { + "epoch": 0.39922, + "grad_norm": 1.1740419876163843, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 39922 + }, + { + "epoch": 0.39923, + "grad_norm": 0.9690652183591144, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 39923 + }, + { + "epoch": 0.39924, + "grad_norm": 1.1868700354263928, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 39924 + }, + { + "epoch": 0.39925, + "grad_norm": 1.0654514767339969, + "learning_rate": 0.003, + "loss": 4.013, + "step": 39925 + }, + { + "epoch": 0.39926, + "grad_norm": 1.0292596925807447, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 39926 + }, + { + "epoch": 0.39927, + "grad_norm": 1.0642378134831563, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 39927 + }, + { + "epoch": 0.39928, + "grad_norm": 1.019244095225631, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 39928 + }, + { + "epoch": 0.39929, + "grad_norm": 0.9379074439185108, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 39929 + }, + { + "epoch": 0.3993, + "grad_norm": 1.11131038112159, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 39930 + }, + { + "epoch": 0.39931, + "grad_norm": 1.2040438206223418, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 39931 + }, + { + "epoch": 0.39932, + "grad_norm": 0.8992142282644267, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 39932 + }, + { + "epoch": 0.39933, + "grad_norm": 1.0324862748660595, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 39933 + }, + { + "epoch": 0.39934, + "grad_norm": 1.3190490866189073, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 39934 + }, + { + "epoch": 0.39935, + "grad_norm": 0.8085707175545562, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 39935 + }, + { + "epoch": 0.39936, + "grad_norm": 0.7684754753543303, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 39936 + }, + { + "epoch": 0.39937, + "grad_norm": 1.0717942021591764, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 39937 + }, + { + "epoch": 0.39938, + "grad_norm": 1.401419760004852, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 39938 + }, + { + "epoch": 0.39939, + "grad_norm": 0.7592558524967433, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 39939 + }, + { + "epoch": 0.3994, + "grad_norm": 0.8714708603559385, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 39940 + }, + { + "epoch": 0.39941, + "grad_norm": 1.2309974082077237, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 39941 + }, + { + "epoch": 0.39942, + "grad_norm": 1.1756592421248253, + "learning_rate": 0.003, + "loss": 4.038, + "step": 39942 + }, + { + "epoch": 0.39943, + "grad_norm": 1.1011809938097428, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 39943 + }, + { + "epoch": 0.39944, + "grad_norm": 1.1513038146367052, + "learning_rate": 0.003, + "loss": 4.024, + "step": 39944 + }, + { + "epoch": 0.39945, + "grad_norm": 1.1914817880548998, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 39945 + }, + { + "epoch": 0.39946, + "grad_norm": 1.0402103266031868, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 39946 + }, + { + "epoch": 0.39947, + "grad_norm": 1.1043674367483671, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 39947 + }, + { + "epoch": 0.39948, + "grad_norm": 0.9957343739345585, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 39948 + }, + { + "epoch": 0.39949, + "grad_norm": 1.1092392585126587, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 39949 + }, + { + "epoch": 0.3995, + "grad_norm": 1.047147638788959, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 39950 + }, + { + "epoch": 0.39951, + "grad_norm": 1.1111514772129742, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 39951 + }, + { + "epoch": 0.39952, + "grad_norm": 1.0662568607880605, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 39952 + }, + { + "epoch": 0.39953, + "grad_norm": 1.2133092240498462, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 39953 + }, + { + "epoch": 0.39954, + "grad_norm": 0.936471627980977, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 39954 + }, + { + "epoch": 0.39955, + "grad_norm": 0.9762396263832235, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 39955 + }, + { + "epoch": 0.39956, + "grad_norm": 1.1499843726538292, + "learning_rate": 0.003, + "loss": 4.023, + "step": 39956 + }, + { + "epoch": 0.39957, + "grad_norm": 1.1077187473654617, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 39957 + }, + { + "epoch": 0.39958, + "grad_norm": 1.1672253630513634, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 39958 + }, + { + "epoch": 0.39959, + "grad_norm": 1.1345758012390381, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 39959 + }, + { + "epoch": 0.3996, + "grad_norm": 1.064355898269221, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 39960 + }, + { + "epoch": 0.39961, + "grad_norm": 1.2219553173016728, + "learning_rate": 0.003, + "loss": 4.007, + "step": 39961 + }, + { + "epoch": 0.39962, + "grad_norm": 0.933146910877789, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 39962 + }, + { + "epoch": 0.39963, + "grad_norm": 1.047745627545263, + "learning_rate": 0.003, + "loss": 4.039, + "step": 39963 + }, + { + "epoch": 0.39964, + "grad_norm": 0.8690900307835908, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 39964 + }, + { + "epoch": 0.39965, + "grad_norm": 0.8761607254136945, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 39965 + }, + { + "epoch": 0.39966, + "grad_norm": 0.9651014028074766, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 39966 + }, + { + "epoch": 0.39967, + "grad_norm": 1.0623382538073016, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 39967 + }, + { + "epoch": 0.39968, + "grad_norm": 1.1679309647680096, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 39968 + }, + { + "epoch": 0.39969, + "grad_norm": 1.0616327515487223, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 39969 + }, + { + "epoch": 0.3997, + "grad_norm": 1.305861029023507, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 39970 + }, + { + "epoch": 0.39971, + "grad_norm": 1.0655761590626511, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 39971 + }, + { + "epoch": 0.39972, + "grad_norm": 1.1173840744108254, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 39972 + }, + { + "epoch": 0.39973, + "grad_norm": 1.0185761811296898, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 39973 + }, + { + "epoch": 0.39974, + "grad_norm": 1.2434358027843573, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 39974 + }, + { + "epoch": 0.39975, + "grad_norm": 1.0786062766644287, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 39975 + }, + { + "epoch": 0.39976, + "grad_norm": 1.0378870378695186, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 39976 + }, + { + "epoch": 0.39977, + "grad_norm": 1.0490091459554507, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 39977 + }, + { + "epoch": 0.39978, + "grad_norm": 1.1931343821145792, + "learning_rate": 0.003, + "loss": 4.035, + "step": 39978 + }, + { + "epoch": 0.39979, + "grad_norm": 1.1607791813441373, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 39979 + }, + { + "epoch": 0.3998, + "grad_norm": 0.9658856394359351, + "learning_rate": 0.003, + "loss": 4.043, + "step": 39980 + }, + { + "epoch": 0.39981, + "grad_norm": 1.0037408060873756, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 39981 + }, + { + "epoch": 0.39982, + "grad_norm": 1.1595112173812303, + "learning_rate": 0.003, + "loss": 4.032, + "step": 39982 + }, + { + "epoch": 0.39983, + "grad_norm": 0.8188808491395301, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 39983 + }, + { + "epoch": 0.39984, + "grad_norm": 0.8004996989576377, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 39984 + }, + { + "epoch": 0.39985, + "grad_norm": 0.9011379043310025, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 39985 + }, + { + "epoch": 0.39986, + "grad_norm": 1.2332232899979865, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 39986 + }, + { + "epoch": 0.39987, + "grad_norm": 1.1520882697282435, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 39987 + }, + { + "epoch": 0.39988, + "grad_norm": 1.3899501753656656, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 39988 + }, + { + "epoch": 0.39989, + "grad_norm": 0.8747619661649791, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 39989 + }, + { + "epoch": 0.3999, + "grad_norm": 1.0036474257203374, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 39990 + }, + { + "epoch": 0.39991, + "grad_norm": 1.3517506973568174, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 39991 + }, + { + "epoch": 0.39992, + "grad_norm": 1.0875199261359716, + "learning_rate": 0.003, + "loss": 4.036, + "step": 39992 + }, + { + "epoch": 0.39993, + "grad_norm": 1.0061367739157019, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 39993 + }, + { + "epoch": 0.39994, + "grad_norm": 0.8795462748632472, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 39994 + }, + { + "epoch": 0.39995, + "grad_norm": 0.8779399496713277, + "learning_rate": 0.003, + "loss": 4.04, + "step": 39995 + }, + { + "epoch": 0.39996, + "grad_norm": 1.0041740200325149, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 39996 + }, + { + "epoch": 0.39997, + "grad_norm": 0.9548865592065253, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 39997 + }, + { + "epoch": 0.39998, + "grad_norm": 1.0482354305827049, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 39998 + }, + { + "epoch": 0.39999, + "grad_norm": 1.0431627790540419, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 39999 + }, + { + "epoch": 0.4, + "grad_norm": 1.2621801137114688, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 40000 + }, + { + "epoch": 0.40001, + "grad_norm": 1.100340433192385, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 40001 + }, + { + "epoch": 0.40002, + "grad_norm": 1.2444595616967473, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 40002 + }, + { + "epoch": 0.40003, + "grad_norm": 0.8861094675657635, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 40003 + }, + { + "epoch": 0.40004, + "grad_norm": 0.9022324813108271, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 40004 + }, + { + "epoch": 0.40005, + "grad_norm": 1.2182389732837853, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 40005 + }, + { + "epoch": 0.40006, + "grad_norm": 1.4966423777830726, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 40006 + }, + { + "epoch": 0.40007, + "grad_norm": 0.8300884488445457, + "learning_rate": 0.003, + "loss": 4.042, + "step": 40007 + }, + { + "epoch": 0.40008, + "grad_norm": 0.7688656301472613, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 40008 + }, + { + "epoch": 0.40009, + "grad_norm": 1.068372663714648, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 40009 + }, + { + "epoch": 0.4001, + "grad_norm": 1.6528733188406375, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 40010 + }, + { + "epoch": 0.40011, + "grad_norm": 0.7376470878936588, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 40011 + }, + { + "epoch": 0.40012, + "grad_norm": 0.7614922695625664, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 40012 + }, + { + "epoch": 0.40013, + "grad_norm": 0.9475934917947009, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 40013 + }, + { + "epoch": 0.40014, + "grad_norm": 0.9971950033343193, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 40014 + }, + { + "epoch": 0.40015, + "grad_norm": 1.0696704631690377, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 40015 + }, + { + "epoch": 0.40016, + "grad_norm": 0.9839538048166216, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 40016 + }, + { + "epoch": 0.40017, + "grad_norm": 1.314466385257643, + "learning_rate": 0.003, + "loss": 4.018, + "step": 40017 + }, + { + "epoch": 0.40018, + "grad_norm": 1.1822868421690162, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 40018 + }, + { + "epoch": 0.40019, + "grad_norm": 1.1316331645155064, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 40019 + }, + { + "epoch": 0.4002, + "grad_norm": 0.9566515831716306, + "learning_rate": 0.003, + "loss": 4.04, + "step": 40020 + }, + { + "epoch": 0.40021, + "grad_norm": 1.1354654483756692, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 40021 + }, + { + "epoch": 0.40022, + "grad_norm": 1.0770912019070569, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 40022 + }, + { + "epoch": 0.40023, + "grad_norm": 0.9652632166078754, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 40023 + }, + { + "epoch": 0.40024, + "grad_norm": 0.9595576667343156, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 40024 + }, + { + "epoch": 0.40025, + "grad_norm": 1.2801638575049288, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 40025 + }, + { + "epoch": 0.40026, + "grad_norm": 1.1853767692896204, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 40026 + }, + { + "epoch": 0.40027, + "grad_norm": 0.8318273149698427, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 40027 + }, + { + "epoch": 0.40028, + "grad_norm": 0.9222402290463555, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 40028 + }, + { + "epoch": 0.40029, + "grad_norm": 1.2502194067384322, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 40029 + }, + { + "epoch": 0.4003, + "grad_norm": 1.0301952544493203, + "learning_rate": 0.003, + "loss": 4.024, + "step": 40030 + }, + { + "epoch": 0.40031, + "grad_norm": 1.1423214198950054, + "learning_rate": 0.003, + "loss": 3.995, + "step": 40031 + }, + { + "epoch": 0.40032, + "grad_norm": 1.2418414780267872, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 40032 + }, + { + "epoch": 0.40033, + "grad_norm": 1.1794360992053978, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 40033 + }, + { + "epoch": 0.40034, + "grad_norm": 0.8634305963725092, + "learning_rate": 0.003, + "loss": 4.048, + "step": 40034 + }, + { + "epoch": 0.40035, + "grad_norm": 0.9508176541318419, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 40035 + }, + { + "epoch": 0.40036, + "grad_norm": 1.142311944496612, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 40036 + }, + { + "epoch": 0.40037, + "grad_norm": 1.0579443632218448, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 40037 + }, + { + "epoch": 0.40038, + "grad_norm": 1.0519923009617151, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 40038 + }, + { + "epoch": 0.40039, + "grad_norm": 1.046738290464452, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 40039 + }, + { + "epoch": 0.4004, + "grad_norm": 1.2981617640688854, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 40040 + }, + { + "epoch": 0.40041, + "grad_norm": 1.1729747706397333, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 40041 + }, + { + "epoch": 0.40042, + "grad_norm": 1.093267107995101, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 40042 + }, + { + "epoch": 0.40043, + "grad_norm": 1.139601721397015, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 40043 + }, + { + "epoch": 0.40044, + "grad_norm": 1.119327750397834, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 40044 + }, + { + "epoch": 0.40045, + "grad_norm": 1.0395716715665817, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 40045 + }, + { + "epoch": 0.40046, + "grad_norm": 1.183927433991117, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 40046 + }, + { + "epoch": 0.40047, + "grad_norm": 1.0583138051639125, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 40047 + }, + { + "epoch": 0.40048, + "grad_norm": 1.2425036938343297, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 40048 + }, + { + "epoch": 0.40049, + "grad_norm": 1.0938526230916923, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 40049 + }, + { + "epoch": 0.4005, + "grad_norm": 1.1034132150087501, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 40050 + }, + { + "epoch": 0.40051, + "grad_norm": 1.0507971603528843, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 40051 + }, + { + "epoch": 0.40052, + "grad_norm": 1.2694817425912566, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 40052 + }, + { + "epoch": 0.40053, + "grad_norm": 1.0856017556335904, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 40053 + }, + { + "epoch": 0.40054, + "grad_norm": 1.010585504333268, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 40054 + }, + { + "epoch": 0.40055, + "grad_norm": 1.2111438315474388, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 40055 + }, + { + "epoch": 0.40056, + "grad_norm": 0.894205452931771, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 40056 + }, + { + "epoch": 0.40057, + "grad_norm": 0.9228076840751761, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 40057 + }, + { + "epoch": 0.40058, + "grad_norm": 0.9996805135502699, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 40058 + }, + { + "epoch": 0.40059, + "grad_norm": 1.2322049428183597, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 40059 + }, + { + "epoch": 0.4006, + "grad_norm": 0.8967282688522696, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 40060 + }, + { + "epoch": 0.40061, + "grad_norm": 0.8564406341371787, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 40061 + }, + { + "epoch": 0.40062, + "grad_norm": 1.040399906360578, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 40062 + }, + { + "epoch": 0.40063, + "grad_norm": 1.1302028228402166, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 40063 + }, + { + "epoch": 0.40064, + "grad_norm": 1.0146925497724186, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 40064 + }, + { + "epoch": 0.40065, + "grad_norm": 1.2509055175931736, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 40065 + }, + { + "epoch": 0.40066, + "grad_norm": 1.149079931603739, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 40066 + }, + { + "epoch": 0.40067, + "grad_norm": 1.1100068313420122, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 40067 + }, + { + "epoch": 0.40068, + "grad_norm": 1.24770206980221, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 40068 + }, + { + "epoch": 0.40069, + "grad_norm": 0.9989347328805772, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 40069 + }, + { + "epoch": 0.4007, + "grad_norm": 1.3063739314217602, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 40070 + }, + { + "epoch": 0.40071, + "grad_norm": 0.9216969110136722, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 40071 + }, + { + "epoch": 0.40072, + "grad_norm": 0.9311665473788305, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 40072 + }, + { + "epoch": 0.40073, + "grad_norm": 1.226500851042059, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 40073 + }, + { + "epoch": 0.40074, + "grad_norm": 1.0708873584734377, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 40074 + }, + { + "epoch": 0.40075, + "grad_norm": 0.9946208452760404, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 40075 + }, + { + "epoch": 0.40076, + "grad_norm": 1.0260917491233208, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 40076 + }, + { + "epoch": 0.40077, + "grad_norm": 1.1201614799602135, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 40077 + }, + { + "epoch": 0.40078, + "grad_norm": 1.0236990505182775, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 40078 + }, + { + "epoch": 0.40079, + "grad_norm": 1.215463324659275, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 40079 + }, + { + "epoch": 0.4008, + "grad_norm": 1.0441316871034936, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 40080 + }, + { + "epoch": 0.40081, + "grad_norm": 1.2108456915139374, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 40081 + }, + { + "epoch": 0.40082, + "grad_norm": 1.3590486485431659, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 40082 + }, + { + "epoch": 0.40083, + "grad_norm": 0.8215622999997734, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 40083 + }, + { + "epoch": 0.40084, + "grad_norm": 0.8298093878403794, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 40084 + }, + { + "epoch": 0.40085, + "grad_norm": 1.0852040341018343, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 40085 + }, + { + "epoch": 0.40086, + "grad_norm": 1.2273728258289638, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 40086 + }, + { + "epoch": 0.40087, + "grad_norm": 0.9648295338477229, + "learning_rate": 0.003, + "loss": 4.016, + "step": 40087 + }, + { + "epoch": 0.40088, + "grad_norm": 1.1835211713356544, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 40088 + }, + { + "epoch": 0.40089, + "grad_norm": 1.0201292113407259, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 40089 + }, + { + "epoch": 0.4009, + "grad_norm": 1.0954500054992973, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 40090 + }, + { + "epoch": 0.40091, + "grad_norm": 0.9075034884344351, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 40091 + }, + { + "epoch": 0.40092, + "grad_norm": 0.8185968938930399, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 40092 + }, + { + "epoch": 0.40093, + "grad_norm": 1.015331955439344, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 40093 + }, + { + "epoch": 0.40094, + "grad_norm": 1.343790903268252, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 40094 + }, + { + "epoch": 0.40095, + "grad_norm": 0.7753477797523042, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 40095 + }, + { + "epoch": 0.40096, + "grad_norm": 0.9731752694896494, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 40096 + }, + { + "epoch": 0.40097, + "grad_norm": 1.2650935494842257, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 40097 + }, + { + "epoch": 0.40098, + "grad_norm": 0.9889231268017881, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 40098 + }, + { + "epoch": 0.40099, + "grad_norm": 1.1847158723995768, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 40099 + }, + { + "epoch": 0.401, + "grad_norm": 1.235878825332868, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 40100 + }, + { + "epoch": 0.40101, + "grad_norm": 0.991336859714067, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 40101 + }, + { + "epoch": 0.40102, + "grad_norm": 1.0041672127398393, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 40102 + }, + { + "epoch": 0.40103, + "grad_norm": 1.135606314348352, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 40103 + }, + { + "epoch": 0.40104, + "grad_norm": 1.0093449438471171, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 40104 + }, + { + "epoch": 0.40105, + "grad_norm": 1.1669102991970837, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 40105 + }, + { + "epoch": 0.40106, + "grad_norm": 0.8707857686352813, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 40106 + }, + { + "epoch": 0.40107, + "grad_norm": 1.0271698179202142, + "learning_rate": 0.003, + "loss": 4.063, + "step": 40107 + }, + { + "epoch": 0.40108, + "grad_norm": 1.537050838747002, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 40108 + }, + { + "epoch": 0.40109, + "grad_norm": 0.8626147626365306, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 40109 + }, + { + "epoch": 0.4011, + "grad_norm": 1.0842017951903793, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 40110 + }, + { + "epoch": 0.40111, + "grad_norm": 1.2495937856156447, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 40111 + }, + { + "epoch": 0.40112, + "grad_norm": 0.8944653839157941, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 40112 + }, + { + "epoch": 0.40113, + "grad_norm": 1.1410415652222996, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 40113 + }, + { + "epoch": 0.40114, + "grad_norm": 1.1716345229283323, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 40114 + }, + { + "epoch": 0.40115, + "grad_norm": 0.8941437765894917, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 40115 + }, + { + "epoch": 0.40116, + "grad_norm": 1.0551755148341433, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 40116 + }, + { + "epoch": 0.40117, + "grad_norm": 1.321937512668069, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 40117 + }, + { + "epoch": 0.40118, + "grad_norm": 0.9368895447559609, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 40118 + }, + { + "epoch": 0.40119, + "grad_norm": 1.0453306263064683, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 40119 + }, + { + "epoch": 0.4012, + "grad_norm": 1.3878316547155718, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 40120 + }, + { + "epoch": 0.40121, + "grad_norm": 0.968997209791894, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 40121 + }, + { + "epoch": 0.40122, + "grad_norm": 0.9350705158467226, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 40122 + }, + { + "epoch": 0.40123, + "grad_norm": 1.0229235784014634, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 40123 + }, + { + "epoch": 0.40124, + "grad_norm": 1.2788103815473926, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 40124 + }, + { + "epoch": 0.40125, + "grad_norm": 0.9649283219400387, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 40125 + }, + { + "epoch": 0.40126, + "grad_norm": 0.8872126554482302, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 40126 + }, + { + "epoch": 0.40127, + "grad_norm": 0.9770406460815962, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 40127 + }, + { + "epoch": 0.40128, + "grad_norm": 1.1444589093291646, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 40128 + }, + { + "epoch": 0.40129, + "grad_norm": 1.1139150318020983, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 40129 + }, + { + "epoch": 0.4013, + "grad_norm": 1.0606636648779204, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 40130 + }, + { + "epoch": 0.40131, + "grad_norm": 1.035588566680443, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 40131 + }, + { + "epoch": 0.40132, + "grad_norm": 0.9226357603653633, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 40132 + }, + { + "epoch": 0.40133, + "grad_norm": 1.0406631849674617, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 40133 + }, + { + "epoch": 0.40134, + "grad_norm": 1.1355823591437082, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 40134 + }, + { + "epoch": 0.40135, + "grad_norm": 0.9312869401938688, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 40135 + }, + { + "epoch": 0.40136, + "grad_norm": 1.223665725079107, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 40136 + }, + { + "epoch": 0.40137, + "grad_norm": 0.9317160286325831, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 40137 + }, + { + "epoch": 0.40138, + "grad_norm": 0.9113062390344799, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 40138 + }, + { + "epoch": 0.40139, + "grad_norm": 1.0990622546335949, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 40139 + }, + { + "epoch": 0.4014, + "grad_norm": 1.0921702526422399, + "learning_rate": 0.003, + "loss": 4.0845, + "step": 40140 + }, + { + "epoch": 0.40141, + "grad_norm": 1.0449258537303325, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 40141 + }, + { + "epoch": 0.40142, + "grad_norm": 1.0180073969770231, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 40142 + }, + { + "epoch": 0.40143, + "grad_norm": 0.9982194008472087, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 40143 + }, + { + "epoch": 0.40144, + "grad_norm": 1.2078787432361913, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 40144 + }, + { + "epoch": 0.40145, + "grad_norm": 1.0361879465233794, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 40145 + }, + { + "epoch": 0.40146, + "grad_norm": 1.143132558689495, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 40146 + }, + { + "epoch": 0.40147, + "grad_norm": 1.2377973161504363, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 40147 + }, + { + "epoch": 0.40148, + "grad_norm": 1.0454630196263364, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 40148 + }, + { + "epoch": 0.40149, + "grad_norm": 1.392769198068972, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 40149 + }, + { + "epoch": 0.4015, + "grad_norm": 0.767232329677476, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 40150 + }, + { + "epoch": 0.40151, + "grad_norm": 0.7846443289115904, + "learning_rate": 0.003, + "loss": 4.079, + "step": 40151 + }, + { + "epoch": 0.40152, + "grad_norm": 0.9443416856098131, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 40152 + }, + { + "epoch": 0.40153, + "grad_norm": 1.0669612940986513, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 40153 + }, + { + "epoch": 0.40154, + "grad_norm": 0.9135376554296457, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 40154 + }, + { + "epoch": 0.40155, + "grad_norm": 0.9493166937943828, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 40155 + }, + { + "epoch": 0.40156, + "grad_norm": 0.9961974853526689, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 40156 + }, + { + "epoch": 0.40157, + "grad_norm": 1.1325774892394935, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 40157 + }, + { + "epoch": 0.40158, + "grad_norm": 1.0929027833288303, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 40158 + }, + { + "epoch": 0.40159, + "grad_norm": 1.195713795816155, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 40159 + }, + { + "epoch": 0.4016, + "grad_norm": 0.9665231083027076, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 40160 + }, + { + "epoch": 0.40161, + "grad_norm": 1.0935374105341462, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 40161 + }, + { + "epoch": 0.40162, + "grad_norm": 1.1042629026850697, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 40162 + }, + { + "epoch": 0.40163, + "grad_norm": 1.2737988551096098, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 40163 + }, + { + "epoch": 0.40164, + "grad_norm": 0.9743994093384529, + "learning_rate": 0.003, + "loss": 4.05, + "step": 40164 + }, + { + "epoch": 0.40165, + "grad_norm": 1.1519061362646617, + "learning_rate": 0.003, + "loss": 4.046, + "step": 40165 + }, + { + "epoch": 0.40166, + "grad_norm": 0.9807061066248993, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 40166 + }, + { + "epoch": 0.40167, + "grad_norm": 1.1968353056739323, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 40167 + }, + { + "epoch": 0.40168, + "grad_norm": 0.9209409531508934, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 40168 + }, + { + "epoch": 0.40169, + "grad_norm": 0.9865684782690186, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 40169 + }, + { + "epoch": 0.4017, + "grad_norm": 1.1329847385322378, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 40170 + }, + { + "epoch": 0.40171, + "grad_norm": 0.9335023875812838, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 40171 + }, + { + "epoch": 0.40172, + "grad_norm": 1.0554507159199995, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 40172 + }, + { + "epoch": 0.40173, + "grad_norm": 0.9033632794658877, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 40173 + }, + { + "epoch": 0.40174, + "grad_norm": 0.8771355628397988, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 40174 + }, + { + "epoch": 0.40175, + "grad_norm": 0.8794956228794645, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 40175 + }, + { + "epoch": 0.40176, + "grad_norm": 1.3174185075134799, + "learning_rate": 0.003, + "loss": 4.0903, + "step": 40176 + }, + { + "epoch": 0.40177, + "grad_norm": 1.0963246013034236, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 40177 + }, + { + "epoch": 0.40178, + "grad_norm": 1.090841458577172, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 40178 + }, + { + "epoch": 0.40179, + "grad_norm": 1.1804156953480052, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 40179 + }, + { + "epoch": 0.4018, + "grad_norm": 0.9093923217415072, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 40180 + }, + { + "epoch": 0.40181, + "grad_norm": 0.9634736383805959, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 40181 + }, + { + "epoch": 0.40182, + "grad_norm": 1.293799893291249, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 40182 + }, + { + "epoch": 0.40183, + "grad_norm": 0.9602595509535036, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 40183 + }, + { + "epoch": 0.40184, + "grad_norm": 1.063070437343679, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 40184 + }, + { + "epoch": 0.40185, + "grad_norm": 0.9668477682938633, + "learning_rate": 0.003, + "loss": 4.042, + "step": 40185 + }, + { + "epoch": 0.40186, + "grad_norm": 1.1637319692319108, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 40186 + }, + { + "epoch": 0.40187, + "grad_norm": 0.8661178009582424, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 40187 + }, + { + "epoch": 0.40188, + "grad_norm": 1.066777485917895, + "learning_rate": 0.003, + "loss": 4.038, + "step": 40188 + }, + { + "epoch": 0.40189, + "grad_norm": 1.2141615324512856, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 40189 + }, + { + "epoch": 0.4019, + "grad_norm": 0.8906484870236018, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 40190 + }, + { + "epoch": 0.40191, + "grad_norm": 1.1574630547463138, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 40191 + }, + { + "epoch": 0.40192, + "grad_norm": 1.2705053553144114, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 40192 + }, + { + "epoch": 0.40193, + "grad_norm": 1.1202460860718348, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 40193 + }, + { + "epoch": 0.40194, + "grad_norm": 0.8472887218840452, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 40194 + }, + { + "epoch": 0.40195, + "grad_norm": 0.8001711779029237, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 40195 + }, + { + "epoch": 0.40196, + "grad_norm": 1.0380374835225554, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 40196 + }, + { + "epoch": 0.40197, + "grad_norm": 1.16783806319071, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 40197 + }, + { + "epoch": 0.40198, + "grad_norm": 0.7820934100157345, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 40198 + }, + { + "epoch": 0.40199, + "grad_norm": 0.9092919562433867, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 40199 + }, + { + "epoch": 0.402, + "grad_norm": 1.127702124830796, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 40200 + }, + { + "epoch": 0.40201, + "grad_norm": 1.2896813178603825, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 40201 + }, + { + "epoch": 0.40202, + "grad_norm": 1.0873575511850435, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 40202 + }, + { + "epoch": 0.40203, + "grad_norm": 1.182213381355236, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 40203 + }, + { + "epoch": 0.40204, + "grad_norm": 1.1111265539726025, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 40204 + }, + { + "epoch": 0.40205, + "grad_norm": 1.0612307387721973, + "learning_rate": 0.003, + "loss": 4.058, + "step": 40205 + }, + { + "epoch": 0.40206, + "grad_norm": 1.2724067789484603, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 40206 + }, + { + "epoch": 0.40207, + "grad_norm": 1.0967335724995053, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 40207 + }, + { + "epoch": 0.40208, + "grad_norm": 1.1497207714327173, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 40208 + }, + { + "epoch": 0.40209, + "grad_norm": 0.9936264564626829, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 40209 + }, + { + "epoch": 0.4021, + "grad_norm": 1.1731816748278279, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 40210 + }, + { + "epoch": 0.40211, + "grad_norm": 1.0823649078745838, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 40211 + }, + { + "epoch": 0.40212, + "grad_norm": 1.106721456127448, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 40212 + }, + { + "epoch": 0.40213, + "grad_norm": 1.1245099083515793, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 40213 + }, + { + "epoch": 0.40214, + "grad_norm": 1.2012462806987874, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 40214 + }, + { + "epoch": 0.40215, + "grad_norm": 1.0432060943830006, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 40215 + }, + { + "epoch": 0.40216, + "grad_norm": 1.1886034076272163, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 40216 + }, + { + "epoch": 0.40217, + "grad_norm": 0.9155000497312048, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 40217 + }, + { + "epoch": 0.40218, + "grad_norm": 0.9391032716240438, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 40218 + }, + { + "epoch": 0.40219, + "grad_norm": 1.0779828717151771, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 40219 + }, + { + "epoch": 0.4022, + "grad_norm": 1.2588549133762519, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 40220 + }, + { + "epoch": 0.40221, + "grad_norm": 0.910050512331598, + "learning_rate": 0.003, + "loss": 4.029, + "step": 40221 + }, + { + "epoch": 0.40222, + "grad_norm": 0.8709481391221964, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 40222 + }, + { + "epoch": 0.40223, + "grad_norm": 0.9074108176683577, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 40223 + }, + { + "epoch": 0.40224, + "grad_norm": 1.1618973226961877, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 40224 + }, + { + "epoch": 0.40225, + "grad_norm": 0.96883212060405, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 40225 + }, + { + "epoch": 0.40226, + "grad_norm": 1.1663999651911257, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 40226 + }, + { + "epoch": 0.40227, + "grad_norm": 1.217188861003572, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 40227 + }, + { + "epoch": 0.40228, + "grad_norm": 1.0186800983129076, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 40228 + }, + { + "epoch": 0.40229, + "grad_norm": 1.236550835832247, + "learning_rate": 0.003, + "loss": 4.011, + "step": 40229 + }, + { + "epoch": 0.4023, + "grad_norm": 1.0259451939997184, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 40230 + }, + { + "epoch": 0.40231, + "grad_norm": 1.0327810116043976, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 40231 + }, + { + "epoch": 0.40232, + "grad_norm": 1.1909309233079768, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 40232 + }, + { + "epoch": 0.40233, + "grad_norm": 0.9637483662722476, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 40233 + }, + { + "epoch": 0.40234, + "grad_norm": 0.8863273662370054, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 40234 + }, + { + "epoch": 0.40235, + "grad_norm": 0.9738831297782025, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 40235 + }, + { + "epoch": 0.40236, + "grad_norm": 1.0296411592237333, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 40236 + }, + { + "epoch": 0.40237, + "grad_norm": 1.0394412223844363, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 40237 + }, + { + "epoch": 0.40238, + "grad_norm": 1.1337744898703659, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 40238 + }, + { + "epoch": 0.40239, + "grad_norm": 1.2716537806971964, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 40239 + }, + { + "epoch": 0.4024, + "grad_norm": 0.9298074659789278, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 40240 + }, + { + "epoch": 0.40241, + "grad_norm": 1.0471679672993823, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 40241 + }, + { + "epoch": 0.40242, + "grad_norm": 1.1416723969670601, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 40242 + }, + { + "epoch": 0.40243, + "grad_norm": 1.0974479891735, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 40243 + }, + { + "epoch": 0.40244, + "grad_norm": 1.4388098501673625, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 40244 + }, + { + "epoch": 0.40245, + "grad_norm": 1.337640391383474, + "learning_rate": 0.003, + "loss": 4.047, + "step": 40245 + }, + { + "epoch": 0.40246, + "grad_norm": 0.8596907186636022, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 40246 + }, + { + "epoch": 0.40247, + "grad_norm": 0.8461998616602096, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 40247 + }, + { + "epoch": 0.40248, + "grad_norm": 1.0143599971673865, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 40248 + }, + { + "epoch": 0.40249, + "grad_norm": 1.1513760279203247, + "learning_rate": 0.003, + "loss": 4.046, + "step": 40249 + }, + { + "epoch": 0.4025, + "grad_norm": 1.0740947629738196, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 40250 + }, + { + "epoch": 0.40251, + "grad_norm": 1.5030841294931578, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 40251 + }, + { + "epoch": 0.40252, + "grad_norm": 0.8379708687325271, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 40252 + }, + { + "epoch": 0.40253, + "grad_norm": 0.7205026168958343, + "learning_rate": 0.003, + "loss": 4.035, + "step": 40253 + }, + { + "epoch": 0.40254, + "grad_norm": 0.7769585751481212, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 40254 + }, + { + "epoch": 0.40255, + "grad_norm": 0.8871390302782819, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 40255 + }, + { + "epoch": 0.40256, + "grad_norm": 1.0679813273439078, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 40256 + }, + { + "epoch": 0.40257, + "grad_norm": 1.1353363004730719, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 40257 + }, + { + "epoch": 0.40258, + "grad_norm": 1.2192718548561967, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 40258 + }, + { + "epoch": 0.40259, + "grad_norm": 1.1240634736004929, + "learning_rate": 0.003, + "loss": 4.056, + "step": 40259 + }, + { + "epoch": 0.4026, + "grad_norm": 1.0897343794760048, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 40260 + }, + { + "epoch": 0.40261, + "grad_norm": 1.1448186415492012, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 40261 + }, + { + "epoch": 0.40262, + "grad_norm": 1.0527934235325827, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 40262 + }, + { + "epoch": 0.40263, + "grad_norm": 1.11719415834504, + "learning_rate": 0.003, + "loss": 4.07, + "step": 40263 + }, + { + "epoch": 0.40264, + "grad_norm": 1.0772592954526319, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 40264 + }, + { + "epoch": 0.40265, + "grad_norm": 0.8542459410703974, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 40265 + }, + { + "epoch": 0.40266, + "grad_norm": 0.814802389649749, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 40266 + }, + { + "epoch": 0.40267, + "grad_norm": 0.9975006372346678, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 40267 + }, + { + "epoch": 0.40268, + "grad_norm": 1.313364610070864, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 40268 + }, + { + "epoch": 0.40269, + "grad_norm": 0.8536429920069115, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 40269 + }, + { + "epoch": 0.4027, + "grad_norm": 1.0562790404604139, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 40270 + }, + { + "epoch": 0.40271, + "grad_norm": 1.2175346356285015, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 40271 + }, + { + "epoch": 0.40272, + "grad_norm": 0.880503050676389, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 40272 + }, + { + "epoch": 0.40273, + "grad_norm": 0.9493729691587881, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 40273 + }, + { + "epoch": 0.40274, + "grad_norm": 1.2397328692898377, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 40274 + }, + { + "epoch": 0.40275, + "grad_norm": 1.037996852825064, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 40275 + }, + { + "epoch": 0.40276, + "grad_norm": 0.9932464027964191, + "learning_rate": 0.003, + "loss": 4.041, + "step": 40276 + }, + { + "epoch": 0.40277, + "grad_norm": 1.2035715161696972, + "learning_rate": 0.003, + "loss": 3.997, + "step": 40277 + }, + { + "epoch": 0.40278, + "grad_norm": 1.0412489885101008, + "learning_rate": 0.003, + "loss": 4.041, + "step": 40278 + }, + { + "epoch": 0.40279, + "grad_norm": 1.2839127987746466, + "learning_rate": 0.003, + "loss": 4.05, + "step": 40279 + }, + { + "epoch": 0.4028, + "grad_norm": 1.0189515308314063, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 40280 + }, + { + "epoch": 0.40281, + "grad_norm": 1.1431749069932426, + "learning_rate": 0.003, + "loss": 4.042, + "step": 40281 + }, + { + "epoch": 0.40282, + "grad_norm": 1.090002558050043, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 40282 + }, + { + "epoch": 0.40283, + "grad_norm": 1.0709552919811265, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 40283 + }, + { + "epoch": 0.40284, + "grad_norm": 1.2966750483890572, + "learning_rate": 0.003, + "loss": 4.017, + "step": 40284 + }, + { + "epoch": 0.40285, + "grad_norm": 1.1045618805028137, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 40285 + }, + { + "epoch": 0.40286, + "grad_norm": 1.2338293761157222, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 40286 + }, + { + "epoch": 0.40287, + "grad_norm": 0.986796978115378, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 40287 + }, + { + "epoch": 0.40288, + "grad_norm": 1.2095447643540669, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 40288 + }, + { + "epoch": 0.40289, + "grad_norm": 1.0561622153303067, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 40289 + }, + { + "epoch": 0.4029, + "grad_norm": 1.131583516433851, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 40290 + }, + { + "epoch": 0.40291, + "grad_norm": 1.0738296713451783, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 40291 + }, + { + "epoch": 0.40292, + "grad_norm": 1.1062882357746955, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 40292 + }, + { + "epoch": 0.40293, + "grad_norm": 1.1144182152111808, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 40293 + }, + { + "epoch": 0.40294, + "grad_norm": 1.3390144533350357, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 40294 + }, + { + "epoch": 0.40295, + "grad_norm": 1.0386288412770919, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 40295 + }, + { + "epoch": 0.40296, + "grad_norm": 1.2502447515556703, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 40296 + }, + { + "epoch": 0.40297, + "grad_norm": 0.9928184567169397, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 40297 + }, + { + "epoch": 0.40298, + "grad_norm": 1.1696416761709914, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 40298 + }, + { + "epoch": 0.40299, + "grad_norm": 1.0496795177753933, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 40299 + }, + { + "epoch": 0.403, + "grad_norm": 1.1417358637954582, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 40300 + }, + { + "epoch": 0.40301, + "grad_norm": 1.0196350621282133, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 40301 + }, + { + "epoch": 0.40302, + "grad_norm": 1.184877455668576, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 40302 + }, + { + "epoch": 0.40303, + "grad_norm": 1.0580953911607196, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 40303 + }, + { + "epoch": 0.40304, + "grad_norm": 1.0202117758887577, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 40304 + }, + { + "epoch": 0.40305, + "grad_norm": 1.1504544473648333, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 40305 + }, + { + "epoch": 0.40306, + "grad_norm": 0.9376334587408103, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 40306 + }, + { + "epoch": 0.40307, + "grad_norm": 0.993365517751816, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 40307 + }, + { + "epoch": 0.40308, + "grad_norm": 1.299364589617953, + "learning_rate": 0.003, + "loss": 4.046, + "step": 40308 + }, + { + "epoch": 0.40309, + "grad_norm": 1.0906226560363048, + "learning_rate": 0.003, + "loss": 4.017, + "step": 40309 + }, + { + "epoch": 0.4031, + "grad_norm": 0.917190309291686, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 40310 + }, + { + "epoch": 0.40311, + "grad_norm": 0.9426244461225897, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 40311 + }, + { + "epoch": 0.40312, + "grad_norm": 1.1047655568403048, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 40312 + }, + { + "epoch": 0.40313, + "grad_norm": 1.1099276054957992, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 40313 + }, + { + "epoch": 0.40314, + "grad_norm": 1.1900582834125104, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 40314 + }, + { + "epoch": 0.40315, + "grad_norm": 1.0832717849229734, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 40315 + }, + { + "epoch": 0.40316, + "grad_norm": 1.2000341784850939, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 40316 + }, + { + "epoch": 0.40317, + "grad_norm": 1.0407041552307845, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 40317 + }, + { + "epoch": 0.40318, + "grad_norm": 1.1242983503357282, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 40318 + }, + { + "epoch": 0.40319, + "grad_norm": 0.9914987982685441, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 40319 + }, + { + "epoch": 0.4032, + "grad_norm": 1.0624563676342729, + "learning_rate": 0.003, + "loss": 4.016, + "step": 40320 + }, + { + "epoch": 0.40321, + "grad_norm": 1.3670323966985445, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 40321 + }, + { + "epoch": 0.40322, + "grad_norm": 0.7283780678638837, + "learning_rate": 0.003, + "loss": 4.004, + "step": 40322 + }, + { + "epoch": 0.40323, + "grad_norm": 0.6759154549704198, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 40323 + }, + { + "epoch": 0.40324, + "grad_norm": 0.8138677310158022, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 40324 + }, + { + "epoch": 0.40325, + "grad_norm": 0.9106236890322771, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 40325 + }, + { + "epoch": 0.40326, + "grad_norm": 0.9685255383919852, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 40326 + }, + { + "epoch": 0.40327, + "grad_norm": 1.0340515224538143, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 40327 + }, + { + "epoch": 0.40328, + "grad_norm": 1.33770254712846, + "learning_rate": 0.003, + "loss": 4.069, + "step": 40328 + }, + { + "epoch": 0.40329, + "grad_norm": 0.8790843763557192, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 40329 + }, + { + "epoch": 0.4033, + "grad_norm": 0.8495031619772185, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 40330 + }, + { + "epoch": 0.40331, + "grad_norm": 0.9801121841960064, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 40331 + }, + { + "epoch": 0.40332, + "grad_norm": 1.4252884859020027, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 40332 + }, + { + "epoch": 0.40333, + "grad_norm": 0.9459677152388385, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 40333 + }, + { + "epoch": 0.40334, + "grad_norm": 1.0792412901660264, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 40334 + }, + { + "epoch": 0.40335, + "grad_norm": 1.1995204835046198, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 40335 + }, + { + "epoch": 0.40336, + "grad_norm": 1.0730399367588956, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 40336 + }, + { + "epoch": 0.40337, + "grad_norm": 1.0526927237020027, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 40337 + }, + { + "epoch": 0.40338, + "grad_norm": 1.0750865405845096, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 40338 + }, + { + "epoch": 0.40339, + "grad_norm": 1.1720624749128088, + "learning_rate": 0.003, + "loss": 4.032, + "step": 40339 + }, + { + "epoch": 0.4034, + "grad_norm": 1.016057734199857, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 40340 + }, + { + "epoch": 0.40341, + "grad_norm": 1.180830669461844, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 40341 + }, + { + "epoch": 0.40342, + "grad_norm": 0.9993601622687385, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 40342 + }, + { + "epoch": 0.40343, + "grad_norm": 1.117604710436444, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 40343 + }, + { + "epoch": 0.40344, + "grad_norm": 1.169654222212871, + "learning_rate": 0.003, + "loss": 4.033, + "step": 40344 + }, + { + "epoch": 0.40345, + "grad_norm": 1.204389100143462, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 40345 + }, + { + "epoch": 0.40346, + "grad_norm": 1.1290668202872403, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 40346 + }, + { + "epoch": 0.40347, + "grad_norm": 0.9993963482254756, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 40347 + }, + { + "epoch": 0.40348, + "grad_norm": 1.3693655261172741, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 40348 + }, + { + "epoch": 0.40349, + "grad_norm": 0.94043072945144, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 40349 + }, + { + "epoch": 0.4035, + "grad_norm": 0.9069868250715004, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 40350 + }, + { + "epoch": 0.40351, + "grad_norm": 1.0391320960768702, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 40351 + }, + { + "epoch": 0.40352, + "grad_norm": 1.4374307143334826, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 40352 + }, + { + "epoch": 0.40353, + "grad_norm": 0.8501750460258755, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 40353 + }, + { + "epoch": 0.40354, + "grad_norm": 0.9991779163597023, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 40354 + }, + { + "epoch": 0.40355, + "grad_norm": 1.123026556828248, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 40355 + }, + { + "epoch": 0.40356, + "grad_norm": 1.0732121127534537, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 40356 + }, + { + "epoch": 0.40357, + "grad_norm": 1.1987551473772844, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 40357 + }, + { + "epoch": 0.40358, + "grad_norm": 0.7504650303013845, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 40358 + }, + { + "epoch": 0.40359, + "grad_norm": 0.7239774207731158, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 40359 + }, + { + "epoch": 0.4036, + "grad_norm": 0.9029929494476542, + "learning_rate": 0.003, + "loss": 4.069, + "step": 40360 + }, + { + "epoch": 0.40361, + "grad_norm": 1.079131380870964, + "learning_rate": 0.003, + "loss": 4.069, + "step": 40361 + }, + { + "epoch": 0.40362, + "grad_norm": 0.9527302992273317, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 40362 + }, + { + "epoch": 0.40363, + "grad_norm": 1.2252189065111148, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 40363 + }, + { + "epoch": 0.40364, + "grad_norm": 1.1723232830918473, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 40364 + }, + { + "epoch": 0.40365, + "grad_norm": 0.876497874627625, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 40365 + }, + { + "epoch": 0.40366, + "grad_norm": 0.9302623966088207, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 40366 + }, + { + "epoch": 0.40367, + "grad_norm": 1.1955444719843338, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 40367 + }, + { + "epoch": 0.40368, + "grad_norm": 1.1288260936853067, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 40368 + }, + { + "epoch": 0.40369, + "grad_norm": 1.1208398302500808, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 40369 + }, + { + "epoch": 0.4037, + "grad_norm": 1.0560662546397228, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 40370 + }, + { + "epoch": 0.40371, + "grad_norm": 1.2139686989729461, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 40371 + }, + { + "epoch": 0.40372, + "grad_norm": 0.9481583751966158, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 40372 + }, + { + "epoch": 0.40373, + "grad_norm": 1.2194386253294045, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 40373 + }, + { + "epoch": 0.40374, + "grad_norm": 1.0095731327164794, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 40374 + }, + { + "epoch": 0.40375, + "grad_norm": 1.4165968814393302, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 40375 + }, + { + "epoch": 0.40376, + "grad_norm": 0.8723227502624011, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 40376 + }, + { + "epoch": 0.40377, + "grad_norm": 0.9041011194080665, + "learning_rate": 0.003, + "loss": 4.054, + "step": 40377 + }, + { + "epoch": 0.40378, + "grad_norm": 1.0105145247900473, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 40378 + }, + { + "epoch": 0.40379, + "grad_norm": 1.0704517581668913, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 40379 + }, + { + "epoch": 0.4038, + "grad_norm": 1.0358268151414265, + "learning_rate": 0.003, + "loss": 4.025, + "step": 40380 + }, + { + "epoch": 0.40381, + "grad_norm": 0.979825630689927, + "learning_rate": 0.003, + "loss": 4.044, + "step": 40381 + }, + { + "epoch": 0.40382, + "grad_norm": 1.1275653680481632, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 40382 + }, + { + "epoch": 0.40383, + "grad_norm": 1.0602975669022514, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 40383 + }, + { + "epoch": 0.40384, + "grad_norm": 1.0901869297516789, + "learning_rate": 0.003, + "loss": 4.031, + "step": 40384 + }, + { + "epoch": 0.40385, + "grad_norm": 0.8016474048772823, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 40385 + }, + { + "epoch": 0.40386, + "grad_norm": 0.8167890459941153, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 40386 + }, + { + "epoch": 0.40387, + "grad_norm": 1.0349543071844898, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 40387 + }, + { + "epoch": 0.40388, + "grad_norm": 1.5961256119945628, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 40388 + }, + { + "epoch": 0.40389, + "grad_norm": 0.865256573734014, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 40389 + }, + { + "epoch": 0.4039, + "grad_norm": 1.0201606494170747, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 40390 + }, + { + "epoch": 0.40391, + "grad_norm": 1.3750494876883055, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 40391 + }, + { + "epoch": 0.40392, + "grad_norm": 1.3246312458980867, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 40392 + }, + { + "epoch": 0.40393, + "grad_norm": 1.1683874225607556, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 40393 + }, + { + "epoch": 0.40394, + "grad_norm": 1.2153480742746947, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 40394 + }, + { + "epoch": 0.40395, + "grad_norm": 0.9778606652636275, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 40395 + }, + { + "epoch": 0.40396, + "grad_norm": 1.0792764757505084, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 40396 + }, + { + "epoch": 0.40397, + "grad_norm": 1.2417472821577975, + "learning_rate": 0.003, + "loss": 4.042, + "step": 40397 + }, + { + "epoch": 0.40398, + "grad_norm": 0.836967308691416, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 40398 + }, + { + "epoch": 0.40399, + "grad_norm": 0.7490086765086593, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 40399 + }, + { + "epoch": 0.404, + "grad_norm": 0.8241523987721062, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 40400 + }, + { + "epoch": 0.40401, + "grad_norm": 0.8748210071928908, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 40401 + }, + { + "epoch": 0.40402, + "grad_norm": 0.883477463398195, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 40402 + }, + { + "epoch": 0.40403, + "grad_norm": 0.9311397391562283, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 40403 + }, + { + "epoch": 0.40404, + "grad_norm": 1.068063616556707, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 40404 + }, + { + "epoch": 0.40405, + "grad_norm": 1.163049425709725, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 40405 + }, + { + "epoch": 0.40406, + "grad_norm": 1.2887757918084932, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 40406 + }, + { + "epoch": 0.40407, + "grad_norm": 1.086896112609094, + "learning_rate": 0.003, + "loss": 4.035, + "step": 40407 + }, + { + "epoch": 0.40408, + "grad_norm": 1.1931842715403953, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 40408 + }, + { + "epoch": 0.40409, + "grad_norm": 1.1240137728743904, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 40409 + }, + { + "epoch": 0.4041, + "grad_norm": 1.2834495342224317, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 40410 + }, + { + "epoch": 0.40411, + "grad_norm": 1.1962882581743486, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 40411 + }, + { + "epoch": 0.40412, + "grad_norm": 1.0791097108713246, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 40412 + }, + { + "epoch": 0.40413, + "grad_norm": 0.8819054807915268, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 40413 + }, + { + "epoch": 0.40414, + "grad_norm": 0.7161770198810027, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 40414 + }, + { + "epoch": 0.40415, + "grad_norm": 0.8465897922000032, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 40415 + }, + { + "epoch": 0.40416, + "grad_norm": 1.2009295378136513, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 40416 + }, + { + "epoch": 0.40417, + "grad_norm": 1.2288764507721626, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 40417 + }, + { + "epoch": 0.40418, + "grad_norm": 1.0577676347567193, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 40418 + }, + { + "epoch": 0.40419, + "grad_norm": 0.9333285762148948, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 40419 + }, + { + "epoch": 0.4042, + "grad_norm": 0.7508655995350295, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 40420 + }, + { + "epoch": 0.40421, + "grad_norm": 0.728085657340058, + "learning_rate": 0.003, + "loss": 4.021, + "step": 40421 + }, + { + "epoch": 0.40422, + "grad_norm": 0.8167170760484276, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 40422 + }, + { + "epoch": 0.40423, + "grad_norm": 0.9700091292205753, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 40423 + }, + { + "epoch": 0.40424, + "grad_norm": 1.2503108743776885, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 40424 + }, + { + "epoch": 0.40425, + "grad_norm": 1.0200507061389084, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 40425 + }, + { + "epoch": 0.40426, + "grad_norm": 1.1746977640321667, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 40426 + }, + { + "epoch": 0.40427, + "grad_norm": 1.0330808150775415, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 40427 + }, + { + "epoch": 0.40428, + "grad_norm": 1.2615164448883491, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 40428 + }, + { + "epoch": 0.40429, + "grad_norm": 1.00032615174138, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 40429 + }, + { + "epoch": 0.4043, + "grad_norm": 1.0257859018702873, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 40430 + }, + { + "epoch": 0.40431, + "grad_norm": 1.1914999663056873, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 40431 + }, + { + "epoch": 0.40432, + "grad_norm": 1.0497525458608425, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 40432 + }, + { + "epoch": 0.40433, + "grad_norm": 1.209390762314148, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 40433 + }, + { + "epoch": 0.40434, + "grad_norm": 1.3418714441204005, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 40434 + }, + { + "epoch": 0.40435, + "grad_norm": 1.1925070498329835, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 40435 + }, + { + "epoch": 0.40436, + "grad_norm": 1.1023799931095004, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 40436 + }, + { + "epoch": 0.40437, + "grad_norm": 1.431601619880322, + "learning_rate": 0.003, + "loss": 3.992, + "step": 40437 + }, + { + "epoch": 0.40438, + "grad_norm": 0.7627466599473016, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 40438 + }, + { + "epoch": 0.40439, + "grad_norm": 0.7371185344600684, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 40439 + }, + { + "epoch": 0.4044, + "grad_norm": 0.9362040825696113, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 40440 + }, + { + "epoch": 0.40441, + "grad_norm": 1.4371133324313996, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 40441 + }, + { + "epoch": 0.40442, + "grad_norm": 0.7465609304676875, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 40442 + }, + { + "epoch": 0.40443, + "grad_norm": 0.8109794857773741, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 40443 + }, + { + "epoch": 0.40444, + "grad_norm": 1.1156646353908624, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 40444 + }, + { + "epoch": 0.40445, + "grad_norm": 1.4459452243986843, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 40445 + }, + { + "epoch": 0.40446, + "grad_norm": 0.6698888942953923, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 40446 + }, + { + "epoch": 0.40447, + "grad_norm": 0.7354111646408126, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 40447 + }, + { + "epoch": 0.40448, + "grad_norm": 0.9803187061179277, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 40448 + }, + { + "epoch": 0.40449, + "grad_norm": 1.229720445226168, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 40449 + }, + { + "epoch": 0.4045, + "grad_norm": 1.1330151893222327, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 40450 + }, + { + "epoch": 0.40451, + "grad_norm": 1.0549416095601987, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 40451 + }, + { + "epoch": 0.40452, + "grad_norm": 0.9965907580311847, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 40452 + }, + { + "epoch": 0.40453, + "grad_norm": 0.9524392433966806, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 40453 + }, + { + "epoch": 0.40454, + "grad_norm": 0.9345556631940012, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 40454 + }, + { + "epoch": 0.40455, + "grad_norm": 1.04009334697629, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 40455 + }, + { + "epoch": 0.40456, + "grad_norm": 1.0864195090928572, + "learning_rate": 0.003, + "loss": 4.04, + "step": 40456 + }, + { + "epoch": 0.40457, + "grad_norm": 0.8877998023340606, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 40457 + }, + { + "epoch": 0.40458, + "grad_norm": 1.0368463254576397, + "learning_rate": 0.003, + "loss": 4.025, + "step": 40458 + }, + { + "epoch": 0.40459, + "grad_norm": 1.3417895166556455, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 40459 + }, + { + "epoch": 0.4046, + "grad_norm": 0.8780675107539043, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 40460 + }, + { + "epoch": 0.40461, + "grad_norm": 0.8127252282368129, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 40461 + }, + { + "epoch": 0.40462, + "grad_norm": 1.0716075331779233, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 40462 + }, + { + "epoch": 0.40463, + "grad_norm": 1.6446409188227433, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 40463 + }, + { + "epoch": 0.40464, + "grad_norm": 0.7512429083676372, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 40464 + }, + { + "epoch": 0.40465, + "grad_norm": 0.9187067131659575, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 40465 + }, + { + "epoch": 0.40466, + "grad_norm": 1.2749631074974732, + "learning_rate": 0.003, + "loss": 4.01, + "step": 40466 + }, + { + "epoch": 0.40467, + "grad_norm": 1.1567069956739209, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 40467 + }, + { + "epoch": 0.40468, + "grad_norm": 1.0557677539927435, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 40468 + }, + { + "epoch": 0.40469, + "grad_norm": 0.9595636386175677, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 40469 + }, + { + "epoch": 0.4047, + "grad_norm": 0.9421608274844294, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 40470 + }, + { + "epoch": 0.40471, + "grad_norm": 0.9956604821502114, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 40471 + }, + { + "epoch": 0.40472, + "grad_norm": 1.1044149923890376, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 40472 + }, + { + "epoch": 0.40473, + "grad_norm": 1.140457646663214, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 40473 + }, + { + "epoch": 0.40474, + "grad_norm": 0.9951306579925377, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 40474 + }, + { + "epoch": 0.40475, + "grad_norm": 1.245543045342693, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 40475 + }, + { + "epoch": 0.40476, + "grad_norm": 0.9011027831118225, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 40476 + }, + { + "epoch": 0.40477, + "grad_norm": 1.0332490143622797, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 40477 + }, + { + "epoch": 0.40478, + "grad_norm": 1.4977592461692357, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 40478 + }, + { + "epoch": 0.40479, + "grad_norm": 0.9678279372205941, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 40479 + }, + { + "epoch": 0.4048, + "grad_norm": 1.2096713368434593, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 40480 + }, + { + "epoch": 0.40481, + "grad_norm": 1.2845674940226794, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 40481 + }, + { + "epoch": 0.40482, + "grad_norm": 1.0605048344275303, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 40482 + }, + { + "epoch": 0.40483, + "grad_norm": 1.0331788840188205, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 40483 + }, + { + "epoch": 0.40484, + "grad_norm": 1.2133464657144348, + "learning_rate": 0.003, + "loss": 4.055, + "step": 40484 + }, + { + "epoch": 0.40485, + "grad_norm": 1.1420762393927764, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 40485 + }, + { + "epoch": 0.40486, + "grad_norm": 1.2065968227956625, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 40486 + }, + { + "epoch": 0.40487, + "grad_norm": 0.9560394125617896, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 40487 + }, + { + "epoch": 0.40488, + "grad_norm": 1.0591528128577883, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 40488 + }, + { + "epoch": 0.40489, + "grad_norm": 1.24974116610272, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 40489 + }, + { + "epoch": 0.4049, + "grad_norm": 0.8520541424349533, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 40490 + }, + { + "epoch": 0.40491, + "grad_norm": 0.835690376251949, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 40491 + }, + { + "epoch": 0.40492, + "grad_norm": 1.0244715759103937, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 40492 + }, + { + "epoch": 0.40493, + "grad_norm": 1.2650044929919257, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 40493 + }, + { + "epoch": 0.40494, + "grad_norm": 0.9246099755390911, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 40494 + }, + { + "epoch": 0.40495, + "grad_norm": 1.0785166590656854, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 40495 + }, + { + "epoch": 0.40496, + "grad_norm": 1.0276635765346678, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 40496 + }, + { + "epoch": 0.40497, + "grad_norm": 1.1900415225793506, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 40497 + }, + { + "epoch": 0.40498, + "grad_norm": 1.0095583650645643, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 40498 + }, + { + "epoch": 0.40499, + "grad_norm": 1.1634134445658981, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 40499 + }, + { + "epoch": 0.405, + "grad_norm": 1.3133106109950659, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 40500 + }, + { + "epoch": 0.40501, + "grad_norm": 1.050638490077326, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 40501 + }, + { + "epoch": 0.40502, + "grad_norm": 1.1864059322528435, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 40502 + }, + { + "epoch": 0.40503, + "grad_norm": 0.9982575035957656, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 40503 + }, + { + "epoch": 0.40504, + "grad_norm": 1.2778568350760442, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 40504 + }, + { + "epoch": 0.40505, + "grad_norm": 2.6563065336373213, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 40505 + }, + { + "epoch": 0.40506, + "grad_norm": 1.6659433838281332, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 40506 + }, + { + "epoch": 0.40507, + "grad_norm": 0.9632459030363888, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 40507 + }, + { + "epoch": 0.40508, + "grad_norm": 2.2712288760661687, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 40508 + }, + { + "epoch": 0.40509, + "grad_norm": 0.9862474832061908, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 40509 + }, + { + "epoch": 0.4051, + "grad_norm": 1.4407439835386089, + "learning_rate": 0.003, + "loss": 4.096, + "step": 40510 + }, + { + "epoch": 0.40511, + "grad_norm": 1.037779890723844, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 40511 + }, + { + "epoch": 0.40512, + "grad_norm": 1.090927130154026, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 40512 + }, + { + "epoch": 0.40513, + "grad_norm": 1.163252028650588, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 40513 + }, + { + "epoch": 0.40514, + "grad_norm": 1.1062543887528478, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 40514 + }, + { + "epoch": 0.40515, + "grad_norm": 0.9873024218402113, + "learning_rate": 0.003, + "loss": 4.064, + "step": 40515 + }, + { + "epoch": 0.40516, + "grad_norm": 1.206578839940886, + "learning_rate": 0.003, + "loss": 4.1087, + "step": 40516 + }, + { + "epoch": 0.40517, + "grad_norm": 0.9659762383085762, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 40517 + }, + { + "epoch": 0.40518, + "grad_norm": 1.2045142002685902, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 40518 + }, + { + "epoch": 0.40519, + "grad_norm": 1.2179899238112923, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 40519 + }, + { + "epoch": 0.4052, + "grad_norm": 0.9463291535993084, + "learning_rate": 0.003, + "loss": 4.08, + "step": 40520 + }, + { + "epoch": 0.40521, + "grad_norm": 0.9773298230090055, + "learning_rate": 0.003, + "loss": 4.045, + "step": 40521 + }, + { + "epoch": 0.40522, + "grad_norm": 1.114150658375948, + "learning_rate": 0.003, + "loss": 4.079, + "step": 40522 + }, + { + "epoch": 0.40523, + "grad_norm": 0.9777225356394277, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 40523 + }, + { + "epoch": 0.40524, + "grad_norm": 1.251827309202729, + "learning_rate": 0.003, + "loss": 4.0855, + "step": 40524 + }, + { + "epoch": 0.40525, + "grad_norm": 0.943151441592245, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 40525 + }, + { + "epoch": 0.40526, + "grad_norm": 0.9813175849164043, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 40526 + }, + { + "epoch": 0.40527, + "grad_norm": 0.8907314586186352, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 40527 + }, + { + "epoch": 0.40528, + "grad_norm": 1.19318508449329, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 40528 + }, + { + "epoch": 0.40529, + "grad_norm": 1.4331765589728471, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 40529 + }, + { + "epoch": 0.4053, + "grad_norm": 0.8272122094164606, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 40530 + }, + { + "epoch": 0.40531, + "grad_norm": 0.8041891070025344, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 40531 + }, + { + "epoch": 0.40532, + "grad_norm": 0.9235397169224153, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 40532 + }, + { + "epoch": 0.40533, + "grad_norm": 1.1128942869355123, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 40533 + }, + { + "epoch": 0.40534, + "grad_norm": 1.2095660126692964, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 40534 + }, + { + "epoch": 0.40535, + "grad_norm": 1.0639627096459974, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 40535 + }, + { + "epoch": 0.40536, + "grad_norm": 1.183487435683211, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 40536 + }, + { + "epoch": 0.40537, + "grad_norm": 0.8778339500112322, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 40537 + }, + { + "epoch": 0.40538, + "grad_norm": 0.8322668654805738, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 40538 + }, + { + "epoch": 0.40539, + "grad_norm": 0.8511408354427451, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 40539 + }, + { + "epoch": 0.4054, + "grad_norm": 0.9640166630509459, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 40540 + }, + { + "epoch": 0.40541, + "grad_norm": 0.9706029950496049, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 40541 + }, + { + "epoch": 0.40542, + "grad_norm": 1.0160805803086888, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 40542 + }, + { + "epoch": 0.40543, + "grad_norm": 0.9342035680183199, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 40543 + }, + { + "epoch": 0.40544, + "grad_norm": 0.8142736657786295, + "learning_rate": 0.003, + "loss": 4.048, + "step": 40544 + }, + { + "epoch": 0.40545, + "grad_norm": 0.9138659859459901, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 40545 + }, + { + "epoch": 0.40546, + "grad_norm": 0.9953384558414368, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 40546 + }, + { + "epoch": 0.40547, + "grad_norm": 1.4056910656733128, + "learning_rate": 0.003, + "loss": 4.051, + "step": 40547 + }, + { + "epoch": 0.40548, + "grad_norm": 1.1303965860189165, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 40548 + }, + { + "epoch": 0.40549, + "grad_norm": 1.3037524233643658, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 40549 + }, + { + "epoch": 0.4055, + "grad_norm": 0.589519494717059, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 40550 + }, + { + "epoch": 0.40551, + "grad_norm": 0.6975890161904825, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 40551 + }, + { + "epoch": 0.40552, + "grad_norm": 0.8759078580297351, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 40552 + }, + { + "epoch": 0.40553, + "grad_norm": 1.5941299912290512, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 40553 + }, + { + "epoch": 0.40554, + "grad_norm": 0.72861577616181, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 40554 + }, + { + "epoch": 0.40555, + "grad_norm": 0.7585074534788525, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 40555 + }, + { + "epoch": 0.40556, + "grad_norm": 1.0297175853431517, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 40556 + }, + { + "epoch": 0.40557, + "grad_norm": 1.466692896468228, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 40557 + }, + { + "epoch": 0.40558, + "grad_norm": 0.9149140439200047, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 40558 + }, + { + "epoch": 0.40559, + "grad_norm": 1.0227947013093954, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 40559 + }, + { + "epoch": 0.4056, + "grad_norm": 1.2458012760527053, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 40560 + }, + { + "epoch": 0.40561, + "grad_norm": 1.0861775171167936, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 40561 + }, + { + "epoch": 0.40562, + "grad_norm": 1.146828272918262, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 40562 + }, + { + "epoch": 0.40563, + "grad_norm": 0.9582521650760281, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 40563 + }, + { + "epoch": 0.40564, + "grad_norm": 0.9029996505962937, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 40564 + }, + { + "epoch": 0.40565, + "grad_norm": 0.9674900688304645, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 40565 + }, + { + "epoch": 0.40566, + "grad_norm": 1.150113565409574, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 40566 + }, + { + "epoch": 0.40567, + "grad_norm": 1.0570262288717474, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 40567 + }, + { + "epoch": 0.40568, + "grad_norm": 1.1588727448797187, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 40568 + }, + { + "epoch": 0.40569, + "grad_norm": 1.2360723115389405, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 40569 + }, + { + "epoch": 0.4057, + "grad_norm": 1.0036765836193287, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 40570 + }, + { + "epoch": 0.40571, + "grad_norm": 0.8888184397819062, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 40571 + }, + { + "epoch": 0.40572, + "grad_norm": 0.8992962257851326, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 40572 + }, + { + "epoch": 0.40573, + "grad_norm": 1.0877561229994783, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 40573 + }, + { + "epoch": 0.40574, + "grad_norm": 1.1280237853855095, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 40574 + }, + { + "epoch": 0.40575, + "grad_norm": 0.8798058703887786, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 40575 + }, + { + "epoch": 0.40576, + "grad_norm": 0.8486380830883872, + "learning_rate": 0.003, + "loss": 4.027, + "step": 40576 + }, + { + "epoch": 0.40577, + "grad_norm": 1.0809113672115382, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 40577 + }, + { + "epoch": 0.40578, + "grad_norm": 1.1599849178847461, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 40578 + }, + { + "epoch": 0.40579, + "grad_norm": 0.8700058507978624, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 40579 + }, + { + "epoch": 0.4058, + "grad_norm": 0.8444339558390485, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 40580 + }, + { + "epoch": 0.40581, + "grad_norm": 0.8614131443414239, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 40581 + }, + { + "epoch": 0.40582, + "grad_norm": 0.907967262929301, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 40582 + }, + { + "epoch": 0.40583, + "grad_norm": 0.9571922237836805, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 40583 + }, + { + "epoch": 0.40584, + "grad_norm": 1.0976631279466031, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 40584 + }, + { + "epoch": 0.40585, + "grad_norm": 1.1715368892662377, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 40585 + }, + { + "epoch": 0.40586, + "grad_norm": 1.1983967454731541, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 40586 + }, + { + "epoch": 0.40587, + "grad_norm": 0.954011597291079, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 40587 + }, + { + "epoch": 0.40588, + "grad_norm": 1.1051701335002548, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 40588 + }, + { + "epoch": 0.40589, + "grad_norm": 1.104701677197118, + "learning_rate": 0.003, + "loss": 4.038, + "step": 40589 + }, + { + "epoch": 0.4059, + "grad_norm": 1.275013745572039, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 40590 + }, + { + "epoch": 0.40591, + "grad_norm": 0.9682434238820444, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 40591 + }, + { + "epoch": 0.40592, + "grad_norm": 1.2930056775855876, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 40592 + }, + { + "epoch": 0.40593, + "grad_norm": 1.0171416968922318, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 40593 + }, + { + "epoch": 0.40594, + "grad_norm": 0.9671670562543603, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 40594 + }, + { + "epoch": 0.40595, + "grad_norm": 1.1468510105678993, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 40595 + }, + { + "epoch": 0.40596, + "grad_norm": 1.4468631666250678, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 40596 + }, + { + "epoch": 0.40597, + "grad_norm": 0.998722583205008, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 40597 + }, + { + "epoch": 0.40598, + "grad_norm": 1.1085035865016468, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 40598 + }, + { + "epoch": 0.40599, + "grad_norm": 1.1115734219065854, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 40599 + }, + { + "epoch": 0.406, + "grad_norm": 1.1050873474350096, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 40600 + }, + { + "epoch": 0.40601, + "grad_norm": 0.9363859898289204, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 40601 + }, + { + "epoch": 0.40602, + "grad_norm": 1.0467173425880345, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 40602 + }, + { + "epoch": 0.40603, + "grad_norm": 1.3913556389226884, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 40603 + }, + { + "epoch": 0.40604, + "grad_norm": 0.7628394133563784, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 40604 + }, + { + "epoch": 0.40605, + "grad_norm": 0.8082012861386821, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 40605 + }, + { + "epoch": 0.40606, + "grad_norm": 0.8821019728015702, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 40606 + }, + { + "epoch": 0.40607, + "grad_norm": 0.9037111217422046, + "learning_rate": 0.003, + "loss": 4.024, + "step": 40607 + }, + { + "epoch": 0.40608, + "grad_norm": 0.8786162438163029, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 40608 + }, + { + "epoch": 0.40609, + "grad_norm": 0.953967614628948, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 40609 + }, + { + "epoch": 0.4061, + "grad_norm": 1.1899914090780628, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 40610 + }, + { + "epoch": 0.40611, + "grad_norm": 0.8511078968409675, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 40611 + }, + { + "epoch": 0.40612, + "grad_norm": 0.9384431544277916, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 40612 + }, + { + "epoch": 0.40613, + "grad_norm": 1.2280198638354733, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 40613 + }, + { + "epoch": 0.40614, + "grad_norm": 1.2947792455334741, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 40614 + }, + { + "epoch": 0.40615, + "grad_norm": 1.0600723044892282, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 40615 + }, + { + "epoch": 0.40616, + "grad_norm": 1.1053339051719628, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 40616 + }, + { + "epoch": 0.40617, + "grad_norm": 1.0066034759373497, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 40617 + }, + { + "epoch": 0.40618, + "grad_norm": 1.1133761475639958, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 40618 + }, + { + "epoch": 0.40619, + "grad_norm": 0.9883767838106811, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 40619 + }, + { + "epoch": 0.4062, + "grad_norm": 1.0847804154634157, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 40620 + }, + { + "epoch": 0.40621, + "grad_norm": 1.1426668453739621, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 40621 + }, + { + "epoch": 0.40622, + "grad_norm": 0.9221237543280941, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 40622 + }, + { + "epoch": 0.40623, + "grad_norm": 0.9062478197720261, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 40623 + }, + { + "epoch": 0.40624, + "grad_norm": 1.180593125898486, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 40624 + }, + { + "epoch": 0.40625, + "grad_norm": 1.158653022742318, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 40625 + }, + { + "epoch": 0.40626, + "grad_norm": 0.9446510273820464, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 40626 + }, + { + "epoch": 0.40627, + "grad_norm": 1.227938966634996, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 40627 + }, + { + "epoch": 0.40628, + "grad_norm": 1.1846356439445174, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 40628 + }, + { + "epoch": 0.40629, + "grad_norm": 0.9250299907539623, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 40629 + }, + { + "epoch": 0.4063, + "grad_norm": 1.2874647610793464, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 40630 + }, + { + "epoch": 0.40631, + "grad_norm": 1.3640822722470496, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 40631 + }, + { + "epoch": 0.40632, + "grad_norm": 0.8717309252833497, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 40632 + }, + { + "epoch": 0.40633, + "grad_norm": 0.688930623262783, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 40633 + }, + { + "epoch": 0.40634, + "grad_norm": 0.8241202238669261, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 40634 + }, + { + "epoch": 0.40635, + "grad_norm": 0.8136513884962313, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 40635 + }, + { + "epoch": 0.40636, + "grad_norm": 0.8698573489044984, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 40636 + }, + { + "epoch": 0.40637, + "grad_norm": 1.061656339695936, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 40637 + }, + { + "epoch": 0.40638, + "grad_norm": 1.2962617574432667, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 40638 + }, + { + "epoch": 0.40639, + "grad_norm": 0.9970998442762908, + "learning_rate": 0.003, + "loss": 4.043, + "step": 40639 + }, + { + "epoch": 0.4064, + "grad_norm": 1.2953928676612285, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 40640 + }, + { + "epoch": 0.40641, + "grad_norm": 1.0005688577198264, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 40641 + }, + { + "epoch": 0.40642, + "grad_norm": 1.203863713923529, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 40642 + }, + { + "epoch": 0.40643, + "grad_norm": 0.9569967915759877, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 40643 + }, + { + "epoch": 0.40644, + "grad_norm": 0.9983513493178072, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 40644 + }, + { + "epoch": 0.40645, + "grad_norm": 1.4014040119441171, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 40645 + }, + { + "epoch": 0.40646, + "grad_norm": 0.7029146296304535, + "learning_rate": 0.003, + "loss": 4.045, + "step": 40646 + }, + { + "epoch": 0.40647, + "grad_norm": 0.7873312384321083, + "learning_rate": 0.003, + "loss": 4.027, + "step": 40647 + }, + { + "epoch": 0.40648, + "grad_norm": 1.1247039754245995, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 40648 + }, + { + "epoch": 0.40649, + "grad_norm": 1.1127675741088916, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 40649 + }, + { + "epoch": 0.4065, + "grad_norm": 1.1565120047931807, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 40650 + }, + { + "epoch": 0.40651, + "grad_norm": 0.9937418017443775, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 40651 + }, + { + "epoch": 0.40652, + "grad_norm": 0.9184900131106533, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 40652 + }, + { + "epoch": 0.40653, + "grad_norm": 1.0504663736578534, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 40653 + }, + { + "epoch": 0.40654, + "grad_norm": 1.1621183339856986, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 40654 + }, + { + "epoch": 0.40655, + "grad_norm": 0.9116545302158149, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 40655 + }, + { + "epoch": 0.40656, + "grad_norm": 0.9526924754265126, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 40656 + }, + { + "epoch": 0.40657, + "grad_norm": 1.222130358892007, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 40657 + }, + { + "epoch": 0.40658, + "grad_norm": 1.1228278982011486, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 40658 + }, + { + "epoch": 0.40659, + "grad_norm": 1.46926471012703, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 40659 + }, + { + "epoch": 0.4066, + "grad_norm": 0.6897071370001082, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 40660 + }, + { + "epoch": 0.40661, + "grad_norm": 0.9820841664418085, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 40661 + }, + { + "epoch": 0.40662, + "grad_norm": 1.4161700201290481, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 40662 + }, + { + "epoch": 0.40663, + "grad_norm": 0.9255083758872037, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 40663 + }, + { + "epoch": 0.40664, + "grad_norm": 1.2432193418455233, + "learning_rate": 0.003, + "loss": 4.024, + "step": 40664 + }, + { + "epoch": 0.40665, + "grad_norm": 1.1067547231158228, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 40665 + }, + { + "epoch": 0.40666, + "grad_norm": 0.7799848354548488, + "learning_rate": 0.003, + "loss": 4.007, + "step": 40666 + }, + { + "epoch": 0.40667, + "grad_norm": 0.8505561697883274, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 40667 + }, + { + "epoch": 0.40668, + "grad_norm": 1.1019695161860255, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 40668 + }, + { + "epoch": 0.40669, + "grad_norm": 0.9801928173656965, + "learning_rate": 0.003, + "loss": 4.064, + "step": 40669 + }, + { + "epoch": 0.4067, + "grad_norm": 1.1222967825496977, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 40670 + }, + { + "epoch": 0.40671, + "grad_norm": 1.2590570403687074, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 40671 + }, + { + "epoch": 0.40672, + "grad_norm": 0.8603091408767158, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 40672 + }, + { + "epoch": 0.40673, + "grad_norm": 0.8933073774611615, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 40673 + }, + { + "epoch": 0.40674, + "grad_norm": 1.0183661474686998, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 40674 + }, + { + "epoch": 0.40675, + "grad_norm": 0.9873366647166003, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 40675 + }, + { + "epoch": 0.40676, + "grad_norm": 0.985271414517974, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 40676 + }, + { + "epoch": 0.40677, + "grad_norm": 1.1065222668330104, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 40677 + }, + { + "epoch": 0.40678, + "grad_norm": 1.155090261479002, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 40678 + }, + { + "epoch": 0.40679, + "grad_norm": 1.0693183615219957, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 40679 + }, + { + "epoch": 0.4068, + "grad_norm": 1.0696336124755974, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 40680 + }, + { + "epoch": 0.40681, + "grad_norm": 1.2527468277117055, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 40681 + }, + { + "epoch": 0.40682, + "grad_norm": 1.402978642050376, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 40682 + }, + { + "epoch": 0.40683, + "grad_norm": 0.8788259161932231, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 40683 + }, + { + "epoch": 0.40684, + "grad_norm": 0.8405377291850866, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 40684 + }, + { + "epoch": 0.40685, + "grad_norm": 1.113155943576139, + "learning_rate": 0.003, + "loss": 4.063, + "step": 40685 + }, + { + "epoch": 0.40686, + "grad_norm": 1.2488837310033662, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 40686 + }, + { + "epoch": 0.40687, + "grad_norm": 1.011815012517753, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 40687 + }, + { + "epoch": 0.40688, + "grad_norm": 1.137990664555404, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 40688 + }, + { + "epoch": 0.40689, + "grad_norm": 1.3775357734180154, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 40689 + }, + { + "epoch": 0.4069, + "grad_norm": 1.3956553316128693, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 40690 + }, + { + "epoch": 0.40691, + "grad_norm": 0.7110944127302834, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 40691 + }, + { + "epoch": 0.40692, + "grad_norm": 0.8181912149170566, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 40692 + }, + { + "epoch": 0.40693, + "grad_norm": 0.9328244216962431, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 40693 + }, + { + "epoch": 0.40694, + "grad_norm": 0.9637221316689879, + "learning_rate": 0.003, + "loss": 4.042, + "step": 40694 + }, + { + "epoch": 0.40695, + "grad_norm": 1.078318514636331, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 40695 + }, + { + "epoch": 0.40696, + "grad_norm": 1.4462078804047862, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 40696 + }, + { + "epoch": 0.40697, + "grad_norm": 1.1404141484684676, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 40697 + }, + { + "epoch": 0.40698, + "grad_norm": 1.0140607414309735, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 40698 + }, + { + "epoch": 0.40699, + "grad_norm": 1.0670364825863048, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 40699 + }, + { + "epoch": 0.407, + "grad_norm": 1.3256197807404695, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 40700 + }, + { + "epoch": 0.40701, + "grad_norm": 0.9369360149728705, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 40701 + }, + { + "epoch": 0.40702, + "grad_norm": 1.0643964524401954, + "learning_rate": 0.003, + "loss": 4.032, + "step": 40702 + }, + { + "epoch": 0.40703, + "grad_norm": 1.4001146112307943, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 40703 + }, + { + "epoch": 0.40704, + "grad_norm": 0.8278775988034895, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 40704 + }, + { + "epoch": 0.40705, + "grad_norm": 0.6606823091867571, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 40705 + }, + { + "epoch": 0.40706, + "grad_norm": 0.8550831835543423, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 40706 + }, + { + "epoch": 0.40707, + "grad_norm": 1.2107209361961941, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 40707 + }, + { + "epoch": 0.40708, + "grad_norm": 1.3487479287082786, + "learning_rate": 0.003, + "loss": 4.063, + "step": 40708 + }, + { + "epoch": 0.40709, + "grad_norm": 0.8460884005285113, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 40709 + }, + { + "epoch": 0.4071, + "grad_norm": 0.7761545892691055, + "learning_rate": 0.003, + "loss": 4.036, + "step": 40710 + }, + { + "epoch": 0.40711, + "grad_norm": 0.9070761129498249, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 40711 + }, + { + "epoch": 0.40712, + "grad_norm": 1.1088662677593224, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 40712 + }, + { + "epoch": 0.40713, + "grad_norm": 1.0570414005289421, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 40713 + }, + { + "epoch": 0.40714, + "grad_norm": 1.0410625718089368, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 40714 + }, + { + "epoch": 0.40715, + "grad_norm": 1.0104562787526294, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 40715 + }, + { + "epoch": 0.40716, + "grad_norm": 1.0866347308805953, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 40716 + }, + { + "epoch": 0.40717, + "grad_norm": 1.0724126926582875, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 40717 + }, + { + "epoch": 0.40718, + "grad_norm": 1.1206794447131647, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 40718 + }, + { + "epoch": 0.40719, + "grad_norm": 1.231796725260632, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 40719 + }, + { + "epoch": 0.4072, + "grad_norm": 0.9441633051226499, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 40720 + }, + { + "epoch": 0.40721, + "grad_norm": 0.9908735778395974, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 40721 + }, + { + "epoch": 0.40722, + "grad_norm": 1.4394366316006995, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 40722 + }, + { + "epoch": 0.40723, + "grad_norm": 1.0118353391123371, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 40723 + }, + { + "epoch": 0.40724, + "grad_norm": 1.2431434540256996, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 40724 + }, + { + "epoch": 0.40725, + "grad_norm": 0.9302195760924563, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 40725 + }, + { + "epoch": 0.40726, + "grad_norm": 1.1344396674919581, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 40726 + }, + { + "epoch": 0.40727, + "grad_norm": 1.2988586576588257, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 40727 + }, + { + "epoch": 0.40728, + "grad_norm": 1.0153330398486633, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 40728 + }, + { + "epoch": 0.40729, + "grad_norm": 1.2728652484334133, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 40729 + }, + { + "epoch": 0.4073, + "grad_norm": 0.9711789647169907, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 40730 + }, + { + "epoch": 0.40731, + "grad_norm": 1.0068398087423334, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 40731 + }, + { + "epoch": 0.40732, + "grad_norm": 1.1889366677975917, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 40732 + }, + { + "epoch": 0.40733, + "grad_norm": 1.2924326024233619, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 40733 + }, + { + "epoch": 0.40734, + "grad_norm": 0.9263620357553181, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 40734 + }, + { + "epoch": 0.40735, + "grad_norm": 0.9698370360714099, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 40735 + }, + { + "epoch": 0.40736, + "grad_norm": 1.315890856903035, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 40736 + }, + { + "epoch": 0.40737, + "grad_norm": 0.858143219877298, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 40737 + }, + { + "epoch": 0.40738, + "grad_norm": 0.8397291592902834, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 40738 + }, + { + "epoch": 0.40739, + "grad_norm": 0.9092792726194555, + "learning_rate": 0.003, + "loss": 4.034, + "step": 40739 + }, + { + "epoch": 0.4074, + "grad_norm": 0.9956709643987418, + "learning_rate": 0.003, + "loss": 4.045, + "step": 40740 + }, + { + "epoch": 0.40741, + "grad_norm": 1.1607852332716202, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 40741 + }, + { + "epoch": 0.40742, + "grad_norm": 1.138332053111871, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 40742 + }, + { + "epoch": 0.40743, + "grad_norm": 0.9771385654751203, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 40743 + }, + { + "epoch": 0.40744, + "grad_norm": 0.9715591336726043, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 40744 + }, + { + "epoch": 0.40745, + "grad_norm": 1.1249951774120328, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 40745 + }, + { + "epoch": 0.40746, + "grad_norm": 1.1995856003767387, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 40746 + }, + { + "epoch": 0.40747, + "grad_norm": 0.997209081878877, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 40747 + }, + { + "epoch": 0.40748, + "grad_norm": 1.6923195784055536, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 40748 + }, + { + "epoch": 0.40749, + "grad_norm": 0.8249961967840708, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 40749 + }, + { + "epoch": 0.4075, + "grad_norm": 0.8127107601878526, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 40750 + }, + { + "epoch": 0.40751, + "grad_norm": 0.8816292468908672, + "learning_rate": 0.003, + "loss": 4.052, + "step": 40751 + }, + { + "epoch": 0.40752, + "grad_norm": 0.8669136113942759, + "learning_rate": 0.003, + "loss": 4.049, + "step": 40752 + }, + { + "epoch": 0.40753, + "grad_norm": 1.024060815741023, + "learning_rate": 0.003, + "loss": 4.035, + "step": 40753 + }, + { + "epoch": 0.40754, + "grad_norm": 1.2461628652479617, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 40754 + }, + { + "epoch": 0.40755, + "grad_norm": 1.106753293279856, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 40755 + }, + { + "epoch": 0.40756, + "grad_norm": 1.2046630961000842, + "learning_rate": 0.003, + "loss": 4.07, + "step": 40756 + }, + { + "epoch": 0.40757, + "grad_norm": 1.190005889646082, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 40757 + }, + { + "epoch": 0.40758, + "grad_norm": 1.0049937782798173, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 40758 + }, + { + "epoch": 0.40759, + "grad_norm": 1.1126867805293406, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 40759 + }, + { + "epoch": 0.4076, + "grad_norm": 1.0673992554957414, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 40760 + }, + { + "epoch": 0.40761, + "grad_norm": 1.1643363135733684, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 40761 + }, + { + "epoch": 0.40762, + "grad_norm": 1.0587650370405057, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 40762 + }, + { + "epoch": 0.40763, + "grad_norm": 1.0607452909015225, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 40763 + }, + { + "epoch": 0.40764, + "grad_norm": 1.0620766865650724, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 40764 + }, + { + "epoch": 0.40765, + "grad_norm": 1.2146714976608732, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 40765 + }, + { + "epoch": 0.40766, + "grad_norm": 1.1076146532502362, + "learning_rate": 0.003, + "loss": 4.047, + "step": 40766 + }, + { + "epoch": 0.40767, + "grad_norm": 1.167784409028509, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 40767 + }, + { + "epoch": 0.40768, + "grad_norm": 0.9396100115685684, + "learning_rate": 0.003, + "loss": 4.037, + "step": 40768 + }, + { + "epoch": 0.40769, + "grad_norm": 0.9825792714813089, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 40769 + }, + { + "epoch": 0.4077, + "grad_norm": 1.143659858560104, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 40770 + }, + { + "epoch": 0.40771, + "grad_norm": 1.2348041919845905, + "learning_rate": 0.003, + "loss": 4.003, + "step": 40771 + }, + { + "epoch": 0.40772, + "grad_norm": 0.9929073473746659, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 40772 + }, + { + "epoch": 0.40773, + "grad_norm": 1.1864893071327582, + "learning_rate": 0.003, + "loss": 4.042, + "step": 40773 + }, + { + "epoch": 0.40774, + "grad_norm": 0.9661857607243343, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 40774 + }, + { + "epoch": 0.40775, + "grad_norm": 0.954908834662701, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 40775 + }, + { + "epoch": 0.40776, + "grad_norm": 1.0333610874333656, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 40776 + }, + { + "epoch": 0.40777, + "grad_norm": 0.9726483049714751, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 40777 + }, + { + "epoch": 0.40778, + "grad_norm": 1.165794091299309, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 40778 + }, + { + "epoch": 0.40779, + "grad_norm": 1.1535585781623565, + "learning_rate": 0.003, + "loss": 4.054, + "step": 40779 + }, + { + "epoch": 0.4078, + "grad_norm": 0.9651888728917077, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 40780 + }, + { + "epoch": 0.40781, + "grad_norm": 0.9661910429330726, + "learning_rate": 0.003, + "loss": 4.013, + "step": 40781 + }, + { + "epoch": 0.40782, + "grad_norm": 1.2725836553234255, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 40782 + }, + { + "epoch": 0.40783, + "grad_norm": 0.9768753119569844, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 40783 + }, + { + "epoch": 0.40784, + "grad_norm": 1.09972861750637, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 40784 + }, + { + "epoch": 0.40785, + "grad_norm": 0.8967278396784361, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 40785 + }, + { + "epoch": 0.40786, + "grad_norm": 0.8808390608630816, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 40786 + }, + { + "epoch": 0.40787, + "grad_norm": 1.120875772429093, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 40787 + }, + { + "epoch": 0.40788, + "grad_norm": 0.9798740414770696, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 40788 + }, + { + "epoch": 0.40789, + "grad_norm": 1.2469156762591722, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 40789 + }, + { + "epoch": 0.4079, + "grad_norm": 1.0300575200065782, + "learning_rate": 0.003, + "loss": 4.021, + "step": 40790 + }, + { + "epoch": 0.40791, + "grad_norm": 1.133789004808889, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 40791 + }, + { + "epoch": 0.40792, + "grad_norm": 0.9589694272069129, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 40792 + }, + { + "epoch": 0.40793, + "grad_norm": 1.0816253937438536, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 40793 + }, + { + "epoch": 0.40794, + "grad_norm": 1.2768716838035126, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 40794 + }, + { + "epoch": 0.40795, + "grad_norm": 1.0157077681703888, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 40795 + }, + { + "epoch": 0.40796, + "grad_norm": 1.1177473413630599, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 40796 + }, + { + "epoch": 0.40797, + "grad_norm": 0.9629615217426917, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 40797 + }, + { + "epoch": 0.40798, + "grad_norm": 1.2171415602676892, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 40798 + }, + { + "epoch": 0.40799, + "grad_norm": 1.0753906187297038, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 40799 + }, + { + "epoch": 0.408, + "grad_norm": 0.9305502456018223, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 40800 + }, + { + "epoch": 0.40801, + "grad_norm": 1.0502125779036493, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 40801 + }, + { + "epoch": 0.40802, + "grad_norm": 1.1996118978202028, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 40802 + }, + { + "epoch": 0.40803, + "grad_norm": 1.0489703883710568, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 40803 + }, + { + "epoch": 0.40804, + "grad_norm": 1.1435954575068068, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 40804 + }, + { + "epoch": 0.40805, + "grad_norm": 1.1609751897992664, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 40805 + }, + { + "epoch": 0.40806, + "grad_norm": 0.9593623948398567, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 40806 + }, + { + "epoch": 0.40807, + "grad_norm": 0.8614329558828198, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 40807 + }, + { + "epoch": 0.40808, + "grad_norm": 0.9238167235491063, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 40808 + }, + { + "epoch": 0.40809, + "grad_norm": 1.064844023949492, + "learning_rate": 0.003, + "loss": 4.035, + "step": 40809 + }, + { + "epoch": 0.4081, + "grad_norm": 1.1784034384289528, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 40810 + }, + { + "epoch": 0.40811, + "grad_norm": 1.0041847586113741, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 40811 + }, + { + "epoch": 0.40812, + "grad_norm": 1.1226244454083114, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 40812 + }, + { + "epoch": 0.40813, + "grad_norm": 1.2664360904426437, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 40813 + }, + { + "epoch": 0.40814, + "grad_norm": 1.1546663259042673, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 40814 + }, + { + "epoch": 0.40815, + "grad_norm": 0.9036941629017081, + "learning_rate": 0.003, + "loss": 4.042, + "step": 40815 + }, + { + "epoch": 0.40816, + "grad_norm": 1.025212772365007, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 40816 + }, + { + "epoch": 0.40817, + "grad_norm": 1.2458739665042458, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 40817 + }, + { + "epoch": 0.40818, + "grad_norm": 0.9609004291491335, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 40818 + }, + { + "epoch": 0.40819, + "grad_norm": 1.1186765926948252, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 40819 + }, + { + "epoch": 0.4082, + "grad_norm": 1.2375692792873765, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 40820 + }, + { + "epoch": 0.40821, + "grad_norm": 0.8781912535985548, + "learning_rate": 0.003, + "loss": 4.033, + "step": 40821 + }, + { + "epoch": 0.40822, + "grad_norm": 0.8420386991167131, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 40822 + }, + { + "epoch": 0.40823, + "grad_norm": 0.9425032336566739, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 40823 + }, + { + "epoch": 0.40824, + "grad_norm": 1.0929408316343157, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 40824 + }, + { + "epoch": 0.40825, + "grad_norm": 1.1630053853154225, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 40825 + }, + { + "epoch": 0.40826, + "grad_norm": 1.1283358631949634, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 40826 + }, + { + "epoch": 0.40827, + "grad_norm": 1.1798225026223534, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 40827 + }, + { + "epoch": 0.40828, + "grad_norm": 1.0056393563674977, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 40828 + }, + { + "epoch": 0.40829, + "grad_norm": 1.1492926885166246, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 40829 + }, + { + "epoch": 0.4083, + "grad_norm": 1.325937953564833, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 40830 + }, + { + "epoch": 0.40831, + "grad_norm": 1.0212365851692538, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 40831 + }, + { + "epoch": 0.40832, + "grad_norm": 1.1637723823686419, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 40832 + }, + { + "epoch": 0.40833, + "grad_norm": 1.1434591515733068, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 40833 + }, + { + "epoch": 0.40834, + "grad_norm": 0.9021900741625088, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 40834 + }, + { + "epoch": 0.40835, + "grad_norm": 1.0108027804236792, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 40835 + }, + { + "epoch": 0.40836, + "grad_norm": 1.1752005363182174, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 40836 + }, + { + "epoch": 0.40837, + "grad_norm": 0.9970695096361344, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 40837 + }, + { + "epoch": 0.40838, + "grad_norm": 1.1520453599375333, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 40838 + }, + { + "epoch": 0.40839, + "grad_norm": 0.8866864966058302, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 40839 + }, + { + "epoch": 0.4084, + "grad_norm": 1.0027649484627974, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 40840 + }, + { + "epoch": 0.40841, + "grad_norm": 1.593817266706805, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 40841 + }, + { + "epoch": 0.40842, + "grad_norm": 0.6312269754966543, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 40842 + }, + { + "epoch": 0.40843, + "grad_norm": 0.8552702997874458, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 40843 + }, + { + "epoch": 0.40844, + "grad_norm": 1.1691519167445035, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 40844 + }, + { + "epoch": 0.40845, + "grad_norm": 1.0723029082170252, + "learning_rate": 0.003, + "loss": 4.027, + "step": 40845 + }, + { + "epoch": 0.40846, + "grad_norm": 1.2305459047467684, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 40846 + }, + { + "epoch": 0.40847, + "grad_norm": 0.9326659716201073, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 40847 + }, + { + "epoch": 0.40848, + "grad_norm": 0.951955307141545, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 40848 + }, + { + "epoch": 0.40849, + "grad_norm": 1.0062002347366352, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 40849 + }, + { + "epoch": 0.4085, + "grad_norm": 1.0980851731760004, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 40850 + }, + { + "epoch": 0.40851, + "grad_norm": 0.950045887014325, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 40851 + }, + { + "epoch": 0.40852, + "grad_norm": 1.0909615430857167, + "learning_rate": 0.003, + "loss": 4.014, + "step": 40852 + }, + { + "epoch": 0.40853, + "grad_norm": 1.057011009566548, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 40853 + }, + { + "epoch": 0.40854, + "grad_norm": 1.0478726240517897, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 40854 + }, + { + "epoch": 0.40855, + "grad_norm": 1.216509258898252, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 40855 + }, + { + "epoch": 0.40856, + "grad_norm": 0.933278104799428, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 40856 + }, + { + "epoch": 0.40857, + "grad_norm": 0.9662373483775711, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 40857 + }, + { + "epoch": 0.40858, + "grad_norm": 0.9863412827149053, + "learning_rate": 0.003, + "loss": 4.031, + "step": 40858 + }, + { + "epoch": 0.40859, + "grad_norm": 1.2672032342135822, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 40859 + }, + { + "epoch": 0.4086, + "grad_norm": 0.9468192816957822, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 40860 + }, + { + "epoch": 0.40861, + "grad_norm": 1.1004873423831123, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 40861 + }, + { + "epoch": 0.40862, + "grad_norm": 1.0904830231070963, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 40862 + }, + { + "epoch": 0.40863, + "grad_norm": 1.0854942053908545, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 40863 + }, + { + "epoch": 0.40864, + "grad_norm": 1.0528435395068676, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 40864 + }, + { + "epoch": 0.40865, + "grad_norm": 1.0150515680676986, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 40865 + }, + { + "epoch": 0.40866, + "grad_norm": 1.1101368171464046, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 40866 + }, + { + "epoch": 0.40867, + "grad_norm": 1.2282093440748325, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 40867 + }, + { + "epoch": 0.40868, + "grad_norm": 1.2111676894887413, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 40868 + }, + { + "epoch": 0.40869, + "grad_norm": 0.8787390422081278, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 40869 + }, + { + "epoch": 0.4087, + "grad_norm": 1.0295318652798364, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 40870 + }, + { + "epoch": 0.40871, + "grad_norm": 1.2539137421831879, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 40871 + }, + { + "epoch": 0.40872, + "grad_norm": 0.9264748750373235, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 40872 + }, + { + "epoch": 0.40873, + "grad_norm": 0.8776560779316889, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 40873 + }, + { + "epoch": 0.40874, + "grad_norm": 0.8660903161965275, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 40874 + }, + { + "epoch": 0.40875, + "grad_norm": 1.0893770219176546, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 40875 + }, + { + "epoch": 0.40876, + "grad_norm": 1.1376785492616788, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 40876 + }, + { + "epoch": 0.40877, + "grad_norm": 1.2420372584446338, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 40877 + }, + { + "epoch": 0.40878, + "grad_norm": 0.8107032561618805, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 40878 + }, + { + "epoch": 0.40879, + "grad_norm": 0.8047954402645815, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 40879 + }, + { + "epoch": 0.4088, + "grad_norm": 1.1946616883166963, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 40880 + }, + { + "epoch": 0.40881, + "grad_norm": 1.1829555458124983, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 40881 + }, + { + "epoch": 0.40882, + "grad_norm": 0.9634242741202859, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 40882 + }, + { + "epoch": 0.40883, + "grad_norm": 1.0317471419952715, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 40883 + }, + { + "epoch": 0.40884, + "grad_norm": 1.1116742817939884, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 40884 + }, + { + "epoch": 0.40885, + "grad_norm": 1.008866595619115, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 40885 + }, + { + "epoch": 0.40886, + "grad_norm": 1.2226888047363942, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 40886 + }, + { + "epoch": 0.40887, + "grad_norm": 0.8395872934515421, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 40887 + }, + { + "epoch": 0.40888, + "grad_norm": 0.8892648695982368, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 40888 + }, + { + "epoch": 0.40889, + "grad_norm": 0.9886952212810075, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 40889 + }, + { + "epoch": 0.4089, + "grad_norm": 0.9699040355549081, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 40890 + }, + { + "epoch": 0.40891, + "grad_norm": 1.231736695635059, + "learning_rate": 0.003, + "loss": 4.048, + "step": 40891 + }, + { + "epoch": 0.40892, + "grad_norm": 1.2412186217851797, + "learning_rate": 0.003, + "loss": 4.009, + "step": 40892 + }, + { + "epoch": 0.40893, + "grad_norm": 1.2859062952499283, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 40893 + }, + { + "epoch": 0.40894, + "grad_norm": 1.1630758467697901, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 40894 + }, + { + "epoch": 0.40895, + "grad_norm": 1.0454813481787086, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 40895 + }, + { + "epoch": 0.40896, + "grad_norm": 1.0897641403884692, + "learning_rate": 0.003, + "loss": 4.073, + "step": 40896 + }, + { + "epoch": 0.40897, + "grad_norm": 1.1475851481699442, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 40897 + }, + { + "epoch": 0.40898, + "grad_norm": 1.0891702510660461, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 40898 + }, + { + "epoch": 0.40899, + "grad_norm": 0.9261882259742024, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 40899 + }, + { + "epoch": 0.409, + "grad_norm": 0.8718781799703895, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 40900 + }, + { + "epoch": 0.40901, + "grad_norm": 1.0410147372341454, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 40901 + }, + { + "epoch": 0.40902, + "grad_norm": 1.1438284013482707, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 40902 + }, + { + "epoch": 0.40903, + "grad_norm": 1.0175373016550355, + "learning_rate": 0.003, + "loss": 4.023, + "step": 40903 + }, + { + "epoch": 0.40904, + "grad_norm": 1.2422327149304997, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 40904 + }, + { + "epoch": 0.40905, + "grad_norm": 1.114726824199648, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 40905 + }, + { + "epoch": 0.40906, + "grad_norm": 1.2634451113536749, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 40906 + }, + { + "epoch": 0.40907, + "grad_norm": 0.8433872120063, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 40907 + }, + { + "epoch": 0.40908, + "grad_norm": 0.8385949385815775, + "learning_rate": 0.003, + "loss": 4.051, + "step": 40908 + }, + { + "epoch": 0.40909, + "grad_norm": 0.9747987906774352, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 40909 + }, + { + "epoch": 0.4091, + "grad_norm": 1.1138306046754116, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 40910 + }, + { + "epoch": 0.40911, + "grad_norm": 1.0310184580733734, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 40911 + }, + { + "epoch": 0.40912, + "grad_norm": 1.1501918030433658, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 40912 + }, + { + "epoch": 0.40913, + "grad_norm": 1.1776250141501603, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 40913 + }, + { + "epoch": 0.40914, + "grad_norm": 1.180669283086804, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 40914 + }, + { + "epoch": 0.40915, + "grad_norm": 0.9260086182993037, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 40915 + }, + { + "epoch": 0.40916, + "grad_norm": 0.9332282410287327, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 40916 + }, + { + "epoch": 0.40917, + "grad_norm": 1.1755871922416503, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 40917 + }, + { + "epoch": 0.40918, + "grad_norm": 1.0422909013892103, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 40918 + }, + { + "epoch": 0.40919, + "grad_norm": 1.0522828833083224, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 40919 + }, + { + "epoch": 0.4092, + "grad_norm": 0.9857138739185575, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 40920 + }, + { + "epoch": 0.40921, + "grad_norm": 1.184476367903165, + "learning_rate": 0.003, + "loss": 4.009, + "step": 40921 + }, + { + "epoch": 0.40922, + "grad_norm": 1.288164353159181, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 40922 + }, + { + "epoch": 0.40923, + "grad_norm": 0.8873750172289145, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 40923 + }, + { + "epoch": 0.40924, + "grad_norm": 0.7985303185485144, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 40924 + }, + { + "epoch": 0.40925, + "grad_norm": 0.8981224541814002, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 40925 + }, + { + "epoch": 0.40926, + "grad_norm": 1.0014159560135933, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 40926 + }, + { + "epoch": 0.40927, + "grad_norm": 1.1528104106659183, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 40927 + }, + { + "epoch": 0.40928, + "grad_norm": 1.077530802768206, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 40928 + }, + { + "epoch": 0.40929, + "grad_norm": 0.9134277539373763, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 40929 + }, + { + "epoch": 0.4093, + "grad_norm": 1.0212898999218738, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 40930 + }, + { + "epoch": 0.40931, + "grad_norm": 1.1886544132932135, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 40931 + }, + { + "epoch": 0.40932, + "grad_norm": 1.0367422751578728, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 40932 + }, + { + "epoch": 0.40933, + "grad_norm": 1.4082051217055163, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 40933 + }, + { + "epoch": 0.40934, + "grad_norm": 1.0379690737434832, + "learning_rate": 0.003, + "loss": 4.024, + "step": 40934 + }, + { + "epoch": 0.40935, + "grad_norm": 1.1612825337300132, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 40935 + }, + { + "epoch": 0.40936, + "grad_norm": 0.8900128018877548, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 40936 + }, + { + "epoch": 0.40937, + "grad_norm": 1.0350516563939618, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 40937 + }, + { + "epoch": 0.40938, + "grad_norm": 1.5330484429163342, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 40938 + }, + { + "epoch": 0.40939, + "grad_norm": 0.8010097781902149, + "learning_rate": 0.003, + "loss": 4.024, + "step": 40939 + }, + { + "epoch": 0.4094, + "grad_norm": 0.7722855905003154, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 40940 + }, + { + "epoch": 0.40941, + "grad_norm": 0.9007435260685063, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 40941 + }, + { + "epoch": 0.40942, + "grad_norm": 1.236686663739773, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 40942 + }, + { + "epoch": 0.40943, + "grad_norm": 0.9066542316416157, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 40943 + }, + { + "epoch": 0.40944, + "grad_norm": 1.0462062143279214, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 40944 + }, + { + "epoch": 0.40945, + "grad_norm": 1.3267559116055199, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 40945 + }, + { + "epoch": 0.40946, + "grad_norm": 0.984436118225778, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 40946 + }, + { + "epoch": 0.40947, + "grad_norm": 1.0879463531463842, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 40947 + }, + { + "epoch": 0.40948, + "grad_norm": 1.0685324172885526, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 40948 + }, + { + "epoch": 0.40949, + "grad_norm": 1.132219618426926, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 40949 + }, + { + "epoch": 0.4095, + "grad_norm": 1.2423722726640156, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 40950 + }, + { + "epoch": 0.40951, + "grad_norm": 0.9535983112323408, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 40951 + }, + { + "epoch": 0.40952, + "grad_norm": 0.9644038087661427, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 40952 + }, + { + "epoch": 0.40953, + "grad_norm": 1.1089690827581633, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 40953 + }, + { + "epoch": 0.40954, + "grad_norm": 0.9954864226327984, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 40954 + }, + { + "epoch": 0.40955, + "grad_norm": 1.1290997653446089, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 40955 + }, + { + "epoch": 0.40956, + "grad_norm": 0.9698806910333939, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 40956 + }, + { + "epoch": 0.40957, + "grad_norm": 1.0646275902424096, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 40957 + }, + { + "epoch": 0.40958, + "grad_norm": 1.041165536743987, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 40958 + }, + { + "epoch": 0.40959, + "grad_norm": 1.1960009464240593, + "learning_rate": 0.003, + "loss": 4.015, + "step": 40959 + }, + { + "epoch": 0.4096, + "grad_norm": 1.0055987553988124, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 40960 + }, + { + "epoch": 0.40961, + "grad_norm": 1.140262170311156, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 40961 + }, + { + "epoch": 0.40962, + "grad_norm": 0.9306996426706864, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 40962 + }, + { + "epoch": 0.40963, + "grad_norm": 0.9352205411846674, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 40963 + }, + { + "epoch": 0.40964, + "grad_norm": 0.9869151484199574, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 40964 + }, + { + "epoch": 0.40965, + "grad_norm": 1.143374843689886, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 40965 + }, + { + "epoch": 0.40966, + "grad_norm": 1.335538358165866, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 40966 + }, + { + "epoch": 0.40967, + "grad_norm": 1.2184530840245273, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 40967 + }, + { + "epoch": 0.40968, + "grad_norm": 1.0467294556561855, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 40968 + }, + { + "epoch": 0.40969, + "grad_norm": 1.027847384157023, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 40969 + }, + { + "epoch": 0.4097, + "grad_norm": 1.3065715950195746, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 40970 + }, + { + "epoch": 0.40971, + "grad_norm": 1.1117908622892805, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 40971 + }, + { + "epoch": 0.40972, + "grad_norm": 1.035982133813377, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 40972 + }, + { + "epoch": 0.40973, + "grad_norm": 1.1620392461864082, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 40973 + }, + { + "epoch": 0.40974, + "grad_norm": 0.9239286727135053, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 40974 + }, + { + "epoch": 0.40975, + "grad_norm": 0.8618744265331247, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 40975 + }, + { + "epoch": 0.40976, + "grad_norm": 1.023506331767386, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 40976 + }, + { + "epoch": 0.40977, + "grad_norm": 1.3706735203305025, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 40977 + }, + { + "epoch": 0.40978, + "grad_norm": 0.7440676278536612, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 40978 + }, + { + "epoch": 0.40979, + "grad_norm": 0.7361732664400842, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 40979 + }, + { + "epoch": 0.4098, + "grad_norm": 0.7781595715190481, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 40980 + }, + { + "epoch": 0.40981, + "grad_norm": 0.8627844848313678, + "learning_rate": 0.003, + "loss": 4.023, + "step": 40981 + }, + { + "epoch": 0.40982, + "grad_norm": 1.1066483286422875, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 40982 + }, + { + "epoch": 0.40983, + "grad_norm": 1.0795820218881378, + "learning_rate": 0.003, + "loss": 4.059, + "step": 40983 + }, + { + "epoch": 0.40984, + "grad_norm": 1.295892350548295, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 40984 + }, + { + "epoch": 0.40985, + "grad_norm": 1.0965319383075571, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 40985 + }, + { + "epoch": 0.40986, + "grad_norm": 1.1900214992549574, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 40986 + }, + { + "epoch": 0.40987, + "grad_norm": 0.9069969126943567, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 40987 + }, + { + "epoch": 0.40988, + "grad_norm": 1.1635564365216307, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 40988 + }, + { + "epoch": 0.40989, + "grad_norm": 1.3239428525590937, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 40989 + }, + { + "epoch": 0.4099, + "grad_norm": 0.8617735285387589, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 40990 + }, + { + "epoch": 0.40991, + "grad_norm": 0.8341631349990561, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 40991 + }, + { + "epoch": 0.40992, + "grad_norm": 0.8688875602756315, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 40992 + }, + { + "epoch": 0.40993, + "grad_norm": 0.9938049189080334, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 40993 + }, + { + "epoch": 0.40994, + "grad_norm": 1.244409686635115, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 40994 + }, + { + "epoch": 0.40995, + "grad_norm": 1.071872852414368, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 40995 + }, + { + "epoch": 0.40996, + "grad_norm": 1.31352810925012, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 40996 + }, + { + "epoch": 0.40997, + "grad_norm": 1.071151019567857, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 40997 + }, + { + "epoch": 0.40998, + "grad_norm": 1.168038574917678, + "learning_rate": 0.003, + "loss": 4.037, + "step": 40998 + }, + { + "epoch": 0.40999, + "grad_norm": 1.057074714995987, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 40999 + }, + { + "epoch": 0.41, + "grad_norm": 0.9856688761174721, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 41000 + }, + { + "epoch": 0.41001, + "grad_norm": 1.1248881466387706, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 41001 + }, + { + "epoch": 0.41002, + "grad_norm": 1.1021261635830792, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 41002 + }, + { + "epoch": 0.41003, + "grad_norm": 1.09436379693237, + "learning_rate": 0.003, + "loss": 4.046, + "step": 41003 + }, + { + "epoch": 0.41004, + "grad_norm": 0.9283191922560484, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 41004 + }, + { + "epoch": 0.41005, + "grad_norm": 1.0858729340959525, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 41005 + }, + { + "epoch": 0.41006, + "grad_norm": 0.8366160872536866, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 41006 + }, + { + "epoch": 0.41007, + "grad_norm": 0.8357842056429061, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 41007 + }, + { + "epoch": 0.41008, + "grad_norm": 0.8669970867358748, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 41008 + }, + { + "epoch": 0.41009, + "grad_norm": 0.9415105233331126, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 41009 + }, + { + "epoch": 0.4101, + "grad_norm": 1.124024688710482, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 41010 + }, + { + "epoch": 0.41011, + "grad_norm": 1.2905555194846305, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 41011 + }, + { + "epoch": 0.41012, + "grad_norm": 1.1798066625686583, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 41012 + }, + { + "epoch": 0.41013, + "grad_norm": 0.9130746393966519, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 41013 + }, + { + "epoch": 0.41014, + "grad_norm": 1.3331544366618742, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 41014 + }, + { + "epoch": 0.41015, + "grad_norm": 1.2001522832564486, + "learning_rate": 0.003, + "loss": 4.067, + "step": 41015 + }, + { + "epoch": 0.41016, + "grad_norm": 0.8084619118997044, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 41016 + }, + { + "epoch": 0.41017, + "grad_norm": 0.8938411375154939, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 41017 + }, + { + "epoch": 0.41018, + "grad_norm": 1.4139873022825773, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 41018 + }, + { + "epoch": 0.41019, + "grad_norm": 0.991092539549531, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 41019 + }, + { + "epoch": 0.4102, + "grad_norm": 1.0875319062364612, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 41020 + }, + { + "epoch": 0.41021, + "grad_norm": 1.2770149700840059, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 41021 + }, + { + "epoch": 0.41022, + "grad_norm": 0.8763344421481445, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 41022 + }, + { + "epoch": 0.41023, + "grad_norm": 0.8488404231307977, + "learning_rate": 0.003, + "loss": 4.043, + "step": 41023 + }, + { + "epoch": 0.41024, + "grad_norm": 0.874400033172728, + "learning_rate": 0.003, + "loss": 4.029, + "step": 41024 + }, + { + "epoch": 0.41025, + "grad_norm": 1.0030690915782654, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 41025 + }, + { + "epoch": 0.41026, + "grad_norm": 1.0814766812125562, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 41026 + }, + { + "epoch": 0.41027, + "grad_norm": 1.1809492786912235, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 41027 + }, + { + "epoch": 0.41028, + "grad_norm": 1.0801817989871474, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 41028 + }, + { + "epoch": 0.41029, + "grad_norm": 1.0894829045108065, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 41029 + }, + { + "epoch": 0.4103, + "grad_norm": 1.2645480353756395, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 41030 + }, + { + "epoch": 0.41031, + "grad_norm": 1.2212162345113002, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 41031 + }, + { + "epoch": 0.41032, + "grad_norm": 0.9437899517450534, + "learning_rate": 0.003, + "loss": 4.017, + "step": 41032 + }, + { + "epoch": 0.41033, + "grad_norm": 0.9026979040642206, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 41033 + }, + { + "epoch": 0.41034, + "grad_norm": 1.052280115706445, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 41034 + }, + { + "epoch": 0.41035, + "grad_norm": 1.0670399766381284, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 41035 + }, + { + "epoch": 0.41036, + "grad_norm": 0.978855192409276, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 41036 + }, + { + "epoch": 0.41037, + "grad_norm": 0.9044192248390404, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 41037 + }, + { + "epoch": 0.41038, + "grad_norm": 0.9757079928047061, + "learning_rate": 0.003, + "loss": 4.045, + "step": 41038 + }, + { + "epoch": 0.41039, + "grad_norm": 1.2311062736978935, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 41039 + }, + { + "epoch": 0.4104, + "grad_norm": 1.3053241459971556, + "learning_rate": 0.003, + "loss": 4.0848, + "step": 41040 + }, + { + "epoch": 0.41041, + "grad_norm": 0.8041580983266724, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 41041 + }, + { + "epoch": 0.41042, + "grad_norm": 0.9007321241025572, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 41042 + }, + { + "epoch": 0.41043, + "grad_norm": 1.2037066761708555, + "learning_rate": 0.003, + "loss": 4.05, + "step": 41043 + }, + { + "epoch": 0.41044, + "grad_norm": 1.0943815986465355, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 41044 + }, + { + "epoch": 0.41045, + "grad_norm": 0.9368810468610071, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 41045 + }, + { + "epoch": 0.41046, + "grad_norm": 1.2212105888353755, + "learning_rate": 0.003, + "loss": 4.046, + "step": 41046 + }, + { + "epoch": 0.41047, + "grad_norm": 1.2224756183199246, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 41047 + }, + { + "epoch": 0.41048, + "grad_norm": 0.8307108188048467, + "learning_rate": 0.003, + "loss": 3.997, + "step": 41048 + }, + { + "epoch": 0.41049, + "grad_norm": 0.935996649580873, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 41049 + }, + { + "epoch": 0.4105, + "grad_norm": 1.1080097147687589, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 41050 + }, + { + "epoch": 0.41051, + "grad_norm": 1.2085624902896925, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 41051 + }, + { + "epoch": 0.41052, + "grad_norm": 1.1348796847227698, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 41052 + }, + { + "epoch": 0.41053, + "grad_norm": 1.3350982189867793, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 41053 + }, + { + "epoch": 0.41054, + "grad_norm": 0.844950574843246, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 41054 + }, + { + "epoch": 0.41055, + "grad_norm": 0.7626237211985046, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 41055 + }, + { + "epoch": 0.41056, + "grad_norm": 0.9177896511054638, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 41056 + }, + { + "epoch": 0.41057, + "grad_norm": 1.2468445387390594, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 41057 + }, + { + "epoch": 0.41058, + "grad_norm": 1.441595512667646, + "learning_rate": 0.003, + "loss": 4.044, + "step": 41058 + }, + { + "epoch": 0.41059, + "grad_norm": 0.7244110367700357, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 41059 + }, + { + "epoch": 0.4106, + "grad_norm": 0.8424577609750289, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 41060 + }, + { + "epoch": 0.41061, + "grad_norm": 0.953818477831467, + "learning_rate": 0.003, + "loss": 4.032, + "step": 41061 + }, + { + "epoch": 0.41062, + "grad_norm": 1.1293735801018838, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 41062 + }, + { + "epoch": 0.41063, + "grad_norm": 1.408917450014383, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 41063 + }, + { + "epoch": 0.41064, + "grad_norm": 0.6986944201545965, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 41064 + }, + { + "epoch": 0.41065, + "grad_norm": 0.79952106580061, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 41065 + }, + { + "epoch": 0.41066, + "grad_norm": 1.0327991969819692, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 41066 + }, + { + "epoch": 0.41067, + "grad_norm": 1.3274614578461268, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 41067 + }, + { + "epoch": 0.41068, + "grad_norm": 1.0420972073804964, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 41068 + }, + { + "epoch": 0.41069, + "grad_norm": 1.2566024527168282, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 41069 + }, + { + "epoch": 0.4107, + "grad_norm": 0.8958440355203422, + "learning_rate": 0.003, + "loss": 4.018, + "step": 41070 + }, + { + "epoch": 0.41071, + "grad_norm": 1.1261333105203004, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 41071 + }, + { + "epoch": 0.41072, + "grad_norm": 0.9444204926518827, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 41072 + }, + { + "epoch": 0.41073, + "grad_norm": 0.931848831670046, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 41073 + }, + { + "epoch": 0.41074, + "grad_norm": 1.2033662474012286, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 41074 + }, + { + "epoch": 0.41075, + "grad_norm": 1.2924178604792107, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 41075 + }, + { + "epoch": 0.41076, + "grad_norm": 1.0330886544410958, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 41076 + }, + { + "epoch": 0.41077, + "grad_norm": 1.1750366596094806, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 41077 + }, + { + "epoch": 0.41078, + "grad_norm": 1.1388266879342188, + "learning_rate": 0.003, + "loss": 4.03, + "step": 41078 + }, + { + "epoch": 0.41079, + "grad_norm": 1.2063804099700979, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 41079 + }, + { + "epoch": 0.4108, + "grad_norm": 0.9942116211874804, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 41080 + }, + { + "epoch": 0.41081, + "grad_norm": 1.0656893733862065, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 41081 + }, + { + "epoch": 0.41082, + "grad_norm": 1.1536307538231387, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 41082 + }, + { + "epoch": 0.41083, + "grad_norm": 1.134009297565916, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 41083 + }, + { + "epoch": 0.41084, + "grad_norm": 1.0705541529417304, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 41084 + }, + { + "epoch": 0.41085, + "grad_norm": 0.9399653760948029, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 41085 + }, + { + "epoch": 0.41086, + "grad_norm": 0.9894329080029194, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 41086 + }, + { + "epoch": 0.41087, + "grad_norm": 1.1838558074636254, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 41087 + }, + { + "epoch": 0.41088, + "grad_norm": 1.062497287169947, + "learning_rate": 0.003, + "loss": 4.05, + "step": 41088 + }, + { + "epoch": 0.41089, + "grad_norm": 1.1552571352120466, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 41089 + }, + { + "epoch": 0.4109, + "grad_norm": 0.9009579574991687, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 41090 + }, + { + "epoch": 0.41091, + "grad_norm": 0.9575370356650822, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 41091 + }, + { + "epoch": 0.41092, + "grad_norm": 1.2049230555797985, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 41092 + }, + { + "epoch": 0.41093, + "grad_norm": 1.038004476860152, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 41093 + }, + { + "epoch": 0.41094, + "grad_norm": 1.1149770814477336, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 41094 + }, + { + "epoch": 0.41095, + "grad_norm": 0.9794241549323573, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 41095 + }, + { + "epoch": 0.41096, + "grad_norm": 1.1426894793327704, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 41096 + }, + { + "epoch": 0.41097, + "grad_norm": 1.1605196349809703, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 41097 + }, + { + "epoch": 0.41098, + "grad_norm": 1.5083414414867564, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 41098 + }, + { + "epoch": 0.41099, + "grad_norm": 0.6455967685493578, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 41099 + }, + { + "epoch": 0.411, + "grad_norm": 0.7525977521762529, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 41100 + }, + { + "epoch": 0.41101, + "grad_norm": 0.8478959481766727, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 41101 + }, + { + "epoch": 0.41102, + "grad_norm": 1.0980608355901618, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 41102 + }, + { + "epoch": 0.41103, + "grad_norm": 1.2927334260692118, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 41103 + }, + { + "epoch": 0.41104, + "grad_norm": 0.9920293927207804, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 41104 + }, + { + "epoch": 0.41105, + "grad_norm": 1.0470655508804234, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 41105 + }, + { + "epoch": 0.41106, + "grad_norm": 0.9761390195041431, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 41106 + }, + { + "epoch": 0.41107, + "grad_norm": 1.0982228327833514, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 41107 + }, + { + "epoch": 0.41108, + "grad_norm": 1.0085228620271005, + "learning_rate": 0.003, + "loss": 4.035, + "step": 41108 + }, + { + "epoch": 0.41109, + "grad_norm": 1.2677151044415775, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 41109 + }, + { + "epoch": 0.4111, + "grad_norm": 0.9145049001286114, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 41110 + }, + { + "epoch": 0.41111, + "grad_norm": 1.060161612608011, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 41111 + }, + { + "epoch": 0.41112, + "grad_norm": 1.1407542486375561, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 41112 + }, + { + "epoch": 0.41113, + "grad_norm": 1.4129552867146566, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 41113 + }, + { + "epoch": 0.41114, + "grad_norm": 0.9445141976755232, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 41114 + }, + { + "epoch": 0.41115, + "grad_norm": 1.1620982873356793, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 41115 + }, + { + "epoch": 0.41116, + "grad_norm": 1.1654296053279092, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 41116 + }, + { + "epoch": 0.41117, + "grad_norm": 1.3305801344690633, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 41117 + }, + { + "epoch": 0.41118, + "grad_norm": 0.9595847476814047, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 41118 + }, + { + "epoch": 0.41119, + "grad_norm": 1.0021522572824442, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 41119 + }, + { + "epoch": 0.4112, + "grad_norm": 1.5126182771210541, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 41120 + }, + { + "epoch": 0.41121, + "grad_norm": 1.1521917115523597, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 41121 + }, + { + "epoch": 0.41122, + "grad_norm": 0.9988043105295829, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 41122 + }, + { + "epoch": 0.41123, + "grad_norm": 0.9883680896419007, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 41123 + }, + { + "epoch": 0.41124, + "grad_norm": 1.2512459236731563, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 41124 + }, + { + "epoch": 0.41125, + "grad_norm": 0.9475212458153255, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 41125 + }, + { + "epoch": 0.41126, + "grad_norm": 1.0226990645286065, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 41126 + }, + { + "epoch": 0.41127, + "grad_norm": 1.0839500947527216, + "learning_rate": 0.003, + "loss": 4.059, + "step": 41127 + }, + { + "epoch": 0.41128, + "grad_norm": 1.0048987161011669, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 41128 + }, + { + "epoch": 0.41129, + "grad_norm": 0.9484901754141489, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 41129 + }, + { + "epoch": 0.4113, + "grad_norm": 0.9658828627797338, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 41130 + }, + { + "epoch": 0.41131, + "grad_norm": 1.1504203665251054, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 41131 + }, + { + "epoch": 0.41132, + "grad_norm": 0.9035617463075605, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 41132 + }, + { + "epoch": 0.41133, + "grad_norm": 0.9116596067104048, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 41133 + }, + { + "epoch": 0.41134, + "grad_norm": 1.3936696669194122, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 41134 + }, + { + "epoch": 0.41135, + "grad_norm": 1.2488022227892628, + "learning_rate": 0.003, + "loss": 4.037, + "step": 41135 + }, + { + "epoch": 0.41136, + "grad_norm": 0.9780466898558995, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 41136 + }, + { + "epoch": 0.41137, + "grad_norm": 1.0015586786100565, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 41137 + }, + { + "epoch": 0.41138, + "grad_norm": 1.1841276428149734, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 41138 + }, + { + "epoch": 0.41139, + "grad_norm": 1.1235864877970096, + "learning_rate": 0.003, + "loss": 4.024, + "step": 41139 + }, + { + "epoch": 0.4114, + "grad_norm": 1.145899966496784, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 41140 + }, + { + "epoch": 0.41141, + "grad_norm": 0.8876447508257422, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 41141 + }, + { + "epoch": 0.41142, + "grad_norm": 0.8724356197579298, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 41142 + }, + { + "epoch": 0.41143, + "grad_norm": 1.0685999617139232, + "learning_rate": 0.003, + "loss": 4.058, + "step": 41143 + }, + { + "epoch": 0.41144, + "grad_norm": 1.2046181146685633, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 41144 + }, + { + "epoch": 0.41145, + "grad_norm": 0.9781998471268812, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 41145 + }, + { + "epoch": 0.41146, + "grad_norm": 1.1249998348826535, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 41146 + }, + { + "epoch": 0.41147, + "grad_norm": 1.048419397523499, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 41147 + }, + { + "epoch": 0.41148, + "grad_norm": 1.1886168414425893, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 41148 + }, + { + "epoch": 0.41149, + "grad_norm": 0.9473476225579317, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 41149 + }, + { + "epoch": 0.4115, + "grad_norm": 0.9856239385414856, + "learning_rate": 0.003, + "loss": 4.044, + "step": 41150 + }, + { + "epoch": 0.41151, + "grad_norm": 1.320444907818893, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 41151 + }, + { + "epoch": 0.41152, + "grad_norm": 1.0060557856845327, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 41152 + }, + { + "epoch": 0.41153, + "grad_norm": 0.9291019990369731, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 41153 + }, + { + "epoch": 0.41154, + "grad_norm": 1.0105033245409605, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 41154 + }, + { + "epoch": 0.41155, + "grad_norm": 1.0876993432033124, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 41155 + }, + { + "epoch": 0.41156, + "grad_norm": 0.919036739717761, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 41156 + }, + { + "epoch": 0.41157, + "grad_norm": 1.0317076062894646, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 41157 + }, + { + "epoch": 0.41158, + "grad_norm": 1.220029352703334, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 41158 + }, + { + "epoch": 0.41159, + "grad_norm": 0.8740418492154208, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 41159 + }, + { + "epoch": 0.4116, + "grad_norm": 0.771865305691347, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 41160 + }, + { + "epoch": 0.41161, + "grad_norm": 0.8869279302651512, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 41161 + }, + { + "epoch": 0.41162, + "grad_norm": 1.1662271031869647, + "learning_rate": 0.003, + "loss": 4.059, + "step": 41162 + }, + { + "epoch": 0.41163, + "grad_norm": 0.9983186928146097, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 41163 + }, + { + "epoch": 0.41164, + "grad_norm": 1.3005731738101416, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 41164 + }, + { + "epoch": 0.41165, + "grad_norm": 0.9900146787924892, + "learning_rate": 0.003, + "loss": 4.037, + "step": 41165 + }, + { + "epoch": 0.41166, + "grad_norm": 1.0710595998572825, + "learning_rate": 0.003, + "loss": 4.046, + "step": 41166 + }, + { + "epoch": 0.41167, + "grad_norm": 1.1790249477985209, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 41167 + }, + { + "epoch": 0.41168, + "grad_norm": 1.1109206255214645, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 41168 + }, + { + "epoch": 0.41169, + "grad_norm": 1.032733951813728, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 41169 + }, + { + "epoch": 0.4117, + "grad_norm": 1.0978402518701489, + "learning_rate": 0.003, + "loss": 4.042, + "step": 41170 + }, + { + "epoch": 0.41171, + "grad_norm": 1.4894222423553647, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 41171 + }, + { + "epoch": 0.41172, + "grad_norm": 0.8554725791643636, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 41172 + }, + { + "epoch": 0.41173, + "grad_norm": 0.8807451755795628, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 41173 + }, + { + "epoch": 0.41174, + "grad_norm": 0.9767538453346423, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 41174 + }, + { + "epoch": 0.41175, + "grad_norm": 1.191965429698328, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 41175 + }, + { + "epoch": 0.41176, + "grad_norm": 1.2134742444287019, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 41176 + }, + { + "epoch": 0.41177, + "grad_norm": 1.0531983109369696, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 41177 + }, + { + "epoch": 0.41178, + "grad_norm": 1.093149600917377, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 41178 + }, + { + "epoch": 0.41179, + "grad_norm": 1.196160645897177, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 41179 + }, + { + "epoch": 0.4118, + "grad_norm": 1.0004627004669004, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 41180 + }, + { + "epoch": 0.41181, + "grad_norm": 1.2722665613985513, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 41181 + }, + { + "epoch": 0.41182, + "grad_norm": 0.9388806321555506, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 41182 + }, + { + "epoch": 0.41183, + "grad_norm": 1.0787102917875313, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 41183 + }, + { + "epoch": 0.41184, + "grad_norm": 1.2699739617724493, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 41184 + }, + { + "epoch": 0.41185, + "grad_norm": 0.8557865583192215, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 41185 + }, + { + "epoch": 0.41186, + "grad_norm": 0.9616905053237442, + "learning_rate": 0.003, + "loss": 4.039, + "step": 41186 + }, + { + "epoch": 0.41187, + "grad_norm": 1.2460440877096222, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 41187 + }, + { + "epoch": 0.41188, + "grad_norm": 1.1082193200916233, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 41188 + }, + { + "epoch": 0.41189, + "grad_norm": 0.9367709704237391, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 41189 + }, + { + "epoch": 0.4119, + "grad_norm": 0.9692145786642133, + "learning_rate": 0.003, + "loss": 4.017, + "step": 41190 + }, + { + "epoch": 0.41191, + "grad_norm": 1.0156258282923367, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 41191 + }, + { + "epoch": 0.41192, + "grad_norm": 0.9548067654957902, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 41192 + }, + { + "epoch": 0.41193, + "grad_norm": 1.0693969694914733, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 41193 + }, + { + "epoch": 0.41194, + "grad_norm": 1.0512911149089543, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 41194 + }, + { + "epoch": 0.41195, + "grad_norm": 1.1191406349134834, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 41195 + }, + { + "epoch": 0.41196, + "grad_norm": 0.9462848741609076, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 41196 + }, + { + "epoch": 0.41197, + "grad_norm": 0.954850281513718, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 41197 + }, + { + "epoch": 0.41198, + "grad_norm": 0.9855148398955298, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 41198 + }, + { + "epoch": 0.41199, + "grad_norm": 1.1627103302550756, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 41199 + }, + { + "epoch": 0.412, + "grad_norm": 1.233639649030151, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 41200 + }, + { + "epoch": 0.41201, + "grad_norm": 1.189604689173424, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 41201 + }, + { + "epoch": 0.41202, + "grad_norm": 1.0981548568299522, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 41202 + }, + { + "epoch": 0.41203, + "grad_norm": 1.2776902708506295, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 41203 + }, + { + "epoch": 0.41204, + "grad_norm": 0.9862944706100325, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 41204 + }, + { + "epoch": 0.41205, + "grad_norm": 0.995567442484338, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 41205 + }, + { + "epoch": 0.41206, + "grad_norm": 1.1869067894122314, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 41206 + }, + { + "epoch": 0.41207, + "grad_norm": 1.1305108357100266, + "learning_rate": 0.003, + "loss": 4.056, + "step": 41207 + }, + { + "epoch": 0.41208, + "grad_norm": 1.0284992928601977, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 41208 + }, + { + "epoch": 0.41209, + "grad_norm": 1.0624028771481517, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 41209 + }, + { + "epoch": 0.4121, + "grad_norm": 1.0622327483603438, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 41210 + }, + { + "epoch": 0.41211, + "grad_norm": 1.1781646813228548, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 41211 + }, + { + "epoch": 0.41212, + "grad_norm": 1.0544055269417427, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 41212 + }, + { + "epoch": 0.41213, + "grad_norm": 0.9332164380661774, + "learning_rate": 0.003, + "loss": 4.008, + "step": 41213 + }, + { + "epoch": 0.41214, + "grad_norm": 0.8918612091715094, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 41214 + }, + { + "epoch": 0.41215, + "grad_norm": 0.8178961198938791, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 41215 + }, + { + "epoch": 0.41216, + "grad_norm": 1.0325692893609484, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 41216 + }, + { + "epoch": 0.41217, + "grad_norm": 1.0941989022486767, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 41217 + }, + { + "epoch": 0.41218, + "grad_norm": 1.074064071393525, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 41218 + }, + { + "epoch": 0.41219, + "grad_norm": 1.2597068810533427, + "learning_rate": 0.003, + "loss": 4.073, + "step": 41219 + }, + { + "epoch": 0.4122, + "grad_norm": 1.213746142235882, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 41220 + }, + { + "epoch": 0.41221, + "grad_norm": 1.0116008598167736, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 41221 + }, + { + "epoch": 0.41222, + "grad_norm": 1.0350465709566754, + "learning_rate": 0.003, + "loss": 4.028, + "step": 41222 + }, + { + "epoch": 0.41223, + "grad_norm": 1.050185276723443, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 41223 + }, + { + "epoch": 0.41224, + "grad_norm": 1.2562754804704128, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 41224 + }, + { + "epoch": 0.41225, + "grad_norm": 1.144575758535222, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 41225 + }, + { + "epoch": 0.41226, + "grad_norm": 1.0997654693731203, + "learning_rate": 0.003, + "loss": 4.031, + "step": 41226 + }, + { + "epoch": 0.41227, + "grad_norm": 0.9098066771055909, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 41227 + }, + { + "epoch": 0.41228, + "grad_norm": 1.0326009918738324, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 41228 + }, + { + "epoch": 0.41229, + "grad_norm": 1.0569420081844865, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 41229 + }, + { + "epoch": 0.4123, + "grad_norm": 1.1541018543714192, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 41230 + }, + { + "epoch": 0.41231, + "grad_norm": 1.1205320367581864, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 41231 + }, + { + "epoch": 0.41232, + "grad_norm": 0.921773746817902, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 41232 + }, + { + "epoch": 0.41233, + "grad_norm": 1.1160582225324178, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 41233 + }, + { + "epoch": 0.41234, + "grad_norm": 1.279045570840561, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 41234 + }, + { + "epoch": 0.41235, + "grad_norm": 1.0070474830450968, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 41235 + }, + { + "epoch": 0.41236, + "grad_norm": 1.1127619035726122, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 41236 + }, + { + "epoch": 0.41237, + "grad_norm": 1.2128857509160407, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 41237 + }, + { + "epoch": 0.41238, + "grad_norm": 0.9110441320263384, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 41238 + }, + { + "epoch": 0.41239, + "grad_norm": 0.9489035656089011, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 41239 + }, + { + "epoch": 0.4124, + "grad_norm": 1.1775205258799895, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 41240 + }, + { + "epoch": 0.41241, + "grad_norm": 0.8805286735857397, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 41241 + }, + { + "epoch": 0.41242, + "grad_norm": 1.0595205505131091, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 41242 + }, + { + "epoch": 0.41243, + "grad_norm": 1.358132731175539, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 41243 + }, + { + "epoch": 0.41244, + "grad_norm": 0.9966376294364961, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 41244 + }, + { + "epoch": 0.41245, + "grad_norm": 1.3135708840841478, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 41245 + }, + { + "epoch": 0.41246, + "grad_norm": 0.9423560267491656, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 41246 + }, + { + "epoch": 0.41247, + "grad_norm": 0.8429173159670108, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 41247 + }, + { + "epoch": 0.41248, + "grad_norm": 1.0045916881927934, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 41248 + }, + { + "epoch": 0.41249, + "grad_norm": 1.410108476620839, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 41249 + }, + { + "epoch": 0.4125, + "grad_norm": 1.0470272784912769, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 41250 + }, + { + "epoch": 0.41251, + "grad_norm": 0.9475289920819301, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 41251 + }, + { + "epoch": 0.41252, + "grad_norm": 1.0807192661062228, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 41252 + }, + { + "epoch": 0.41253, + "grad_norm": 1.1255845698809899, + "learning_rate": 0.003, + "loss": 4.028, + "step": 41253 + }, + { + "epoch": 0.41254, + "grad_norm": 0.9716845431421445, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 41254 + }, + { + "epoch": 0.41255, + "grad_norm": 0.9292352536359207, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 41255 + }, + { + "epoch": 0.41256, + "grad_norm": 1.0211715007517157, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 41256 + }, + { + "epoch": 0.41257, + "grad_norm": 1.13383302484551, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 41257 + }, + { + "epoch": 0.41258, + "grad_norm": 1.16342686379131, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 41258 + }, + { + "epoch": 0.41259, + "grad_norm": 1.0905136998314602, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 41259 + }, + { + "epoch": 0.4126, + "grad_norm": 1.3055433768236508, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 41260 + }, + { + "epoch": 0.41261, + "grad_norm": 0.8753543698364668, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 41261 + }, + { + "epoch": 0.41262, + "grad_norm": 1.0623340161758403, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 41262 + }, + { + "epoch": 0.41263, + "grad_norm": 1.3531681514880225, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 41263 + }, + { + "epoch": 0.41264, + "grad_norm": 0.8437637349960507, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 41264 + }, + { + "epoch": 0.41265, + "grad_norm": 0.7818433203610888, + "learning_rate": 0.003, + "loss": 4.038, + "step": 41265 + }, + { + "epoch": 0.41266, + "grad_norm": 0.8518759720366358, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 41266 + }, + { + "epoch": 0.41267, + "grad_norm": 1.1571439214637056, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 41267 + }, + { + "epoch": 0.41268, + "grad_norm": 1.2813029624945838, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 41268 + }, + { + "epoch": 0.41269, + "grad_norm": 0.9794356943598219, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 41269 + }, + { + "epoch": 0.4127, + "grad_norm": 1.002690593544184, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 41270 + }, + { + "epoch": 0.41271, + "grad_norm": 1.0169577879368785, + "learning_rate": 0.003, + "loss": 4.049, + "step": 41271 + }, + { + "epoch": 0.41272, + "grad_norm": 1.136043742497766, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 41272 + }, + { + "epoch": 0.41273, + "grad_norm": 0.9742747305319283, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 41273 + }, + { + "epoch": 0.41274, + "grad_norm": 1.0915901360238387, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 41274 + }, + { + "epoch": 0.41275, + "grad_norm": 0.827175609240378, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 41275 + }, + { + "epoch": 0.41276, + "grad_norm": 0.8969957820517913, + "learning_rate": 0.003, + "loss": 4.06, + "step": 41276 + }, + { + "epoch": 0.41277, + "grad_norm": 0.9978848921827517, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 41277 + }, + { + "epoch": 0.41278, + "grad_norm": 1.33111290660784, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 41278 + }, + { + "epoch": 0.41279, + "grad_norm": 0.8236080119160918, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 41279 + }, + { + "epoch": 0.4128, + "grad_norm": 0.9391497826387315, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 41280 + }, + { + "epoch": 0.41281, + "grad_norm": 1.1222121731522094, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 41281 + }, + { + "epoch": 0.41282, + "grad_norm": 1.1144646476124194, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 41282 + }, + { + "epoch": 0.41283, + "grad_norm": 1.2490921817344165, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 41283 + }, + { + "epoch": 0.41284, + "grad_norm": 1.1290657606083823, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 41284 + }, + { + "epoch": 0.41285, + "grad_norm": 1.2551147692091584, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 41285 + }, + { + "epoch": 0.41286, + "grad_norm": 1.0487102400192192, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 41286 + }, + { + "epoch": 0.41287, + "grad_norm": 0.92359132120383, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 41287 + }, + { + "epoch": 0.41288, + "grad_norm": 1.3179809376375287, + "learning_rate": 0.003, + "loss": 4.081, + "step": 41288 + }, + { + "epoch": 0.41289, + "grad_norm": 0.8562684525530531, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 41289 + }, + { + "epoch": 0.4129, + "grad_norm": 1.0152055568331264, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 41290 + }, + { + "epoch": 0.41291, + "grad_norm": 1.4634206130427594, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 41291 + }, + { + "epoch": 0.41292, + "grad_norm": 0.8792447344947575, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 41292 + }, + { + "epoch": 0.41293, + "grad_norm": 0.8882473804313002, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 41293 + }, + { + "epoch": 0.41294, + "grad_norm": 1.1738838650760206, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 41294 + }, + { + "epoch": 0.41295, + "grad_norm": 1.0175750016317027, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 41295 + }, + { + "epoch": 0.41296, + "grad_norm": 1.0562480288805425, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 41296 + }, + { + "epoch": 0.41297, + "grad_norm": 1.2572875368930465, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 41297 + }, + { + "epoch": 0.41298, + "grad_norm": 1.044624424606174, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 41298 + }, + { + "epoch": 0.41299, + "grad_norm": 1.098757063762601, + "learning_rate": 0.003, + "loss": 4.03, + "step": 41299 + }, + { + "epoch": 0.413, + "grad_norm": 1.0146223191267314, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 41300 + }, + { + "epoch": 0.41301, + "grad_norm": 1.0021259676766743, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 41301 + }, + { + "epoch": 0.41302, + "grad_norm": 0.9986770077963099, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 41302 + }, + { + "epoch": 0.41303, + "grad_norm": 1.09731822349368, + "learning_rate": 0.003, + "loss": 4.034, + "step": 41303 + }, + { + "epoch": 0.41304, + "grad_norm": 1.15147833111571, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 41304 + }, + { + "epoch": 0.41305, + "grad_norm": 1.2478484405340913, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 41305 + }, + { + "epoch": 0.41306, + "grad_norm": 0.8602433574294321, + "learning_rate": 0.003, + "loss": 4.043, + "step": 41306 + }, + { + "epoch": 0.41307, + "grad_norm": 1.195436910414514, + "learning_rate": 0.003, + "loss": 4.058, + "step": 41307 + }, + { + "epoch": 0.41308, + "grad_norm": 1.3843840176357876, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 41308 + }, + { + "epoch": 0.41309, + "grad_norm": 0.866943357737719, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 41309 + }, + { + "epoch": 0.4131, + "grad_norm": 0.9033979031906473, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 41310 + }, + { + "epoch": 0.41311, + "grad_norm": 1.0741203820668388, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 41311 + }, + { + "epoch": 0.41312, + "grad_norm": 1.0723597646474554, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 41312 + }, + { + "epoch": 0.41313, + "grad_norm": 1.0786939932157966, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 41313 + }, + { + "epoch": 0.41314, + "grad_norm": 1.0972553486063394, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 41314 + }, + { + "epoch": 0.41315, + "grad_norm": 1.178448167288582, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 41315 + }, + { + "epoch": 0.41316, + "grad_norm": 0.9748694035914419, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 41316 + }, + { + "epoch": 0.41317, + "grad_norm": 1.0947285607323631, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 41317 + }, + { + "epoch": 0.41318, + "grad_norm": 0.9984769110788237, + "learning_rate": 0.003, + "loss": 4.018, + "step": 41318 + }, + { + "epoch": 0.41319, + "grad_norm": 1.3615370645356346, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 41319 + }, + { + "epoch": 0.4132, + "grad_norm": 0.7974679369706394, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 41320 + }, + { + "epoch": 0.41321, + "grad_norm": 0.7666945666187379, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 41321 + }, + { + "epoch": 0.41322, + "grad_norm": 0.9426006868638915, + "learning_rate": 0.003, + "loss": 4.042, + "step": 41322 + }, + { + "epoch": 0.41323, + "grad_norm": 1.3316673237532684, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 41323 + }, + { + "epoch": 0.41324, + "grad_norm": 0.8857340871685515, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 41324 + }, + { + "epoch": 0.41325, + "grad_norm": 1.0479899229397012, + "learning_rate": 0.003, + "loss": 4.019, + "step": 41325 + }, + { + "epoch": 0.41326, + "grad_norm": 1.7341338903946497, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 41326 + }, + { + "epoch": 0.41327, + "grad_norm": 0.8017928292074618, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 41327 + }, + { + "epoch": 0.41328, + "grad_norm": 0.8279986966377967, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 41328 + }, + { + "epoch": 0.41329, + "grad_norm": 1.003227524639877, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 41329 + }, + { + "epoch": 0.4133, + "grad_norm": 1.2327980489657313, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 41330 + }, + { + "epoch": 0.41331, + "grad_norm": 1.1467873089587974, + "learning_rate": 0.003, + "loss": 4.027, + "step": 41331 + }, + { + "epoch": 0.41332, + "grad_norm": 1.1820043490229053, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 41332 + }, + { + "epoch": 0.41333, + "grad_norm": 1.0134577779232135, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 41333 + }, + { + "epoch": 0.41334, + "grad_norm": 1.0340233044614766, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 41334 + }, + { + "epoch": 0.41335, + "grad_norm": 1.3102216762657581, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 41335 + }, + { + "epoch": 0.41336, + "grad_norm": 0.8872250110140796, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 41336 + }, + { + "epoch": 0.41337, + "grad_norm": 0.9326665862725834, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 41337 + }, + { + "epoch": 0.41338, + "grad_norm": 1.0664178706706156, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 41338 + }, + { + "epoch": 0.41339, + "grad_norm": 1.0141996116055694, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 41339 + }, + { + "epoch": 0.4134, + "grad_norm": 1.1579841835485754, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 41340 + }, + { + "epoch": 0.41341, + "grad_norm": 1.0568655687652486, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 41341 + }, + { + "epoch": 0.41342, + "grad_norm": 1.0729782252623885, + "learning_rate": 0.003, + "loss": 4.008, + "step": 41342 + }, + { + "epoch": 0.41343, + "grad_norm": 1.028098735798463, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 41343 + }, + { + "epoch": 0.41344, + "grad_norm": 1.0425923191224513, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 41344 + }, + { + "epoch": 0.41345, + "grad_norm": 1.0370218944555192, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 41345 + }, + { + "epoch": 0.41346, + "grad_norm": 1.1179656531691786, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 41346 + }, + { + "epoch": 0.41347, + "grad_norm": 1.0953342098315573, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 41347 + }, + { + "epoch": 0.41348, + "grad_norm": 1.1304597926419173, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 41348 + }, + { + "epoch": 0.41349, + "grad_norm": 1.0855564100521056, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 41349 + }, + { + "epoch": 0.4135, + "grad_norm": 1.253103997290262, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 41350 + }, + { + "epoch": 0.41351, + "grad_norm": 0.795426646132106, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 41351 + }, + { + "epoch": 0.41352, + "grad_norm": 0.9214850428449147, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 41352 + }, + { + "epoch": 0.41353, + "grad_norm": 1.127401559856835, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 41353 + }, + { + "epoch": 0.41354, + "grad_norm": 1.1199985811582118, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 41354 + }, + { + "epoch": 0.41355, + "grad_norm": 1.0994197039749858, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 41355 + }, + { + "epoch": 0.41356, + "grad_norm": 1.061446420218049, + "learning_rate": 0.003, + "loss": 4.042, + "step": 41356 + }, + { + "epoch": 0.41357, + "grad_norm": 1.2946170405258328, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 41357 + }, + { + "epoch": 0.41358, + "grad_norm": 0.964338055580123, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 41358 + }, + { + "epoch": 0.41359, + "grad_norm": 1.1520548912794808, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 41359 + }, + { + "epoch": 0.4136, + "grad_norm": 1.1433328951560662, + "learning_rate": 0.003, + "loss": 4.035, + "step": 41360 + }, + { + "epoch": 0.41361, + "grad_norm": 0.9032418350481676, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 41361 + }, + { + "epoch": 0.41362, + "grad_norm": 1.0216933734695945, + "learning_rate": 0.003, + "loss": 4.059, + "step": 41362 + }, + { + "epoch": 0.41363, + "grad_norm": 1.0386610165221921, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 41363 + }, + { + "epoch": 0.41364, + "grad_norm": 1.073877285438678, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 41364 + }, + { + "epoch": 0.41365, + "grad_norm": 1.181840668904054, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 41365 + }, + { + "epoch": 0.41366, + "grad_norm": 1.0192651908207775, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 41366 + }, + { + "epoch": 0.41367, + "grad_norm": 1.3850760302674956, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 41367 + }, + { + "epoch": 0.41368, + "grad_norm": 0.9906275611054097, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 41368 + }, + { + "epoch": 0.41369, + "grad_norm": 0.9252162967998148, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 41369 + }, + { + "epoch": 0.4137, + "grad_norm": 0.9807241239374068, + "learning_rate": 0.003, + "loss": 4.031, + "step": 41370 + }, + { + "epoch": 0.41371, + "grad_norm": 1.175566606391698, + "learning_rate": 0.003, + "loss": 4.023, + "step": 41371 + }, + { + "epoch": 0.41372, + "grad_norm": 1.299701493926495, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 41372 + }, + { + "epoch": 0.41373, + "grad_norm": 0.910055568396593, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 41373 + }, + { + "epoch": 0.41374, + "grad_norm": 0.987932013903189, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 41374 + }, + { + "epoch": 0.41375, + "grad_norm": 1.2877551478636402, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 41375 + }, + { + "epoch": 0.41376, + "grad_norm": 0.8403144672852594, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 41376 + }, + { + "epoch": 0.41377, + "grad_norm": 0.7917130171988989, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 41377 + }, + { + "epoch": 0.41378, + "grad_norm": 0.7818311180390867, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 41378 + }, + { + "epoch": 0.41379, + "grad_norm": 0.731089897697726, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 41379 + }, + { + "epoch": 0.4138, + "grad_norm": 0.7089102164016767, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 41380 + }, + { + "epoch": 0.41381, + "grad_norm": 0.8653068629709149, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 41381 + }, + { + "epoch": 0.41382, + "grad_norm": 1.004428683951494, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 41382 + }, + { + "epoch": 0.41383, + "grad_norm": 1.0736960378656315, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 41383 + }, + { + "epoch": 0.41384, + "grad_norm": 1.043611981074796, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 41384 + }, + { + "epoch": 0.41385, + "grad_norm": 1.2495002953618897, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 41385 + }, + { + "epoch": 0.41386, + "grad_norm": 1.0620116602471144, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 41386 + }, + { + "epoch": 0.41387, + "grad_norm": 0.9503770167003281, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 41387 + }, + { + "epoch": 0.41388, + "grad_norm": 1.0302551409678289, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 41388 + }, + { + "epoch": 0.41389, + "grad_norm": 1.5019212461240514, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 41389 + }, + { + "epoch": 0.4139, + "grad_norm": 0.7240836693265655, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 41390 + }, + { + "epoch": 0.41391, + "grad_norm": 0.851830525651554, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 41391 + }, + { + "epoch": 0.41392, + "grad_norm": 1.225033527890849, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 41392 + }, + { + "epoch": 0.41393, + "grad_norm": 1.0078572142716475, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 41393 + }, + { + "epoch": 0.41394, + "grad_norm": 1.3545989330849462, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 41394 + }, + { + "epoch": 0.41395, + "grad_norm": 0.8937853338199722, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 41395 + }, + { + "epoch": 0.41396, + "grad_norm": 0.8552920044564631, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 41396 + }, + { + "epoch": 0.41397, + "grad_norm": 1.1114143570796589, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 41397 + }, + { + "epoch": 0.41398, + "grad_norm": 1.0069074589902094, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 41398 + }, + { + "epoch": 0.41399, + "grad_norm": 1.2138508239408994, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 41399 + }, + { + "epoch": 0.414, + "grad_norm": 0.991811898589862, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 41400 + }, + { + "epoch": 0.41401, + "grad_norm": 1.2112187581778324, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 41401 + }, + { + "epoch": 0.41402, + "grad_norm": 0.9244223313603315, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 41402 + }, + { + "epoch": 0.41403, + "grad_norm": 1.052360616085121, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 41403 + }, + { + "epoch": 0.41404, + "grad_norm": 1.4290871334251016, + "learning_rate": 0.003, + "loss": 4.059, + "step": 41404 + }, + { + "epoch": 0.41405, + "grad_norm": 0.816098313410408, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 41405 + }, + { + "epoch": 0.41406, + "grad_norm": 0.9365838088118826, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 41406 + }, + { + "epoch": 0.41407, + "grad_norm": 1.5325498046708312, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 41407 + }, + { + "epoch": 0.41408, + "grad_norm": 1.270541266236754, + "learning_rate": 0.003, + "loss": 4.036, + "step": 41408 + }, + { + "epoch": 0.41409, + "grad_norm": 0.7922111728358507, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 41409 + }, + { + "epoch": 0.4141, + "grad_norm": 0.8081018769370574, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 41410 + }, + { + "epoch": 0.41411, + "grad_norm": 0.8975917645454057, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 41411 + }, + { + "epoch": 0.41412, + "grad_norm": 1.0560806290222675, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 41412 + }, + { + "epoch": 0.41413, + "grad_norm": 0.9802660498789245, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 41413 + }, + { + "epoch": 0.41414, + "grad_norm": 1.2555109042068537, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 41414 + }, + { + "epoch": 0.41415, + "grad_norm": 1.0414161425009327, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 41415 + }, + { + "epoch": 0.41416, + "grad_norm": 1.2000635117793215, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 41416 + }, + { + "epoch": 0.41417, + "grad_norm": 1.129303859559035, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 41417 + }, + { + "epoch": 0.41418, + "grad_norm": 0.9783950172301298, + "learning_rate": 0.003, + "loss": 4.047, + "step": 41418 + }, + { + "epoch": 0.41419, + "grad_norm": 1.0002230826986662, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 41419 + }, + { + "epoch": 0.4142, + "grad_norm": 1.0912714874434692, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 41420 + }, + { + "epoch": 0.41421, + "grad_norm": 1.0322893360293806, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 41421 + }, + { + "epoch": 0.41422, + "grad_norm": 1.2303003747279542, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 41422 + }, + { + "epoch": 0.41423, + "grad_norm": 1.0360549132270012, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 41423 + }, + { + "epoch": 0.41424, + "grad_norm": 1.1354407592833264, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 41424 + }, + { + "epoch": 0.41425, + "grad_norm": 1.0585995623721927, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 41425 + }, + { + "epoch": 0.41426, + "grad_norm": 0.9895705113201729, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 41426 + }, + { + "epoch": 0.41427, + "grad_norm": 1.0737440408932195, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 41427 + }, + { + "epoch": 0.41428, + "grad_norm": 1.2709555664991872, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 41428 + }, + { + "epoch": 0.41429, + "grad_norm": 1.2315907672590976, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 41429 + }, + { + "epoch": 0.4143, + "grad_norm": 1.0290184295461031, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 41430 + }, + { + "epoch": 0.41431, + "grad_norm": 1.2717946879606181, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 41431 + }, + { + "epoch": 0.41432, + "grad_norm": 1.0409840695775383, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 41432 + }, + { + "epoch": 0.41433, + "grad_norm": 1.0177136909363151, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 41433 + }, + { + "epoch": 0.41434, + "grad_norm": 0.995403316273746, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 41434 + }, + { + "epoch": 0.41435, + "grad_norm": 1.00325931021238, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 41435 + }, + { + "epoch": 0.41436, + "grad_norm": 1.0594769586344395, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 41436 + }, + { + "epoch": 0.41437, + "grad_norm": 1.1368677512739638, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 41437 + }, + { + "epoch": 0.41438, + "grad_norm": 1.1807597682824542, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 41438 + }, + { + "epoch": 0.41439, + "grad_norm": 1.009275066148526, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 41439 + }, + { + "epoch": 0.4144, + "grad_norm": 1.1574640557400906, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 41440 + }, + { + "epoch": 0.41441, + "grad_norm": 0.9757694660790911, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 41441 + }, + { + "epoch": 0.41442, + "grad_norm": 1.0605766321867687, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 41442 + }, + { + "epoch": 0.41443, + "grad_norm": 1.2934172416787004, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 41443 + }, + { + "epoch": 0.41444, + "grad_norm": 1.0133422684506292, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 41444 + }, + { + "epoch": 0.41445, + "grad_norm": 0.9165266228121438, + "learning_rate": 0.003, + "loss": 4.002, + "step": 41445 + }, + { + "epoch": 0.41446, + "grad_norm": 1.151922670120972, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 41446 + }, + { + "epoch": 0.41447, + "grad_norm": 1.2545458209864748, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 41447 + }, + { + "epoch": 0.41448, + "grad_norm": 1.1357628339384185, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 41448 + }, + { + "epoch": 0.41449, + "grad_norm": 0.782810147629317, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 41449 + }, + { + "epoch": 0.4145, + "grad_norm": 0.7746499913036515, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 41450 + }, + { + "epoch": 0.41451, + "grad_norm": 0.8711303349945595, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 41451 + }, + { + "epoch": 0.41452, + "grad_norm": 0.8893496520503494, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 41452 + }, + { + "epoch": 0.41453, + "grad_norm": 1.0553579495046375, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 41453 + }, + { + "epoch": 0.41454, + "grad_norm": 1.1910721379417624, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 41454 + }, + { + "epoch": 0.41455, + "grad_norm": 0.9839691616159502, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 41455 + }, + { + "epoch": 0.41456, + "grad_norm": 1.0714483505235652, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 41456 + }, + { + "epoch": 0.41457, + "grad_norm": 1.0421759614665829, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 41457 + }, + { + "epoch": 0.41458, + "grad_norm": 0.9048140535446213, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 41458 + }, + { + "epoch": 0.41459, + "grad_norm": 0.9117809600953043, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 41459 + }, + { + "epoch": 0.4146, + "grad_norm": 1.1844307417387998, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 41460 + }, + { + "epoch": 0.41461, + "grad_norm": 1.322924941423145, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 41461 + }, + { + "epoch": 0.41462, + "grad_norm": 0.9883028610550002, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 41462 + }, + { + "epoch": 0.41463, + "grad_norm": 1.143004954642728, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 41463 + }, + { + "epoch": 0.41464, + "grad_norm": 1.0867432878856358, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 41464 + }, + { + "epoch": 0.41465, + "grad_norm": 1.0415431550637082, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 41465 + }, + { + "epoch": 0.41466, + "grad_norm": 1.197128294934922, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 41466 + }, + { + "epoch": 0.41467, + "grad_norm": 1.210438705717191, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 41467 + }, + { + "epoch": 0.41468, + "grad_norm": 0.9914158377481633, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 41468 + }, + { + "epoch": 0.41469, + "grad_norm": 1.2129792910557855, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 41469 + }, + { + "epoch": 0.4147, + "grad_norm": 1.1795823355511785, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 41470 + }, + { + "epoch": 0.41471, + "grad_norm": 0.9037316718826216, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 41471 + }, + { + "epoch": 0.41472, + "grad_norm": 0.9779573831029544, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 41472 + }, + { + "epoch": 0.41473, + "grad_norm": 1.0894546064467832, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 41473 + }, + { + "epoch": 0.41474, + "grad_norm": 1.05459173343724, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 41474 + }, + { + "epoch": 0.41475, + "grad_norm": 1.140183298777585, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 41475 + }, + { + "epoch": 0.41476, + "grad_norm": 1.1085431883281456, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 41476 + }, + { + "epoch": 0.41477, + "grad_norm": 1.026459530900187, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 41477 + }, + { + "epoch": 0.41478, + "grad_norm": 1.2904849974911679, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 41478 + }, + { + "epoch": 0.41479, + "grad_norm": 1.0637714497641482, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 41479 + }, + { + "epoch": 0.4148, + "grad_norm": 1.1334852504816342, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 41480 + }, + { + "epoch": 0.41481, + "grad_norm": 0.9342071418580161, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 41481 + }, + { + "epoch": 0.41482, + "grad_norm": 0.9958825089420165, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 41482 + }, + { + "epoch": 0.41483, + "grad_norm": 1.1690054531379728, + "learning_rate": 0.003, + "loss": 4.029, + "step": 41483 + }, + { + "epoch": 0.41484, + "grad_norm": 1.0770479656364114, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 41484 + }, + { + "epoch": 0.41485, + "grad_norm": 0.9731336576306822, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 41485 + }, + { + "epoch": 0.41486, + "grad_norm": 1.0144996546332292, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 41486 + }, + { + "epoch": 0.41487, + "grad_norm": 1.1415677045170496, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 41487 + }, + { + "epoch": 0.41488, + "grad_norm": 1.1112232773138457, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 41488 + }, + { + "epoch": 0.41489, + "grad_norm": 1.1704015879392566, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 41489 + }, + { + "epoch": 0.4149, + "grad_norm": 0.792869123917844, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 41490 + }, + { + "epoch": 0.41491, + "grad_norm": 0.80069815591956, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 41491 + }, + { + "epoch": 0.41492, + "grad_norm": 1.04373521381403, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 41492 + }, + { + "epoch": 0.41493, + "grad_norm": 0.9912748323663605, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 41493 + }, + { + "epoch": 0.41494, + "grad_norm": 1.209018199661853, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 41494 + }, + { + "epoch": 0.41495, + "grad_norm": 1.1603551334132791, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 41495 + }, + { + "epoch": 0.41496, + "grad_norm": 1.1972371909182156, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 41496 + }, + { + "epoch": 0.41497, + "grad_norm": 0.8293054751436001, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 41497 + }, + { + "epoch": 0.41498, + "grad_norm": 1.0512845585279003, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 41498 + }, + { + "epoch": 0.41499, + "grad_norm": 1.0944581083868563, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 41499 + }, + { + "epoch": 0.415, + "grad_norm": 0.9274819937867548, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 41500 + }, + { + "epoch": 0.41501, + "grad_norm": 1.099808792186417, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 41501 + }, + { + "epoch": 0.41502, + "grad_norm": 1.2162384742357926, + "learning_rate": 0.003, + "loss": 4.036, + "step": 41502 + }, + { + "epoch": 0.41503, + "grad_norm": 1.1469710585343318, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 41503 + }, + { + "epoch": 0.41504, + "grad_norm": 1.0262498862800091, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 41504 + }, + { + "epoch": 0.41505, + "grad_norm": 1.0847309926529969, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 41505 + }, + { + "epoch": 0.41506, + "grad_norm": 1.2240503317339109, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 41506 + }, + { + "epoch": 0.41507, + "grad_norm": 1.2081970082896891, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 41507 + }, + { + "epoch": 0.41508, + "grad_norm": 1.0613051396137567, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 41508 + }, + { + "epoch": 0.41509, + "grad_norm": 0.9676213813092062, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 41509 + }, + { + "epoch": 0.4151, + "grad_norm": 1.0953983992393743, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 41510 + }, + { + "epoch": 0.41511, + "grad_norm": 0.9665773265592925, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 41511 + }, + { + "epoch": 0.41512, + "grad_norm": 1.1531570985332293, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 41512 + }, + { + "epoch": 0.41513, + "grad_norm": 1.2555510919901292, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 41513 + }, + { + "epoch": 0.41514, + "grad_norm": 1.1377621207212565, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 41514 + }, + { + "epoch": 0.41515, + "grad_norm": 0.9130653915766854, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 41515 + }, + { + "epoch": 0.41516, + "grad_norm": 0.9766564540066269, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 41516 + }, + { + "epoch": 0.41517, + "grad_norm": 1.2645956802901555, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 41517 + }, + { + "epoch": 0.41518, + "grad_norm": 0.998164195327988, + "learning_rate": 0.003, + "loss": 4.025, + "step": 41518 + }, + { + "epoch": 0.41519, + "grad_norm": 1.1423643866940796, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 41519 + }, + { + "epoch": 0.4152, + "grad_norm": 0.9618109365168772, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 41520 + }, + { + "epoch": 0.41521, + "grad_norm": 1.019662693473228, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 41521 + }, + { + "epoch": 0.41522, + "grad_norm": 1.311077247139598, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 41522 + }, + { + "epoch": 0.41523, + "grad_norm": 1.0318497188203275, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 41523 + }, + { + "epoch": 0.41524, + "grad_norm": 1.2325517842001248, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 41524 + }, + { + "epoch": 0.41525, + "grad_norm": 0.818909882162279, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 41525 + }, + { + "epoch": 0.41526, + "grad_norm": 0.9056109895417479, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 41526 + }, + { + "epoch": 0.41527, + "grad_norm": 0.9942309271037907, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 41527 + }, + { + "epoch": 0.41528, + "grad_norm": 1.1877766565594499, + "learning_rate": 0.003, + "loss": 4.046, + "step": 41528 + }, + { + "epoch": 0.41529, + "grad_norm": 0.8044307877074001, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 41529 + }, + { + "epoch": 0.4153, + "grad_norm": 0.865599519937339, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 41530 + }, + { + "epoch": 0.41531, + "grad_norm": 1.0935766106990437, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 41531 + }, + { + "epoch": 0.41532, + "grad_norm": 1.2057094256821073, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 41532 + }, + { + "epoch": 0.41533, + "grad_norm": 0.8781291272238684, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 41533 + }, + { + "epoch": 0.41534, + "grad_norm": 0.8117177139066871, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 41534 + }, + { + "epoch": 0.41535, + "grad_norm": 0.8503733933362959, + "learning_rate": 0.003, + "loss": 4.054, + "step": 41535 + }, + { + "epoch": 0.41536, + "grad_norm": 1.0047017888739835, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 41536 + }, + { + "epoch": 0.41537, + "grad_norm": 1.170457857156804, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 41537 + }, + { + "epoch": 0.41538, + "grad_norm": 1.2445082839084511, + "learning_rate": 0.003, + "loss": 4.031, + "step": 41538 + }, + { + "epoch": 0.41539, + "grad_norm": 1.1913413938895532, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 41539 + }, + { + "epoch": 0.4154, + "grad_norm": 1.2735366989257144, + "learning_rate": 0.003, + "loss": 4.003, + "step": 41540 + }, + { + "epoch": 0.41541, + "grad_norm": 0.9185696979291382, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 41541 + }, + { + "epoch": 0.41542, + "grad_norm": 0.957240230061575, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 41542 + }, + { + "epoch": 0.41543, + "grad_norm": 1.3189254082016884, + "learning_rate": 0.003, + "loss": 4.04, + "step": 41543 + }, + { + "epoch": 0.41544, + "grad_norm": 1.1215297332901089, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 41544 + }, + { + "epoch": 0.41545, + "grad_norm": 1.0522589665243596, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 41545 + }, + { + "epoch": 0.41546, + "grad_norm": 1.1126784845165631, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 41546 + }, + { + "epoch": 0.41547, + "grad_norm": 0.9986257746678328, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 41547 + }, + { + "epoch": 0.41548, + "grad_norm": 1.0417160069431393, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 41548 + }, + { + "epoch": 0.41549, + "grad_norm": 0.9968592909271236, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 41549 + }, + { + "epoch": 0.4155, + "grad_norm": 1.1522647122458392, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 41550 + }, + { + "epoch": 0.41551, + "grad_norm": 0.9582541342929843, + "learning_rate": 0.003, + "loss": 4.039, + "step": 41551 + }, + { + "epoch": 0.41552, + "grad_norm": 0.8611904197764454, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 41552 + }, + { + "epoch": 0.41553, + "grad_norm": 0.8065502847631325, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 41553 + }, + { + "epoch": 0.41554, + "grad_norm": 1.075685042085982, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 41554 + }, + { + "epoch": 0.41555, + "grad_norm": 1.4289545606545844, + "learning_rate": 0.003, + "loss": 4.013, + "step": 41555 + }, + { + "epoch": 0.41556, + "grad_norm": 1.011361497662174, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 41556 + }, + { + "epoch": 0.41557, + "grad_norm": 1.301001774359851, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 41557 + }, + { + "epoch": 0.41558, + "grad_norm": 0.8543379968846382, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 41558 + }, + { + "epoch": 0.41559, + "grad_norm": 0.8665315994869061, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 41559 + }, + { + "epoch": 0.4156, + "grad_norm": 0.9825295477106283, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 41560 + }, + { + "epoch": 0.41561, + "grad_norm": 1.056707614364633, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 41561 + }, + { + "epoch": 0.41562, + "grad_norm": 1.3388100940006642, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 41562 + }, + { + "epoch": 0.41563, + "grad_norm": 0.8541987364841278, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 41563 + }, + { + "epoch": 0.41564, + "grad_norm": 0.8905754307254498, + "learning_rate": 0.003, + "loss": 4.052, + "step": 41564 + }, + { + "epoch": 0.41565, + "grad_norm": 1.1772213705637944, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 41565 + }, + { + "epoch": 0.41566, + "grad_norm": 1.0628159252192895, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 41566 + }, + { + "epoch": 0.41567, + "grad_norm": 1.0578422632883473, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 41567 + }, + { + "epoch": 0.41568, + "grad_norm": 1.0002841166704863, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 41568 + }, + { + "epoch": 0.41569, + "grad_norm": 1.06067137567819, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 41569 + }, + { + "epoch": 0.4157, + "grad_norm": 0.9871264018214168, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 41570 + }, + { + "epoch": 0.41571, + "grad_norm": 1.0087513937644994, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 41571 + }, + { + "epoch": 0.41572, + "grad_norm": 1.123496043370526, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 41572 + }, + { + "epoch": 0.41573, + "grad_norm": 0.9944141679274872, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 41573 + }, + { + "epoch": 0.41574, + "grad_norm": 1.2778376615716556, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 41574 + }, + { + "epoch": 0.41575, + "grad_norm": 1.219890830868019, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 41575 + }, + { + "epoch": 0.41576, + "grad_norm": 1.1285556336079294, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 41576 + }, + { + "epoch": 0.41577, + "grad_norm": 1.1642295015452215, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 41577 + }, + { + "epoch": 0.41578, + "grad_norm": 1.0999255136439112, + "learning_rate": 0.003, + "loss": 4.042, + "step": 41578 + }, + { + "epoch": 0.41579, + "grad_norm": 1.2700283376442083, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 41579 + }, + { + "epoch": 0.4158, + "grad_norm": 1.0077594699549552, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 41580 + }, + { + "epoch": 0.41581, + "grad_norm": 0.9868758865644404, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 41581 + }, + { + "epoch": 0.41582, + "grad_norm": 1.4265997196369087, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 41582 + }, + { + "epoch": 0.41583, + "grad_norm": 0.8539868490095384, + "learning_rate": 0.003, + "loss": 4.026, + "step": 41583 + }, + { + "epoch": 0.41584, + "grad_norm": 0.9995178629028173, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 41584 + }, + { + "epoch": 0.41585, + "grad_norm": 1.0348027883130284, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 41585 + }, + { + "epoch": 0.41586, + "grad_norm": 1.2326472376160074, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 41586 + }, + { + "epoch": 0.41587, + "grad_norm": 1.0475800028125404, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 41587 + }, + { + "epoch": 0.41588, + "grad_norm": 1.1476451203157494, + "learning_rate": 0.003, + "loss": 4.039, + "step": 41588 + }, + { + "epoch": 0.41589, + "grad_norm": 1.2422065103480802, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 41589 + }, + { + "epoch": 0.4159, + "grad_norm": 0.8564532166727074, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 41590 + }, + { + "epoch": 0.41591, + "grad_norm": 0.8664014069802592, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 41591 + }, + { + "epoch": 0.41592, + "grad_norm": 0.9592270019671187, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 41592 + }, + { + "epoch": 0.41593, + "grad_norm": 1.0308422941879918, + "learning_rate": 0.003, + "loss": 4.024, + "step": 41593 + }, + { + "epoch": 0.41594, + "grad_norm": 1.0830169111831218, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 41594 + }, + { + "epoch": 0.41595, + "grad_norm": 1.1607018657222523, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 41595 + }, + { + "epoch": 0.41596, + "grad_norm": 1.1936613063322528, + "learning_rate": 0.003, + "loss": 4.048, + "step": 41596 + }, + { + "epoch": 0.41597, + "grad_norm": 1.1703698220468386, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 41597 + }, + { + "epoch": 0.41598, + "grad_norm": 1.0908188059536752, + "learning_rate": 0.003, + "loss": 4.034, + "step": 41598 + }, + { + "epoch": 0.41599, + "grad_norm": 1.0280221443033477, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 41599 + }, + { + "epoch": 0.416, + "grad_norm": 1.0348934691290808, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 41600 + }, + { + "epoch": 0.41601, + "grad_norm": 0.9630291257418975, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 41601 + }, + { + "epoch": 0.41602, + "grad_norm": 1.0134187031715476, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 41602 + }, + { + "epoch": 0.41603, + "grad_norm": 1.1961722962922825, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 41603 + }, + { + "epoch": 0.41604, + "grad_norm": 0.9261393523288702, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 41604 + }, + { + "epoch": 0.41605, + "grad_norm": 1.0929716681384827, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 41605 + }, + { + "epoch": 0.41606, + "grad_norm": 0.9219678406012684, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 41606 + }, + { + "epoch": 0.41607, + "grad_norm": 1.017033870589275, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 41607 + }, + { + "epoch": 0.41608, + "grad_norm": 1.2909512162725083, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 41608 + }, + { + "epoch": 0.41609, + "grad_norm": 0.9027016786009091, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 41609 + }, + { + "epoch": 0.4161, + "grad_norm": 0.9636569916184629, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 41610 + }, + { + "epoch": 0.41611, + "grad_norm": 0.9964313220491239, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 41611 + }, + { + "epoch": 0.41612, + "grad_norm": 0.9995655608917756, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 41612 + }, + { + "epoch": 0.41613, + "grad_norm": 1.0643455174832905, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 41613 + }, + { + "epoch": 0.41614, + "grad_norm": 1.0185590852349515, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 41614 + }, + { + "epoch": 0.41615, + "grad_norm": 1.31399512761402, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 41615 + }, + { + "epoch": 0.41616, + "grad_norm": 1.2461660498423885, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 41616 + }, + { + "epoch": 0.41617, + "grad_norm": 1.2937710581301225, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 41617 + }, + { + "epoch": 0.41618, + "grad_norm": 1.0335028154855765, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 41618 + }, + { + "epoch": 0.41619, + "grad_norm": 1.2681286692158218, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 41619 + }, + { + "epoch": 0.4162, + "grad_norm": 1.0773899728884762, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 41620 + }, + { + "epoch": 0.41621, + "grad_norm": 0.9449812822802709, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 41621 + }, + { + "epoch": 0.41622, + "grad_norm": 0.9391213806711266, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 41622 + }, + { + "epoch": 0.41623, + "grad_norm": 1.0289036988647893, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 41623 + }, + { + "epoch": 0.41624, + "grad_norm": 1.361788090740356, + "learning_rate": 0.003, + "loss": 4.027, + "step": 41624 + }, + { + "epoch": 0.41625, + "grad_norm": 0.7678552714514865, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 41625 + }, + { + "epoch": 0.41626, + "grad_norm": 0.8084545184817378, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 41626 + }, + { + "epoch": 0.41627, + "grad_norm": 0.8575459417909845, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 41627 + }, + { + "epoch": 0.41628, + "grad_norm": 1.0194775506990152, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 41628 + }, + { + "epoch": 0.41629, + "grad_norm": 1.1281734431219463, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 41629 + }, + { + "epoch": 0.4163, + "grad_norm": 1.0296770050359902, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 41630 + }, + { + "epoch": 0.41631, + "grad_norm": 1.237096255977147, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 41631 + }, + { + "epoch": 0.41632, + "grad_norm": 0.9837240044135587, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 41632 + }, + { + "epoch": 0.41633, + "grad_norm": 1.0476981319307552, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 41633 + }, + { + "epoch": 0.41634, + "grad_norm": 1.1046548736830843, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 41634 + }, + { + "epoch": 0.41635, + "grad_norm": 0.9287222622766886, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 41635 + }, + { + "epoch": 0.41636, + "grad_norm": 1.1644795375355794, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 41636 + }, + { + "epoch": 0.41637, + "grad_norm": 0.9374561180459188, + "learning_rate": 0.003, + "loss": 4.025, + "step": 41637 + }, + { + "epoch": 0.41638, + "grad_norm": 1.131114281850117, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 41638 + }, + { + "epoch": 0.41639, + "grad_norm": 1.1133539414953908, + "learning_rate": 0.003, + "loss": 4.031, + "step": 41639 + }, + { + "epoch": 0.4164, + "grad_norm": 1.2251993808849355, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 41640 + }, + { + "epoch": 0.41641, + "grad_norm": 1.0172337753692569, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 41641 + }, + { + "epoch": 0.41642, + "grad_norm": 1.2537479815454153, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 41642 + }, + { + "epoch": 0.41643, + "grad_norm": 1.060257547949241, + "learning_rate": 0.003, + "loss": 4.062, + "step": 41643 + }, + { + "epoch": 0.41644, + "grad_norm": 1.2534605467261313, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 41644 + }, + { + "epoch": 0.41645, + "grad_norm": 1.036836787884051, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 41645 + }, + { + "epoch": 0.41646, + "grad_norm": 1.0215479952300255, + "learning_rate": 0.003, + "loss": 4.0, + "step": 41646 + }, + { + "epoch": 0.41647, + "grad_norm": 1.3547909561688734, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 41647 + }, + { + "epoch": 0.41648, + "grad_norm": 0.7967768058623339, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 41648 + }, + { + "epoch": 0.41649, + "grad_norm": 0.7289090010938976, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 41649 + }, + { + "epoch": 0.4165, + "grad_norm": 0.8097872931624464, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 41650 + }, + { + "epoch": 0.41651, + "grad_norm": 0.9489206262144455, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 41651 + }, + { + "epoch": 0.41652, + "grad_norm": 1.1896128823887986, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 41652 + }, + { + "epoch": 0.41653, + "grad_norm": 0.9184421316122258, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 41653 + }, + { + "epoch": 0.41654, + "grad_norm": 0.8616858466939831, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 41654 + }, + { + "epoch": 0.41655, + "grad_norm": 0.9447552006525667, + "learning_rate": 0.003, + "loss": 4.056, + "step": 41655 + }, + { + "epoch": 0.41656, + "grad_norm": 1.1079383900407953, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 41656 + }, + { + "epoch": 0.41657, + "grad_norm": 1.4584272449633369, + "learning_rate": 0.003, + "loss": 4.051, + "step": 41657 + }, + { + "epoch": 0.41658, + "grad_norm": 1.0171822205599017, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 41658 + }, + { + "epoch": 0.41659, + "grad_norm": 1.2277683704147004, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 41659 + }, + { + "epoch": 0.4166, + "grad_norm": 1.0955890499894971, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 41660 + }, + { + "epoch": 0.41661, + "grad_norm": 1.1289622401334036, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 41661 + }, + { + "epoch": 0.41662, + "grad_norm": 0.8995088262365655, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 41662 + }, + { + "epoch": 0.41663, + "grad_norm": 0.9326991149795117, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 41663 + }, + { + "epoch": 0.41664, + "grad_norm": 1.144640152908402, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 41664 + }, + { + "epoch": 0.41665, + "grad_norm": 1.0200892407555064, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 41665 + }, + { + "epoch": 0.41666, + "grad_norm": 0.99457908887648, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 41666 + }, + { + "epoch": 0.41667, + "grad_norm": 1.1829800210041328, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 41667 + }, + { + "epoch": 0.41668, + "grad_norm": 1.0492974397841361, + "learning_rate": 0.003, + "loss": 4.046, + "step": 41668 + }, + { + "epoch": 0.41669, + "grad_norm": 1.0572427252979293, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 41669 + }, + { + "epoch": 0.4167, + "grad_norm": 0.972112282084282, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 41670 + }, + { + "epoch": 0.41671, + "grad_norm": 1.0321303242856437, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 41671 + }, + { + "epoch": 0.41672, + "grad_norm": 1.04679070496397, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 41672 + }, + { + "epoch": 0.41673, + "grad_norm": 1.1674785723136067, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 41673 + }, + { + "epoch": 0.41674, + "grad_norm": 1.1051923386408768, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 41674 + }, + { + "epoch": 0.41675, + "grad_norm": 1.11652471367937, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 41675 + }, + { + "epoch": 0.41676, + "grad_norm": 1.1386737936444522, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 41676 + }, + { + "epoch": 0.41677, + "grad_norm": 1.0961638210549616, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 41677 + }, + { + "epoch": 0.41678, + "grad_norm": 1.1613065781020446, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 41678 + }, + { + "epoch": 0.41679, + "grad_norm": 0.8753555071537961, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 41679 + }, + { + "epoch": 0.4168, + "grad_norm": 0.9347648693915211, + "learning_rate": 0.003, + "loss": 4.008, + "step": 41680 + }, + { + "epoch": 0.41681, + "grad_norm": 0.8972961538105625, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 41681 + }, + { + "epoch": 0.41682, + "grad_norm": 0.8615461728913865, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 41682 + }, + { + "epoch": 0.41683, + "grad_norm": 1.0090562817015436, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 41683 + }, + { + "epoch": 0.41684, + "grad_norm": 1.1227672898219676, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 41684 + }, + { + "epoch": 0.41685, + "grad_norm": 1.434631900321592, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 41685 + }, + { + "epoch": 0.41686, + "grad_norm": 0.9400219931961488, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 41686 + }, + { + "epoch": 0.41687, + "grad_norm": 1.0948723200307218, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 41687 + }, + { + "epoch": 0.41688, + "grad_norm": 1.2641454489890185, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 41688 + }, + { + "epoch": 0.41689, + "grad_norm": 0.982724791198542, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 41689 + }, + { + "epoch": 0.4169, + "grad_norm": 1.2358558780020483, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 41690 + }, + { + "epoch": 0.41691, + "grad_norm": 1.0638080165708022, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 41691 + }, + { + "epoch": 0.41692, + "grad_norm": 1.0506291373708323, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 41692 + }, + { + "epoch": 0.41693, + "grad_norm": 1.0078827305749745, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 41693 + }, + { + "epoch": 0.41694, + "grad_norm": 1.0863758142723856, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 41694 + }, + { + "epoch": 0.41695, + "grad_norm": 1.0631965553527798, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 41695 + }, + { + "epoch": 0.41696, + "grad_norm": 1.131240246915635, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 41696 + }, + { + "epoch": 0.41697, + "grad_norm": 1.3555160849145909, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 41697 + }, + { + "epoch": 0.41698, + "grad_norm": 0.8988787472673886, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 41698 + }, + { + "epoch": 0.41699, + "grad_norm": 0.8885415873326705, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 41699 + }, + { + "epoch": 0.417, + "grad_norm": 1.2554288221891174, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 41700 + }, + { + "epoch": 0.41701, + "grad_norm": 1.4829320585915329, + "learning_rate": 0.003, + "loss": 4.042, + "step": 41701 + }, + { + "epoch": 0.41702, + "grad_norm": 0.7135251833524857, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 41702 + }, + { + "epoch": 0.41703, + "grad_norm": 0.795560656336441, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 41703 + }, + { + "epoch": 0.41704, + "grad_norm": 0.8680786565042113, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 41704 + }, + { + "epoch": 0.41705, + "grad_norm": 1.0989202592339542, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 41705 + }, + { + "epoch": 0.41706, + "grad_norm": 1.1242105895729564, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 41706 + }, + { + "epoch": 0.41707, + "grad_norm": 1.1062527963901985, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 41707 + }, + { + "epoch": 0.41708, + "grad_norm": 1.3135851021470708, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 41708 + }, + { + "epoch": 0.41709, + "grad_norm": 0.744080934261401, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 41709 + }, + { + "epoch": 0.4171, + "grad_norm": 0.6661140385879757, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 41710 + }, + { + "epoch": 0.41711, + "grad_norm": 0.9341927027467603, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 41711 + }, + { + "epoch": 0.41712, + "grad_norm": 1.1834846010309967, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 41712 + }, + { + "epoch": 0.41713, + "grad_norm": 1.0646742635185913, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 41713 + }, + { + "epoch": 0.41714, + "grad_norm": 1.0532945251045815, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 41714 + }, + { + "epoch": 0.41715, + "grad_norm": 0.9753712708768331, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 41715 + }, + { + "epoch": 0.41716, + "grad_norm": 1.080813895447, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 41716 + }, + { + "epoch": 0.41717, + "grad_norm": 0.9374298766554522, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 41717 + }, + { + "epoch": 0.41718, + "grad_norm": 0.9850162414855228, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 41718 + }, + { + "epoch": 0.41719, + "grad_norm": 1.0806689337727502, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 41719 + }, + { + "epoch": 0.4172, + "grad_norm": 1.0126408255630794, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 41720 + }, + { + "epoch": 0.41721, + "grad_norm": 1.1500401567534668, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 41721 + }, + { + "epoch": 0.41722, + "grad_norm": 1.245495833665472, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 41722 + }, + { + "epoch": 0.41723, + "grad_norm": 1.2531344171569023, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 41723 + }, + { + "epoch": 0.41724, + "grad_norm": 1.0514936057090678, + "learning_rate": 0.003, + "loss": 4.024, + "step": 41724 + }, + { + "epoch": 0.41725, + "grad_norm": 1.1796358008555667, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 41725 + }, + { + "epoch": 0.41726, + "grad_norm": 1.0081483435098646, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 41726 + }, + { + "epoch": 0.41727, + "grad_norm": 1.1036494335764235, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 41727 + }, + { + "epoch": 0.41728, + "grad_norm": 0.9634170895864042, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 41728 + }, + { + "epoch": 0.41729, + "grad_norm": 1.0905778080574324, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 41729 + }, + { + "epoch": 0.4173, + "grad_norm": 1.164406258720973, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 41730 + }, + { + "epoch": 0.41731, + "grad_norm": 1.06143891239501, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 41731 + }, + { + "epoch": 0.41732, + "grad_norm": 1.111724451572271, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 41732 + }, + { + "epoch": 0.41733, + "grad_norm": 1.1093154065536384, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 41733 + }, + { + "epoch": 0.41734, + "grad_norm": 1.1653974646354157, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 41734 + }, + { + "epoch": 0.41735, + "grad_norm": 0.9042114411952707, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 41735 + }, + { + "epoch": 0.41736, + "grad_norm": 0.870452878000975, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 41736 + }, + { + "epoch": 0.41737, + "grad_norm": 0.9135421897067536, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 41737 + }, + { + "epoch": 0.41738, + "grad_norm": 0.9360331573359462, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 41738 + }, + { + "epoch": 0.41739, + "grad_norm": 1.0058202465558488, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 41739 + }, + { + "epoch": 0.4174, + "grad_norm": 1.4813737011973325, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 41740 + }, + { + "epoch": 0.41741, + "grad_norm": 0.7938290303933376, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 41741 + }, + { + "epoch": 0.41742, + "grad_norm": 0.9669801666488497, + "learning_rate": 0.003, + "loss": 4.016, + "step": 41742 + }, + { + "epoch": 0.41743, + "grad_norm": 1.163208143597179, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 41743 + }, + { + "epoch": 0.41744, + "grad_norm": 1.0130321957683408, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 41744 + }, + { + "epoch": 0.41745, + "grad_norm": 1.267407258538966, + "learning_rate": 0.003, + "loss": 4.056, + "step": 41745 + }, + { + "epoch": 0.41746, + "grad_norm": 1.0752878871841973, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 41746 + }, + { + "epoch": 0.41747, + "grad_norm": 1.0054000720438934, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 41747 + }, + { + "epoch": 0.41748, + "grad_norm": 1.18576508623667, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 41748 + }, + { + "epoch": 0.41749, + "grad_norm": 1.2356355448242893, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 41749 + }, + { + "epoch": 0.4175, + "grad_norm": 0.964544524165719, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 41750 + }, + { + "epoch": 0.41751, + "grad_norm": 1.0016678480994448, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 41751 + }, + { + "epoch": 0.41752, + "grad_norm": 1.182002723358865, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 41752 + }, + { + "epoch": 0.41753, + "grad_norm": 1.1196235759103976, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 41753 + }, + { + "epoch": 0.41754, + "grad_norm": 1.1862850943647336, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 41754 + }, + { + "epoch": 0.41755, + "grad_norm": 1.0599609275520026, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 41755 + }, + { + "epoch": 0.41756, + "grad_norm": 1.2950247306137899, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 41756 + }, + { + "epoch": 0.41757, + "grad_norm": 0.9671861304425631, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 41757 + }, + { + "epoch": 0.41758, + "grad_norm": 1.272572023961474, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 41758 + }, + { + "epoch": 0.41759, + "grad_norm": 0.8473470061428636, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 41759 + }, + { + "epoch": 0.4176, + "grad_norm": 0.775391176331998, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 41760 + }, + { + "epoch": 0.41761, + "grad_norm": 0.943121837943406, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 41761 + }, + { + "epoch": 0.41762, + "grad_norm": 1.395683495770378, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 41762 + }, + { + "epoch": 0.41763, + "grad_norm": 0.8558044481518582, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 41763 + }, + { + "epoch": 0.41764, + "grad_norm": 0.991166186547173, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 41764 + }, + { + "epoch": 0.41765, + "grad_norm": 1.5553093922702437, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 41765 + }, + { + "epoch": 0.41766, + "grad_norm": 0.7941987830970484, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 41766 + }, + { + "epoch": 0.41767, + "grad_norm": 0.8312490666872868, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 41767 + }, + { + "epoch": 0.41768, + "grad_norm": 1.1562801782393726, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 41768 + }, + { + "epoch": 0.41769, + "grad_norm": 0.987812319879747, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 41769 + }, + { + "epoch": 0.4177, + "grad_norm": 1.1199998186604097, + "learning_rate": 0.003, + "loss": 4.007, + "step": 41770 + }, + { + "epoch": 0.41771, + "grad_norm": 1.0357126929143867, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 41771 + }, + { + "epoch": 0.41772, + "grad_norm": 1.3759885989345368, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 41772 + }, + { + "epoch": 0.41773, + "grad_norm": 0.8038663334009913, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 41773 + }, + { + "epoch": 0.41774, + "grad_norm": 0.760474742599847, + "learning_rate": 0.003, + "loss": 4.026, + "step": 41774 + }, + { + "epoch": 0.41775, + "grad_norm": 0.9365187547182867, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 41775 + }, + { + "epoch": 0.41776, + "grad_norm": 1.2357006594825015, + "learning_rate": 0.003, + "loss": 4.031, + "step": 41776 + }, + { + "epoch": 0.41777, + "grad_norm": 0.9775205256845433, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 41777 + }, + { + "epoch": 0.41778, + "grad_norm": 1.3156142659053605, + "learning_rate": 0.003, + "loss": 4.0876, + "step": 41778 + }, + { + "epoch": 0.41779, + "grad_norm": 0.8881526738843615, + "learning_rate": 0.003, + "loss": 4.039, + "step": 41779 + }, + { + "epoch": 0.4178, + "grad_norm": 0.9605325506008523, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 41780 + }, + { + "epoch": 0.41781, + "grad_norm": 1.3391075085721091, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 41781 + }, + { + "epoch": 0.41782, + "grad_norm": 0.9810215347201936, + "learning_rate": 0.003, + "loss": 4.024, + "step": 41782 + }, + { + "epoch": 0.41783, + "grad_norm": 1.1207761300169603, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 41783 + }, + { + "epoch": 0.41784, + "grad_norm": 0.928711710933346, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 41784 + }, + { + "epoch": 0.41785, + "grad_norm": 1.0422525532614264, + "learning_rate": 0.003, + "loss": 4.045, + "step": 41785 + }, + { + "epoch": 0.41786, + "grad_norm": 1.0820734942574468, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 41786 + }, + { + "epoch": 0.41787, + "grad_norm": 0.9075908801098322, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 41787 + }, + { + "epoch": 0.41788, + "grad_norm": 0.9997371306739264, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 41788 + }, + { + "epoch": 0.41789, + "grad_norm": 1.2459079461157705, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 41789 + }, + { + "epoch": 0.4179, + "grad_norm": 1.1139458722059152, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 41790 + }, + { + "epoch": 0.41791, + "grad_norm": 1.2959598726888208, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 41791 + }, + { + "epoch": 0.41792, + "grad_norm": 1.201799711474182, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 41792 + }, + { + "epoch": 0.41793, + "grad_norm": 1.192163319602088, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 41793 + }, + { + "epoch": 0.41794, + "grad_norm": 0.9676776025976345, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 41794 + }, + { + "epoch": 0.41795, + "grad_norm": 1.133187326637371, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 41795 + }, + { + "epoch": 0.41796, + "grad_norm": 1.2434800527227, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 41796 + }, + { + "epoch": 0.41797, + "grad_norm": 0.9872984894202974, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 41797 + }, + { + "epoch": 0.41798, + "grad_norm": 0.9617859140484262, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 41798 + }, + { + "epoch": 0.41799, + "grad_norm": 0.8877109612972675, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 41799 + }, + { + "epoch": 0.418, + "grad_norm": 1.0521446518233848, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 41800 + }, + { + "epoch": 0.41801, + "grad_norm": 1.2587632194932576, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 41801 + }, + { + "epoch": 0.41802, + "grad_norm": 0.9457454232619975, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 41802 + }, + { + "epoch": 0.41803, + "grad_norm": 1.033457443323151, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 41803 + }, + { + "epoch": 0.41804, + "grad_norm": 1.2122182963579087, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 41804 + }, + { + "epoch": 0.41805, + "grad_norm": 0.8940663639486556, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 41805 + }, + { + "epoch": 0.41806, + "grad_norm": 0.9299520337130135, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 41806 + }, + { + "epoch": 0.41807, + "grad_norm": 1.0214335296176, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 41807 + }, + { + "epoch": 0.41808, + "grad_norm": 1.136790841090598, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 41808 + }, + { + "epoch": 0.41809, + "grad_norm": 0.9961019789810795, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 41809 + }, + { + "epoch": 0.4181, + "grad_norm": 1.1047930598836664, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 41810 + }, + { + "epoch": 0.41811, + "grad_norm": 0.9853385687280223, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 41811 + }, + { + "epoch": 0.41812, + "grad_norm": 1.3617541545150718, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 41812 + }, + { + "epoch": 0.41813, + "grad_norm": 1.058210173560052, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 41813 + }, + { + "epoch": 0.41814, + "grad_norm": 1.1953894338187578, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 41814 + }, + { + "epoch": 0.41815, + "grad_norm": 1.0020417845948333, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 41815 + }, + { + "epoch": 0.41816, + "grad_norm": 1.2539961590555124, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 41816 + }, + { + "epoch": 0.41817, + "grad_norm": 1.0775632711973404, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 41817 + }, + { + "epoch": 0.41818, + "grad_norm": 1.1777386935551801, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 41818 + }, + { + "epoch": 0.41819, + "grad_norm": 0.9092808949071826, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 41819 + }, + { + "epoch": 0.4182, + "grad_norm": 0.9160099584017246, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 41820 + }, + { + "epoch": 0.41821, + "grad_norm": 0.9333604526155885, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 41821 + }, + { + "epoch": 0.41822, + "grad_norm": 0.9649615902990006, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 41822 + }, + { + "epoch": 0.41823, + "grad_norm": 1.153228983093948, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 41823 + }, + { + "epoch": 0.41824, + "grad_norm": 1.00268032295567, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 41824 + }, + { + "epoch": 0.41825, + "grad_norm": 1.136536358464147, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 41825 + }, + { + "epoch": 0.41826, + "grad_norm": 1.256897591164052, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 41826 + }, + { + "epoch": 0.41827, + "grad_norm": 1.0303472827808837, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 41827 + }, + { + "epoch": 0.41828, + "grad_norm": 0.919625532771164, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 41828 + }, + { + "epoch": 0.41829, + "grad_norm": 0.874874573810031, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 41829 + }, + { + "epoch": 0.4183, + "grad_norm": 0.9444618989995905, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 41830 + }, + { + "epoch": 0.41831, + "grad_norm": 1.1834609601289572, + "learning_rate": 0.003, + "loss": 4.031, + "step": 41831 + }, + { + "epoch": 0.41832, + "grad_norm": 1.10318411502311, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 41832 + }, + { + "epoch": 0.41833, + "grad_norm": 1.0905157741011995, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 41833 + }, + { + "epoch": 0.41834, + "grad_norm": 1.1077225874061205, + "learning_rate": 0.003, + "loss": 4.043, + "step": 41834 + }, + { + "epoch": 0.41835, + "grad_norm": 1.0008071735276078, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 41835 + }, + { + "epoch": 0.41836, + "grad_norm": 1.0712132748037975, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 41836 + }, + { + "epoch": 0.41837, + "grad_norm": 1.135477691567293, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 41837 + }, + { + "epoch": 0.41838, + "grad_norm": 0.9001769581035661, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 41838 + }, + { + "epoch": 0.41839, + "grad_norm": 0.9018865880950482, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 41839 + }, + { + "epoch": 0.4184, + "grad_norm": 1.2669976534162835, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 41840 + }, + { + "epoch": 0.41841, + "grad_norm": 1.3513426510407118, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 41841 + }, + { + "epoch": 0.41842, + "grad_norm": 0.7263143468867462, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 41842 + }, + { + "epoch": 0.41843, + "grad_norm": 0.8726768274109626, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 41843 + }, + { + "epoch": 0.41844, + "grad_norm": 1.0484873453337258, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 41844 + }, + { + "epoch": 0.41845, + "grad_norm": 0.9881291571630996, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 41845 + }, + { + "epoch": 0.41846, + "grad_norm": 1.2020190200862697, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 41846 + }, + { + "epoch": 0.41847, + "grad_norm": 0.9700579335121248, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 41847 + }, + { + "epoch": 0.41848, + "grad_norm": 1.0541734593502632, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 41848 + }, + { + "epoch": 0.41849, + "grad_norm": 1.0580658630567308, + "learning_rate": 0.003, + "loss": 4.047, + "step": 41849 + }, + { + "epoch": 0.4185, + "grad_norm": 1.1907898005702298, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 41850 + }, + { + "epoch": 0.41851, + "grad_norm": 1.1055885487911037, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 41851 + }, + { + "epoch": 0.41852, + "grad_norm": 1.0440141858268888, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 41852 + }, + { + "epoch": 0.41853, + "grad_norm": 1.0496544682395155, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 41853 + }, + { + "epoch": 0.41854, + "grad_norm": 1.0753134573471492, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 41854 + }, + { + "epoch": 0.41855, + "grad_norm": 1.3558153299960638, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 41855 + }, + { + "epoch": 0.41856, + "grad_norm": 1.391702309972483, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 41856 + }, + { + "epoch": 0.41857, + "grad_norm": 0.9145851165103215, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 41857 + }, + { + "epoch": 0.41858, + "grad_norm": 0.9297007351463259, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 41858 + }, + { + "epoch": 0.41859, + "grad_norm": 1.3629058179906977, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 41859 + }, + { + "epoch": 0.4186, + "grad_norm": 1.2169518699345938, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 41860 + }, + { + "epoch": 0.41861, + "grad_norm": 0.9786265088452871, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 41861 + }, + { + "epoch": 0.41862, + "grad_norm": 1.0208320043380197, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 41862 + }, + { + "epoch": 0.41863, + "grad_norm": 1.3620485663472732, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 41863 + }, + { + "epoch": 0.41864, + "grad_norm": 1.0712392880507833, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 41864 + }, + { + "epoch": 0.41865, + "grad_norm": 1.026689264398307, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 41865 + }, + { + "epoch": 0.41866, + "grad_norm": 1.1720912896647262, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 41866 + }, + { + "epoch": 0.41867, + "grad_norm": 0.9501045526663826, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 41867 + }, + { + "epoch": 0.41868, + "grad_norm": 1.0631149463607854, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 41868 + }, + { + "epoch": 0.41869, + "grad_norm": 1.0382380540298841, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 41869 + }, + { + "epoch": 0.4187, + "grad_norm": 0.9886662534660516, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 41870 + }, + { + "epoch": 0.41871, + "grad_norm": 1.1695540161767612, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 41871 + }, + { + "epoch": 0.41872, + "grad_norm": 1.128179171279366, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 41872 + }, + { + "epoch": 0.41873, + "grad_norm": 1.0579798754166874, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 41873 + }, + { + "epoch": 0.41874, + "grad_norm": 1.3253992519754239, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 41874 + }, + { + "epoch": 0.41875, + "grad_norm": 1.0518910986070689, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 41875 + }, + { + "epoch": 0.41876, + "grad_norm": 0.9668509364531039, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 41876 + }, + { + "epoch": 0.41877, + "grad_norm": 1.0758830122122491, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 41877 + }, + { + "epoch": 0.41878, + "grad_norm": 1.1550687851703736, + "learning_rate": 0.003, + "loss": 4.031, + "step": 41878 + }, + { + "epoch": 0.41879, + "grad_norm": 1.1215335897307248, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 41879 + }, + { + "epoch": 0.4188, + "grad_norm": 1.1159487333656402, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 41880 + }, + { + "epoch": 0.41881, + "grad_norm": 0.9851988245030391, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 41881 + }, + { + "epoch": 0.41882, + "grad_norm": 1.2063854596346377, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 41882 + }, + { + "epoch": 0.41883, + "grad_norm": 1.000960999808621, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 41883 + }, + { + "epoch": 0.41884, + "grad_norm": 1.0265097963174787, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 41884 + }, + { + "epoch": 0.41885, + "grad_norm": 0.9792265939992063, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 41885 + }, + { + "epoch": 0.41886, + "grad_norm": 1.2969566439209834, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 41886 + }, + { + "epoch": 0.41887, + "grad_norm": 1.0514376443332738, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 41887 + }, + { + "epoch": 0.41888, + "grad_norm": 0.9482409315741889, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 41888 + }, + { + "epoch": 0.41889, + "grad_norm": 1.1514682388795503, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 41889 + }, + { + "epoch": 0.4189, + "grad_norm": 1.1791151792730674, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 41890 + }, + { + "epoch": 0.41891, + "grad_norm": 0.9847208178238401, + "learning_rate": 0.003, + "loss": 4.031, + "step": 41891 + }, + { + "epoch": 0.41892, + "grad_norm": 0.971520960373586, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 41892 + }, + { + "epoch": 0.41893, + "grad_norm": 1.1612839076708374, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 41893 + }, + { + "epoch": 0.41894, + "grad_norm": 1.0220001878804958, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 41894 + }, + { + "epoch": 0.41895, + "grad_norm": 1.1251974189996254, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 41895 + }, + { + "epoch": 0.41896, + "grad_norm": 1.027060585234623, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 41896 + }, + { + "epoch": 0.41897, + "grad_norm": 1.1647332957666505, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 41897 + }, + { + "epoch": 0.41898, + "grad_norm": 0.9049703655222479, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 41898 + }, + { + "epoch": 0.41899, + "grad_norm": 0.9348653178753809, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 41899 + }, + { + "epoch": 0.419, + "grad_norm": 1.136273491293085, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 41900 + }, + { + "epoch": 0.41901, + "grad_norm": 0.9591778424799284, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 41901 + }, + { + "epoch": 0.41902, + "grad_norm": 1.0706297364840374, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 41902 + }, + { + "epoch": 0.41903, + "grad_norm": 1.2117372011778056, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 41903 + }, + { + "epoch": 0.41904, + "grad_norm": 1.090370003517041, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 41904 + }, + { + "epoch": 0.41905, + "grad_norm": 1.0476949273248546, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 41905 + }, + { + "epoch": 0.41906, + "grad_norm": 1.0507282163507252, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 41906 + }, + { + "epoch": 0.41907, + "grad_norm": 1.1827437200249744, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 41907 + }, + { + "epoch": 0.41908, + "grad_norm": 1.2317661520684726, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 41908 + }, + { + "epoch": 0.41909, + "grad_norm": 1.0044281315158607, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 41909 + }, + { + "epoch": 0.4191, + "grad_norm": 1.1807598660886114, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 41910 + }, + { + "epoch": 0.41911, + "grad_norm": 0.8898336128055293, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 41911 + }, + { + "epoch": 0.41912, + "grad_norm": 0.9601108459614979, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 41912 + }, + { + "epoch": 0.41913, + "grad_norm": 1.444941890388895, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 41913 + }, + { + "epoch": 0.41914, + "grad_norm": 1.0679097227976013, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 41914 + }, + { + "epoch": 0.41915, + "grad_norm": 1.3929389714711586, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 41915 + }, + { + "epoch": 0.41916, + "grad_norm": 0.8853851585195818, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 41916 + }, + { + "epoch": 0.41917, + "grad_norm": 0.9024006748047783, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 41917 + }, + { + "epoch": 0.41918, + "grad_norm": 1.292972963907102, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 41918 + }, + { + "epoch": 0.41919, + "grad_norm": 1.0277967554370364, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 41919 + }, + { + "epoch": 0.4192, + "grad_norm": 1.2951331398341255, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 41920 + }, + { + "epoch": 0.41921, + "grad_norm": 1.032192064564518, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 41921 + }, + { + "epoch": 0.41922, + "grad_norm": 1.0534941428812508, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 41922 + }, + { + "epoch": 0.41923, + "grad_norm": 1.1060304171670792, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 41923 + }, + { + "epoch": 0.41924, + "grad_norm": 0.9619081372748317, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 41924 + }, + { + "epoch": 0.41925, + "grad_norm": 1.1796059297773813, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 41925 + }, + { + "epoch": 0.41926, + "grad_norm": 1.2768690847939816, + "learning_rate": 0.003, + "loss": 4.034, + "step": 41926 + }, + { + "epoch": 0.41927, + "grad_norm": 0.8997343227544267, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 41927 + }, + { + "epoch": 0.41928, + "grad_norm": 1.022975374936245, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 41928 + }, + { + "epoch": 0.41929, + "grad_norm": 0.9661739733879576, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 41929 + }, + { + "epoch": 0.4193, + "grad_norm": 1.1261245449933597, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 41930 + }, + { + "epoch": 0.41931, + "grad_norm": 1.0428590626719256, + "learning_rate": 0.003, + "loss": 4.027, + "step": 41931 + }, + { + "epoch": 0.41932, + "grad_norm": 1.0846756582491022, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 41932 + }, + { + "epoch": 0.41933, + "grad_norm": 1.1622456869343232, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 41933 + }, + { + "epoch": 0.41934, + "grad_norm": 1.1029418495215682, + "learning_rate": 0.003, + "loss": 4.032, + "step": 41934 + }, + { + "epoch": 0.41935, + "grad_norm": 1.222086397622649, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 41935 + }, + { + "epoch": 0.41936, + "grad_norm": 0.9011417978945144, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 41936 + }, + { + "epoch": 0.41937, + "grad_norm": 0.8525305979994902, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 41937 + }, + { + "epoch": 0.41938, + "grad_norm": 0.9107210634150585, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 41938 + }, + { + "epoch": 0.41939, + "grad_norm": 1.0946138282365627, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 41939 + }, + { + "epoch": 0.4194, + "grad_norm": 1.1021915652084222, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 41940 + }, + { + "epoch": 0.41941, + "grad_norm": 1.072191795291337, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 41941 + }, + { + "epoch": 0.41942, + "grad_norm": 1.1432666812990047, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 41942 + }, + { + "epoch": 0.41943, + "grad_norm": 1.0179462797031613, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 41943 + }, + { + "epoch": 0.41944, + "grad_norm": 1.0776551154131706, + "learning_rate": 0.003, + "loss": 4.02, + "step": 41944 + }, + { + "epoch": 0.41945, + "grad_norm": 1.0843923798976642, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 41945 + }, + { + "epoch": 0.41946, + "grad_norm": 1.1822718697944963, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 41946 + }, + { + "epoch": 0.41947, + "grad_norm": 0.8799294088304025, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 41947 + }, + { + "epoch": 0.41948, + "grad_norm": 1.0176640303779483, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 41948 + }, + { + "epoch": 0.41949, + "grad_norm": 1.3035854108144844, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 41949 + }, + { + "epoch": 0.4195, + "grad_norm": 0.9343426130910395, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 41950 + }, + { + "epoch": 0.41951, + "grad_norm": 1.02138191104763, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 41951 + }, + { + "epoch": 0.41952, + "grad_norm": 1.1841524835796653, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 41952 + }, + { + "epoch": 0.41953, + "grad_norm": 1.0744770959213747, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 41953 + }, + { + "epoch": 0.41954, + "grad_norm": 1.0491673632287775, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 41954 + }, + { + "epoch": 0.41955, + "grad_norm": 1.0636140308354383, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 41955 + }, + { + "epoch": 0.41956, + "grad_norm": 1.1647369447031108, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 41956 + }, + { + "epoch": 0.41957, + "grad_norm": 1.0447157606392392, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 41957 + }, + { + "epoch": 0.41958, + "grad_norm": 1.0933177661284612, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 41958 + }, + { + "epoch": 0.41959, + "grad_norm": 1.1490898268883751, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 41959 + }, + { + "epoch": 0.4196, + "grad_norm": 1.0377700512178742, + "learning_rate": 0.003, + "loss": 4.009, + "step": 41960 + }, + { + "epoch": 0.41961, + "grad_norm": 1.2311750707203726, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 41961 + }, + { + "epoch": 0.41962, + "grad_norm": 0.9880470670862601, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 41962 + }, + { + "epoch": 0.41963, + "grad_norm": 1.266617293313973, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 41963 + }, + { + "epoch": 0.41964, + "grad_norm": 1.0158225523092383, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 41964 + }, + { + "epoch": 0.41965, + "grad_norm": 1.0030909998834456, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 41965 + }, + { + "epoch": 0.41966, + "grad_norm": 1.2042803818662282, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 41966 + }, + { + "epoch": 0.41967, + "grad_norm": 1.207633113043871, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 41967 + }, + { + "epoch": 0.41968, + "grad_norm": 0.9026973245918345, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 41968 + }, + { + "epoch": 0.41969, + "grad_norm": 0.9201459811020748, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 41969 + }, + { + "epoch": 0.4197, + "grad_norm": 0.8739481710125794, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 41970 + }, + { + "epoch": 0.41971, + "grad_norm": 0.8153670909197822, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 41971 + }, + { + "epoch": 0.41972, + "grad_norm": 0.8103188460818084, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 41972 + }, + { + "epoch": 0.41973, + "grad_norm": 0.9570113546980344, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 41973 + }, + { + "epoch": 0.41974, + "grad_norm": 1.4637339032187664, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 41974 + }, + { + "epoch": 0.41975, + "grad_norm": 0.9655515872340658, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 41975 + }, + { + "epoch": 0.41976, + "grad_norm": 1.1827517899039768, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 41976 + }, + { + "epoch": 0.41977, + "grad_norm": 1.1260925380293731, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 41977 + }, + { + "epoch": 0.41978, + "grad_norm": 0.9146029905811863, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 41978 + }, + { + "epoch": 0.41979, + "grad_norm": 0.8834778303938065, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 41979 + }, + { + "epoch": 0.4198, + "grad_norm": 0.9692516364643323, + "learning_rate": 0.003, + "loss": 4.03, + "step": 41980 + }, + { + "epoch": 0.41981, + "grad_norm": 1.2911270312412422, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 41981 + }, + { + "epoch": 0.41982, + "grad_norm": 0.8623356098692484, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 41982 + }, + { + "epoch": 0.41983, + "grad_norm": 0.8599605602796467, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 41983 + }, + { + "epoch": 0.41984, + "grad_norm": 0.9144420292398631, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 41984 + }, + { + "epoch": 0.41985, + "grad_norm": 1.0268752322505208, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 41985 + }, + { + "epoch": 0.41986, + "grad_norm": 1.0771164936555504, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 41986 + }, + { + "epoch": 0.41987, + "grad_norm": 0.979946221185764, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 41987 + }, + { + "epoch": 0.41988, + "grad_norm": 0.9758350467033148, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 41988 + }, + { + "epoch": 0.41989, + "grad_norm": 1.0418296706801717, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 41989 + }, + { + "epoch": 0.4199, + "grad_norm": 1.2121031271254987, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 41990 + }, + { + "epoch": 0.41991, + "grad_norm": 1.222935105112375, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 41991 + }, + { + "epoch": 0.41992, + "grad_norm": 0.9950229154581128, + "learning_rate": 0.003, + "loss": 4.042, + "step": 41992 + }, + { + "epoch": 0.41993, + "grad_norm": 1.2973743568066713, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 41993 + }, + { + "epoch": 0.41994, + "grad_norm": 0.8848712659735267, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 41994 + }, + { + "epoch": 0.41995, + "grad_norm": 0.8481274882340895, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 41995 + }, + { + "epoch": 0.41996, + "grad_norm": 0.9897414276625159, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 41996 + }, + { + "epoch": 0.41997, + "grad_norm": 1.1571610570520228, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 41997 + }, + { + "epoch": 0.41998, + "grad_norm": 1.4158092672136864, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 41998 + }, + { + "epoch": 0.41999, + "grad_norm": 0.9396289112431226, + "learning_rate": 0.003, + "loss": 4.043, + "step": 41999 + }, + { + "epoch": 0.42, + "grad_norm": 1.0004628276350382, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 42000 + }, + { + "epoch": 0.42001, + "grad_norm": 1.2638281199597166, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 42001 + }, + { + "epoch": 0.42002, + "grad_norm": 0.9175715107678741, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 42002 + }, + { + "epoch": 0.42003, + "grad_norm": 0.9569890307654938, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 42003 + }, + { + "epoch": 0.42004, + "grad_norm": 1.1608293528429152, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 42004 + }, + { + "epoch": 0.42005, + "grad_norm": 1.2804645272860087, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 42005 + }, + { + "epoch": 0.42006, + "grad_norm": 0.9085252109907554, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 42006 + }, + { + "epoch": 0.42007, + "grad_norm": 1.03461569847127, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 42007 + }, + { + "epoch": 0.42008, + "grad_norm": 1.0572037708064739, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 42008 + }, + { + "epoch": 0.42009, + "grad_norm": 1.0881045909647273, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 42009 + }, + { + "epoch": 0.4201, + "grad_norm": 1.0815107320475934, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 42010 + }, + { + "epoch": 0.42011, + "grad_norm": 1.0676155153639508, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 42011 + }, + { + "epoch": 0.42012, + "grad_norm": 1.1618668458769912, + "learning_rate": 0.003, + "loss": 4.052, + "step": 42012 + }, + { + "epoch": 0.42013, + "grad_norm": 1.1112634051763797, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 42013 + }, + { + "epoch": 0.42014, + "grad_norm": 1.3289729984628635, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 42014 + }, + { + "epoch": 0.42015, + "grad_norm": 0.8151173693246274, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 42015 + }, + { + "epoch": 0.42016, + "grad_norm": 0.8006036394489082, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 42016 + }, + { + "epoch": 0.42017, + "grad_norm": 0.8530667762039513, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 42017 + }, + { + "epoch": 0.42018, + "grad_norm": 0.8889935644261777, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 42018 + }, + { + "epoch": 0.42019, + "grad_norm": 0.9770652408034137, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 42019 + }, + { + "epoch": 0.4202, + "grad_norm": 1.0604631754003133, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 42020 + }, + { + "epoch": 0.42021, + "grad_norm": 1.083928888062653, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 42021 + }, + { + "epoch": 0.42022, + "grad_norm": 0.9078677464284192, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 42022 + }, + { + "epoch": 0.42023, + "grad_norm": 1.1324532525097535, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 42023 + }, + { + "epoch": 0.42024, + "grad_norm": 1.012640919828751, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 42024 + }, + { + "epoch": 0.42025, + "grad_norm": 1.1154971912531608, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 42025 + }, + { + "epoch": 0.42026, + "grad_norm": 0.9755513505425359, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 42026 + }, + { + "epoch": 0.42027, + "grad_norm": 1.2696732238342916, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 42027 + }, + { + "epoch": 0.42028, + "grad_norm": 1.165545893035464, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 42028 + }, + { + "epoch": 0.42029, + "grad_norm": 1.0659631579153783, + "learning_rate": 0.003, + "loss": 4.027, + "step": 42029 + }, + { + "epoch": 0.4203, + "grad_norm": 0.99144676705353, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 42030 + }, + { + "epoch": 0.42031, + "grad_norm": 1.3594552561364177, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 42031 + }, + { + "epoch": 0.42032, + "grad_norm": 0.995945216807129, + "learning_rate": 0.003, + "loss": 4.063, + "step": 42032 + }, + { + "epoch": 0.42033, + "grad_norm": 1.1561885340823401, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 42033 + }, + { + "epoch": 0.42034, + "grad_norm": 0.9401888811974126, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 42034 + }, + { + "epoch": 0.42035, + "grad_norm": 1.052021603944741, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 42035 + }, + { + "epoch": 0.42036, + "grad_norm": 1.136189571174776, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 42036 + }, + { + "epoch": 0.42037, + "grad_norm": 0.9724381141878012, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 42037 + }, + { + "epoch": 0.42038, + "grad_norm": 1.147784759134928, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 42038 + }, + { + "epoch": 0.42039, + "grad_norm": 0.9811842640101435, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 42039 + }, + { + "epoch": 0.4204, + "grad_norm": 1.053631237151014, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 42040 + }, + { + "epoch": 0.42041, + "grad_norm": 1.0003034926187684, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 42041 + }, + { + "epoch": 0.42042, + "grad_norm": 0.992077266374198, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 42042 + }, + { + "epoch": 0.42043, + "grad_norm": 1.4074130714392756, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 42043 + }, + { + "epoch": 0.42044, + "grad_norm": 0.9531454969973591, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 42044 + }, + { + "epoch": 0.42045, + "grad_norm": 1.3510311465733393, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 42045 + }, + { + "epoch": 0.42046, + "grad_norm": 0.9120030965949675, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 42046 + }, + { + "epoch": 0.42047, + "grad_norm": 0.7813603704147798, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 42047 + }, + { + "epoch": 0.42048, + "grad_norm": 0.7577555354181466, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 42048 + }, + { + "epoch": 0.42049, + "grad_norm": 0.8966662053144409, + "learning_rate": 0.003, + "loss": 4.013, + "step": 42049 + }, + { + "epoch": 0.4205, + "grad_norm": 1.1786407464218187, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 42050 + }, + { + "epoch": 0.42051, + "grad_norm": 1.1953662413873574, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 42051 + }, + { + "epoch": 0.42052, + "grad_norm": 1.0497363346189725, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 42052 + }, + { + "epoch": 0.42053, + "grad_norm": 0.885344193730302, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 42053 + }, + { + "epoch": 0.42054, + "grad_norm": 0.786828131233083, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 42054 + }, + { + "epoch": 0.42055, + "grad_norm": 0.773797273334827, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 42055 + }, + { + "epoch": 0.42056, + "grad_norm": 0.8243413408796304, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 42056 + }, + { + "epoch": 0.42057, + "grad_norm": 0.9403672893655552, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 42057 + }, + { + "epoch": 0.42058, + "grad_norm": 1.0182553243437726, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 42058 + }, + { + "epoch": 0.42059, + "grad_norm": 1.1812994315570187, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 42059 + }, + { + "epoch": 0.4206, + "grad_norm": 1.18509021003161, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 42060 + }, + { + "epoch": 0.42061, + "grad_norm": 1.5178087125107789, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 42061 + }, + { + "epoch": 0.42062, + "grad_norm": 0.7072660980477365, + "learning_rate": 0.003, + "loss": 4.031, + "step": 42062 + }, + { + "epoch": 0.42063, + "grad_norm": 0.7888133018897585, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 42063 + }, + { + "epoch": 0.42064, + "grad_norm": 0.8295875722225418, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 42064 + }, + { + "epoch": 0.42065, + "grad_norm": 0.9270468900966125, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 42065 + }, + { + "epoch": 0.42066, + "grad_norm": 1.08874087169847, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 42066 + }, + { + "epoch": 0.42067, + "grad_norm": 1.114957435260073, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 42067 + }, + { + "epoch": 0.42068, + "grad_norm": 1.240874547571618, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 42068 + }, + { + "epoch": 0.42069, + "grad_norm": 1.0668890728639358, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 42069 + }, + { + "epoch": 0.4207, + "grad_norm": 0.9661607859308536, + "learning_rate": 0.003, + "loss": 4.021, + "step": 42070 + }, + { + "epoch": 0.42071, + "grad_norm": 1.1635658518422487, + "learning_rate": 0.003, + "loss": 4.071, + "step": 42071 + }, + { + "epoch": 0.42072, + "grad_norm": 1.1070048999706892, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 42072 + }, + { + "epoch": 0.42073, + "grad_norm": 0.9590502735812532, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 42073 + }, + { + "epoch": 0.42074, + "grad_norm": 1.110391357207208, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 42074 + }, + { + "epoch": 0.42075, + "grad_norm": 1.0565655598922408, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 42075 + }, + { + "epoch": 0.42076, + "grad_norm": 1.233650381180508, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 42076 + }, + { + "epoch": 0.42077, + "grad_norm": 1.2251810226682942, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 42077 + }, + { + "epoch": 0.42078, + "grad_norm": 0.9951664255056312, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 42078 + }, + { + "epoch": 0.42079, + "grad_norm": 1.1382962248695743, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 42079 + }, + { + "epoch": 0.4208, + "grad_norm": 1.0848415093070127, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 42080 + }, + { + "epoch": 0.42081, + "grad_norm": 1.1614127681013373, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 42081 + }, + { + "epoch": 0.42082, + "grad_norm": 0.9951114463579236, + "learning_rate": 0.003, + "loss": 4.049, + "step": 42082 + }, + { + "epoch": 0.42083, + "grad_norm": 1.263105224958204, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 42083 + }, + { + "epoch": 0.42084, + "grad_norm": 0.9164266695436776, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 42084 + }, + { + "epoch": 0.42085, + "grad_norm": 1.0525816103583832, + "learning_rate": 0.003, + "loss": 3.996, + "step": 42085 + }, + { + "epoch": 0.42086, + "grad_norm": 1.0178063352471352, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 42086 + }, + { + "epoch": 0.42087, + "grad_norm": 1.0478377380807926, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 42087 + }, + { + "epoch": 0.42088, + "grad_norm": 1.1419785983937243, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 42088 + }, + { + "epoch": 0.42089, + "grad_norm": 1.1516673441757377, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 42089 + }, + { + "epoch": 0.4209, + "grad_norm": 0.9837490764353723, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 42090 + }, + { + "epoch": 0.42091, + "grad_norm": 1.0300292473028985, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 42091 + }, + { + "epoch": 0.42092, + "grad_norm": 1.1746907617779982, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 42092 + }, + { + "epoch": 0.42093, + "grad_norm": 0.9752838044041939, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 42093 + }, + { + "epoch": 0.42094, + "grad_norm": 1.2402469056347174, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 42094 + }, + { + "epoch": 0.42095, + "grad_norm": 1.1178453896680571, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 42095 + }, + { + "epoch": 0.42096, + "grad_norm": 1.11525008793487, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 42096 + }, + { + "epoch": 0.42097, + "grad_norm": 1.106799092552223, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 42097 + }, + { + "epoch": 0.42098, + "grad_norm": 1.2259514543650156, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 42098 + }, + { + "epoch": 0.42099, + "grad_norm": 0.9601606066351712, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 42099 + }, + { + "epoch": 0.421, + "grad_norm": 1.1772625733545918, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 42100 + }, + { + "epoch": 0.42101, + "grad_norm": 1.1163509715919822, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 42101 + }, + { + "epoch": 0.42102, + "grad_norm": 0.9742477643503067, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 42102 + }, + { + "epoch": 0.42103, + "grad_norm": 0.877660846452723, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 42103 + }, + { + "epoch": 0.42104, + "grad_norm": 0.9664479118268783, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 42104 + }, + { + "epoch": 0.42105, + "grad_norm": 1.2689727167322027, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 42105 + }, + { + "epoch": 0.42106, + "grad_norm": 0.9884244543535396, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 42106 + }, + { + "epoch": 0.42107, + "grad_norm": 1.2179997371751707, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 42107 + }, + { + "epoch": 0.42108, + "grad_norm": 0.9517470223932123, + "learning_rate": 0.003, + "loss": 4.039, + "step": 42108 + }, + { + "epoch": 0.42109, + "grad_norm": 1.0325856578535544, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 42109 + }, + { + "epoch": 0.4211, + "grad_norm": 1.3367668082056925, + "learning_rate": 0.003, + "loss": 4.03, + "step": 42110 + }, + { + "epoch": 0.42111, + "grad_norm": 0.744570858970535, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 42111 + }, + { + "epoch": 0.42112, + "grad_norm": 0.7884686325966401, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 42112 + }, + { + "epoch": 0.42113, + "grad_norm": 1.2008180134957216, + "learning_rate": 0.003, + "loss": 4.044, + "step": 42113 + }, + { + "epoch": 0.42114, + "grad_norm": 1.2743792716017512, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 42114 + }, + { + "epoch": 0.42115, + "grad_norm": 0.8376514117759177, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 42115 + }, + { + "epoch": 0.42116, + "grad_norm": 0.7921088528557729, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 42116 + }, + { + "epoch": 0.42117, + "grad_norm": 0.8658382277717622, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 42117 + }, + { + "epoch": 0.42118, + "grad_norm": 1.0676157210473733, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 42118 + }, + { + "epoch": 0.42119, + "grad_norm": 1.1424368370907085, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 42119 + }, + { + "epoch": 0.4212, + "grad_norm": 1.1355019677770206, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 42120 + }, + { + "epoch": 0.42121, + "grad_norm": 1.0026032078506972, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 42121 + }, + { + "epoch": 0.42122, + "grad_norm": 1.0613177632160116, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 42122 + }, + { + "epoch": 0.42123, + "grad_norm": 1.0496496347418565, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 42123 + }, + { + "epoch": 0.42124, + "grad_norm": 1.089255091059357, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 42124 + }, + { + "epoch": 0.42125, + "grad_norm": 1.1891257183315675, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 42125 + }, + { + "epoch": 0.42126, + "grad_norm": 1.0522027022034757, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 42126 + }, + { + "epoch": 0.42127, + "grad_norm": 1.3705576582843213, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 42127 + }, + { + "epoch": 0.42128, + "grad_norm": 1.084505344011131, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 42128 + }, + { + "epoch": 0.42129, + "grad_norm": 1.0693049477536114, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 42129 + }, + { + "epoch": 0.4213, + "grad_norm": 1.04111820639467, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 42130 + }, + { + "epoch": 0.42131, + "grad_norm": 1.3203094361314331, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 42131 + }, + { + "epoch": 0.42132, + "grad_norm": 1.0114019555692817, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 42132 + }, + { + "epoch": 0.42133, + "grad_norm": 1.0251674982525494, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 42133 + }, + { + "epoch": 0.42134, + "grad_norm": 0.9491080735600644, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 42134 + }, + { + "epoch": 0.42135, + "grad_norm": 1.0784446285509741, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 42135 + }, + { + "epoch": 0.42136, + "grad_norm": 1.0973184066553534, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 42136 + }, + { + "epoch": 0.42137, + "grad_norm": 0.8953964143461484, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 42137 + }, + { + "epoch": 0.42138, + "grad_norm": 1.2428740368468338, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 42138 + }, + { + "epoch": 0.42139, + "grad_norm": 1.3526544355820849, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 42139 + }, + { + "epoch": 0.4214, + "grad_norm": 0.7828643165745209, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 42140 + }, + { + "epoch": 0.42141, + "grad_norm": 0.9283483112233982, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 42141 + }, + { + "epoch": 0.42142, + "grad_norm": 1.1483413826293, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 42142 + }, + { + "epoch": 0.42143, + "grad_norm": 1.0574072671383228, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 42143 + }, + { + "epoch": 0.42144, + "grad_norm": 1.1588012514839492, + "learning_rate": 0.003, + "loss": 4.023, + "step": 42144 + }, + { + "epoch": 0.42145, + "grad_norm": 1.0579748143284773, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 42145 + }, + { + "epoch": 0.42146, + "grad_norm": 1.2391546914888014, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 42146 + }, + { + "epoch": 0.42147, + "grad_norm": 1.0062746660450375, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 42147 + }, + { + "epoch": 0.42148, + "grad_norm": 1.0549114105113466, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 42148 + }, + { + "epoch": 0.42149, + "grad_norm": 1.0624081990776266, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 42149 + }, + { + "epoch": 0.4215, + "grad_norm": 1.2436693454684904, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 42150 + }, + { + "epoch": 0.42151, + "grad_norm": 0.9142472516029985, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 42151 + }, + { + "epoch": 0.42152, + "grad_norm": 0.8566109079597661, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 42152 + }, + { + "epoch": 0.42153, + "grad_norm": 0.8550604254771832, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 42153 + }, + { + "epoch": 0.42154, + "grad_norm": 1.0261961495954437, + "learning_rate": 0.003, + "loss": 4.006, + "step": 42154 + }, + { + "epoch": 0.42155, + "grad_norm": 1.205497121847196, + "learning_rate": 0.003, + "loss": 4.006, + "step": 42155 + }, + { + "epoch": 0.42156, + "grad_norm": 0.9281136837914846, + "learning_rate": 0.003, + "loss": 4.042, + "step": 42156 + }, + { + "epoch": 0.42157, + "grad_norm": 1.0720937765536742, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 42157 + }, + { + "epoch": 0.42158, + "grad_norm": 1.1542874112609587, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 42158 + }, + { + "epoch": 0.42159, + "grad_norm": 0.9497396763522507, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 42159 + }, + { + "epoch": 0.4216, + "grad_norm": 1.1372211474703364, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 42160 + }, + { + "epoch": 0.42161, + "grad_norm": 1.21351748815458, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 42161 + }, + { + "epoch": 0.42162, + "grad_norm": 0.9646317458876073, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 42162 + }, + { + "epoch": 0.42163, + "grad_norm": 1.1124250563993414, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 42163 + }, + { + "epoch": 0.42164, + "grad_norm": 1.1416238284938387, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 42164 + }, + { + "epoch": 0.42165, + "grad_norm": 1.0649248232652162, + "learning_rate": 0.003, + "loss": 4.037, + "step": 42165 + }, + { + "epoch": 0.42166, + "grad_norm": 1.0270870970702286, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 42166 + }, + { + "epoch": 0.42167, + "grad_norm": 0.959871264616895, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 42167 + }, + { + "epoch": 0.42168, + "grad_norm": 0.9740319532764214, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 42168 + }, + { + "epoch": 0.42169, + "grad_norm": 1.1679336524502266, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 42169 + }, + { + "epoch": 0.4217, + "grad_norm": 1.150571457988525, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 42170 + }, + { + "epoch": 0.42171, + "grad_norm": 1.1198397300784044, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 42171 + }, + { + "epoch": 0.42172, + "grad_norm": 1.3307572806551309, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 42172 + }, + { + "epoch": 0.42173, + "grad_norm": 0.6618071026690325, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 42173 + }, + { + "epoch": 0.42174, + "grad_norm": 0.8022494439228298, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 42174 + }, + { + "epoch": 0.42175, + "grad_norm": 1.1809483270876993, + "learning_rate": 0.003, + "loss": 4.058, + "step": 42175 + }, + { + "epoch": 0.42176, + "grad_norm": 1.1876007304584748, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 42176 + }, + { + "epoch": 0.42177, + "grad_norm": 1.0187467142625803, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 42177 + }, + { + "epoch": 0.42178, + "grad_norm": 0.986330839200717, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 42178 + }, + { + "epoch": 0.42179, + "grad_norm": 0.9987677578723149, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 42179 + }, + { + "epoch": 0.4218, + "grad_norm": 1.2052449328958195, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 42180 + }, + { + "epoch": 0.42181, + "grad_norm": 0.9314907109826838, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 42181 + }, + { + "epoch": 0.42182, + "grad_norm": 1.2117398013980025, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 42182 + }, + { + "epoch": 0.42183, + "grad_norm": 1.085777735396466, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 42183 + }, + { + "epoch": 0.42184, + "grad_norm": 1.0051059061826562, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 42184 + }, + { + "epoch": 0.42185, + "grad_norm": 1.0173619427733867, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 42185 + }, + { + "epoch": 0.42186, + "grad_norm": 1.082597356111565, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 42186 + }, + { + "epoch": 0.42187, + "grad_norm": 1.2331400197216391, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 42187 + }, + { + "epoch": 0.42188, + "grad_norm": 0.9964849028149714, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 42188 + }, + { + "epoch": 0.42189, + "grad_norm": 1.17452888255905, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 42189 + }, + { + "epoch": 0.4219, + "grad_norm": 1.0110136767793048, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 42190 + }, + { + "epoch": 0.42191, + "grad_norm": 1.0952903298012828, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 42191 + }, + { + "epoch": 0.42192, + "grad_norm": 1.0788923551206295, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 42192 + }, + { + "epoch": 0.42193, + "grad_norm": 1.0358312549842101, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 42193 + }, + { + "epoch": 0.42194, + "grad_norm": 1.0077324888150219, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 42194 + }, + { + "epoch": 0.42195, + "grad_norm": 1.0444393879715612, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 42195 + }, + { + "epoch": 0.42196, + "grad_norm": 1.2003986739840156, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 42196 + }, + { + "epoch": 0.42197, + "grad_norm": 0.8974862446917719, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 42197 + }, + { + "epoch": 0.42198, + "grad_norm": 1.4527742302578748, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 42198 + }, + { + "epoch": 0.42199, + "grad_norm": 0.8289927470828239, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 42199 + }, + { + "epoch": 0.422, + "grad_norm": 0.6909671080508306, + "learning_rate": 0.003, + "loss": 4.06, + "step": 42200 + }, + { + "epoch": 0.42201, + "grad_norm": 0.8101477025103243, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 42201 + }, + { + "epoch": 0.42202, + "grad_norm": 1.0704905577149095, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 42202 + }, + { + "epoch": 0.42203, + "grad_norm": 1.287040290808506, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 42203 + }, + { + "epoch": 0.42204, + "grad_norm": 0.9695058275605544, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 42204 + }, + { + "epoch": 0.42205, + "grad_norm": 1.1426675701966316, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 42205 + }, + { + "epoch": 0.42206, + "grad_norm": 1.3037260364511247, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 42206 + }, + { + "epoch": 0.42207, + "grad_norm": 0.8834079444133325, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 42207 + }, + { + "epoch": 0.42208, + "grad_norm": 0.8376746793221933, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 42208 + }, + { + "epoch": 0.42209, + "grad_norm": 0.8883977673582657, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 42209 + }, + { + "epoch": 0.4221, + "grad_norm": 0.9713064864864621, + "learning_rate": 0.003, + "loss": 4.041, + "step": 42210 + }, + { + "epoch": 0.42211, + "grad_norm": 1.0043187816876338, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 42211 + }, + { + "epoch": 0.42212, + "grad_norm": 1.2216850343361536, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 42212 + }, + { + "epoch": 0.42213, + "grad_norm": 0.9579892165373993, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 42213 + }, + { + "epoch": 0.42214, + "grad_norm": 1.1534527582395695, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 42214 + }, + { + "epoch": 0.42215, + "grad_norm": 1.1322444765904434, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 42215 + }, + { + "epoch": 0.42216, + "grad_norm": 1.1110070659343578, + "learning_rate": 0.003, + "loss": 3.996, + "step": 42216 + }, + { + "epoch": 0.42217, + "grad_norm": 0.9612669112814161, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 42217 + }, + { + "epoch": 0.42218, + "grad_norm": 1.0315495025508818, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 42218 + }, + { + "epoch": 0.42219, + "grad_norm": 1.2232709159977997, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 42219 + }, + { + "epoch": 0.4222, + "grad_norm": 0.9177294619753124, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 42220 + }, + { + "epoch": 0.42221, + "grad_norm": 0.9698751074461942, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 42221 + }, + { + "epoch": 0.42222, + "grad_norm": 1.036477698372192, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 42222 + }, + { + "epoch": 0.42223, + "grad_norm": 1.0407899464651307, + "learning_rate": 0.003, + "loss": 4.04, + "step": 42223 + }, + { + "epoch": 0.42224, + "grad_norm": 1.2441433648166182, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 42224 + }, + { + "epoch": 0.42225, + "grad_norm": 0.9553447548059498, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 42225 + }, + { + "epoch": 0.42226, + "grad_norm": 1.2476927554447343, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 42226 + }, + { + "epoch": 0.42227, + "grad_norm": 1.0058785614209838, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 42227 + }, + { + "epoch": 0.42228, + "grad_norm": 1.1183634005773224, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 42228 + }, + { + "epoch": 0.42229, + "grad_norm": 0.9808527587709799, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 42229 + }, + { + "epoch": 0.4223, + "grad_norm": 1.397339042595547, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 42230 + }, + { + "epoch": 0.42231, + "grad_norm": 0.843857760086652, + "learning_rate": 0.003, + "loss": 4.028, + "step": 42231 + }, + { + "epoch": 0.42232, + "grad_norm": 0.8593021521456159, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 42232 + }, + { + "epoch": 0.42233, + "grad_norm": 1.1740499712643424, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 42233 + }, + { + "epoch": 0.42234, + "grad_norm": 1.5299527606939518, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 42234 + }, + { + "epoch": 0.42235, + "grad_norm": 0.7691425234667296, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 42235 + }, + { + "epoch": 0.42236, + "grad_norm": 0.7220405135356596, + "learning_rate": 0.003, + "loss": 4.055, + "step": 42236 + }, + { + "epoch": 0.42237, + "grad_norm": 0.8106279185627309, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 42237 + }, + { + "epoch": 0.42238, + "grad_norm": 0.8713249037122986, + "learning_rate": 0.003, + "loss": 4.023, + "step": 42238 + }, + { + "epoch": 0.42239, + "grad_norm": 1.007033939315719, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 42239 + }, + { + "epoch": 0.4224, + "grad_norm": 1.1258218951982963, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 42240 + }, + { + "epoch": 0.42241, + "grad_norm": 0.9397305227047994, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 42241 + }, + { + "epoch": 0.42242, + "grad_norm": 1.0619397841832041, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 42242 + }, + { + "epoch": 0.42243, + "grad_norm": 1.0965324545548552, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 42243 + }, + { + "epoch": 0.42244, + "grad_norm": 1.111036367097732, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 42244 + }, + { + "epoch": 0.42245, + "grad_norm": 1.08701667933292, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 42245 + }, + { + "epoch": 0.42246, + "grad_norm": 1.2287130207989874, + "learning_rate": 0.003, + "loss": 4.026, + "step": 42246 + }, + { + "epoch": 0.42247, + "grad_norm": 1.1868402613651223, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 42247 + }, + { + "epoch": 0.42248, + "grad_norm": 1.1017217419335545, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 42248 + }, + { + "epoch": 0.42249, + "grad_norm": 1.0623736097896082, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 42249 + }, + { + "epoch": 0.4225, + "grad_norm": 1.0492239107764207, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 42250 + }, + { + "epoch": 0.42251, + "grad_norm": 1.0018543260056518, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 42251 + }, + { + "epoch": 0.42252, + "grad_norm": 1.18961987996961, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 42252 + }, + { + "epoch": 0.42253, + "grad_norm": 1.044636901801973, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 42253 + }, + { + "epoch": 0.42254, + "grad_norm": 0.9815345649424465, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 42254 + }, + { + "epoch": 0.42255, + "grad_norm": 1.0247640852936228, + "learning_rate": 0.003, + "loss": 4.042, + "step": 42255 + }, + { + "epoch": 0.42256, + "grad_norm": 0.9969889238406521, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 42256 + }, + { + "epoch": 0.42257, + "grad_norm": 1.2735909998721637, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 42257 + }, + { + "epoch": 0.42258, + "grad_norm": 0.9211033226594737, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 42258 + }, + { + "epoch": 0.42259, + "grad_norm": 1.032234802750657, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 42259 + }, + { + "epoch": 0.4226, + "grad_norm": 1.158657026579932, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 42260 + }, + { + "epoch": 0.42261, + "grad_norm": 1.0425112104806151, + "learning_rate": 0.003, + "loss": 4.057, + "step": 42261 + }, + { + "epoch": 0.42262, + "grad_norm": 0.9622730963146827, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 42262 + }, + { + "epoch": 0.42263, + "grad_norm": 0.9085212423398054, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 42263 + }, + { + "epoch": 0.42264, + "grad_norm": 0.9846308719746175, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 42264 + }, + { + "epoch": 0.42265, + "grad_norm": 1.1526061662146978, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 42265 + }, + { + "epoch": 0.42266, + "grad_norm": 1.3262796096415015, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 42266 + }, + { + "epoch": 0.42267, + "grad_norm": 0.9326229885560156, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 42267 + }, + { + "epoch": 0.42268, + "grad_norm": 0.9954549985625885, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 42268 + }, + { + "epoch": 0.42269, + "grad_norm": 1.2719090222869514, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 42269 + }, + { + "epoch": 0.4227, + "grad_norm": 1.0170291201491928, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 42270 + }, + { + "epoch": 0.42271, + "grad_norm": 1.26615823685422, + "learning_rate": 0.003, + "loss": 4.018, + "step": 42271 + }, + { + "epoch": 0.42272, + "grad_norm": 0.948943204079651, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 42272 + }, + { + "epoch": 0.42273, + "grad_norm": 0.9599726945442044, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 42273 + }, + { + "epoch": 0.42274, + "grad_norm": 1.0580076641536291, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 42274 + }, + { + "epoch": 0.42275, + "grad_norm": 1.236984327084072, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 42275 + }, + { + "epoch": 0.42276, + "grad_norm": 1.181978549010439, + "learning_rate": 0.003, + "loss": 4.041, + "step": 42276 + }, + { + "epoch": 0.42277, + "grad_norm": 1.2317862673329627, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 42277 + }, + { + "epoch": 0.42278, + "grad_norm": 0.9759350375452231, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 42278 + }, + { + "epoch": 0.42279, + "grad_norm": 1.0563701357849966, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 42279 + }, + { + "epoch": 0.4228, + "grad_norm": 1.1056023530807755, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 42280 + }, + { + "epoch": 0.42281, + "grad_norm": 0.9471895908462767, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 42281 + }, + { + "epoch": 0.42282, + "grad_norm": 1.0435669904672606, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 42282 + }, + { + "epoch": 0.42283, + "grad_norm": 1.065699704492761, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 42283 + }, + { + "epoch": 0.42284, + "grad_norm": 1.324555741500661, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 42284 + }, + { + "epoch": 0.42285, + "grad_norm": 0.8951678503749939, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 42285 + }, + { + "epoch": 0.42286, + "grad_norm": 0.9929816043860901, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 42286 + }, + { + "epoch": 0.42287, + "grad_norm": 1.0854962508783712, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 42287 + }, + { + "epoch": 0.42288, + "grad_norm": 1.0355128070557547, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 42288 + }, + { + "epoch": 0.42289, + "grad_norm": 1.277354871424866, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 42289 + }, + { + "epoch": 0.4229, + "grad_norm": 0.8228668880700989, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 42290 + }, + { + "epoch": 0.42291, + "grad_norm": 0.9358257161186457, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 42291 + }, + { + "epoch": 0.42292, + "grad_norm": 1.1345561910551378, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 42292 + }, + { + "epoch": 0.42293, + "grad_norm": 1.0944281606804918, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 42293 + }, + { + "epoch": 0.42294, + "grad_norm": 0.9369850531335271, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 42294 + }, + { + "epoch": 0.42295, + "grad_norm": 1.004257293629476, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 42295 + }, + { + "epoch": 0.42296, + "grad_norm": 1.202157641175635, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 42296 + }, + { + "epoch": 0.42297, + "grad_norm": 1.1584733506702374, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 42297 + }, + { + "epoch": 0.42298, + "grad_norm": 1.0369825960091663, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 42298 + }, + { + "epoch": 0.42299, + "grad_norm": 0.7886941487698582, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 42299 + }, + { + "epoch": 0.423, + "grad_norm": 0.8282854260064912, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 42300 + }, + { + "epoch": 0.42301, + "grad_norm": 0.9581209730049894, + "learning_rate": 0.003, + "loss": 4.026, + "step": 42301 + }, + { + "epoch": 0.42302, + "grad_norm": 1.1245712639611651, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 42302 + }, + { + "epoch": 0.42303, + "grad_norm": 1.1152037496366451, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 42303 + }, + { + "epoch": 0.42304, + "grad_norm": 1.0069815105596505, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 42304 + }, + { + "epoch": 0.42305, + "grad_norm": 1.0339342646258762, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 42305 + }, + { + "epoch": 0.42306, + "grad_norm": 1.1526030925517405, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 42306 + }, + { + "epoch": 0.42307, + "grad_norm": 0.9843049109509134, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 42307 + }, + { + "epoch": 0.42308, + "grad_norm": 1.3573757945975267, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 42308 + }, + { + "epoch": 0.42309, + "grad_norm": 1.1822364561984304, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 42309 + }, + { + "epoch": 0.4231, + "grad_norm": 0.9804684440638475, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 42310 + }, + { + "epoch": 0.42311, + "grad_norm": 1.2126169510579816, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 42311 + }, + { + "epoch": 0.42312, + "grad_norm": 1.0620305873736489, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 42312 + }, + { + "epoch": 0.42313, + "grad_norm": 0.9208811263848848, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 42313 + }, + { + "epoch": 0.42314, + "grad_norm": 1.1839305672957592, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 42314 + }, + { + "epoch": 0.42315, + "grad_norm": 1.2868697765923602, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 42315 + }, + { + "epoch": 0.42316, + "grad_norm": 1.075134443679944, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 42316 + }, + { + "epoch": 0.42317, + "grad_norm": 1.0749083846128122, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 42317 + }, + { + "epoch": 0.42318, + "grad_norm": 0.991295607816747, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 42318 + }, + { + "epoch": 0.42319, + "grad_norm": 1.0745226027782715, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 42319 + }, + { + "epoch": 0.4232, + "grad_norm": 1.144103160433686, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 42320 + }, + { + "epoch": 0.42321, + "grad_norm": 0.9599698142821516, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 42321 + }, + { + "epoch": 0.42322, + "grad_norm": 1.1640490329669613, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 42322 + }, + { + "epoch": 0.42323, + "grad_norm": 0.9192956157096874, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 42323 + }, + { + "epoch": 0.42324, + "grad_norm": 1.1561014092670883, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 42324 + }, + { + "epoch": 0.42325, + "grad_norm": 1.0336917379108512, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 42325 + }, + { + "epoch": 0.42326, + "grad_norm": 1.3186717823539267, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 42326 + }, + { + "epoch": 0.42327, + "grad_norm": 0.913510324826699, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 42327 + }, + { + "epoch": 0.42328, + "grad_norm": 0.9857024584668955, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 42328 + }, + { + "epoch": 0.42329, + "grad_norm": 1.29175635771397, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 42329 + }, + { + "epoch": 0.4233, + "grad_norm": 0.9039548895103388, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 42330 + }, + { + "epoch": 0.42331, + "grad_norm": 1.0978994261973956, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 42331 + }, + { + "epoch": 0.42332, + "grad_norm": 1.328590855224733, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 42332 + }, + { + "epoch": 0.42333, + "grad_norm": 0.857309300280483, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 42333 + }, + { + "epoch": 0.42334, + "grad_norm": 0.805565263315229, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 42334 + }, + { + "epoch": 0.42335, + "grad_norm": 0.8208636210998554, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 42335 + }, + { + "epoch": 0.42336, + "grad_norm": 0.7914592056498579, + "learning_rate": 0.003, + "loss": 4.026, + "step": 42336 + }, + { + "epoch": 0.42337, + "grad_norm": 0.8235729062236137, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 42337 + }, + { + "epoch": 0.42338, + "grad_norm": 1.0207719204654213, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 42338 + }, + { + "epoch": 0.42339, + "grad_norm": 1.1368300594667535, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 42339 + }, + { + "epoch": 0.4234, + "grad_norm": 1.0647533549602237, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 42340 + }, + { + "epoch": 0.42341, + "grad_norm": 1.0087010712815199, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 42341 + }, + { + "epoch": 0.42342, + "grad_norm": 1.2113414183927862, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 42342 + }, + { + "epoch": 0.42343, + "grad_norm": 0.9279857323436299, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 42343 + }, + { + "epoch": 0.42344, + "grad_norm": 1.0696114413399012, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 42344 + }, + { + "epoch": 0.42345, + "grad_norm": 1.0022800722284888, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 42345 + }, + { + "epoch": 0.42346, + "grad_norm": 1.1939868779977456, + "learning_rate": 0.003, + "loss": 4.035, + "step": 42346 + }, + { + "epoch": 0.42347, + "grad_norm": 0.7816107294623525, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 42347 + }, + { + "epoch": 0.42348, + "grad_norm": 0.8400091504405898, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 42348 + }, + { + "epoch": 0.42349, + "grad_norm": 0.9099194283232556, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 42349 + }, + { + "epoch": 0.4235, + "grad_norm": 1.0879761999568032, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 42350 + }, + { + "epoch": 0.42351, + "grad_norm": 1.1490823753234183, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 42351 + }, + { + "epoch": 0.42352, + "grad_norm": 1.178746080691639, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 42352 + }, + { + "epoch": 0.42353, + "grad_norm": 1.4639369101479311, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 42353 + }, + { + "epoch": 0.42354, + "grad_norm": 0.9361819945509229, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 42354 + }, + { + "epoch": 0.42355, + "grad_norm": 0.9622186443906972, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 42355 + }, + { + "epoch": 0.42356, + "grad_norm": 1.280760852549157, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 42356 + }, + { + "epoch": 0.42357, + "grad_norm": 1.2450738655912543, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 42357 + }, + { + "epoch": 0.42358, + "grad_norm": 0.9372111704446634, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 42358 + }, + { + "epoch": 0.42359, + "grad_norm": 0.9747822327765905, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 42359 + }, + { + "epoch": 0.4236, + "grad_norm": 1.1441537790997554, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 42360 + }, + { + "epoch": 0.42361, + "grad_norm": 1.014369974820975, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 42361 + }, + { + "epoch": 0.42362, + "grad_norm": 1.1620867399186856, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 42362 + }, + { + "epoch": 0.42363, + "grad_norm": 1.124723596954547, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 42363 + }, + { + "epoch": 0.42364, + "grad_norm": 1.1118015652415818, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 42364 + }, + { + "epoch": 0.42365, + "grad_norm": 0.9697643499332632, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 42365 + }, + { + "epoch": 0.42366, + "grad_norm": 1.1208217083959926, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 42366 + }, + { + "epoch": 0.42367, + "grad_norm": 1.1060560940712878, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 42367 + }, + { + "epoch": 0.42368, + "grad_norm": 1.2063423208007735, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 42368 + }, + { + "epoch": 0.42369, + "grad_norm": 0.9943040160674125, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 42369 + }, + { + "epoch": 0.4237, + "grad_norm": 1.1856886829516686, + "learning_rate": 0.003, + "loss": 3.994, + "step": 42370 + }, + { + "epoch": 0.42371, + "grad_norm": 0.9167358941752204, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 42371 + }, + { + "epoch": 0.42372, + "grad_norm": 0.8043029494686241, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 42372 + }, + { + "epoch": 0.42373, + "grad_norm": 0.7991303960843196, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 42373 + }, + { + "epoch": 0.42374, + "grad_norm": 0.9042491519703434, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 42374 + }, + { + "epoch": 0.42375, + "grad_norm": 0.9559367361522347, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 42375 + }, + { + "epoch": 0.42376, + "grad_norm": 0.9885016748455351, + "learning_rate": 0.003, + "loss": 4.044, + "step": 42376 + }, + { + "epoch": 0.42377, + "grad_norm": 0.945432524538731, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 42377 + }, + { + "epoch": 0.42378, + "grad_norm": 0.9390643631012785, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 42378 + }, + { + "epoch": 0.42379, + "grad_norm": 1.0612154914978877, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 42379 + }, + { + "epoch": 0.4238, + "grad_norm": 1.1110034767858405, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 42380 + }, + { + "epoch": 0.42381, + "grad_norm": 1.2026115110631623, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 42381 + }, + { + "epoch": 0.42382, + "grad_norm": 1.020291772908924, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 42382 + }, + { + "epoch": 0.42383, + "grad_norm": 1.1392377380981187, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 42383 + }, + { + "epoch": 0.42384, + "grad_norm": 1.1825973314622342, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 42384 + }, + { + "epoch": 0.42385, + "grad_norm": 1.3631365405507183, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 42385 + }, + { + "epoch": 0.42386, + "grad_norm": 0.794846745899579, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 42386 + }, + { + "epoch": 0.42387, + "grad_norm": 0.7281566021161859, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 42387 + }, + { + "epoch": 0.42388, + "grad_norm": 0.9762057205648407, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 42388 + }, + { + "epoch": 0.42389, + "grad_norm": 1.3596195735524383, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 42389 + }, + { + "epoch": 0.4239, + "grad_norm": 0.7057531172598073, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 42390 + }, + { + "epoch": 0.42391, + "grad_norm": 0.8924562931394695, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 42391 + }, + { + "epoch": 0.42392, + "grad_norm": 1.0395658136401478, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 42392 + }, + { + "epoch": 0.42393, + "grad_norm": 1.4125303912056646, + "learning_rate": 0.003, + "loss": 4.014, + "step": 42393 + }, + { + "epoch": 0.42394, + "grad_norm": 1.283336864567314, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 42394 + }, + { + "epoch": 0.42395, + "grad_norm": 1.011184472941377, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 42395 + }, + { + "epoch": 0.42396, + "grad_norm": 1.0391726724926138, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 42396 + }, + { + "epoch": 0.42397, + "grad_norm": 1.0587415186814948, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 42397 + }, + { + "epoch": 0.42398, + "grad_norm": 1.2006586156602896, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 42398 + }, + { + "epoch": 0.42399, + "grad_norm": 1.1421548042151886, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 42399 + }, + { + "epoch": 0.424, + "grad_norm": 1.2011856946656578, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 42400 + }, + { + "epoch": 0.42401, + "grad_norm": 0.8858619697439895, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 42401 + }, + { + "epoch": 0.42402, + "grad_norm": 0.9323513477669095, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 42402 + }, + { + "epoch": 0.42403, + "grad_norm": 1.1806537187108073, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 42403 + }, + { + "epoch": 0.42404, + "grad_norm": 1.169949063731457, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 42404 + }, + { + "epoch": 0.42405, + "grad_norm": 1.1234859988825616, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 42405 + }, + { + "epoch": 0.42406, + "grad_norm": 1.0582974476735927, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 42406 + }, + { + "epoch": 0.42407, + "grad_norm": 1.1836662323759708, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 42407 + }, + { + "epoch": 0.42408, + "grad_norm": 0.9697666678305307, + "learning_rate": 0.003, + "loss": 3.997, + "step": 42408 + }, + { + "epoch": 0.42409, + "grad_norm": 0.9907784335304544, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 42409 + }, + { + "epoch": 0.4241, + "grad_norm": 1.1637540616834845, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 42410 + }, + { + "epoch": 0.42411, + "grad_norm": 1.081685610982198, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 42411 + }, + { + "epoch": 0.42412, + "grad_norm": 1.0583497746912127, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 42412 + }, + { + "epoch": 0.42413, + "grad_norm": 0.9464257181205731, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 42413 + }, + { + "epoch": 0.42414, + "grad_norm": 0.8865236682180044, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 42414 + }, + { + "epoch": 0.42415, + "grad_norm": 1.0658503213171506, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 42415 + }, + { + "epoch": 0.42416, + "grad_norm": 1.2760213571305756, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 42416 + }, + { + "epoch": 0.42417, + "grad_norm": 1.1479591389719181, + "learning_rate": 0.003, + "loss": 4.038, + "step": 42417 + }, + { + "epoch": 0.42418, + "grad_norm": 1.0192251126781666, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 42418 + }, + { + "epoch": 0.42419, + "grad_norm": 1.1994604612545123, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 42419 + }, + { + "epoch": 0.4242, + "grad_norm": 0.9943299290573159, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 42420 + }, + { + "epoch": 0.42421, + "grad_norm": 1.0885992655578292, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 42421 + }, + { + "epoch": 0.42422, + "grad_norm": 0.9950904799629781, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 42422 + }, + { + "epoch": 0.42423, + "grad_norm": 1.1245730367107913, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 42423 + }, + { + "epoch": 0.42424, + "grad_norm": 0.9133803778191405, + "learning_rate": 0.003, + "loss": 4.041, + "step": 42424 + }, + { + "epoch": 0.42425, + "grad_norm": 1.1036662743219103, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 42425 + }, + { + "epoch": 0.42426, + "grad_norm": 1.254178603354237, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 42426 + }, + { + "epoch": 0.42427, + "grad_norm": 1.2518113799617165, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 42427 + }, + { + "epoch": 0.42428, + "grad_norm": 1.0797415339736927, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 42428 + }, + { + "epoch": 0.42429, + "grad_norm": 1.0622201398784312, + "learning_rate": 0.003, + "loss": 4.058, + "step": 42429 + }, + { + "epoch": 0.4243, + "grad_norm": 1.1773081105284455, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 42430 + }, + { + "epoch": 0.42431, + "grad_norm": 1.0876308499761311, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 42431 + }, + { + "epoch": 0.42432, + "grad_norm": 1.1324037241569864, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 42432 + }, + { + "epoch": 0.42433, + "grad_norm": 1.4936313133287682, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 42433 + }, + { + "epoch": 0.42434, + "grad_norm": 0.7573058884324125, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 42434 + }, + { + "epoch": 0.42435, + "grad_norm": 0.8002438194219522, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 42435 + }, + { + "epoch": 0.42436, + "grad_norm": 0.8766949227522927, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 42436 + }, + { + "epoch": 0.42437, + "grad_norm": 1.0384689302803596, + "learning_rate": 0.003, + "loss": 4.047, + "step": 42437 + }, + { + "epoch": 0.42438, + "grad_norm": 1.1322014445920565, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 42438 + }, + { + "epoch": 0.42439, + "grad_norm": 1.1115028607244155, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 42439 + }, + { + "epoch": 0.4244, + "grad_norm": 1.0320697835224057, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 42440 + }, + { + "epoch": 0.42441, + "grad_norm": 1.0719243227497675, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 42441 + }, + { + "epoch": 0.42442, + "grad_norm": 1.1246456757744656, + "learning_rate": 0.003, + "loss": 4.048, + "step": 42442 + }, + { + "epoch": 0.42443, + "grad_norm": 1.214383494635108, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 42443 + }, + { + "epoch": 0.42444, + "grad_norm": 1.1512000127021316, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 42444 + }, + { + "epoch": 0.42445, + "grad_norm": 0.8065224715808449, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 42445 + }, + { + "epoch": 0.42446, + "grad_norm": 0.8187691291333565, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 42446 + }, + { + "epoch": 0.42447, + "grad_norm": 1.0237525377219485, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 42447 + }, + { + "epoch": 0.42448, + "grad_norm": 1.1964234396297093, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 42448 + }, + { + "epoch": 0.42449, + "grad_norm": 1.1996473112696973, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 42449 + }, + { + "epoch": 0.4245, + "grad_norm": 0.8608546617464735, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 42450 + }, + { + "epoch": 0.42451, + "grad_norm": 0.7971842606276915, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 42451 + }, + { + "epoch": 0.42452, + "grad_norm": 0.7642145806299949, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 42452 + }, + { + "epoch": 0.42453, + "grad_norm": 1.0337107551704057, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 42453 + }, + { + "epoch": 0.42454, + "grad_norm": 1.2146069671874666, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 42454 + }, + { + "epoch": 0.42455, + "grad_norm": 1.081275218690866, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 42455 + }, + { + "epoch": 0.42456, + "grad_norm": 1.0617337949985588, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 42456 + }, + { + "epoch": 0.42457, + "grad_norm": 1.1934283721327976, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 42457 + }, + { + "epoch": 0.42458, + "grad_norm": 0.8148963763390339, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 42458 + }, + { + "epoch": 0.42459, + "grad_norm": 0.7859010235674282, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 42459 + }, + { + "epoch": 0.4246, + "grad_norm": 0.8342514128282116, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 42460 + }, + { + "epoch": 0.42461, + "grad_norm": 0.996423827242355, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 42461 + }, + { + "epoch": 0.42462, + "grad_norm": 1.2538386429409194, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 42462 + }, + { + "epoch": 0.42463, + "grad_norm": 1.1534886156073412, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 42463 + }, + { + "epoch": 0.42464, + "grad_norm": 1.2917542588780186, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 42464 + }, + { + "epoch": 0.42465, + "grad_norm": 0.8076483845157171, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 42465 + }, + { + "epoch": 0.42466, + "grad_norm": 0.8403942637161572, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 42466 + }, + { + "epoch": 0.42467, + "grad_norm": 0.9616914543572054, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 42467 + }, + { + "epoch": 0.42468, + "grad_norm": 1.1729699672014624, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 42468 + }, + { + "epoch": 0.42469, + "grad_norm": 1.1734974204799806, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 42469 + }, + { + "epoch": 0.4247, + "grad_norm": 1.3445475595333938, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 42470 + }, + { + "epoch": 0.42471, + "grad_norm": 0.8755603044439588, + "learning_rate": 0.003, + "loss": 4.046, + "step": 42471 + }, + { + "epoch": 0.42472, + "grad_norm": 0.9654881473853499, + "learning_rate": 0.003, + "loss": 4.02, + "step": 42472 + }, + { + "epoch": 0.42473, + "grad_norm": 1.3270878671400526, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 42473 + }, + { + "epoch": 0.42474, + "grad_norm": 1.0627088418281987, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 42474 + }, + { + "epoch": 0.42475, + "grad_norm": 1.5945143222302836, + "learning_rate": 0.003, + "loss": 4.023, + "step": 42475 + }, + { + "epoch": 0.42476, + "grad_norm": 0.7907091234404464, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 42476 + }, + { + "epoch": 0.42477, + "grad_norm": 0.751518286211239, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 42477 + }, + { + "epoch": 0.42478, + "grad_norm": 1.0087588219970616, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 42478 + }, + { + "epoch": 0.42479, + "grad_norm": 1.2219337064373785, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 42479 + }, + { + "epoch": 0.4248, + "grad_norm": 1.0920445642987955, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 42480 + }, + { + "epoch": 0.42481, + "grad_norm": 1.0339002595610058, + "learning_rate": 0.003, + "loss": 4.04, + "step": 42481 + }, + { + "epoch": 0.42482, + "grad_norm": 1.1289715719038387, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 42482 + }, + { + "epoch": 0.42483, + "grad_norm": 1.0543075831370952, + "learning_rate": 0.003, + "loss": 4.04, + "step": 42483 + }, + { + "epoch": 0.42484, + "grad_norm": 1.0041053302972924, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 42484 + }, + { + "epoch": 0.42485, + "grad_norm": 1.2330149970439614, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 42485 + }, + { + "epoch": 0.42486, + "grad_norm": 0.9844898532417976, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 42486 + }, + { + "epoch": 0.42487, + "grad_norm": 1.1797939685716903, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 42487 + }, + { + "epoch": 0.42488, + "grad_norm": 0.9910113034638297, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 42488 + }, + { + "epoch": 0.42489, + "grad_norm": 1.1702472034630038, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 42489 + }, + { + "epoch": 0.4249, + "grad_norm": 1.149161844913745, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 42490 + }, + { + "epoch": 0.42491, + "grad_norm": 0.981163293544716, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 42491 + }, + { + "epoch": 0.42492, + "grad_norm": 1.0908021117407474, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 42492 + }, + { + "epoch": 0.42493, + "grad_norm": 0.8705869992974179, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 42493 + }, + { + "epoch": 0.42494, + "grad_norm": 1.0313622514637617, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 42494 + }, + { + "epoch": 0.42495, + "grad_norm": 1.0004607896234083, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 42495 + }, + { + "epoch": 0.42496, + "grad_norm": 1.1112020570251482, + "learning_rate": 0.003, + "loss": 4.019, + "step": 42496 + }, + { + "epoch": 0.42497, + "grad_norm": 1.0590542333123365, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 42497 + }, + { + "epoch": 0.42498, + "grad_norm": 0.9835132066193139, + "learning_rate": 0.003, + "loss": 4.061, + "step": 42498 + }, + { + "epoch": 0.42499, + "grad_norm": 1.044777313326872, + "learning_rate": 0.003, + "loss": 4.027, + "step": 42499 + }, + { + "epoch": 0.425, + "grad_norm": 1.0032770462428622, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 42500 + }, + { + "epoch": 0.42501, + "grad_norm": 1.175262944150529, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 42501 + }, + { + "epoch": 0.42502, + "grad_norm": 1.068315382898784, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 42502 + }, + { + "epoch": 0.42503, + "grad_norm": 1.4587695424374478, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 42503 + }, + { + "epoch": 0.42504, + "grad_norm": 1.059150573883668, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 42504 + }, + { + "epoch": 0.42505, + "grad_norm": 1.215981477696914, + "learning_rate": 0.003, + "loss": 4.047, + "step": 42505 + }, + { + "epoch": 0.42506, + "grad_norm": 0.8951925952622185, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 42506 + }, + { + "epoch": 0.42507, + "grad_norm": 0.891787573172901, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 42507 + }, + { + "epoch": 0.42508, + "grad_norm": 1.0219981922688894, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 42508 + }, + { + "epoch": 0.42509, + "grad_norm": 1.641198083295391, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 42509 + }, + { + "epoch": 0.4251, + "grad_norm": 0.7638965373236184, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 42510 + }, + { + "epoch": 0.42511, + "grad_norm": 0.9298779947092383, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 42511 + }, + { + "epoch": 0.42512, + "grad_norm": 1.150271963523183, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 42512 + }, + { + "epoch": 0.42513, + "grad_norm": 0.9010491146054889, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 42513 + }, + { + "epoch": 0.42514, + "grad_norm": 0.9664353616807151, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 42514 + }, + { + "epoch": 0.42515, + "grad_norm": 1.1383310766044585, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 42515 + }, + { + "epoch": 0.42516, + "grad_norm": 1.32866621581628, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 42516 + }, + { + "epoch": 0.42517, + "grad_norm": 0.8061473278572548, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 42517 + }, + { + "epoch": 0.42518, + "grad_norm": 0.8007538533494015, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 42518 + }, + { + "epoch": 0.42519, + "grad_norm": 1.0634774740880133, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 42519 + }, + { + "epoch": 0.4252, + "grad_norm": 1.3521176162983108, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 42520 + }, + { + "epoch": 0.42521, + "grad_norm": 0.879227326188695, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 42521 + }, + { + "epoch": 0.42522, + "grad_norm": 1.0749965423781767, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 42522 + }, + { + "epoch": 0.42523, + "grad_norm": 1.311573431711792, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 42523 + }, + { + "epoch": 0.42524, + "grad_norm": 0.9472500620213419, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 42524 + }, + { + "epoch": 0.42525, + "grad_norm": 1.1746952984372665, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 42525 + }, + { + "epoch": 0.42526, + "grad_norm": 1.1023851891979863, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 42526 + }, + { + "epoch": 0.42527, + "grad_norm": 1.0998725366950477, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 42527 + }, + { + "epoch": 0.42528, + "grad_norm": 1.155497559934494, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 42528 + }, + { + "epoch": 0.42529, + "grad_norm": 1.0348441333812815, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 42529 + }, + { + "epoch": 0.4253, + "grad_norm": 1.21464279514076, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 42530 + }, + { + "epoch": 0.42531, + "grad_norm": 0.9665575261057068, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 42531 + }, + { + "epoch": 0.42532, + "grad_norm": 1.0636289829129324, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 42532 + }, + { + "epoch": 0.42533, + "grad_norm": 1.2433282825025949, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 42533 + }, + { + "epoch": 0.42534, + "grad_norm": 1.076548153046899, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 42534 + }, + { + "epoch": 0.42535, + "grad_norm": 0.9168240867061213, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 42535 + }, + { + "epoch": 0.42536, + "grad_norm": 0.9639320715957947, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 42536 + }, + { + "epoch": 0.42537, + "grad_norm": 1.2159126786110446, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 42537 + }, + { + "epoch": 0.42538, + "grad_norm": 1.3167702461785915, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 42538 + }, + { + "epoch": 0.42539, + "grad_norm": 0.8829444214287933, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 42539 + }, + { + "epoch": 0.4254, + "grad_norm": 0.8815406969407675, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 42540 + }, + { + "epoch": 0.42541, + "grad_norm": 1.0266381100742508, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 42541 + }, + { + "epoch": 0.42542, + "grad_norm": 1.3433402698727595, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 42542 + }, + { + "epoch": 0.42543, + "grad_norm": 0.9087087316536545, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 42543 + }, + { + "epoch": 0.42544, + "grad_norm": 1.18972329623036, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 42544 + }, + { + "epoch": 0.42545, + "grad_norm": 1.2054488227014701, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 42545 + }, + { + "epoch": 0.42546, + "grad_norm": 0.992096007265317, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 42546 + }, + { + "epoch": 0.42547, + "grad_norm": 0.9670109261720755, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 42547 + }, + { + "epoch": 0.42548, + "grad_norm": 1.0012468528146021, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 42548 + }, + { + "epoch": 0.42549, + "grad_norm": 1.2169130190048754, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 42549 + }, + { + "epoch": 0.4255, + "grad_norm": 1.1476863036304097, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 42550 + }, + { + "epoch": 0.42551, + "grad_norm": 1.0229698800433713, + "learning_rate": 0.003, + "loss": 4.025, + "step": 42551 + }, + { + "epoch": 0.42552, + "grad_norm": 1.1461332199836862, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 42552 + }, + { + "epoch": 0.42553, + "grad_norm": 0.9552696003640881, + "learning_rate": 0.003, + "loss": 4.082, + "step": 42553 + }, + { + "epoch": 0.42554, + "grad_norm": 0.9881968721122979, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 42554 + }, + { + "epoch": 0.42555, + "grad_norm": 1.1001094251498578, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 42555 + }, + { + "epoch": 0.42556, + "grad_norm": 1.1954282094601474, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 42556 + }, + { + "epoch": 0.42557, + "grad_norm": 0.9434948635448327, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 42557 + }, + { + "epoch": 0.42558, + "grad_norm": 1.1517605979151777, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 42558 + }, + { + "epoch": 0.42559, + "grad_norm": 1.3225423418711804, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 42559 + }, + { + "epoch": 0.4256, + "grad_norm": 0.7655122185023288, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 42560 + }, + { + "epoch": 0.42561, + "grad_norm": 0.7188503365017326, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 42561 + }, + { + "epoch": 0.42562, + "grad_norm": 0.8142372804009679, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 42562 + }, + { + "epoch": 0.42563, + "grad_norm": 1.0239765399626213, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 42563 + }, + { + "epoch": 0.42564, + "grad_norm": 1.5221446078289467, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 42564 + }, + { + "epoch": 0.42565, + "grad_norm": 0.7055086970088801, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 42565 + }, + { + "epoch": 0.42566, + "grad_norm": 0.8428647607758508, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 42566 + }, + { + "epoch": 0.42567, + "grad_norm": 1.0473667738526244, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 42567 + }, + { + "epoch": 0.42568, + "grad_norm": 1.2548407095657312, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 42568 + }, + { + "epoch": 0.42569, + "grad_norm": 0.96743385566155, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 42569 + }, + { + "epoch": 0.4257, + "grad_norm": 0.9521619636891782, + "learning_rate": 0.003, + "loss": 4.027, + "step": 42570 + }, + { + "epoch": 0.42571, + "grad_norm": 1.3778248819862238, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 42571 + }, + { + "epoch": 0.42572, + "grad_norm": 1.0715671133673321, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 42572 + }, + { + "epoch": 0.42573, + "grad_norm": 1.0007328421383037, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 42573 + }, + { + "epoch": 0.42574, + "grad_norm": 1.0912780555937711, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 42574 + }, + { + "epoch": 0.42575, + "grad_norm": 1.1576359816744284, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 42575 + }, + { + "epoch": 0.42576, + "grad_norm": 0.971196916541871, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 42576 + }, + { + "epoch": 0.42577, + "grad_norm": 1.0395596490417665, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 42577 + }, + { + "epoch": 0.42578, + "grad_norm": 1.3418543072915685, + "learning_rate": 0.003, + "loss": 4.036, + "step": 42578 + }, + { + "epoch": 0.42579, + "grad_norm": 0.9138326445787074, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 42579 + }, + { + "epoch": 0.4258, + "grad_norm": 0.9793910841169685, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 42580 + }, + { + "epoch": 0.42581, + "grad_norm": 1.3433862229502442, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 42581 + }, + { + "epoch": 0.42582, + "grad_norm": 0.8494709139253378, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 42582 + }, + { + "epoch": 0.42583, + "grad_norm": 0.8262713216747117, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 42583 + }, + { + "epoch": 0.42584, + "grad_norm": 1.0145829715141215, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 42584 + }, + { + "epoch": 0.42585, + "grad_norm": 1.263847409933273, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 42585 + }, + { + "epoch": 0.42586, + "grad_norm": 1.086135375705896, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 42586 + }, + { + "epoch": 0.42587, + "grad_norm": 1.37223133846214, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 42587 + }, + { + "epoch": 0.42588, + "grad_norm": 0.794819495529008, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 42588 + }, + { + "epoch": 0.42589, + "grad_norm": 0.7095506618312641, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 42589 + }, + { + "epoch": 0.4259, + "grad_norm": 1.005542141282413, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 42590 + }, + { + "epoch": 0.42591, + "grad_norm": 1.3416618636330542, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 42591 + }, + { + "epoch": 0.42592, + "grad_norm": 1.052676096475043, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 42592 + }, + { + "epoch": 0.42593, + "grad_norm": 1.111773251503904, + "learning_rate": 0.003, + "loss": 4.049, + "step": 42593 + }, + { + "epoch": 0.42594, + "grad_norm": 1.0262838144770112, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 42594 + }, + { + "epoch": 0.42595, + "grad_norm": 1.1671236816023474, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 42595 + }, + { + "epoch": 0.42596, + "grad_norm": 0.9094801428574251, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 42596 + }, + { + "epoch": 0.42597, + "grad_norm": 1.0389134045152302, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 42597 + }, + { + "epoch": 0.42598, + "grad_norm": 0.9792030326478394, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 42598 + }, + { + "epoch": 0.42599, + "grad_norm": 1.000158521270124, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 42599 + }, + { + "epoch": 0.426, + "grad_norm": 1.1792247186395433, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 42600 + }, + { + "epoch": 0.42601, + "grad_norm": 0.9670630369686303, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 42601 + }, + { + "epoch": 0.42602, + "grad_norm": 1.274510306383337, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 42602 + }, + { + "epoch": 0.42603, + "grad_norm": 0.8739123301589996, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 42603 + }, + { + "epoch": 0.42604, + "grad_norm": 0.9977416372006914, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 42604 + }, + { + "epoch": 0.42605, + "grad_norm": 1.2677141012100237, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 42605 + }, + { + "epoch": 0.42606, + "grad_norm": 1.0426445144370307, + "learning_rate": 0.003, + "loss": 4.028, + "step": 42606 + }, + { + "epoch": 0.42607, + "grad_norm": 0.9984577201925157, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 42607 + }, + { + "epoch": 0.42608, + "grad_norm": 1.0218103659522353, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 42608 + }, + { + "epoch": 0.42609, + "grad_norm": 1.2657644123321428, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 42609 + }, + { + "epoch": 0.4261, + "grad_norm": 0.9100562760235377, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 42610 + }, + { + "epoch": 0.42611, + "grad_norm": 1.0955542370803812, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 42611 + }, + { + "epoch": 0.42612, + "grad_norm": 1.6323754407982396, + "learning_rate": 0.003, + "loss": 4.048, + "step": 42612 + }, + { + "epoch": 0.42613, + "grad_norm": 1.0302955722329414, + "learning_rate": 0.003, + "loss": 4.039, + "step": 42613 + }, + { + "epoch": 0.42614, + "grad_norm": 1.137661397184626, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 42614 + }, + { + "epoch": 0.42615, + "grad_norm": 1.185107695112903, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 42615 + }, + { + "epoch": 0.42616, + "grad_norm": 1.3313873567612986, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 42616 + }, + { + "epoch": 0.42617, + "grad_norm": 0.9426226914220577, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 42617 + }, + { + "epoch": 0.42618, + "grad_norm": 1.2467362646899074, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 42618 + }, + { + "epoch": 0.42619, + "grad_norm": 1.2049695493603743, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 42619 + }, + { + "epoch": 0.4262, + "grad_norm": 0.8573396134384669, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 42620 + }, + { + "epoch": 0.42621, + "grad_norm": 1.0845460642667266, + "learning_rate": 0.003, + "loss": 4.013, + "step": 42621 + }, + { + "epoch": 0.42622, + "grad_norm": 1.020793695135295, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 42622 + }, + { + "epoch": 0.42623, + "grad_norm": 0.9200662937526883, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 42623 + }, + { + "epoch": 0.42624, + "grad_norm": 1.0480180890678803, + "learning_rate": 0.003, + "loss": 4.023, + "step": 42624 + }, + { + "epoch": 0.42625, + "grad_norm": 1.2371879829528074, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 42625 + }, + { + "epoch": 0.42626, + "grad_norm": 0.9421680772809784, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 42626 + }, + { + "epoch": 0.42627, + "grad_norm": 1.2704642148858305, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 42627 + }, + { + "epoch": 0.42628, + "grad_norm": 1.1628875783677814, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 42628 + }, + { + "epoch": 0.42629, + "grad_norm": 1.0757822589580943, + "learning_rate": 0.003, + "loss": 4.029, + "step": 42629 + }, + { + "epoch": 0.4263, + "grad_norm": 1.0207741062604936, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 42630 + }, + { + "epoch": 0.42631, + "grad_norm": 1.0520130970360895, + "learning_rate": 0.003, + "loss": 4.033, + "step": 42631 + }, + { + "epoch": 0.42632, + "grad_norm": 1.011221202416192, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 42632 + }, + { + "epoch": 0.42633, + "grad_norm": 1.0775913810777686, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 42633 + }, + { + "epoch": 0.42634, + "grad_norm": 0.9512792568502407, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 42634 + }, + { + "epoch": 0.42635, + "grad_norm": 0.9395894130589807, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 42635 + }, + { + "epoch": 0.42636, + "grad_norm": 1.0116581100901743, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 42636 + }, + { + "epoch": 0.42637, + "grad_norm": 0.950582065338779, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 42637 + }, + { + "epoch": 0.42638, + "grad_norm": 0.9263813168180461, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 42638 + }, + { + "epoch": 0.42639, + "grad_norm": 1.2713740695414544, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 42639 + }, + { + "epoch": 0.4264, + "grad_norm": 1.205307298231729, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 42640 + }, + { + "epoch": 0.42641, + "grad_norm": 1.1164607128485642, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 42641 + }, + { + "epoch": 0.42642, + "grad_norm": 1.1193460450181332, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 42642 + }, + { + "epoch": 0.42643, + "grad_norm": 1.1443053721032452, + "learning_rate": 0.003, + "loss": 4.03, + "step": 42643 + }, + { + "epoch": 0.42644, + "grad_norm": 1.0649640985287534, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 42644 + }, + { + "epoch": 0.42645, + "grad_norm": 1.143379664592755, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 42645 + }, + { + "epoch": 0.42646, + "grad_norm": 1.1463343921548148, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 42646 + }, + { + "epoch": 0.42647, + "grad_norm": 0.9829652237575083, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 42647 + }, + { + "epoch": 0.42648, + "grad_norm": 1.0434317417647048, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 42648 + }, + { + "epoch": 0.42649, + "grad_norm": 1.0227099459980424, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 42649 + }, + { + "epoch": 0.4265, + "grad_norm": 1.0364155414521172, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 42650 + }, + { + "epoch": 0.42651, + "grad_norm": 0.9562079584772162, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 42651 + }, + { + "epoch": 0.42652, + "grad_norm": 1.1144494527836584, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 42652 + }, + { + "epoch": 0.42653, + "grad_norm": 0.9884525474526575, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 42653 + }, + { + "epoch": 0.42654, + "grad_norm": 1.0589524005500524, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 42654 + }, + { + "epoch": 0.42655, + "grad_norm": 1.1133365890400249, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 42655 + }, + { + "epoch": 0.42656, + "grad_norm": 1.1118029324715188, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 42656 + }, + { + "epoch": 0.42657, + "grad_norm": 0.895969616792602, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 42657 + }, + { + "epoch": 0.42658, + "grad_norm": 0.915833943765889, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 42658 + }, + { + "epoch": 0.42659, + "grad_norm": 1.0306251678829748, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 42659 + }, + { + "epoch": 0.4266, + "grad_norm": 0.9988228720542374, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 42660 + }, + { + "epoch": 0.42661, + "grad_norm": 0.9568879226798238, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 42661 + }, + { + "epoch": 0.42662, + "grad_norm": 0.9265029642165963, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 42662 + }, + { + "epoch": 0.42663, + "grad_norm": 1.0150522732670406, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 42663 + }, + { + "epoch": 0.42664, + "grad_norm": 1.1357938399179233, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 42664 + }, + { + "epoch": 0.42665, + "grad_norm": 1.1010708889810459, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 42665 + }, + { + "epoch": 0.42666, + "grad_norm": 1.1069643129604194, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 42666 + }, + { + "epoch": 0.42667, + "grad_norm": 1.4235266648578138, + "learning_rate": 0.003, + "loss": 4.033, + "step": 42667 + }, + { + "epoch": 0.42668, + "grad_norm": 1.042798413233834, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 42668 + }, + { + "epoch": 0.42669, + "grad_norm": 0.9919468410632152, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 42669 + }, + { + "epoch": 0.4267, + "grad_norm": 1.221308669530015, + "learning_rate": 0.003, + "loss": 4.051, + "step": 42670 + }, + { + "epoch": 0.42671, + "grad_norm": 1.2412198722350505, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 42671 + }, + { + "epoch": 0.42672, + "grad_norm": 0.9951185209146731, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 42672 + }, + { + "epoch": 0.42673, + "grad_norm": 1.049621330963539, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 42673 + }, + { + "epoch": 0.42674, + "grad_norm": 1.079575156308222, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 42674 + }, + { + "epoch": 0.42675, + "grad_norm": 0.902571273171723, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 42675 + }, + { + "epoch": 0.42676, + "grad_norm": 1.13706681775121, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 42676 + }, + { + "epoch": 0.42677, + "grad_norm": 1.1621905183867587, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 42677 + }, + { + "epoch": 0.42678, + "grad_norm": 0.9561203682358462, + "learning_rate": 0.003, + "loss": 4.001, + "step": 42678 + }, + { + "epoch": 0.42679, + "grad_norm": 1.0143349153406411, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 42679 + }, + { + "epoch": 0.4268, + "grad_norm": 1.080948993123185, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 42680 + }, + { + "epoch": 0.42681, + "grad_norm": 1.0375359780969324, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 42681 + }, + { + "epoch": 0.42682, + "grad_norm": 1.0921448368019087, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 42682 + }, + { + "epoch": 0.42683, + "grad_norm": 1.1254084205017905, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 42683 + }, + { + "epoch": 0.42684, + "grad_norm": 1.40108244794107, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 42684 + }, + { + "epoch": 0.42685, + "grad_norm": 1.0219290138340271, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 42685 + }, + { + "epoch": 0.42686, + "grad_norm": 1.1308837417363318, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 42686 + }, + { + "epoch": 0.42687, + "grad_norm": 1.2636368383066543, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 42687 + }, + { + "epoch": 0.42688, + "grad_norm": 1.0473835261501367, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 42688 + }, + { + "epoch": 0.42689, + "grad_norm": 1.365221666831375, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 42689 + }, + { + "epoch": 0.4269, + "grad_norm": 0.7904389068116512, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 42690 + }, + { + "epoch": 0.42691, + "grad_norm": 0.722855819032758, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 42691 + }, + { + "epoch": 0.42692, + "grad_norm": 0.7445534709682808, + "learning_rate": 0.003, + "loss": 4.027, + "step": 42692 + }, + { + "epoch": 0.42693, + "grad_norm": 0.9657211942387255, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 42693 + }, + { + "epoch": 0.42694, + "grad_norm": 1.0710135637392286, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 42694 + }, + { + "epoch": 0.42695, + "grad_norm": 1.0194609248789561, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 42695 + }, + { + "epoch": 0.42696, + "grad_norm": 1.0220924264976146, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 42696 + }, + { + "epoch": 0.42697, + "grad_norm": 0.9747390272004394, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 42697 + }, + { + "epoch": 0.42698, + "grad_norm": 1.007429915330194, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 42698 + }, + { + "epoch": 0.42699, + "grad_norm": 1.1087281370390514, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 42699 + }, + { + "epoch": 0.427, + "grad_norm": 1.1088527206549235, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 42700 + }, + { + "epoch": 0.42701, + "grad_norm": 1.0662284002876208, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 42701 + }, + { + "epoch": 0.42702, + "grad_norm": 1.1003666418940072, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 42702 + }, + { + "epoch": 0.42703, + "grad_norm": 0.8835524415149618, + "learning_rate": 0.003, + "loss": 4.029, + "step": 42703 + }, + { + "epoch": 0.42704, + "grad_norm": 1.0176799785809865, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 42704 + }, + { + "epoch": 0.42705, + "grad_norm": 1.1269957961134354, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 42705 + }, + { + "epoch": 0.42706, + "grad_norm": 0.9464589234190084, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 42706 + }, + { + "epoch": 0.42707, + "grad_norm": 1.0787095356697114, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 42707 + }, + { + "epoch": 0.42708, + "grad_norm": 0.9371037309469175, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 42708 + }, + { + "epoch": 0.42709, + "grad_norm": 0.9654085794551579, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 42709 + }, + { + "epoch": 0.4271, + "grad_norm": 1.1295311364703373, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 42710 + }, + { + "epoch": 0.42711, + "grad_norm": 1.0706578204585029, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 42711 + }, + { + "epoch": 0.42712, + "grad_norm": 0.9236100809298715, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 42712 + }, + { + "epoch": 0.42713, + "grad_norm": 1.1030555189062785, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 42713 + }, + { + "epoch": 0.42714, + "grad_norm": 1.1540692193133506, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 42714 + }, + { + "epoch": 0.42715, + "grad_norm": 1.295432731047244, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 42715 + }, + { + "epoch": 0.42716, + "grad_norm": 0.8589696871367923, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 42716 + }, + { + "epoch": 0.42717, + "grad_norm": 0.8648963600218567, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 42717 + }, + { + "epoch": 0.42718, + "grad_norm": 1.2076230718818055, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 42718 + }, + { + "epoch": 0.42719, + "grad_norm": 1.5560710444535766, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 42719 + }, + { + "epoch": 0.4272, + "grad_norm": 0.7663745967713846, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 42720 + }, + { + "epoch": 0.42721, + "grad_norm": 0.9075212483080595, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 42721 + }, + { + "epoch": 0.42722, + "grad_norm": 1.2058254328448437, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 42722 + }, + { + "epoch": 0.42723, + "grad_norm": 1.1273623953325773, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 42723 + }, + { + "epoch": 0.42724, + "grad_norm": 0.9622438205906119, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 42724 + }, + { + "epoch": 0.42725, + "grad_norm": 1.0314493809614236, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 42725 + }, + { + "epoch": 0.42726, + "grad_norm": 1.1273813993970463, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 42726 + }, + { + "epoch": 0.42727, + "grad_norm": 1.117512942669241, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 42727 + }, + { + "epoch": 0.42728, + "grad_norm": 1.191593358681951, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 42728 + }, + { + "epoch": 0.42729, + "grad_norm": 1.0453931272640251, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 42729 + }, + { + "epoch": 0.4273, + "grad_norm": 1.0716578189344625, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 42730 + }, + { + "epoch": 0.42731, + "grad_norm": 1.2348513400505479, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 42731 + }, + { + "epoch": 0.42732, + "grad_norm": 1.0132337313474138, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 42732 + }, + { + "epoch": 0.42733, + "grad_norm": 1.1020935196964625, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 42733 + }, + { + "epoch": 0.42734, + "grad_norm": 1.0949127308654223, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 42734 + }, + { + "epoch": 0.42735, + "grad_norm": 1.0034210568584199, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 42735 + }, + { + "epoch": 0.42736, + "grad_norm": 1.1930349105879476, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 42736 + }, + { + "epoch": 0.42737, + "grad_norm": 0.8973597476418976, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 42737 + }, + { + "epoch": 0.42738, + "grad_norm": 0.8768522893485892, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 42738 + }, + { + "epoch": 0.42739, + "grad_norm": 0.9248541996230437, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 42739 + }, + { + "epoch": 0.4274, + "grad_norm": 0.9209204760303297, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 42740 + }, + { + "epoch": 0.42741, + "grad_norm": 1.05194502724585, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 42741 + }, + { + "epoch": 0.42742, + "grad_norm": 1.2053866804435154, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 42742 + }, + { + "epoch": 0.42743, + "grad_norm": 0.909540011256576, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 42743 + }, + { + "epoch": 0.42744, + "grad_norm": 1.0379691896970755, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 42744 + }, + { + "epoch": 0.42745, + "grad_norm": 1.5421490173978372, + "learning_rate": 0.003, + "loss": 4.1, + "step": 42745 + }, + { + "epoch": 0.42746, + "grad_norm": 0.7871118235935087, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 42746 + }, + { + "epoch": 0.42747, + "grad_norm": 0.8316990601800719, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 42747 + }, + { + "epoch": 0.42748, + "grad_norm": 1.0091708893016076, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 42748 + }, + { + "epoch": 0.42749, + "grad_norm": 1.0880268387774812, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 42749 + }, + { + "epoch": 0.4275, + "grad_norm": 1.0360926224524667, + "learning_rate": 0.003, + "loss": 4.044, + "step": 42750 + }, + { + "epoch": 0.42751, + "grad_norm": 1.1106094398993627, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 42751 + }, + { + "epoch": 0.42752, + "grad_norm": 1.1737122786529517, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 42752 + }, + { + "epoch": 0.42753, + "grad_norm": 1.1651513072853328, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 42753 + }, + { + "epoch": 0.42754, + "grad_norm": 1.2185224210056358, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 42754 + }, + { + "epoch": 0.42755, + "grad_norm": 0.9055486923819954, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 42755 + }, + { + "epoch": 0.42756, + "grad_norm": 1.0678762172785594, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 42756 + }, + { + "epoch": 0.42757, + "grad_norm": 1.2840735670473198, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 42757 + }, + { + "epoch": 0.42758, + "grad_norm": 1.0105136395037082, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 42758 + }, + { + "epoch": 0.42759, + "grad_norm": 1.224958782597126, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 42759 + }, + { + "epoch": 0.4276, + "grad_norm": 1.1018724239116764, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 42760 + }, + { + "epoch": 0.42761, + "grad_norm": 1.1967709578808008, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 42761 + }, + { + "epoch": 0.42762, + "grad_norm": 0.9732543618350747, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 42762 + }, + { + "epoch": 0.42763, + "grad_norm": 1.1823832897509545, + "learning_rate": 0.003, + "loss": 4.046, + "step": 42763 + }, + { + "epoch": 0.42764, + "grad_norm": 0.9684176957282703, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 42764 + }, + { + "epoch": 0.42765, + "grad_norm": 0.8992047835159295, + "learning_rate": 0.003, + "loss": 4.032, + "step": 42765 + }, + { + "epoch": 0.42766, + "grad_norm": 0.9453187373189816, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 42766 + }, + { + "epoch": 0.42767, + "grad_norm": 1.0415760045614517, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 42767 + }, + { + "epoch": 0.42768, + "grad_norm": 1.062371066038506, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 42768 + }, + { + "epoch": 0.42769, + "grad_norm": 1.1263254231541797, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 42769 + }, + { + "epoch": 0.4277, + "grad_norm": 1.0440905706546686, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 42770 + }, + { + "epoch": 0.42771, + "grad_norm": 1.2481137263272617, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 42771 + }, + { + "epoch": 0.42772, + "grad_norm": 0.8293729996794204, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 42772 + }, + { + "epoch": 0.42773, + "grad_norm": 0.8112000539362804, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 42773 + }, + { + "epoch": 0.42774, + "grad_norm": 0.8936958904069536, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 42774 + }, + { + "epoch": 0.42775, + "grad_norm": 0.9752204659450803, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 42775 + }, + { + "epoch": 0.42776, + "grad_norm": 0.9900810204256811, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 42776 + }, + { + "epoch": 0.42777, + "grad_norm": 0.9646074939487091, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 42777 + }, + { + "epoch": 0.42778, + "grad_norm": 1.072838575583213, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 42778 + }, + { + "epoch": 0.42779, + "grad_norm": 1.0137907840096032, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 42779 + }, + { + "epoch": 0.4278, + "grad_norm": 1.2966206443795898, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 42780 + }, + { + "epoch": 0.42781, + "grad_norm": 0.9166694173214953, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 42781 + }, + { + "epoch": 0.42782, + "grad_norm": 0.954954096948651, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 42782 + }, + { + "epoch": 0.42783, + "grad_norm": 1.053398383743642, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 42783 + }, + { + "epoch": 0.42784, + "grad_norm": 1.1145581614222322, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 42784 + }, + { + "epoch": 0.42785, + "grad_norm": 1.0885847876161066, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 42785 + }, + { + "epoch": 0.42786, + "grad_norm": 1.0794632069582322, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 42786 + }, + { + "epoch": 0.42787, + "grad_norm": 0.959306609147431, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 42787 + }, + { + "epoch": 0.42788, + "grad_norm": 1.0689488486451184, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 42788 + }, + { + "epoch": 0.42789, + "grad_norm": 1.1146237548161428, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 42789 + }, + { + "epoch": 0.4279, + "grad_norm": 0.9737335337261365, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 42790 + }, + { + "epoch": 0.42791, + "grad_norm": 1.0898186377503787, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 42791 + }, + { + "epoch": 0.42792, + "grad_norm": 1.2771343549899035, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 42792 + }, + { + "epoch": 0.42793, + "grad_norm": 1.3726996481040767, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 42793 + }, + { + "epoch": 0.42794, + "grad_norm": 0.7833084237821929, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 42794 + }, + { + "epoch": 0.42795, + "grad_norm": 0.7167090752741165, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 42795 + }, + { + "epoch": 0.42796, + "grad_norm": 0.7874554094796617, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 42796 + }, + { + "epoch": 0.42797, + "grad_norm": 0.9353000503025556, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 42797 + }, + { + "epoch": 0.42798, + "grad_norm": 1.123830788809233, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 42798 + }, + { + "epoch": 0.42799, + "grad_norm": 1.0419954688929205, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 42799 + }, + { + "epoch": 0.428, + "grad_norm": 1.235436604906963, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 42800 + }, + { + "epoch": 0.42801, + "grad_norm": 0.9094507497878276, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 42801 + }, + { + "epoch": 0.42802, + "grad_norm": 1.0971770434455057, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 42802 + }, + { + "epoch": 0.42803, + "grad_norm": 1.0811814031849953, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 42803 + }, + { + "epoch": 0.42804, + "grad_norm": 1.1709606756359476, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 42804 + }, + { + "epoch": 0.42805, + "grad_norm": 0.9541161946668697, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 42805 + }, + { + "epoch": 0.42806, + "grad_norm": 1.073884022746885, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 42806 + }, + { + "epoch": 0.42807, + "grad_norm": 1.2451465472431016, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 42807 + }, + { + "epoch": 0.42808, + "grad_norm": 0.9983064777262888, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 42808 + }, + { + "epoch": 0.42809, + "grad_norm": 1.1675389507052742, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 42809 + }, + { + "epoch": 0.4281, + "grad_norm": 1.0717382392589831, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 42810 + }, + { + "epoch": 0.42811, + "grad_norm": 1.0602040314289256, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 42811 + }, + { + "epoch": 0.42812, + "grad_norm": 1.1136122417441263, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 42812 + }, + { + "epoch": 0.42813, + "grad_norm": 1.0672758864799579, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 42813 + }, + { + "epoch": 0.42814, + "grad_norm": 1.0258704808085277, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 42814 + }, + { + "epoch": 0.42815, + "grad_norm": 1.2801176589547858, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 42815 + }, + { + "epoch": 0.42816, + "grad_norm": 0.7563782773125344, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 42816 + }, + { + "epoch": 0.42817, + "grad_norm": 0.7759710776989848, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 42817 + }, + { + "epoch": 0.42818, + "grad_norm": 1.123516609500514, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 42818 + }, + { + "epoch": 0.42819, + "grad_norm": 1.3265369424737448, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 42819 + }, + { + "epoch": 0.4282, + "grad_norm": 0.9771135012594552, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 42820 + }, + { + "epoch": 0.42821, + "grad_norm": 1.0408719199271752, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 42821 + }, + { + "epoch": 0.42822, + "grad_norm": 1.1195417749821261, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 42822 + }, + { + "epoch": 0.42823, + "grad_norm": 1.043542195553955, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 42823 + }, + { + "epoch": 0.42824, + "grad_norm": 1.057506031187412, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 42824 + }, + { + "epoch": 0.42825, + "grad_norm": 0.9968337165777739, + "learning_rate": 0.003, + "loss": 4.018, + "step": 42825 + }, + { + "epoch": 0.42826, + "grad_norm": 1.2453711635772897, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 42826 + }, + { + "epoch": 0.42827, + "grad_norm": 0.9605780159402698, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 42827 + }, + { + "epoch": 0.42828, + "grad_norm": 1.0527291720927752, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 42828 + }, + { + "epoch": 0.42829, + "grad_norm": 1.1161504122251764, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 42829 + }, + { + "epoch": 0.4283, + "grad_norm": 0.9705915128377152, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 42830 + }, + { + "epoch": 0.42831, + "grad_norm": 1.0493334796669775, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 42831 + }, + { + "epoch": 0.42832, + "grad_norm": 1.1510061437728152, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 42832 + }, + { + "epoch": 0.42833, + "grad_norm": 0.9568473282348294, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 42833 + }, + { + "epoch": 0.42834, + "grad_norm": 1.0487473527058846, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 42834 + }, + { + "epoch": 0.42835, + "grad_norm": 1.145272036948661, + "learning_rate": 0.003, + "loss": 4.048, + "step": 42835 + }, + { + "epoch": 0.42836, + "grad_norm": 1.2201113048060943, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 42836 + }, + { + "epoch": 0.42837, + "grad_norm": 0.9662126213380429, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 42837 + }, + { + "epoch": 0.42838, + "grad_norm": 1.1872531550219116, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 42838 + }, + { + "epoch": 0.42839, + "grad_norm": 0.961186754865018, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 42839 + }, + { + "epoch": 0.4284, + "grad_norm": 1.2334873304512977, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 42840 + }, + { + "epoch": 0.42841, + "grad_norm": 1.074161094569911, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 42841 + }, + { + "epoch": 0.42842, + "grad_norm": 1.0303769608282145, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 42842 + }, + { + "epoch": 0.42843, + "grad_norm": 1.2935060923640511, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 42843 + }, + { + "epoch": 0.42844, + "grad_norm": 1.1520292430451888, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 42844 + }, + { + "epoch": 0.42845, + "grad_norm": 1.1307704120285424, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 42845 + }, + { + "epoch": 0.42846, + "grad_norm": 1.1626399174515067, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 42846 + }, + { + "epoch": 0.42847, + "grad_norm": 1.0541952661946201, + "learning_rate": 0.003, + "loss": 4.035, + "step": 42847 + }, + { + "epoch": 0.42848, + "grad_norm": 1.1254229680292753, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 42848 + }, + { + "epoch": 0.42849, + "grad_norm": 1.020899280002007, + "learning_rate": 0.003, + "loss": 4.013, + "step": 42849 + }, + { + "epoch": 0.4285, + "grad_norm": 1.163477037336982, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 42850 + }, + { + "epoch": 0.42851, + "grad_norm": 0.949040141136751, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 42851 + }, + { + "epoch": 0.42852, + "grad_norm": 1.0802012116059787, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 42852 + }, + { + "epoch": 0.42853, + "grad_norm": 1.2787415446328634, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 42853 + }, + { + "epoch": 0.42854, + "grad_norm": 0.8554715726064728, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 42854 + }, + { + "epoch": 0.42855, + "grad_norm": 0.9784477989648804, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 42855 + }, + { + "epoch": 0.42856, + "grad_norm": 1.005900282713771, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 42856 + }, + { + "epoch": 0.42857, + "grad_norm": 1.1101610638542874, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 42857 + }, + { + "epoch": 0.42858, + "grad_norm": 0.8978256295145235, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 42858 + }, + { + "epoch": 0.42859, + "grad_norm": 0.8879801510889187, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 42859 + }, + { + "epoch": 0.4286, + "grad_norm": 0.9136565905199483, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 42860 + }, + { + "epoch": 0.42861, + "grad_norm": 1.057678304762161, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 42861 + }, + { + "epoch": 0.42862, + "grad_norm": 1.5086695978847358, + "learning_rate": 0.003, + "loss": 4.02, + "step": 42862 + }, + { + "epoch": 0.42863, + "grad_norm": 0.7377955354293786, + "learning_rate": 0.003, + "loss": 3.991, + "step": 42863 + }, + { + "epoch": 0.42864, + "grad_norm": 0.8096757294867396, + "learning_rate": 0.003, + "loss": 4.023, + "step": 42864 + }, + { + "epoch": 0.42865, + "grad_norm": 1.0907675001086532, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 42865 + }, + { + "epoch": 0.42866, + "grad_norm": 0.9713636934508909, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 42866 + }, + { + "epoch": 0.42867, + "grad_norm": 1.2866670559447229, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 42867 + }, + { + "epoch": 0.42868, + "grad_norm": 1.49060456191392, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 42868 + }, + { + "epoch": 0.42869, + "grad_norm": 0.8179951742961051, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 42869 + }, + { + "epoch": 0.4287, + "grad_norm": 0.8302917284721371, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 42870 + }, + { + "epoch": 0.42871, + "grad_norm": 0.9279783836538066, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 42871 + }, + { + "epoch": 0.42872, + "grad_norm": 1.360155641675323, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 42872 + }, + { + "epoch": 0.42873, + "grad_norm": 1.320159460012257, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 42873 + }, + { + "epoch": 0.42874, + "grad_norm": 1.0236508590677917, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 42874 + }, + { + "epoch": 0.42875, + "grad_norm": 0.8424564395199767, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 42875 + }, + { + "epoch": 0.42876, + "grad_norm": 0.8353887354606085, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 42876 + }, + { + "epoch": 0.42877, + "grad_norm": 0.9583436408891102, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 42877 + }, + { + "epoch": 0.42878, + "grad_norm": 1.0656028651619567, + "learning_rate": 0.003, + "loss": 4.045, + "step": 42878 + }, + { + "epoch": 0.42879, + "grad_norm": 1.019326736240404, + "learning_rate": 0.003, + "loss": 4.035, + "step": 42879 + }, + { + "epoch": 0.4288, + "grad_norm": 0.9521169735269289, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 42880 + }, + { + "epoch": 0.42881, + "grad_norm": 0.9734838157653151, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 42881 + }, + { + "epoch": 0.42882, + "grad_norm": 1.1363645576834465, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 42882 + }, + { + "epoch": 0.42883, + "grad_norm": 1.1444233893948965, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 42883 + }, + { + "epoch": 0.42884, + "grad_norm": 1.0631751181561715, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 42884 + }, + { + "epoch": 0.42885, + "grad_norm": 1.2063055889087335, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 42885 + }, + { + "epoch": 0.42886, + "grad_norm": 1.316094389119299, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 42886 + }, + { + "epoch": 0.42887, + "grad_norm": 1.1736563446925492, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 42887 + }, + { + "epoch": 0.42888, + "grad_norm": 0.9647083412464655, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 42888 + }, + { + "epoch": 0.42889, + "grad_norm": 1.1314678833099179, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 42889 + }, + { + "epoch": 0.4289, + "grad_norm": 1.2441839201050369, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 42890 + }, + { + "epoch": 0.42891, + "grad_norm": 0.9842971738558441, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 42891 + }, + { + "epoch": 0.42892, + "grad_norm": 1.0853473749707472, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 42892 + }, + { + "epoch": 0.42893, + "grad_norm": 1.07851357817616, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 42893 + }, + { + "epoch": 0.42894, + "grad_norm": 1.036133809358422, + "learning_rate": 0.003, + "loss": 4.027, + "step": 42894 + }, + { + "epoch": 0.42895, + "grad_norm": 0.9942510265192475, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 42895 + }, + { + "epoch": 0.42896, + "grad_norm": 1.117629679871869, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 42896 + }, + { + "epoch": 0.42897, + "grad_norm": 0.9638960535692009, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 42897 + }, + { + "epoch": 0.42898, + "grad_norm": 1.1804726980968192, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 42898 + }, + { + "epoch": 0.42899, + "grad_norm": 0.9526126295687252, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 42899 + }, + { + "epoch": 0.429, + "grad_norm": 0.8949013410526329, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 42900 + }, + { + "epoch": 0.42901, + "grad_norm": 0.9326923902486671, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 42901 + }, + { + "epoch": 0.42902, + "grad_norm": 1.0705669303390626, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 42902 + }, + { + "epoch": 0.42903, + "grad_norm": 1.1805617018140342, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 42903 + }, + { + "epoch": 0.42904, + "grad_norm": 1.2129128850400142, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 42904 + }, + { + "epoch": 0.42905, + "grad_norm": 1.1149940193677534, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 42905 + }, + { + "epoch": 0.42906, + "grad_norm": 0.9445242621869387, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 42906 + }, + { + "epoch": 0.42907, + "grad_norm": 1.0786508940592985, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 42907 + }, + { + "epoch": 0.42908, + "grad_norm": 1.1511949545965035, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 42908 + }, + { + "epoch": 0.42909, + "grad_norm": 1.1975298088945636, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 42909 + }, + { + "epoch": 0.4291, + "grad_norm": 1.1466642800157254, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 42910 + }, + { + "epoch": 0.42911, + "grad_norm": 0.9121464837221626, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 42911 + }, + { + "epoch": 0.42912, + "grad_norm": 0.8467385813873506, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 42912 + }, + { + "epoch": 0.42913, + "grad_norm": 1.0018729627284553, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 42913 + }, + { + "epoch": 0.42914, + "grad_norm": 1.3182097420098156, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 42914 + }, + { + "epoch": 0.42915, + "grad_norm": 0.769565411921804, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 42915 + }, + { + "epoch": 0.42916, + "grad_norm": 0.7469973158667916, + "learning_rate": 0.003, + "loss": 4.038, + "step": 42916 + }, + { + "epoch": 0.42917, + "grad_norm": 0.8201874050525334, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 42917 + }, + { + "epoch": 0.42918, + "grad_norm": 0.9617590860712677, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 42918 + }, + { + "epoch": 0.42919, + "grad_norm": 1.0720019653161916, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 42919 + }, + { + "epoch": 0.4292, + "grad_norm": 1.0103350581470767, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 42920 + }, + { + "epoch": 0.42921, + "grad_norm": 1.0287216100390524, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 42921 + }, + { + "epoch": 0.42922, + "grad_norm": 1.0577837775051544, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 42922 + }, + { + "epoch": 0.42923, + "grad_norm": 1.158856292853576, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 42923 + }, + { + "epoch": 0.42924, + "grad_norm": 1.2662365381449852, + "learning_rate": 0.003, + "loss": 4.036, + "step": 42924 + }, + { + "epoch": 0.42925, + "grad_norm": 1.0079225766608402, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 42925 + }, + { + "epoch": 0.42926, + "grad_norm": 1.1291772499289583, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 42926 + }, + { + "epoch": 0.42927, + "grad_norm": 1.2766860108786322, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 42927 + }, + { + "epoch": 0.42928, + "grad_norm": 1.078290818431975, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 42928 + }, + { + "epoch": 0.42929, + "grad_norm": 1.1211049282748193, + "learning_rate": 0.003, + "loss": 4.027, + "step": 42929 + }, + { + "epoch": 0.4293, + "grad_norm": 0.9870681711040398, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 42930 + }, + { + "epoch": 0.42931, + "grad_norm": 1.1216545854449955, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 42931 + }, + { + "epoch": 0.42932, + "grad_norm": 1.1311631661109416, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 42932 + }, + { + "epoch": 0.42933, + "grad_norm": 1.1372097808786275, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 42933 + }, + { + "epoch": 0.42934, + "grad_norm": 1.4250344663715477, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 42934 + }, + { + "epoch": 0.42935, + "grad_norm": 0.9255283927281408, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 42935 + }, + { + "epoch": 0.42936, + "grad_norm": 1.0320941412448466, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 42936 + }, + { + "epoch": 0.42937, + "grad_norm": 1.1309936329869974, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 42937 + }, + { + "epoch": 0.42938, + "grad_norm": 0.9495398250495708, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 42938 + }, + { + "epoch": 0.42939, + "grad_norm": 1.0571126917472178, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 42939 + }, + { + "epoch": 0.4294, + "grad_norm": 1.080441643453445, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 42940 + }, + { + "epoch": 0.42941, + "grad_norm": 1.0817672195021517, + "learning_rate": 0.003, + "loss": 4.028, + "step": 42941 + }, + { + "epoch": 0.42942, + "grad_norm": 1.0978726347870058, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 42942 + }, + { + "epoch": 0.42943, + "grad_norm": 1.2146104608158295, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 42943 + }, + { + "epoch": 0.42944, + "grad_norm": 0.8780566722406492, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 42944 + }, + { + "epoch": 0.42945, + "grad_norm": 0.9106743198139539, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 42945 + }, + { + "epoch": 0.42946, + "grad_norm": 1.1368557058508109, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 42946 + }, + { + "epoch": 0.42947, + "grad_norm": 1.1098947267735844, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 42947 + }, + { + "epoch": 0.42948, + "grad_norm": 1.4272854605869894, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 42948 + }, + { + "epoch": 0.42949, + "grad_norm": 1.0602116676086069, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 42949 + }, + { + "epoch": 0.4295, + "grad_norm": 1.2587608672518942, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 42950 + }, + { + "epoch": 0.42951, + "grad_norm": 0.8468591169298879, + "learning_rate": 0.003, + "loss": 4.035, + "step": 42951 + }, + { + "epoch": 0.42952, + "grad_norm": 0.8699649775282055, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 42952 + }, + { + "epoch": 0.42953, + "grad_norm": 1.0493737630866653, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 42953 + }, + { + "epoch": 0.42954, + "grad_norm": 1.3637450751759832, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 42954 + }, + { + "epoch": 0.42955, + "grad_norm": 0.6785265702560039, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 42955 + }, + { + "epoch": 0.42956, + "grad_norm": 0.7664371073549562, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 42956 + }, + { + "epoch": 0.42957, + "grad_norm": 0.8128713916418533, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 42957 + }, + { + "epoch": 0.42958, + "grad_norm": 0.8820362431291755, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 42958 + }, + { + "epoch": 0.42959, + "grad_norm": 0.8708196170674717, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 42959 + }, + { + "epoch": 0.4296, + "grad_norm": 0.9057407433274304, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 42960 + }, + { + "epoch": 0.42961, + "grad_norm": 1.0365190811182206, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 42961 + }, + { + "epoch": 0.42962, + "grad_norm": 1.352905343673179, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 42962 + }, + { + "epoch": 0.42963, + "grad_norm": 1.3835643387668295, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 42963 + }, + { + "epoch": 0.42964, + "grad_norm": 0.9973621265897294, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 42964 + }, + { + "epoch": 0.42965, + "grad_norm": 1.2856591012612428, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 42965 + }, + { + "epoch": 0.42966, + "grad_norm": 1.1577741243516217, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 42966 + }, + { + "epoch": 0.42967, + "grad_norm": 0.9310696357182002, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 42967 + }, + { + "epoch": 0.42968, + "grad_norm": 0.9739564182119471, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 42968 + }, + { + "epoch": 0.42969, + "grad_norm": 1.1332369835918976, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 42969 + }, + { + "epoch": 0.4297, + "grad_norm": 1.2495266121703732, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 42970 + }, + { + "epoch": 0.42971, + "grad_norm": 1.139046055732347, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 42971 + }, + { + "epoch": 0.42972, + "grad_norm": 0.9394874843325464, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 42972 + }, + { + "epoch": 0.42973, + "grad_norm": 1.0354107993096973, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 42973 + }, + { + "epoch": 0.42974, + "grad_norm": 1.2464878963950727, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 42974 + }, + { + "epoch": 0.42975, + "grad_norm": 0.9461898651914072, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 42975 + }, + { + "epoch": 0.42976, + "grad_norm": 1.0883980604509818, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 42976 + }, + { + "epoch": 0.42977, + "grad_norm": 1.1664496516876421, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 42977 + }, + { + "epoch": 0.42978, + "grad_norm": 0.8960820147924279, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 42978 + }, + { + "epoch": 0.42979, + "grad_norm": 0.9149499564173731, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 42979 + }, + { + "epoch": 0.4298, + "grad_norm": 0.937171536577243, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 42980 + }, + { + "epoch": 0.42981, + "grad_norm": 1.1071670316349338, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 42981 + }, + { + "epoch": 0.42982, + "grad_norm": 1.106686517739155, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 42982 + }, + { + "epoch": 0.42983, + "grad_norm": 1.1887374083406927, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 42983 + }, + { + "epoch": 0.42984, + "grad_norm": 1.1767839511240101, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 42984 + }, + { + "epoch": 0.42985, + "grad_norm": 0.8978651160790089, + "learning_rate": 0.003, + "loss": 4.031, + "step": 42985 + }, + { + "epoch": 0.42986, + "grad_norm": 0.9361289086773213, + "learning_rate": 0.003, + "loss": 4.027, + "step": 42986 + }, + { + "epoch": 0.42987, + "grad_norm": 1.2663523757032824, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 42987 + }, + { + "epoch": 0.42988, + "grad_norm": 0.8019595463262259, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 42988 + }, + { + "epoch": 0.42989, + "grad_norm": 1.0224014182412033, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 42989 + }, + { + "epoch": 0.4299, + "grad_norm": 1.257762663819449, + "learning_rate": 0.003, + "loss": 4.05, + "step": 42990 + }, + { + "epoch": 0.42991, + "grad_norm": 0.9603876267854446, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 42991 + }, + { + "epoch": 0.42992, + "grad_norm": 1.3413908446376355, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 42992 + }, + { + "epoch": 0.42993, + "grad_norm": 1.1363249825268906, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 42993 + }, + { + "epoch": 0.42994, + "grad_norm": 1.050505563446824, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 42994 + }, + { + "epoch": 0.42995, + "grad_norm": 1.2082205130053858, + "learning_rate": 0.003, + "loss": 4.021, + "step": 42995 + }, + { + "epoch": 0.42996, + "grad_norm": 1.090400129363021, + "learning_rate": 0.003, + "loss": 4.032, + "step": 42996 + }, + { + "epoch": 0.42997, + "grad_norm": 1.2414099694263034, + "learning_rate": 0.003, + "loss": 4.056, + "step": 42997 + }, + { + "epoch": 0.42998, + "grad_norm": 0.9156363787789202, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 42998 + }, + { + "epoch": 0.42999, + "grad_norm": 0.9429996139372037, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 42999 + }, + { + "epoch": 0.43, + "grad_norm": 1.0738159114629295, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 43000 + }, + { + "epoch": 0.43001, + "grad_norm": 1.2060502113424607, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 43001 + }, + { + "epoch": 0.43002, + "grad_norm": 0.7464898558563298, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 43002 + }, + { + "epoch": 0.43003, + "grad_norm": 0.6883182979805527, + "learning_rate": 0.003, + "loss": 4.047, + "step": 43003 + }, + { + "epoch": 0.43004, + "grad_norm": 0.6410119217194415, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 43004 + }, + { + "epoch": 0.43005, + "grad_norm": 0.8031248418046163, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 43005 + }, + { + "epoch": 0.43006, + "grad_norm": 0.850431602907568, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 43006 + }, + { + "epoch": 0.43007, + "grad_norm": 0.8388337451731693, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 43007 + }, + { + "epoch": 0.43008, + "grad_norm": 0.9230058886044437, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 43008 + }, + { + "epoch": 0.43009, + "grad_norm": 1.1349084437685526, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 43009 + }, + { + "epoch": 0.4301, + "grad_norm": 1.4120742037144938, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 43010 + }, + { + "epoch": 0.43011, + "grad_norm": 1.313551023106489, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 43011 + }, + { + "epoch": 0.43012, + "grad_norm": 0.9521478025956859, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 43012 + }, + { + "epoch": 0.43013, + "grad_norm": 1.096254032279823, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 43013 + }, + { + "epoch": 0.43014, + "grad_norm": 1.3038652309538539, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 43014 + }, + { + "epoch": 0.43015, + "grad_norm": 0.9192399471601405, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 43015 + }, + { + "epoch": 0.43016, + "grad_norm": 1.2118064987136559, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 43016 + }, + { + "epoch": 0.43017, + "grad_norm": 1.0301508322193829, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 43017 + }, + { + "epoch": 0.43018, + "grad_norm": 1.2060297034598977, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 43018 + }, + { + "epoch": 0.43019, + "grad_norm": 1.016004032163254, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 43019 + }, + { + "epoch": 0.4302, + "grad_norm": 1.2311141609179406, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 43020 + }, + { + "epoch": 0.43021, + "grad_norm": 1.122356755691172, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 43021 + }, + { + "epoch": 0.43022, + "grad_norm": 1.2219686856676462, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 43022 + }, + { + "epoch": 0.43023, + "grad_norm": 1.1136077853701045, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 43023 + }, + { + "epoch": 0.43024, + "grad_norm": 1.041247581436865, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 43024 + }, + { + "epoch": 0.43025, + "grad_norm": 1.1748922699075304, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 43025 + }, + { + "epoch": 0.43026, + "grad_norm": 0.9614219652715568, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 43026 + }, + { + "epoch": 0.43027, + "grad_norm": 1.2644744426372034, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 43027 + }, + { + "epoch": 0.43028, + "grad_norm": 1.3804975571076537, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 43028 + }, + { + "epoch": 0.43029, + "grad_norm": 0.8498872175361555, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 43029 + }, + { + "epoch": 0.4303, + "grad_norm": 0.805107860223512, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 43030 + }, + { + "epoch": 0.43031, + "grad_norm": 0.8815958256457983, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 43031 + }, + { + "epoch": 0.43032, + "grad_norm": 0.9536431112587127, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 43032 + }, + { + "epoch": 0.43033, + "grad_norm": 0.9606778624539108, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 43033 + }, + { + "epoch": 0.43034, + "grad_norm": 1.1083867302177666, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 43034 + }, + { + "epoch": 0.43035, + "grad_norm": 1.2549805025274654, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 43035 + }, + { + "epoch": 0.43036, + "grad_norm": 1.2283277020752492, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 43036 + }, + { + "epoch": 0.43037, + "grad_norm": 0.9499435836078415, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 43037 + }, + { + "epoch": 0.43038, + "grad_norm": 1.0427192029329606, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 43038 + }, + { + "epoch": 0.43039, + "grad_norm": 1.2109074952831607, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 43039 + }, + { + "epoch": 0.4304, + "grad_norm": 1.020756712369319, + "learning_rate": 0.003, + "loss": 4.045, + "step": 43040 + }, + { + "epoch": 0.43041, + "grad_norm": 1.1727934515511544, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 43041 + }, + { + "epoch": 0.43042, + "grad_norm": 0.9889782589926792, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 43042 + }, + { + "epoch": 0.43043, + "grad_norm": 1.1503433196387869, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 43043 + }, + { + "epoch": 0.43044, + "grad_norm": 1.1007826478037184, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 43044 + }, + { + "epoch": 0.43045, + "grad_norm": 1.0466798676836764, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 43045 + }, + { + "epoch": 0.43046, + "grad_norm": 1.099560988364938, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 43046 + }, + { + "epoch": 0.43047, + "grad_norm": 1.037587916540768, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 43047 + }, + { + "epoch": 0.43048, + "grad_norm": 1.1583721730387253, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 43048 + }, + { + "epoch": 0.43049, + "grad_norm": 1.0122224960750343, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 43049 + }, + { + "epoch": 0.4305, + "grad_norm": 1.2675587798552805, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 43050 + }, + { + "epoch": 0.43051, + "grad_norm": 1.1554234926469682, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 43051 + }, + { + "epoch": 0.43052, + "grad_norm": 1.2131578829673362, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 43052 + }, + { + "epoch": 0.43053, + "grad_norm": 0.9054542426134572, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 43053 + }, + { + "epoch": 0.43054, + "grad_norm": 0.8729996013293962, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 43054 + }, + { + "epoch": 0.43055, + "grad_norm": 1.0606570855445994, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 43055 + }, + { + "epoch": 0.43056, + "grad_norm": 1.0733997925410588, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 43056 + }, + { + "epoch": 0.43057, + "grad_norm": 1.119029301047445, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 43057 + }, + { + "epoch": 0.43058, + "grad_norm": 1.0044960611502611, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 43058 + }, + { + "epoch": 0.43059, + "grad_norm": 1.2224951446576142, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 43059 + }, + { + "epoch": 0.4306, + "grad_norm": 0.9305429505983327, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 43060 + }, + { + "epoch": 0.43061, + "grad_norm": 1.3595796563433287, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 43061 + }, + { + "epoch": 0.43062, + "grad_norm": 1.0007371598937203, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 43062 + }, + { + "epoch": 0.43063, + "grad_norm": 1.120519230305488, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 43063 + }, + { + "epoch": 0.43064, + "grad_norm": 1.1366016105918282, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 43064 + }, + { + "epoch": 0.43065, + "grad_norm": 1.0865476917616166, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 43065 + }, + { + "epoch": 0.43066, + "grad_norm": 1.0740026501201452, + "learning_rate": 0.003, + "loss": 4.045, + "step": 43066 + }, + { + "epoch": 0.43067, + "grad_norm": 1.0342310540983244, + "learning_rate": 0.003, + "loss": 4.021, + "step": 43067 + }, + { + "epoch": 0.43068, + "grad_norm": 1.1886537412074911, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 43068 + }, + { + "epoch": 0.43069, + "grad_norm": 1.0148077289651714, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 43069 + }, + { + "epoch": 0.4307, + "grad_norm": 1.1254296971596274, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 43070 + }, + { + "epoch": 0.43071, + "grad_norm": 1.2447580966545115, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 43071 + }, + { + "epoch": 0.43072, + "grad_norm": 1.067793037584533, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 43072 + }, + { + "epoch": 0.43073, + "grad_norm": 0.947613126405093, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 43073 + }, + { + "epoch": 0.43074, + "grad_norm": 0.9683880766827837, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 43074 + }, + { + "epoch": 0.43075, + "grad_norm": 1.299598875541202, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 43075 + }, + { + "epoch": 0.43076, + "grad_norm": 1.0275538137829066, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 43076 + }, + { + "epoch": 0.43077, + "grad_norm": 1.0729681293906, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 43077 + }, + { + "epoch": 0.43078, + "grad_norm": 1.2589846788771157, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 43078 + }, + { + "epoch": 0.43079, + "grad_norm": 0.9324505833734112, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 43079 + }, + { + "epoch": 0.4308, + "grad_norm": 0.7170679113283639, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 43080 + }, + { + "epoch": 0.43081, + "grad_norm": 0.8344221314590353, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 43081 + }, + { + "epoch": 0.43082, + "grad_norm": 1.073303067641245, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 43082 + }, + { + "epoch": 0.43083, + "grad_norm": 1.3051797268635787, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 43083 + }, + { + "epoch": 0.43084, + "grad_norm": 1.0880055546181355, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 43084 + }, + { + "epoch": 0.43085, + "grad_norm": 0.9160081974474812, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 43085 + }, + { + "epoch": 0.43086, + "grad_norm": 0.9997937491093233, + "learning_rate": 0.003, + "loss": 4.055, + "step": 43086 + }, + { + "epoch": 0.43087, + "grad_norm": 1.2752176408034286, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 43087 + }, + { + "epoch": 0.43088, + "grad_norm": 1.0276838767386172, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 43088 + }, + { + "epoch": 0.43089, + "grad_norm": 0.9804589549047278, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 43089 + }, + { + "epoch": 0.4309, + "grad_norm": 1.2668287942069347, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 43090 + }, + { + "epoch": 0.43091, + "grad_norm": 0.7923509762812674, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 43091 + }, + { + "epoch": 0.43092, + "grad_norm": 0.8440991194173681, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 43092 + }, + { + "epoch": 0.43093, + "grad_norm": 1.1256923685026372, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 43093 + }, + { + "epoch": 0.43094, + "grad_norm": 0.9489658237694867, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 43094 + }, + { + "epoch": 0.43095, + "grad_norm": 1.01758105384071, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 43095 + }, + { + "epoch": 0.43096, + "grad_norm": 1.1398119098363926, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 43096 + }, + { + "epoch": 0.43097, + "grad_norm": 1.0948385715833306, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 43097 + }, + { + "epoch": 0.43098, + "grad_norm": 1.0570655719540507, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 43098 + }, + { + "epoch": 0.43099, + "grad_norm": 1.0950005787710007, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 43099 + }, + { + "epoch": 0.431, + "grad_norm": 1.4444453255330367, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 43100 + }, + { + "epoch": 0.43101, + "grad_norm": 1.2040467544285787, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 43101 + }, + { + "epoch": 0.43102, + "grad_norm": 0.9667621325577012, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 43102 + }, + { + "epoch": 0.43103, + "grad_norm": 1.2244010014387632, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 43103 + }, + { + "epoch": 0.43104, + "grad_norm": 0.8960729247618476, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 43104 + }, + { + "epoch": 0.43105, + "grad_norm": 0.7443731437145137, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 43105 + }, + { + "epoch": 0.43106, + "grad_norm": 0.8071663591136296, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 43106 + }, + { + "epoch": 0.43107, + "grad_norm": 0.9301540746717647, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 43107 + }, + { + "epoch": 0.43108, + "grad_norm": 1.009675484987691, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 43108 + }, + { + "epoch": 0.43109, + "grad_norm": 1.0789073436508207, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 43109 + }, + { + "epoch": 0.4311, + "grad_norm": 1.011852717825852, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 43110 + }, + { + "epoch": 0.43111, + "grad_norm": 1.1650290455513344, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 43111 + }, + { + "epoch": 0.43112, + "grad_norm": 1.2266894987013683, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 43112 + }, + { + "epoch": 0.43113, + "grad_norm": 1.3728675006126219, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 43113 + }, + { + "epoch": 0.43114, + "grad_norm": 1.134664477028703, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 43114 + }, + { + "epoch": 0.43115, + "grad_norm": 1.02358504794835, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 43115 + }, + { + "epoch": 0.43116, + "grad_norm": 1.110193020899599, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 43116 + }, + { + "epoch": 0.43117, + "grad_norm": 1.0392187875640158, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 43117 + }, + { + "epoch": 0.43118, + "grad_norm": 1.1941007144294582, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 43118 + }, + { + "epoch": 0.43119, + "grad_norm": 1.0098193973810383, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 43119 + }, + { + "epoch": 0.4312, + "grad_norm": 1.1361974135721784, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 43120 + }, + { + "epoch": 0.43121, + "grad_norm": 0.9442034567281047, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 43121 + }, + { + "epoch": 0.43122, + "grad_norm": 0.9432017108341317, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 43122 + }, + { + "epoch": 0.43123, + "grad_norm": 1.088045423672423, + "learning_rate": 0.003, + "loss": 4.035, + "step": 43123 + }, + { + "epoch": 0.43124, + "grad_norm": 1.2947910663096573, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 43124 + }, + { + "epoch": 0.43125, + "grad_norm": 1.1394352944093133, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 43125 + }, + { + "epoch": 0.43126, + "grad_norm": 1.2965455030918704, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 43126 + }, + { + "epoch": 0.43127, + "grad_norm": 0.9348943327677854, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 43127 + }, + { + "epoch": 0.43128, + "grad_norm": 1.1114615803962538, + "learning_rate": 0.003, + "loss": 4.039, + "step": 43128 + }, + { + "epoch": 0.43129, + "grad_norm": 1.094081534124305, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 43129 + }, + { + "epoch": 0.4313, + "grad_norm": 1.2003609498221754, + "learning_rate": 0.003, + "loss": 4.047, + "step": 43130 + }, + { + "epoch": 0.43131, + "grad_norm": 0.9243215569337223, + "learning_rate": 0.003, + "loss": 4.016, + "step": 43131 + }, + { + "epoch": 0.43132, + "grad_norm": 0.9388653360782576, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 43132 + }, + { + "epoch": 0.43133, + "grad_norm": 1.0094971384538376, + "learning_rate": 0.003, + "loss": 4.077, + "step": 43133 + }, + { + "epoch": 0.43134, + "grad_norm": 1.1615505128478665, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 43134 + }, + { + "epoch": 0.43135, + "grad_norm": 1.158140423101913, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 43135 + }, + { + "epoch": 0.43136, + "grad_norm": 1.162497034111156, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 43136 + }, + { + "epoch": 0.43137, + "grad_norm": 1.1343669692950307, + "learning_rate": 0.003, + "loss": 4.011, + "step": 43137 + }, + { + "epoch": 0.43138, + "grad_norm": 0.9879121922747852, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 43138 + }, + { + "epoch": 0.43139, + "grad_norm": 1.1557278942923699, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 43139 + }, + { + "epoch": 0.4314, + "grad_norm": 0.9971927935395892, + "learning_rate": 0.003, + "loss": 4.009, + "step": 43140 + }, + { + "epoch": 0.43141, + "grad_norm": 1.0714111688760914, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 43141 + }, + { + "epoch": 0.43142, + "grad_norm": 1.059652394495864, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 43142 + }, + { + "epoch": 0.43143, + "grad_norm": 1.0354396118999445, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 43143 + }, + { + "epoch": 0.43144, + "grad_norm": 1.3215805750940177, + "learning_rate": 0.003, + "loss": 4.053, + "step": 43144 + }, + { + "epoch": 0.43145, + "grad_norm": 0.9642204238688221, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 43145 + }, + { + "epoch": 0.43146, + "grad_norm": 1.1946806010214361, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 43146 + }, + { + "epoch": 0.43147, + "grad_norm": 0.9657842029311233, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 43147 + }, + { + "epoch": 0.43148, + "grad_norm": 0.8895703431755686, + "learning_rate": 0.003, + "loss": 4.047, + "step": 43148 + }, + { + "epoch": 0.43149, + "grad_norm": 0.8973461737852688, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 43149 + }, + { + "epoch": 0.4315, + "grad_norm": 1.0045800190334566, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 43150 + }, + { + "epoch": 0.43151, + "grad_norm": 1.278473065689272, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 43151 + }, + { + "epoch": 0.43152, + "grad_norm": 1.1606974251296058, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 43152 + }, + { + "epoch": 0.43153, + "grad_norm": 1.282725074547535, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 43153 + }, + { + "epoch": 0.43154, + "grad_norm": 0.9670131540797368, + "learning_rate": 0.003, + "loss": 4.022, + "step": 43154 + }, + { + "epoch": 0.43155, + "grad_norm": 1.05488539908673, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 43155 + }, + { + "epoch": 0.43156, + "grad_norm": 1.0941030926489335, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 43156 + }, + { + "epoch": 0.43157, + "grad_norm": 1.1996354234210105, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 43157 + }, + { + "epoch": 0.43158, + "grad_norm": 0.8885841869657848, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 43158 + }, + { + "epoch": 0.43159, + "grad_norm": 0.9077977950638776, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 43159 + }, + { + "epoch": 0.4316, + "grad_norm": 1.0064895000447394, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 43160 + }, + { + "epoch": 0.43161, + "grad_norm": 1.020991291491783, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 43161 + }, + { + "epoch": 0.43162, + "grad_norm": 1.0397846481185848, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 43162 + }, + { + "epoch": 0.43163, + "grad_norm": 1.2901668147347405, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 43163 + }, + { + "epoch": 0.43164, + "grad_norm": 1.141775835202137, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 43164 + }, + { + "epoch": 0.43165, + "grad_norm": 0.9717738556662792, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 43165 + }, + { + "epoch": 0.43166, + "grad_norm": 1.113370258399939, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 43166 + }, + { + "epoch": 0.43167, + "grad_norm": 1.2357600140072986, + "learning_rate": 0.003, + "loss": 4.025, + "step": 43167 + }, + { + "epoch": 0.43168, + "grad_norm": 0.891147723588548, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 43168 + }, + { + "epoch": 0.43169, + "grad_norm": 1.0249904719632401, + "learning_rate": 0.003, + "loss": 4.005, + "step": 43169 + }, + { + "epoch": 0.4317, + "grad_norm": 1.3207480776177316, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 43170 + }, + { + "epoch": 0.43171, + "grad_norm": 0.8020753833660479, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 43171 + }, + { + "epoch": 0.43172, + "grad_norm": 0.929992232326976, + "learning_rate": 0.003, + "loss": 4.039, + "step": 43172 + }, + { + "epoch": 0.43173, + "grad_norm": 0.9836015585181123, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 43173 + }, + { + "epoch": 0.43174, + "grad_norm": 1.2202460793954701, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 43174 + }, + { + "epoch": 0.43175, + "grad_norm": 0.9292850225839513, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 43175 + }, + { + "epoch": 0.43176, + "grad_norm": 1.227081181188628, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 43176 + }, + { + "epoch": 0.43177, + "grad_norm": 1.078475794756675, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 43177 + }, + { + "epoch": 0.43178, + "grad_norm": 1.0929634969613842, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 43178 + }, + { + "epoch": 0.43179, + "grad_norm": 0.960025425852625, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 43179 + }, + { + "epoch": 0.4318, + "grad_norm": 1.1443212112525656, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 43180 + }, + { + "epoch": 0.43181, + "grad_norm": 1.0094834670420105, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 43181 + }, + { + "epoch": 0.43182, + "grad_norm": 1.1096071977971633, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 43182 + }, + { + "epoch": 0.43183, + "grad_norm": 0.9874026935965192, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 43183 + }, + { + "epoch": 0.43184, + "grad_norm": 1.0846569020719887, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 43184 + }, + { + "epoch": 0.43185, + "grad_norm": 1.194749307861579, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 43185 + }, + { + "epoch": 0.43186, + "grad_norm": 1.106051983446227, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 43186 + }, + { + "epoch": 0.43187, + "grad_norm": 1.1248644847958336, + "learning_rate": 0.003, + "loss": 4.025, + "step": 43187 + }, + { + "epoch": 0.43188, + "grad_norm": 1.0163188146948485, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 43188 + }, + { + "epoch": 0.43189, + "grad_norm": 1.250674251370452, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 43189 + }, + { + "epoch": 0.4319, + "grad_norm": 0.9817400330506985, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 43190 + }, + { + "epoch": 0.43191, + "grad_norm": 1.0187609696548345, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 43191 + }, + { + "epoch": 0.43192, + "grad_norm": 1.4065099231229135, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 43192 + }, + { + "epoch": 0.43193, + "grad_norm": 0.8629126298592452, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 43193 + }, + { + "epoch": 0.43194, + "grad_norm": 0.8730179689218959, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 43194 + }, + { + "epoch": 0.43195, + "grad_norm": 0.9486546183080189, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 43195 + }, + { + "epoch": 0.43196, + "grad_norm": 1.1928314678606413, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 43196 + }, + { + "epoch": 0.43197, + "grad_norm": 1.1100271120400502, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 43197 + }, + { + "epoch": 0.43198, + "grad_norm": 1.0465230858162675, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 43198 + }, + { + "epoch": 0.43199, + "grad_norm": 1.1400500226724364, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 43199 + }, + { + "epoch": 0.432, + "grad_norm": 1.0574559859757848, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 43200 + }, + { + "epoch": 0.43201, + "grad_norm": 0.9591086070875652, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 43201 + }, + { + "epoch": 0.43202, + "grad_norm": 0.9789813455043452, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 43202 + }, + { + "epoch": 0.43203, + "grad_norm": 1.185958570169855, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 43203 + }, + { + "epoch": 0.43204, + "grad_norm": 1.049969650360067, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 43204 + }, + { + "epoch": 0.43205, + "grad_norm": 1.269129473362498, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 43205 + }, + { + "epoch": 0.43206, + "grad_norm": 0.901745810408102, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 43206 + }, + { + "epoch": 0.43207, + "grad_norm": 1.0704280571340563, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 43207 + }, + { + "epoch": 0.43208, + "grad_norm": 1.4326030708358377, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 43208 + }, + { + "epoch": 0.43209, + "grad_norm": 1.0027794424066476, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 43209 + }, + { + "epoch": 0.4321, + "grad_norm": 1.0322463064215222, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 43210 + }, + { + "epoch": 0.43211, + "grad_norm": 0.9240439537371141, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 43211 + }, + { + "epoch": 0.43212, + "grad_norm": 1.0137518274700525, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 43212 + }, + { + "epoch": 0.43213, + "grad_norm": 1.132179986620845, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 43213 + }, + { + "epoch": 0.43214, + "grad_norm": 0.7808765480216987, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 43214 + }, + { + "epoch": 0.43215, + "grad_norm": 0.8803571290733068, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 43215 + }, + { + "epoch": 0.43216, + "grad_norm": 1.0169348810875003, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 43216 + }, + { + "epoch": 0.43217, + "grad_norm": 1.1693304770484847, + "learning_rate": 0.003, + "loss": 4.059, + "step": 43217 + }, + { + "epoch": 0.43218, + "grad_norm": 1.0154768469579396, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 43218 + }, + { + "epoch": 0.43219, + "grad_norm": 1.1063589105511786, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 43219 + }, + { + "epoch": 0.4322, + "grad_norm": 0.9211388527546825, + "learning_rate": 0.003, + "loss": 4.033, + "step": 43220 + }, + { + "epoch": 0.43221, + "grad_norm": 0.939890676367297, + "learning_rate": 0.003, + "loss": 4.042, + "step": 43221 + }, + { + "epoch": 0.43222, + "grad_norm": 1.1582732509397928, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 43222 + }, + { + "epoch": 0.43223, + "grad_norm": 1.1196876563133389, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 43223 + }, + { + "epoch": 0.43224, + "grad_norm": 1.010687531157231, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 43224 + }, + { + "epoch": 0.43225, + "grad_norm": 1.3048445178203314, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 43225 + }, + { + "epoch": 0.43226, + "grad_norm": 0.8757477316338214, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 43226 + }, + { + "epoch": 0.43227, + "grad_norm": 1.0095222560305988, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 43227 + }, + { + "epoch": 0.43228, + "grad_norm": 1.3173642467560231, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 43228 + }, + { + "epoch": 0.43229, + "grad_norm": 0.9392582188593093, + "learning_rate": 0.003, + "loss": 4.05, + "step": 43229 + }, + { + "epoch": 0.4323, + "grad_norm": 1.009556555791412, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 43230 + }, + { + "epoch": 0.43231, + "grad_norm": 1.15539512375305, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 43231 + }, + { + "epoch": 0.43232, + "grad_norm": 0.8329381528417154, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 43232 + }, + { + "epoch": 0.43233, + "grad_norm": 0.8840393822069098, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 43233 + }, + { + "epoch": 0.43234, + "grad_norm": 0.9152342030742875, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 43234 + }, + { + "epoch": 0.43235, + "grad_norm": 0.8353070805808412, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 43235 + }, + { + "epoch": 0.43236, + "grad_norm": 1.145695894168856, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 43236 + }, + { + "epoch": 0.43237, + "grad_norm": 1.4522753201322554, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 43237 + }, + { + "epoch": 0.43238, + "grad_norm": 0.9929847927954069, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 43238 + }, + { + "epoch": 0.43239, + "grad_norm": 1.166833049799732, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 43239 + }, + { + "epoch": 0.4324, + "grad_norm": 0.9375791776196901, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 43240 + }, + { + "epoch": 0.43241, + "grad_norm": 1.2409926082700284, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 43241 + }, + { + "epoch": 0.43242, + "grad_norm": 1.2399212028460331, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 43242 + }, + { + "epoch": 0.43243, + "grad_norm": 0.8879316838623668, + "learning_rate": 0.003, + "loss": 4.009, + "step": 43243 + }, + { + "epoch": 0.43244, + "grad_norm": 0.9824499816524132, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 43244 + }, + { + "epoch": 0.43245, + "grad_norm": 1.3411801530197, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 43245 + }, + { + "epoch": 0.43246, + "grad_norm": 0.8516413755741334, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 43246 + }, + { + "epoch": 0.43247, + "grad_norm": 0.8983291150655542, + "learning_rate": 0.003, + "loss": 4.036, + "step": 43247 + }, + { + "epoch": 0.43248, + "grad_norm": 0.9484376107937813, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 43248 + }, + { + "epoch": 0.43249, + "grad_norm": 1.157624460591003, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 43249 + }, + { + "epoch": 0.4325, + "grad_norm": 1.2873327431341437, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 43250 + }, + { + "epoch": 0.43251, + "grad_norm": 0.9862902403804271, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 43251 + }, + { + "epoch": 0.43252, + "grad_norm": 1.2286598929518442, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 43252 + }, + { + "epoch": 0.43253, + "grad_norm": 1.0928786449210803, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 43253 + }, + { + "epoch": 0.43254, + "grad_norm": 0.8983899916051873, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 43254 + }, + { + "epoch": 0.43255, + "grad_norm": 1.0939804256608505, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 43255 + }, + { + "epoch": 0.43256, + "grad_norm": 1.1019710783345211, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 43256 + }, + { + "epoch": 0.43257, + "grad_norm": 0.9118366815007588, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 43257 + }, + { + "epoch": 0.43258, + "grad_norm": 0.9851538559782892, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 43258 + }, + { + "epoch": 0.43259, + "grad_norm": 1.303491276393207, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 43259 + }, + { + "epoch": 0.4326, + "grad_norm": 0.9677358279374239, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 43260 + }, + { + "epoch": 0.43261, + "grad_norm": 1.1848665801392158, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 43261 + }, + { + "epoch": 0.43262, + "grad_norm": 1.0931328968535714, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 43262 + }, + { + "epoch": 0.43263, + "grad_norm": 1.1080961413612025, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 43263 + }, + { + "epoch": 0.43264, + "grad_norm": 0.9841253409950991, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 43264 + }, + { + "epoch": 0.43265, + "grad_norm": 1.037357677176202, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 43265 + }, + { + "epoch": 0.43266, + "grad_norm": 1.0590476290657636, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 43266 + }, + { + "epoch": 0.43267, + "grad_norm": 0.9110805036531393, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 43267 + }, + { + "epoch": 0.43268, + "grad_norm": 0.9720304025942753, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 43268 + }, + { + "epoch": 0.43269, + "grad_norm": 1.1921491038253225, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 43269 + }, + { + "epoch": 0.4327, + "grad_norm": 1.2231502380205617, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 43270 + }, + { + "epoch": 0.43271, + "grad_norm": 1.0763714988731277, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 43271 + }, + { + "epoch": 0.43272, + "grad_norm": 0.9831416318141285, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 43272 + }, + { + "epoch": 0.43273, + "grad_norm": 1.0291362619004367, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 43273 + }, + { + "epoch": 0.43274, + "grad_norm": 1.035146294499252, + "learning_rate": 0.003, + "loss": 4.032, + "step": 43274 + }, + { + "epoch": 0.43275, + "grad_norm": 1.0625458670585068, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 43275 + }, + { + "epoch": 0.43276, + "grad_norm": 1.155778649913486, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 43276 + }, + { + "epoch": 0.43277, + "grad_norm": 1.0728845630759705, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 43277 + }, + { + "epoch": 0.43278, + "grad_norm": 0.9907787952204242, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 43278 + }, + { + "epoch": 0.43279, + "grad_norm": 1.0862649010782772, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 43279 + }, + { + "epoch": 0.4328, + "grad_norm": 1.1566461342637155, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 43280 + }, + { + "epoch": 0.43281, + "grad_norm": 1.092461185990487, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 43281 + }, + { + "epoch": 0.43282, + "grad_norm": 1.2508557338025887, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 43282 + }, + { + "epoch": 0.43283, + "grad_norm": 1.1103313014769045, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 43283 + }, + { + "epoch": 0.43284, + "grad_norm": 0.9293691885048782, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 43284 + }, + { + "epoch": 0.43285, + "grad_norm": 0.9972474199866266, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 43285 + }, + { + "epoch": 0.43286, + "grad_norm": 1.4081408235335582, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 43286 + }, + { + "epoch": 0.43287, + "grad_norm": 0.9161132777390255, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 43287 + }, + { + "epoch": 0.43288, + "grad_norm": 0.9664050235321552, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 43288 + }, + { + "epoch": 0.43289, + "grad_norm": 1.0598073294928134, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 43289 + }, + { + "epoch": 0.4329, + "grad_norm": 1.0666051659933868, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 43290 + }, + { + "epoch": 0.43291, + "grad_norm": 0.977945203704579, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 43291 + }, + { + "epoch": 0.43292, + "grad_norm": 0.9347395562232594, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 43292 + }, + { + "epoch": 0.43293, + "grad_norm": 1.0118677757355499, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 43293 + }, + { + "epoch": 0.43294, + "grad_norm": 1.5863030812465393, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 43294 + }, + { + "epoch": 0.43295, + "grad_norm": 0.7484485220589964, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 43295 + }, + { + "epoch": 0.43296, + "grad_norm": 0.7814389488308449, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 43296 + }, + { + "epoch": 0.43297, + "grad_norm": 0.9096543264527772, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 43297 + }, + { + "epoch": 0.43298, + "grad_norm": 1.050431779712842, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 43298 + }, + { + "epoch": 0.43299, + "grad_norm": 1.233602779149469, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 43299 + }, + { + "epoch": 0.433, + "grad_norm": 0.994574716179539, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 43300 + }, + { + "epoch": 0.43301, + "grad_norm": 1.3790547040474952, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 43301 + }, + { + "epoch": 0.43302, + "grad_norm": 0.99230744919306, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 43302 + }, + { + "epoch": 0.43303, + "grad_norm": 1.1791629694765735, + "learning_rate": 0.003, + "loss": 4.02, + "step": 43303 + }, + { + "epoch": 0.43304, + "grad_norm": 0.9086605410088296, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 43304 + }, + { + "epoch": 0.43305, + "grad_norm": 0.9442501733123383, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 43305 + }, + { + "epoch": 0.43306, + "grad_norm": 1.1849914661855099, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 43306 + }, + { + "epoch": 0.43307, + "grad_norm": 0.8911774861494905, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 43307 + }, + { + "epoch": 0.43308, + "grad_norm": 1.0161742249853778, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 43308 + }, + { + "epoch": 0.43309, + "grad_norm": 1.16419524182046, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 43309 + }, + { + "epoch": 0.4331, + "grad_norm": 0.856675258380683, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 43310 + }, + { + "epoch": 0.43311, + "grad_norm": 1.1412808487190114, + "learning_rate": 0.003, + "loss": 4.048, + "step": 43311 + }, + { + "epoch": 0.43312, + "grad_norm": 0.9816128173015213, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 43312 + }, + { + "epoch": 0.43313, + "grad_norm": 1.3089720754921756, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 43313 + }, + { + "epoch": 0.43314, + "grad_norm": 1.4009953496339582, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 43314 + }, + { + "epoch": 0.43315, + "grad_norm": 0.9228147957445745, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 43315 + }, + { + "epoch": 0.43316, + "grad_norm": 0.8265038503714455, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 43316 + }, + { + "epoch": 0.43317, + "grad_norm": 0.9869788255469762, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 43317 + }, + { + "epoch": 0.43318, + "grad_norm": 1.2779396422889635, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 43318 + }, + { + "epoch": 0.43319, + "grad_norm": 1.075793059628353, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 43319 + }, + { + "epoch": 0.4332, + "grad_norm": 1.3975485373306682, + "learning_rate": 0.003, + "loss": 4.037, + "step": 43320 + }, + { + "epoch": 0.43321, + "grad_norm": 0.9883277841840782, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 43321 + }, + { + "epoch": 0.43322, + "grad_norm": 1.036376317278012, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 43322 + }, + { + "epoch": 0.43323, + "grad_norm": 1.063339755135244, + "learning_rate": 0.003, + "loss": 4.043, + "step": 43323 + }, + { + "epoch": 0.43324, + "grad_norm": 1.136742755385901, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 43324 + }, + { + "epoch": 0.43325, + "grad_norm": 1.0457517782050123, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 43325 + }, + { + "epoch": 0.43326, + "grad_norm": 1.2677016853322938, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 43326 + }, + { + "epoch": 0.43327, + "grad_norm": 1.1432927190204658, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 43327 + }, + { + "epoch": 0.43328, + "grad_norm": 0.8875983587356193, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 43328 + }, + { + "epoch": 0.43329, + "grad_norm": 0.9196649667048147, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 43329 + }, + { + "epoch": 0.4333, + "grad_norm": 0.9790542273030396, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 43330 + }, + { + "epoch": 0.43331, + "grad_norm": 1.081904400146533, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 43331 + }, + { + "epoch": 0.43332, + "grad_norm": 1.2760126831201861, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 43332 + }, + { + "epoch": 0.43333, + "grad_norm": 1.1287000718214235, + "learning_rate": 0.003, + "loss": 4.016, + "step": 43333 + }, + { + "epoch": 0.43334, + "grad_norm": 1.2732844744551062, + "learning_rate": 0.003, + "loss": 4.048, + "step": 43334 + }, + { + "epoch": 0.43335, + "grad_norm": 1.0816558052207896, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 43335 + }, + { + "epoch": 0.43336, + "grad_norm": 1.1886872863184794, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 43336 + }, + { + "epoch": 0.43337, + "grad_norm": 1.098182000966335, + "learning_rate": 0.003, + "loss": 4.036, + "step": 43337 + }, + { + "epoch": 0.43338, + "grad_norm": 1.271299422251038, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 43338 + }, + { + "epoch": 0.43339, + "grad_norm": 0.8605809037982292, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 43339 + }, + { + "epoch": 0.4334, + "grad_norm": 0.8422243685476579, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 43340 + }, + { + "epoch": 0.43341, + "grad_norm": 1.247414763024792, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 43341 + }, + { + "epoch": 0.43342, + "grad_norm": 0.873665184252342, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 43342 + }, + { + "epoch": 0.43343, + "grad_norm": 1.0427159349557322, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 43343 + }, + { + "epoch": 0.43344, + "grad_norm": 1.2879480806839352, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 43344 + }, + { + "epoch": 0.43345, + "grad_norm": 0.8793451542287058, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 43345 + }, + { + "epoch": 0.43346, + "grad_norm": 0.9906679429308373, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 43346 + }, + { + "epoch": 0.43347, + "grad_norm": 1.1530539366147015, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 43347 + }, + { + "epoch": 0.43348, + "grad_norm": 0.9341522588391478, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 43348 + }, + { + "epoch": 0.43349, + "grad_norm": 1.1030719890611023, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 43349 + }, + { + "epoch": 0.4335, + "grad_norm": 1.290402680805891, + "learning_rate": 0.003, + "loss": 4.04, + "step": 43350 + }, + { + "epoch": 0.43351, + "grad_norm": 1.0030478149656359, + "learning_rate": 0.003, + "loss": 4.069, + "step": 43351 + }, + { + "epoch": 0.43352, + "grad_norm": 1.3568366984957234, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 43352 + }, + { + "epoch": 0.43353, + "grad_norm": 0.9458000246169228, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 43353 + }, + { + "epoch": 0.43354, + "grad_norm": 0.9143339307679166, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 43354 + }, + { + "epoch": 0.43355, + "grad_norm": 1.332190617101876, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 43355 + }, + { + "epoch": 0.43356, + "grad_norm": 1.3450928944186005, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 43356 + }, + { + "epoch": 0.43357, + "grad_norm": 0.7499654314010836, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 43357 + }, + { + "epoch": 0.43358, + "grad_norm": 0.8356192414836082, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 43358 + }, + { + "epoch": 0.43359, + "grad_norm": 0.8465356084074441, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 43359 + }, + { + "epoch": 0.4336, + "grad_norm": 0.9222141465719332, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 43360 + }, + { + "epoch": 0.43361, + "grad_norm": 0.9696394299569299, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 43361 + }, + { + "epoch": 0.43362, + "grad_norm": 1.343128849846907, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 43362 + }, + { + "epoch": 0.43363, + "grad_norm": 1.0880743632744343, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 43363 + }, + { + "epoch": 0.43364, + "grad_norm": 1.057832146539535, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 43364 + }, + { + "epoch": 0.43365, + "grad_norm": 1.1510510714102216, + "learning_rate": 0.003, + "loss": 4.05, + "step": 43365 + }, + { + "epoch": 0.43366, + "grad_norm": 1.0889065403743576, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 43366 + }, + { + "epoch": 0.43367, + "grad_norm": 1.4701638126007508, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 43367 + }, + { + "epoch": 0.43368, + "grad_norm": 1.0274940815870899, + "learning_rate": 0.003, + "loss": 4.043, + "step": 43368 + }, + { + "epoch": 0.43369, + "grad_norm": 0.9076088436031681, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 43369 + }, + { + "epoch": 0.4337, + "grad_norm": 1.0207026190360677, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 43370 + }, + { + "epoch": 0.43371, + "grad_norm": 1.3928719722727905, + "learning_rate": 0.003, + "loss": 4.047, + "step": 43371 + }, + { + "epoch": 0.43372, + "grad_norm": 0.9360036567737715, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 43372 + }, + { + "epoch": 0.43373, + "grad_norm": 1.1415554001788517, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 43373 + }, + { + "epoch": 0.43374, + "grad_norm": 1.3432068559552184, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 43374 + }, + { + "epoch": 0.43375, + "grad_norm": 1.1921942373757195, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 43375 + }, + { + "epoch": 0.43376, + "grad_norm": 0.8736867925405726, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 43376 + }, + { + "epoch": 0.43377, + "grad_norm": 0.7287406795814537, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 43377 + }, + { + "epoch": 0.43378, + "grad_norm": 0.7768563925510399, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 43378 + }, + { + "epoch": 0.43379, + "grad_norm": 1.0488330838392526, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 43379 + }, + { + "epoch": 0.4338, + "grad_norm": 1.0084274509567235, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 43380 + }, + { + "epoch": 0.43381, + "grad_norm": 1.0837002596820973, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 43381 + }, + { + "epoch": 0.43382, + "grad_norm": 1.0692323309711818, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 43382 + }, + { + "epoch": 0.43383, + "grad_norm": 1.2229403319347323, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 43383 + }, + { + "epoch": 0.43384, + "grad_norm": 0.9603250598455493, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 43384 + }, + { + "epoch": 0.43385, + "grad_norm": 1.0129315056190975, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 43385 + }, + { + "epoch": 0.43386, + "grad_norm": 1.1403221154948144, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 43386 + }, + { + "epoch": 0.43387, + "grad_norm": 0.9120097858760786, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 43387 + }, + { + "epoch": 0.43388, + "grad_norm": 0.9104779298780653, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 43388 + }, + { + "epoch": 0.43389, + "grad_norm": 0.9824295641445987, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 43389 + }, + { + "epoch": 0.4339, + "grad_norm": 0.9720898347886148, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 43390 + }, + { + "epoch": 0.43391, + "grad_norm": 0.9645236920900963, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 43391 + }, + { + "epoch": 0.43392, + "grad_norm": 1.0691633485564067, + "learning_rate": 0.003, + "loss": 4.01, + "step": 43392 + }, + { + "epoch": 0.43393, + "grad_norm": 1.202189752417543, + "learning_rate": 0.003, + "loss": 4.034, + "step": 43393 + }, + { + "epoch": 0.43394, + "grad_norm": 1.2142846158466618, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 43394 + }, + { + "epoch": 0.43395, + "grad_norm": 0.9470199316938455, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 43395 + }, + { + "epoch": 0.43396, + "grad_norm": 1.0280594257377744, + "learning_rate": 0.003, + "loss": 4.019, + "step": 43396 + }, + { + "epoch": 0.43397, + "grad_norm": 1.2842770579533542, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 43397 + }, + { + "epoch": 0.43398, + "grad_norm": 1.1046370980360465, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 43398 + }, + { + "epoch": 0.43399, + "grad_norm": 1.2017961045604586, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 43399 + }, + { + "epoch": 0.434, + "grad_norm": 1.0725295017825152, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 43400 + }, + { + "epoch": 0.43401, + "grad_norm": 1.0684658520411052, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 43401 + }, + { + "epoch": 0.43402, + "grad_norm": 1.2876025244917109, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 43402 + }, + { + "epoch": 0.43403, + "grad_norm": 1.23084615095955, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 43403 + }, + { + "epoch": 0.43404, + "grad_norm": 0.999257222270434, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 43404 + }, + { + "epoch": 0.43405, + "grad_norm": 1.1114377872412282, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 43405 + }, + { + "epoch": 0.43406, + "grad_norm": 0.8922264395776316, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 43406 + }, + { + "epoch": 0.43407, + "grad_norm": 0.9562416353902963, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 43407 + }, + { + "epoch": 0.43408, + "grad_norm": 1.0422890533447886, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 43408 + }, + { + "epoch": 0.43409, + "grad_norm": 1.1360757323253026, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 43409 + }, + { + "epoch": 0.4341, + "grad_norm": 1.2273746172718742, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 43410 + }, + { + "epoch": 0.43411, + "grad_norm": 1.155905971639686, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 43411 + }, + { + "epoch": 0.43412, + "grad_norm": 1.1422728179482358, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 43412 + }, + { + "epoch": 0.43413, + "grad_norm": 0.9891584716642463, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 43413 + }, + { + "epoch": 0.43414, + "grad_norm": 1.1867571681611115, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 43414 + }, + { + "epoch": 0.43415, + "grad_norm": 1.2830605416991105, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 43415 + }, + { + "epoch": 0.43416, + "grad_norm": 1.052536493313379, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 43416 + }, + { + "epoch": 0.43417, + "grad_norm": 0.9220806780135432, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 43417 + }, + { + "epoch": 0.43418, + "grad_norm": 0.9919554282465752, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 43418 + }, + { + "epoch": 0.43419, + "grad_norm": 1.069792845538301, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 43419 + }, + { + "epoch": 0.4342, + "grad_norm": 1.0961815825413803, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 43420 + }, + { + "epoch": 0.43421, + "grad_norm": 1.1426574603794857, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 43421 + }, + { + "epoch": 0.43422, + "grad_norm": 0.9797847442626265, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 43422 + }, + { + "epoch": 0.43423, + "grad_norm": 0.9606251182463287, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 43423 + }, + { + "epoch": 0.43424, + "grad_norm": 1.08453801809083, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 43424 + }, + { + "epoch": 0.43425, + "grad_norm": 0.9984575290163156, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 43425 + }, + { + "epoch": 0.43426, + "grad_norm": 1.0980276388832801, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 43426 + }, + { + "epoch": 0.43427, + "grad_norm": 0.9938659229854376, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 43427 + }, + { + "epoch": 0.43428, + "grad_norm": 1.1379914675434002, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 43428 + }, + { + "epoch": 0.43429, + "grad_norm": 1.072888143645586, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 43429 + }, + { + "epoch": 0.4343, + "grad_norm": 1.1653353833794566, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 43430 + }, + { + "epoch": 0.43431, + "grad_norm": 1.0608361188157456, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 43431 + }, + { + "epoch": 0.43432, + "grad_norm": 1.0822549931311254, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 43432 + }, + { + "epoch": 0.43433, + "grad_norm": 0.9015007688174053, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 43433 + }, + { + "epoch": 0.43434, + "grad_norm": 0.8605796027078779, + "learning_rate": 0.003, + "loss": 4.021, + "step": 43434 + }, + { + "epoch": 0.43435, + "grad_norm": 1.1019719698599857, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 43435 + }, + { + "epoch": 0.43436, + "grad_norm": 1.6096372167999058, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 43436 + }, + { + "epoch": 0.43437, + "grad_norm": 0.7779001622315966, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 43437 + }, + { + "epoch": 0.43438, + "grad_norm": 1.113218210633954, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 43438 + }, + { + "epoch": 0.43439, + "grad_norm": 1.0738199720471049, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 43439 + }, + { + "epoch": 0.4344, + "grad_norm": 1.032568471485378, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 43440 + }, + { + "epoch": 0.43441, + "grad_norm": 1.0942349435070058, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 43441 + }, + { + "epoch": 0.43442, + "grad_norm": 1.2522207041962905, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 43442 + }, + { + "epoch": 0.43443, + "grad_norm": 1.0589067884526444, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 43443 + }, + { + "epoch": 0.43444, + "grad_norm": 1.1718019294706075, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 43444 + }, + { + "epoch": 0.43445, + "grad_norm": 0.9446110087432451, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 43445 + }, + { + "epoch": 0.43446, + "grad_norm": 1.184731946663093, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 43446 + }, + { + "epoch": 0.43447, + "grad_norm": 1.2653811382500155, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 43447 + }, + { + "epoch": 0.43448, + "grad_norm": 0.7650275911005119, + "learning_rate": 0.003, + "loss": 4.044, + "step": 43448 + }, + { + "epoch": 0.43449, + "grad_norm": 0.7869108247140462, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 43449 + }, + { + "epoch": 0.4345, + "grad_norm": 1.0777525953723506, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 43450 + }, + { + "epoch": 0.43451, + "grad_norm": 1.307074895555246, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 43451 + }, + { + "epoch": 0.43452, + "grad_norm": 1.0929371875456138, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 43452 + }, + { + "epoch": 0.43453, + "grad_norm": 1.150894177449205, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 43453 + }, + { + "epoch": 0.43454, + "grad_norm": 0.9785662703760133, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 43454 + }, + { + "epoch": 0.43455, + "grad_norm": 1.0083429351951403, + "learning_rate": 0.003, + "loss": 4.019, + "step": 43455 + }, + { + "epoch": 0.43456, + "grad_norm": 1.082442486319093, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 43456 + }, + { + "epoch": 0.43457, + "grad_norm": 0.98512903441198, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 43457 + }, + { + "epoch": 0.43458, + "grad_norm": 1.2356472519209398, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 43458 + }, + { + "epoch": 0.43459, + "grad_norm": 1.0690122731293388, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 43459 + }, + { + "epoch": 0.4346, + "grad_norm": 1.403778189114136, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 43460 + }, + { + "epoch": 0.43461, + "grad_norm": 0.8637121526958383, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 43461 + }, + { + "epoch": 0.43462, + "grad_norm": 0.9454150540098843, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 43462 + }, + { + "epoch": 0.43463, + "grad_norm": 0.966184206168107, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 43463 + }, + { + "epoch": 0.43464, + "grad_norm": 1.002830046499269, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 43464 + }, + { + "epoch": 0.43465, + "grad_norm": 1.0892375675141321, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 43465 + }, + { + "epoch": 0.43466, + "grad_norm": 1.0392742167461884, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 43466 + }, + { + "epoch": 0.43467, + "grad_norm": 1.141326479742143, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 43467 + }, + { + "epoch": 0.43468, + "grad_norm": 1.0963094017782449, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 43468 + }, + { + "epoch": 0.43469, + "grad_norm": 1.3350303712791305, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 43469 + }, + { + "epoch": 0.4347, + "grad_norm": 0.8083805239226673, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 43470 + }, + { + "epoch": 0.43471, + "grad_norm": 0.8811250719088949, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 43471 + }, + { + "epoch": 0.43472, + "grad_norm": 1.0951292306230123, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 43472 + }, + { + "epoch": 0.43473, + "grad_norm": 1.2319862741026408, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 43473 + }, + { + "epoch": 0.43474, + "grad_norm": 0.9300398500151985, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 43474 + }, + { + "epoch": 0.43475, + "grad_norm": 1.2119679193782102, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 43475 + }, + { + "epoch": 0.43476, + "grad_norm": 1.100395706873132, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 43476 + }, + { + "epoch": 0.43477, + "grad_norm": 0.9453960763987092, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 43477 + }, + { + "epoch": 0.43478, + "grad_norm": 1.1028328472732034, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 43478 + }, + { + "epoch": 0.43479, + "grad_norm": 1.1372074215779155, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 43479 + }, + { + "epoch": 0.4348, + "grad_norm": 1.0193147061218035, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 43480 + }, + { + "epoch": 0.43481, + "grad_norm": 1.1059401563311573, + "learning_rate": 0.003, + "loss": 4.03, + "step": 43481 + }, + { + "epoch": 0.43482, + "grad_norm": 1.212901242059986, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 43482 + }, + { + "epoch": 0.43483, + "grad_norm": 1.0282429667462796, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 43483 + }, + { + "epoch": 0.43484, + "grad_norm": 1.1757321547511916, + "learning_rate": 0.003, + "loss": 4.024, + "step": 43484 + }, + { + "epoch": 0.43485, + "grad_norm": 1.0587213180024724, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 43485 + }, + { + "epoch": 0.43486, + "grad_norm": 1.0676985132637833, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 43486 + }, + { + "epoch": 0.43487, + "grad_norm": 1.0347529120424481, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 43487 + }, + { + "epoch": 0.43488, + "grad_norm": 1.1224243532346097, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 43488 + }, + { + "epoch": 0.43489, + "grad_norm": 0.9601642407520316, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 43489 + }, + { + "epoch": 0.4349, + "grad_norm": 1.1920287621047265, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 43490 + }, + { + "epoch": 0.43491, + "grad_norm": 1.2084238009665105, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 43491 + }, + { + "epoch": 0.43492, + "grad_norm": 0.9832880897123176, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 43492 + }, + { + "epoch": 0.43493, + "grad_norm": 1.3144803042603606, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 43493 + }, + { + "epoch": 0.43494, + "grad_norm": 1.0912448759223654, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 43494 + }, + { + "epoch": 0.43495, + "grad_norm": 1.0196175198994872, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 43495 + }, + { + "epoch": 0.43496, + "grad_norm": 1.0212554030507688, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 43496 + }, + { + "epoch": 0.43497, + "grad_norm": 1.0330092386824463, + "learning_rate": 0.003, + "loss": 4.02, + "step": 43497 + }, + { + "epoch": 0.43498, + "grad_norm": 1.0262451262023644, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 43498 + }, + { + "epoch": 0.43499, + "grad_norm": 0.9906928309856117, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 43499 + }, + { + "epoch": 0.435, + "grad_norm": 1.2095925823186677, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 43500 + }, + { + "epoch": 0.43501, + "grad_norm": 1.1344341648147314, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 43501 + }, + { + "epoch": 0.43502, + "grad_norm": 1.17384228405086, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 43502 + }, + { + "epoch": 0.43503, + "grad_norm": 0.8820064866011122, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 43503 + }, + { + "epoch": 0.43504, + "grad_norm": 1.0291473768212758, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 43504 + }, + { + "epoch": 0.43505, + "grad_norm": 1.5162093444751261, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 43505 + }, + { + "epoch": 0.43506, + "grad_norm": 0.8823409377613134, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 43506 + }, + { + "epoch": 0.43507, + "grad_norm": 0.9543610511780168, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 43507 + }, + { + "epoch": 0.43508, + "grad_norm": 1.0262470915099355, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 43508 + }, + { + "epoch": 0.43509, + "grad_norm": 1.115398245285325, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 43509 + }, + { + "epoch": 0.4351, + "grad_norm": 1.1871047038740044, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 43510 + }, + { + "epoch": 0.43511, + "grad_norm": 1.144408739871852, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 43511 + }, + { + "epoch": 0.43512, + "grad_norm": 1.0780548541835178, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 43512 + }, + { + "epoch": 0.43513, + "grad_norm": 1.165226823375269, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 43513 + }, + { + "epoch": 0.43514, + "grad_norm": 1.0200905362210422, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 43514 + }, + { + "epoch": 0.43515, + "grad_norm": 1.1112116408074246, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 43515 + }, + { + "epoch": 0.43516, + "grad_norm": 0.8713291980324014, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 43516 + }, + { + "epoch": 0.43517, + "grad_norm": 0.8497800341923704, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 43517 + }, + { + "epoch": 0.43518, + "grad_norm": 0.9951423382453715, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 43518 + }, + { + "epoch": 0.43519, + "grad_norm": 1.1563601786458593, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 43519 + }, + { + "epoch": 0.4352, + "grad_norm": 1.0456123294559032, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 43520 + }, + { + "epoch": 0.43521, + "grad_norm": 1.2724781918060932, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 43521 + }, + { + "epoch": 0.43522, + "grad_norm": 0.9878204604768374, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 43522 + }, + { + "epoch": 0.43523, + "grad_norm": 1.2868328374569253, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 43523 + }, + { + "epoch": 0.43524, + "grad_norm": 0.9847185444092016, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 43524 + }, + { + "epoch": 0.43525, + "grad_norm": 1.230442296221705, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 43525 + }, + { + "epoch": 0.43526, + "grad_norm": 1.0538451934528517, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 43526 + }, + { + "epoch": 0.43527, + "grad_norm": 0.9677293424640486, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 43527 + }, + { + "epoch": 0.43528, + "grad_norm": 1.0211052837714123, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 43528 + }, + { + "epoch": 0.43529, + "grad_norm": 1.4542455766210376, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 43529 + }, + { + "epoch": 0.4353, + "grad_norm": 1.032220227370103, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 43530 + }, + { + "epoch": 0.43531, + "grad_norm": 1.1552265529754506, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 43531 + }, + { + "epoch": 0.43532, + "grad_norm": 1.033453764947266, + "learning_rate": 0.003, + "loss": 4.016, + "step": 43532 + }, + { + "epoch": 0.43533, + "grad_norm": 1.2411493309342359, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 43533 + }, + { + "epoch": 0.43534, + "grad_norm": 1.0396979751646778, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 43534 + }, + { + "epoch": 0.43535, + "grad_norm": 1.0595470700356164, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 43535 + }, + { + "epoch": 0.43536, + "grad_norm": 1.0307672105973769, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 43536 + }, + { + "epoch": 0.43537, + "grad_norm": 1.0687487618479556, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 43537 + }, + { + "epoch": 0.43538, + "grad_norm": 1.0705862065510179, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 43538 + }, + { + "epoch": 0.43539, + "grad_norm": 1.0585277506556874, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 43539 + }, + { + "epoch": 0.4354, + "grad_norm": 1.0346250709838745, + "learning_rate": 0.003, + "loss": 4.019, + "step": 43540 + }, + { + "epoch": 0.43541, + "grad_norm": 1.4115590442217631, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 43541 + }, + { + "epoch": 0.43542, + "grad_norm": 0.8075554884969343, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 43542 + }, + { + "epoch": 0.43543, + "grad_norm": 0.7661530123197074, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 43543 + }, + { + "epoch": 0.43544, + "grad_norm": 0.802794329017145, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 43544 + }, + { + "epoch": 0.43545, + "grad_norm": 1.0028360073228353, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 43545 + }, + { + "epoch": 0.43546, + "grad_norm": 1.1926211282080805, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 43546 + }, + { + "epoch": 0.43547, + "grad_norm": 0.8990033681006022, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 43547 + }, + { + "epoch": 0.43548, + "grad_norm": 1.22324230273949, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 43548 + }, + { + "epoch": 0.43549, + "grad_norm": 1.2089363257309909, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 43549 + }, + { + "epoch": 0.4355, + "grad_norm": 1.1135428173301385, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 43550 + }, + { + "epoch": 0.43551, + "grad_norm": 1.0731347092502252, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 43551 + }, + { + "epoch": 0.43552, + "grad_norm": 1.228016165840674, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 43552 + }, + { + "epoch": 0.43553, + "grad_norm": 1.0148397911659306, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 43553 + }, + { + "epoch": 0.43554, + "grad_norm": 1.1731286631525366, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 43554 + }, + { + "epoch": 0.43555, + "grad_norm": 0.8469375244132795, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 43555 + }, + { + "epoch": 0.43556, + "grad_norm": 0.8181379685218291, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 43556 + }, + { + "epoch": 0.43557, + "grad_norm": 1.0360228050143845, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 43557 + }, + { + "epoch": 0.43558, + "grad_norm": 1.2108974740842096, + "learning_rate": 0.003, + "loss": 4.04, + "step": 43558 + }, + { + "epoch": 0.43559, + "grad_norm": 0.7752678146947485, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 43559 + }, + { + "epoch": 0.4356, + "grad_norm": 0.8573590580959194, + "learning_rate": 0.003, + "loss": 4.016, + "step": 43560 + }, + { + "epoch": 0.43561, + "grad_norm": 0.7937087856490457, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 43561 + }, + { + "epoch": 0.43562, + "grad_norm": 0.8700419446447727, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 43562 + }, + { + "epoch": 0.43563, + "grad_norm": 1.0134748392940427, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 43563 + }, + { + "epoch": 0.43564, + "grad_norm": 1.198635064783647, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 43564 + }, + { + "epoch": 0.43565, + "grad_norm": 0.7660830508470607, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 43565 + }, + { + "epoch": 0.43566, + "grad_norm": 1.103685092716043, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 43566 + }, + { + "epoch": 0.43567, + "grad_norm": 1.654345777740366, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 43567 + }, + { + "epoch": 0.43568, + "grad_norm": 0.6748301508025176, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 43568 + }, + { + "epoch": 0.43569, + "grad_norm": 0.8056851181583566, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 43569 + }, + { + "epoch": 0.4357, + "grad_norm": 0.7595313500884966, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 43570 + }, + { + "epoch": 0.43571, + "grad_norm": 0.7548321583094858, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 43571 + }, + { + "epoch": 0.43572, + "grad_norm": 0.8170192434035697, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 43572 + }, + { + "epoch": 0.43573, + "grad_norm": 1.152390622071344, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 43573 + }, + { + "epoch": 0.43574, + "grad_norm": 1.1783040194851147, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 43574 + }, + { + "epoch": 0.43575, + "grad_norm": 1.074256375531784, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 43575 + }, + { + "epoch": 0.43576, + "grad_norm": 1.1971932263194367, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 43576 + }, + { + "epoch": 0.43577, + "grad_norm": 1.3346767751172026, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 43577 + }, + { + "epoch": 0.43578, + "grad_norm": 1.0812817332101858, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 43578 + }, + { + "epoch": 0.43579, + "grad_norm": 1.0891084674824618, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 43579 + }, + { + "epoch": 0.4358, + "grad_norm": 1.1262753338237637, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 43580 + }, + { + "epoch": 0.43581, + "grad_norm": 1.2268552747429753, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 43581 + }, + { + "epoch": 0.43582, + "grad_norm": 0.8776687898985402, + "learning_rate": 0.003, + "loss": 4.03, + "step": 43582 + }, + { + "epoch": 0.43583, + "grad_norm": 0.8219061408641452, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 43583 + }, + { + "epoch": 0.43584, + "grad_norm": 0.9399512645514136, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 43584 + }, + { + "epoch": 0.43585, + "grad_norm": 1.0804539718304291, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 43585 + }, + { + "epoch": 0.43586, + "grad_norm": 1.228154258237157, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 43586 + }, + { + "epoch": 0.43587, + "grad_norm": 0.9660638253534514, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 43587 + }, + { + "epoch": 0.43588, + "grad_norm": 1.0608826875556376, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 43588 + }, + { + "epoch": 0.43589, + "grad_norm": 1.1586386637490578, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 43589 + }, + { + "epoch": 0.4359, + "grad_norm": 1.1109969886423763, + "learning_rate": 0.003, + "loss": 4.036, + "step": 43590 + }, + { + "epoch": 0.43591, + "grad_norm": 1.05677648847504, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 43591 + }, + { + "epoch": 0.43592, + "grad_norm": 1.0497069131178653, + "learning_rate": 0.003, + "loss": 4.029, + "step": 43592 + }, + { + "epoch": 0.43593, + "grad_norm": 1.1382297376116997, + "learning_rate": 0.003, + "loss": 4.045, + "step": 43593 + }, + { + "epoch": 0.43594, + "grad_norm": 1.0630959610331998, + "learning_rate": 0.003, + "loss": 4.037, + "step": 43594 + }, + { + "epoch": 0.43595, + "grad_norm": 1.0747938932157812, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 43595 + }, + { + "epoch": 0.43596, + "grad_norm": 1.2226300664735628, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 43596 + }, + { + "epoch": 0.43597, + "grad_norm": 1.1980468442049068, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 43597 + }, + { + "epoch": 0.43598, + "grad_norm": 1.1952515481039143, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 43598 + }, + { + "epoch": 0.43599, + "grad_norm": 0.8963660866379642, + "learning_rate": 0.003, + "loss": 4.023, + "step": 43599 + }, + { + "epoch": 0.436, + "grad_norm": 0.8671434025297168, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 43600 + }, + { + "epoch": 0.43601, + "grad_norm": 1.0588841123222958, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 43601 + }, + { + "epoch": 0.43602, + "grad_norm": 1.218243469008554, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 43602 + }, + { + "epoch": 0.43603, + "grad_norm": 0.9593741739434603, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 43603 + }, + { + "epoch": 0.43604, + "grad_norm": 1.3136413879588935, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 43604 + }, + { + "epoch": 0.43605, + "grad_norm": 0.8362610681359154, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 43605 + }, + { + "epoch": 0.43606, + "grad_norm": 0.8328945065786232, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 43606 + }, + { + "epoch": 0.43607, + "grad_norm": 0.9993949063926937, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 43607 + }, + { + "epoch": 0.43608, + "grad_norm": 1.422393886264785, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 43608 + }, + { + "epoch": 0.43609, + "grad_norm": 0.8838280101797079, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 43609 + }, + { + "epoch": 0.4361, + "grad_norm": 1.0644932909440092, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 43610 + }, + { + "epoch": 0.43611, + "grad_norm": 1.2593336118535328, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 43611 + }, + { + "epoch": 0.43612, + "grad_norm": 1.2192001622870128, + "learning_rate": 0.003, + "loss": 4.056, + "step": 43612 + }, + { + "epoch": 0.43613, + "grad_norm": 0.7789849310284294, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 43613 + }, + { + "epoch": 0.43614, + "grad_norm": 0.6926547455601463, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 43614 + }, + { + "epoch": 0.43615, + "grad_norm": 0.8865656703690064, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 43615 + }, + { + "epoch": 0.43616, + "grad_norm": 1.0637846270589555, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 43616 + }, + { + "epoch": 0.43617, + "grad_norm": 1.044003408693503, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 43617 + }, + { + "epoch": 0.43618, + "grad_norm": 1.012212719512513, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 43618 + }, + { + "epoch": 0.43619, + "grad_norm": 1.1895442056174617, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 43619 + }, + { + "epoch": 0.4362, + "grad_norm": 1.0326748092024023, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 43620 + }, + { + "epoch": 0.43621, + "grad_norm": 1.238738817963252, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 43621 + }, + { + "epoch": 0.43622, + "grad_norm": 1.0832480206275252, + "learning_rate": 0.003, + "loss": 4.032, + "step": 43622 + }, + { + "epoch": 0.43623, + "grad_norm": 1.13270862508782, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 43623 + }, + { + "epoch": 0.43624, + "grad_norm": 1.1894591005224682, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 43624 + }, + { + "epoch": 0.43625, + "grad_norm": 1.1227119307154794, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 43625 + }, + { + "epoch": 0.43626, + "grad_norm": 0.968429506609206, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 43626 + }, + { + "epoch": 0.43627, + "grad_norm": 0.9112148120652818, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 43627 + }, + { + "epoch": 0.43628, + "grad_norm": 0.9318842733561754, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 43628 + }, + { + "epoch": 0.43629, + "grad_norm": 0.983947252300882, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 43629 + }, + { + "epoch": 0.4363, + "grad_norm": 1.2370021502459312, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 43630 + }, + { + "epoch": 0.43631, + "grad_norm": 0.9905047557406356, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 43631 + }, + { + "epoch": 0.43632, + "grad_norm": 1.5553552758004865, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 43632 + }, + { + "epoch": 0.43633, + "grad_norm": 0.7507295378830195, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 43633 + }, + { + "epoch": 0.43634, + "grad_norm": 0.8182484639215346, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 43634 + }, + { + "epoch": 0.43635, + "grad_norm": 1.0171872823649017, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 43635 + }, + { + "epoch": 0.43636, + "grad_norm": 1.1630754066644402, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 43636 + }, + { + "epoch": 0.43637, + "grad_norm": 1.0022061462803191, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 43637 + }, + { + "epoch": 0.43638, + "grad_norm": 1.2847803335471877, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 43638 + }, + { + "epoch": 0.43639, + "grad_norm": 1.0407172128174276, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 43639 + }, + { + "epoch": 0.4364, + "grad_norm": 0.9001829260532196, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 43640 + }, + { + "epoch": 0.43641, + "grad_norm": 1.0876120097685538, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 43641 + }, + { + "epoch": 0.43642, + "grad_norm": 1.284151614425097, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 43642 + }, + { + "epoch": 0.43643, + "grad_norm": 1.0367250077018197, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 43643 + }, + { + "epoch": 0.43644, + "grad_norm": 1.0834524629129603, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 43644 + }, + { + "epoch": 0.43645, + "grad_norm": 1.2662450601726727, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 43645 + }, + { + "epoch": 0.43646, + "grad_norm": 1.2687152885872428, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 43646 + }, + { + "epoch": 0.43647, + "grad_norm": 0.9287663478583098, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 43647 + }, + { + "epoch": 0.43648, + "grad_norm": 1.0580778981492724, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 43648 + }, + { + "epoch": 0.43649, + "grad_norm": 1.539472184108297, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 43649 + }, + { + "epoch": 0.4365, + "grad_norm": 0.7504030307294962, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 43650 + }, + { + "epoch": 0.43651, + "grad_norm": 0.8947923644401077, + "learning_rate": 0.003, + "loss": 4.035, + "step": 43651 + }, + { + "epoch": 0.43652, + "grad_norm": 1.066643568672229, + "learning_rate": 0.003, + "loss": 4.03, + "step": 43652 + }, + { + "epoch": 0.43653, + "grad_norm": 1.2636390110455764, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 43653 + }, + { + "epoch": 0.43654, + "grad_norm": 0.8765343872714663, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 43654 + }, + { + "epoch": 0.43655, + "grad_norm": 1.1183178369807325, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 43655 + }, + { + "epoch": 0.43656, + "grad_norm": 1.1612672484781994, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 43656 + }, + { + "epoch": 0.43657, + "grad_norm": 1.074430352294401, + "learning_rate": 0.003, + "loss": 3.996, + "step": 43657 + }, + { + "epoch": 0.43658, + "grad_norm": 1.1670528057137375, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 43658 + }, + { + "epoch": 0.43659, + "grad_norm": 0.9603541505076038, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 43659 + }, + { + "epoch": 0.4366, + "grad_norm": 1.2396479412675931, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 43660 + }, + { + "epoch": 0.43661, + "grad_norm": 1.3024022993625104, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 43661 + }, + { + "epoch": 0.43662, + "grad_norm": 1.1605381913035349, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 43662 + }, + { + "epoch": 0.43663, + "grad_norm": 0.8827230502393094, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 43663 + }, + { + "epoch": 0.43664, + "grad_norm": 0.8036391219472266, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 43664 + }, + { + "epoch": 0.43665, + "grad_norm": 1.0135583472474914, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 43665 + }, + { + "epoch": 0.43666, + "grad_norm": 1.268420824959316, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 43666 + }, + { + "epoch": 0.43667, + "grad_norm": 0.8695107410144163, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 43667 + }, + { + "epoch": 0.43668, + "grad_norm": 0.8735095817501685, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 43668 + }, + { + "epoch": 0.43669, + "grad_norm": 1.1088514787791137, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 43669 + }, + { + "epoch": 0.4367, + "grad_norm": 1.3370769021119049, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 43670 + }, + { + "epoch": 0.43671, + "grad_norm": 0.850502955986534, + "learning_rate": 0.003, + "loss": 4.055, + "step": 43671 + }, + { + "epoch": 0.43672, + "grad_norm": 0.8104483746758531, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 43672 + }, + { + "epoch": 0.43673, + "grad_norm": 0.861248478179829, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 43673 + }, + { + "epoch": 0.43674, + "grad_norm": 0.8546601285295826, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 43674 + }, + { + "epoch": 0.43675, + "grad_norm": 0.992391436592568, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 43675 + }, + { + "epoch": 0.43676, + "grad_norm": 1.2567187169893244, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 43676 + }, + { + "epoch": 0.43677, + "grad_norm": 1.1516719977433092, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 43677 + }, + { + "epoch": 0.43678, + "grad_norm": 1.021042263020072, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 43678 + }, + { + "epoch": 0.43679, + "grad_norm": 1.1073812527883393, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 43679 + }, + { + "epoch": 0.4368, + "grad_norm": 1.0051273402039538, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 43680 + }, + { + "epoch": 0.43681, + "grad_norm": 1.13813142424424, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 43681 + }, + { + "epoch": 0.43682, + "grad_norm": 1.0005175645108513, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 43682 + }, + { + "epoch": 0.43683, + "grad_norm": 1.467962494458747, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 43683 + }, + { + "epoch": 0.43684, + "grad_norm": 1.1944870047559024, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 43684 + }, + { + "epoch": 0.43685, + "grad_norm": 1.1483787820566385, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 43685 + }, + { + "epoch": 0.43686, + "grad_norm": 0.9582278500993777, + "learning_rate": 0.003, + "loss": 4.048, + "step": 43686 + }, + { + "epoch": 0.43687, + "grad_norm": 0.9899190410672841, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 43687 + }, + { + "epoch": 0.43688, + "grad_norm": 1.1566378852275623, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 43688 + }, + { + "epoch": 0.43689, + "grad_norm": 1.210377718320136, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 43689 + }, + { + "epoch": 0.4369, + "grad_norm": 1.026404894729774, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 43690 + }, + { + "epoch": 0.43691, + "grad_norm": 1.117332840693603, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 43691 + }, + { + "epoch": 0.43692, + "grad_norm": 0.8145101416133221, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 43692 + }, + { + "epoch": 0.43693, + "grad_norm": 0.9034755526531739, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 43693 + }, + { + "epoch": 0.43694, + "grad_norm": 1.1076304829598849, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 43694 + }, + { + "epoch": 0.43695, + "grad_norm": 1.0383176878159022, + "learning_rate": 0.003, + "loss": 4.025, + "step": 43695 + }, + { + "epoch": 0.43696, + "grad_norm": 1.0151263791187548, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 43696 + }, + { + "epoch": 0.43697, + "grad_norm": 1.1571607887886775, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 43697 + }, + { + "epoch": 0.43698, + "grad_norm": 1.122984776917964, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 43698 + }, + { + "epoch": 0.43699, + "grad_norm": 1.0527586169665495, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 43699 + }, + { + "epoch": 0.437, + "grad_norm": 1.0836470067138901, + "learning_rate": 0.003, + "loss": 4.045, + "step": 43700 + }, + { + "epoch": 0.43701, + "grad_norm": 1.2128688878154184, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 43701 + }, + { + "epoch": 0.43702, + "grad_norm": 0.9769132221079466, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 43702 + }, + { + "epoch": 0.43703, + "grad_norm": 1.1709660839965705, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 43703 + }, + { + "epoch": 0.43704, + "grad_norm": 1.086519550017568, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 43704 + }, + { + "epoch": 0.43705, + "grad_norm": 1.0265551718204344, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 43705 + }, + { + "epoch": 0.43706, + "grad_norm": 1.1650168177513123, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 43706 + }, + { + "epoch": 0.43707, + "grad_norm": 0.9484123097359446, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 43707 + }, + { + "epoch": 0.43708, + "grad_norm": 1.0317290993161008, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 43708 + }, + { + "epoch": 0.43709, + "grad_norm": 1.1578996537952744, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 43709 + }, + { + "epoch": 0.4371, + "grad_norm": 0.9726992243328847, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 43710 + }, + { + "epoch": 0.43711, + "grad_norm": 0.9538621063918388, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 43711 + }, + { + "epoch": 0.43712, + "grad_norm": 1.1354799410765064, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 43712 + }, + { + "epoch": 0.43713, + "grad_norm": 1.1198798398096963, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 43713 + }, + { + "epoch": 0.43714, + "grad_norm": 1.2358841080099652, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 43714 + }, + { + "epoch": 0.43715, + "grad_norm": 1.12078593071531, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 43715 + }, + { + "epoch": 0.43716, + "grad_norm": 0.986220294656883, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 43716 + }, + { + "epoch": 0.43717, + "grad_norm": 1.151186112888998, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 43717 + }, + { + "epoch": 0.43718, + "grad_norm": 1.145313573353755, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 43718 + }, + { + "epoch": 0.43719, + "grad_norm": 1.0755980342939702, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 43719 + }, + { + "epoch": 0.4372, + "grad_norm": 1.4227559837760062, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 43720 + }, + { + "epoch": 0.43721, + "grad_norm": 0.959983706980432, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 43721 + }, + { + "epoch": 0.43722, + "grad_norm": 1.1181105151693964, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 43722 + }, + { + "epoch": 0.43723, + "grad_norm": 0.8813777075075969, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 43723 + }, + { + "epoch": 0.43724, + "grad_norm": 0.948987488354536, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 43724 + }, + { + "epoch": 0.43725, + "grad_norm": 1.013528475895237, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 43725 + }, + { + "epoch": 0.43726, + "grad_norm": 0.9724315783434199, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 43726 + }, + { + "epoch": 0.43727, + "grad_norm": 1.1552453834665861, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 43727 + }, + { + "epoch": 0.43728, + "grad_norm": 1.1881061599715208, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 43728 + }, + { + "epoch": 0.43729, + "grad_norm": 1.163730474622556, + "learning_rate": 0.003, + "loss": 4.035, + "step": 43729 + }, + { + "epoch": 0.4373, + "grad_norm": 0.8954729618273741, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 43730 + }, + { + "epoch": 0.43731, + "grad_norm": 0.9084233079089724, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 43731 + }, + { + "epoch": 0.43732, + "grad_norm": 1.163837072573458, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 43732 + }, + { + "epoch": 0.43733, + "grad_norm": 1.4228538997364057, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 43733 + }, + { + "epoch": 0.43734, + "grad_norm": 0.8447333842924808, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 43734 + }, + { + "epoch": 0.43735, + "grad_norm": 0.8383954800375036, + "learning_rate": 0.003, + "loss": 4.046, + "step": 43735 + }, + { + "epoch": 0.43736, + "grad_norm": 0.8665299952318504, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 43736 + }, + { + "epoch": 0.43737, + "grad_norm": 1.1436063611033762, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 43737 + }, + { + "epoch": 0.43738, + "grad_norm": 0.9914851317730213, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 43738 + }, + { + "epoch": 0.43739, + "grad_norm": 0.945610215059818, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 43739 + }, + { + "epoch": 0.4374, + "grad_norm": 0.9924677361016572, + "learning_rate": 0.003, + "loss": 4.015, + "step": 43740 + }, + { + "epoch": 0.43741, + "grad_norm": 1.2045260343330506, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 43741 + }, + { + "epoch": 0.43742, + "grad_norm": 0.9555439579952273, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 43742 + }, + { + "epoch": 0.43743, + "grad_norm": 0.9977207567783644, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 43743 + }, + { + "epoch": 0.43744, + "grad_norm": 1.0461516149830146, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 43744 + }, + { + "epoch": 0.43745, + "grad_norm": 1.1076311443243387, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 43745 + }, + { + "epoch": 0.43746, + "grad_norm": 1.3179978188595238, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 43746 + }, + { + "epoch": 0.43747, + "grad_norm": 1.1619188717116538, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 43747 + }, + { + "epoch": 0.43748, + "grad_norm": 1.1937076155142443, + "learning_rate": 0.003, + "loss": 4.064, + "step": 43748 + }, + { + "epoch": 0.43749, + "grad_norm": 0.9574609168857913, + "learning_rate": 0.003, + "loss": 4.017, + "step": 43749 + }, + { + "epoch": 0.4375, + "grad_norm": 0.7604815866912914, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 43750 + }, + { + "epoch": 0.43751, + "grad_norm": 0.8228393638628926, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 43751 + }, + { + "epoch": 0.43752, + "grad_norm": 1.075256698460484, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 43752 + }, + { + "epoch": 0.43753, + "grad_norm": 1.3135929743156205, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 43753 + }, + { + "epoch": 0.43754, + "grad_norm": 1.2065639696757586, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 43754 + }, + { + "epoch": 0.43755, + "grad_norm": 1.0601676973071157, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 43755 + }, + { + "epoch": 0.43756, + "grad_norm": 1.168083409188499, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 43756 + }, + { + "epoch": 0.43757, + "grad_norm": 1.046252772035688, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 43757 + }, + { + "epoch": 0.43758, + "grad_norm": 1.0834245022039661, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 43758 + }, + { + "epoch": 0.43759, + "grad_norm": 1.0494012218767763, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 43759 + }, + { + "epoch": 0.4376, + "grad_norm": 1.137226542715762, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 43760 + }, + { + "epoch": 0.43761, + "grad_norm": 0.9842847129673543, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 43761 + }, + { + "epoch": 0.43762, + "grad_norm": 1.2437789268835546, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 43762 + }, + { + "epoch": 0.43763, + "grad_norm": 0.9858028754841256, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 43763 + }, + { + "epoch": 0.43764, + "grad_norm": 1.0976508197919066, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 43764 + }, + { + "epoch": 0.43765, + "grad_norm": 0.9664392356358046, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 43765 + }, + { + "epoch": 0.43766, + "grad_norm": 1.0243135256013665, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 43766 + }, + { + "epoch": 0.43767, + "grad_norm": 1.134577352476687, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 43767 + }, + { + "epoch": 0.43768, + "grad_norm": 1.2415921402610106, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 43768 + }, + { + "epoch": 0.43769, + "grad_norm": 0.8883098567954054, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 43769 + }, + { + "epoch": 0.4377, + "grad_norm": 1.1690718411462229, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 43770 + }, + { + "epoch": 0.43771, + "grad_norm": 1.1251950862346947, + "learning_rate": 0.003, + "loss": 3.997, + "step": 43771 + }, + { + "epoch": 0.43772, + "grad_norm": 0.9456931921728627, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 43772 + }, + { + "epoch": 0.43773, + "grad_norm": 1.0422119488277275, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 43773 + }, + { + "epoch": 0.43774, + "grad_norm": 1.200143483028237, + "learning_rate": 0.003, + "loss": 4.067, + "step": 43774 + }, + { + "epoch": 0.43775, + "grad_norm": 0.9638742924822222, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 43775 + }, + { + "epoch": 0.43776, + "grad_norm": 0.9733545087629261, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 43776 + }, + { + "epoch": 0.43777, + "grad_norm": 1.0037321549788754, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 43777 + }, + { + "epoch": 0.43778, + "grad_norm": 1.1648302596307363, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 43778 + }, + { + "epoch": 0.43779, + "grad_norm": 1.0577278486513284, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 43779 + }, + { + "epoch": 0.4378, + "grad_norm": 0.9738629939210426, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 43780 + }, + { + "epoch": 0.43781, + "grad_norm": 1.1013070424061546, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 43781 + }, + { + "epoch": 0.43782, + "grad_norm": 1.2083227124617355, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 43782 + }, + { + "epoch": 0.43783, + "grad_norm": 1.035341978968132, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 43783 + }, + { + "epoch": 0.43784, + "grad_norm": 1.121636992363356, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 43784 + }, + { + "epoch": 0.43785, + "grad_norm": 1.0856325625248624, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 43785 + }, + { + "epoch": 0.43786, + "grad_norm": 0.9644380953578744, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 43786 + }, + { + "epoch": 0.43787, + "grad_norm": 1.17683156497457, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 43787 + }, + { + "epoch": 0.43788, + "grad_norm": 0.9695207971015359, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 43788 + }, + { + "epoch": 0.43789, + "grad_norm": 1.2776782852225725, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 43789 + }, + { + "epoch": 0.4379, + "grad_norm": 1.1867981421232814, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 43790 + }, + { + "epoch": 0.43791, + "grad_norm": 1.1389750873392481, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 43791 + }, + { + "epoch": 0.43792, + "grad_norm": 1.1080277094750521, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 43792 + }, + { + "epoch": 0.43793, + "grad_norm": 1.0913073656757526, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 43793 + }, + { + "epoch": 0.43794, + "grad_norm": 1.1329208186259916, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 43794 + }, + { + "epoch": 0.43795, + "grad_norm": 0.9687910522082632, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 43795 + }, + { + "epoch": 0.43796, + "grad_norm": 1.0576012270017332, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 43796 + }, + { + "epoch": 0.43797, + "grad_norm": 1.1534919337403629, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 43797 + }, + { + "epoch": 0.43798, + "grad_norm": 0.9306782133047198, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 43798 + }, + { + "epoch": 0.43799, + "grad_norm": 1.02986803597724, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 43799 + }, + { + "epoch": 0.438, + "grad_norm": 1.0728884775224177, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 43800 + }, + { + "epoch": 0.43801, + "grad_norm": 0.996472523097025, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 43801 + }, + { + "epoch": 0.43802, + "grad_norm": 1.0584276062553395, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 43802 + }, + { + "epoch": 0.43803, + "grad_norm": 1.1507936069087439, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 43803 + }, + { + "epoch": 0.43804, + "grad_norm": 1.0085669353179145, + "learning_rate": 0.003, + "loss": 4.07, + "step": 43804 + }, + { + "epoch": 0.43805, + "grad_norm": 1.0889164339781667, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 43805 + }, + { + "epoch": 0.43806, + "grad_norm": 1.036816186188455, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 43806 + }, + { + "epoch": 0.43807, + "grad_norm": 1.1103773908301846, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 43807 + }, + { + "epoch": 0.43808, + "grad_norm": 0.9625388822826688, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 43808 + }, + { + "epoch": 0.43809, + "grad_norm": 1.1234756980496026, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 43809 + }, + { + "epoch": 0.4381, + "grad_norm": 1.4175678441069297, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 43810 + }, + { + "epoch": 0.43811, + "grad_norm": 1.0118974641593388, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 43811 + }, + { + "epoch": 0.43812, + "grad_norm": 1.4086109360756045, + "learning_rate": 0.003, + "loss": 4.057, + "step": 43812 + }, + { + "epoch": 0.43813, + "grad_norm": 0.823182614733845, + "learning_rate": 0.003, + "loss": 4.029, + "step": 43813 + }, + { + "epoch": 0.43814, + "grad_norm": 0.931306035963521, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 43814 + }, + { + "epoch": 0.43815, + "grad_norm": 1.2140418999649627, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 43815 + }, + { + "epoch": 0.43816, + "grad_norm": 1.168405707823527, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 43816 + }, + { + "epoch": 0.43817, + "grad_norm": 0.9006230174893332, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 43817 + }, + { + "epoch": 0.43818, + "grad_norm": 0.9375957704716693, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 43818 + }, + { + "epoch": 0.43819, + "grad_norm": 1.0777844952802766, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 43819 + }, + { + "epoch": 0.4382, + "grad_norm": 1.2552107700625377, + "learning_rate": 0.003, + "loss": 4.037, + "step": 43820 + }, + { + "epoch": 0.43821, + "grad_norm": 1.016913147244642, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 43821 + }, + { + "epoch": 0.43822, + "grad_norm": 1.2615874803498361, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 43822 + }, + { + "epoch": 0.43823, + "grad_norm": 1.129542373061986, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 43823 + }, + { + "epoch": 0.43824, + "grad_norm": 0.9435423192641802, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 43824 + }, + { + "epoch": 0.43825, + "grad_norm": 1.112292111077354, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 43825 + }, + { + "epoch": 0.43826, + "grad_norm": 1.137873554235583, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 43826 + }, + { + "epoch": 0.43827, + "grad_norm": 1.0434546313042832, + "learning_rate": 0.003, + "loss": 4.024, + "step": 43827 + }, + { + "epoch": 0.43828, + "grad_norm": 1.0568143736619589, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 43828 + }, + { + "epoch": 0.43829, + "grad_norm": 1.0569028015955193, + "learning_rate": 0.003, + "loss": 4.034, + "step": 43829 + }, + { + "epoch": 0.4383, + "grad_norm": 0.977614172717292, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 43830 + }, + { + "epoch": 0.43831, + "grad_norm": 1.1013109027248802, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 43831 + }, + { + "epoch": 0.43832, + "grad_norm": 1.252324899996786, + "learning_rate": 0.003, + "loss": 4.041, + "step": 43832 + }, + { + "epoch": 0.43833, + "grad_norm": 0.8973331441339119, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 43833 + }, + { + "epoch": 0.43834, + "grad_norm": 0.9881850645191913, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 43834 + }, + { + "epoch": 0.43835, + "grad_norm": 1.040005351160585, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 43835 + }, + { + "epoch": 0.43836, + "grad_norm": 1.1668921447509752, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 43836 + }, + { + "epoch": 0.43837, + "grad_norm": 1.0085377527890313, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 43837 + }, + { + "epoch": 0.43838, + "grad_norm": 1.0316443129176383, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 43838 + }, + { + "epoch": 0.43839, + "grad_norm": 1.2867811082093326, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 43839 + }, + { + "epoch": 0.4384, + "grad_norm": 1.2058663443677922, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 43840 + }, + { + "epoch": 0.43841, + "grad_norm": 1.0676878125205338, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 43841 + }, + { + "epoch": 0.43842, + "grad_norm": 0.8600856873032805, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 43842 + }, + { + "epoch": 0.43843, + "grad_norm": 0.9172553605544728, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 43843 + }, + { + "epoch": 0.43844, + "grad_norm": 1.0078601005292596, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 43844 + }, + { + "epoch": 0.43845, + "grad_norm": 1.192923741211192, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 43845 + }, + { + "epoch": 0.43846, + "grad_norm": 0.981090468579566, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 43846 + }, + { + "epoch": 0.43847, + "grad_norm": 1.001353570611885, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 43847 + }, + { + "epoch": 0.43848, + "grad_norm": 1.3990507362694882, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 43848 + }, + { + "epoch": 0.43849, + "grad_norm": 1.1721717432341026, + "learning_rate": 0.003, + "loss": 4.044, + "step": 43849 + }, + { + "epoch": 0.4385, + "grad_norm": 1.0825065770137967, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 43850 + }, + { + "epoch": 0.43851, + "grad_norm": 1.2424621680958772, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 43851 + }, + { + "epoch": 0.43852, + "grad_norm": 0.8585781754132805, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 43852 + }, + { + "epoch": 0.43853, + "grad_norm": 0.8521586323760577, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 43853 + }, + { + "epoch": 0.43854, + "grad_norm": 1.1944974151334702, + "learning_rate": 0.003, + "loss": 4.042, + "step": 43854 + }, + { + "epoch": 0.43855, + "grad_norm": 1.0031872555142467, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 43855 + }, + { + "epoch": 0.43856, + "grad_norm": 1.2859121620519076, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 43856 + }, + { + "epoch": 0.43857, + "grad_norm": 0.966886728622694, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 43857 + }, + { + "epoch": 0.43858, + "grad_norm": 1.1435257176413824, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 43858 + }, + { + "epoch": 0.43859, + "grad_norm": 1.114767604765179, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 43859 + }, + { + "epoch": 0.4386, + "grad_norm": 0.9979439995008124, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 43860 + }, + { + "epoch": 0.43861, + "grad_norm": 1.0461180417548428, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 43861 + }, + { + "epoch": 0.43862, + "grad_norm": 1.1011014110686521, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 43862 + }, + { + "epoch": 0.43863, + "grad_norm": 1.1807034732368562, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 43863 + }, + { + "epoch": 0.43864, + "grad_norm": 1.1315081633806383, + "learning_rate": 0.003, + "loss": 4.019, + "step": 43864 + }, + { + "epoch": 0.43865, + "grad_norm": 1.1381068581095721, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 43865 + }, + { + "epoch": 0.43866, + "grad_norm": 0.960194332068134, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 43866 + }, + { + "epoch": 0.43867, + "grad_norm": 1.0745148816251127, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 43867 + }, + { + "epoch": 0.43868, + "grad_norm": 1.2088919858690312, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 43868 + }, + { + "epoch": 0.43869, + "grad_norm": 0.8854239414885864, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 43869 + }, + { + "epoch": 0.4387, + "grad_norm": 0.9363961655594862, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 43870 + }, + { + "epoch": 0.43871, + "grad_norm": 0.9908984229709181, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 43871 + }, + { + "epoch": 0.43872, + "grad_norm": 1.208479003422296, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 43872 + }, + { + "epoch": 0.43873, + "grad_norm": 0.9758172824362247, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 43873 + }, + { + "epoch": 0.43874, + "grad_norm": 0.9767164067937657, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 43874 + }, + { + "epoch": 0.43875, + "grad_norm": 1.0485671445925302, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 43875 + }, + { + "epoch": 0.43876, + "grad_norm": 0.9670440867539266, + "learning_rate": 0.003, + "loss": 4.04, + "step": 43876 + }, + { + "epoch": 0.43877, + "grad_norm": 1.0026659654639651, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 43877 + }, + { + "epoch": 0.43878, + "grad_norm": 1.0303712021356042, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 43878 + }, + { + "epoch": 0.43879, + "grad_norm": 1.1616192911333374, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 43879 + }, + { + "epoch": 0.4388, + "grad_norm": 1.2030098942452152, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 43880 + }, + { + "epoch": 0.43881, + "grad_norm": 1.2011900084855223, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 43881 + }, + { + "epoch": 0.43882, + "grad_norm": 1.0767680513850477, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 43882 + }, + { + "epoch": 0.43883, + "grad_norm": 1.198919573352605, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 43883 + }, + { + "epoch": 0.43884, + "grad_norm": 1.2611330600765716, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 43884 + }, + { + "epoch": 0.43885, + "grad_norm": 0.8744005655216666, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 43885 + }, + { + "epoch": 0.43886, + "grad_norm": 0.8558632029096376, + "learning_rate": 0.003, + "loss": 4.036, + "step": 43886 + }, + { + "epoch": 0.43887, + "grad_norm": 0.8779940467350043, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 43887 + }, + { + "epoch": 0.43888, + "grad_norm": 0.9687464961251581, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 43888 + }, + { + "epoch": 0.43889, + "grad_norm": 1.0582134281262938, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 43889 + }, + { + "epoch": 0.4389, + "grad_norm": 1.1573195085899066, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 43890 + }, + { + "epoch": 0.43891, + "grad_norm": 1.1617415491510845, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 43891 + }, + { + "epoch": 0.43892, + "grad_norm": 1.1192139933734713, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 43892 + }, + { + "epoch": 0.43893, + "grad_norm": 0.8735099695789823, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 43893 + }, + { + "epoch": 0.43894, + "grad_norm": 0.8448581794735196, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 43894 + }, + { + "epoch": 0.43895, + "grad_norm": 0.9885588778048956, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 43895 + }, + { + "epoch": 0.43896, + "grad_norm": 1.4531649234783846, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 43896 + }, + { + "epoch": 0.43897, + "grad_norm": 0.6925001501098703, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 43897 + }, + { + "epoch": 0.43898, + "grad_norm": 0.8804993242077825, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 43898 + }, + { + "epoch": 0.43899, + "grad_norm": 1.1633304525497241, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 43899 + }, + { + "epoch": 0.439, + "grad_norm": 1.2427077288287876, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 43900 + }, + { + "epoch": 0.43901, + "grad_norm": 0.844870466644891, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 43901 + }, + { + "epoch": 0.43902, + "grad_norm": 0.8421956591432241, + "learning_rate": 0.003, + "loss": 4.032, + "step": 43902 + }, + { + "epoch": 0.43903, + "grad_norm": 0.8764783527316078, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 43903 + }, + { + "epoch": 0.43904, + "grad_norm": 1.2026105426398224, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 43904 + }, + { + "epoch": 0.43905, + "grad_norm": 1.3341011648085477, + "learning_rate": 0.003, + "loss": 4.042, + "step": 43905 + }, + { + "epoch": 0.43906, + "grad_norm": 0.8523227574048385, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 43906 + }, + { + "epoch": 0.43907, + "grad_norm": 0.6832711253946205, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 43907 + }, + { + "epoch": 0.43908, + "grad_norm": 0.8416321818959018, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 43908 + }, + { + "epoch": 0.43909, + "grad_norm": 1.1707258352519767, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 43909 + }, + { + "epoch": 0.4391, + "grad_norm": 1.070341588694884, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 43910 + }, + { + "epoch": 0.43911, + "grad_norm": 0.9879681840137321, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 43911 + }, + { + "epoch": 0.43912, + "grad_norm": 1.1433083059699227, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 43912 + }, + { + "epoch": 0.43913, + "grad_norm": 0.9844643410900737, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 43913 + }, + { + "epoch": 0.43914, + "grad_norm": 0.9651666811210222, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 43914 + }, + { + "epoch": 0.43915, + "grad_norm": 1.052073777633555, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 43915 + }, + { + "epoch": 0.43916, + "grad_norm": 1.0580400986147929, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 43916 + }, + { + "epoch": 0.43917, + "grad_norm": 1.0088763410872423, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 43917 + }, + { + "epoch": 0.43918, + "grad_norm": 0.9846919168430126, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 43918 + }, + { + "epoch": 0.43919, + "grad_norm": 1.0072939863896357, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 43919 + }, + { + "epoch": 0.4392, + "grad_norm": 1.3153202946615519, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 43920 + }, + { + "epoch": 0.43921, + "grad_norm": 1.137482696921022, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 43921 + }, + { + "epoch": 0.43922, + "grad_norm": 1.2252820820607295, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 43922 + }, + { + "epoch": 0.43923, + "grad_norm": 1.2295124021567474, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 43923 + }, + { + "epoch": 0.43924, + "grad_norm": 1.0887681041287158, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 43924 + }, + { + "epoch": 0.43925, + "grad_norm": 1.2542550835093962, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 43925 + }, + { + "epoch": 0.43926, + "grad_norm": 0.7916907454822253, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 43926 + }, + { + "epoch": 0.43927, + "grad_norm": 0.8483311941968394, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 43927 + }, + { + "epoch": 0.43928, + "grad_norm": 1.0335541811889237, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 43928 + }, + { + "epoch": 0.43929, + "grad_norm": 1.1411833278118115, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 43929 + }, + { + "epoch": 0.4393, + "grad_norm": 0.8813020802804987, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 43930 + }, + { + "epoch": 0.43931, + "grad_norm": 0.9510495467354841, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 43931 + }, + { + "epoch": 0.43932, + "grad_norm": 1.2471924355702928, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 43932 + }, + { + "epoch": 0.43933, + "grad_norm": 0.8991290933538594, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 43933 + }, + { + "epoch": 0.43934, + "grad_norm": 1.0062320227803239, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 43934 + }, + { + "epoch": 0.43935, + "grad_norm": 1.6574870574558873, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 43935 + }, + { + "epoch": 0.43936, + "grad_norm": 0.7643709124006869, + "learning_rate": 0.003, + "loss": 4.059, + "step": 43936 + }, + { + "epoch": 0.43937, + "grad_norm": 0.9089058392814352, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 43937 + }, + { + "epoch": 0.43938, + "grad_norm": 1.1111370445977045, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 43938 + }, + { + "epoch": 0.43939, + "grad_norm": 0.9045114516875238, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 43939 + }, + { + "epoch": 0.4394, + "grad_norm": 1.0097485958150298, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 43940 + }, + { + "epoch": 0.43941, + "grad_norm": 1.5592537260460269, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 43941 + }, + { + "epoch": 0.43942, + "grad_norm": 0.8970770281496707, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 43942 + }, + { + "epoch": 0.43943, + "grad_norm": 0.9588162441318003, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 43943 + }, + { + "epoch": 0.43944, + "grad_norm": 1.1296435626197687, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 43944 + }, + { + "epoch": 0.43945, + "grad_norm": 1.144152802322663, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 43945 + }, + { + "epoch": 0.43946, + "grad_norm": 1.003233747447363, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 43946 + }, + { + "epoch": 0.43947, + "grad_norm": 1.199758758280721, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 43947 + }, + { + "epoch": 0.43948, + "grad_norm": 1.1563126448542125, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 43948 + }, + { + "epoch": 0.43949, + "grad_norm": 1.0708569411012945, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 43949 + }, + { + "epoch": 0.4395, + "grad_norm": 0.9945043532593238, + "learning_rate": 0.003, + "loss": 4.034, + "step": 43950 + }, + { + "epoch": 0.43951, + "grad_norm": 1.0891029862860186, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 43951 + }, + { + "epoch": 0.43952, + "grad_norm": 1.2339422797547341, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 43952 + }, + { + "epoch": 0.43953, + "grad_norm": 1.029476923531424, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 43953 + }, + { + "epoch": 0.43954, + "grad_norm": 1.2768825984891443, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 43954 + }, + { + "epoch": 0.43955, + "grad_norm": 0.9809831935334908, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 43955 + }, + { + "epoch": 0.43956, + "grad_norm": 1.1349382403508377, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 43956 + }, + { + "epoch": 0.43957, + "grad_norm": 1.0144879406654213, + "learning_rate": 0.003, + "loss": 4.055, + "step": 43957 + }, + { + "epoch": 0.43958, + "grad_norm": 1.2945607470764782, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 43958 + }, + { + "epoch": 0.43959, + "grad_norm": 0.9630352376879913, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 43959 + }, + { + "epoch": 0.4396, + "grad_norm": 1.1046187974773094, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 43960 + }, + { + "epoch": 0.43961, + "grad_norm": 1.1357891884018692, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 43961 + }, + { + "epoch": 0.43962, + "grad_norm": 0.944313321275537, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 43962 + }, + { + "epoch": 0.43963, + "grad_norm": 1.1090538688767122, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 43963 + }, + { + "epoch": 0.43964, + "grad_norm": 0.9575381164768844, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 43964 + }, + { + "epoch": 0.43965, + "grad_norm": 1.037782141923561, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 43965 + }, + { + "epoch": 0.43966, + "grad_norm": 1.177266462255421, + "learning_rate": 0.003, + "loss": 4.005, + "step": 43966 + }, + { + "epoch": 0.43967, + "grad_norm": 1.2179321041732174, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 43967 + }, + { + "epoch": 0.43968, + "grad_norm": 1.2087912837879078, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 43968 + }, + { + "epoch": 0.43969, + "grad_norm": 0.7990415225329481, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 43969 + }, + { + "epoch": 0.4397, + "grad_norm": 0.7546357178871292, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 43970 + }, + { + "epoch": 0.43971, + "grad_norm": 0.8641438888616265, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 43971 + }, + { + "epoch": 0.43972, + "grad_norm": 0.9427145362731166, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 43972 + }, + { + "epoch": 0.43973, + "grad_norm": 1.2202526019562114, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 43973 + }, + { + "epoch": 0.43974, + "grad_norm": 1.2018644455068406, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 43974 + }, + { + "epoch": 0.43975, + "grad_norm": 0.9468975369008061, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 43975 + }, + { + "epoch": 0.43976, + "grad_norm": 1.0244054355423287, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 43976 + }, + { + "epoch": 0.43977, + "grad_norm": 1.1864907248027774, + "learning_rate": 0.003, + "loss": 4.051, + "step": 43977 + }, + { + "epoch": 0.43978, + "grad_norm": 0.7959831373139846, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 43978 + }, + { + "epoch": 0.43979, + "grad_norm": 0.8863517332939205, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 43979 + }, + { + "epoch": 0.4398, + "grad_norm": 1.1367980006507163, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 43980 + }, + { + "epoch": 0.43981, + "grad_norm": 1.0871492018060294, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 43981 + }, + { + "epoch": 0.43982, + "grad_norm": 1.1441741661153293, + "learning_rate": 0.003, + "loss": 4.037, + "step": 43982 + }, + { + "epoch": 0.43983, + "grad_norm": 1.0480054097752542, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 43983 + }, + { + "epoch": 0.43984, + "grad_norm": 1.1679555657460998, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 43984 + }, + { + "epoch": 0.43985, + "grad_norm": 1.0692593105801618, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 43985 + }, + { + "epoch": 0.43986, + "grad_norm": 1.177868157333052, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 43986 + }, + { + "epoch": 0.43987, + "grad_norm": 0.9287711448150725, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 43987 + }, + { + "epoch": 0.43988, + "grad_norm": 0.866816608101617, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 43988 + }, + { + "epoch": 0.43989, + "grad_norm": 0.937034906584837, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 43989 + }, + { + "epoch": 0.4399, + "grad_norm": 1.1596841370961033, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 43990 + }, + { + "epoch": 0.43991, + "grad_norm": 1.044423494847447, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 43991 + }, + { + "epoch": 0.43992, + "grad_norm": 1.1102982069182856, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 43992 + }, + { + "epoch": 0.43993, + "grad_norm": 1.120565307289818, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 43993 + }, + { + "epoch": 0.43994, + "grad_norm": 1.12150123631106, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 43994 + }, + { + "epoch": 0.43995, + "grad_norm": 0.9257014362837658, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 43995 + }, + { + "epoch": 0.43996, + "grad_norm": 1.1826680575266864, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 43996 + }, + { + "epoch": 0.43997, + "grad_norm": 1.1348988482287081, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 43997 + }, + { + "epoch": 0.43998, + "grad_norm": 0.9481286572501041, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 43998 + }, + { + "epoch": 0.43999, + "grad_norm": 0.9574738956098188, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 43999 + }, + { + "epoch": 0.44, + "grad_norm": 1.145407880427967, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 44000 + }, + { + "epoch": 0.44001, + "grad_norm": 1.189206063387876, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 44001 + }, + { + "epoch": 0.44002, + "grad_norm": 0.9949940196667237, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 44002 + }, + { + "epoch": 0.44003, + "grad_norm": 0.9924540982530762, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 44003 + }, + { + "epoch": 0.44004, + "grad_norm": 1.1402974428841013, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 44004 + }, + { + "epoch": 0.44005, + "grad_norm": 1.2573571913553754, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 44005 + }, + { + "epoch": 0.44006, + "grad_norm": 0.7727310973653491, + "learning_rate": 0.003, + "loss": 4.013, + "step": 44006 + }, + { + "epoch": 0.44007, + "grad_norm": 0.8299795961001388, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 44007 + }, + { + "epoch": 0.44008, + "grad_norm": 1.037882464855054, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 44008 + }, + { + "epoch": 0.44009, + "grad_norm": 1.2586989321044448, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 44009 + }, + { + "epoch": 0.4401, + "grad_norm": 0.90403322647741, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 44010 + }, + { + "epoch": 0.44011, + "grad_norm": 1.0370146492045569, + "learning_rate": 0.003, + "loss": 4.013, + "step": 44011 + }, + { + "epoch": 0.44012, + "grad_norm": 1.2704718690507342, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 44012 + }, + { + "epoch": 0.44013, + "grad_norm": 1.1561636185310389, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 44013 + }, + { + "epoch": 0.44014, + "grad_norm": 0.9504830926675488, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 44014 + }, + { + "epoch": 0.44015, + "grad_norm": 0.9640645896851497, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 44015 + }, + { + "epoch": 0.44016, + "grad_norm": 1.017551181545891, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 44016 + }, + { + "epoch": 0.44017, + "grad_norm": 1.0713142081586269, + "learning_rate": 0.003, + "loss": 4.026, + "step": 44017 + }, + { + "epoch": 0.44018, + "grad_norm": 0.8618507898889811, + "learning_rate": 0.003, + "loss": 4.02, + "step": 44018 + }, + { + "epoch": 0.44019, + "grad_norm": 0.9553155593682614, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 44019 + }, + { + "epoch": 0.4402, + "grad_norm": 1.081409577964404, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 44020 + }, + { + "epoch": 0.44021, + "grad_norm": 0.9793731097237786, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 44021 + }, + { + "epoch": 0.44022, + "grad_norm": 1.2236738741882143, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 44022 + }, + { + "epoch": 0.44023, + "grad_norm": 1.4182440590390706, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 44023 + }, + { + "epoch": 0.44024, + "grad_norm": 0.996181369460147, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 44024 + }, + { + "epoch": 0.44025, + "grad_norm": 1.043120403015833, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 44025 + }, + { + "epoch": 0.44026, + "grad_norm": 1.1873758640437984, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 44026 + }, + { + "epoch": 0.44027, + "grad_norm": 1.1267170575374024, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 44027 + }, + { + "epoch": 0.44028, + "grad_norm": 1.0635106604521578, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 44028 + }, + { + "epoch": 0.44029, + "grad_norm": 1.122792330714898, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 44029 + }, + { + "epoch": 0.4403, + "grad_norm": 0.991165102182839, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 44030 + }, + { + "epoch": 0.44031, + "grad_norm": 1.0564373728786676, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 44031 + }, + { + "epoch": 0.44032, + "grad_norm": 0.9961671663720837, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 44032 + }, + { + "epoch": 0.44033, + "grad_norm": 1.1061176588279467, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 44033 + }, + { + "epoch": 0.44034, + "grad_norm": 1.0760305461474038, + "learning_rate": 0.003, + "loss": 4.027, + "step": 44034 + }, + { + "epoch": 0.44035, + "grad_norm": 1.143695003531761, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 44035 + }, + { + "epoch": 0.44036, + "grad_norm": 1.0417854011310652, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 44036 + }, + { + "epoch": 0.44037, + "grad_norm": 1.085171463604108, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 44037 + }, + { + "epoch": 0.44038, + "grad_norm": 1.1056502454605446, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 44038 + }, + { + "epoch": 0.44039, + "grad_norm": 1.2052682409168147, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 44039 + }, + { + "epoch": 0.4404, + "grad_norm": 0.9926156925140698, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 44040 + }, + { + "epoch": 0.44041, + "grad_norm": 1.1081563106038332, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 44041 + }, + { + "epoch": 0.44042, + "grad_norm": 1.0429628349479576, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 44042 + }, + { + "epoch": 0.44043, + "grad_norm": 1.062572102543067, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 44043 + }, + { + "epoch": 0.44044, + "grad_norm": 1.0265573044849585, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 44044 + }, + { + "epoch": 0.44045, + "grad_norm": 0.9920159757973186, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 44045 + }, + { + "epoch": 0.44046, + "grad_norm": 1.0797105284305253, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 44046 + }, + { + "epoch": 0.44047, + "grad_norm": 1.0026726487322897, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 44047 + }, + { + "epoch": 0.44048, + "grad_norm": 1.3159349578378645, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 44048 + }, + { + "epoch": 0.44049, + "grad_norm": 1.0048080051504558, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 44049 + }, + { + "epoch": 0.4405, + "grad_norm": 1.1792156536636411, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 44050 + }, + { + "epoch": 0.44051, + "grad_norm": 0.9776026184504532, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 44051 + }, + { + "epoch": 0.44052, + "grad_norm": 1.0561655414237248, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 44052 + }, + { + "epoch": 0.44053, + "grad_norm": 1.0844130947230552, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 44053 + }, + { + "epoch": 0.44054, + "grad_norm": 0.9825810888798419, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 44054 + }, + { + "epoch": 0.44055, + "grad_norm": 1.0417796122016503, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 44055 + }, + { + "epoch": 0.44056, + "grad_norm": 1.101216535851065, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 44056 + }, + { + "epoch": 0.44057, + "grad_norm": 1.049646468000422, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 44057 + }, + { + "epoch": 0.44058, + "grad_norm": 1.132959906417624, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 44058 + }, + { + "epoch": 0.44059, + "grad_norm": 1.1443563183790442, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 44059 + }, + { + "epoch": 0.4406, + "grad_norm": 1.05595531531078, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 44060 + }, + { + "epoch": 0.44061, + "grad_norm": 1.1236914099475275, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 44061 + }, + { + "epoch": 0.44062, + "grad_norm": 1.1299387459033212, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 44062 + }, + { + "epoch": 0.44063, + "grad_norm": 0.9928236717326593, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 44063 + }, + { + "epoch": 0.44064, + "grad_norm": 1.1422534671589615, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 44064 + }, + { + "epoch": 0.44065, + "grad_norm": 1.0360968898171468, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 44065 + }, + { + "epoch": 0.44066, + "grad_norm": 1.2316995367458072, + "learning_rate": 0.003, + "loss": 4.045, + "step": 44066 + }, + { + "epoch": 0.44067, + "grad_norm": 0.9531526663667218, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 44067 + }, + { + "epoch": 0.44068, + "grad_norm": 1.040475622719196, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 44068 + }, + { + "epoch": 0.44069, + "grad_norm": 1.2138713404455008, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 44069 + }, + { + "epoch": 0.4407, + "grad_norm": 0.9166036326966542, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 44070 + }, + { + "epoch": 0.44071, + "grad_norm": 1.1472738690945208, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 44071 + }, + { + "epoch": 0.44072, + "grad_norm": 1.188898523528245, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 44072 + }, + { + "epoch": 0.44073, + "grad_norm": 0.9568519162897869, + "learning_rate": 0.003, + "loss": 4.068, + "step": 44073 + }, + { + "epoch": 0.44074, + "grad_norm": 1.053948566816374, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 44074 + }, + { + "epoch": 0.44075, + "grad_norm": 1.3325279492864708, + "learning_rate": 0.003, + "loss": 4.043, + "step": 44075 + }, + { + "epoch": 0.44076, + "grad_norm": 0.9588568967485963, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 44076 + }, + { + "epoch": 0.44077, + "grad_norm": 1.1281041971030317, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 44077 + }, + { + "epoch": 0.44078, + "grad_norm": 1.205040369783287, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 44078 + }, + { + "epoch": 0.44079, + "grad_norm": 0.9829775467601148, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 44079 + }, + { + "epoch": 0.4408, + "grad_norm": 1.0569793667795493, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 44080 + }, + { + "epoch": 0.44081, + "grad_norm": 1.0742544633190747, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 44081 + }, + { + "epoch": 0.44082, + "grad_norm": 1.1708631705489494, + "learning_rate": 0.003, + "loss": 4.03, + "step": 44082 + }, + { + "epoch": 0.44083, + "grad_norm": 1.069376281461237, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 44083 + }, + { + "epoch": 0.44084, + "grad_norm": 0.9069136938568804, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 44084 + }, + { + "epoch": 0.44085, + "grad_norm": 0.8944160792230699, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 44085 + }, + { + "epoch": 0.44086, + "grad_norm": 0.9711003140710972, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 44086 + }, + { + "epoch": 0.44087, + "grad_norm": 1.1547677607292648, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 44087 + }, + { + "epoch": 0.44088, + "grad_norm": 0.7864032578220538, + "learning_rate": 0.003, + "loss": 4.036, + "step": 44088 + }, + { + "epoch": 0.44089, + "grad_norm": 1.0128555654869347, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 44089 + }, + { + "epoch": 0.4409, + "grad_norm": 1.2595208393238533, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 44090 + }, + { + "epoch": 0.44091, + "grad_norm": 0.9426883720452509, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 44091 + }, + { + "epoch": 0.44092, + "grad_norm": 1.1082012776500585, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 44092 + }, + { + "epoch": 0.44093, + "grad_norm": 1.112927846739963, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 44093 + }, + { + "epoch": 0.44094, + "grad_norm": 0.9891882703177156, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 44094 + }, + { + "epoch": 0.44095, + "grad_norm": 1.1564957598330978, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 44095 + }, + { + "epoch": 0.44096, + "grad_norm": 1.1305336964239594, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 44096 + }, + { + "epoch": 0.44097, + "grad_norm": 0.946767641986737, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 44097 + }, + { + "epoch": 0.44098, + "grad_norm": 0.8701738458329233, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 44098 + }, + { + "epoch": 0.44099, + "grad_norm": 0.9120657736054822, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 44099 + }, + { + "epoch": 0.441, + "grad_norm": 1.0348702671465306, + "learning_rate": 0.003, + "loss": 4.037, + "step": 44100 + }, + { + "epoch": 0.44101, + "grad_norm": 1.177027372365076, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 44101 + }, + { + "epoch": 0.44102, + "grad_norm": 0.9050212102395557, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 44102 + }, + { + "epoch": 0.44103, + "grad_norm": 1.0776651372621406, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 44103 + }, + { + "epoch": 0.44104, + "grad_norm": 1.006580969201562, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 44104 + }, + { + "epoch": 0.44105, + "grad_norm": 1.1861151436136659, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 44105 + }, + { + "epoch": 0.44106, + "grad_norm": 0.8850116414830312, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 44106 + }, + { + "epoch": 0.44107, + "grad_norm": 1.3096296357001034, + "learning_rate": 0.003, + "loss": 4.012, + "step": 44107 + }, + { + "epoch": 0.44108, + "grad_norm": 1.2295459339301824, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 44108 + }, + { + "epoch": 0.44109, + "grad_norm": 1.0613541698843139, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 44109 + }, + { + "epoch": 0.4411, + "grad_norm": 1.3167173135138759, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 44110 + }, + { + "epoch": 0.44111, + "grad_norm": 0.9305965159529362, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 44111 + }, + { + "epoch": 0.44112, + "grad_norm": 0.772872253941486, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 44112 + }, + { + "epoch": 0.44113, + "grad_norm": 1.0021507286891842, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 44113 + }, + { + "epoch": 0.44114, + "grad_norm": 1.5017881962674975, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 44114 + }, + { + "epoch": 0.44115, + "grad_norm": 0.7388770526951657, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 44115 + }, + { + "epoch": 0.44116, + "grad_norm": 0.8296852287084138, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 44116 + }, + { + "epoch": 0.44117, + "grad_norm": 1.071286471109705, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 44117 + }, + { + "epoch": 0.44118, + "grad_norm": 1.3787170025373043, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 44118 + }, + { + "epoch": 0.44119, + "grad_norm": 0.8077967968412368, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 44119 + }, + { + "epoch": 0.4412, + "grad_norm": 0.79345280577412, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 44120 + }, + { + "epoch": 0.44121, + "grad_norm": 1.0065395495922755, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 44121 + }, + { + "epoch": 0.44122, + "grad_norm": 1.1958175031033538, + "learning_rate": 0.003, + "loss": 4.025, + "step": 44122 + }, + { + "epoch": 0.44123, + "grad_norm": 0.9847282003801755, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 44123 + }, + { + "epoch": 0.44124, + "grad_norm": 1.0265819192976329, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 44124 + }, + { + "epoch": 0.44125, + "grad_norm": 1.133367547336492, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 44125 + }, + { + "epoch": 0.44126, + "grad_norm": 0.9242558321877242, + "learning_rate": 0.003, + "loss": 4.049, + "step": 44126 + }, + { + "epoch": 0.44127, + "grad_norm": 1.028900150629591, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 44127 + }, + { + "epoch": 0.44128, + "grad_norm": 1.0731599944627463, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 44128 + }, + { + "epoch": 0.44129, + "grad_norm": 1.2245002388892914, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 44129 + }, + { + "epoch": 0.4413, + "grad_norm": 0.9987582688881468, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 44130 + }, + { + "epoch": 0.44131, + "grad_norm": 1.191623921002882, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 44131 + }, + { + "epoch": 0.44132, + "grad_norm": 1.0812407902640524, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 44132 + }, + { + "epoch": 0.44133, + "grad_norm": 0.8312672194607077, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 44133 + }, + { + "epoch": 0.44134, + "grad_norm": 0.7905894961024322, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 44134 + }, + { + "epoch": 0.44135, + "grad_norm": 0.9653200752493984, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 44135 + }, + { + "epoch": 0.44136, + "grad_norm": 1.3201048540192912, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 44136 + }, + { + "epoch": 0.44137, + "grad_norm": 0.8865766324091965, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 44137 + }, + { + "epoch": 0.44138, + "grad_norm": 1.1039799062180804, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 44138 + }, + { + "epoch": 0.44139, + "grad_norm": 1.1246903472114913, + "learning_rate": 0.003, + "loss": 4.047, + "step": 44139 + }, + { + "epoch": 0.4414, + "grad_norm": 1.1819631329093783, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 44140 + }, + { + "epoch": 0.44141, + "grad_norm": 0.8936160241652402, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 44141 + }, + { + "epoch": 0.44142, + "grad_norm": 0.993285569944545, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 44142 + }, + { + "epoch": 0.44143, + "grad_norm": 1.0912531409034778, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 44143 + }, + { + "epoch": 0.44144, + "grad_norm": 1.2076563812817098, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 44144 + }, + { + "epoch": 0.44145, + "grad_norm": 0.9951458550288906, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 44145 + }, + { + "epoch": 0.44146, + "grad_norm": 1.308595863012127, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 44146 + }, + { + "epoch": 0.44147, + "grad_norm": 0.9483944202876005, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 44147 + }, + { + "epoch": 0.44148, + "grad_norm": 0.9253978733273824, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 44148 + }, + { + "epoch": 0.44149, + "grad_norm": 0.9970091194037666, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 44149 + }, + { + "epoch": 0.4415, + "grad_norm": 1.2056366425722358, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 44150 + }, + { + "epoch": 0.44151, + "grad_norm": 0.9380267676651546, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 44151 + }, + { + "epoch": 0.44152, + "grad_norm": 1.0634171688425549, + "learning_rate": 0.003, + "loss": 4.028, + "step": 44152 + }, + { + "epoch": 0.44153, + "grad_norm": 1.1170017518690678, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 44153 + }, + { + "epoch": 0.44154, + "grad_norm": 0.9757068181676621, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 44154 + }, + { + "epoch": 0.44155, + "grad_norm": 1.214187037636958, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 44155 + }, + { + "epoch": 0.44156, + "grad_norm": 0.8250351993215073, + "learning_rate": 0.003, + "loss": 4.028, + "step": 44156 + }, + { + "epoch": 0.44157, + "grad_norm": 0.8929850579949296, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 44157 + }, + { + "epoch": 0.44158, + "grad_norm": 1.232066902263953, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 44158 + }, + { + "epoch": 0.44159, + "grad_norm": 1.2275299262411155, + "learning_rate": 0.003, + "loss": 4.047, + "step": 44159 + }, + { + "epoch": 0.4416, + "grad_norm": 0.9684925602357473, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 44160 + }, + { + "epoch": 0.44161, + "grad_norm": 1.0881318621414435, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 44161 + }, + { + "epoch": 0.44162, + "grad_norm": 1.146111624930338, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 44162 + }, + { + "epoch": 0.44163, + "grad_norm": 0.8516563145246632, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 44163 + }, + { + "epoch": 0.44164, + "grad_norm": 0.7677312934404261, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 44164 + }, + { + "epoch": 0.44165, + "grad_norm": 0.8784046043941183, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 44165 + }, + { + "epoch": 0.44166, + "grad_norm": 1.0555054875776761, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 44166 + }, + { + "epoch": 0.44167, + "grad_norm": 1.1888851312427187, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 44167 + }, + { + "epoch": 0.44168, + "grad_norm": 0.9644921301536734, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 44168 + }, + { + "epoch": 0.44169, + "grad_norm": 1.1776278202023056, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 44169 + }, + { + "epoch": 0.4417, + "grad_norm": 1.13810930761508, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 44170 + }, + { + "epoch": 0.44171, + "grad_norm": 1.0754362889002496, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 44171 + }, + { + "epoch": 0.44172, + "grad_norm": 1.2764536600158707, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 44172 + }, + { + "epoch": 0.44173, + "grad_norm": 0.9613734993076755, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 44173 + }, + { + "epoch": 0.44174, + "grad_norm": 1.1483007755730446, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 44174 + }, + { + "epoch": 0.44175, + "grad_norm": 1.195989798613463, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 44175 + }, + { + "epoch": 0.44176, + "grad_norm": 0.8201147724352547, + "learning_rate": 0.003, + "loss": 4.017, + "step": 44176 + }, + { + "epoch": 0.44177, + "grad_norm": 0.7424545327251291, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 44177 + }, + { + "epoch": 0.44178, + "grad_norm": 0.8246450381441969, + "learning_rate": 0.003, + "loss": 3.997, + "step": 44178 + }, + { + "epoch": 0.44179, + "grad_norm": 1.0416997238496986, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 44179 + }, + { + "epoch": 0.4418, + "grad_norm": 1.28448897572881, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 44180 + }, + { + "epoch": 0.44181, + "grad_norm": 0.9283382562083118, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 44181 + }, + { + "epoch": 0.44182, + "grad_norm": 1.2144764981434313, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 44182 + }, + { + "epoch": 0.44183, + "grad_norm": 0.9857940583992173, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 44183 + }, + { + "epoch": 0.44184, + "grad_norm": 1.088937603351259, + "learning_rate": 0.003, + "loss": 4.034, + "step": 44184 + }, + { + "epoch": 0.44185, + "grad_norm": 1.2928387609180558, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 44185 + }, + { + "epoch": 0.44186, + "grad_norm": 0.9687360519924846, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 44186 + }, + { + "epoch": 0.44187, + "grad_norm": 1.1927013927919787, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 44187 + }, + { + "epoch": 0.44188, + "grad_norm": 1.0424054269912333, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 44188 + }, + { + "epoch": 0.44189, + "grad_norm": 1.0350483675943873, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 44189 + }, + { + "epoch": 0.4419, + "grad_norm": 1.1254336646756649, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 44190 + }, + { + "epoch": 0.44191, + "grad_norm": 0.9949777760673163, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 44191 + }, + { + "epoch": 0.44192, + "grad_norm": 1.0381462168992746, + "learning_rate": 0.003, + "loss": 4.033, + "step": 44192 + }, + { + "epoch": 0.44193, + "grad_norm": 1.1055555804952384, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 44193 + }, + { + "epoch": 0.44194, + "grad_norm": 1.081437785874723, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 44194 + }, + { + "epoch": 0.44195, + "grad_norm": 0.977516094820464, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 44195 + }, + { + "epoch": 0.44196, + "grad_norm": 1.1335560536126148, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 44196 + }, + { + "epoch": 0.44197, + "grad_norm": 0.9475782117200137, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 44197 + }, + { + "epoch": 0.44198, + "grad_norm": 1.2279751260349445, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 44198 + }, + { + "epoch": 0.44199, + "grad_norm": 1.0629649164991644, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 44199 + }, + { + "epoch": 0.442, + "grad_norm": 1.1019454235946389, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 44200 + }, + { + "epoch": 0.44201, + "grad_norm": 0.9440875175447574, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 44201 + }, + { + "epoch": 0.44202, + "grad_norm": 1.0420768642946316, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 44202 + }, + { + "epoch": 0.44203, + "grad_norm": 1.0437192886530555, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 44203 + }, + { + "epoch": 0.44204, + "grad_norm": 1.0127680293036248, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 44204 + }, + { + "epoch": 0.44205, + "grad_norm": 0.9944050743672039, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 44205 + }, + { + "epoch": 0.44206, + "grad_norm": 1.0272494491723334, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 44206 + }, + { + "epoch": 0.44207, + "grad_norm": 1.1330319106777738, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 44207 + }, + { + "epoch": 0.44208, + "grad_norm": 1.0303522516353893, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 44208 + }, + { + "epoch": 0.44209, + "grad_norm": 1.0212929005519245, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 44209 + }, + { + "epoch": 0.4421, + "grad_norm": 1.097092231860814, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 44210 + }, + { + "epoch": 0.44211, + "grad_norm": 1.0937894619453585, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 44211 + }, + { + "epoch": 0.44212, + "grad_norm": 1.0456703085827037, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 44212 + }, + { + "epoch": 0.44213, + "grad_norm": 1.100096014972956, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 44213 + }, + { + "epoch": 0.44214, + "grad_norm": 1.2214610836423607, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 44214 + }, + { + "epoch": 0.44215, + "grad_norm": 1.010978357082391, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 44215 + }, + { + "epoch": 0.44216, + "grad_norm": 1.270467526549627, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 44216 + }, + { + "epoch": 0.44217, + "grad_norm": 0.9060103161934165, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 44217 + }, + { + "epoch": 0.44218, + "grad_norm": 0.9887916399666049, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 44218 + }, + { + "epoch": 0.44219, + "grad_norm": 1.055161954114557, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 44219 + }, + { + "epoch": 0.4422, + "grad_norm": 0.9522978650297864, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 44220 + }, + { + "epoch": 0.44221, + "grad_norm": 1.021179785469154, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 44221 + }, + { + "epoch": 0.44222, + "grad_norm": 1.1221913591523411, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 44222 + }, + { + "epoch": 0.44223, + "grad_norm": 1.077607193974599, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 44223 + }, + { + "epoch": 0.44224, + "grad_norm": 1.2990793339003794, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 44224 + }, + { + "epoch": 0.44225, + "grad_norm": 1.1488924685720847, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 44225 + }, + { + "epoch": 0.44226, + "grad_norm": 1.050528690348902, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 44226 + }, + { + "epoch": 0.44227, + "grad_norm": 1.1273568572742638, + "learning_rate": 0.003, + "loss": 4.073, + "step": 44227 + }, + { + "epoch": 0.44228, + "grad_norm": 0.9439041097643873, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 44228 + }, + { + "epoch": 0.44229, + "grad_norm": 1.1409139974840472, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 44229 + }, + { + "epoch": 0.4423, + "grad_norm": 1.1130136767774919, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 44230 + }, + { + "epoch": 0.44231, + "grad_norm": 1.2146409558717373, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 44231 + }, + { + "epoch": 0.44232, + "grad_norm": 0.8628212868271508, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 44232 + }, + { + "epoch": 0.44233, + "grad_norm": 0.7643478832397944, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 44233 + }, + { + "epoch": 0.44234, + "grad_norm": 0.8562931446727308, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 44234 + }, + { + "epoch": 0.44235, + "grad_norm": 1.0918074373968434, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 44235 + }, + { + "epoch": 0.44236, + "grad_norm": 1.3501231832995553, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 44236 + }, + { + "epoch": 0.44237, + "grad_norm": 0.836028452718404, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 44237 + }, + { + "epoch": 0.44238, + "grad_norm": 1.0649920311950836, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 44238 + }, + { + "epoch": 0.44239, + "grad_norm": 1.167304329470371, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 44239 + }, + { + "epoch": 0.4424, + "grad_norm": 0.9281943132630318, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 44240 + }, + { + "epoch": 0.44241, + "grad_norm": 0.9128888998561623, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 44241 + }, + { + "epoch": 0.44242, + "grad_norm": 0.9899170097077707, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 44242 + }, + { + "epoch": 0.44243, + "grad_norm": 1.4167741755009613, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 44243 + }, + { + "epoch": 0.44244, + "grad_norm": 0.845000097899636, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 44244 + }, + { + "epoch": 0.44245, + "grad_norm": 1.0001892787212663, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 44245 + }, + { + "epoch": 0.44246, + "grad_norm": 1.1927169413552972, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 44246 + }, + { + "epoch": 0.44247, + "grad_norm": 0.9227562154487856, + "learning_rate": 0.003, + "loss": 4.033, + "step": 44247 + }, + { + "epoch": 0.44248, + "grad_norm": 1.17616451052051, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 44248 + }, + { + "epoch": 0.44249, + "grad_norm": 0.9161475735548187, + "learning_rate": 0.003, + "loss": 4.036, + "step": 44249 + }, + { + "epoch": 0.4425, + "grad_norm": 0.9925700122159321, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 44250 + }, + { + "epoch": 0.44251, + "grad_norm": 1.1541827992039406, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 44251 + }, + { + "epoch": 0.44252, + "grad_norm": 0.7229199966484298, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 44252 + }, + { + "epoch": 0.44253, + "grad_norm": 0.7358149748333895, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 44253 + }, + { + "epoch": 0.44254, + "grad_norm": 0.8556751067765218, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 44254 + }, + { + "epoch": 0.44255, + "grad_norm": 1.3665849645961072, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 44255 + }, + { + "epoch": 0.44256, + "grad_norm": 1.1275884810202832, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 44256 + }, + { + "epoch": 0.44257, + "grad_norm": 1.1536643201190042, + "learning_rate": 0.003, + "loss": 4.043, + "step": 44257 + }, + { + "epoch": 0.44258, + "grad_norm": 1.1400485112244407, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 44258 + }, + { + "epoch": 0.44259, + "grad_norm": 1.266492051920303, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 44259 + }, + { + "epoch": 0.4426, + "grad_norm": 0.8171208293062652, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 44260 + }, + { + "epoch": 0.44261, + "grad_norm": 0.7464640825262666, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 44261 + }, + { + "epoch": 0.44262, + "grad_norm": 0.985626991953503, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 44262 + }, + { + "epoch": 0.44263, + "grad_norm": 1.263927203516455, + "learning_rate": 0.003, + "loss": 4.016, + "step": 44263 + }, + { + "epoch": 0.44264, + "grad_norm": 0.8593766376728097, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 44264 + }, + { + "epoch": 0.44265, + "grad_norm": 0.948009511823534, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 44265 + }, + { + "epoch": 0.44266, + "grad_norm": 1.2708834416503254, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 44266 + }, + { + "epoch": 0.44267, + "grad_norm": 1.1253531052938837, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 44267 + }, + { + "epoch": 0.44268, + "grad_norm": 0.9350388434785066, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 44268 + }, + { + "epoch": 0.44269, + "grad_norm": 0.9545399569678856, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 44269 + }, + { + "epoch": 0.4427, + "grad_norm": 1.1479767396970433, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 44270 + }, + { + "epoch": 0.44271, + "grad_norm": 1.1445742230754903, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 44271 + }, + { + "epoch": 0.44272, + "grad_norm": 1.2083829090803277, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 44272 + }, + { + "epoch": 0.44273, + "grad_norm": 0.8449370680748348, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 44273 + }, + { + "epoch": 0.44274, + "grad_norm": 0.805432399516268, + "learning_rate": 0.003, + "loss": 4.033, + "step": 44274 + }, + { + "epoch": 0.44275, + "grad_norm": 1.024806120187861, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 44275 + }, + { + "epoch": 0.44276, + "grad_norm": 1.2347625303361174, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 44276 + }, + { + "epoch": 0.44277, + "grad_norm": 1.194114143193584, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 44277 + }, + { + "epoch": 0.44278, + "grad_norm": 0.9344057672823656, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 44278 + }, + { + "epoch": 0.44279, + "grad_norm": 1.1342759855110065, + "learning_rate": 0.003, + "loss": 3.997, + "step": 44279 + }, + { + "epoch": 0.4428, + "grad_norm": 1.3907160722259628, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 44280 + }, + { + "epoch": 0.44281, + "grad_norm": 0.9485697488290997, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 44281 + }, + { + "epoch": 0.44282, + "grad_norm": 0.8383888939910824, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 44282 + }, + { + "epoch": 0.44283, + "grad_norm": 0.9556791761852325, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 44283 + }, + { + "epoch": 0.44284, + "grad_norm": 1.1769071928072972, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 44284 + }, + { + "epoch": 0.44285, + "grad_norm": 1.079819440206096, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 44285 + }, + { + "epoch": 0.44286, + "grad_norm": 0.9800230141883098, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 44286 + }, + { + "epoch": 0.44287, + "grad_norm": 1.301844600294766, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 44287 + }, + { + "epoch": 0.44288, + "grad_norm": 1.0132606829707123, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 44288 + }, + { + "epoch": 0.44289, + "grad_norm": 1.0685318496721647, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 44289 + }, + { + "epoch": 0.4429, + "grad_norm": 1.0556223777862843, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 44290 + }, + { + "epoch": 0.44291, + "grad_norm": 1.008311491634872, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 44291 + }, + { + "epoch": 0.44292, + "grad_norm": 1.461636006908275, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 44292 + }, + { + "epoch": 0.44293, + "grad_norm": 0.8778478237220741, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 44293 + }, + { + "epoch": 0.44294, + "grad_norm": 1.0962685597003412, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 44294 + }, + { + "epoch": 0.44295, + "grad_norm": 1.2385792275038405, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 44295 + }, + { + "epoch": 0.44296, + "grad_norm": 0.9715565725409163, + "learning_rate": 0.003, + "loss": 4.029, + "step": 44296 + }, + { + "epoch": 0.44297, + "grad_norm": 1.162706435491631, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 44297 + }, + { + "epoch": 0.44298, + "grad_norm": 1.0520516654922598, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 44298 + }, + { + "epoch": 0.44299, + "grad_norm": 1.1906589911361334, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 44299 + }, + { + "epoch": 0.443, + "grad_norm": 1.081824004062726, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 44300 + }, + { + "epoch": 0.44301, + "grad_norm": 0.9632266785894847, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 44301 + }, + { + "epoch": 0.44302, + "grad_norm": 1.3188182513600606, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 44302 + }, + { + "epoch": 0.44303, + "grad_norm": 0.9772337560370783, + "learning_rate": 0.003, + "loss": 4.048, + "step": 44303 + }, + { + "epoch": 0.44304, + "grad_norm": 1.217098647411705, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 44304 + }, + { + "epoch": 0.44305, + "grad_norm": 0.9508571379931616, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 44305 + }, + { + "epoch": 0.44306, + "grad_norm": 0.957092803662866, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 44306 + }, + { + "epoch": 0.44307, + "grad_norm": 0.9415824378713653, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 44307 + }, + { + "epoch": 0.44308, + "grad_norm": 1.0709564166738468, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 44308 + }, + { + "epoch": 0.44309, + "grad_norm": 1.142902074234839, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 44309 + }, + { + "epoch": 0.4431, + "grad_norm": 1.242669491828082, + "learning_rate": 0.003, + "loss": 4.042, + "step": 44310 + }, + { + "epoch": 0.44311, + "grad_norm": 0.9450590674286824, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 44311 + }, + { + "epoch": 0.44312, + "grad_norm": 0.9385166919164736, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 44312 + }, + { + "epoch": 0.44313, + "grad_norm": 1.3295281395079097, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 44313 + }, + { + "epoch": 0.44314, + "grad_norm": 0.8333554082701715, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 44314 + }, + { + "epoch": 0.44315, + "grad_norm": 0.8330883823658206, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 44315 + }, + { + "epoch": 0.44316, + "grad_norm": 0.9858948391653983, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 44316 + }, + { + "epoch": 0.44317, + "grad_norm": 1.2810488198384526, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 44317 + }, + { + "epoch": 0.44318, + "grad_norm": 1.0305884232289093, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 44318 + }, + { + "epoch": 0.44319, + "grad_norm": 1.2387548827340342, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 44319 + }, + { + "epoch": 0.4432, + "grad_norm": 0.9627208233631522, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 44320 + }, + { + "epoch": 0.44321, + "grad_norm": 1.0750101529974276, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 44321 + }, + { + "epoch": 0.44322, + "grad_norm": 1.2790806452645505, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 44322 + }, + { + "epoch": 0.44323, + "grad_norm": 1.0551505947925979, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 44323 + }, + { + "epoch": 0.44324, + "grad_norm": 0.9248552459579337, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 44324 + }, + { + "epoch": 0.44325, + "grad_norm": 0.9409486379627912, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 44325 + }, + { + "epoch": 0.44326, + "grad_norm": 1.033248409494907, + "learning_rate": 0.003, + "loss": 4.016, + "step": 44326 + }, + { + "epoch": 0.44327, + "grad_norm": 1.0180172235299212, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 44327 + }, + { + "epoch": 0.44328, + "grad_norm": 0.9553476975743466, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 44328 + }, + { + "epoch": 0.44329, + "grad_norm": 1.0300305967223966, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 44329 + }, + { + "epoch": 0.4433, + "grad_norm": 1.1570646914851221, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 44330 + }, + { + "epoch": 0.44331, + "grad_norm": 1.1520692443312177, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 44331 + }, + { + "epoch": 0.44332, + "grad_norm": 0.8978382820674081, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 44332 + }, + { + "epoch": 0.44333, + "grad_norm": 1.0295972539623715, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 44333 + }, + { + "epoch": 0.44334, + "grad_norm": 1.3051121728701613, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 44334 + }, + { + "epoch": 0.44335, + "grad_norm": 0.8877913740828329, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 44335 + }, + { + "epoch": 0.44336, + "grad_norm": 1.0687940312766844, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 44336 + }, + { + "epoch": 0.44337, + "grad_norm": 1.1558176574470325, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 44337 + }, + { + "epoch": 0.44338, + "grad_norm": 1.1407674994853383, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 44338 + }, + { + "epoch": 0.44339, + "grad_norm": 1.1247297834848842, + "learning_rate": 0.003, + "loss": 4.004, + "step": 44339 + }, + { + "epoch": 0.4434, + "grad_norm": 1.0084112211686704, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 44340 + }, + { + "epoch": 0.44341, + "grad_norm": 1.1285887137559674, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 44341 + }, + { + "epoch": 0.44342, + "grad_norm": 1.0495326470589894, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 44342 + }, + { + "epoch": 0.44343, + "grad_norm": 1.1663436704689716, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 44343 + }, + { + "epoch": 0.44344, + "grad_norm": 1.0021740607187275, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 44344 + }, + { + "epoch": 0.44345, + "grad_norm": 1.0872236655459426, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 44345 + }, + { + "epoch": 0.44346, + "grad_norm": 0.9371199078697292, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 44346 + }, + { + "epoch": 0.44347, + "grad_norm": 0.9918450616366655, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 44347 + }, + { + "epoch": 0.44348, + "grad_norm": 1.2656276560320814, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 44348 + }, + { + "epoch": 0.44349, + "grad_norm": 1.0378981831850236, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 44349 + }, + { + "epoch": 0.4435, + "grad_norm": 1.5251521898562204, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 44350 + }, + { + "epoch": 0.44351, + "grad_norm": 0.8777376266074547, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 44351 + }, + { + "epoch": 0.44352, + "grad_norm": 0.9317046913687898, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 44352 + }, + { + "epoch": 0.44353, + "grad_norm": 1.0813118979010876, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 44353 + }, + { + "epoch": 0.44354, + "grad_norm": 1.1996394169398028, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 44354 + }, + { + "epoch": 0.44355, + "grad_norm": 1.286892991331832, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 44355 + }, + { + "epoch": 0.44356, + "grad_norm": 1.1860079015698266, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 44356 + }, + { + "epoch": 0.44357, + "grad_norm": 1.0954695465113755, + "learning_rate": 0.003, + "loss": 4.053, + "step": 44357 + }, + { + "epoch": 0.44358, + "grad_norm": 1.0939373039032239, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 44358 + }, + { + "epoch": 0.44359, + "grad_norm": 0.9682510084643269, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 44359 + }, + { + "epoch": 0.4436, + "grad_norm": 0.8730320009444199, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 44360 + }, + { + "epoch": 0.44361, + "grad_norm": 0.8163059688765629, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 44361 + }, + { + "epoch": 0.44362, + "grad_norm": 0.9327907174108221, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 44362 + }, + { + "epoch": 0.44363, + "grad_norm": 1.250921165659358, + "learning_rate": 0.003, + "loss": 4.041, + "step": 44363 + }, + { + "epoch": 0.44364, + "grad_norm": 1.0029531414025965, + "learning_rate": 0.003, + "loss": 4.047, + "step": 44364 + }, + { + "epoch": 0.44365, + "grad_norm": 1.0257601901131643, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 44365 + }, + { + "epoch": 0.44366, + "grad_norm": 1.2469635669332817, + "learning_rate": 0.003, + "loss": 4.104, + "step": 44366 + }, + { + "epoch": 0.44367, + "grad_norm": 1.0984416048396966, + "learning_rate": 0.003, + "loss": 4.056, + "step": 44367 + }, + { + "epoch": 0.44368, + "grad_norm": 1.2904117571854892, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 44368 + }, + { + "epoch": 0.44369, + "grad_norm": 0.9594463340395802, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 44369 + }, + { + "epoch": 0.4437, + "grad_norm": 1.006407292604191, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 44370 + }, + { + "epoch": 0.44371, + "grad_norm": 1.198695185191512, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 44371 + }, + { + "epoch": 0.44372, + "grad_norm": 1.1290516724935404, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 44372 + }, + { + "epoch": 0.44373, + "grad_norm": 0.7699255194223649, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 44373 + }, + { + "epoch": 0.44374, + "grad_norm": 0.7745972485812583, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 44374 + }, + { + "epoch": 0.44375, + "grad_norm": 0.9238266285366988, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 44375 + }, + { + "epoch": 0.44376, + "grad_norm": 1.2834717716150343, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 44376 + }, + { + "epoch": 0.44377, + "grad_norm": 0.9629458855270493, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 44377 + }, + { + "epoch": 0.44378, + "grad_norm": 0.8439240659797884, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 44378 + }, + { + "epoch": 0.44379, + "grad_norm": 0.8080480723667662, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 44379 + }, + { + "epoch": 0.4438, + "grad_norm": 0.8473425894841069, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 44380 + }, + { + "epoch": 0.44381, + "grad_norm": 1.0271934918113776, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 44381 + }, + { + "epoch": 0.44382, + "grad_norm": 1.1252711937661313, + "learning_rate": 0.003, + "loss": 4.022, + "step": 44382 + }, + { + "epoch": 0.44383, + "grad_norm": 1.0010000865496493, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 44383 + }, + { + "epoch": 0.44384, + "grad_norm": 1.227071329721213, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 44384 + }, + { + "epoch": 0.44385, + "grad_norm": 1.0284091517736613, + "learning_rate": 0.003, + "loss": 4.022, + "step": 44385 + }, + { + "epoch": 0.44386, + "grad_norm": 1.1069541083985501, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 44386 + }, + { + "epoch": 0.44387, + "grad_norm": 1.0857296325610981, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 44387 + }, + { + "epoch": 0.44388, + "grad_norm": 0.8949253714508959, + "learning_rate": 0.003, + "loss": 4.033, + "step": 44388 + }, + { + "epoch": 0.44389, + "grad_norm": 0.8507418892798485, + "learning_rate": 0.003, + "loss": 4.03, + "step": 44389 + }, + { + "epoch": 0.4439, + "grad_norm": 0.7846085586996612, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 44390 + }, + { + "epoch": 0.44391, + "grad_norm": 0.7304753294913608, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 44391 + }, + { + "epoch": 0.44392, + "grad_norm": 0.7481812955236256, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 44392 + }, + { + "epoch": 0.44393, + "grad_norm": 0.8080277683301773, + "learning_rate": 0.003, + "loss": 4.035, + "step": 44393 + }, + { + "epoch": 0.44394, + "grad_norm": 0.9464185219405705, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 44394 + }, + { + "epoch": 0.44395, + "grad_norm": 1.1276259191505655, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 44395 + }, + { + "epoch": 0.44396, + "grad_norm": 1.0402040693044978, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 44396 + }, + { + "epoch": 0.44397, + "grad_norm": 1.0899366632466954, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 44397 + }, + { + "epoch": 0.44398, + "grad_norm": 0.9925383190466217, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 44398 + }, + { + "epoch": 0.44399, + "grad_norm": 1.2985439058920325, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 44399 + }, + { + "epoch": 0.444, + "grad_norm": 1.210849235087844, + "learning_rate": 0.003, + "loss": 4.021, + "step": 44400 + }, + { + "epoch": 0.44401, + "grad_norm": 1.3171496182970042, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 44401 + }, + { + "epoch": 0.44402, + "grad_norm": 0.8766937467583926, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 44402 + }, + { + "epoch": 0.44403, + "grad_norm": 0.9834967865626391, + "learning_rate": 0.003, + "loss": 4.039, + "step": 44403 + }, + { + "epoch": 0.44404, + "grad_norm": 1.4442265805169638, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 44404 + }, + { + "epoch": 0.44405, + "grad_norm": 1.0337432649793998, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 44405 + }, + { + "epoch": 0.44406, + "grad_norm": 1.3390797400855023, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 44406 + }, + { + "epoch": 0.44407, + "grad_norm": 1.0682770642580834, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 44407 + }, + { + "epoch": 0.44408, + "grad_norm": 1.0738740381395235, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 44408 + }, + { + "epoch": 0.44409, + "grad_norm": 1.0391389715401784, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 44409 + }, + { + "epoch": 0.4441, + "grad_norm": 1.1199942582574571, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 44410 + }, + { + "epoch": 0.44411, + "grad_norm": 0.930261287247159, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 44411 + }, + { + "epoch": 0.44412, + "grad_norm": 0.9327610304642652, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 44412 + }, + { + "epoch": 0.44413, + "grad_norm": 1.0061422023415734, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 44413 + }, + { + "epoch": 0.44414, + "grad_norm": 1.1109658412221581, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 44414 + }, + { + "epoch": 0.44415, + "grad_norm": 1.0841757893335875, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 44415 + }, + { + "epoch": 0.44416, + "grad_norm": 1.0109898306255451, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 44416 + }, + { + "epoch": 0.44417, + "grad_norm": 0.9373451963860793, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 44417 + }, + { + "epoch": 0.44418, + "grad_norm": 1.0908896810011623, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 44418 + }, + { + "epoch": 0.44419, + "grad_norm": 0.9849640824588648, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 44419 + }, + { + "epoch": 0.4442, + "grad_norm": 0.9788230345796035, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 44420 + }, + { + "epoch": 0.44421, + "grad_norm": 1.0685617489660908, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 44421 + }, + { + "epoch": 0.44422, + "grad_norm": 1.181621179181294, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 44422 + }, + { + "epoch": 0.44423, + "grad_norm": 1.129093422044906, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 44423 + }, + { + "epoch": 0.44424, + "grad_norm": 1.516089461614981, + "learning_rate": 0.003, + "loss": 4.028, + "step": 44424 + }, + { + "epoch": 0.44425, + "grad_norm": 0.6840242482691588, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 44425 + }, + { + "epoch": 0.44426, + "grad_norm": 0.7145600714574281, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 44426 + }, + { + "epoch": 0.44427, + "grad_norm": 0.8746594753108143, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 44427 + }, + { + "epoch": 0.44428, + "grad_norm": 0.9277385951798863, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 44428 + }, + { + "epoch": 0.44429, + "grad_norm": 1.069193295986879, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 44429 + }, + { + "epoch": 0.4443, + "grad_norm": 1.4359713365112419, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 44430 + }, + { + "epoch": 0.44431, + "grad_norm": 0.9406630973149377, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 44431 + }, + { + "epoch": 0.44432, + "grad_norm": 0.8767319833077137, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 44432 + }, + { + "epoch": 0.44433, + "grad_norm": 0.8675667303538179, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 44433 + }, + { + "epoch": 0.44434, + "grad_norm": 1.0852603923055923, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 44434 + }, + { + "epoch": 0.44435, + "grad_norm": 1.178773191003315, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 44435 + }, + { + "epoch": 0.44436, + "grad_norm": 1.1059435661578172, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 44436 + }, + { + "epoch": 0.44437, + "grad_norm": 1.036028563261645, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 44437 + }, + { + "epoch": 0.44438, + "grad_norm": 1.1129069195251575, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 44438 + }, + { + "epoch": 0.44439, + "grad_norm": 0.9606140604463393, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 44439 + }, + { + "epoch": 0.4444, + "grad_norm": 1.0273410060484018, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 44440 + }, + { + "epoch": 0.44441, + "grad_norm": 0.9945576635765081, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 44441 + }, + { + "epoch": 0.44442, + "grad_norm": 0.9250736840362564, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 44442 + }, + { + "epoch": 0.44443, + "grad_norm": 1.000548980678877, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 44443 + }, + { + "epoch": 0.44444, + "grad_norm": 1.1434909773745525, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 44444 + }, + { + "epoch": 0.44445, + "grad_norm": 1.030659618116655, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 44445 + }, + { + "epoch": 0.44446, + "grad_norm": 1.1603016355517595, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 44446 + }, + { + "epoch": 0.44447, + "grad_norm": 0.9555143152537112, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 44447 + }, + { + "epoch": 0.44448, + "grad_norm": 1.16534170176119, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 44448 + }, + { + "epoch": 0.44449, + "grad_norm": 1.122923092055082, + "learning_rate": 0.003, + "loss": 4.033, + "step": 44449 + }, + { + "epoch": 0.4445, + "grad_norm": 1.1268202078146474, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 44450 + }, + { + "epoch": 0.44451, + "grad_norm": 0.9784357005217676, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 44451 + }, + { + "epoch": 0.44452, + "grad_norm": 1.2884456522487024, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 44452 + }, + { + "epoch": 0.44453, + "grad_norm": 1.0990384885865012, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 44453 + }, + { + "epoch": 0.44454, + "grad_norm": 1.295223207114721, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 44454 + }, + { + "epoch": 0.44455, + "grad_norm": 1.1328390632163423, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 44455 + }, + { + "epoch": 0.44456, + "grad_norm": 0.8603841085247557, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 44456 + }, + { + "epoch": 0.44457, + "grad_norm": 0.908384955667415, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 44457 + }, + { + "epoch": 0.44458, + "grad_norm": 0.995764216261399, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 44458 + }, + { + "epoch": 0.44459, + "grad_norm": 1.2543942066462441, + "learning_rate": 0.003, + "loss": 4.008, + "step": 44459 + }, + { + "epoch": 0.4446, + "grad_norm": 0.9643318145345913, + "learning_rate": 0.003, + "loss": 4.004, + "step": 44460 + }, + { + "epoch": 0.44461, + "grad_norm": 1.1705285267833516, + "learning_rate": 0.003, + "loss": 4.018, + "step": 44461 + }, + { + "epoch": 0.44462, + "grad_norm": 0.9725621362407529, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 44462 + }, + { + "epoch": 0.44463, + "grad_norm": 0.9461032814117669, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 44463 + }, + { + "epoch": 0.44464, + "grad_norm": 1.1377865897139283, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 44464 + }, + { + "epoch": 0.44465, + "grad_norm": 0.9706523083453203, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 44465 + }, + { + "epoch": 0.44466, + "grad_norm": 0.9961343664617244, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 44466 + }, + { + "epoch": 0.44467, + "grad_norm": 1.2342824016215952, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 44467 + }, + { + "epoch": 0.44468, + "grad_norm": 1.0968410045597623, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 44468 + }, + { + "epoch": 0.44469, + "grad_norm": 1.1912510610542757, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 44469 + }, + { + "epoch": 0.4447, + "grad_norm": 1.253776782332147, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 44470 + }, + { + "epoch": 0.44471, + "grad_norm": 1.0919181597722833, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 44471 + }, + { + "epoch": 0.44472, + "grad_norm": 0.995480775426333, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 44472 + }, + { + "epoch": 0.44473, + "grad_norm": 1.1866028898645529, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 44473 + }, + { + "epoch": 0.44474, + "grad_norm": 1.1380236125808885, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 44474 + }, + { + "epoch": 0.44475, + "grad_norm": 0.8886364434765797, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 44475 + }, + { + "epoch": 0.44476, + "grad_norm": 0.8163843588651846, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 44476 + }, + { + "epoch": 0.44477, + "grad_norm": 0.8522717952957916, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 44477 + }, + { + "epoch": 0.44478, + "grad_norm": 1.1294567936957451, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 44478 + }, + { + "epoch": 0.44479, + "grad_norm": 1.3111295213563898, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 44479 + }, + { + "epoch": 0.4448, + "grad_norm": 0.9854587586977861, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 44480 + }, + { + "epoch": 0.44481, + "grad_norm": 1.0804009016882825, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 44481 + }, + { + "epoch": 0.44482, + "grad_norm": 1.250913450663011, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 44482 + }, + { + "epoch": 0.44483, + "grad_norm": 1.1918680047274561, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 44483 + }, + { + "epoch": 0.44484, + "grad_norm": 0.8547198065344793, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 44484 + }, + { + "epoch": 0.44485, + "grad_norm": 0.7518296092952754, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 44485 + }, + { + "epoch": 0.44486, + "grad_norm": 0.8451043085499431, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 44486 + }, + { + "epoch": 0.44487, + "grad_norm": 0.9812707129959456, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 44487 + }, + { + "epoch": 0.44488, + "grad_norm": 1.221750848152882, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 44488 + }, + { + "epoch": 0.44489, + "grad_norm": 1.1633501828104975, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 44489 + }, + { + "epoch": 0.4449, + "grad_norm": 0.9441722101822637, + "learning_rate": 0.003, + "loss": 4.032, + "step": 44490 + }, + { + "epoch": 0.44491, + "grad_norm": 1.174222778814598, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 44491 + }, + { + "epoch": 0.44492, + "grad_norm": 1.1321602567873692, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 44492 + }, + { + "epoch": 0.44493, + "grad_norm": 0.9155295251827802, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 44493 + }, + { + "epoch": 0.44494, + "grad_norm": 1.0348017331626893, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 44494 + }, + { + "epoch": 0.44495, + "grad_norm": 1.180000340721708, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 44495 + }, + { + "epoch": 0.44496, + "grad_norm": 0.9197071933447679, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 44496 + }, + { + "epoch": 0.44497, + "grad_norm": 0.879444788811553, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 44497 + }, + { + "epoch": 0.44498, + "grad_norm": 1.0540865010947431, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 44498 + }, + { + "epoch": 0.44499, + "grad_norm": 1.18965077047988, + "learning_rate": 0.003, + "loss": 4.0, + "step": 44499 + }, + { + "epoch": 0.445, + "grad_norm": 1.2388003704866715, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 44500 + }, + { + "epoch": 0.44501, + "grad_norm": 0.8890681580515073, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 44501 + }, + { + "epoch": 0.44502, + "grad_norm": 0.8284431690517662, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 44502 + }, + { + "epoch": 0.44503, + "grad_norm": 0.7959034972863184, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 44503 + }, + { + "epoch": 0.44504, + "grad_norm": 0.9313730155389874, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 44504 + }, + { + "epoch": 0.44505, + "grad_norm": 0.9573499235983745, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 44505 + }, + { + "epoch": 0.44506, + "grad_norm": 1.0949479921711407, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 44506 + }, + { + "epoch": 0.44507, + "grad_norm": 1.0622823163506545, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 44507 + }, + { + "epoch": 0.44508, + "grad_norm": 1.2474989479109901, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 44508 + }, + { + "epoch": 0.44509, + "grad_norm": 1.0820447991615292, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 44509 + }, + { + "epoch": 0.4451, + "grad_norm": 1.0630587983373776, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 44510 + }, + { + "epoch": 0.44511, + "grad_norm": 1.1618082777128476, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 44511 + }, + { + "epoch": 0.44512, + "grad_norm": 1.0873589727357782, + "learning_rate": 0.003, + "loss": 4.029, + "step": 44512 + }, + { + "epoch": 0.44513, + "grad_norm": 0.9438681537702255, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 44513 + }, + { + "epoch": 0.44514, + "grad_norm": 1.084410101050218, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 44514 + }, + { + "epoch": 0.44515, + "grad_norm": 1.3840917672576567, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 44515 + }, + { + "epoch": 0.44516, + "grad_norm": 0.9821547086185602, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 44516 + }, + { + "epoch": 0.44517, + "grad_norm": 1.0143154510212011, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 44517 + }, + { + "epoch": 0.44518, + "grad_norm": 1.0070056053137744, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 44518 + }, + { + "epoch": 0.44519, + "grad_norm": 1.0282809754561242, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 44519 + }, + { + "epoch": 0.4452, + "grad_norm": 1.1347958268607645, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 44520 + }, + { + "epoch": 0.44521, + "grad_norm": 1.2187261876094424, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 44521 + }, + { + "epoch": 0.44522, + "grad_norm": 1.1376375559032186, + "learning_rate": 0.003, + "loss": 4.032, + "step": 44522 + }, + { + "epoch": 0.44523, + "grad_norm": 1.0025885095820688, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 44523 + }, + { + "epoch": 0.44524, + "grad_norm": 1.2560817383898322, + "learning_rate": 0.003, + "loss": 4.028, + "step": 44524 + }, + { + "epoch": 0.44525, + "grad_norm": 0.9356450880711115, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 44525 + }, + { + "epoch": 0.44526, + "grad_norm": 1.1018172369828476, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 44526 + }, + { + "epoch": 0.44527, + "grad_norm": 1.2084846828233304, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 44527 + }, + { + "epoch": 0.44528, + "grad_norm": 1.1639888270783494, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 44528 + }, + { + "epoch": 0.44529, + "grad_norm": 0.9010840479074372, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 44529 + }, + { + "epoch": 0.4453, + "grad_norm": 0.9806376140274964, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 44530 + }, + { + "epoch": 0.44531, + "grad_norm": 1.1195444231082259, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 44531 + }, + { + "epoch": 0.44532, + "grad_norm": 1.0549940104093898, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 44532 + }, + { + "epoch": 0.44533, + "grad_norm": 0.8561375993773652, + "learning_rate": 0.003, + "loss": 4.031, + "step": 44533 + }, + { + "epoch": 0.44534, + "grad_norm": 1.0010191274218752, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 44534 + }, + { + "epoch": 0.44535, + "grad_norm": 1.1720170943049097, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 44535 + }, + { + "epoch": 0.44536, + "grad_norm": 1.0909753369400006, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 44536 + }, + { + "epoch": 0.44537, + "grad_norm": 1.2140355619777523, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 44537 + }, + { + "epoch": 0.44538, + "grad_norm": 1.0302510987449374, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 44538 + }, + { + "epoch": 0.44539, + "grad_norm": 1.1596278393600505, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 44539 + }, + { + "epoch": 0.4454, + "grad_norm": 1.017065900605249, + "learning_rate": 0.003, + "loss": 4.023, + "step": 44540 + }, + { + "epoch": 0.44541, + "grad_norm": 1.1533293981443082, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 44541 + }, + { + "epoch": 0.44542, + "grad_norm": 1.3692942031387096, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 44542 + }, + { + "epoch": 0.44543, + "grad_norm": 1.0974210243489353, + "learning_rate": 0.003, + "loss": 4.039, + "step": 44543 + }, + { + "epoch": 0.44544, + "grad_norm": 1.1783645090175892, + "learning_rate": 0.003, + "loss": 4.048, + "step": 44544 + }, + { + "epoch": 0.44545, + "grad_norm": 0.9560914084002182, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 44545 + }, + { + "epoch": 0.44546, + "grad_norm": 1.3042127251213074, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 44546 + }, + { + "epoch": 0.44547, + "grad_norm": 0.9735465712378856, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 44547 + }, + { + "epoch": 0.44548, + "grad_norm": 1.019190344566993, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 44548 + }, + { + "epoch": 0.44549, + "grad_norm": 1.1550129954039927, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 44549 + }, + { + "epoch": 0.4455, + "grad_norm": 1.2394838573597993, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 44550 + }, + { + "epoch": 0.44551, + "grad_norm": 1.1486079550532255, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 44551 + }, + { + "epoch": 0.44552, + "grad_norm": 0.8232451476152736, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 44552 + }, + { + "epoch": 0.44553, + "grad_norm": 0.7490280309069804, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 44553 + }, + { + "epoch": 0.44554, + "grad_norm": 0.850172564858055, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 44554 + }, + { + "epoch": 0.44555, + "grad_norm": 1.0936789403948188, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 44555 + }, + { + "epoch": 0.44556, + "grad_norm": 1.155424472975886, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 44556 + }, + { + "epoch": 0.44557, + "grad_norm": 1.2091233957607754, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 44557 + }, + { + "epoch": 0.44558, + "grad_norm": 0.879166794928107, + "learning_rate": 0.003, + "loss": 4.0, + "step": 44558 + }, + { + "epoch": 0.44559, + "grad_norm": 0.7295491944533693, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 44559 + }, + { + "epoch": 0.4456, + "grad_norm": 0.7009854289179598, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 44560 + }, + { + "epoch": 0.44561, + "grad_norm": 0.8435153126265875, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 44561 + }, + { + "epoch": 0.44562, + "grad_norm": 0.9356545346644082, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 44562 + }, + { + "epoch": 0.44563, + "grad_norm": 1.1200306282941872, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 44563 + }, + { + "epoch": 0.44564, + "grad_norm": 1.341685710264536, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 44564 + }, + { + "epoch": 0.44565, + "grad_norm": 0.9660584695014808, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 44565 + }, + { + "epoch": 0.44566, + "grad_norm": 1.2018703019043229, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 44566 + }, + { + "epoch": 0.44567, + "grad_norm": 1.1591405657296794, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 44567 + }, + { + "epoch": 0.44568, + "grad_norm": 0.6826852930386835, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 44568 + }, + { + "epoch": 0.44569, + "grad_norm": 0.5881810380331017, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 44569 + }, + { + "epoch": 0.4457, + "grad_norm": 0.6353759265749633, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 44570 + }, + { + "epoch": 0.44571, + "grad_norm": 0.8920721176052533, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 44571 + }, + { + "epoch": 0.44572, + "grad_norm": 1.4321711834900106, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 44572 + }, + { + "epoch": 0.44573, + "grad_norm": 0.7778918580461341, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 44573 + }, + { + "epoch": 0.44574, + "grad_norm": 0.8167630340873371, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 44574 + }, + { + "epoch": 0.44575, + "grad_norm": 0.9474190354059007, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 44575 + }, + { + "epoch": 0.44576, + "grad_norm": 1.2729078852658429, + "learning_rate": 0.003, + "loss": 4.042, + "step": 44576 + }, + { + "epoch": 0.44577, + "grad_norm": 0.8419562478443806, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 44577 + }, + { + "epoch": 0.44578, + "grad_norm": 1.0017276956556376, + "learning_rate": 0.003, + "loss": 4.069, + "step": 44578 + }, + { + "epoch": 0.44579, + "grad_norm": 1.1513290854459017, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 44579 + }, + { + "epoch": 0.4458, + "grad_norm": 1.3955636003203258, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 44580 + }, + { + "epoch": 0.44581, + "grad_norm": 0.9429282237639527, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 44581 + }, + { + "epoch": 0.44582, + "grad_norm": 1.3380326559500646, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 44582 + }, + { + "epoch": 0.44583, + "grad_norm": 1.0367712802389197, + "learning_rate": 0.003, + "loss": 4.026, + "step": 44583 + }, + { + "epoch": 0.44584, + "grad_norm": 1.0581587966595563, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 44584 + }, + { + "epoch": 0.44585, + "grad_norm": 1.0957728671215854, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 44585 + }, + { + "epoch": 0.44586, + "grad_norm": 1.5452514999484175, + "learning_rate": 0.003, + "loss": 4.052, + "step": 44586 + }, + { + "epoch": 0.44587, + "grad_norm": 0.9223744882506811, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 44587 + }, + { + "epoch": 0.44588, + "grad_norm": 1.1315573920394904, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 44588 + }, + { + "epoch": 0.44589, + "grad_norm": 1.6007395467650762, + "learning_rate": 0.003, + "loss": 4.049, + "step": 44589 + }, + { + "epoch": 0.4459, + "grad_norm": 0.7963775875094157, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 44590 + }, + { + "epoch": 0.44591, + "grad_norm": 0.8600962006307338, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 44591 + }, + { + "epoch": 0.44592, + "grad_norm": 1.0838197147818673, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 44592 + }, + { + "epoch": 0.44593, + "grad_norm": 0.8984491318623539, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 44593 + }, + { + "epoch": 0.44594, + "grad_norm": 1.012508563723094, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 44594 + }, + { + "epoch": 0.44595, + "grad_norm": 1.3583451888392035, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 44595 + }, + { + "epoch": 0.44596, + "grad_norm": 0.9724279964652295, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 44596 + }, + { + "epoch": 0.44597, + "grad_norm": 1.0846263276589456, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 44597 + }, + { + "epoch": 0.44598, + "grad_norm": 1.2522569982487917, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 44598 + }, + { + "epoch": 0.44599, + "grad_norm": 1.0950320513707157, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 44599 + }, + { + "epoch": 0.446, + "grad_norm": 1.0672281562892845, + "learning_rate": 0.003, + "loss": 4.046, + "step": 44600 + }, + { + "epoch": 0.44601, + "grad_norm": 1.1127491893798365, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 44601 + }, + { + "epoch": 0.44602, + "grad_norm": 1.0119223812680336, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 44602 + }, + { + "epoch": 0.44603, + "grad_norm": 1.0657158795915536, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 44603 + }, + { + "epoch": 0.44604, + "grad_norm": 0.9833455695567389, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 44604 + }, + { + "epoch": 0.44605, + "grad_norm": 1.009815855052152, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 44605 + }, + { + "epoch": 0.44606, + "grad_norm": 1.2324432002246324, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 44606 + }, + { + "epoch": 0.44607, + "grad_norm": 0.9486998361775879, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 44607 + }, + { + "epoch": 0.44608, + "grad_norm": 0.9599043731226717, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 44608 + }, + { + "epoch": 0.44609, + "grad_norm": 1.0296210763034166, + "learning_rate": 0.003, + "loss": 4.025, + "step": 44609 + }, + { + "epoch": 0.4461, + "grad_norm": 1.3051441723264838, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 44610 + }, + { + "epoch": 0.44611, + "grad_norm": 1.0544850290786378, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 44611 + }, + { + "epoch": 0.44612, + "grad_norm": 1.0281893734255707, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 44612 + }, + { + "epoch": 0.44613, + "grad_norm": 0.9714585403526769, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 44613 + }, + { + "epoch": 0.44614, + "grad_norm": 1.2615211957335206, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 44614 + }, + { + "epoch": 0.44615, + "grad_norm": 1.0368699281896616, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 44615 + }, + { + "epoch": 0.44616, + "grad_norm": 1.1811331168273282, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 44616 + }, + { + "epoch": 0.44617, + "grad_norm": 0.8135085557087081, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 44617 + }, + { + "epoch": 0.44618, + "grad_norm": 1.0041965101573112, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 44618 + }, + { + "epoch": 0.44619, + "grad_norm": 1.2557494773441331, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 44619 + }, + { + "epoch": 0.4462, + "grad_norm": 1.01888828468945, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 44620 + }, + { + "epoch": 0.44621, + "grad_norm": 1.301024745259451, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 44621 + }, + { + "epoch": 0.44622, + "grad_norm": 1.1531069679421793, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 44622 + }, + { + "epoch": 0.44623, + "grad_norm": 0.9765084707561038, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 44623 + }, + { + "epoch": 0.44624, + "grad_norm": 1.227810778069981, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 44624 + }, + { + "epoch": 0.44625, + "grad_norm": 1.0251736663106545, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 44625 + }, + { + "epoch": 0.44626, + "grad_norm": 1.0400399941252207, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 44626 + }, + { + "epoch": 0.44627, + "grad_norm": 1.0855812595425975, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 44627 + }, + { + "epoch": 0.44628, + "grad_norm": 0.9337109187223201, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 44628 + }, + { + "epoch": 0.44629, + "grad_norm": 1.2051949293696511, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 44629 + }, + { + "epoch": 0.4463, + "grad_norm": 1.2701548628913817, + "learning_rate": 0.003, + "loss": 4.036, + "step": 44630 + }, + { + "epoch": 0.44631, + "grad_norm": 0.8399622400413675, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 44631 + }, + { + "epoch": 0.44632, + "grad_norm": 0.8732843459307577, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 44632 + }, + { + "epoch": 0.44633, + "grad_norm": 1.0513398582397715, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 44633 + }, + { + "epoch": 0.44634, + "grad_norm": 1.1024768205708895, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 44634 + }, + { + "epoch": 0.44635, + "grad_norm": 0.9858094960623831, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 44635 + }, + { + "epoch": 0.44636, + "grad_norm": 1.0877625881350332, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 44636 + }, + { + "epoch": 0.44637, + "grad_norm": 1.198493446755909, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 44637 + }, + { + "epoch": 0.44638, + "grad_norm": 1.028982395095585, + "learning_rate": 0.003, + "loss": 4.042, + "step": 44638 + }, + { + "epoch": 0.44639, + "grad_norm": 1.032134587104425, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 44639 + }, + { + "epoch": 0.4464, + "grad_norm": 1.1369135481572414, + "learning_rate": 0.003, + "loss": 4.017, + "step": 44640 + }, + { + "epoch": 0.44641, + "grad_norm": 1.1150490085819533, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 44641 + }, + { + "epoch": 0.44642, + "grad_norm": 1.0113008097804457, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 44642 + }, + { + "epoch": 0.44643, + "grad_norm": 1.1650092613507768, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 44643 + }, + { + "epoch": 0.44644, + "grad_norm": 1.0622711939423723, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 44644 + }, + { + "epoch": 0.44645, + "grad_norm": 1.3466151628225715, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 44645 + }, + { + "epoch": 0.44646, + "grad_norm": 0.8006757364170912, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 44646 + }, + { + "epoch": 0.44647, + "grad_norm": 0.8165254708874774, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 44647 + }, + { + "epoch": 0.44648, + "grad_norm": 1.0138564496391602, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 44648 + }, + { + "epoch": 0.44649, + "grad_norm": 1.2616073279471924, + "learning_rate": 0.003, + "loss": 4.027, + "step": 44649 + }, + { + "epoch": 0.4465, + "grad_norm": 1.036553780577205, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 44650 + }, + { + "epoch": 0.44651, + "grad_norm": 1.0712192023822247, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 44651 + }, + { + "epoch": 0.44652, + "grad_norm": 1.18808344906809, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 44652 + }, + { + "epoch": 0.44653, + "grad_norm": 1.2286488921379768, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 44653 + }, + { + "epoch": 0.44654, + "grad_norm": 0.9539829384775118, + "learning_rate": 0.003, + "loss": 4.043, + "step": 44654 + }, + { + "epoch": 0.44655, + "grad_norm": 0.9873760236228796, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 44655 + }, + { + "epoch": 0.44656, + "grad_norm": 1.2821142972033013, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 44656 + }, + { + "epoch": 0.44657, + "grad_norm": 0.815431170906312, + "learning_rate": 0.003, + "loss": 4.012, + "step": 44657 + }, + { + "epoch": 0.44658, + "grad_norm": 0.9622084219999194, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 44658 + }, + { + "epoch": 0.44659, + "grad_norm": 1.3343252827876617, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 44659 + }, + { + "epoch": 0.4466, + "grad_norm": 0.8842992853780457, + "learning_rate": 0.003, + "loss": 4.029, + "step": 44660 + }, + { + "epoch": 0.44661, + "grad_norm": 0.8217083540343323, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 44661 + }, + { + "epoch": 0.44662, + "grad_norm": 0.8314311188957536, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 44662 + }, + { + "epoch": 0.44663, + "grad_norm": 0.9455888824952191, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 44663 + }, + { + "epoch": 0.44664, + "grad_norm": 1.269745540617019, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 44664 + }, + { + "epoch": 0.44665, + "grad_norm": 0.9258191051016781, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 44665 + }, + { + "epoch": 0.44666, + "grad_norm": 1.0881026180256714, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 44666 + }, + { + "epoch": 0.44667, + "grad_norm": 1.20283070053246, + "learning_rate": 0.003, + "loss": 4.044, + "step": 44667 + }, + { + "epoch": 0.44668, + "grad_norm": 1.2009980062268906, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 44668 + }, + { + "epoch": 0.44669, + "grad_norm": 0.8771179170093242, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 44669 + }, + { + "epoch": 0.4467, + "grad_norm": 0.9172436956928477, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 44670 + }, + { + "epoch": 0.44671, + "grad_norm": 1.2441453514270098, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 44671 + }, + { + "epoch": 0.44672, + "grad_norm": 1.2460086404894333, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 44672 + }, + { + "epoch": 0.44673, + "grad_norm": 0.8038080803408904, + "learning_rate": 0.003, + "loss": 4.052, + "step": 44673 + }, + { + "epoch": 0.44674, + "grad_norm": 0.7518963703956492, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 44674 + }, + { + "epoch": 0.44675, + "grad_norm": 0.8575484908430027, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 44675 + }, + { + "epoch": 0.44676, + "grad_norm": 0.9330605540525504, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 44676 + }, + { + "epoch": 0.44677, + "grad_norm": 0.8720133078481842, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 44677 + }, + { + "epoch": 0.44678, + "grad_norm": 1.0296039106107415, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 44678 + }, + { + "epoch": 0.44679, + "grad_norm": 1.3217853398951493, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 44679 + }, + { + "epoch": 0.4468, + "grad_norm": 0.7113212222841819, + "learning_rate": 0.003, + "loss": 4.024, + "step": 44680 + }, + { + "epoch": 0.44681, + "grad_norm": 0.813072809069434, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 44681 + }, + { + "epoch": 0.44682, + "grad_norm": 0.9230586468785797, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 44682 + }, + { + "epoch": 0.44683, + "grad_norm": 1.0051490082715546, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 44683 + }, + { + "epoch": 0.44684, + "grad_norm": 1.3282296622192378, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 44684 + }, + { + "epoch": 0.44685, + "grad_norm": 1.013657803673048, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 44685 + }, + { + "epoch": 0.44686, + "grad_norm": 1.1975535068382601, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 44686 + }, + { + "epoch": 0.44687, + "grad_norm": 1.367258126081377, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 44687 + }, + { + "epoch": 0.44688, + "grad_norm": 1.0305885837551172, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 44688 + }, + { + "epoch": 0.44689, + "grad_norm": 1.1189380880098094, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 44689 + }, + { + "epoch": 0.4469, + "grad_norm": 1.3093946722651584, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 44690 + }, + { + "epoch": 0.44691, + "grad_norm": 0.9012134648720082, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 44691 + }, + { + "epoch": 0.44692, + "grad_norm": 0.881300626177909, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 44692 + }, + { + "epoch": 0.44693, + "grad_norm": 1.2490063398282174, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 44693 + }, + { + "epoch": 0.44694, + "grad_norm": 1.0679714485641945, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 44694 + }, + { + "epoch": 0.44695, + "grad_norm": 1.0668795183249056, + "learning_rate": 0.003, + "loss": 4.015, + "step": 44695 + }, + { + "epoch": 0.44696, + "grad_norm": 1.178464953061859, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 44696 + }, + { + "epoch": 0.44697, + "grad_norm": 1.0982152489923782, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 44697 + }, + { + "epoch": 0.44698, + "grad_norm": 1.226472870122142, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 44698 + }, + { + "epoch": 0.44699, + "grad_norm": 1.2509525478492032, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 44699 + }, + { + "epoch": 0.447, + "grad_norm": 0.9063072039917868, + "learning_rate": 0.003, + "loss": 4.037, + "step": 44700 + }, + { + "epoch": 0.44701, + "grad_norm": 0.9039423749710495, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 44701 + }, + { + "epoch": 0.44702, + "grad_norm": 1.1883687444668676, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 44702 + }, + { + "epoch": 0.44703, + "grad_norm": 1.2153675712270589, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 44703 + }, + { + "epoch": 0.44704, + "grad_norm": 0.9038237641806595, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 44704 + }, + { + "epoch": 0.44705, + "grad_norm": 0.8104303257783578, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 44705 + }, + { + "epoch": 0.44706, + "grad_norm": 0.9092480713309052, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 44706 + }, + { + "epoch": 0.44707, + "grad_norm": 1.4079814589916813, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 44707 + }, + { + "epoch": 0.44708, + "grad_norm": 1.0194839850350916, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 44708 + }, + { + "epoch": 0.44709, + "grad_norm": 1.0093785122125085, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 44709 + }, + { + "epoch": 0.4471, + "grad_norm": 1.1452549623067556, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 44710 + }, + { + "epoch": 0.44711, + "grad_norm": 1.2797966434967685, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 44711 + }, + { + "epoch": 0.44712, + "grad_norm": 0.9908942921497663, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 44712 + }, + { + "epoch": 0.44713, + "grad_norm": 1.0263919683258262, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 44713 + }, + { + "epoch": 0.44714, + "grad_norm": 1.2143998625907524, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 44714 + }, + { + "epoch": 0.44715, + "grad_norm": 0.8979161111078435, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 44715 + }, + { + "epoch": 0.44716, + "grad_norm": 0.9442415226475489, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 44716 + }, + { + "epoch": 0.44717, + "grad_norm": 0.9563538172798441, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 44717 + }, + { + "epoch": 0.44718, + "grad_norm": 1.2278697204832578, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 44718 + }, + { + "epoch": 0.44719, + "grad_norm": 0.9816275373931328, + "learning_rate": 0.003, + "loss": 4.055, + "step": 44719 + }, + { + "epoch": 0.4472, + "grad_norm": 1.1212520456707211, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 44720 + }, + { + "epoch": 0.44721, + "grad_norm": 0.9792153628064221, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 44721 + }, + { + "epoch": 0.44722, + "grad_norm": 1.1252664174605005, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 44722 + }, + { + "epoch": 0.44723, + "grad_norm": 1.1525569993293614, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 44723 + }, + { + "epoch": 0.44724, + "grad_norm": 1.123487459286113, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 44724 + }, + { + "epoch": 0.44725, + "grad_norm": 1.0391036879322042, + "learning_rate": 0.003, + "loss": 4.022, + "step": 44725 + }, + { + "epoch": 0.44726, + "grad_norm": 1.2988291596233947, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 44726 + }, + { + "epoch": 0.44727, + "grad_norm": 1.0122209952877361, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 44727 + }, + { + "epoch": 0.44728, + "grad_norm": 1.2244597349810793, + "learning_rate": 0.003, + "loss": 4.016, + "step": 44728 + }, + { + "epoch": 0.44729, + "grad_norm": 1.1136296803845036, + "learning_rate": 0.003, + "loss": 4.04, + "step": 44729 + }, + { + "epoch": 0.4473, + "grad_norm": 0.9299435382865001, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 44730 + }, + { + "epoch": 0.44731, + "grad_norm": 1.1372435809091368, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 44731 + }, + { + "epoch": 0.44732, + "grad_norm": 1.1007369133264493, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 44732 + }, + { + "epoch": 0.44733, + "grad_norm": 1.0518870854978382, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 44733 + }, + { + "epoch": 0.44734, + "grad_norm": 1.0096309016920457, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 44734 + }, + { + "epoch": 0.44735, + "grad_norm": 0.9284319641851015, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 44735 + }, + { + "epoch": 0.44736, + "grad_norm": 0.9225092109417733, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 44736 + }, + { + "epoch": 0.44737, + "grad_norm": 0.9751313816676143, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 44737 + }, + { + "epoch": 0.44738, + "grad_norm": 1.1876795928403943, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 44738 + }, + { + "epoch": 0.44739, + "grad_norm": 1.0833306768628372, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 44739 + }, + { + "epoch": 0.4474, + "grad_norm": 1.0792129967360278, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 44740 + }, + { + "epoch": 0.44741, + "grad_norm": 1.2065860206974033, + "learning_rate": 0.003, + "loss": 4.024, + "step": 44741 + }, + { + "epoch": 0.44742, + "grad_norm": 0.9477671605095405, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 44742 + }, + { + "epoch": 0.44743, + "grad_norm": 1.1425100971369124, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 44743 + }, + { + "epoch": 0.44744, + "grad_norm": 1.2161679882632106, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 44744 + }, + { + "epoch": 0.44745, + "grad_norm": 0.9787509791379762, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 44745 + }, + { + "epoch": 0.44746, + "grad_norm": 0.8846909598713498, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 44746 + }, + { + "epoch": 0.44747, + "grad_norm": 0.8656200799318406, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 44747 + }, + { + "epoch": 0.44748, + "grad_norm": 0.9427453266120417, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 44748 + }, + { + "epoch": 0.44749, + "grad_norm": 1.1913103257423785, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 44749 + }, + { + "epoch": 0.4475, + "grad_norm": 1.2528296993648964, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 44750 + }, + { + "epoch": 0.44751, + "grad_norm": 1.183413517864566, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 44751 + }, + { + "epoch": 0.44752, + "grad_norm": 0.9109843576100272, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 44752 + }, + { + "epoch": 0.44753, + "grad_norm": 0.9815950926951957, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 44753 + }, + { + "epoch": 0.44754, + "grad_norm": 1.0367867255481995, + "learning_rate": 0.003, + "loss": 4.032, + "step": 44754 + }, + { + "epoch": 0.44755, + "grad_norm": 0.9841529822461238, + "learning_rate": 0.003, + "loss": 4.02, + "step": 44755 + }, + { + "epoch": 0.44756, + "grad_norm": 1.1273077262728637, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 44756 + }, + { + "epoch": 0.44757, + "grad_norm": 1.0681510827753133, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 44757 + }, + { + "epoch": 0.44758, + "grad_norm": 1.0558999102607618, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 44758 + }, + { + "epoch": 0.44759, + "grad_norm": 1.2228461611138337, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 44759 + }, + { + "epoch": 0.4476, + "grad_norm": 1.0754196097232191, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 44760 + }, + { + "epoch": 0.44761, + "grad_norm": 0.9043413137628373, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 44761 + }, + { + "epoch": 0.44762, + "grad_norm": 0.945499570707807, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 44762 + }, + { + "epoch": 0.44763, + "grad_norm": 0.9480251459677245, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 44763 + }, + { + "epoch": 0.44764, + "grad_norm": 1.1704396363195606, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 44764 + }, + { + "epoch": 0.44765, + "grad_norm": 1.2156380595413294, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 44765 + }, + { + "epoch": 0.44766, + "grad_norm": 1.2071556677962445, + "learning_rate": 0.003, + "loss": 4.053, + "step": 44766 + }, + { + "epoch": 0.44767, + "grad_norm": 1.0160191860562353, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 44767 + }, + { + "epoch": 0.44768, + "grad_norm": 1.0178064849543518, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 44768 + }, + { + "epoch": 0.44769, + "grad_norm": 1.230291153979386, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 44769 + }, + { + "epoch": 0.4477, + "grad_norm": 1.0220422698727767, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 44770 + }, + { + "epoch": 0.44771, + "grad_norm": 1.0811619288422067, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 44771 + }, + { + "epoch": 0.44772, + "grad_norm": 1.2352325950351826, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 44772 + }, + { + "epoch": 0.44773, + "grad_norm": 1.14722431320917, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 44773 + }, + { + "epoch": 0.44774, + "grad_norm": 1.1355407730381006, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 44774 + }, + { + "epoch": 0.44775, + "grad_norm": 1.0150999089713486, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 44775 + }, + { + "epoch": 0.44776, + "grad_norm": 0.9899081049228905, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 44776 + }, + { + "epoch": 0.44777, + "grad_norm": 1.334973626120232, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 44777 + }, + { + "epoch": 0.44778, + "grad_norm": 0.8434489233787698, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 44778 + }, + { + "epoch": 0.44779, + "grad_norm": 0.7878472918746652, + "learning_rate": 0.003, + "loss": 4.048, + "step": 44779 + }, + { + "epoch": 0.4478, + "grad_norm": 0.9773605962504239, + "learning_rate": 0.003, + "loss": 4.015, + "step": 44780 + }, + { + "epoch": 0.44781, + "grad_norm": 1.2351789943329017, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 44781 + }, + { + "epoch": 0.44782, + "grad_norm": 0.8721535045716633, + "learning_rate": 0.003, + "loss": 4.043, + "step": 44782 + }, + { + "epoch": 0.44783, + "grad_norm": 0.9718176417172056, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 44783 + }, + { + "epoch": 0.44784, + "grad_norm": 1.1546744262754085, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 44784 + }, + { + "epoch": 0.44785, + "grad_norm": 0.8598411001585069, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 44785 + }, + { + "epoch": 0.44786, + "grad_norm": 0.9987534180035523, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 44786 + }, + { + "epoch": 0.44787, + "grad_norm": 1.2838996720862357, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 44787 + }, + { + "epoch": 0.44788, + "grad_norm": 0.8181000612884972, + "learning_rate": 0.003, + "loss": 4.049, + "step": 44788 + }, + { + "epoch": 0.44789, + "grad_norm": 0.872682068021396, + "learning_rate": 0.003, + "loss": 4.009, + "step": 44789 + }, + { + "epoch": 0.4479, + "grad_norm": 1.0576301232465934, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 44790 + }, + { + "epoch": 0.44791, + "grad_norm": 1.1085692732818586, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 44791 + }, + { + "epoch": 0.44792, + "grad_norm": 1.0518721137774518, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 44792 + }, + { + "epoch": 0.44793, + "grad_norm": 1.212417335920578, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 44793 + }, + { + "epoch": 0.44794, + "grad_norm": 0.9131613851106605, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 44794 + }, + { + "epoch": 0.44795, + "grad_norm": 1.124803396049686, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 44795 + }, + { + "epoch": 0.44796, + "grad_norm": 1.3056071507799845, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 44796 + }, + { + "epoch": 0.44797, + "grad_norm": 0.8522696458916402, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 44797 + }, + { + "epoch": 0.44798, + "grad_norm": 0.9077770106816513, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 44798 + }, + { + "epoch": 0.44799, + "grad_norm": 0.960507303702146, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 44799 + }, + { + "epoch": 0.448, + "grad_norm": 1.3705229336236389, + "learning_rate": 0.003, + "loss": 4.056, + "step": 44800 + }, + { + "epoch": 0.44801, + "grad_norm": 0.8836923911055441, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 44801 + }, + { + "epoch": 0.44802, + "grad_norm": 0.86192769185258, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 44802 + }, + { + "epoch": 0.44803, + "grad_norm": 1.0871428750122771, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 44803 + }, + { + "epoch": 0.44804, + "grad_norm": 1.293387743753302, + "learning_rate": 0.003, + "loss": 3.982, + "step": 44804 + }, + { + "epoch": 0.44805, + "grad_norm": 1.1004076630792738, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 44805 + }, + { + "epoch": 0.44806, + "grad_norm": 1.0058363243584254, + "learning_rate": 0.003, + "loss": 4.011, + "step": 44806 + }, + { + "epoch": 0.44807, + "grad_norm": 1.2004640980691315, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 44807 + }, + { + "epoch": 0.44808, + "grad_norm": 0.8841967916381662, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 44808 + }, + { + "epoch": 0.44809, + "grad_norm": 0.8611875060320192, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 44809 + }, + { + "epoch": 0.4481, + "grad_norm": 0.9496086711058591, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 44810 + }, + { + "epoch": 0.44811, + "grad_norm": 1.0183860127067355, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 44811 + }, + { + "epoch": 0.44812, + "grad_norm": 1.051342107329849, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 44812 + }, + { + "epoch": 0.44813, + "grad_norm": 1.0406164421028514, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 44813 + }, + { + "epoch": 0.44814, + "grad_norm": 1.0690008571847724, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 44814 + }, + { + "epoch": 0.44815, + "grad_norm": 1.010001012297076, + "learning_rate": 0.003, + "loss": 4.017, + "step": 44815 + }, + { + "epoch": 0.44816, + "grad_norm": 1.2779597960117925, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 44816 + }, + { + "epoch": 0.44817, + "grad_norm": 1.0099419159466674, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 44817 + }, + { + "epoch": 0.44818, + "grad_norm": 1.2240896990139822, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 44818 + }, + { + "epoch": 0.44819, + "grad_norm": 1.0742578401385943, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 44819 + }, + { + "epoch": 0.4482, + "grad_norm": 1.0229092090116276, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 44820 + }, + { + "epoch": 0.44821, + "grad_norm": 1.123923277421345, + "learning_rate": 0.003, + "loss": 4.043, + "step": 44821 + }, + { + "epoch": 0.44822, + "grad_norm": 1.1932610963677182, + "learning_rate": 0.003, + "loss": 4.031, + "step": 44822 + }, + { + "epoch": 0.44823, + "grad_norm": 1.1774886383115122, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 44823 + }, + { + "epoch": 0.44824, + "grad_norm": 1.00008327940341, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 44824 + }, + { + "epoch": 0.44825, + "grad_norm": 1.163634698241442, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 44825 + }, + { + "epoch": 0.44826, + "grad_norm": 0.9675975301190218, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 44826 + }, + { + "epoch": 0.44827, + "grad_norm": 0.9237235174050565, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 44827 + }, + { + "epoch": 0.44828, + "grad_norm": 1.103710431396244, + "learning_rate": 0.003, + "loss": 4.033, + "step": 44828 + }, + { + "epoch": 0.44829, + "grad_norm": 1.0294847965542937, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 44829 + }, + { + "epoch": 0.4483, + "grad_norm": 1.352725886896396, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 44830 + }, + { + "epoch": 0.44831, + "grad_norm": 0.9712561957916906, + "learning_rate": 0.003, + "loss": 4.048, + "step": 44831 + }, + { + "epoch": 0.44832, + "grad_norm": 1.3055030945952035, + "learning_rate": 0.003, + "loss": 4.028, + "step": 44832 + }, + { + "epoch": 0.44833, + "grad_norm": 0.9795354891783017, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 44833 + }, + { + "epoch": 0.44834, + "grad_norm": 1.2282263852295952, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 44834 + }, + { + "epoch": 0.44835, + "grad_norm": 1.175972821337585, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 44835 + }, + { + "epoch": 0.44836, + "grad_norm": 0.8402268740842793, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 44836 + }, + { + "epoch": 0.44837, + "grad_norm": 1.090839602120492, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 44837 + }, + { + "epoch": 0.44838, + "grad_norm": 1.1597561680564235, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 44838 + }, + { + "epoch": 0.44839, + "grad_norm": 0.9358049885622975, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 44839 + }, + { + "epoch": 0.4484, + "grad_norm": 1.0002077625918522, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 44840 + }, + { + "epoch": 0.44841, + "grad_norm": 1.4131918037231668, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 44841 + }, + { + "epoch": 0.44842, + "grad_norm": 0.908826307559977, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 44842 + }, + { + "epoch": 0.44843, + "grad_norm": 0.9177877849615217, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 44843 + }, + { + "epoch": 0.44844, + "grad_norm": 0.9774119973517855, + "learning_rate": 0.003, + "loss": 4.027, + "step": 44844 + }, + { + "epoch": 0.44845, + "grad_norm": 0.8933359070384038, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 44845 + }, + { + "epoch": 0.44846, + "grad_norm": 1.027720656446259, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 44846 + }, + { + "epoch": 0.44847, + "grad_norm": 1.3914557981329108, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 44847 + }, + { + "epoch": 0.44848, + "grad_norm": 1.0985050842140414, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 44848 + }, + { + "epoch": 0.44849, + "grad_norm": 1.2791676479886538, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 44849 + }, + { + "epoch": 0.4485, + "grad_norm": 0.9496163276157359, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 44850 + }, + { + "epoch": 0.44851, + "grad_norm": 0.993700654855454, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 44851 + }, + { + "epoch": 0.44852, + "grad_norm": 1.1041394552090849, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 44852 + }, + { + "epoch": 0.44853, + "grad_norm": 1.0836616562813988, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 44853 + }, + { + "epoch": 0.44854, + "grad_norm": 1.2285195886434206, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 44854 + }, + { + "epoch": 0.44855, + "grad_norm": 0.89918263849289, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 44855 + }, + { + "epoch": 0.44856, + "grad_norm": 1.0791550000859071, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 44856 + }, + { + "epoch": 0.44857, + "grad_norm": 1.1550966808454428, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 44857 + }, + { + "epoch": 0.44858, + "grad_norm": 1.03039363009361, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 44858 + }, + { + "epoch": 0.44859, + "grad_norm": 1.0238565637289874, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 44859 + }, + { + "epoch": 0.4486, + "grad_norm": 0.9444596221078959, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 44860 + }, + { + "epoch": 0.44861, + "grad_norm": 1.0895974829766373, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 44861 + }, + { + "epoch": 0.44862, + "grad_norm": 1.3570772434593967, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 44862 + }, + { + "epoch": 0.44863, + "grad_norm": 0.8378766210935679, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 44863 + }, + { + "epoch": 0.44864, + "grad_norm": 1.033186314755237, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 44864 + }, + { + "epoch": 0.44865, + "grad_norm": 1.5803875091769126, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 44865 + }, + { + "epoch": 0.44866, + "grad_norm": 0.9088078850487163, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 44866 + }, + { + "epoch": 0.44867, + "grad_norm": 0.8899325318045774, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 44867 + }, + { + "epoch": 0.44868, + "grad_norm": 1.098881449543579, + "learning_rate": 0.003, + "loss": 4.039, + "step": 44868 + }, + { + "epoch": 0.44869, + "grad_norm": 0.9234230467726111, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 44869 + }, + { + "epoch": 0.4487, + "grad_norm": 1.1193523556664051, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 44870 + }, + { + "epoch": 0.44871, + "grad_norm": 1.1129406830154627, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 44871 + }, + { + "epoch": 0.44872, + "grad_norm": 1.0559184242984818, + "learning_rate": 0.003, + "loss": 4.04, + "step": 44872 + }, + { + "epoch": 0.44873, + "grad_norm": 0.9894157971495477, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 44873 + }, + { + "epoch": 0.44874, + "grad_norm": 1.3178673263433018, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 44874 + }, + { + "epoch": 0.44875, + "grad_norm": 1.173543109238927, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 44875 + }, + { + "epoch": 0.44876, + "grad_norm": 1.1368787456375007, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 44876 + }, + { + "epoch": 0.44877, + "grad_norm": 1.0027470053226006, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 44877 + }, + { + "epoch": 0.44878, + "grad_norm": 1.1293045857936783, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 44878 + }, + { + "epoch": 0.44879, + "grad_norm": 1.1218297385507898, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 44879 + }, + { + "epoch": 0.4488, + "grad_norm": 1.1981873289104594, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 44880 + }, + { + "epoch": 0.44881, + "grad_norm": 1.0043488941777443, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 44881 + }, + { + "epoch": 0.44882, + "grad_norm": 1.0995530950038086, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 44882 + }, + { + "epoch": 0.44883, + "grad_norm": 0.9394255388645252, + "learning_rate": 0.003, + "loss": 3.988, + "step": 44883 + }, + { + "epoch": 0.44884, + "grad_norm": 0.9982302762851062, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 44884 + }, + { + "epoch": 0.44885, + "grad_norm": 1.2193242043931762, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 44885 + }, + { + "epoch": 0.44886, + "grad_norm": 1.0722232053258736, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 44886 + }, + { + "epoch": 0.44887, + "grad_norm": 1.0288660358787696, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 44887 + }, + { + "epoch": 0.44888, + "grad_norm": 1.1038688610325085, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 44888 + }, + { + "epoch": 0.44889, + "grad_norm": 1.136908725223507, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 44889 + }, + { + "epoch": 0.4489, + "grad_norm": 1.1974836391141135, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 44890 + }, + { + "epoch": 0.44891, + "grad_norm": 0.9555161334840833, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 44891 + }, + { + "epoch": 0.44892, + "grad_norm": 1.0638869183203608, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 44892 + }, + { + "epoch": 0.44893, + "grad_norm": 1.3015664757470258, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 44893 + }, + { + "epoch": 0.44894, + "grad_norm": 1.20013445044124, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 44894 + }, + { + "epoch": 0.44895, + "grad_norm": 0.8868206567135678, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 44895 + }, + { + "epoch": 0.44896, + "grad_norm": 0.8903268536478024, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 44896 + }, + { + "epoch": 0.44897, + "grad_norm": 1.0709813044794805, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 44897 + }, + { + "epoch": 0.44898, + "grad_norm": 1.315963331587594, + "learning_rate": 0.003, + "loss": 4.016, + "step": 44898 + }, + { + "epoch": 0.44899, + "grad_norm": 0.8692076501559662, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 44899 + }, + { + "epoch": 0.449, + "grad_norm": 0.8979456946890935, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 44900 + }, + { + "epoch": 0.44901, + "grad_norm": 1.0421027365220237, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 44901 + }, + { + "epoch": 0.44902, + "grad_norm": 1.2386512710233373, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 44902 + }, + { + "epoch": 0.44903, + "grad_norm": 0.8625790641758554, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 44903 + }, + { + "epoch": 0.44904, + "grad_norm": 0.9822120114517479, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 44904 + }, + { + "epoch": 0.44905, + "grad_norm": 1.1121404241178576, + "learning_rate": 0.003, + "loss": 4.072, + "step": 44905 + }, + { + "epoch": 0.44906, + "grad_norm": 1.1717695852814853, + "learning_rate": 0.003, + "loss": 4.038, + "step": 44906 + }, + { + "epoch": 0.44907, + "grad_norm": 1.0100246580682692, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 44907 + }, + { + "epoch": 0.44908, + "grad_norm": 1.21144438015461, + "learning_rate": 0.003, + "loss": 4.043, + "step": 44908 + }, + { + "epoch": 0.44909, + "grad_norm": 1.0200278765180437, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 44909 + }, + { + "epoch": 0.4491, + "grad_norm": 1.0077127767962495, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 44910 + }, + { + "epoch": 0.44911, + "grad_norm": 0.9795198539511956, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 44911 + }, + { + "epoch": 0.44912, + "grad_norm": 0.9088016580291456, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 44912 + }, + { + "epoch": 0.44913, + "grad_norm": 1.1500537951238858, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 44913 + }, + { + "epoch": 0.44914, + "grad_norm": 1.0401505183190538, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 44914 + }, + { + "epoch": 0.44915, + "grad_norm": 1.28362667587413, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 44915 + }, + { + "epoch": 0.44916, + "grad_norm": 1.2845811892109624, + "learning_rate": 0.003, + "loss": 4.026, + "step": 44916 + }, + { + "epoch": 0.44917, + "grad_norm": 0.9546613894381892, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 44917 + }, + { + "epoch": 0.44918, + "grad_norm": 0.977557200464471, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 44918 + }, + { + "epoch": 0.44919, + "grad_norm": 1.1978777440622195, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 44919 + }, + { + "epoch": 0.4492, + "grad_norm": 1.0229257140934451, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 44920 + }, + { + "epoch": 0.44921, + "grad_norm": 1.0576741731864638, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 44921 + }, + { + "epoch": 0.44922, + "grad_norm": 1.177189945131432, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 44922 + }, + { + "epoch": 0.44923, + "grad_norm": 1.0579981440208093, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 44923 + }, + { + "epoch": 0.44924, + "grad_norm": 1.2027775979199438, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 44924 + }, + { + "epoch": 0.44925, + "grad_norm": 0.8485950781211912, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 44925 + }, + { + "epoch": 0.44926, + "grad_norm": 0.9066350298812521, + "learning_rate": 0.003, + "loss": 4.044, + "step": 44926 + }, + { + "epoch": 0.44927, + "grad_norm": 1.2029646208949218, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 44927 + }, + { + "epoch": 0.44928, + "grad_norm": 1.1403977555679625, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 44928 + }, + { + "epoch": 0.44929, + "grad_norm": 1.0223742869138106, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 44929 + }, + { + "epoch": 0.4493, + "grad_norm": 1.1360911660988, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 44930 + }, + { + "epoch": 0.44931, + "grad_norm": 1.0048152015624545, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 44931 + }, + { + "epoch": 0.44932, + "grad_norm": 1.280717745614132, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 44932 + }, + { + "epoch": 0.44933, + "grad_norm": 1.1325212926915984, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 44933 + }, + { + "epoch": 0.44934, + "grad_norm": 1.016018704404967, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 44934 + }, + { + "epoch": 0.44935, + "grad_norm": 1.232387242211279, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 44935 + }, + { + "epoch": 0.44936, + "grad_norm": 1.0531020790180925, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 44936 + }, + { + "epoch": 0.44937, + "grad_norm": 1.174857130806944, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 44937 + }, + { + "epoch": 0.44938, + "grad_norm": 0.9781189583742885, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 44938 + }, + { + "epoch": 0.44939, + "grad_norm": 1.2498868777597607, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 44939 + }, + { + "epoch": 0.4494, + "grad_norm": 0.9407355481426319, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 44940 + }, + { + "epoch": 0.44941, + "grad_norm": 1.0502469070716396, + "learning_rate": 0.003, + "loss": 4.015, + "step": 44941 + }, + { + "epoch": 0.44942, + "grad_norm": 1.1270796993137118, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 44942 + }, + { + "epoch": 0.44943, + "grad_norm": 1.0082655053800786, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 44943 + }, + { + "epoch": 0.44944, + "grad_norm": 1.1440620400254995, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 44944 + }, + { + "epoch": 0.44945, + "grad_norm": 1.0433912830155565, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 44945 + }, + { + "epoch": 0.44946, + "grad_norm": 1.0592141901300445, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 44946 + }, + { + "epoch": 0.44947, + "grad_norm": 1.0131618908179996, + "learning_rate": 0.003, + "loss": 4.022, + "step": 44947 + }, + { + "epoch": 0.44948, + "grad_norm": 1.1194636845334496, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 44948 + }, + { + "epoch": 0.44949, + "grad_norm": 1.0704444193076705, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 44949 + }, + { + "epoch": 0.4495, + "grad_norm": 1.0632380416221079, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 44950 + }, + { + "epoch": 0.44951, + "grad_norm": 1.1433734052898072, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 44951 + }, + { + "epoch": 0.44952, + "grad_norm": 1.2240733011074292, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 44952 + }, + { + "epoch": 0.44953, + "grad_norm": 0.9567967345548315, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 44953 + }, + { + "epoch": 0.44954, + "grad_norm": 1.1100736222952443, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 44954 + }, + { + "epoch": 0.44955, + "grad_norm": 0.9950892472688448, + "learning_rate": 0.003, + "loss": 4.01, + "step": 44955 + }, + { + "epoch": 0.44956, + "grad_norm": 1.072383945254462, + "learning_rate": 0.003, + "loss": 4.048, + "step": 44956 + }, + { + "epoch": 0.44957, + "grad_norm": 0.9520686345079378, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 44957 + }, + { + "epoch": 0.44958, + "grad_norm": 0.9795326049523264, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 44958 + }, + { + "epoch": 0.44959, + "grad_norm": 1.3509709176293663, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 44959 + }, + { + "epoch": 0.4496, + "grad_norm": 1.1589503432449857, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 44960 + }, + { + "epoch": 0.44961, + "grad_norm": 0.9619358067921869, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 44961 + }, + { + "epoch": 0.44962, + "grad_norm": 1.0019546018936094, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 44962 + }, + { + "epoch": 0.44963, + "grad_norm": 1.2482675777867451, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 44963 + }, + { + "epoch": 0.44964, + "grad_norm": 1.215699187652039, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 44964 + }, + { + "epoch": 0.44965, + "grad_norm": 1.0811137253621654, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 44965 + }, + { + "epoch": 0.44966, + "grad_norm": 1.1908342233118874, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 44966 + }, + { + "epoch": 0.44967, + "grad_norm": 1.0296347191516237, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 44967 + }, + { + "epoch": 0.44968, + "grad_norm": 1.020981781010401, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 44968 + }, + { + "epoch": 0.44969, + "grad_norm": 1.0869730278258356, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 44969 + }, + { + "epoch": 0.4497, + "grad_norm": 1.2053007389170258, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 44970 + }, + { + "epoch": 0.44971, + "grad_norm": 1.2115811761445887, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 44971 + }, + { + "epoch": 0.44972, + "grad_norm": 1.1508518909816843, + "learning_rate": 0.003, + "loss": 4.011, + "step": 44972 + }, + { + "epoch": 0.44973, + "grad_norm": 1.0495569400573586, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 44973 + }, + { + "epoch": 0.44974, + "grad_norm": 1.2166772165113289, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 44974 + }, + { + "epoch": 0.44975, + "grad_norm": 1.2447829691615475, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 44975 + }, + { + "epoch": 0.44976, + "grad_norm": 0.8918917701360525, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 44976 + }, + { + "epoch": 0.44977, + "grad_norm": 0.8744776765405522, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 44977 + }, + { + "epoch": 0.44978, + "grad_norm": 1.0620782106022435, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 44978 + }, + { + "epoch": 0.44979, + "grad_norm": 1.3010723204272907, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 44979 + }, + { + "epoch": 0.4498, + "grad_norm": 0.9724848461592108, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 44980 + }, + { + "epoch": 0.44981, + "grad_norm": 1.0623396698098313, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 44981 + }, + { + "epoch": 0.44982, + "grad_norm": 0.986206091732325, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 44982 + }, + { + "epoch": 0.44983, + "grad_norm": 1.2207011754746209, + "learning_rate": 0.003, + "loss": 4.025, + "step": 44983 + }, + { + "epoch": 0.44984, + "grad_norm": 0.882152926636427, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 44984 + }, + { + "epoch": 0.44985, + "grad_norm": 1.0439141271359598, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 44985 + }, + { + "epoch": 0.44986, + "grad_norm": 1.1393087102975292, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 44986 + }, + { + "epoch": 0.44987, + "grad_norm": 1.1753823600864004, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 44987 + }, + { + "epoch": 0.44988, + "grad_norm": 1.2016066508725554, + "learning_rate": 0.003, + "loss": 4.032, + "step": 44988 + }, + { + "epoch": 0.44989, + "grad_norm": 1.0686828984707537, + "learning_rate": 0.003, + "loss": 4.038, + "step": 44989 + }, + { + "epoch": 0.4499, + "grad_norm": 1.1738776710800238, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 44990 + }, + { + "epoch": 0.44991, + "grad_norm": 0.922748085285623, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 44991 + }, + { + "epoch": 0.44992, + "grad_norm": 0.8690215628360138, + "learning_rate": 0.003, + "loss": 4.067, + "step": 44992 + }, + { + "epoch": 0.44993, + "grad_norm": 0.8709797574454309, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 44993 + }, + { + "epoch": 0.44994, + "grad_norm": 0.9791759660889897, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 44994 + }, + { + "epoch": 0.44995, + "grad_norm": 1.0970799145641272, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 44995 + }, + { + "epoch": 0.44996, + "grad_norm": 0.94083431636341, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 44996 + }, + { + "epoch": 0.44997, + "grad_norm": 1.1342280151764004, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 44997 + }, + { + "epoch": 0.44998, + "grad_norm": 1.1327113798421147, + "learning_rate": 0.003, + "loss": 4.05, + "step": 44998 + }, + { + "epoch": 0.44999, + "grad_norm": 1.118559252233198, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 44999 + }, + { + "epoch": 0.45, + "grad_norm": 1.270064555794198, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 45000 + }, + { + "epoch": 0.45001, + "grad_norm": 0.9677396689252453, + "learning_rate": 0.003, + "loss": 4.02, + "step": 45001 + }, + { + "epoch": 0.45002, + "grad_norm": 1.0467190947514942, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 45002 + }, + { + "epoch": 0.45003, + "grad_norm": 1.076393396106475, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 45003 + }, + { + "epoch": 0.45004, + "grad_norm": 0.9874735834581218, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 45004 + }, + { + "epoch": 0.45005, + "grad_norm": 0.9786225592021064, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 45005 + }, + { + "epoch": 0.45006, + "grad_norm": 1.19052766116725, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 45006 + }, + { + "epoch": 0.45007, + "grad_norm": 1.1345652431433186, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 45007 + }, + { + "epoch": 0.45008, + "grad_norm": 1.0570405752255885, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 45008 + }, + { + "epoch": 0.45009, + "grad_norm": 1.0337227681462675, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 45009 + }, + { + "epoch": 0.4501, + "grad_norm": 1.0980454181801063, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 45010 + }, + { + "epoch": 0.45011, + "grad_norm": 1.2557068688446984, + "learning_rate": 0.003, + "loss": 4.037, + "step": 45011 + }, + { + "epoch": 0.45012, + "grad_norm": 0.7838821663897987, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 45012 + }, + { + "epoch": 0.45013, + "grad_norm": 0.8313447637184707, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 45013 + }, + { + "epoch": 0.45014, + "grad_norm": 0.9880729484054028, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 45014 + }, + { + "epoch": 0.45015, + "grad_norm": 1.0570489235215763, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 45015 + }, + { + "epoch": 0.45016, + "grad_norm": 0.9755800517891178, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 45016 + }, + { + "epoch": 0.45017, + "grad_norm": 1.1153892163560781, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 45017 + }, + { + "epoch": 0.45018, + "grad_norm": 1.1572455228641707, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 45018 + }, + { + "epoch": 0.45019, + "grad_norm": 0.888672641407827, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 45019 + }, + { + "epoch": 0.4502, + "grad_norm": 0.9301256377045819, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 45020 + }, + { + "epoch": 0.45021, + "grad_norm": 0.926639403117289, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 45021 + }, + { + "epoch": 0.45022, + "grad_norm": 0.920554111039553, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 45022 + }, + { + "epoch": 0.45023, + "grad_norm": 1.0219130402487595, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 45023 + }, + { + "epoch": 0.45024, + "grad_norm": 1.1739935212799202, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 45024 + }, + { + "epoch": 0.45025, + "grad_norm": 0.9480232884934844, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 45025 + }, + { + "epoch": 0.45026, + "grad_norm": 0.9870108457214739, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 45026 + }, + { + "epoch": 0.45027, + "grad_norm": 1.3249771438754983, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 45027 + }, + { + "epoch": 0.45028, + "grad_norm": 1.1626010808597769, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 45028 + }, + { + "epoch": 0.45029, + "grad_norm": 1.4836129539258887, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 45029 + }, + { + "epoch": 0.4503, + "grad_norm": 0.7135386255398298, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 45030 + }, + { + "epoch": 0.45031, + "grad_norm": 0.6826271677047104, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 45031 + }, + { + "epoch": 0.45032, + "grad_norm": 0.8905133639842638, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 45032 + }, + { + "epoch": 0.45033, + "grad_norm": 1.134924175000244, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 45033 + }, + { + "epoch": 0.45034, + "grad_norm": 1.2863004670184648, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 45034 + }, + { + "epoch": 0.45035, + "grad_norm": 1.0092597079402856, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 45035 + }, + { + "epoch": 0.45036, + "grad_norm": 1.0986975474812597, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 45036 + }, + { + "epoch": 0.45037, + "grad_norm": 1.1530893177130352, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 45037 + }, + { + "epoch": 0.45038, + "grad_norm": 1.031998598088034, + "learning_rate": 0.003, + "loss": 4.004, + "step": 45038 + }, + { + "epoch": 0.45039, + "grad_norm": 1.1064041488717038, + "learning_rate": 0.003, + "loss": 4.049, + "step": 45039 + }, + { + "epoch": 0.4504, + "grad_norm": 1.293657479411942, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 45040 + }, + { + "epoch": 0.45041, + "grad_norm": 0.9577072673454382, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 45041 + }, + { + "epoch": 0.45042, + "grad_norm": 1.3721146495269552, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 45042 + }, + { + "epoch": 0.45043, + "grad_norm": 1.033121265981305, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 45043 + }, + { + "epoch": 0.45044, + "grad_norm": 0.8503758489808454, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 45044 + }, + { + "epoch": 0.45045, + "grad_norm": 1.024608813010871, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 45045 + }, + { + "epoch": 0.45046, + "grad_norm": 1.3336868717269312, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 45046 + }, + { + "epoch": 0.45047, + "grad_norm": 0.9186990684660908, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 45047 + }, + { + "epoch": 0.45048, + "grad_norm": 1.035706128734697, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 45048 + }, + { + "epoch": 0.45049, + "grad_norm": 1.0093126937199015, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 45049 + }, + { + "epoch": 0.4505, + "grad_norm": 1.108170574581438, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 45050 + }, + { + "epoch": 0.45051, + "grad_norm": 0.9571040579729609, + "learning_rate": 0.003, + "loss": 4.055, + "step": 45051 + }, + { + "epoch": 0.45052, + "grad_norm": 0.9816696561652725, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 45052 + }, + { + "epoch": 0.45053, + "grad_norm": 0.9815659514168451, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 45053 + }, + { + "epoch": 0.45054, + "grad_norm": 0.96946760031267, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 45054 + }, + { + "epoch": 0.45055, + "grad_norm": 1.2367924324443689, + "learning_rate": 0.003, + "loss": 4.06, + "step": 45055 + }, + { + "epoch": 0.45056, + "grad_norm": 0.9214307622507291, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 45056 + }, + { + "epoch": 0.45057, + "grad_norm": 0.9294716590222875, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 45057 + }, + { + "epoch": 0.45058, + "grad_norm": 1.1176462608271953, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 45058 + }, + { + "epoch": 0.45059, + "grad_norm": 1.109903908484556, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 45059 + }, + { + "epoch": 0.4506, + "grad_norm": 1.1172558432352153, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 45060 + }, + { + "epoch": 0.45061, + "grad_norm": 1.1446510568978978, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 45061 + }, + { + "epoch": 0.45062, + "grad_norm": 0.8479301903329304, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 45062 + }, + { + "epoch": 0.45063, + "grad_norm": 0.8280056873438041, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 45063 + }, + { + "epoch": 0.45064, + "grad_norm": 1.0462323433824603, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 45064 + }, + { + "epoch": 0.45065, + "grad_norm": 1.227095378664637, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 45065 + }, + { + "epoch": 0.45066, + "grad_norm": 1.0185902073397366, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 45066 + }, + { + "epoch": 0.45067, + "grad_norm": 1.2896750303725657, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 45067 + }, + { + "epoch": 0.45068, + "grad_norm": 0.9383960188912445, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 45068 + }, + { + "epoch": 0.45069, + "grad_norm": 0.9266795619370426, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 45069 + }, + { + "epoch": 0.4507, + "grad_norm": 1.0339465189533932, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 45070 + }, + { + "epoch": 0.45071, + "grad_norm": 1.2340203993112862, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 45071 + }, + { + "epoch": 0.45072, + "grad_norm": 0.8656436513419281, + "learning_rate": 0.003, + "loss": 4.03, + "step": 45072 + }, + { + "epoch": 0.45073, + "grad_norm": 0.8621842693700745, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 45073 + }, + { + "epoch": 0.45074, + "grad_norm": 0.8912132336166785, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 45074 + }, + { + "epoch": 0.45075, + "grad_norm": 0.9395706530492721, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 45075 + }, + { + "epoch": 0.45076, + "grad_norm": 0.9946809289840254, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 45076 + }, + { + "epoch": 0.45077, + "grad_norm": 1.0673391390680897, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 45077 + }, + { + "epoch": 0.45078, + "grad_norm": 0.9376860747312297, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 45078 + }, + { + "epoch": 0.45079, + "grad_norm": 1.0534654813042894, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 45079 + }, + { + "epoch": 0.4508, + "grad_norm": 1.1353212056773452, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 45080 + }, + { + "epoch": 0.45081, + "grad_norm": 1.0744096371312186, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 45081 + }, + { + "epoch": 0.45082, + "grad_norm": 1.4848161812847736, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 45082 + }, + { + "epoch": 0.45083, + "grad_norm": 0.8846040670998206, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 45083 + }, + { + "epoch": 0.45084, + "grad_norm": 1.1049741325255005, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 45084 + }, + { + "epoch": 0.45085, + "grad_norm": 1.3068383328116535, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 45085 + }, + { + "epoch": 0.45086, + "grad_norm": 1.0152200044615092, + "learning_rate": 0.003, + "loss": 4.027, + "step": 45086 + }, + { + "epoch": 0.45087, + "grad_norm": 1.2024392303154725, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 45087 + }, + { + "epoch": 0.45088, + "grad_norm": 1.3032020131513817, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 45088 + }, + { + "epoch": 0.45089, + "grad_norm": 0.8875270509669538, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 45089 + }, + { + "epoch": 0.4509, + "grad_norm": 0.7873549313572161, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 45090 + }, + { + "epoch": 0.45091, + "grad_norm": 0.8732083783022949, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 45091 + }, + { + "epoch": 0.45092, + "grad_norm": 1.0578362584068857, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 45092 + }, + { + "epoch": 0.45093, + "grad_norm": 0.9755289631371957, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 45093 + }, + { + "epoch": 0.45094, + "grad_norm": 0.9989994365926055, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 45094 + }, + { + "epoch": 0.45095, + "grad_norm": 1.1774774631959322, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 45095 + }, + { + "epoch": 0.45096, + "grad_norm": 0.9149604539825026, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 45096 + }, + { + "epoch": 0.45097, + "grad_norm": 0.9924580382150675, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 45097 + }, + { + "epoch": 0.45098, + "grad_norm": 1.4885124210922889, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 45098 + }, + { + "epoch": 0.45099, + "grad_norm": 1.203448584049763, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 45099 + }, + { + "epoch": 0.451, + "grad_norm": 0.9415999350171699, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 45100 + }, + { + "epoch": 0.45101, + "grad_norm": 0.9022868488488278, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 45101 + }, + { + "epoch": 0.45102, + "grad_norm": 1.1332563970361995, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 45102 + }, + { + "epoch": 0.45103, + "grad_norm": 1.0094901872457391, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 45103 + }, + { + "epoch": 0.45104, + "grad_norm": 0.9228380794700881, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 45104 + }, + { + "epoch": 0.45105, + "grad_norm": 0.9261862522973102, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 45105 + }, + { + "epoch": 0.45106, + "grad_norm": 0.9784210070989043, + "learning_rate": 0.003, + "loss": 4.032, + "step": 45106 + }, + { + "epoch": 0.45107, + "grad_norm": 1.3602579564447441, + "learning_rate": 0.003, + "loss": 4.063, + "step": 45107 + }, + { + "epoch": 0.45108, + "grad_norm": 1.305753720683656, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 45108 + }, + { + "epoch": 0.45109, + "grad_norm": 0.9445062240575707, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 45109 + }, + { + "epoch": 0.4511, + "grad_norm": 1.087529739486476, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 45110 + }, + { + "epoch": 0.45111, + "grad_norm": 0.9399811341048633, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 45111 + }, + { + "epoch": 0.45112, + "grad_norm": 1.0776146568213072, + "learning_rate": 0.003, + "loss": 4.047, + "step": 45112 + }, + { + "epoch": 0.45113, + "grad_norm": 1.199580111654328, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 45113 + }, + { + "epoch": 0.45114, + "grad_norm": 1.1274067317856777, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 45114 + }, + { + "epoch": 0.45115, + "grad_norm": 0.9346482715928998, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 45115 + }, + { + "epoch": 0.45116, + "grad_norm": 1.0581755217851587, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 45116 + }, + { + "epoch": 0.45117, + "grad_norm": 1.2870021140962469, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 45117 + }, + { + "epoch": 0.45118, + "grad_norm": 0.9985480314724451, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 45118 + }, + { + "epoch": 0.45119, + "grad_norm": 1.399425851328258, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 45119 + }, + { + "epoch": 0.4512, + "grad_norm": 1.0130855215294463, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 45120 + }, + { + "epoch": 0.45121, + "grad_norm": 1.0698533950694422, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 45121 + }, + { + "epoch": 0.45122, + "grad_norm": 0.9987813132376939, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 45122 + }, + { + "epoch": 0.45123, + "grad_norm": 1.4042755011351316, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 45123 + }, + { + "epoch": 0.45124, + "grad_norm": 0.8204239176438151, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 45124 + }, + { + "epoch": 0.45125, + "grad_norm": 0.8127370807363729, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 45125 + }, + { + "epoch": 0.45126, + "grad_norm": 0.8368844591657578, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 45126 + }, + { + "epoch": 0.45127, + "grad_norm": 0.9863688681293883, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 45127 + }, + { + "epoch": 0.45128, + "grad_norm": 1.4544301301092992, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 45128 + }, + { + "epoch": 0.45129, + "grad_norm": 0.8630367854738238, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 45129 + }, + { + "epoch": 0.4513, + "grad_norm": 1.0097466449739594, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 45130 + }, + { + "epoch": 0.45131, + "grad_norm": 1.0776459556528541, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 45131 + }, + { + "epoch": 0.45132, + "grad_norm": 1.160390501314177, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 45132 + }, + { + "epoch": 0.45133, + "grad_norm": 1.2050436030668248, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 45133 + }, + { + "epoch": 0.45134, + "grad_norm": 0.9587018535090848, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 45134 + }, + { + "epoch": 0.45135, + "grad_norm": 1.0980030610578142, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 45135 + }, + { + "epoch": 0.45136, + "grad_norm": 0.933416404131801, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 45136 + }, + { + "epoch": 0.45137, + "grad_norm": 1.0106083035842317, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 45137 + }, + { + "epoch": 0.45138, + "grad_norm": 1.079653563830709, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 45138 + }, + { + "epoch": 0.45139, + "grad_norm": 0.9899292876516144, + "learning_rate": 0.003, + "loss": 4.037, + "step": 45139 + }, + { + "epoch": 0.4514, + "grad_norm": 1.2617086409641511, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 45140 + }, + { + "epoch": 0.45141, + "grad_norm": 0.9314491450071094, + "learning_rate": 0.003, + "loss": 3.991, + "step": 45141 + }, + { + "epoch": 0.45142, + "grad_norm": 1.112106277981932, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 45142 + }, + { + "epoch": 0.45143, + "grad_norm": 1.2719878551107504, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 45143 + }, + { + "epoch": 0.45144, + "grad_norm": 0.9091703230629609, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 45144 + }, + { + "epoch": 0.45145, + "grad_norm": 0.9459706866283076, + "learning_rate": 0.003, + "loss": 4.031, + "step": 45145 + }, + { + "epoch": 0.45146, + "grad_norm": 1.1130573139355455, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 45146 + }, + { + "epoch": 0.45147, + "grad_norm": 1.1896406972991231, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 45147 + }, + { + "epoch": 0.45148, + "grad_norm": 1.0792065450052244, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 45148 + }, + { + "epoch": 0.45149, + "grad_norm": 1.089613220825506, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 45149 + }, + { + "epoch": 0.4515, + "grad_norm": 1.020474485701699, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 45150 + }, + { + "epoch": 0.45151, + "grad_norm": 1.1552835842782485, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 45151 + }, + { + "epoch": 0.45152, + "grad_norm": 0.9833131266898845, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 45152 + }, + { + "epoch": 0.45153, + "grad_norm": 1.0289872356735101, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 45153 + }, + { + "epoch": 0.45154, + "grad_norm": 1.1571427539936103, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 45154 + }, + { + "epoch": 0.45155, + "grad_norm": 1.0221764822682022, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 45155 + }, + { + "epoch": 0.45156, + "grad_norm": 1.1532837306941182, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 45156 + }, + { + "epoch": 0.45157, + "grad_norm": 1.0072760423382305, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 45157 + }, + { + "epoch": 0.45158, + "grad_norm": 1.290437349861052, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 45158 + }, + { + "epoch": 0.45159, + "grad_norm": 1.0154892580491113, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 45159 + }, + { + "epoch": 0.4516, + "grad_norm": 1.1939447817579243, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 45160 + }, + { + "epoch": 0.45161, + "grad_norm": 1.1002406646469414, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 45161 + }, + { + "epoch": 0.45162, + "grad_norm": 1.0634935275719333, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 45162 + }, + { + "epoch": 0.45163, + "grad_norm": 0.9840003807483003, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 45163 + }, + { + "epoch": 0.45164, + "grad_norm": 1.1339844060879691, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 45164 + }, + { + "epoch": 0.45165, + "grad_norm": 1.0736246910225182, + "learning_rate": 0.003, + "loss": 4.037, + "step": 45165 + }, + { + "epoch": 0.45166, + "grad_norm": 1.0313295875010338, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 45166 + }, + { + "epoch": 0.45167, + "grad_norm": 0.8979628923799892, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 45167 + }, + { + "epoch": 0.45168, + "grad_norm": 0.9268873683815725, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 45168 + }, + { + "epoch": 0.45169, + "grad_norm": 1.0748845245528447, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 45169 + }, + { + "epoch": 0.4517, + "grad_norm": 1.1263443619355138, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 45170 + }, + { + "epoch": 0.45171, + "grad_norm": 0.9192632775387212, + "learning_rate": 0.003, + "loss": 4.056, + "step": 45171 + }, + { + "epoch": 0.45172, + "grad_norm": 1.0317670207535927, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 45172 + }, + { + "epoch": 0.45173, + "grad_norm": 1.2019855568702333, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 45173 + }, + { + "epoch": 0.45174, + "grad_norm": 0.9009830115058902, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 45174 + }, + { + "epoch": 0.45175, + "grad_norm": 1.0170687446107682, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 45175 + }, + { + "epoch": 0.45176, + "grad_norm": 1.0763251976635113, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 45176 + }, + { + "epoch": 0.45177, + "grad_norm": 0.9303876139762087, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 45177 + }, + { + "epoch": 0.45178, + "grad_norm": 1.0214383300643817, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 45178 + }, + { + "epoch": 0.45179, + "grad_norm": 1.1201645181098745, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 45179 + }, + { + "epoch": 0.4518, + "grad_norm": 1.1459664994549197, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 45180 + }, + { + "epoch": 0.45181, + "grad_norm": 1.2878463092159718, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 45181 + }, + { + "epoch": 0.45182, + "grad_norm": 0.9619603907896677, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 45182 + }, + { + "epoch": 0.45183, + "grad_norm": 1.0935910896408676, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 45183 + }, + { + "epoch": 0.45184, + "grad_norm": 1.1893229816838216, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 45184 + }, + { + "epoch": 0.45185, + "grad_norm": 0.9488580066033744, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 45185 + }, + { + "epoch": 0.45186, + "grad_norm": 0.8964996667411004, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 45186 + }, + { + "epoch": 0.45187, + "grad_norm": 1.051764692395411, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 45187 + }, + { + "epoch": 0.45188, + "grad_norm": 1.1273509375975213, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 45188 + }, + { + "epoch": 0.45189, + "grad_norm": 0.8365031843516629, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 45189 + }, + { + "epoch": 0.4519, + "grad_norm": 0.8452138435430552, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 45190 + }, + { + "epoch": 0.45191, + "grad_norm": 0.9669073062097442, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 45191 + }, + { + "epoch": 0.45192, + "grad_norm": 1.1093110202352667, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 45192 + }, + { + "epoch": 0.45193, + "grad_norm": 1.0728109656733675, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 45193 + }, + { + "epoch": 0.45194, + "grad_norm": 1.1410461534176886, + "learning_rate": 0.003, + "loss": 4.027, + "step": 45194 + }, + { + "epoch": 0.45195, + "grad_norm": 1.2942389809766865, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 45195 + }, + { + "epoch": 0.45196, + "grad_norm": 1.0524793809397355, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 45196 + }, + { + "epoch": 0.45197, + "grad_norm": 1.1494923841086362, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 45197 + }, + { + "epoch": 0.45198, + "grad_norm": 0.9331164879028486, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 45198 + }, + { + "epoch": 0.45199, + "grad_norm": 1.0513815623914842, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 45199 + }, + { + "epoch": 0.452, + "grad_norm": 1.2731067417721957, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 45200 + }, + { + "epoch": 0.45201, + "grad_norm": 0.790558710913539, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 45201 + }, + { + "epoch": 0.45202, + "grad_norm": 0.7847823088323609, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 45202 + }, + { + "epoch": 0.45203, + "grad_norm": 0.9202111270901864, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 45203 + }, + { + "epoch": 0.45204, + "grad_norm": 1.1347611451747253, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 45204 + }, + { + "epoch": 0.45205, + "grad_norm": 1.0465553704591086, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 45205 + }, + { + "epoch": 0.45206, + "grad_norm": 1.0773029685646986, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 45206 + }, + { + "epoch": 0.45207, + "grad_norm": 1.0365134035025936, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 45207 + }, + { + "epoch": 0.45208, + "grad_norm": 1.020033576971651, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 45208 + }, + { + "epoch": 0.45209, + "grad_norm": 1.2631079314836362, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 45209 + }, + { + "epoch": 0.4521, + "grad_norm": 1.1426751777707351, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 45210 + }, + { + "epoch": 0.45211, + "grad_norm": 1.1879199870362815, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 45211 + }, + { + "epoch": 0.45212, + "grad_norm": 1.0169022927407658, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 45212 + }, + { + "epoch": 0.45213, + "grad_norm": 1.0940230697763085, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 45213 + }, + { + "epoch": 0.45214, + "grad_norm": 1.2247785069587185, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 45214 + }, + { + "epoch": 0.45215, + "grad_norm": 0.8574477882598062, + "learning_rate": 0.003, + "loss": 4.033, + "step": 45215 + }, + { + "epoch": 0.45216, + "grad_norm": 0.8795046215320529, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 45216 + }, + { + "epoch": 0.45217, + "grad_norm": 1.0439624204098426, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 45217 + }, + { + "epoch": 0.45218, + "grad_norm": 1.1507708272871389, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 45218 + }, + { + "epoch": 0.45219, + "grad_norm": 1.003205540883924, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 45219 + }, + { + "epoch": 0.4522, + "grad_norm": 1.1509361530389373, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 45220 + }, + { + "epoch": 0.45221, + "grad_norm": 1.0763986798821266, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 45221 + }, + { + "epoch": 0.45222, + "grad_norm": 1.2764168459697074, + "learning_rate": 0.003, + "loss": 4.049, + "step": 45222 + }, + { + "epoch": 0.45223, + "grad_norm": 0.9971168816758702, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 45223 + }, + { + "epoch": 0.45224, + "grad_norm": 0.9866587926972225, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 45224 + }, + { + "epoch": 0.45225, + "grad_norm": 0.9613443595757908, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 45225 + }, + { + "epoch": 0.45226, + "grad_norm": 1.1686339244891506, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 45226 + }, + { + "epoch": 0.45227, + "grad_norm": 0.9999259638622968, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 45227 + }, + { + "epoch": 0.45228, + "grad_norm": 1.066830827545876, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 45228 + }, + { + "epoch": 0.45229, + "grad_norm": 1.1250030435506233, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 45229 + }, + { + "epoch": 0.4523, + "grad_norm": 0.9311048164353679, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 45230 + }, + { + "epoch": 0.45231, + "grad_norm": 1.1400496686619332, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 45231 + }, + { + "epoch": 0.45232, + "grad_norm": 0.9479957309658729, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 45232 + }, + { + "epoch": 0.45233, + "grad_norm": 1.2161561687304345, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 45233 + }, + { + "epoch": 0.45234, + "grad_norm": 1.114659595769527, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 45234 + }, + { + "epoch": 0.45235, + "grad_norm": 1.119885337758001, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 45235 + }, + { + "epoch": 0.45236, + "grad_norm": 1.403532262440014, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 45236 + }, + { + "epoch": 0.45237, + "grad_norm": 0.9206014541151409, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 45237 + }, + { + "epoch": 0.45238, + "grad_norm": 1.1142006767970862, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 45238 + }, + { + "epoch": 0.45239, + "grad_norm": 1.284401523278454, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 45239 + }, + { + "epoch": 0.4524, + "grad_norm": 0.908056917719069, + "learning_rate": 0.003, + "loss": 4.042, + "step": 45240 + }, + { + "epoch": 0.45241, + "grad_norm": 0.890821277757702, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 45241 + }, + { + "epoch": 0.45242, + "grad_norm": 1.0022834648003418, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 45242 + }, + { + "epoch": 0.45243, + "grad_norm": 1.4121705242681235, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 45243 + }, + { + "epoch": 0.45244, + "grad_norm": 0.8103760194938083, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 45244 + }, + { + "epoch": 0.45245, + "grad_norm": 0.9635199045995202, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 45245 + }, + { + "epoch": 0.45246, + "grad_norm": 1.16620777864411, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 45246 + }, + { + "epoch": 0.45247, + "grad_norm": 1.2605255972648857, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 45247 + }, + { + "epoch": 0.45248, + "grad_norm": 1.1345720094891547, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 45248 + }, + { + "epoch": 0.45249, + "grad_norm": 0.9675766636400385, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 45249 + }, + { + "epoch": 0.4525, + "grad_norm": 1.0471234683609536, + "learning_rate": 0.003, + "loss": 4.052, + "step": 45250 + }, + { + "epoch": 0.45251, + "grad_norm": 1.0470745749175774, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 45251 + }, + { + "epoch": 0.45252, + "grad_norm": 0.9946995526865805, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 45252 + }, + { + "epoch": 0.45253, + "grad_norm": 1.0184188656996735, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 45253 + }, + { + "epoch": 0.45254, + "grad_norm": 1.1822044276509707, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 45254 + }, + { + "epoch": 0.45255, + "grad_norm": 1.127021724880334, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 45255 + }, + { + "epoch": 0.45256, + "grad_norm": 0.9941958949873803, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 45256 + }, + { + "epoch": 0.45257, + "grad_norm": 0.9488264976439483, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 45257 + }, + { + "epoch": 0.45258, + "grad_norm": 0.99142880439294, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 45258 + }, + { + "epoch": 0.45259, + "grad_norm": 0.9765310280167915, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 45259 + }, + { + "epoch": 0.4526, + "grad_norm": 1.1815248346377138, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 45260 + }, + { + "epoch": 0.45261, + "grad_norm": 1.16452434373928, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 45261 + }, + { + "epoch": 0.45262, + "grad_norm": 1.09404481487077, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 45262 + }, + { + "epoch": 0.45263, + "grad_norm": 1.0087380639167753, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 45263 + }, + { + "epoch": 0.45264, + "grad_norm": 1.3368614165115251, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 45264 + }, + { + "epoch": 0.45265, + "grad_norm": 0.784033786226818, + "learning_rate": 0.003, + "loss": 4.021, + "step": 45265 + }, + { + "epoch": 0.45266, + "grad_norm": 0.7948343645855142, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 45266 + }, + { + "epoch": 0.45267, + "grad_norm": 0.8427003828624416, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 45267 + }, + { + "epoch": 0.45268, + "grad_norm": 0.7864130217956652, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 45268 + }, + { + "epoch": 0.45269, + "grad_norm": 0.8262726914781803, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 45269 + }, + { + "epoch": 0.4527, + "grad_norm": 0.920866585923303, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 45270 + }, + { + "epoch": 0.45271, + "grad_norm": 1.095769013008382, + "learning_rate": 0.003, + "loss": 3.999, + "step": 45271 + }, + { + "epoch": 0.45272, + "grad_norm": 1.1494261453404464, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 45272 + }, + { + "epoch": 0.45273, + "grad_norm": 0.9968127967291269, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 45273 + }, + { + "epoch": 0.45274, + "grad_norm": 1.469438881974641, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 45274 + }, + { + "epoch": 0.45275, + "grad_norm": 0.9839312761915127, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 45275 + }, + { + "epoch": 0.45276, + "grad_norm": 1.2101140823657222, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 45276 + }, + { + "epoch": 0.45277, + "grad_norm": 0.9057531395610209, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 45277 + }, + { + "epoch": 0.45278, + "grad_norm": 0.9004713450017651, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 45278 + }, + { + "epoch": 0.45279, + "grad_norm": 1.240594194972776, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 45279 + }, + { + "epoch": 0.4528, + "grad_norm": 0.9900420467053569, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 45280 + }, + { + "epoch": 0.45281, + "grad_norm": 0.9052820403360505, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 45281 + }, + { + "epoch": 0.45282, + "grad_norm": 0.8221503989011519, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 45282 + }, + { + "epoch": 0.45283, + "grad_norm": 0.923095363001211, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 45283 + }, + { + "epoch": 0.45284, + "grad_norm": 1.1379479376264938, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 45284 + }, + { + "epoch": 0.45285, + "grad_norm": 0.972658309946476, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 45285 + }, + { + "epoch": 0.45286, + "grad_norm": 1.0048095831086004, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 45286 + }, + { + "epoch": 0.45287, + "grad_norm": 1.303133245902698, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 45287 + }, + { + "epoch": 0.45288, + "grad_norm": 1.1536400893443701, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 45288 + }, + { + "epoch": 0.45289, + "grad_norm": 1.0171185158857567, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 45289 + }, + { + "epoch": 0.4529, + "grad_norm": 1.2979538633350034, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 45290 + }, + { + "epoch": 0.45291, + "grad_norm": 1.3003717253679206, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 45291 + }, + { + "epoch": 0.45292, + "grad_norm": 0.9228288062070491, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 45292 + }, + { + "epoch": 0.45293, + "grad_norm": 0.9387409162573558, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 45293 + }, + { + "epoch": 0.45294, + "grad_norm": 1.3285650957769999, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 45294 + }, + { + "epoch": 0.45295, + "grad_norm": 1.136096298252301, + "learning_rate": 0.003, + "loss": 4.025, + "step": 45295 + }, + { + "epoch": 0.45296, + "grad_norm": 0.9435656830910315, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 45296 + }, + { + "epoch": 0.45297, + "grad_norm": 1.1164671992651984, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 45297 + }, + { + "epoch": 0.45298, + "grad_norm": 1.3919537077418462, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 45298 + }, + { + "epoch": 0.45299, + "grad_norm": 0.996132899242782, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 45299 + }, + { + "epoch": 0.453, + "grad_norm": 1.32393056904914, + "learning_rate": 0.003, + "loss": 4.063, + "step": 45300 + }, + { + "epoch": 0.45301, + "grad_norm": 0.9000193983694668, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 45301 + }, + { + "epoch": 0.45302, + "grad_norm": 0.9599506702530974, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 45302 + }, + { + "epoch": 0.45303, + "grad_norm": 1.1118109740044777, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 45303 + }, + { + "epoch": 0.45304, + "grad_norm": 0.9451323783268616, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 45304 + }, + { + "epoch": 0.45305, + "grad_norm": 1.120369972421738, + "learning_rate": 0.003, + "loss": 4.052, + "step": 45305 + }, + { + "epoch": 0.45306, + "grad_norm": 1.096254761714889, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 45306 + }, + { + "epoch": 0.45307, + "grad_norm": 1.0474594011263176, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 45307 + }, + { + "epoch": 0.45308, + "grad_norm": 1.1402477528562236, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 45308 + }, + { + "epoch": 0.45309, + "grad_norm": 1.1096848106684873, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 45309 + }, + { + "epoch": 0.4531, + "grad_norm": 1.1642905398149044, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 45310 + }, + { + "epoch": 0.45311, + "grad_norm": 1.2299068198665337, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 45311 + }, + { + "epoch": 0.45312, + "grad_norm": 0.9676475662327694, + "learning_rate": 0.003, + "loss": 4.028, + "step": 45312 + }, + { + "epoch": 0.45313, + "grad_norm": 1.1318972796117863, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 45313 + }, + { + "epoch": 0.45314, + "grad_norm": 1.0090466236041604, + "learning_rate": 0.003, + "loss": 4.028, + "step": 45314 + }, + { + "epoch": 0.45315, + "grad_norm": 1.2044786032943129, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 45315 + }, + { + "epoch": 0.45316, + "grad_norm": 0.9570406838539651, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 45316 + }, + { + "epoch": 0.45317, + "grad_norm": 1.1651772408733463, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 45317 + }, + { + "epoch": 0.45318, + "grad_norm": 0.9650113440026773, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 45318 + }, + { + "epoch": 0.45319, + "grad_norm": 1.0512040130857379, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 45319 + }, + { + "epoch": 0.4532, + "grad_norm": 1.265082627803366, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 45320 + }, + { + "epoch": 0.45321, + "grad_norm": 0.8999725441123968, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 45321 + }, + { + "epoch": 0.45322, + "grad_norm": 0.8068687449963476, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 45322 + }, + { + "epoch": 0.45323, + "grad_norm": 0.9218760925753432, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 45323 + }, + { + "epoch": 0.45324, + "grad_norm": 1.1762074271401002, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 45324 + }, + { + "epoch": 0.45325, + "grad_norm": 0.8690221546878164, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 45325 + }, + { + "epoch": 0.45326, + "grad_norm": 0.9231702764094725, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 45326 + }, + { + "epoch": 0.45327, + "grad_norm": 1.3095943927229499, + "learning_rate": 0.003, + "loss": 4.029, + "step": 45327 + }, + { + "epoch": 0.45328, + "grad_norm": 1.1138249900684627, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 45328 + }, + { + "epoch": 0.45329, + "grad_norm": 1.116586871967308, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 45329 + }, + { + "epoch": 0.4533, + "grad_norm": 1.0325893214709687, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 45330 + }, + { + "epoch": 0.45331, + "grad_norm": 1.042135837193755, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 45331 + }, + { + "epoch": 0.45332, + "grad_norm": 1.0350706132781455, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 45332 + }, + { + "epoch": 0.45333, + "grad_norm": 1.0974437404257242, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 45333 + }, + { + "epoch": 0.45334, + "grad_norm": 0.9320666455615689, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 45334 + }, + { + "epoch": 0.45335, + "grad_norm": 0.8721637828201338, + "learning_rate": 0.003, + "loss": 4.023, + "step": 45335 + }, + { + "epoch": 0.45336, + "grad_norm": 1.0262832709132779, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 45336 + }, + { + "epoch": 0.45337, + "grad_norm": 1.2454131648399944, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 45337 + }, + { + "epoch": 0.45338, + "grad_norm": 0.894397541937679, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 45338 + }, + { + "epoch": 0.45339, + "grad_norm": 1.1402653727477745, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 45339 + }, + { + "epoch": 0.4534, + "grad_norm": 1.313494107835662, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 45340 + }, + { + "epoch": 0.45341, + "grad_norm": 0.614791772002659, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 45341 + }, + { + "epoch": 0.45342, + "grad_norm": 0.9189795547846458, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 45342 + }, + { + "epoch": 0.45343, + "grad_norm": 1.149613631589156, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 45343 + }, + { + "epoch": 0.45344, + "grad_norm": 1.0052062362461962, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 45344 + }, + { + "epoch": 0.45345, + "grad_norm": 1.1535321996085128, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 45345 + }, + { + "epoch": 0.45346, + "grad_norm": 1.2563692869373573, + "learning_rate": 0.003, + "loss": 4.034, + "step": 45346 + }, + { + "epoch": 0.45347, + "grad_norm": 0.8739106087690489, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 45347 + }, + { + "epoch": 0.45348, + "grad_norm": 0.8775048436024935, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 45348 + }, + { + "epoch": 0.45349, + "grad_norm": 1.0787755755710824, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 45349 + }, + { + "epoch": 0.4535, + "grad_norm": 1.1427632931121985, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 45350 + }, + { + "epoch": 0.45351, + "grad_norm": 0.9114150650150762, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 45351 + }, + { + "epoch": 0.45352, + "grad_norm": 1.1537932382445721, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 45352 + }, + { + "epoch": 0.45353, + "grad_norm": 1.05018305141726, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 45353 + }, + { + "epoch": 0.45354, + "grad_norm": 1.0559684561311926, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 45354 + }, + { + "epoch": 0.45355, + "grad_norm": 1.1522956021284034, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 45355 + }, + { + "epoch": 0.45356, + "grad_norm": 1.1666896221921945, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 45356 + }, + { + "epoch": 0.45357, + "grad_norm": 0.857912106895958, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 45357 + }, + { + "epoch": 0.45358, + "grad_norm": 0.8662432478587436, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 45358 + }, + { + "epoch": 0.45359, + "grad_norm": 0.9708473543391264, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 45359 + }, + { + "epoch": 0.4536, + "grad_norm": 1.2392649624147434, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 45360 + }, + { + "epoch": 0.45361, + "grad_norm": 0.9377282850615613, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 45361 + }, + { + "epoch": 0.45362, + "grad_norm": 1.2903468930610036, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 45362 + }, + { + "epoch": 0.45363, + "grad_norm": 0.9526580936244707, + "learning_rate": 0.003, + "loss": 4.024, + "step": 45363 + }, + { + "epoch": 0.45364, + "grad_norm": 1.198762201064503, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 45364 + }, + { + "epoch": 0.45365, + "grad_norm": 0.9804847094506016, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 45365 + }, + { + "epoch": 0.45366, + "grad_norm": 1.3447557203900282, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 45366 + }, + { + "epoch": 0.45367, + "grad_norm": 1.1016052392759765, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 45367 + }, + { + "epoch": 0.45368, + "grad_norm": 0.8504044046829261, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 45368 + }, + { + "epoch": 0.45369, + "grad_norm": 1.07268657540944, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 45369 + }, + { + "epoch": 0.4537, + "grad_norm": 1.3922334588823284, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 45370 + }, + { + "epoch": 0.45371, + "grad_norm": 1.3775989379272944, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 45371 + }, + { + "epoch": 0.45372, + "grad_norm": 1.011750395904183, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 45372 + }, + { + "epoch": 0.45373, + "grad_norm": 0.8958680916208532, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 45373 + }, + { + "epoch": 0.45374, + "grad_norm": 1.1755539393012058, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 45374 + }, + { + "epoch": 0.45375, + "grad_norm": 1.1956730931632273, + "learning_rate": 0.003, + "loss": 4.024, + "step": 45375 + }, + { + "epoch": 0.45376, + "grad_norm": 0.8048432288558075, + "learning_rate": 0.003, + "loss": 4.024, + "step": 45376 + }, + { + "epoch": 0.45377, + "grad_norm": 0.9180683665401959, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 45377 + }, + { + "epoch": 0.45378, + "grad_norm": 0.9427641721035446, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 45378 + }, + { + "epoch": 0.45379, + "grad_norm": 0.9750914442693862, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 45379 + }, + { + "epoch": 0.4538, + "grad_norm": 1.0521522756155222, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 45380 + }, + { + "epoch": 0.45381, + "grad_norm": 1.2045835465867227, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 45381 + }, + { + "epoch": 0.45382, + "grad_norm": 1.1286045192222378, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 45382 + }, + { + "epoch": 0.45383, + "grad_norm": 1.0849585841241731, + "learning_rate": 0.003, + "loss": 3.995, + "step": 45383 + }, + { + "epoch": 0.45384, + "grad_norm": 1.071274039504891, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 45384 + }, + { + "epoch": 0.45385, + "grad_norm": 1.2044896379548342, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 45385 + }, + { + "epoch": 0.45386, + "grad_norm": 1.0176553321328037, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 45386 + }, + { + "epoch": 0.45387, + "grad_norm": 1.030614354213593, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 45387 + }, + { + "epoch": 0.45388, + "grad_norm": 0.9875468723697058, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 45388 + }, + { + "epoch": 0.45389, + "grad_norm": 1.0064299985247576, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 45389 + }, + { + "epoch": 0.4539, + "grad_norm": 1.2820569282903223, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 45390 + }, + { + "epoch": 0.45391, + "grad_norm": 0.8972508583111344, + "learning_rate": 0.003, + "loss": 4.052, + "step": 45391 + }, + { + "epoch": 0.45392, + "grad_norm": 0.9335099097423782, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 45392 + }, + { + "epoch": 0.45393, + "grad_norm": 1.1902107254278653, + "learning_rate": 0.003, + "loss": 4.062, + "step": 45393 + }, + { + "epoch": 0.45394, + "grad_norm": 1.3209719913870952, + "learning_rate": 0.003, + "loss": 4.055, + "step": 45394 + }, + { + "epoch": 0.45395, + "grad_norm": 0.8794960690700507, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 45395 + }, + { + "epoch": 0.45396, + "grad_norm": 0.7789532557424602, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 45396 + }, + { + "epoch": 0.45397, + "grad_norm": 1.0007233702132592, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 45397 + }, + { + "epoch": 0.45398, + "grad_norm": 1.3066139439106437, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 45398 + }, + { + "epoch": 0.45399, + "grad_norm": 0.9032033570089977, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 45399 + }, + { + "epoch": 0.454, + "grad_norm": 0.9810040314520817, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 45400 + }, + { + "epoch": 0.45401, + "grad_norm": 1.3642583455453248, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 45401 + }, + { + "epoch": 0.45402, + "grad_norm": 1.293934139438529, + "learning_rate": 0.003, + "loss": 4.034, + "step": 45402 + }, + { + "epoch": 0.45403, + "grad_norm": 0.8240924044032426, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 45403 + }, + { + "epoch": 0.45404, + "grad_norm": 0.8173902036890027, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 45404 + }, + { + "epoch": 0.45405, + "grad_norm": 1.0184950361093679, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 45405 + }, + { + "epoch": 0.45406, + "grad_norm": 1.295495660802748, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 45406 + }, + { + "epoch": 0.45407, + "grad_norm": 0.7857245761848577, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 45407 + }, + { + "epoch": 0.45408, + "grad_norm": 0.9229367223628694, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 45408 + }, + { + "epoch": 0.45409, + "grad_norm": 1.407590848008784, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 45409 + }, + { + "epoch": 0.4541, + "grad_norm": 0.9916231854329384, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 45410 + }, + { + "epoch": 0.45411, + "grad_norm": 1.0964204618855375, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 45411 + }, + { + "epoch": 0.45412, + "grad_norm": 1.0237506513042067, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 45412 + }, + { + "epoch": 0.45413, + "grad_norm": 0.9634666918646051, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 45413 + }, + { + "epoch": 0.45414, + "grad_norm": 0.890644517125729, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 45414 + }, + { + "epoch": 0.45415, + "grad_norm": 1.0480319891487428, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 45415 + }, + { + "epoch": 0.45416, + "grad_norm": 1.0337589421467057, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 45416 + }, + { + "epoch": 0.45417, + "grad_norm": 1.2163280509740106, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 45417 + }, + { + "epoch": 0.45418, + "grad_norm": 1.0164269772099084, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 45418 + }, + { + "epoch": 0.45419, + "grad_norm": 1.1279602490476264, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 45419 + }, + { + "epoch": 0.4542, + "grad_norm": 1.0496887724870974, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 45420 + }, + { + "epoch": 0.45421, + "grad_norm": 1.1489078144384852, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 45421 + }, + { + "epoch": 0.45422, + "grad_norm": 1.1886343610442351, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 45422 + }, + { + "epoch": 0.45423, + "grad_norm": 1.0148642790979885, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 45423 + }, + { + "epoch": 0.45424, + "grad_norm": 1.204895170680873, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 45424 + }, + { + "epoch": 0.45425, + "grad_norm": 0.9127088347361177, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 45425 + }, + { + "epoch": 0.45426, + "grad_norm": 1.0400489193428366, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 45426 + }, + { + "epoch": 0.45427, + "grad_norm": 1.2924162550985157, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 45427 + }, + { + "epoch": 0.45428, + "grad_norm": 0.8143426176998583, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 45428 + }, + { + "epoch": 0.45429, + "grad_norm": 0.856013144110507, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 45429 + }, + { + "epoch": 0.4543, + "grad_norm": 1.191994148038468, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 45430 + }, + { + "epoch": 0.45431, + "grad_norm": 1.0561218457876722, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 45431 + }, + { + "epoch": 0.45432, + "grad_norm": 0.9757079502276041, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 45432 + }, + { + "epoch": 0.45433, + "grad_norm": 1.1041917071413516, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 45433 + }, + { + "epoch": 0.45434, + "grad_norm": 1.0024379771707927, + "learning_rate": 0.003, + "loss": 4.013, + "step": 45434 + }, + { + "epoch": 0.45435, + "grad_norm": 0.9979292219479663, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 45435 + }, + { + "epoch": 0.45436, + "grad_norm": 1.1831218389412133, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 45436 + }, + { + "epoch": 0.45437, + "grad_norm": 0.9011617967461146, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 45437 + }, + { + "epoch": 0.45438, + "grad_norm": 0.8597839618075573, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 45438 + }, + { + "epoch": 0.45439, + "grad_norm": 0.8715879056716375, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 45439 + }, + { + "epoch": 0.4544, + "grad_norm": 0.9327238363941995, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 45440 + }, + { + "epoch": 0.45441, + "grad_norm": 1.2816541765790086, + "learning_rate": 0.003, + "loss": 4.034, + "step": 45441 + }, + { + "epoch": 0.45442, + "grad_norm": 1.1748807479193162, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 45442 + }, + { + "epoch": 0.45443, + "grad_norm": 0.9786747809760403, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 45443 + }, + { + "epoch": 0.45444, + "grad_norm": 1.0775946457183967, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 45444 + }, + { + "epoch": 0.45445, + "grad_norm": 1.215750766650053, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 45445 + }, + { + "epoch": 0.45446, + "grad_norm": 1.0526740404315114, + "learning_rate": 0.003, + "loss": 4.04, + "step": 45446 + }, + { + "epoch": 0.45447, + "grad_norm": 1.10302113168332, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 45447 + }, + { + "epoch": 0.45448, + "grad_norm": 0.9979337268988857, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 45448 + }, + { + "epoch": 0.45449, + "grad_norm": 1.0214195328477107, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 45449 + }, + { + "epoch": 0.4545, + "grad_norm": 1.0636240380198305, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 45450 + }, + { + "epoch": 0.45451, + "grad_norm": 1.0103967598515153, + "learning_rate": 0.003, + "loss": 4.041, + "step": 45451 + }, + { + "epoch": 0.45452, + "grad_norm": 0.9636684945473989, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 45452 + }, + { + "epoch": 0.45453, + "grad_norm": 1.1357943809213558, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 45453 + }, + { + "epoch": 0.45454, + "grad_norm": 0.9034612719500063, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 45454 + }, + { + "epoch": 0.45455, + "grad_norm": 0.8970549444150087, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 45455 + }, + { + "epoch": 0.45456, + "grad_norm": 0.8875001208514135, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 45456 + }, + { + "epoch": 0.45457, + "grad_norm": 1.1588134570868656, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 45457 + }, + { + "epoch": 0.45458, + "grad_norm": 1.0015649352613687, + "learning_rate": 0.003, + "loss": 4.047, + "step": 45458 + }, + { + "epoch": 0.45459, + "grad_norm": 1.105733369696113, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 45459 + }, + { + "epoch": 0.4546, + "grad_norm": 1.267023210779307, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 45460 + }, + { + "epoch": 0.45461, + "grad_norm": 1.0949793781055464, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 45461 + }, + { + "epoch": 0.45462, + "grad_norm": 1.2392119806198678, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 45462 + }, + { + "epoch": 0.45463, + "grad_norm": 0.9959540733005465, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 45463 + }, + { + "epoch": 0.45464, + "grad_norm": 1.0067381431321742, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 45464 + }, + { + "epoch": 0.45465, + "grad_norm": 1.0343227490234859, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 45465 + }, + { + "epoch": 0.45466, + "grad_norm": 0.9455984832321532, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 45466 + }, + { + "epoch": 0.45467, + "grad_norm": 1.2485362245441471, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 45467 + }, + { + "epoch": 0.45468, + "grad_norm": 1.017241020325741, + "learning_rate": 0.003, + "loss": 4.044, + "step": 45468 + }, + { + "epoch": 0.45469, + "grad_norm": 1.1831872844165066, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 45469 + }, + { + "epoch": 0.4547, + "grad_norm": 0.9133787020153316, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 45470 + }, + { + "epoch": 0.45471, + "grad_norm": 0.9778034115797584, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 45471 + }, + { + "epoch": 0.45472, + "grad_norm": 1.4150649028854128, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 45472 + }, + { + "epoch": 0.45473, + "grad_norm": 1.2250965347891514, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 45473 + }, + { + "epoch": 0.45474, + "grad_norm": 0.8836144548848184, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 45474 + }, + { + "epoch": 0.45475, + "grad_norm": 0.8062461817474432, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 45475 + }, + { + "epoch": 0.45476, + "grad_norm": 0.8501906256532943, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 45476 + }, + { + "epoch": 0.45477, + "grad_norm": 0.9050165659158864, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 45477 + }, + { + "epoch": 0.45478, + "grad_norm": 1.150434706080054, + "learning_rate": 0.003, + "loss": 4.045, + "step": 45478 + }, + { + "epoch": 0.45479, + "grad_norm": 1.319342136688368, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 45479 + }, + { + "epoch": 0.4548, + "grad_norm": 0.8826487218064517, + "learning_rate": 0.003, + "loss": 4.029, + "step": 45480 + }, + { + "epoch": 0.45481, + "grad_norm": 0.9309848566804885, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 45481 + }, + { + "epoch": 0.45482, + "grad_norm": 0.9804987742558317, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 45482 + }, + { + "epoch": 0.45483, + "grad_norm": 1.207440400658715, + "learning_rate": 0.003, + "loss": 4.035, + "step": 45483 + }, + { + "epoch": 0.45484, + "grad_norm": 0.9784311032061043, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 45484 + }, + { + "epoch": 0.45485, + "grad_norm": 1.0470844119855403, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 45485 + }, + { + "epoch": 0.45486, + "grad_norm": 1.2192676572094472, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 45486 + }, + { + "epoch": 0.45487, + "grad_norm": 1.0909762652797155, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 45487 + }, + { + "epoch": 0.45488, + "grad_norm": 0.912786378348586, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 45488 + }, + { + "epoch": 0.45489, + "grad_norm": 1.0841847361325025, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 45489 + }, + { + "epoch": 0.4549, + "grad_norm": 1.1803361670475894, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 45490 + }, + { + "epoch": 0.45491, + "grad_norm": 1.0155867052569048, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 45491 + }, + { + "epoch": 0.45492, + "grad_norm": 1.0879754268663258, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 45492 + }, + { + "epoch": 0.45493, + "grad_norm": 1.0012901450174634, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 45493 + }, + { + "epoch": 0.45494, + "grad_norm": 1.2819365604231598, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 45494 + }, + { + "epoch": 0.45495, + "grad_norm": 1.0167277376087978, + "learning_rate": 0.003, + "loss": 4.029, + "step": 45495 + }, + { + "epoch": 0.45496, + "grad_norm": 1.055072530110987, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 45496 + }, + { + "epoch": 0.45497, + "grad_norm": 1.3006149019620659, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 45497 + }, + { + "epoch": 0.45498, + "grad_norm": 0.8379297822843356, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 45498 + }, + { + "epoch": 0.45499, + "grad_norm": 0.9513370917471194, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 45499 + }, + { + "epoch": 0.455, + "grad_norm": 1.4352453626895574, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 45500 + }, + { + "epoch": 0.45501, + "grad_norm": 1.056246965679729, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 45501 + }, + { + "epoch": 0.45502, + "grad_norm": 0.9944564428467577, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 45502 + }, + { + "epoch": 0.45503, + "grad_norm": 1.254317896187832, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 45503 + }, + { + "epoch": 0.45504, + "grad_norm": 0.9403515545792391, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 45504 + }, + { + "epoch": 0.45505, + "grad_norm": 1.0211390005461252, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 45505 + }, + { + "epoch": 0.45506, + "grad_norm": 1.0493858881220417, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 45506 + }, + { + "epoch": 0.45507, + "grad_norm": 0.9668087996352948, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 45507 + }, + { + "epoch": 0.45508, + "grad_norm": 1.1794868436452586, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 45508 + }, + { + "epoch": 0.45509, + "grad_norm": 0.9319296363680265, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 45509 + }, + { + "epoch": 0.4551, + "grad_norm": 1.2803631387982797, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 45510 + }, + { + "epoch": 0.45511, + "grad_norm": 0.9459531704803518, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 45511 + }, + { + "epoch": 0.45512, + "grad_norm": 0.8845910187420446, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 45512 + }, + { + "epoch": 0.45513, + "grad_norm": 1.0023561214566667, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 45513 + }, + { + "epoch": 0.45514, + "grad_norm": 1.0748474120030584, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 45514 + }, + { + "epoch": 0.45515, + "grad_norm": 1.0339125313226136, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 45515 + }, + { + "epoch": 0.45516, + "grad_norm": 1.2488389308047358, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 45516 + }, + { + "epoch": 0.45517, + "grad_norm": 0.9768515288819634, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 45517 + }, + { + "epoch": 0.45518, + "grad_norm": 1.0098733264947888, + "learning_rate": 0.003, + "loss": 4.042, + "step": 45518 + }, + { + "epoch": 0.45519, + "grad_norm": 1.3584652198567129, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 45519 + }, + { + "epoch": 0.4552, + "grad_norm": 0.8748075468577239, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 45520 + }, + { + "epoch": 0.45521, + "grad_norm": 0.8480530408483541, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 45521 + }, + { + "epoch": 0.45522, + "grad_norm": 0.9659322870521512, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 45522 + }, + { + "epoch": 0.45523, + "grad_norm": 1.5117976362256256, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 45523 + }, + { + "epoch": 0.45524, + "grad_norm": 0.9720558451749709, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 45524 + }, + { + "epoch": 0.45525, + "grad_norm": 0.9467777419689254, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 45525 + }, + { + "epoch": 0.45526, + "grad_norm": 1.0536226124355454, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 45526 + }, + { + "epoch": 0.45527, + "grad_norm": 1.579686140693516, + "learning_rate": 0.003, + "loss": 4.047, + "step": 45527 + }, + { + "epoch": 0.45528, + "grad_norm": 0.8800989307199865, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 45528 + }, + { + "epoch": 0.45529, + "grad_norm": 0.8721138573137516, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 45529 + }, + { + "epoch": 0.4553, + "grad_norm": 0.7518314907652569, + "learning_rate": 0.003, + "loss": 4.02, + "step": 45530 + }, + { + "epoch": 0.45531, + "grad_norm": 0.8459570508948322, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 45531 + }, + { + "epoch": 0.45532, + "grad_norm": 0.8614621795879163, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 45532 + }, + { + "epoch": 0.45533, + "grad_norm": 1.2419183552152533, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 45533 + }, + { + "epoch": 0.45534, + "grad_norm": 1.2908734535127815, + "learning_rate": 0.003, + "loss": 4.038, + "step": 45534 + }, + { + "epoch": 0.45535, + "grad_norm": 0.7616866350715233, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 45535 + }, + { + "epoch": 0.45536, + "grad_norm": 0.7393831123737427, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 45536 + }, + { + "epoch": 0.45537, + "grad_norm": 0.9646865383483825, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 45537 + }, + { + "epoch": 0.45538, + "grad_norm": 1.0516341144722314, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 45538 + }, + { + "epoch": 0.45539, + "grad_norm": 1.1767374507865191, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 45539 + }, + { + "epoch": 0.4554, + "grad_norm": 0.9140325170291753, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 45540 + }, + { + "epoch": 0.45541, + "grad_norm": 0.975728781516209, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 45541 + }, + { + "epoch": 0.45542, + "grad_norm": 1.2873998335590202, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 45542 + }, + { + "epoch": 0.45543, + "grad_norm": 1.2083914968414662, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 45543 + }, + { + "epoch": 0.45544, + "grad_norm": 0.9045085329359082, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 45544 + }, + { + "epoch": 0.45545, + "grad_norm": 0.9824934673199971, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 45545 + }, + { + "epoch": 0.45546, + "grad_norm": 1.4247944425720385, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 45546 + }, + { + "epoch": 0.45547, + "grad_norm": 0.9831694703046003, + "learning_rate": 0.003, + "loss": 4.009, + "step": 45547 + }, + { + "epoch": 0.45548, + "grad_norm": 1.2660004075274476, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 45548 + }, + { + "epoch": 0.45549, + "grad_norm": 1.0489290012476415, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 45549 + }, + { + "epoch": 0.4555, + "grad_norm": 1.1095523515402328, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 45550 + }, + { + "epoch": 0.45551, + "grad_norm": 1.1229251053864533, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 45551 + }, + { + "epoch": 0.45552, + "grad_norm": 0.9084542602266907, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 45552 + }, + { + "epoch": 0.45553, + "grad_norm": 0.8504987179534116, + "learning_rate": 0.003, + "loss": 4.051, + "step": 45553 + }, + { + "epoch": 0.45554, + "grad_norm": 0.976108862764421, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 45554 + }, + { + "epoch": 0.45555, + "grad_norm": 1.0777565538823073, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 45555 + }, + { + "epoch": 0.45556, + "grad_norm": 1.1616299795721368, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 45556 + }, + { + "epoch": 0.45557, + "grad_norm": 1.0739998121896195, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 45557 + }, + { + "epoch": 0.45558, + "grad_norm": 0.9555420460580545, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 45558 + }, + { + "epoch": 0.45559, + "grad_norm": 1.0343042544961893, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 45559 + }, + { + "epoch": 0.4556, + "grad_norm": 0.9792589479778832, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 45560 + }, + { + "epoch": 0.45561, + "grad_norm": 1.4516134348393694, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 45561 + }, + { + "epoch": 0.45562, + "grad_norm": 0.9987245536378857, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 45562 + }, + { + "epoch": 0.45563, + "grad_norm": 1.1859358458293963, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 45563 + }, + { + "epoch": 0.45564, + "grad_norm": 0.8512796660092239, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 45564 + }, + { + "epoch": 0.45565, + "grad_norm": 0.9771706929178459, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 45565 + }, + { + "epoch": 0.45566, + "grad_norm": 0.9356337338704115, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 45566 + }, + { + "epoch": 0.45567, + "grad_norm": 1.0924657950424053, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 45567 + }, + { + "epoch": 0.45568, + "grad_norm": 1.1440396516389861, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 45568 + }, + { + "epoch": 0.45569, + "grad_norm": 1.079122422689615, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 45569 + }, + { + "epoch": 0.4557, + "grad_norm": 1.0566979209418557, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 45570 + }, + { + "epoch": 0.45571, + "grad_norm": 1.3763954520842157, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 45571 + }, + { + "epoch": 0.45572, + "grad_norm": 0.9778929513684821, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 45572 + }, + { + "epoch": 0.45573, + "grad_norm": 0.9324159543717866, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 45573 + }, + { + "epoch": 0.45574, + "grad_norm": 1.2521646730753406, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 45574 + }, + { + "epoch": 0.45575, + "grad_norm": 0.9345790842026706, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 45575 + }, + { + "epoch": 0.45576, + "grad_norm": 1.120161770830728, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 45576 + }, + { + "epoch": 0.45577, + "grad_norm": 1.3937137892757256, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 45577 + }, + { + "epoch": 0.45578, + "grad_norm": 0.8971965729673895, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 45578 + }, + { + "epoch": 0.45579, + "grad_norm": 0.8924364804073874, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 45579 + }, + { + "epoch": 0.4558, + "grad_norm": 0.9944594512754434, + "learning_rate": 0.003, + "loss": 4.012, + "step": 45580 + }, + { + "epoch": 0.45581, + "grad_norm": 1.1860097086776367, + "learning_rate": 0.003, + "loss": 4.049, + "step": 45581 + }, + { + "epoch": 0.45582, + "grad_norm": 1.14399372280873, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 45582 + }, + { + "epoch": 0.45583, + "grad_norm": 1.0805962270658847, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 45583 + }, + { + "epoch": 0.45584, + "grad_norm": 1.1805511532166975, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 45584 + }, + { + "epoch": 0.45585, + "grad_norm": 1.0402896713323053, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 45585 + }, + { + "epoch": 0.45586, + "grad_norm": 1.2016644339681497, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 45586 + }, + { + "epoch": 0.45587, + "grad_norm": 0.9666781299491927, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 45587 + }, + { + "epoch": 0.45588, + "grad_norm": 1.0994902815018375, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 45588 + }, + { + "epoch": 0.45589, + "grad_norm": 1.0716434180349905, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 45589 + }, + { + "epoch": 0.4559, + "grad_norm": 1.2135908261855126, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 45590 + }, + { + "epoch": 0.45591, + "grad_norm": 0.9189160040033104, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 45591 + }, + { + "epoch": 0.45592, + "grad_norm": 1.0266066426261786, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 45592 + }, + { + "epoch": 0.45593, + "grad_norm": 1.1481031216183535, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 45593 + }, + { + "epoch": 0.45594, + "grad_norm": 0.9227855359171125, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 45594 + }, + { + "epoch": 0.45595, + "grad_norm": 0.9545814746065845, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 45595 + }, + { + "epoch": 0.45596, + "grad_norm": 1.177669592894156, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 45596 + }, + { + "epoch": 0.45597, + "grad_norm": 1.1222751775410735, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 45597 + }, + { + "epoch": 0.45598, + "grad_norm": 0.9744435594740423, + "learning_rate": 0.003, + "loss": 4.012, + "step": 45598 + }, + { + "epoch": 0.45599, + "grad_norm": 1.002138981601541, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 45599 + }, + { + "epoch": 0.456, + "grad_norm": 0.913882992920237, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 45600 + }, + { + "epoch": 0.45601, + "grad_norm": 0.9328781734770419, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 45601 + }, + { + "epoch": 0.45602, + "grad_norm": 1.059448457642761, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 45602 + }, + { + "epoch": 0.45603, + "grad_norm": 1.2534710146616244, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 45603 + }, + { + "epoch": 0.45604, + "grad_norm": 1.4347387610505253, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 45604 + }, + { + "epoch": 0.45605, + "grad_norm": 0.9114030086465705, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 45605 + }, + { + "epoch": 0.45606, + "grad_norm": 1.0825089516492616, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 45606 + }, + { + "epoch": 0.45607, + "grad_norm": 1.1317105872287463, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 45607 + }, + { + "epoch": 0.45608, + "grad_norm": 1.2454893306922425, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 45608 + }, + { + "epoch": 0.45609, + "grad_norm": 0.9554938563912866, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 45609 + }, + { + "epoch": 0.4561, + "grad_norm": 1.0509461931859378, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 45610 + }, + { + "epoch": 0.45611, + "grad_norm": 1.1379519378748888, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 45611 + }, + { + "epoch": 0.45612, + "grad_norm": 0.9732500133032614, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 45612 + }, + { + "epoch": 0.45613, + "grad_norm": 1.170725960206192, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 45613 + }, + { + "epoch": 0.45614, + "grad_norm": 0.7879822055787815, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 45614 + }, + { + "epoch": 0.45615, + "grad_norm": 0.952237285988855, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 45615 + }, + { + "epoch": 0.45616, + "grad_norm": 1.2567202668264383, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 45616 + }, + { + "epoch": 0.45617, + "grad_norm": 1.2084914905340078, + "learning_rate": 0.003, + "loss": 4.06, + "step": 45617 + }, + { + "epoch": 0.45618, + "grad_norm": 0.9050758712983523, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 45618 + }, + { + "epoch": 0.45619, + "grad_norm": 0.9189163108089722, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 45619 + }, + { + "epoch": 0.4562, + "grad_norm": 0.9209799145218402, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 45620 + }, + { + "epoch": 0.45621, + "grad_norm": 1.0761207453733872, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 45621 + }, + { + "epoch": 0.45622, + "grad_norm": 1.080338959828679, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 45622 + }, + { + "epoch": 0.45623, + "grad_norm": 1.0250929527094403, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 45623 + }, + { + "epoch": 0.45624, + "grad_norm": 1.2567845317380564, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 45624 + }, + { + "epoch": 0.45625, + "grad_norm": 1.1125806923178276, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 45625 + }, + { + "epoch": 0.45626, + "grad_norm": 1.511581532711166, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 45626 + }, + { + "epoch": 0.45627, + "grad_norm": 0.9478960466027248, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 45627 + }, + { + "epoch": 0.45628, + "grad_norm": 1.1791404747635121, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 45628 + }, + { + "epoch": 0.45629, + "grad_norm": 0.9610438483527975, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 45629 + }, + { + "epoch": 0.4563, + "grad_norm": 1.1941006468235762, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 45630 + }, + { + "epoch": 0.45631, + "grad_norm": 1.0406898332879189, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 45631 + }, + { + "epoch": 0.45632, + "grad_norm": 1.1804174630533255, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 45632 + }, + { + "epoch": 0.45633, + "grad_norm": 0.9702555870187252, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 45633 + }, + { + "epoch": 0.45634, + "grad_norm": 1.2250059006293215, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 45634 + }, + { + "epoch": 0.45635, + "grad_norm": 1.278505372304201, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 45635 + }, + { + "epoch": 0.45636, + "grad_norm": 0.9998042059321081, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 45636 + }, + { + "epoch": 0.45637, + "grad_norm": 1.0691115270943876, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 45637 + }, + { + "epoch": 0.45638, + "grad_norm": 1.12011269562547, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 45638 + }, + { + "epoch": 0.45639, + "grad_norm": 1.0369794108945611, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 45639 + }, + { + "epoch": 0.4564, + "grad_norm": 1.3042382842996283, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 45640 + }, + { + "epoch": 0.45641, + "grad_norm": 0.9113622915050419, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 45641 + }, + { + "epoch": 0.45642, + "grad_norm": 0.9246724065498491, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 45642 + }, + { + "epoch": 0.45643, + "grad_norm": 1.169739451952915, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 45643 + }, + { + "epoch": 0.45644, + "grad_norm": 1.2441619265912243, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 45644 + }, + { + "epoch": 0.45645, + "grad_norm": 1.0541316964385519, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 45645 + }, + { + "epoch": 0.45646, + "grad_norm": 1.021573198127877, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 45646 + }, + { + "epoch": 0.45647, + "grad_norm": 0.9802202069030909, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 45647 + }, + { + "epoch": 0.45648, + "grad_norm": 0.8835809063714432, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 45648 + }, + { + "epoch": 0.45649, + "grad_norm": 1.0165061810353802, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 45649 + }, + { + "epoch": 0.4565, + "grad_norm": 1.061490883503742, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 45650 + }, + { + "epoch": 0.45651, + "grad_norm": 1.322618479852048, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 45651 + }, + { + "epoch": 0.45652, + "grad_norm": 0.9795945600161899, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 45652 + }, + { + "epoch": 0.45653, + "grad_norm": 1.3403785619069803, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 45653 + }, + { + "epoch": 0.45654, + "grad_norm": 0.9480266733124425, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 45654 + }, + { + "epoch": 0.45655, + "grad_norm": 0.8232962929132803, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 45655 + }, + { + "epoch": 0.45656, + "grad_norm": 0.963815314924635, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 45656 + }, + { + "epoch": 0.45657, + "grad_norm": 1.152070400237176, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 45657 + }, + { + "epoch": 0.45658, + "grad_norm": 1.0761210494183129, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 45658 + }, + { + "epoch": 0.45659, + "grad_norm": 1.140650043435263, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 45659 + }, + { + "epoch": 0.4566, + "grad_norm": 1.028842407758734, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 45660 + }, + { + "epoch": 0.45661, + "grad_norm": 1.2331641361659045, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 45661 + }, + { + "epoch": 0.45662, + "grad_norm": 0.89078916268411, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 45662 + }, + { + "epoch": 0.45663, + "grad_norm": 0.9465031966712224, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 45663 + }, + { + "epoch": 0.45664, + "grad_norm": 1.3712786451242178, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 45664 + }, + { + "epoch": 0.45665, + "grad_norm": 0.7799089656734709, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 45665 + }, + { + "epoch": 0.45666, + "grad_norm": 0.9199115898857314, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 45666 + }, + { + "epoch": 0.45667, + "grad_norm": 0.989993091721622, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 45667 + }, + { + "epoch": 0.45668, + "grad_norm": 1.1672924187254503, + "learning_rate": 0.003, + "loss": 4.052, + "step": 45668 + }, + { + "epoch": 0.45669, + "grad_norm": 1.1797634968310478, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 45669 + }, + { + "epoch": 0.4567, + "grad_norm": 0.9444067074115313, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 45670 + }, + { + "epoch": 0.45671, + "grad_norm": 1.0634256406043436, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 45671 + }, + { + "epoch": 0.45672, + "grad_norm": 1.0083714089248954, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 45672 + }, + { + "epoch": 0.45673, + "grad_norm": 1.0871504681584923, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 45673 + }, + { + "epoch": 0.45674, + "grad_norm": 1.1210733924778111, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 45674 + }, + { + "epoch": 0.45675, + "grad_norm": 0.9218627580183182, + "learning_rate": 0.003, + "loss": 4.026, + "step": 45675 + }, + { + "epoch": 0.45676, + "grad_norm": 0.9602600276991, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 45676 + }, + { + "epoch": 0.45677, + "grad_norm": 1.0599471410231274, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 45677 + }, + { + "epoch": 0.45678, + "grad_norm": 1.218448373626708, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 45678 + }, + { + "epoch": 0.45679, + "grad_norm": 0.985893672484656, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 45679 + }, + { + "epoch": 0.4568, + "grad_norm": 1.2303272537982037, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 45680 + }, + { + "epoch": 0.45681, + "grad_norm": 0.9352936432074335, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 45681 + }, + { + "epoch": 0.45682, + "grad_norm": 0.9429106858256802, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 45682 + }, + { + "epoch": 0.45683, + "grad_norm": 1.2146120622844334, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 45683 + }, + { + "epoch": 0.45684, + "grad_norm": 1.0844386814518805, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 45684 + }, + { + "epoch": 0.45685, + "grad_norm": 1.2159542399881231, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 45685 + }, + { + "epoch": 0.45686, + "grad_norm": 0.9450623554073683, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 45686 + }, + { + "epoch": 0.45687, + "grad_norm": 1.2147914281381182, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 45687 + }, + { + "epoch": 0.45688, + "grad_norm": 1.0916057355529538, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 45688 + }, + { + "epoch": 0.45689, + "grad_norm": 1.2263526694296838, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 45689 + }, + { + "epoch": 0.4569, + "grad_norm": 0.9836392843383652, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 45690 + }, + { + "epoch": 0.45691, + "grad_norm": 1.1794851870150063, + "learning_rate": 0.003, + "loss": 4.034, + "step": 45691 + }, + { + "epoch": 0.45692, + "grad_norm": 1.0067311796841647, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 45692 + }, + { + "epoch": 0.45693, + "grad_norm": 1.2143435252952577, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 45693 + }, + { + "epoch": 0.45694, + "grad_norm": 0.8667343938135692, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 45694 + }, + { + "epoch": 0.45695, + "grad_norm": 0.9121805322724016, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 45695 + }, + { + "epoch": 0.45696, + "grad_norm": 1.211600905940955, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 45696 + }, + { + "epoch": 0.45697, + "grad_norm": 1.0083471067500809, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 45697 + }, + { + "epoch": 0.45698, + "grad_norm": 1.1500125375171684, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 45698 + }, + { + "epoch": 0.45699, + "grad_norm": 1.0504947366796056, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 45699 + }, + { + "epoch": 0.457, + "grad_norm": 1.1058105444881288, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 45700 + }, + { + "epoch": 0.45701, + "grad_norm": 1.120629969708723, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 45701 + }, + { + "epoch": 0.45702, + "grad_norm": 1.0295530434667568, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 45702 + }, + { + "epoch": 0.45703, + "grad_norm": 1.0474373688460932, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 45703 + }, + { + "epoch": 0.45704, + "grad_norm": 1.1849493391753594, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 45704 + }, + { + "epoch": 0.45705, + "grad_norm": 1.0157569292124171, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 45705 + }, + { + "epoch": 0.45706, + "grad_norm": 1.1606466988084507, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 45706 + }, + { + "epoch": 0.45707, + "grad_norm": 1.0940470871573058, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 45707 + }, + { + "epoch": 0.45708, + "grad_norm": 0.8989542488540834, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 45708 + }, + { + "epoch": 0.45709, + "grad_norm": 0.9027988012913664, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 45709 + }, + { + "epoch": 0.4571, + "grad_norm": 0.9840931177504356, + "learning_rate": 0.003, + "loss": 4.046, + "step": 45710 + }, + { + "epoch": 0.45711, + "grad_norm": 1.1018757649439994, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 45711 + }, + { + "epoch": 0.45712, + "grad_norm": 0.9400304169258288, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 45712 + }, + { + "epoch": 0.45713, + "grad_norm": 1.1150114351103455, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 45713 + }, + { + "epoch": 0.45714, + "grad_norm": 1.1046917762654545, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 45714 + }, + { + "epoch": 0.45715, + "grad_norm": 1.1859581042994123, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 45715 + }, + { + "epoch": 0.45716, + "grad_norm": 1.1315740758206208, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 45716 + }, + { + "epoch": 0.45717, + "grad_norm": 0.9617385219445964, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 45717 + }, + { + "epoch": 0.45718, + "grad_norm": 1.1813682105474979, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 45718 + }, + { + "epoch": 0.45719, + "grad_norm": 1.0373037659506485, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 45719 + }, + { + "epoch": 0.4572, + "grad_norm": 1.3119677496870548, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 45720 + }, + { + "epoch": 0.45721, + "grad_norm": 1.061035053507475, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 45721 + }, + { + "epoch": 0.45722, + "grad_norm": 1.136501672523863, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 45722 + }, + { + "epoch": 0.45723, + "grad_norm": 1.1637827105852716, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 45723 + }, + { + "epoch": 0.45724, + "grad_norm": 0.8209352888531964, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 45724 + }, + { + "epoch": 0.45725, + "grad_norm": 0.9170993976429448, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 45725 + }, + { + "epoch": 0.45726, + "grad_norm": 1.1322927540005314, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 45726 + }, + { + "epoch": 0.45727, + "grad_norm": 1.0448115654312844, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 45727 + }, + { + "epoch": 0.45728, + "grad_norm": 1.262107165120052, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 45728 + }, + { + "epoch": 0.45729, + "grad_norm": 1.0812857768139685, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 45729 + }, + { + "epoch": 0.4573, + "grad_norm": 0.8491058436542708, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 45730 + }, + { + "epoch": 0.45731, + "grad_norm": 0.9317694068930383, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 45731 + }, + { + "epoch": 0.45732, + "grad_norm": 1.1394378571666148, + "learning_rate": 0.003, + "loss": 4.058, + "step": 45732 + }, + { + "epoch": 0.45733, + "grad_norm": 0.9677429576664485, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 45733 + }, + { + "epoch": 0.45734, + "grad_norm": 1.2095117702427294, + "learning_rate": 0.003, + "loss": 4.035, + "step": 45734 + }, + { + "epoch": 0.45735, + "grad_norm": 1.1038068654321043, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 45735 + }, + { + "epoch": 0.45736, + "grad_norm": 1.0787105683042453, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 45736 + }, + { + "epoch": 0.45737, + "grad_norm": 1.2691492239994595, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 45737 + }, + { + "epoch": 0.45738, + "grad_norm": 0.7578213350344218, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 45738 + }, + { + "epoch": 0.45739, + "grad_norm": 0.8349017418019307, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 45739 + }, + { + "epoch": 0.4574, + "grad_norm": 1.058584015554166, + "learning_rate": 0.003, + "loss": 4.033, + "step": 45740 + }, + { + "epoch": 0.45741, + "grad_norm": 1.1401605905908845, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 45741 + }, + { + "epoch": 0.45742, + "grad_norm": 1.1407326294678977, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 45742 + }, + { + "epoch": 0.45743, + "grad_norm": 1.1664600740336015, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 45743 + }, + { + "epoch": 0.45744, + "grad_norm": 0.9065189919416486, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 45744 + }, + { + "epoch": 0.45745, + "grad_norm": 0.9091224870074527, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 45745 + }, + { + "epoch": 0.45746, + "grad_norm": 1.0615562050786087, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 45746 + }, + { + "epoch": 0.45747, + "grad_norm": 1.324281989767866, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 45747 + }, + { + "epoch": 0.45748, + "grad_norm": 0.8004190216957491, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 45748 + }, + { + "epoch": 0.45749, + "grad_norm": 0.9798513587079062, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 45749 + }, + { + "epoch": 0.4575, + "grad_norm": 1.1864717796035353, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 45750 + }, + { + "epoch": 0.45751, + "grad_norm": 0.9841360378515313, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 45751 + }, + { + "epoch": 0.45752, + "grad_norm": 1.1336194158170492, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 45752 + }, + { + "epoch": 0.45753, + "grad_norm": 1.0318767579655423, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 45753 + }, + { + "epoch": 0.45754, + "grad_norm": 1.2392506670966346, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 45754 + }, + { + "epoch": 0.45755, + "grad_norm": 1.0782616992640859, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 45755 + }, + { + "epoch": 0.45756, + "grad_norm": 1.2355979674369284, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 45756 + }, + { + "epoch": 0.45757, + "grad_norm": 0.947664496206375, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 45757 + }, + { + "epoch": 0.45758, + "grad_norm": 1.0941967232112564, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 45758 + }, + { + "epoch": 0.45759, + "grad_norm": 1.023630885728467, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 45759 + }, + { + "epoch": 0.4576, + "grad_norm": 1.1299321365497712, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 45760 + }, + { + "epoch": 0.45761, + "grad_norm": 1.1765357929858635, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 45761 + }, + { + "epoch": 0.45762, + "grad_norm": 0.9080500230841679, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 45762 + }, + { + "epoch": 0.45763, + "grad_norm": 0.9335861476725859, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 45763 + }, + { + "epoch": 0.45764, + "grad_norm": 1.0911404663667454, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 45764 + }, + { + "epoch": 0.45765, + "grad_norm": 1.2330712266482924, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 45765 + }, + { + "epoch": 0.45766, + "grad_norm": 1.0476521997858732, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 45766 + }, + { + "epoch": 0.45767, + "grad_norm": 1.264365832926271, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 45767 + }, + { + "epoch": 0.45768, + "grad_norm": 0.9676932597060096, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 45768 + }, + { + "epoch": 0.45769, + "grad_norm": 1.0092120271709322, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 45769 + }, + { + "epoch": 0.4577, + "grad_norm": 1.1127683349256519, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 45770 + }, + { + "epoch": 0.45771, + "grad_norm": 0.9155675491928722, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 45771 + }, + { + "epoch": 0.45772, + "grad_norm": 0.9911652337215583, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 45772 + }, + { + "epoch": 0.45773, + "grad_norm": 1.1181975386326717, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 45773 + }, + { + "epoch": 0.45774, + "grad_norm": 1.0624524483529028, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 45774 + }, + { + "epoch": 0.45775, + "grad_norm": 1.0567500827243095, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 45775 + }, + { + "epoch": 0.45776, + "grad_norm": 1.0521876093532856, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 45776 + }, + { + "epoch": 0.45777, + "grad_norm": 1.1390424142315057, + "learning_rate": 0.003, + "loss": 4.019, + "step": 45777 + }, + { + "epoch": 0.45778, + "grad_norm": 1.3473263250158596, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 45778 + }, + { + "epoch": 0.45779, + "grad_norm": 1.0184047950330835, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 45779 + }, + { + "epoch": 0.4578, + "grad_norm": 0.9632250562521268, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 45780 + }, + { + "epoch": 0.45781, + "grad_norm": 1.1569454157918166, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 45781 + }, + { + "epoch": 0.45782, + "grad_norm": 0.965980749296808, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 45782 + }, + { + "epoch": 0.45783, + "grad_norm": 1.1384612191293444, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 45783 + }, + { + "epoch": 0.45784, + "grad_norm": 1.1140188249126932, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 45784 + }, + { + "epoch": 0.45785, + "grad_norm": 1.0795222602953614, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 45785 + }, + { + "epoch": 0.45786, + "grad_norm": 1.0812440786651285, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 45786 + }, + { + "epoch": 0.45787, + "grad_norm": 0.9372703518307007, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 45787 + }, + { + "epoch": 0.45788, + "grad_norm": 0.9888379939928045, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 45788 + }, + { + "epoch": 0.45789, + "grad_norm": 1.0999216059244483, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 45789 + }, + { + "epoch": 0.4579, + "grad_norm": 0.9218944755166598, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 45790 + }, + { + "epoch": 0.45791, + "grad_norm": 1.196245745756781, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 45791 + }, + { + "epoch": 0.45792, + "grad_norm": 1.2088321391633428, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 45792 + }, + { + "epoch": 0.45793, + "grad_norm": 0.9914810109852135, + "learning_rate": 0.003, + "loss": 4.028, + "step": 45793 + }, + { + "epoch": 0.45794, + "grad_norm": 1.2065481455042908, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 45794 + }, + { + "epoch": 0.45795, + "grad_norm": 1.0718745970259214, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 45795 + }, + { + "epoch": 0.45796, + "grad_norm": 1.1512298386995472, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 45796 + }, + { + "epoch": 0.45797, + "grad_norm": 1.012012095036231, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 45797 + }, + { + "epoch": 0.45798, + "grad_norm": 1.4185487676693012, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 45798 + }, + { + "epoch": 0.45799, + "grad_norm": 0.8761477409205021, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 45799 + }, + { + "epoch": 0.458, + "grad_norm": 1.0205808990347411, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 45800 + }, + { + "epoch": 0.45801, + "grad_norm": 1.1048390805107655, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 45801 + }, + { + "epoch": 0.45802, + "grad_norm": 1.1696905550765562, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 45802 + }, + { + "epoch": 0.45803, + "grad_norm": 0.8936732683946945, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 45803 + }, + { + "epoch": 0.45804, + "grad_norm": 1.0720687768550083, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 45804 + }, + { + "epoch": 0.45805, + "grad_norm": 0.9651586312444139, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 45805 + }, + { + "epoch": 0.45806, + "grad_norm": 0.9636578141501072, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 45806 + }, + { + "epoch": 0.45807, + "grad_norm": 1.2338131175154998, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 45807 + }, + { + "epoch": 0.45808, + "grad_norm": 1.1466328380965565, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 45808 + }, + { + "epoch": 0.45809, + "grad_norm": 1.0252397419309918, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 45809 + }, + { + "epoch": 0.4581, + "grad_norm": 0.9009161122184944, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 45810 + }, + { + "epoch": 0.45811, + "grad_norm": 0.9941425799435483, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 45811 + }, + { + "epoch": 0.45812, + "grad_norm": 1.2675301990454086, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 45812 + }, + { + "epoch": 0.45813, + "grad_norm": 0.8997661519637996, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 45813 + }, + { + "epoch": 0.45814, + "grad_norm": 0.9285073768666484, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 45814 + }, + { + "epoch": 0.45815, + "grad_norm": 1.1181367708833323, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 45815 + }, + { + "epoch": 0.45816, + "grad_norm": 1.2659874553148205, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 45816 + }, + { + "epoch": 0.45817, + "grad_norm": 0.9015886177784491, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 45817 + }, + { + "epoch": 0.45818, + "grad_norm": 0.944367806512296, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 45818 + }, + { + "epoch": 0.45819, + "grad_norm": 0.8880924165417217, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 45819 + }, + { + "epoch": 0.4582, + "grad_norm": 0.8343190802199303, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 45820 + }, + { + "epoch": 0.45821, + "grad_norm": 0.9128451261696922, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 45821 + }, + { + "epoch": 0.45822, + "grad_norm": 1.0416913935973506, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 45822 + }, + { + "epoch": 0.45823, + "grad_norm": 1.1623940160366255, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 45823 + }, + { + "epoch": 0.45824, + "grad_norm": 0.9813015472387507, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 45824 + }, + { + "epoch": 0.45825, + "grad_norm": 1.2153715790619302, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 45825 + }, + { + "epoch": 0.45826, + "grad_norm": 1.0505492090837067, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 45826 + }, + { + "epoch": 0.45827, + "grad_norm": 1.165249873772667, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 45827 + }, + { + "epoch": 0.45828, + "grad_norm": 1.1354097617425367, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 45828 + }, + { + "epoch": 0.45829, + "grad_norm": 0.9911181513572817, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 45829 + }, + { + "epoch": 0.4583, + "grad_norm": 1.3090488673787029, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 45830 + }, + { + "epoch": 0.45831, + "grad_norm": 0.8404290497053939, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 45831 + }, + { + "epoch": 0.45832, + "grad_norm": 0.8861850970998166, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 45832 + }, + { + "epoch": 0.45833, + "grad_norm": 1.184680404642842, + "learning_rate": 0.003, + "loss": 4.054, + "step": 45833 + }, + { + "epoch": 0.45834, + "grad_norm": 1.0172637939946647, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 45834 + }, + { + "epoch": 0.45835, + "grad_norm": 1.1814731857038334, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 45835 + }, + { + "epoch": 0.45836, + "grad_norm": 1.0339704103112672, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 45836 + }, + { + "epoch": 0.45837, + "grad_norm": 1.1045585123884394, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 45837 + }, + { + "epoch": 0.45838, + "grad_norm": 0.7781015813336972, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 45838 + }, + { + "epoch": 0.45839, + "grad_norm": 0.823629077989373, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 45839 + }, + { + "epoch": 0.4584, + "grad_norm": 1.0003457361009511, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 45840 + }, + { + "epoch": 0.45841, + "grad_norm": 1.4569878140543586, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 45841 + }, + { + "epoch": 0.45842, + "grad_norm": 0.8938151825497361, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 45842 + }, + { + "epoch": 0.45843, + "grad_norm": 0.9386017173954191, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 45843 + }, + { + "epoch": 0.45844, + "grad_norm": 1.1614715576556152, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 45844 + }, + { + "epoch": 0.45845, + "grad_norm": 1.5361039050466316, + "learning_rate": 0.003, + "loss": 4.0905, + "step": 45845 + }, + { + "epoch": 0.45846, + "grad_norm": 0.7911938268608186, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 45846 + }, + { + "epoch": 0.45847, + "grad_norm": 0.8315172557869512, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 45847 + }, + { + "epoch": 0.45848, + "grad_norm": 1.079664215782551, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 45848 + }, + { + "epoch": 0.45849, + "grad_norm": 1.3449920116764693, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 45849 + }, + { + "epoch": 0.4585, + "grad_norm": 1.150592541967816, + "learning_rate": 0.003, + "loss": 4.021, + "step": 45850 + }, + { + "epoch": 0.45851, + "grad_norm": 1.0718661230615538, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 45851 + }, + { + "epoch": 0.45852, + "grad_norm": 1.0491241221119942, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 45852 + }, + { + "epoch": 0.45853, + "grad_norm": 1.0643177019151209, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 45853 + }, + { + "epoch": 0.45854, + "grad_norm": 1.132057984555531, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 45854 + }, + { + "epoch": 0.45855, + "grad_norm": 0.9250549174377901, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 45855 + }, + { + "epoch": 0.45856, + "grad_norm": 0.9713444512483076, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 45856 + }, + { + "epoch": 0.45857, + "grad_norm": 1.0016104764661244, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 45857 + }, + { + "epoch": 0.45858, + "grad_norm": 1.0564753648132557, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 45858 + }, + { + "epoch": 0.45859, + "grad_norm": 0.9874544906131635, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 45859 + }, + { + "epoch": 0.4586, + "grad_norm": 1.0848749714416575, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 45860 + }, + { + "epoch": 0.45861, + "grad_norm": 1.0351753664710084, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 45861 + }, + { + "epoch": 0.45862, + "grad_norm": 1.132628922362136, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 45862 + }, + { + "epoch": 0.45863, + "grad_norm": 1.1829182856153928, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 45863 + }, + { + "epoch": 0.45864, + "grad_norm": 1.195882895265105, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 45864 + }, + { + "epoch": 0.45865, + "grad_norm": 1.0267801379417631, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 45865 + }, + { + "epoch": 0.45866, + "grad_norm": 1.2524027139035347, + "learning_rate": 0.003, + "loss": 4.039, + "step": 45866 + }, + { + "epoch": 0.45867, + "grad_norm": 1.023213079945863, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 45867 + }, + { + "epoch": 0.45868, + "grad_norm": 1.3360116575480911, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 45868 + }, + { + "epoch": 0.45869, + "grad_norm": 1.009512787341644, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 45869 + }, + { + "epoch": 0.4587, + "grad_norm": 1.1253058014078883, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 45870 + }, + { + "epoch": 0.45871, + "grad_norm": 1.258033944984937, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 45871 + }, + { + "epoch": 0.45872, + "grad_norm": 0.869506448730146, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 45872 + }, + { + "epoch": 0.45873, + "grad_norm": 0.8028187895075257, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 45873 + }, + { + "epoch": 0.45874, + "grad_norm": 0.868077492931787, + "learning_rate": 0.003, + "loss": 4.04, + "step": 45874 + }, + { + "epoch": 0.45875, + "grad_norm": 0.9603539918792448, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 45875 + }, + { + "epoch": 0.45876, + "grad_norm": 1.0885620515452379, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 45876 + }, + { + "epoch": 0.45877, + "grad_norm": 1.2137099466678796, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 45877 + }, + { + "epoch": 0.45878, + "grad_norm": 0.9700413088939633, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 45878 + }, + { + "epoch": 0.45879, + "grad_norm": 1.1569441642609812, + "learning_rate": 0.003, + "loss": 4.032, + "step": 45879 + }, + { + "epoch": 0.4588, + "grad_norm": 1.0013725337455128, + "learning_rate": 0.003, + "loss": 4.037, + "step": 45880 + }, + { + "epoch": 0.45881, + "grad_norm": 1.2762842496455362, + "learning_rate": 0.003, + "loss": 4.018, + "step": 45881 + }, + { + "epoch": 0.45882, + "grad_norm": 0.955423982348063, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 45882 + }, + { + "epoch": 0.45883, + "grad_norm": 1.1078069562413126, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 45883 + }, + { + "epoch": 0.45884, + "grad_norm": 1.122598081240139, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 45884 + }, + { + "epoch": 0.45885, + "grad_norm": 1.130185211861581, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 45885 + }, + { + "epoch": 0.45886, + "grad_norm": 0.9740813363912915, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 45886 + }, + { + "epoch": 0.45887, + "grad_norm": 1.0747150501165954, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 45887 + }, + { + "epoch": 0.45888, + "grad_norm": 1.1408184416597587, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 45888 + }, + { + "epoch": 0.45889, + "grad_norm": 0.8857505409687821, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 45889 + }, + { + "epoch": 0.4589, + "grad_norm": 0.816326776514859, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 45890 + }, + { + "epoch": 0.45891, + "grad_norm": 0.9370225555769743, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 45891 + }, + { + "epoch": 0.45892, + "grad_norm": 1.206704288094177, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 45892 + }, + { + "epoch": 0.45893, + "grad_norm": 0.955017107220451, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 45893 + }, + { + "epoch": 0.45894, + "grad_norm": 1.0271307028856531, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 45894 + }, + { + "epoch": 0.45895, + "grad_norm": 0.9864851084084844, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 45895 + }, + { + "epoch": 0.45896, + "grad_norm": 0.7832735373153087, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 45896 + }, + { + "epoch": 0.45897, + "grad_norm": 0.9827558006050635, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 45897 + }, + { + "epoch": 0.45898, + "grad_norm": 1.4648305550646787, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 45898 + }, + { + "epoch": 0.45899, + "grad_norm": 1.0650661899097804, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 45899 + }, + { + "epoch": 0.459, + "grad_norm": 1.028909885450094, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 45900 + }, + { + "epoch": 0.45901, + "grad_norm": 0.8245063341572909, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 45901 + }, + { + "epoch": 0.45902, + "grad_norm": 0.9698695724181013, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 45902 + }, + { + "epoch": 0.45903, + "grad_norm": 1.4543100526686061, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 45903 + }, + { + "epoch": 0.45904, + "grad_norm": 0.9389266825795456, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 45904 + }, + { + "epoch": 0.45905, + "grad_norm": 0.8963427643506768, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 45905 + }, + { + "epoch": 0.45906, + "grad_norm": 1.0095773149628002, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 45906 + }, + { + "epoch": 0.45907, + "grad_norm": 1.1600827093600672, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 45907 + }, + { + "epoch": 0.45908, + "grad_norm": 1.0307278938375664, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 45908 + }, + { + "epoch": 0.45909, + "grad_norm": 1.4957976837494942, + "learning_rate": 0.003, + "loss": 3.986, + "step": 45909 + }, + { + "epoch": 0.4591, + "grad_norm": 0.7441612801084305, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 45910 + }, + { + "epoch": 0.45911, + "grad_norm": 0.7705230397046363, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 45911 + }, + { + "epoch": 0.45912, + "grad_norm": 0.8674868994532677, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 45912 + }, + { + "epoch": 0.45913, + "grad_norm": 1.0078912948434928, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 45913 + }, + { + "epoch": 0.45914, + "grad_norm": 1.0214069598697935, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 45914 + }, + { + "epoch": 0.45915, + "grad_norm": 0.9991368475124691, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 45915 + }, + { + "epoch": 0.45916, + "grad_norm": 1.2426800211192877, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 45916 + }, + { + "epoch": 0.45917, + "grad_norm": 1.0938849744213965, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 45917 + }, + { + "epoch": 0.45918, + "grad_norm": 0.9283334772939268, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 45918 + }, + { + "epoch": 0.45919, + "grad_norm": 0.9939146372419407, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 45919 + }, + { + "epoch": 0.4592, + "grad_norm": 1.166908054366695, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 45920 + }, + { + "epoch": 0.45921, + "grad_norm": 1.0477913280582878, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 45921 + }, + { + "epoch": 0.45922, + "grad_norm": 1.0731799953881025, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 45922 + }, + { + "epoch": 0.45923, + "grad_norm": 0.9320431783702728, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 45923 + }, + { + "epoch": 0.45924, + "grad_norm": 1.1140368115501325, + "learning_rate": 0.003, + "loss": 4.04, + "step": 45924 + }, + { + "epoch": 0.45925, + "grad_norm": 0.9871318247543561, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 45925 + }, + { + "epoch": 0.45926, + "grad_norm": 1.14069577287473, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 45926 + }, + { + "epoch": 0.45927, + "grad_norm": 1.2774817722594032, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 45927 + }, + { + "epoch": 0.45928, + "grad_norm": 1.2013422935538356, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 45928 + }, + { + "epoch": 0.45929, + "grad_norm": 1.131723001478634, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 45929 + }, + { + "epoch": 0.4593, + "grad_norm": 0.7766025194605773, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 45930 + }, + { + "epoch": 0.45931, + "grad_norm": 0.7654827770066144, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 45931 + }, + { + "epoch": 0.45932, + "grad_norm": 0.9370561058618003, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 45932 + }, + { + "epoch": 0.45933, + "grad_norm": 0.9309782162868451, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 45933 + }, + { + "epoch": 0.45934, + "grad_norm": 0.9933987601256105, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 45934 + }, + { + "epoch": 0.45935, + "grad_norm": 1.4881528130455168, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 45935 + }, + { + "epoch": 0.45936, + "grad_norm": 0.9568416768231625, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 45936 + }, + { + "epoch": 0.45937, + "grad_norm": 1.1695883759488381, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 45937 + }, + { + "epoch": 0.45938, + "grad_norm": 1.2300808439277677, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 45938 + }, + { + "epoch": 0.45939, + "grad_norm": 1.02928876878782, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 45939 + }, + { + "epoch": 0.4594, + "grad_norm": 1.2690124075818046, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 45940 + }, + { + "epoch": 0.45941, + "grad_norm": 0.8984875330276901, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 45941 + }, + { + "epoch": 0.45942, + "grad_norm": 0.7892461098504577, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 45942 + }, + { + "epoch": 0.45943, + "grad_norm": 0.8828305864790933, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 45943 + }, + { + "epoch": 0.45944, + "grad_norm": 1.0459734779929273, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 45944 + }, + { + "epoch": 0.45945, + "grad_norm": 1.0895930387724793, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 45945 + }, + { + "epoch": 0.45946, + "grad_norm": 1.0221682579706195, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 45946 + }, + { + "epoch": 0.45947, + "grad_norm": 0.960866266241462, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 45947 + }, + { + "epoch": 0.45948, + "grad_norm": 0.9132863259739191, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 45948 + }, + { + "epoch": 0.45949, + "grad_norm": 1.208544539697202, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 45949 + }, + { + "epoch": 0.4595, + "grad_norm": 1.0991016113855445, + "learning_rate": 0.003, + "loss": 4.038, + "step": 45950 + }, + { + "epoch": 0.45951, + "grad_norm": 1.2361269313355208, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 45951 + }, + { + "epoch": 0.45952, + "grad_norm": 0.9279862057882333, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 45952 + }, + { + "epoch": 0.45953, + "grad_norm": 0.9199924955928822, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 45953 + }, + { + "epoch": 0.45954, + "grad_norm": 0.9226339146034876, + "learning_rate": 0.003, + "loss": 4.003, + "step": 45954 + }, + { + "epoch": 0.45955, + "grad_norm": 1.2185424658711408, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 45955 + }, + { + "epoch": 0.45956, + "grad_norm": 1.1043603441435546, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 45956 + }, + { + "epoch": 0.45957, + "grad_norm": 1.0901582931465366, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 45957 + }, + { + "epoch": 0.45958, + "grad_norm": 1.1980423420699167, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 45958 + }, + { + "epoch": 0.45959, + "grad_norm": 0.9148665390868154, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 45959 + }, + { + "epoch": 0.4596, + "grad_norm": 0.861725685763828, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 45960 + }, + { + "epoch": 0.45961, + "grad_norm": 1.032381439131911, + "learning_rate": 0.003, + "loss": 4.004, + "step": 45961 + }, + { + "epoch": 0.45962, + "grad_norm": 1.4421508933929779, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 45962 + }, + { + "epoch": 0.45963, + "grad_norm": 0.8244969221303526, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 45963 + }, + { + "epoch": 0.45964, + "grad_norm": 0.8008276371289725, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 45964 + }, + { + "epoch": 0.45965, + "grad_norm": 0.842489518705725, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 45965 + }, + { + "epoch": 0.45966, + "grad_norm": 1.1533423763583335, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 45966 + }, + { + "epoch": 0.45967, + "grad_norm": 1.3091724041653183, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 45967 + }, + { + "epoch": 0.45968, + "grad_norm": 0.7358813598811758, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 45968 + }, + { + "epoch": 0.45969, + "grad_norm": 0.7860178582819948, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 45969 + }, + { + "epoch": 0.4597, + "grad_norm": 0.979613181552971, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 45970 + }, + { + "epoch": 0.45971, + "grad_norm": 1.5551037142169943, + "learning_rate": 0.003, + "loss": 4.037, + "step": 45971 + }, + { + "epoch": 0.45972, + "grad_norm": 0.7348535598386955, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 45972 + }, + { + "epoch": 0.45973, + "grad_norm": 0.7712328827300194, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 45973 + }, + { + "epoch": 0.45974, + "grad_norm": 0.9013613475389782, + "learning_rate": 0.003, + "loss": 4.032, + "step": 45974 + }, + { + "epoch": 0.45975, + "grad_norm": 1.1816942818634282, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 45975 + }, + { + "epoch": 0.45976, + "grad_norm": 1.1563592734372177, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 45976 + }, + { + "epoch": 0.45977, + "grad_norm": 1.1103505447545565, + "learning_rate": 0.003, + "loss": 4.047, + "step": 45977 + }, + { + "epoch": 0.45978, + "grad_norm": 1.2414688770479996, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 45978 + }, + { + "epoch": 0.45979, + "grad_norm": 0.967583778838064, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 45979 + }, + { + "epoch": 0.4598, + "grad_norm": 0.9285016108743058, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 45980 + }, + { + "epoch": 0.45981, + "grad_norm": 1.1633437605799621, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 45981 + }, + { + "epoch": 0.45982, + "grad_norm": 1.2143355490504462, + "learning_rate": 0.003, + "loss": 4.028, + "step": 45982 + }, + { + "epoch": 0.45983, + "grad_norm": 1.0442870901059358, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 45983 + }, + { + "epoch": 0.45984, + "grad_norm": 1.1604186318952543, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 45984 + }, + { + "epoch": 0.45985, + "grad_norm": 1.0213827662188475, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 45985 + }, + { + "epoch": 0.45986, + "grad_norm": 1.095265218669474, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 45986 + }, + { + "epoch": 0.45987, + "grad_norm": 0.9720125573389405, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 45987 + }, + { + "epoch": 0.45988, + "grad_norm": 0.965767153373626, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 45988 + }, + { + "epoch": 0.45989, + "grad_norm": 1.0854127417722759, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 45989 + }, + { + "epoch": 0.4599, + "grad_norm": 1.11721487836905, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 45990 + }, + { + "epoch": 0.45991, + "grad_norm": 0.9843415434965764, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 45991 + }, + { + "epoch": 0.45992, + "grad_norm": 1.024908911497259, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 45992 + }, + { + "epoch": 0.45993, + "grad_norm": 1.067954276463941, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 45993 + }, + { + "epoch": 0.45994, + "grad_norm": 1.033104637983808, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 45994 + }, + { + "epoch": 0.45995, + "grad_norm": 1.6523373529714085, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 45995 + }, + { + "epoch": 0.45996, + "grad_norm": 0.7471491851629981, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 45996 + }, + { + "epoch": 0.45997, + "grad_norm": 0.8611996140625984, + "learning_rate": 0.003, + "loss": 4.038, + "step": 45997 + }, + { + "epoch": 0.45998, + "grad_norm": 1.0439045238876288, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 45998 + }, + { + "epoch": 0.45999, + "grad_norm": 1.2673484157999146, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 45999 + }, + { + "epoch": 0.46, + "grad_norm": 1.092455691840197, + "learning_rate": 0.003, + "loss": 4.043, + "step": 46000 + }, + { + "epoch": 0.46001, + "grad_norm": 1.0917927158815046, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 46001 + }, + { + "epoch": 0.46002, + "grad_norm": 1.2253902902236666, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 46002 + }, + { + "epoch": 0.46003, + "grad_norm": 1.1393499412449222, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 46003 + }, + { + "epoch": 0.46004, + "grad_norm": 0.9255758209098675, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 46004 + }, + { + "epoch": 0.46005, + "grad_norm": 1.0220078726663495, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 46005 + }, + { + "epoch": 0.46006, + "grad_norm": 1.211065489396454, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 46006 + }, + { + "epoch": 0.46007, + "grad_norm": 0.8635039862749985, + "learning_rate": 0.003, + "loss": 4.006, + "step": 46007 + }, + { + "epoch": 0.46008, + "grad_norm": 0.8635170576374656, + "learning_rate": 0.003, + "loss": 4.029, + "step": 46008 + }, + { + "epoch": 0.46009, + "grad_norm": 0.9399790380120802, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 46009 + }, + { + "epoch": 0.4601, + "grad_norm": 0.8926647124121495, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 46010 + }, + { + "epoch": 0.46011, + "grad_norm": 1.2421514442026815, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 46011 + }, + { + "epoch": 0.46012, + "grad_norm": 0.9227465229847773, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 46012 + }, + { + "epoch": 0.46013, + "grad_norm": 0.9707252947070996, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 46013 + }, + { + "epoch": 0.46014, + "grad_norm": 0.9874468282544588, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 46014 + }, + { + "epoch": 0.46015, + "grad_norm": 1.198289349183322, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 46015 + }, + { + "epoch": 0.46016, + "grad_norm": 0.9284932249087957, + "learning_rate": 0.003, + "loss": 4.05, + "step": 46016 + }, + { + "epoch": 0.46017, + "grad_norm": 0.9824122555337418, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 46017 + }, + { + "epoch": 0.46018, + "grad_norm": 1.084873713066318, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 46018 + }, + { + "epoch": 0.46019, + "grad_norm": 1.265884304823973, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 46019 + }, + { + "epoch": 0.4602, + "grad_norm": 1.2959301897696478, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 46020 + }, + { + "epoch": 0.46021, + "grad_norm": 1.2468735669320465, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 46021 + }, + { + "epoch": 0.46022, + "grad_norm": 1.112011981222518, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 46022 + }, + { + "epoch": 0.46023, + "grad_norm": 1.0791475463149225, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 46023 + }, + { + "epoch": 0.46024, + "grad_norm": 1.2167656968761729, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 46024 + }, + { + "epoch": 0.46025, + "grad_norm": 0.9813801362183567, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 46025 + }, + { + "epoch": 0.46026, + "grad_norm": 1.0365900564593982, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 46026 + }, + { + "epoch": 0.46027, + "grad_norm": 1.2689259767853323, + "learning_rate": 0.003, + "loss": 4.025, + "step": 46027 + }, + { + "epoch": 0.46028, + "grad_norm": 0.9867986062805204, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 46028 + }, + { + "epoch": 0.46029, + "grad_norm": 1.1250341839202695, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 46029 + }, + { + "epoch": 0.4603, + "grad_norm": 1.0297629624969848, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 46030 + }, + { + "epoch": 0.46031, + "grad_norm": 1.3789600983413364, + "learning_rate": 0.003, + "loss": 4.04, + "step": 46031 + }, + { + "epoch": 0.46032, + "grad_norm": 0.9777810317420313, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 46032 + }, + { + "epoch": 0.46033, + "grad_norm": 1.063521625237807, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 46033 + }, + { + "epoch": 0.46034, + "grad_norm": 1.1134077892278065, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 46034 + }, + { + "epoch": 0.46035, + "grad_norm": 1.0710012178455617, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 46035 + }, + { + "epoch": 0.46036, + "grad_norm": 1.0404273555848267, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 46036 + }, + { + "epoch": 0.46037, + "grad_norm": 0.88788349163557, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 46037 + }, + { + "epoch": 0.46038, + "grad_norm": 0.9096043178214666, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 46038 + }, + { + "epoch": 0.46039, + "grad_norm": 1.0289080773455064, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 46039 + }, + { + "epoch": 0.4604, + "grad_norm": 1.1737816091095326, + "learning_rate": 0.003, + "loss": 4.058, + "step": 46040 + }, + { + "epoch": 0.46041, + "grad_norm": 1.4246862627907757, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 46041 + }, + { + "epoch": 0.46042, + "grad_norm": 0.7973822307171218, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 46042 + }, + { + "epoch": 0.46043, + "grad_norm": 0.8542337056400673, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 46043 + }, + { + "epoch": 0.46044, + "grad_norm": 1.1804387147472146, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 46044 + }, + { + "epoch": 0.46045, + "grad_norm": 1.4777466244200819, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 46045 + }, + { + "epoch": 0.46046, + "grad_norm": 0.6696053459400724, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 46046 + }, + { + "epoch": 0.46047, + "grad_norm": 0.8546907491661905, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 46047 + }, + { + "epoch": 0.46048, + "grad_norm": 0.9420334304874141, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 46048 + }, + { + "epoch": 0.46049, + "grad_norm": 1.061706456362875, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 46049 + }, + { + "epoch": 0.4605, + "grad_norm": 1.225068243235443, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 46050 + }, + { + "epoch": 0.46051, + "grad_norm": 1.2039183301153016, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 46051 + }, + { + "epoch": 0.46052, + "grad_norm": 0.9463887644866726, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 46052 + }, + { + "epoch": 0.46053, + "grad_norm": 1.1818015867900997, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 46053 + }, + { + "epoch": 0.46054, + "grad_norm": 0.9936855852396331, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 46054 + }, + { + "epoch": 0.46055, + "grad_norm": 1.1137363917389782, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 46055 + }, + { + "epoch": 0.46056, + "grad_norm": 0.928860265989577, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 46056 + }, + { + "epoch": 0.46057, + "grad_norm": 1.0219182314982653, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 46057 + }, + { + "epoch": 0.46058, + "grad_norm": 1.3306633472886156, + "learning_rate": 0.003, + "loss": 4.034, + "step": 46058 + }, + { + "epoch": 0.46059, + "grad_norm": 0.9235541811202727, + "learning_rate": 0.003, + "loss": 4.01, + "step": 46059 + }, + { + "epoch": 0.4606, + "grad_norm": 1.0659492934424069, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 46060 + }, + { + "epoch": 0.46061, + "grad_norm": 1.1374098596412698, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 46061 + }, + { + "epoch": 0.46062, + "grad_norm": 1.0887254055354318, + "learning_rate": 0.003, + "loss": 4.024, + "step": 46062 + }, + { + "epoch": 0.46063, + "grad_norm": 1.2133275978759437, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 46063 + }, + { + "epoch": 0.46064, + "grad_norm": 0.8409873865092347, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 46064 + }, + { + "epoch": 0.46065, + "grad_norm": 0.9643693824333854, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 46065 + }, + { + "epoch": 0.46066, + "grad_norm": 1.052089594098514, + "learning_rate": 0.003, + "loss": 4.06, + "step": 46066 + }, + { + "epoch": 0.46067, + "grad_norm": 1.16975904518424, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 46067 + }, + { + "epoch": 0.46068, + "grad_norm": 1.1175143869044388, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 46068 + }, + { + "epoch": 0.46069, + "grad_norm": 1.2156494547155832, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 46069 + }, + { + "epoch": 0.4607, + "grad_norm": 1.042021536570283, + "learning_rate": 0.003, + "loss": 3.982, + "step": 46070 + }, + { + "epoch": 0.46071, + "grad_norm": 1.1299002024764142, + "learning_rate": 0.003, + "loss": 3.996, + "step": 46071 + }, + { + "epoch": 0.46072, + "grad_norm": 1.0796211342905981, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 46072 + }, + { + "epoch": 0.46073, + "grad_norm": 1.0786485319175505, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 46073 + }, + { + "epoch": 0.46074, + "grad_norm": 1.062343215959134, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 46074 + }, + { + "epoch": 0.46075, + "grad_norm": 1.142726308938457, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 46075 + }, + { + "epoch": 0.46076, + "grad_norm": 0.9045610602573578, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 46076 + }, + { + "epoch": 0.46077, + "grad_norm": 0.9544949202409677, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 46077 + }, + { + "epoch": 0.46078, + "grad_norm": 0.9326793241294937, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 46078 + }, + { + "epoch": 0.46079, + "grad_norm": 1.065707201707584, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 46079 + }, + { + "epoch": 0.4608, + "grad_norm": 1.3231799178811854, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 46080 + }, + { + "epoch": 0.46081, + "grad_norm": 0.7892762229876548, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 46081 + }, + { + "epoch": 0.46082, + "grad_norm": 0.837895228481627, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 46082 + }, + { + "epoch": 0.46083, + "grad_norm": 0.7997610098295449, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 46083 + }, + { + "epoch": 0.46084, + "grad_norm": 1.0063426273531944, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 46084 + }, + { + "epoch": 0.46085, + "grad_norm": 1.1176009148088983, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 46085 + }, + { + "epoch": 0.46086, + "grad_norm": 1.0433465048323585, + "learning_rate": 0.003, + "loss": 4.055, + "step": 46086 + }, + { + "epoch": 0.46087, + "grad_norm": 1.1863264892778387, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 46087 + }, + { + "epoch": 0.46088, + "grad_norm": 0.9088080361009924, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 46088 + }, + { + "epoch": 0.46089, + "grad_norm": 1.0273553523479741, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 46089 + }, + { + "epoch": 0.4609, + "grad_norm": 0.8597395336408608, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 46090 + }, + { + "epoch": 0.46091, + "grad_norm": 0.9253206632574132, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 46091 + }, + { + "epoch": 0.46092, + "grad_norm": 0.9675207920467879, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 46092 + }, + { + "epoch": 0.46093, + "grad_norm": 1.1978469963818636, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 46093 + }, + { + "epoch": 0.46094, + "grad_norm": 1.1689959559352787, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 46094 + }, + { + "epoch": 0.46095, + "grad_norm": 1.3658136012514395, + "learning_rate": 0.003, + "loss": 4.044, + "step": 46095 + }, + { + "epoch": 0.46096, + "grad_norm": 1.0892185058115562, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 46096 + }, + { + "epoch": 0.46097, + "grad_norm": 1.2130081771859418, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 46097 + }, + { + "epoch": 0.46098, + "grad_norm": 1.2326023449464207, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 46098 + }, + { + "epoch": 0.46099, + "grad_norm": 0.7906880110332741, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 46099 + }, + { + "epoch": 0.461, + "grad_norm": 0.6256711699271218, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 46100 + }, + { + "epoch": 0.46101, + "grad_norm": 0.6921326134545688, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 46101 + }, + { + "epoch": 0.46102, + "grad_norm": 0.8489633260882326, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 46102 + }, + { + "epoch": 0.46103, + "grad_norm": 1.0121441438545307, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 46103 + }, + { + "epoch": 0.46104, + "grad_norm": 1.1105905945986423, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 46104 + }, + { + "epoch": 0.46105, + "grad_norm": 1.1023759783427622, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 46105 + }, + { + "epoch": 0.46106, + "grad_norm": 1.255860569029664, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 46106 + }, + { + "epoch": 0.46107, + "grad_norm": 0.9098972788826193, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 46107 + }, + { + "epoch": 0.46108, + "grad_norm": 1.1592529205846136, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 46108 + }, + { + "epoch": 0.46109, + "grad_norm": 1.340093197682869, + "learning_rate": 0.003, + "loss": 4.001, + "step": 46109 + }, + { + "epoch": 0.4611, + "grad_norm": 1.1541457726301356, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 46110 + }, + { + "epoch": 0.46111, + "grad_norm": 1.0210224427826002, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 46111 + }, + { + "epoch": 0.46112, + "grad_norm": 1.1712352103855983, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 46112 + }, + { + "epoch": 0.46113, + "grad_norm": 1.0513111454414221, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 46113 + }, + { + "epoch": 0.46114, + "grad_norm": 1.238595190235626, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 46114 + }, + { + "epoch": 0.46115, + "grad_norm": 1.0198253035504496, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 46115 + }, + { + "epoch": 0.46116, + "grad_norm": 1.0486408701162542, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 46116 + }, + { + "epoch": 0.46117, + "grad_norm": 0.941735657260333, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 46117 + }, + { + "epoch": 0.46118, + "grad_norm": 1.1469343583348401, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 46118 + }, + { + "epoch": 0.46119, + "grad_norm": 1.0447879617337947, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 46119 + }, + { + "epoch": 0.4612, + "grad_norm": 1.090169500948542, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 46120 + }, + { + "epoch": 0.46121, + "grad_norm": 1.007918541884802, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 46121 + }, + { + "epoch": 0.46122, + "grad_norm": 1.5535007708873594, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 46122 + }, + { + "epoch": 0.46123, + "grad_norm": 1.0640480493143964, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 46123 + }, + { + "epoch": 0.46124, + "grad_norm": 1.1916496040692857, + "learning_rate": 0.003, + "loss": 4.034, + "step": 46124 + }, + { + "epoch": 0.46125, + "grad_norm": 1.0328183997974596, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 46125 + }, + { + "epoch": 0.46126, + "grad_norm": 1.2101257864128414, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 46126 + }, + { + "epoch": 0.46127, + "grad_norm": 1.012347669465698, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 46127 + }, + { + "epoch": 0.46128, + "grad_norm": 1.0716608211269973, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 46128 + }, + { + "epoch": 0.46129, + "grad_norm": 1.1394983649161192, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 46129 + }, + { + "epoch": 0.4613, + "grad_norm": 0.92664936878226, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 46130 + }, + { + "epoch": 0.46131, + "grad_norm": 0.799404095887877, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 46131 + }, + { + "epoch": 0.46132, + "grad_norm": 0.789981811187575, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 46132 + }, + { + "epoch": 0.46133, + "grad_norm": 0.9028480038953959, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 46133 + }, + { + "epoch": 0.46134, + "grad_norm": 1.020313724089687, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 46134 + }, + { + "epoch": 0.46135, + "grad_norm": 1.1646319534737715, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 46135 + }, + { + "epoch": 0.46136, + "grad_norm": 1.1997996348165638, + "learning_rate": 0.003, + "loss": 4.038, + "step": 46136 + }, + { + "epoch": 0.46137, + "grad_norm": 1.029417826759055, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 46137 + }, + { + "epoch": 0.46138, + "grad_norm": 1.2271711752310257, + "learning_rate": 0.003, + "loss": 4.05, + "step": 46138 + }, + { + "epoch": 0.46139, + "grad_norm": 1.050256757790238, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 46139 + }, + { + "epoch": 0.4614, + "grad_norm": 0.9927812726931422, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 46140 + }, + { + "epoch": 0.46141, + "grad_norm": 1.320609376874102, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 46141 + }, + { + "epoch": 0.46142, + "grad_norm": 0.9679640774526101, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 46142 + }, + { + "epoch": 0.46143, + "grad_norm": 0.9248622501366979, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 46143 + }, + { + "epoch": 0.46144, + "grad_norm": 1.2876630740698425, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 46144 + }, + { + "epoch": 0.46145, + "grad_norm": 0.9901381109026297, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 46145 + }, + { + "epoch": 0.46146, + "grad_norm": 1.290224997493041, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 46146 + }, + { + "epoch": 0.46147, + "grad_norm": 0.975235072157531, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 46147 + }, + { + "epoch": 0.46148, + "grad_norm": 1.0969696871573886, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 46148 + }, + { + "epoch": 0.46149, + "grad_norm": 0.9716334627677276, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 46149 + }, + { + "epoch": 0.4615, + "grad_norm": 1.0757463553952724, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 46150 + }, + { + "epoch": 0.46151, + "grad_norm": 0.9543646369564979, + "learning_rate": 0.003, + "loss": 4.051, + "step": 46151 + }, + { + "epoch": 0.46152, + "grad_norm": 0.8283887684116008, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 46152 + }, + { + "epoch": 0.46153, + "grad_norm": 0.9405877619809262, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 46153 + }, + { + "epoch": 0.46154, + "grad_norm": 1.0154093522358723, + "learning_rate": 0.003, + "loss": 4.017, + "step": 46154 + }, + { + "epoch": 0.46155, + "grad_norm": 1.197474483443573, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 46155 + }, + { + "epoch": 0.46156, + "grad_norm": 1.291041426005145, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 46156 + }, + { + "epoch": 0.46157, + "grad_norm": 0.8900862740749786, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 46157 + }, + { + "epoch": 0.46158, + "grad_norm": 0.7956130728088207, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 46158 + }, + { + "epoch": 0.46159, + "grad_norm": 0.9452368726719426, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 46159 + }, + { + "epoch": 0.4616, + "grad_norm": 1.2630516215568557, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 46160 + }, + { + "epoch": 0.46161, + "grad_norm": 1.0612744478020213, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 46161 + }, + { + "epoch": 0.46162, + "grad_norm": 1.396818895111486, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 46162 + }, + { + "epoch": 0.46163, + "grad_norm": 1.0612308243077988, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 46163 + }, + { + "epoch": 0.46164, + "grad_norm": 0.9620077454136261, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 46164 + }, + { + "epoch": 0.46165, + "grad_norm": 0.9828088124058251, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 46165 + }, + { + "epoch": 0.46166, + "grad_norm": 1.1615352321483674, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 46166 + }, + { + "epoch": 0.46167, + "grad_norm": 1.0120141605985566, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 46167 + }, + { + "epoch": 0.46168, + "grad_norm": 1.1104637116870273, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 46168 + }, + { + "epoch": 0.46169, + "grad_norm": 1.0469227590173469, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 46169 + }, + { + "epoch": 0.4617, + "grad_norm": 1.1896937330634139, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 46170 + }, + { + "epoch": 0.46171, + "grad_norm": 1.0284427048689884, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 46171 + }, + { + "epoch": 0.46172, + "grad_norm": 1.3214005757942529, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 46172 + }, + { + "epoch": 0.46173, + "grad_norm": 0.9832348588608695, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 46173 + }, + { + "epoch": 0.46174, + "grad_norm": 1.0693956616263436, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 46174 + }, + { + "epoch": 0.46175, + "grad_norm": 0.940316934989616, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 46175 + }, + { + "epoch": 0.46176, + "grad_norm": 0.8640058956949864, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 46176 + }, + { + "epoch": 0.46177, + "grad_norm": 0.987719668401159, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 46177 + }, + { + "epoch": 0.46178, + "grad_norm": 1.1405425643703229, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 46178 + }, + { + "epoch": 0.46179, + "grad_norm": 0.7765431166505723, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 46179 + }, + { + "epoch": 0.4618, + "grad_norm": 0.7384520481761429, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 46180 + }, + { + "epoch": 0.46181, + "grad_norm": 0.6781677430973745, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 46181 + }, + { + "epoch": 0.46182, + "grad_norm": 0.7585771146761845, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 46182 + }, + { + "epoch": 0.46183, + "grad_norm": 0.9510396605131343, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 46183 + }, + { + "epoch": 0.46184, + "grad_norm": 1.0989705632909692, + "learning_rate": 0.003, + "loss": 4.035, + "step": 46184 + }, + { + "epoch": 0.46185, + "grad_norm": 1.1049896241528774, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 46185 + }, + { + "epoch": 0.46186, + "grad_norm": 1.2327889303894033, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 46186 + }, + { + "epoch": 0.46187, + "grad_norm": 0.9751098646464919, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 46187 + }, + { + "epoch": 0.46188, + "grad_norm": 1.0394601219478359, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 46188 + }, + { + "epoch": 0.46189, + "grad_norm": 1.1713993402951581, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 46189 + }, + { + "epoch": 0.4619, + "grad_norm": 1.249511511703673, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 46190 + }, + { + "epoch": 0.46191, + "grad_norm": 0.9629457287849982, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 46191 + }, + { + "epoch": 0.46192, + "grad_norm": 1.2609798049719505, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 46192 + }, + { + "epoch": 0.46193, + "grad_norm": 1.1001496456626902, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 46193 + }, + { + "epoch": 0.46194, + "grad_norm": 1.090542708727239, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 46194 + }, + { + "epoch": 0.46195, + "grad_norm": 1.0309665219681268, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 46195 + }, + { + "epoch": 0.46196, + "grad_norm": 1.2420665575433436, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 46196 + }, + { + "epoch": 0.46197, + "grad_norm": 0.9613273098711725, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 46197 + }, + { + "epoch": 0.46198, + "grad_norm": 1.1586144278034267, + "learning_rate": 0.003, + "loss": 4.082, + "step": 46198 + }, + { + "epoch": 0.46199, + "grad_norm": 1.0578254318042757, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 46199 + }, + { + "epoch": 0.462, + "grad_norm": 1.1483203796883519, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 46200 + }, + { + "epoch": 0.46201, + "grad_norm": 1.063940618772463, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 46201 + }, + { + "epoch": 0.46202, + "grad_norm": 0.9824555438451897, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 46202 + }, + { + "epoch": 0.46203, + "grad_norm": 1.128358408098246, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 46203 + }, + { + "epoch": 0.46204, + "grad_norm": 1.3013299680724397, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 46204 + }, + { + "epoch": 0.46205, + "grad_norm": 1.0828283195090544, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 46205 + }, + { + "epoch": 0.46206, + "grad_norm": 1.3280585869016972, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 46206 + }, + { + "epoch": 0.46207, + "grad_norm": 1.0908702982671707, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 46207 + }, + { + "epoch": 0.46208, + "grad_norm": 1.0014991472004076, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 46208 + }, + { + "epoch": 0.46209, + "grad_norm": 1.1072477628172575, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 46209 + }, + { + "epoch": 0.4621, + "grad_norm": 1.0771063020458425, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 46210 + }, + { + "epoch": 0.46211, + "grad_norm": 0.9859113434592894, + "learning_rate": 0.003, + "loss": 4.035, + "step": 46211 + }, + { + "epoch": 0.46212, + "grad_norm": 1.2605687647337664, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 46212 + }, + { + "epoch": 0.46213, + "grad_norm": 0.8889349582885533, + "learning_rate": 0.003, + "loss": 4.04, + "step": 46213 + }, + { + "epoch": 0.46214, + "grad_norm": 0.886198785470355, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 46214 + }, + { + "epoch": 0.46215, + "grad_norm": 1.0314712963113686, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 46215 + }, + { + "epoch": 0.46216, + "grad_norm": 1.1885849866137357, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 46216 + }, + { + "epoch": 0.46217, + "grad_norm": 1.1409826732005766, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 46217 + }, + { + "epoch": 0.46218, + "grad_norm": 1.0263576240326717, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 46218 + }, + { + "epoch": 0.46219, + "grad_norm": 1.1858577780916189, + "learning_rate": 0.003, + "loss": 4.044, + "step": 46219 + }, + { + "epoch": 0.4622, + "grad_norm": 0.9571786336707214, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 46220 + }, + { + "epoch": 0.46221, + "grad_norm": 0.9237240112752786, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 46221 + }, + { + "epoch": 0.46222, + "grad_norm": 0.8091076144813452, + "learning_rate": 0.003, + "loss": 4.026, + "step": 46222 + }, + { + "epoch": 0.46223, + "grad_norm": 0.8590824013796482, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 46223 + }, + { + "epoch": 0.46224, + "grad_norm": 1.1384380453408476, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 46224 + }, + { + "epoch": 0.46225, + "grad_norm": 0.9717113888747416, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 46225 + }, + { + "epoch": 0.46226, + "grad_norm": 1.3248217316542714, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 46226 + }, + { + "epoch": 0.46227, + "grad_norm": 0.9296407349011138, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 46227 + }, + { + "epoch": 0.46228, + "grad_norm": 1.1126314479935202, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 46228 + }, + { + "epoch": 0.46229, + "grad_norm": 1.2076020438498647, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 46229 + }, + { + "epoch": 0.4623, + "grad_norm": 1.0544714438404763, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 46230 + }, + { + "epoch": 0.46231, + "grad_norm": 0.886132615546103, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 46231 + }, + { + "epoch": 0.46232, + "grad_norm": 0.9018075441087422, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 46232 + }, + { + "epoch": 0.46233, + "grad_norm": 1.0936024528094095, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 46233 + }, + { + "epoch": 0.46234, + "grad_norm": 1.0778080763446873, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 46234 + }, + { + "epoch": 0.46235, + "grad_norm": 1.1191719061663363, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 46235 + }, + { + "epoch": 0.46236, + "grad_norm": 1.005497158656382, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 46236 + }, + { + "epoch": 0.46237, + "grad_norm": 1.1244568520414833, + "learning_rate": 0.003, + "loss": 4.008, + "step": 46237 + }, + { + "epoch": 0.46238, + "grad_norm": 0.8859996373498125, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 46238 + }, + { + "epoch": 0.46239, + "grad_norm": 0.8667818304087389, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 46239 + }, + { + "epoch": 0.4624, + "grad_norm": 0.9021896587214469, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 46240 + }, + { + "epoch": 0.46241, + "grad_norm": 1.131750764064151, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 46241 + }, + { + "epoch": 0.46242, + "grad_norm": 1.1690602509695533, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 46242 + }, + { + "epoch": 0.46243, + "grad_norm": 0.918160562550634, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 46243 + }, + { + "epoch": 0.46244, + "grad_norm": 1.2190424761417888, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 46244 + }, + { + "epoch": 0.46245, + "grad_norm": 1.0537916550558222, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 46245 + }, + { + "epoch": 0.46246, + "grad_norm": 1.0788476369071651, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 46246 + }, + { + "epoch": 0.46247, + "grad_norm": 1.3247969076894828, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 46247 + }, + { + "epoch": 0.46248, + "grad_norm": 0.884216822628466, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 46248 + }, + { + "epoch": 0.46249, + "grad_norm": 0.9004859250604393, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 46249 + }, + { + "epoch": 0.4625, + "grad_norm": 1.1348227220016507, + "learning_rate": 0.003, + "loss": 4.046, + "step": 46250 + }, + { + "epoch": 0.46251, + "grad_norm": 1.1741185627892514, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 46251 + }, + { + "epoch": 0.46252, + "grad_norm": 0.9432918976930201, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 46252 + }, + { + "epoch": 0.46253, + "grad_norm": 1.1924523711857302, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 46253 + }, + { + "epoch": 0.46254, + "grad_norm": 1.212126903380922, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 46254 + }, + { + "epoch": 0.46255, + "grad_norm": 1.0437995913578297, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 46255 + }, + { + "epoch": 0.46256, + "grad_norm": 0.9946432371776704, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 46256 + }, + { + "epoch": 0.46257, + "grad_norm": 1.330904926514894, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 46257 + }, + { + "epoch": 0.46258, + "grad_norm": 0.891473636051254, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 46258 + }, + { + "epoch": 0.46259, + "grad_norm": 1.0978115653747984, + "learning_rate": 0.003, + "loss": 4.057, + "step": 46259 + }, + { + "epoch": 0.4626, + "grad_norm": 1.1078791160958446, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 46260 + }, + { + "epoch": 0.46261, + "grad_norm": 0.8349486064575591, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 46261 + }, + { + "epoch": 0.46262, + "grad_norm": 0.9498596112512729, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 46262 + }, + { + "epoch": 0.46263, + "grad_norm": 1.1933470735855563, + "learning_rate": 0.003, + "loss": 4.034, + "step": 46263 + }, + { + "epoch": 0.46264, + "grad_norm": 1.038877437832309, + "learning_rate": 0.003, + "loss": 4.036, + "step": 46264 + }, + { + "epoch": 0.46265, + "grad_norm": 1.1794032855787646, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 46265 + }, + { + "epoch": 0.46266, + "grad_norm": 0.9510415319621804, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 46266 + }, + { + "epoch": 0.46267, + "grad_norm": 1.0850661036258524, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 46267 + }, + { + "epoch": 0.46268, + "grad_norm": 1.309646433937659, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 46268 + }, + { + "epoch": 0.46269, + "grad_norm": 1.2246473582926327, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 46269 + }, + { + "epoch": 0.4627, + "grad_norm": 0.9468889764646952, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 46270 + }, + { + "epoch": 0.46271, + "grad_norm": 1.1201090327238856, + "learning_rate": 0.003, + "loss": 4.029, + "step": 46271 + }, + { + "epoch": 0.46272, + "grad_norm": 1.0529833931407158, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 46272 + }, + { + "epoch": 0.46273, + "grad_norm": 1.1357244727653997, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 46273 + }, + { + "epoch": 0.46274, + "grad_norm": 1.1023965021806257, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 46274 + }, + { + "epoch": 0.46275, + "grad_norm": 1.0430991887377787, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 46275 + }, + { + "epoch": 0.46276, + "grad_norm": 1.199343002171924, + "learning_rate": 0.003, + "loss": 4.076, + "step": 46276 + }, + { + "epoch": 0.46277, + "grad_norm": 0.9381911108092224, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 46277 + }, + { + "epoch": 0.46278, + "grad_norm": 1.0397971737933878, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 46278 + }, + { + "epoch": 0.46279, + "grad_norm": 1.0395432443087622, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 46279 + }, + { + "epoch": 0.4628, + "grad_norm": 1.3201286652305824, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 46280 + }, + { + "epoch": 0.46281, + "grad_norm": 1.0943926836934221, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 46281 + }, + { + "epoch": 0.46282, + "grad_norm": 1.0195914958710666, + "learning_rate": 0.003, + "loss": 4.014, + "step": 46282 + }, + { + "epoch": 0.46283, + "grad_norm": 1.0767540335463879, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 46283 + }, + { + "epoch": 0.46284, + "grad_norm": 1.2280432741936544, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 46284 + }, + { + "epoch": 0.46285, + "grad_norm": 0.8124095496141561, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 46285 + }, + { + "epoch": 0.46286, + "grad_norm": 0.9435226038587367, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 46286 + }, + { + "epoch": 0.46287, + "grad_norm": 1.1939511093061035, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 46287 + }, + { + "epoch": 0.46288, + "grad_norm": 1.0574336282582502, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 46288 + }, + { + "epoch": 0.46289, + "grad_norm": 0.9382570275602233, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 46289 + }, + { + "epoch": 0.4629, + "grad_norm": 1.0887483620528964, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 46290 + }, + { + "epoch": 0.46291, + "grad_norm": 1.3210867235319972, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 46291 + }, + { + "epoch": 0.46292, + "grad_norm": 0.985845820554608, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 46292 + }, + { + "epoch": 0.46293, + "grad_norm": 1.155361402335321, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 46293 + }, + { + "epoch": 0.46294, + "grad_norm": 1.1085305527260234, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 46294 + }, + { + "epoch": 0.46295, + "grad_norm": 1.1476473498637414, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 46295 + }, + { + "epoch": 0.46296, + "grad_norm": 1.0004516119248625, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 46296 + }, + { + "epoch": 0.46297, + "grad_norm": 1.038120216035452, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 46297 + }, + { + "epoch": 0.46298, + "grad_norm": 0.9744230951538873, + "learning_rate": 0.003, + "loss": 4.046, + "step": 46298 + }, + { + "epoch": 0.46299, + "grad_norm": 1.048650549561915, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 46299 + }, + { + "epoch": 0.463, + "grad_norm": 1.1651372184258593, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 46300 + }, + { + "epoch": 0.46301, + "grad_norm": 1.2482395164317357, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 46301 + }, + { + "epoch": 0.46302, + "grad_norm": 0.8463809818934351, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 46302 + }, + { + "epoch": 0.46303, + "grad_norm": 0.8221874053359404, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 46303 + }, + { + "epoch": 0.46304, + "grad_norm": 1.1019572353564198, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 46304 + }, + { + "epoch": 0.46305, + "grad_norm": 1.2259715487909295, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 46305 + }, + { + "epoch": 0.46306, + "grad_norm": 0.9847539634805147, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 46306 + }, + { + "epoch": 0.46307, + "grad_norm": 1.2205314438923265, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 46307 + }, + { + "epoch": 0.46308, + "grad_norm": 0.9441683570428093, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 46308 + }, + { + "epoch": 0.46309, + "grad_norm": 0.9743961294142448, + "learning_rate": 0.003, + "loss": 4.025, + "step": 46309 + }, + { + "epoch": 0.4631, + "grad_norm": 1.2485886820440242, + "learning_rate": 0.003, + "loss": 4.054, + "step": 46310 + }, + { + "epoch": 0.46311, + "grad_norm": 1.1941767419242124, + "learning_rate": 0.003, + "loss": 4.026, + "step": 46311 + }, + { + "epoch": 0.46312, + "grad_norm": 1.0942089335773553, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 46312 + }, + { + "epoch": 0.46313, + "grad_norm": 1.149613924186116, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 46313 + }, + { + "epoch": 0.46314, + "grad_norm": 1.0558764260488964, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 46314 + }, + { + "epoch": 0.46315, + "grad_norm": 1.0065835645969192, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 46315 + }, + { + "epoch": 0.46316, + "grad_norm": 1.0275004578931608, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 46316 + }, + { + "epoch": 0.46317, + "grad_norm": 1.0448397618195362, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 46317 + }, + { + "epoch": 0.46318, + "grad_norm": 1.0853400983248354, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 46318 + }, + { + "epoch": 0.46319, + "grad_norm": 1.1946904349120078, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 46319 + }, + { + "epoch": 0.4632, + "grad_norm": 0.8990526309285329, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 46320 + }, + { + "epoch": 0.46321, + "grad_norm": 1.047736730960609, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 46321 + }, + { + "epoch": 0.46322, + "grad_norm": 1.2921570338262682, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 46322 + }, + { + "epoch": 0.46323, + "grad_norm": 1.0895144614314969, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 46323 + }, + { + "epoch": 0.46324, + "grad_norm": 1.0775725374690825, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 46324 + }, + { + "epoch": 0.46325, + "grad_norm": 1.0427652987846021, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 46325 + }, + { + "epoch": 0.46326, + "grad_norm": 1.1761815488248752, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 46326 + }, + { + "epoch": 0.46327, + "grad_norm": 0.8928310861779158, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 46327 + }, + { + "epoch": 0.46328, + "grad_norm": 1.0777517331739046, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 46328 + }, + { + "epoch": 0.46329, + "grad_norm": 1.2772488230724939, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 46329 + }, + { + "epoch": 0.4633, + "grad_norm": 1.0137928467005672, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 46330 + }, + { + "epoch": 0.46331, + "grad_norm": 1.1372349134948583, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 46331 + }, + { + "epoch": 0.46332, + "grad_norm": 1.0455075198898451, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 46332 + }, + { + "epoch": 0.46333, + "grad_norm": 1.2882265717003962, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 46333 + }, + { + "epoch": 0.46334, + "grad_norm": 1.1226790397037931, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 46334 + }, + { + "epoch": 0.46335, + "grad_norm": 0.8237426173546102, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 46335 + }, + { + "epoch": 0.46336, + "grad_norm": 0.7165749392669197, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 46336 + }, + { + "epoch": 0.46337, + "grad_norm": 0.7111456479488522, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 46337 + }, + { + "epoch": 0.46338, + "grad_norm": 0.8749226597513722, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 46338 + }, + { + "epoch": 0.46339, + "grad_norm": 1.0307677364129755, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 46339 + }, + { + "epoch": 0.4634, + "grad_norm": 1.2791609899539647, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 46340 + }, + { + "epoch": 0.46341, + "grad_norm": 0.9236253119050006, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 46341 + }, + { + "epoch": 0.46342, + "grad_norm": 1.0175366253981692, + "learning_rate": 0.003, + "loss": 4.035, + "step": 46342 + }, + { + "epoch": 0.46343, + "grad_norm": 1.179258919397755, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 46343 + }, + { + "epoch": 0.46344, + "grad_norm": 0.9975795449804554, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 46344 + }, + { + "epoch": 0.46345, + "grad_norm": 1.2620674325926735, + "learning_rate": 0.003, + "loss": 4.068, + "step": 46345 + }, + { + "epoch": 0.46346, + "grad_norm": 0.9151938815834804, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 46346 + }, + { + "epoch": 0.46347, + "grad_norm": 0.983416291819887, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 46347 + }, + { + "epoch": 0.46348, + "grad_norm": 0.9467203952686452, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 46348 + }, + { + "epoch": 0.46349, + "grad_norm": 1.0858568921191376, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 46349 + }, + { + "epoch": 0.4635, + "grad_norm": 1.0790358529130468, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 46350 + }, + { + "epoch": 0.46351, + "grad_norm": 1.132305430477665, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 46351 + }, + { + "epoch": 0.46352, + "grad_norm": 1.1300651170145954, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 46352 + }, + { + "epoch": 0.46353, + "grad_norm": 1.4987762114568948, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 46353 + }, + { + "epoch": 0.46354, + "grad_norm": 0.8038189132227392, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 46354 + }, + { + "epoch": 0.46355, + "grad_norm": 0.8013509984479823, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 46355 + }, + { + "epoch": 0.46356, + "grad_norm": 0.940687583790039, + "learning_rate": 0.003, + "loss": 4.028, + "step": 46356 + }, + { + "epoch": 0.46357, + "grad_norm": 1.160188469171704, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 46357 + }, + { + "epoch": 0.46358, + "grad_norm": 1.0412100324271623, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 46358 + }, + { + "epoch": 0.46359, + "grad_norm": 1.1881219086344434, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 46359 + }, + { + "epoch": 0.4636, + "grad_norm": 1.0851273023131618, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 46360 + }, + { + "epoch": 0.46361, + "grad_norm": 1.3005104525790512, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 46361 + }, + { + "epoch": 0.46362, + "grad_norm": 0.9067472712928416, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 46362 + }, + { + "epoch": 0.46363, + "grad_norm": 0.9755246070529096, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 46363 + }, + { + "epoch": 0.46364, + "grad_norm": 1.3344797274957876, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 46364 + }, + { + "epoch": 0.46365, + "grad_norm": 0.8894857947481591, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 46365 + }, + { + "epoch": 0.46366, + "grad_norm": 0.9147419801893456, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 46366 + }, + { + "epoch": 0.46367, + "grad_norm": 1.0563701962629004, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 46367 + }, + { + "epoch": 0.46368, + "grad_norm": 1.0386212036404012, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 46368 + }, + { + "epoch": 0.46369, + "grad_norm": 1.2414289201466309, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 46369 + }, + { + "epoch": 0.4637, + "grad_norm": 1.1501360513321663, + "learning_rate": 0.003, + "loss": 4.029, + "step": 46370 + }, + { + "epoch": 0.46371, + "grad_norm": 1.1882448318475138, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 46371 + }, + { + "epoch": 0.46372, + "grad_norm": 1.2893846443332904, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 46372 + }, + { + "epoch": 0.46373, + "grad_norm": 0.9096465842318022, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 46373 + }, + { + "epoch": 0.46374, + "grad_norm": 0.8355908141260394, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 46374 + }, + { + "epoch": 0.46375, + "grad_norm": 0.8945928471327829, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 46375 + }, + { + "epoch": 0.46376, + "grad_norm": 0.9269015713666763, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 46376 + }, + { + "epoch": 0.46377, + "grad_norm": 0.9451402288016122, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 46377 + }, + { + "epoch": 0.46378, + "grad_norm": 1.0207408178766968, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 46378 + }, + { + "epoch": 0.46379, + "grad_norm": 1.127398075927535, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 46379 + }, + { + "epoch": 0.4638, + "grad_norm": 1.082238969588087, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 46380 + }, + { + "epoch": 0.46381, + "grad_norm": 1.1028438256303568, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 46381 + }, + { + "epoch": 0.46382, + "grad_norm": 1.1033139743020401, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 46382 + }, + { + "epoch": 0.46383, + "grad_norm": 1.073600777363018, + "learning_rate": 0.003, + "loss": 3.998, + "step": 46383 + }, + { + "epoch": 0.46384, + "grad_norm": 1.17132147364399, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 46384 + }, + { + "epoch": 0.46385, + "grad_norm": 1.164084218702746, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 46385 + }, + { + "epoch": 0.46386, + "grad_norm": 1.0230577970456531, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 46386 + }, + { + "epoch": 0.46387, + "grad_norm": 1.151003492463692, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 46387 + }, + { + "epoch": 0.46388, + "grad_norm": 0.9067444191992098, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 46388 + }, + { + "epoch": 0.46389, + "grad_norm": 1.040839335376295, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 46389 + }, + { + "epoch": 0.4639, + "grad_norm": 1.3143819385903837, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 46390 + }, + { + "epoch": 0.46391, + "grad_norm": 0.783334545415341, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 46391 + }, + { + "epoch": 0.46392, + "grad_norm": 0.7397336215377457, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 46392 + }, + { + "epoch": 0.46393, + "grad_norm": 0.9259477666528009, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 46393 + }, + { + "epoch": 0.46394, + "grad_norm": 1.0442335599720918, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 46394 + }, + { + "epoch": 0.46395, + "grad_norm": 1.1794346316086006, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 46395 + }, + { + "epoch": 0.46396, + "grad_norm": 1.1785305358278817, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 46396 + }, + { + "epoch": 0.46397, + "grad_norm": 1.0876917033277547, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 46397 + }, + { + "epoch": 0.46398, + "grad_norm": 1.0243676315258157, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 46398 + }, + { + "epoch": 0.46399, + "grad_norm": 1.1442387925814372, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 46399 + }, + { + "epoch": 0.464, + "grad_norm": 0.9748139692204988, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 46400 + }, + { + "epoch": 0.46401, + "grad_norm": 0.9367088077443403, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 46401 + }, + { + "epoch": 0.46402, + "grad_norm": 1.0154610307005858, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 46402 + }, + { + "epoch": 0.46403, + "grad_norm": 1.200714950247328, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 46403 + }, + { + "epoch": 0.46404, + "grad_norm": 0.9532315004800196, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 46404 + }, + { + "epoch": 0.46405, + "grad_norm": 0.9797422004026564, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 46405 + }, + { + "epoch": 0.46406, + "grad_norm": 1.2396786220616722, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 46406 + }, + { + "epoch": 0.46407, + "grad_norm": 0.8983781670461596, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 46407 + }, + { + "epoch": 0.46408, + "grad_norm": 0.9305526042075858, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 46408 + }, + { + "epoch": 0.46409, + "grad_norm": 0.9934389183148983, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 46409 + }, + { + "epoch": 0.4641, + "grad_norm": 1.2658388052693406, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 46410 + }, + { + "epoch": 0.46411, + "grad_norm": 0.9684393372741775, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 46411 + }, + { + "epoch": 0.46412, + "grad_norm": 0.9365902414752273, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 46412 + }, + { + "epoch": 0.46413, + "grad_norm": 1.1561119901769803, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 46413 + }, + { + "epoch": 0.46414, + "grad_norm": 1.2727290403089924, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 46414 + }, + { + "epoch": 0.46415, + "grad_norm": 0.9187644431700737, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 46415 + }, + { + "epoch": 0.46416, + "grad_norm": 1.194791928406179, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 46416 + }, + { + "epoch": 0.46417, + "grad_norm": 1.2767152028860698, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 46417 + }, + { + "epoch": 0.46418, + "grad_norm": 1.0795737126957983, + "learning_rate": 0.003, + "loss": 4.07, + "step": 46418 + }, + { + "epoch": 0.46419, + "grad_norm": 1.292381256191984, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 46419 + }, + { + "epoch": 0.4642, + "grad_norm": 0.937891109347911, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 46420 + }, + { + "epoch": 0.46421, + "grad_norm": 0.8649324933257405, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 46421 + }, + { + "epoch": 0.46422, + "grad_norm": 1.0643298401773142, + "learning_rate": 0.003, + "loss": 4.031, + "step": 46422 + }, + { + "epoch": 0.46423, + "grad_norm": 1.3930276357051516, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 46423 + }, + { + "epoch": 0.46424, + "grad_norm": 0.8108498623306873, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 46424 + }, + { + "epoch": 0.46425, + "grad_norm": 0.6542749546263773, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 46425 + }, + { + "epoch": 0.46426, + "grad_norm": 0.6969582361724288, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 46426 + }, + { + "epoch": 0.46427, + "grad_norm": 0.6758695195829157, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 46427 + }, + { + "epoch": 0.46428, + "grad_norm": 0.8023820299094765, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 46428 + }, + { + "epoch": 0.46429, + "grad_norm": 1.1665468655029672, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 46429 + }, + { + "epoch": 0.4643, + "grad_norm": 1.2176921234306586, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 46430 + }, + { + "epoch": 0.46431, + "grad_norm": 1.0253411211137993, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 46431 + }, + { + "epoch": 0.46432, + "grad_norm": 1.111316722289323, + "learning_rate": 0.003, + "loss": 4.025, + "step": 46432 + }, + { + "epoch": 0.46433, + "grad_norm": 1.2217809122501027, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 46433 + }, + { + "epoch": 0.46434, + "grad_norm": 1.0727724025783436, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 46434 + }, + { + "epoch": 0.46435, + "grad_norm": 1.0749421850090473, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 46435 + }, + { + "epoch": 0.46436, + "grad_norm": 0.8712185772714388, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 46436 + }, + { + "epoch": 0.46437, + "grad_norm": 0.8757572890729431, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 46437 + }, + { + "epoch": 0.46438, + "grad_norm": 1.0505495826035218, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 46438 + }, + { + "epoch": 0.46439, + "grad_norm": 0.9732242293180604, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 46439 + }, + { + "epoch": 0.4644, + "grad_norm": 1.0635007174190216, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 46440 + }, + { + "epoch": 0.46441, + "grad_norm": 1.3091660561520395, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 46441 + }, + { + "epoch": 0.46442, + "grad_norm": 1.1045552259733624, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 46442 + }, + { + "epoch": 0.46443, + "grad_norm": 1.1572808500282763, + "learning_rate": 0.003, + "loss": 4.052, + "step": 46443 + }, + { + "epoch": 0.46444, + "grad_norm": 1.0350227183954657, + "learning_rate": 0.003, + "loss": 4.037, + "step": 46444 + }, + { + "epoch": 0.46445, + "grad_norm": 1.2148600051253637, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 46445 + }, + { + "epoch": 0.46446, + "grad_norm": 1.1985225173173448, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 46446 + }, + { + "epoch": 0.46447, + "grad_norm": 1.2146358438971787, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 46447 + }, + { + "epoch": 0.46448, + "grad_norm": 1.093904555863501, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 46448 + }, + { + "epoch": 0.46449, + "grad_norm": 0.8698892307939837, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 46449 + }, + { + "epoch": 0.4645, + "grad_norm": 0.8012882955062596, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 46450 + }, + { + "epoch": 0.46451, + "grad_norm": 0.8037182752592091, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 46451 + }, + { + "epoch": 0.46452, + "grad_norm": 1.065235419342574, + "learning_rate": 0.003, + "loss": 4.026, + "step": 46452 + }, + { + "epoch": 0.46453, + "grad_norm": 1.235855130768875, + "learning_rate": 0.003, + "loss": 4.04, + "step": 46453 + }, + { + "epoch": 0.46454, + "grad_norm": 1.0400707117051489, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 46454 + }, + { + "epoch": 0.46455, + "grad_norm": 1.0178131189929638, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 46455 + }, + { + "epoch": 0.46456, + "grad_norm": 1.2775084446961074, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 46456 + }, + { + "epoch": 0.46457, + "grad_norm": 1.1091872786720165, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 46457 + }, + { + "epoch": 0.46458, + "grad_norm": 0.9794340487911527, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 46458 + }, + { + "epoch": 0.46459, + "grad_norm": 1.0000437238750084, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 46459 + }, + { + "epoch": 0.4646, + "grad_norm": 1.0726757062763088, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 46460 + }, + { + "epoch": 0.46461, + "grad_norm": 1.3142346936648264, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 46461 + }, + { + "epoch": 0.46462, + "grad_norm": 0.9510265250384469, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 46462 + }, + { + "epoch": 0.46463, + "grad_norm": 1.1884746281740957, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 46463 + }, + { + "epoch": 0.46464, + "grad_norm": 1.299279568061839, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 46464 + }, + { + "epoch": 0.46465, + "grad_norm": 0.965742354389979, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 46465 + }, + { + "epoch": 0.46466, + "grad_norm": 0.9994483736665078, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 46466 + }, + { + "epoch": 0.46467, + "grad_norm": 1.372197859848804, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 46467 + }, + { + "epoch": 0.46468, + "grad_norm": 0.8420820344343111, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 46468 + }, + { + "epoch": 0.46469, + "grad_norm": 0.9161877443561106, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 46469 + }, + { + "epoch": 0.4647, + "grad_norm": 1.1718627110178874, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 46470 + }, + { + "epoch": 0.46471, + "grad_norm": 1.0162021291842953, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 46471 + }, + { + "epoch": 0.46472, + "grad_norm": 1.1009573779184485, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 46472 + }, + { + "epoch": 0.46473, + "grad_norm": 0.9762625719782411, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 46473 + }, + { + "epoch": 0.46474, + "grad_norm": 1.08826708105795, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 46474 + }, + { + "epoch": 0.46475, + "grad_norm": 0.9110418354680925, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 46475 + }, + { + "epoch": 0.46476, + "grad_norm": 0.9832812577079363, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 46476 + }, + { + "epoch": 0.46477, + "grad_norm": 1.688878194832363, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 46477 + }, + { + "epoch": 0.46478, + "grad_norm": 0.8639378634526023, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 46478 + }, + { + "epoch": 0.46479, + "grad_norm": 0.9158599646144387, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 46479 + }, + { + "epoch": 0.4648, + "grad_norm": 1.0584872642238845, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 46480 + }, + { + "epoch": 0.46481, + "grad_norm": 1.1451376929621246, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 46481 + }, + { + "epoch": 0.46482, + "grad_norm": 0.9566118051188883, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 46482 + }, + { + "epoch": 0.46483, + "grad_norm": 1.1482185474543802, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 46483 + }, + { + "epoch": 0.46484, + "grad_norm": 1.1335708949613452, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 46484 + }, + { + "epoch": 0.46485, + "grad_norm": 1.239018304818821, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 46485 + }, + { + "epoch": 0.46486, + "grad_norm": 1.0596111008920173, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 46486 + }, + { + "epoch": 0.46487, + "grad_norm": 1.2197009065739441, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 46487 + }, + { + "epoch": 0.46488, + "grad_norm": 1.1148083483623583, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 46488 + }, + { + "epoch": 0.46489, + "grad_norm": 0.9511600151858601, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 46489 + }, + { + "epoch": 0.4649, + "grad_norm": 1.106547065745151, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 46490 + }, + { + "epoch": 0.46491, + "grad_norm": 1.1646894197649258, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 46491 + }, + { + "epoch": 0.46492, + "grad_norm": 0.9836031674137408, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 46492 + }, + { + "epoch": 0.46493, + "grad_norm": 1.2519890163033565, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 46493 + }, + { + "epoch": 0.46494, + "grad_norm": 0.7573787461423938, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 46494 + }, + { + "epoch": 0.46495, + "grad_norm": 0.9760293742907354, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 46495 + }, + { + "epoch": 0.46496, + "grad_norm": 1.2010033242844846, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 46496 + }, + { + "epoch": 0.46497, + "grad_norm": 0.8978028783477212, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 46497 + }, + { + "epoch": 0.46498, + "grad_norm": 1.3190193472644167, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 46498 + }, + { + "epoch": 0.46499, + "grad_norm": 0.939023743266933, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 46499 + }, + { + "epoch": 0.465, + "grad_norm": 0.979598732978843, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 46500 + }, + { + "epoch": 0.46501, + "grad_norm": 1.1411554459400446, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 46501 + }, + { + "epoch": 0.46502, + "grad_norm": 0.8640315346650833, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 46502 + }, + { + "epoch": 0.46503, + "grad_norm": 0.9639210434015985, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 46503 + }, + { + "epoch": 0.46504, + "grad_norm": 1.2376192091951643, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 46504 + }, + { + "epoch": 0.46505, + "grad_norm": 0.8576680619477509, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 46505 + }, + { + "epoch": 0.46506, + "grad_norm": 0.8811745566072156, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 46506 + }, + { + "epoch": 0.46507, + "grad_norm": 1.1257191804083215, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 46507 + }, + { + "epoch": 0.46508, + "grad_norm": 0.9987861037178993, + "learning_rate": 0.003, + "loss": 4.055, + "step": 46508 + }, + { + "epoch": 0.46509, + "grad_norm": 1.0686793374324457, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 46509 + }, + { + "epoch": 0.4651, + "grad_norm": 1.0764530836988324, + "learning_rate": 0.003, + "loss": 4.041, + "step": 46510 + }, + { + "epoch": 0.46511, + "grad_norm": 0.9416744993692994, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 46511 + }, + { + "epoch": 0.46512, + "grad_norm": 0.993464006821849, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 46512 + }, + { + "epoch": 0.46513, + "grad_norm": 1.0239325693142955, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 46513 + }, + { + "epoch": 0.46514, + "grad_norm": 1.48013329627476, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 46514 + }, + { + "epoch": 0.46515, + "grad_norm": 0.9487500450438149, + "learning_rate": 0.003, + "loss": 4.02, + "step": 46515 + }, + { + "epoch": 0.46516, + "grad_norm": 1.0709002220479504, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 46516 + }, + { + "epoch": 0.46517, + "grad_norm": 1.469140382891928, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 46517 + }, + { + "epoch": 0.46518, + "grad_norm": 0.921378184167737, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 46518 + }, + { + "epoch": 0.46519, + "grad_norm": 0.9175663709222773, + "learning_rate": 0.003, + "loss": 4.036, + "step": 46519 + }, + { + "epoch": 0.4652, + "grad_norm": 1.2361561267023204, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 46520 + }, + { + "epoch": 0.46521, + "grad_norm": 1.3075173765244998, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 46521 + }, + { + "epoch": 0.46522, + "grad_norm": 1.1388013051641377, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 46522 + }, + { + "epoch": 0.46523, + "grad_norm": 1.1474781298656618, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 46523 + }, + { + "epoch": 0.46524, + "grad_norm": 1.1476821665605665, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 46524 + }, + { + "epoch": 0.46525, + "grad_norm": 0.9011872977518762, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 46525 + }, + { + "epoch": 0.46526, + "grad_norm": 1.0556521438861788, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 46526 + }, + { + "epoch": 0.46527, + "grad_norm": 0.9275735630630918, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 46527 + }, + { + "epoch": 0.46528, + "grad_norm": 1.1324305523727372, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 46528 + }, + { + "epoch": 0.46529, + "grad_norm": 1.3368607385291928, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 46529 + }, + { + "epoch": 0.4653, + "grad_norm": 1.0334186130106078, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 46530 + }, + { + "epoch": 0.46531, + "grad_norm": 1.0329922170078947, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 46531 + }, + { + "epoch": 0.46532, + "grad_norm": 1.1160259790776994, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 46532 + }, + { + "epoch": 0.46533, + "grad_norm": 1.009694989235886, + "learning_rate": 0.003, + "loss": 4.01, + "step": 46533 + }, + { + "epoch": 0.46534, + "grad_norm": 1.1337052036545958, + "learning_rate": 0.003, + "loss": 4.062, + "step": 46534 + }, + { + "epoch": 0.46535, + "grad_norm": 0.9702997026757493, + "learning_rate": 0.003, + "loss": 4.026, + "step": 46535 + }, + { + "epoch": 0.46536, + "grad_norm": 1.116569631752525, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 46536 + }, + { + "epoch": 0.46537, + "grad_norm": 0.9590331100691956, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 46537 + }, + { + "epoch": 0.46538, + "grad_norm": 1.0626192707224524, + "learning_rate": 0.003, + "loss": 4.045, + "step": 46538 + }, + { + "epoch": 0.46539, + "grad_norm": 1.1038212590397156, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 46539 + }, + { + "epoch": 0.4654, + "grad_norm": 1.1913667785083824, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 46540 + }, + { + "epoch": 0.46541, + "grad_norm": 0.9078915705928052, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 46541 + }, + { + "epoch": 0.46542, + "grad_norm": 1.0401121892166274, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 46542 + }, + { + "epoch": 0.46543, + "grad_norm": 1.328348071016244, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 46543 + }, + { + "epoch": 0.46544, + "grad_norm": 0.9022599698511766, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 46544 + }, + { + "epoch": 0.46545, + "grad_norm": 1.1106475523613715, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 46545 + }, + { + "epoch": 0.46546, + "grad_norm": 1.2291831276940168, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 46546 + }, + { + "epoch": 0.46547, + "grad_norm": 0.8539317360283393, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 46547 + }, + { + "epoch": 0.46548, + "grad_norm": 0.8416521196041803, + "learning_rate": 0.003, + "loss": 4.035, + "step": 46548 + }, + { + "epoch": 0.46549, + "grad_norm": 0.9602975994580165, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 46549 + }, + { + "epoch": 0.4655, + "grad_norm": 1.1083451203853947, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 46550 + }, + { + "epoch": 0.46551, + "grad_norm": 1.3140518649486044, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 46551 + }, + { + "epoch": 0.46552, + "grad_norm": 1.0307527640902614, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 46552 + }, + { + "epoch": 0.46553, + "grad_norm": 1.0314066162787816, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 46553 + }, + { + "epoch": 0.46554, + "grad_norm": 1.0645547160849682, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 46554 + }, + { + "epoch": 0.46555, + "grad_norm": 1.0155579317817334, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 46555 + }, + { + "epoch": 0.46556, + "grad_norm": 1.1877317651432593, + "learning_rate": 0.003, + "loss": 4.041, + "step": 46556 + }, + { + "epoch": 0.46557, + "grad_norm": 0.9644622549761408, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 46557 + }, + { + "epoch": 0.46558, + "grad_norm": 1.116155125454559, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 46558 + }, + { + "epoch": 0.46559, + "grad_norm": 1.053908804577595, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 46559 + }, + { + "epoch": 0.4656, + "grad_norm": 1.2849673833070252, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 46560 + }, + { + "epoch": 0.46561, + "grad_norm": 1.0624651558095028, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 46561 + }, + { + "epoch": 0.46562, + "grad_norm": 1.17632281338543, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 46562 + }, + { + "epoch": 0.46563, + "grad_norm": 0.9722523322771007, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 46563 + }, + { + "epoch": 0.46564, + "grad_norm": 1.0715845104557988, + "learning_rate": 0.003, + "loss": 4.046, + "step": 46564 + }, + { + "epoch": 0.46565, + "grad_norm": 1.0242481400551853, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 46565 + }, + { + "epoch": 0.46566, + "grad_norm": 1.1720145543081808, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 46566 + }, + { + "epoch": 0.46567, + "grad_norm": 0.994773523650883, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 46567 + }, + { + "epoch": 0.46568, + "grad_norm": 1.142200312682565, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 46568 + }, + { + "epoch": 0.46569, + "grad_norm": 1.1785419749321702, + "learning_rate": 0.003, + "loss": 4.043, + "step": 46569 + }, + { + "epoch": 0.4657, + "grad_norm": 0.8497455308450613, + "learning_rate": 0.003, + "loss": 4.048, + "step": 46570 + }, + { + "epoch": 0.46571, + "grad_norm": 0.8868355419728186, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 46571 + }, + { + "epoch": 0.46572, + "grad_norm": 1.0289595859550766, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 46572 + }, + { + "epoch": 0.46573, + "grad_norm": 1.3025048456640642, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 46573 + }, + { + "epoch": 0.46574, + "grad_norm": 0.8887876369798423, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 46574 + }, + { + "epoch": 0.46575, + "grad_norm": 1.0273549973013574, + "learning_rate": 0.003, + "loss": 4.058, + "step": 46575 + }, + { + "epoch": 0.46576, + "grad_norm": 1.1900244922536591, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 46576 + }, + { + "epoch": 0.46577, + "grad_norm": 0.9172954083148558, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 46577 + }, + { + "epoch": 0.46578, + "grad_norm": 1.1256968158638416, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 46578 + }, + { + "epoch": 0.46579, + "grad_norm": 1.1906639044975385, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 46579 + }, + { + "epoch": 0.4658, + "grad_norm": 0.9457771876867831, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 46580 + }, + { + "epoch": 0.46581, + "grad_norm": 1.0533022850315175, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 46581 + }, + { + "epoch": 0.46582, + "grad_norm": 1.0379653640269648, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 46582 + }, + { + "epoch": 0.46583, + "grad_norm": 0.9726212388767715, + "learning_rate": 0.003, + "loss": 4.036, + "step": 46583 + }, + { + "epoch": 0.46584, + "grad_norm": 1.0749591731419328, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 46584 + }, + { + "epoch": 0.46585, + "grad_norm": 1.2752637004146812, + "learning_rate": 0.003, + "loss": 4.022, + "step": 46585 + }, + { + "epoch": 0.46586, + "grad_norm": 0.8846293854118655, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 46586 + }, + { + "epoch": 0.46587, + "grad_norm": 0.8414554017816905, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 46587 + }, + { + "epoch": 0.46588, + "grad_norm": 1.1076267041716592, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 46588 + }, + { + "epoch": 0.46589, + "grad_norm": 1.3351650759076317, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 46589 + }, + { + "epoch": 0.4659, + "grad_norm": 1.02038989569386, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 46590 + }, + { + "epoch": 0.46591, + "grad_norm": 1.117437924799319, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 46591 + }, + { + "epoch": 0.46592, + "grad_norm": 1.0866781934176297, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 46592 + }, + { + "epoch": 0.46593, + "grad_norm": 0.8798468571223352, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 46593 + }, + { + "epoch": 0.46594, + "grad_norm": 0.8032225151107014, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 46594 + }, + { + "epoch": 0.46595, + "grad_norm": 0.748430731312727, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 46595 + }, + { + "epoch": 0.46596, + "grad_norm": 0.6945902619356868, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 46596 + }, + { + "epoch": 0.46597, + "grad_norm": 0.7270601634545792, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 46597 + }, + { + "epoch": 0.46598, + "grad_norm": 0.8649639665268871, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 46598 + }, + { + "epoch": 0.46599, + "grad_norm": 1.0660847425888207, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 46599 + }, + { + "epoch": 0.466, + "grad_norm": 1.0891248021146642, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 46600 + }, + { + "epoch": 0.46601, + "grad_norm": 1.3220261434842213, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 46601 + }, + { + "epoch": 0.46602, + "grad_norm": 0.7790010165770354, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 46602 + }, + { + "epoch": 0.46603, + "grad_norm": 0.6672948111759329, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 46603 + }, + { + "epoch": 0.46604, + "grad_norm": 0.7392592127250542, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 46604 + }, + { + "epoch": 0.46605, + "grad_norm": 0.8991151676710426, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 46605 + }, + { + "epoch": 0.46606, + "grad_norm": 1.2296064381996397, + "learning_rate": 0.003, + "loss": 4.051, + "step": 46606 + }, + { + "epoch": 0.46607, + "grad_norm": 0.931749454512497, + "learning_rate": 0.003, + "loss": 4.045, + "step": 46607 + }, + { + "epoch": 0.46608, + "grad_norm": 1.1465016751133472, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 46608 + }, + { + "epoch": 0.46609, + "grad_norm": 0.98726501969278, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 46609 + }, + { + "epoch": 0.4661, + "grad_norm": 0.9130732692194538, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 46610 + }, + { + "epoch": 0.46611, + "grad_norm": 1.2546029649094634, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 46611 + }, + { + "epoch": 0.46612, + "grad_norm": 1.010066159226664, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 46612 + }, + { + "epoch": 0.46613, + "grad_norm": 1.2889514676719147, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 46613 + }, + { + "epoch": 0.46614, + "grad_norm": 1.0405366980782271, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 46614 + }, + { + "epoch": 0.46615, + "grad_norm": 1.31537617372955, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 46615 + }, + { + "epoch": 0.46616, + "grad_norm": 1.330076397906051, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 46616 + }, + { + "epoch": 0.46617, + "grad_norm": 0.8297746109711972, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 46617 + }, + { + "epoch": 0.46618, + "grad_norm": 0.7953172552876963, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 46618 + }, + { + "epoch": 0.46619, + "grad_norm": 1.0259281376790612, + "learning_rate": 0.003, + "loss": 4.036, + "step": 46619 + }, + { + "epoch": 0.4662, + "grad_norm": 1.6081190141610728, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 46620 + }, + { + "epoch": 0.46621, + "grad_norm": 0.9528013517561519, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 46621 + }, + { + "epoch": 0.46622, + "grad_norm": 1.1959883304379815, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 46622 + }, + { + "epoch": 0.46623, + "grad_norm": 1.1054814661365737, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 46623 + }, + { + "epoch": 0.46624, + "grad_norm": 1.1785672170039196, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 46624 + }, + { + "epoch": 0.46625, + "grad_norm": 0.9150724720449803, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 46625 + }, + { + "epoch": 0.46626, + "grad_norm": 0.8817136311613796, + "learning_rate": 0.003, + "loss": 4.057, + "step": 46626 + }, + { + "epoch": 0.46627, + "grad_norm": 0.9625024889348848, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 46627 + }, + { + "epoch": 0.46628, + "grad_norm": 1.2266509261669203, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 46628 + }, + { + "epoch": 0.46629, + "grad_norm": 0.9329383453813184, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 46629 + }, + { + "epoch": 0.4663, + "grad_norm": 1.1655529271361103, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 46630 + }, + { + "epoch": 0.46631, + "grad_norm": 1.1089747452809646, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 46631 + }, + { + "epoch": 0.46632, + "grad_norm": 0.9969921687855463, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 46632 + }, + { + "epoch": 0.46633, + "grad_norm": 1.018898861121458, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 46633 + }, + { + "epoch": 0.46634, + "grad_norm": 1.0472265060678454, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 46634 + }, + { + "epoch": 0.46635, + "grad_norm": 1.0097634449024555, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 46635 + }, + { + "epoch": 0.46636, + "grad_norm": 1.267196319399178, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 46636 + }, + { + "epoch": 0.46637, + "grad_norm": 0.9662781095595093, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 46637 + }, + { + "epoch": 0.46638, + "grad_norm": 0.9495983835591474, + "learning_rate": 0.003, + "loss": 4.043, + "step": 46638 + }, + { + "epoch": 0.46639, + "grad_norm": 0.9537839552402765, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 46639 + }, + { + "epoch": 0.4664, + "grad_norm": 1.2388921911262694, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 46640 + }, + { + "epoch": 0.46641, + "grad_norm": 1.0128366524908554, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 46641 + }, + { + "epoch": 0.46642, + "grad_norm": 0.9638557275795356, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 46642 + }, + { + "epoch": 0.46643, + "grad_norm": 0.9713232759569081, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 46643 + }, + { + "epoch": 0.46644, + "grad_norm": 0.9549947182655396, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 46644 + }, + { + "epoch": 0.46645, + "grad_norm": 0.9192840854233916, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 46645 + }, + { + "epoch": 0.46646, + "grad_norm": 1.0935609388814314, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 46646 + }, + { + "epoch": 0.46647, + "grad_norm": 0.9712918567714289, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 46647 + }, + { + "epoch": 0.46648, + "grad_norm": 1.2269700367213119, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 46648 + }, + { + "epoch": 0.46649, + "grad_norm": 1.0695286072872574, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 46649 + }, + { + "epoch": 0.4665, + "grad_norm": 1.1071958845664054, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 46650 + }, + { + "epoch": 0.46651, + "grad_norm": 1.1417939849606948, + "learning_rate": 0.003, + "loss": 4.027, + "step": 46651 + }, + { + "epoch": 0.46652, + "grad_norm": 1.1168721878171506, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 46652 + }, + { + "epoch": 0.46653, + "grad_norm": 1.133863114946586, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 46653 + }, + { + "epoch": 0.46654, + "grad_norm": 1.078330535573663, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 46654 + }, + { + "epoch": 0.46655, + "grad_norm": 1.1494607962485555, + "learning_rate": 0.003, + "loss": 4.044, + "step": 46655 + }, + { + "epoch": 0.46656, + "grad_norm": 1.4835046473656865, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 46656 + }, + { + "epoch": 0.46657, + "grad_norm": 0.9721603911939395, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 46657 + }, + { + "epoch": 0.46658, + "grad_norm": 1.2678062193868667, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 46658 + }, + { + "epoch": 0.46659, + "grad_norm": 1.1327548999847388, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 46659 + }, + { + "epoch": 0.4666, + "grad_norm": 1.076836573468753, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 46660 + }, + { + "epoch": 0.46661, + "grad_norm": 1.0914637703514625, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 46661 + }, + { + "epoch": 0.46662, + "grad_norm": 0.9668627156131295, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 46662 + }, + { + "epoch": 0.46663, + "grad_norm": 0.9932588280843861, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 46663 + }, + { + "epoch": 0.46664, + "grad_norm": 1.1659147977444146, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 46664 + }, + { + "epoch": 0.46665, + "grad_norm": 1.1376367023243061, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 46665 + }, + { + "epoch": 0.46666, + "grad_norm": 1.0956201124272476, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 46666 + }, + { + "epoch": 0.46667, + "grad_norm": 1.1717146485996974, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 46667 + }, + { + "epoch": 0.46668, + "grad_norm": 1.2930726574088116, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 46668 + }, + { + "epoch": 0.46669, + "grad_norm": 0.9608488519922476, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 46669 + }, + { + "epoch": 0.4667, + "grad_norm": 1.122860754552751, + "learning_rate": 0.003, + "loss": 4.033, + "step": 46670 + }, + { + "epoch": 0.46671, + "grad_norm": 1.1636353648872602, + "learning_rate": 0.003, + "loss": 4.053, + "step": 46671 + }, + { + "epoch": 0.46672, + "grad_norm": 0.9512013892697577, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 46672 + }, + { + "epoch": 0.46673, + "grad_norm": 0.8613787435241542, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 46673 + }, + { + "epoch": 0.46674, + "grad_norm": 1.0157132673351685, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 46674 + }, + { + "epoch": 0.46675, + "grad_norm": 1.3295563702339404, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 46675 + }, + { + "epoch": 0.46676, + "grad_norm": 0.8573526431885661, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 46676 + }, + { + "epoch": 0.46677, + "grad_norm": 0.8999919487427105, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 46677 + }, + { + "epoch": 0.46678, + "grad_norm": 1.0374920341548655, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 46678 + }, + { + "epoch": 0.46679, + "grad_norm": 1.1128055584041363, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 46679 + }, + { + "epoch": 0.4668, + "grad_norm": 1.0191961175956101, + "learning_rate": 0.003, + "loss": 4.038, + "step": 46680 + }, + { + "epoch": 0.46681, + "grad_norm": 1.0977250662450497, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 46681 + }, + { + "epoch": 0.46682, + "grad_norm": 1.0523655472124258, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 46682 + }, + { + "epoch": 0.46683, + "grad_norm": 1.0038350145074915, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 46683 + }, + { + "epoch": 0.46684, + "grad_norm": 1.2282999027875752, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 46684 + }, + { + "epoch": 0.46685, + "grad_norm": 1.2115321948343136, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 46685 + }, + { + "epoch": 0.46686, + "grad_norm": 1.0907725916752722, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 46686 + }, + { + "epoch": 0.46687, + "grad_norm": 1.037795134910069, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 46687 + }, + { + "epoch": 0.46688, + "grad_norm": 1.0043610796910574, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 46688 + }, + { + "epoch": 0.46689, + "grad_norm": 1.0473526381395872, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 46689 + }, + { + "epoch": 0.4669, + "grad_norm": 0.8986663288657971, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 46690 + }, + { + "epoch": 0.46691, + "grad_norm": 1.0145962571111062, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 46691 + }, + { + "epoch": 0.46692, + "grad_norm": 1.0992709329677581, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 46692 + }, + { + "epoch": 0.46693, + "grad_norm": 0.9306078319200837, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 46693 + }, + { + "epoch": 0.46694, + "grad_norm": 1.0646563167035918, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 46694 + }, + { + "epoch": 0.46695, + "grad_norm": 1.0565115454343326, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 46695 + }, + { + "epoch": 0.46696, + "grad_norm": 1.4451828991260083, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 46696 + }, + { + "epoch": 0.46697, + "grad_norm": 1.0824162423692971, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 46697 + }, + { + "epoch": 0.46698, + "grad_norm": 0.9807197754871095, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 46698 + }, + { + "epoch": 0.46699, + "grad_norm": 1.1394785245646488, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 46699 + }, + { + "epoch": 0.467, + "grad_norm": 1.1255065487608231, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 46700 + }, + { + "epoch": 0.46701, + "grad_norm": 1.2791926264367295, + "learning_rate": 0.003, + "loss": 4.02, + "step": 46701 + }, + { + "epoch": 0.46702, + "grad_norm": 0.8373524394032235, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 46702 + }, + { + "epoch": 0.46703, + "grad_norm": 0.8290575149663222, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 46703 + }, + { + "epoch": 0.46704, + "grad_norm": 0.9461405582916625, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 46704 + }, + { + "epoch": 0.46705, + "grad_norm": 1.2255575152344789, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 46705 + }, + { + "epoch": 0.46706, + "grad_norm": 0.9539953741545478, + "learning_rate": 0.003, + "loss": 4.034, + "step": 46706 + }, + { + "epoch": 0.46707, + "grad_norm": 0.8191808430728434, + "learning_rate": 0.003, + "loss": 3.978, + "step": 46707 + }, + { + "epoch": 0.46708, + "grad_norm": 0.8442659696284279, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 46708 + }, + { + "epoch": 0.46709, + "grad_norm": 0.9721811828200235, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 46709 + }, + { + "epoch": 0.4671, + "grad_norm": 1.08997944999847, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 46710 + }, + { + "epoch": 0.46711, + "grad_norm": 1.1659402930606584, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 46711 + }, + { + "epoch": 0.46712, + "grad_norm": 1.090210082522445, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 46712 + }, + { + "epoch": 0.46713, + "grad_norm": 1.2181785104301275, + "learning_rate": 0.003, + "loss": 4.028, + "step": 46713 + }, + { + "epoch": 0.46714, + "grad_norm": 1.0908624083177825, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 46714 + }, + { + "epoch": 0.46715, + "grad_norm": 1.3994773384863195, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 46715 + }, + { + "epoch": 0.46716, + "grad_norm": 1.035318531113978, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 46716 + }, + { + "epoch": 0.46717, + "grad_norm": 1.0358277022286329, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 46717 + }, + { + "epoch": 0.46718, + "grad_norm": 1.0463844037140022, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 46718 + }, + { + "epoch": 0.46719, + "grad_norm": 1.1109285904956179, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 46719 + }, + { + "epoch": 0.4672, + "grad_norm": 1.0542582284680495, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 46720 + }, + { + "epoch": 0.46721, + "grad_norm": 0.9928906497058608, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 46721 + }, + { + "epoch": 0.46722, + "grad_norm": 1.1075222912061826, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 46722 + }, + { + "epoch": 0.46723, + "grad_norm": 0.8470341242115114, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 46723 + }, + { + "epoch": 0.46724, + "grad_norm": 0.934407360287406, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 46724 + }, + { + "epoch": 0.46725, + "grad_norm": 1.003134509962348, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 46725 + }, + { + "epoch": 0.46726, + "grad_norm": 1.406198370395076, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 46726 + }, + { + "epoch": 0.46727, + "grad_norm": 1.156151877752012, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 46727 + }, + { + "epoch": 0.46728, + "grad_norm": 0.9553570625716621, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 46728 + }, + { + "epoch": 0.46729, + "grad_norm": 1.0292308902351341, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 46729 + }, + { + "epoch": 0.4673, + "grad_norm": 1.3245155181256385, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 46730 + }, + { + "epoch": 0.46731, + "grad_norm": 0.8318681239381157, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 46731 + }, + { + "epoch": 0.46732, + "grad_norm": 0.9248909134546256, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 46732 + }, + { + "epoch": 0.46733, + "grad_norm": 1.2822257787946327, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 46733 + }, + { + "epoch": 0.46734, + "grad_norm": 1.0082113923278344, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 46734 + }, + { + "epoch": 0.46735, + "grad_norm": 1.0927181974718785, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 46735 + }, + { + "epoch": 0.46736, + "grad_norm": 1.2527980152038787, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 46736 + }, + { + "epoch": 0.46737, + "grad_norm": 0.831680571034005, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 46737 + }, + { + "epoch": 0.46738, + "grad_norm": 0.9598677493216463, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 46738 + }, + { + "epoch": 0.46739, + "grad_norm": 1.125592502411446, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 46739 + }, + { + "epoch": 0.4674, + "grad_norm": 1.0603335198750916, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 46740 + }, + { + "epoch": 0.46741, + "grad_norm": 0.9151460742864715, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 46741 + }, + { + "epoch": 0.46742, + "grad_norm": 0.847002265404357, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 46742 + }, + { + "epoch": 0.46743, + "grad_norm": 1.0497898544693758, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 46743 + }, + { + "epoch": 0.46744, + "grad_norm": 1.2981226645066604, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 46744 + }, + { + "epoch": 0.46745, + "grad_norm": 0.928067633757231, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 46745 + }, + { + "epoch": 0.46746, + "grad_norm": 1.0003713568502355, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 46746 + }, + { + "epoch": 0.46747, + "grad_norm": 1.403651977354227, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 46747 + }, + { + "epoch": 0.46748, + "grad_norm": 0.8112149237470166, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 46748 + }, + { + "epoch": 0.46749, + "grad_norm": 0.81165883696304, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 46749 + }, + { + "epoch": 0.4675, + "grad_norm": 0.8658244734444479, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 46750 + }, + { + "epoch": 0.46751, + "grad_norm": 1.0640311872785198, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 46751 + }, + { + "epoch": 0.46752, + "grad_norm": 1.2391121064779944, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 46752 + }, + { + "epoch": 0.46753, + "grad_norm": 1.011343564673627, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 46753 + }, + { + "epoch": 0.46754, + "grad_norm": 1.5044547540850541, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 46754 + }, + { + "epoch": 0.46755, + "grad_norm": 0.7281602501650878, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 46755 + }, + { + "epoch": 0.46756, + "grad_norm": 0.8198878197284111, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 46756 + }, + { + "epoch": 0.46757, + "grad_norm": 1.0923601034829273, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 46757 + }, + { + "epoch": 0.46758, + "grad_norm": 1.3472943137586546, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 46758 + }, + { + "epoch": 0.46759, + "grad_norm": 1.0519502303502744, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 46759 + }, + { + "epoch": 0.4676, + "grad_norm": 1.0025384570838844, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 46760 + }, + { + "epoch": 0.46761, + "grad_norm": 1.1039336318850235, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 46761 + }, + { + "epoch": 0.46762, + "grad_norm": 0.9305065453572996, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 46762 + }, + { + "epoch": 0.46763, + "grad_norm": 1.2237897847640984, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 46763 + }, + { + "epoch": 0.46764, + "grad_norm": 1.036635772268757, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 46764 + }, + { + "epoch": 0.46765, + "grad_norm": 0.9718299686865921, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 46765 + }, + { + "epoch": 0.46766, + "grad_norm": 1.2322754329348315, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 46766 + }, + { + "epoch": 0.46767, + "grad_norm": 1.3604899722008899, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 46767 + }, + { + "epoch": 0.46768, + "grad_norm": 0.8636618573631635, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 46768 + }, + { + "epoch": 0.46769, + "grad_norm": 1.0718439829718878, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 46769 + }, + { + "epoch": 0.4677, + "grad_norm": 1.2448396949755427, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 46770 + }, + { + "epoch": 0.46771, + "grad_norm": 0.9853792462951896, + "learning_rate": 0.003, + "loss": 4.027, + "step": 46771 + }, + { + "epoch": 0.46772, + "grad_norm": 1.122332628092207, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 46772 + }, + { + "epoch": 0.46773, + "grad_norm": 1.146255861524195, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 46773 + }, + { + "epoch": 0.46774, + "grad_norm": 1.0691423442414048, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 46774 + }, + { + "epoch": 0.46775, + "grad_norm": 1.0852209067060214, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 46775 + }, + { + "epoch": 0.46776, + "grad_norm": 0.9152906724419871, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 46776 + }, + { + "epoch": 0.46777, + "grad_norm": 0.8847806519304069, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 46777 + }, + { + "epoch": 0.46778, + "grad_norm": 1.0423365052443034, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 46778 + }, + { + "epoch": 0.46779, + "grad_norm": 1.2398843925110619, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 46779 + }, + { + "epoch": 0.4678, + "grad_norm": 1.0215586429956078, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 46780 + }, + { + "epoch": 0.46781, + "grad_norm": 1.154999139868981, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 46781 + }, + { + "epoch": 0.46782, + "grad_norm": 1.128084994765999, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 46782 + }, + { + "epoch": 0.46783, + "grad_norm": 1.2375701063470141, + "learning_rate": 0.003, + "loss": 4.028, + "step": 46783 + }, + { + "epoch": 0.46784, + "grad_norm": 1.1863168549613574, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 46784 + }, + { + "epoch": 0.46785, + "grad_norm": 1.009552985411251, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 46785 + }, + { + "epoch": 0.46786, + "grad_norm": 1.0999663801272064, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 46786 + }, + { + "epoch": 0.46787, + "grad_norm": 0.9238401414349614, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 46787 + }, + { + "epoch": 0.46788, + "grad_norm": 0.9163738015818826, + "learning_rate": 0.003, + "loss": 4.031, + "step": 46788 + }, + { + "epoch": 0.46789, + "grad_norm": 1.0229558455671395, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 46789 + }, + { + "epoch": 0.4679, + "grad_norm": 1.2813996622145423, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 46790 + }, + { + "epoch": 0.46791, + "grad_norm": 1.0354015792852926, + "learning_rate": 0.003, + "loss": 4.03, + "step": 46791 + }, + { + "epoch": 0.46792, + "grad_norm": 0.9672364186399688, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 46792 + }, + { + "epoch": 0.46793, + "grad_norm": 0.9891940146373028, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 46793 + }, + { + "epoch": 0.46794, + "grad_norm": 1.09288750453215, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 46794 + }, + { + "epoch": 0.46795, + "grad_norm": 1.080777616612828, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 46795 + }, + { + "epoch": 0.46796, + "grad_norm": 1.455602449886524, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 46796 + }, + { + "epoch": 0.46797, + "grad_norm": 0.7743373490461938, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 46797 + }, + { + "epoch": 0.46798, + "grad_norm": 0.8173508825832866, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 46798 + }, + { + "epoch": 0.46799, + "grad_norm": 0.9868090996919716, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 46799 + }, + { + "epoch": 0.468, + "grad_norm": 1.0412883237411057, + "learning_rate": 0.003, + "loss": 4.048, + "step": 46800 + }, + { + "epoch": 0.46801, + "grad_norm": 1.1615947253965875, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 46801 + }, + { + "epoch": 0.46802, + "grad_norm": 1.0970239519638185, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 46802 + }, + { + "epoch": 0.46803, + "grad_norm": 1.4759827952880356, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 46803 + }, + { + "epoch": 0.46804, + "grad_norm": 0.7521432708387432, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 46804 + }, + { + "epoch": 0.46805, + "grad_norm": 0.7294079959808589, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 46805 + }, + { + "epoch": 0.46806, + "grad_norm": 0.8708817806711274, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 46806 + }, + { + "epoch": 0.46807, + "grad_norm": 1.1560035346221942, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 46807 + }, + { + "epoch": 0.46808, + "grad_norm": 1.019206079084952, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 46808 + }, + { + "epoch": 0.46809, + "grad_norm": 1.1271494016490669, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 46809 + }, + { + "epoch": 0.4681, + "grad_norm": 0.9580946305408445, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 46810 + }, + { + "epoch": 0.46811, + "grad_norm": 1.02364834437713, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 46811 + }, + { + "epoch": 0.46812, + "grad_norm": 1.3972720639821428, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 46812 + }, + { + "epoch": 0.46813, + "grad_norm": 0.7884825683736798, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 46813 + }, + { + "epoch": 0.46814, + "grad_norm": 0.7993716679248967, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 46814 + }, + { + "epoch": 0.46815, + "grad_norm": 0.8362435251888497, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 46815 + }, + { + "epoch": 0.46816, + "grad_norm": 1.1881567766904921, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 46816 + }, + { + "epoch": 0.46817, + "grad_norm": 1.1580784238580846, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 46817 + }, + { + "epoch": 0.46818, + "grad_norm": 1.1436114339223868, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 46818 + }, + { + "epoch": 0.46819, + "grad_norm": 1.2569302621735132, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 46819 + }, + { + "epoch": 0.4682, + "grad_norm": 0.9301782440343385, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 46820 + }, + { + "epoch": 0.46821, + "grad_norm": 1.0180718348554967, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 46821 + }, + { + "epoch": 0.46822, + "grad_norm": 1.260225342546046, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 46822 + }, + { + "epoch": 0.46823, + "grad_norm": 0.8774447588207753, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 46823 + }, + { + "epoch": 0.46824, + "grad_norm": 1.0357469794749354, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 46824 + }, + { + "epoch": 0.46825, + "grad_norm": 1.163410841600912, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 46825 + }, + { + "epoch": 0.46826, + "grad_norm": 1.0278003486083516, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 46826 + }, + { + "epoch": 0.46827, + "grad_norm": 1.2497178984439838, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 46827 + }, + { + "epoch": 0.46828, + "grad_norm": 0.9780341433554378, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 46828 + }, + { + "epoch": 0.46829, + "grad_norm": 1.2426328829469326, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 46829 + }, + { + "epoch": 0.4683, + "grad_norm": 1.041049365323944, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 46830 + }, + { + "epoch": 0.46831, + "grad_norm": 1.0326293349209925, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 46831 + }, + { + "epoch": 0.46832, + "grad_norm": 1.1247797602472571, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 46832 + }, + { + "epoch": 0.46833, + "grad_norm": 0.9156252825361716, + "learning_rate": 0.003, + "loss": 4.055, + "step": 46833 + }, + { + "epoch": 0.46834, + "grad_norm": 1.0520233544474875, + "learning_rate": 0.003, + "loss": 4.034, + "step": 46834 + }, + { + "epoch": 0.46835, + "grad_norm": 1.138438581200076, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 46835 + }, + { + "epoch": 0.46836, + "grad_norm": 1.1111222678801955, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 46836 + }, + { + "epoch": 0.46837, + "grad_norm": 1.0560566092868426, + "learning_rate": 0.003, + "loss": 4.006, + "step": 46837 + }, + { + "epoch": 0.46838, + "grad_norm": 1.2080748518777218, + "learning_rate": 0.003, + "loss": 4.028, + "step": 46838 + }, + { + "epoch": 0.46839, + "grad_norm": 1.1261039278259481, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 46839 + }, + { + "epoch": 0.4684, + "grad_norm": 1.131750665790347, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 46840 + }, + { + "epoch": 0.46841, + "grad_norm": 1.0623356798685841, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 46841 + }, + { + "epoch": 0.46842, + "grad_norm": 0.9398308648241623, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 46842 + }, + { + "epoch": 0.46843, + "grad_norm": 1.1924503515385028, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 46843 + }, + { + "epoch": 0.46844, + "grad_norm": 1.101335158263482, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 46844 + }, + { + "epoch": 0.46845, + "grad_norm": 1.074773965278515, + "learning_rate": 0.003, + "loss": 4.035, + "step": 46845 + }, + { + "epoch": 0.46846, + "grad_norm": 0.9491521344950171, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 46846 + }, + { + "epoch": 0.46847, + "grad_norm": 1.23789633380621, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 46847 + }, + { + "epoch": 0.46848, + "grad_norm": 1.1165119312924943, + "learning_rate": 0.003, + "loss": 4.037, + "step": 46848 + }, + { + "epoch": 0.46849, + "grad_norm": 1.0884178636626012, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 46849 + }, + { + "epoch": 0.4685, + "grad_norm": 0.8385340642176272, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 46850 + }, + { + "epoch": 0.46851, + "grad_norm": 0.973727328866319, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 46851 + }, + { + "epoch": 0.46852, + "grad_norm": 1.036535579602446, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 46852 + }, + { + "epoch": 0.46853, + "grad_norm": 1.1251878792448577, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 46853 + }, + { + "epoch": 0.46854, + "grad_norm": 1.264003433113297, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 46854 + }, + { + "epoch": 0.46855, + "grad_norm": 1.0009853360204939, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 46855 + }, + { + "epoch": 0.46856, + "grad_norm": 1.2065105722176723, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 46856 + }, + { + "epoch": 0.46857, + "grad_norm": 1.0245914320978136, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 46857 + }, + { + "epoch": 0.46858, + "grad_norm": 1.3048256933061224, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 46858 + }, + { + "epoch": 0.46859, + "grad_norm": 0.9441271554334885, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 46859 + }, + { + "epoch": 0.4686, + "grad_norm": 0.9704434794271077, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 46860 + }, + { + "epoch": 0.46861, + "grad_norm": 1.2515014124400665, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 46861 + }, + { + "epoch": 0.46862, + "grad_norm": 0.9523563443621351, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 46862 + }, + { + "epoch": 0.46863, + "grad_norm": 0.9992429253290724, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 46863 + }, + { + "epoch": 0.46864, + "grad_norm": 1.1848072547322175, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 46864 + }, + { + "epoch": 0.46865, + "grad_norm": 1.0895071632687638, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 46865 + }, + { + "epoch": 0.46866, + "grad_norm": 1.0061988751070436, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 46866 + }, + { + "epoch": 0.46867, + "grad_norm": 1.047929695659627, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 46867 + }, + { + "epoch": 0.46868, + "grad_norm": 1.096609047499437, + "learning_rate": 0.003, + "loss": 4.06, + "step": 46868 + }, + { + "epoch": 0.46869, + "grad_norm": 0.9677672543524036, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 46869 + }, + { + "epoch": 0.4687, + "grad_norm": 0.8885019047485784, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 46870 + }, + { + "epoch": 0.46871, + "grad_norm": 0.9056357439405407, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 46871 + }, + { + "epoch": 0.46872, + "grad_norm": 0.9742165255404818, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 46872 + }, + { + "epoch": 0.46873, + "grad_norm": 1.0135329097406434, + "learning_rate": 0.003, + "loss": 4.035, + "step": 46873 + }, + { + "epoch": 0.46874, + "grad_norm": 1.3775837978354566, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 46874 + }, + { + "epoch": 0.46875, + "grad_norm": 0.924306765979343, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 46875 + }, + { + "epoch": 0.46876, + "grad_norm": 0.9979431911933595, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 46876 + }, + { + "epoch": 0.46877, + "grad_norm": 1.341406364372382, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 46877 + }, + { + "epoch": 0.46878, + "grad_norm": 1.2611173545478154, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 46878 + }, + { + "epoch": 0.46879, + "grad_norm": 1.054245507333645, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 46879 + }, + { + "epoch": 0.4688, + "grad_norm": 1.1485669731490886, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 46880 + }, + { + "epoch": 0.46881, + "grad_norm": 0.9498413415423206, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 46881 + }, + { + "epoch": 0.46882, + "grad_norm": 1.1348795675440713, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 46882 + }, + { + "epoch": 0.46883, + "grad_norm": 1.0605503163533252, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 46883 + }, + { + "epoch": 0.46884, + "grad_norm": 1.2319779944617284, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 46884 + }, + { + "epoch": 0.46885, + "grad_norm": 0.940305892471008, + "learning_rate": 0.003, + "loss": 4.069, + "step": 46885 + }, + { + "epoch": 0.46886, + "grad_norm": 0.9848309479396198, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 46886 + }, + { + "epoch": 0.46887, + "grad_norm": 1.2476395498004391, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 46887 + }, + { + "epoch": 0.46888, + "grad_norm": 1.1608069112559976, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 46888 + }, + { + "epoch": 0.46889, + "grad_norm": 1.041585024567791, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 46889 + }, + { + "epoch": 0.4689, + "grad_norm": 1.046076736246221, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 46890 + }, + { + "epoch": 0.46891, + "grad_norm": 1.0626171148598027, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 46891 + }, + { + "epoch": 0.46892, + "grad_norm": 0.8935517192816217, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 46892 + }, + { + "epoch": 0.46893, + "grad_norm": 0.8771266460469778, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 46893 + }, + { + "epoch": 0.46894, + "grad_norm": 1.0572696336296041, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 46894 + }, + { + "epoch": 0.46895, + "grad_norm": 1.08386264232727, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 46895 + }, + { + "epoch": 0.46896, + "grad_norm": 0.938983840500144, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 46896 + }, + { + "epoch": 0.46897, + "grad_norm": 0.9683381636563494, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 46897 + }, + { + "epoch": 0.46898, + "grad_norm": 1.2775825640092215, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 46898 + }, + { + "epoch": 0.46899, + "grad_norm": 0.7220805791709981, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 46899 + }, + { + "epoch": 0.469, + "grad_norm": 0.7393261204582248, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 46900 + }, + { + "epoch": 0.46901, + "grad_norm": 0.8774616536867127, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 46901 + }, + { + "epoch": 0.46902, + "grad_norm": 1.0277723348482977, + "learning_rate": 0.003, + "loss": 4.033, + "step": 46902 + }, + { + "epoch": 0.46903, + "grad_norm": 1.1597061094811514, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 46903 + }, + { + "epoch": 0.46904, + "grad_norm": 1.1200262370700595, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 46904 + }, + { + "epoch": 0.46905, + "grad_norm": 1.2542728534290748, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 46905 + }, + { + "epoch": 0.46906, + "grad_norm": 1.0074622195017633, + "learning_rate": 0.003, + "loss": 4.016, + "step": 46906 + }, + { + "epoch": 0.46907, + "grad_norm": 1.1478018725989512, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 46907 + }, + { + "epoch": 0.46908, + "grad_norm": 1.1160483167838369, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 46908 + }, + { + "epoch": 0.46909, + "grad_norm": 1.0715551458424328, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 46909 + }, + { + "epoch": 0.4691, + "grad_norm": 1.0918234906359092, + "learning_rate": 0.003, + "loss": 4.035, + "step": 46910 + }, + { + "epoch": 0.46911, + "grad_norm": 1.2753456016202152, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 46911 + }, + { + "epoch": 0.46912, + "grad_norm": 1.0462418844815902, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 46912 + }, + { + "epoch": 0.46913, + "grad_norm": 1.3471325542186556, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 46913 + }, + { + "epoch": 0.46914, + "grad_norm": 0.9390726884257209, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 46914 + }, + { + "epoch": 0.46915, + "grad_norm": 0.8413084232276994, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 46915 + }, + { + "epoch": 0.46916, + "grad_norm": 0.9864939240760546, + "learning_rate": 0.003, + "loss": 4.016, + "step": 46916 + }, + { + "epoch": 0.46917, + "grad_norm": 1.0878971165248772, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 46917 + }, + { + "epoch": 0.46918, + "grad_norm": 0.9986517116450863, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 46918 + }, + { + "epoch": 0.46919, + "grad_norm": 1.1937329717051148, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 46919 + }, + { + "epoch": 0.4692, + "grad_norm": 0.9217796616683009, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 46920 + }, + { + "epoch": 0.46921, + "grad_norm": 1.0260431624787407, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 46921 + }, + { + "epoch": 0.46922, + "grad_norm": 1.175694651958549, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 46922 + }, + { + "epoch": 0.46923, + "grad_norm": 1.0762622252329554, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 46923 + }, + { + "epoch": 0.46924, + "grad_norm": 1.2527118339888528, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 46924 + }, + { + "epoch": 0.46925, + "grad_norm": 1.0161047056244659, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 46925 + }, + { + "epoch": 0.46926, + "grad_norm": 1.1800715025680801, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 46926 + }, + { + "epoch": 0.46927, + "grad_norm": 1.1466574601217405, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 46927 + }, + { + "epoch": 0.46928, + "grad_norm": 1.0801419028898376, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 46928 + }, + { + "epoch": 0.46929, + "grad_norm": 0.8973992906241901, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 46929 + }, + { + "epoch": 0.4693, + "grad_norm": 0.9375098919911677, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 46930 + }, + { + "epoch": 0.46931, + "grad_norm": 1.1503563781863073, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 46931 + }, + { + "epoch": 0.46932, + "grad_norm": 0.9451624953679633, + "learning_rate": 0.003, + "loss": 4.068, + "step": 46932 + }, + { + "epoch": 0.46933, + "grad_norm": 1.2041023213922777, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 46933 + }, + { + "epoch": 0.46934, + "grad_norm": 1.2043136377247379, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 46934 + }, + { + "epoch": 0.46935, + "grad_norm": 1.1202467155361135, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 46935 + }, + { + "epoch": 0.46936, + "grad_norm": 0.998879094307563, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 46936 + }, + { + "epoch": 0.46937, + "grad_norm": 1.018349090981993, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 46937 + }, + { + "epoch": 0.46938, + "grad_norm": 1.1290409928187364, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 46938 + }, + { + "epoch": 0.46939, + "grad_norm": 1.0927996514615719, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 46939 + }, + { + "epoch": 0.4694, + "grad_norm": 1.140560628095361, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 46940 + }, + { + "epoch": 0.46941, + "grad_norm": 1.198120509146442, + "learning_rate": 0.003, + "loss": 4.051, + "step": 46941 + }, + { + "epoch": 0.46942, + "grad_norm": 0.9829984106872426, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 46942 + }, + { + "epoch": 0.46943, + "grad_norm": 1.02895651606843, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 46943 + }, + { + "epoch": 0.46944, + "grad_norm": 1.0332392364244085, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 46944 + }, + { + "epoch": 0.46945, + "grad_norm": 0.9233494544731673, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 46945 + }, + { + "epoch": 0.46946, + "grad_norm": 0.873744707405634, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 46946 + }, + { + "epoch": 0.46947, + "grad_norm": 0.8421400867836066, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 46947 + }, + { + "epoch": 0.46948, + "grad_norm": 0.9542854982712484, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 46948 + }, + { + "epoch": 0.46949, + "grad_norm": 1.0419851919508882, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 46949 + }, + { + "epoch": 0.4695, + "grad_norm": 1.1030654544262462, + "learning_rate": 0.003, + "loss": 4.047, + "step": 46950 + }, + { + "epoch": 0.46951, + "grad_norm": 1.0827939260251307, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 46951 + }, + { + "epoch": 0.46952, + "grad_norm": 1.0473722338694387, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 46952 + }, + { + "epoch": 0.46953, + "grad_norm": 1.6003149222220816, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 46953 + }, + { + "epoch": 0.46954, + "grad_norm": 0.8194219280254647, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 46954 + }, + { + "epoch": 0.46955, + "grad_norm": 0.8749659991403884, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 46955 + }, + { + "epoch": 0.46956, + "grad_norm": 0.9627233240867954, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 46956 + }, + { + "epoch": 0.46957, + "grad_norm": 0.8950786415738246, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 46957 + }, + { + "epoch": 0.46958, + "grad_norm": 1.067767848322479, + "learning_rate": 0.003, + "loss": 4.049, + "step": 46958 + }, + { + "epoch": 0.46959, + "grad_norm": 0.9884108980786164, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 46959 + }, + { + "epoch": 0.4696, + "grad_norm": 0.9912817462306213, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 46960 + }, + { + "epoch": 0.46961, + "grad_norm": 0.9449584095410177, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 46961 + }, + { + "epoch": 0.46962, + "grad_norm": 1.1008486388573624, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 46962 + }, + { + "epoch": 0.46963, + "grad_norm": 1.2136047857508792, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 46963 + }, + { + "epoch": 0.46964, + "grad_norm": 1.1945085035521212, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 46964 + }, + { + "epoch": 0.46965, + "grad_norm": 1.2555113687441757, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 46965 + }, + { + "epoch": 0.46966, + "grad_norm": 1.0087066857079552, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 46966 + }, + { + "epoch": 0.46967, + "grad_norm": 0.894893525387343, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 46967 + }, + { + "epoch": 0.46968, + "grad_norm": 1.0820691967415224, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 46968 + }, + { + "epoch": 0.46969, + "grad_norm": 1.4077465010055736, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 46969 + }, + { + "epoch": 0.4697, + "grad_norm": 0.8193042582673117, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 46970 + }, + { + "epoch": 0.46971, + "grad_norm": 0.8535142220687872, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 46971 + }, + { + "epoch": 0.46972, + "grad_norm": 0.9786582398183373, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 46972 + }, + { + "epoch": 0.46973, + "grad_norm": 1.4712812071757064, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 46973 + }, + { + "epoch": 0.46974, + "grad_norm": 0.7235827822168228, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 46974 + }, + { + "epoch": 0.46975, + "grad_norm": 0.8347056827490715, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 46975 + }, + { + "epoch": 0.46976, + "grad_norm": 1.1346391938480427, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 46976 + }, + { + "epoch": 0.46977, + "grad_norm": 1.385040834009163, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 46977 + }, + { + "epoch": 0.46978, + "grad_norm": 0.7677343580660407, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 46978 + }, + { + "epoch": 0.46979, + "grad_norm": 0.8797833585232953, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 46979 + }, + { + "epoch": 0.4698, + "grad_norm": 1.4872758322077761, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 46980 + }, + { + "epoch": 0.46981, + "grad_norm": 1.2025546378481542, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 46981 + }, + { + "epoch": 0.46982, + "grad_norm": 1.0318817653768304, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 46982 + }, + { + "epoch": 0.46983, + "grad_norm": 1.292073418368407, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 46983 + }, + { + "epoch": 0.46984, + "grad_norm": 1.0579670368034118, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 46984 + }, + { + "epoch": 0.46985, + "grad_norm": 1.290645115302989, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 46985 + }, + { + "epoch": 0.46986, + "grad_norm": 0.9273036875844688, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 46986 + }, + { + "epoch": 0.46987, + "grad_norm": 0.9810463614509226, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 46987 + }, + { + "epoch": 0.46988, + "grad_norm": 1.3500250115453396, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 46988 + }, + { + "epoch": 0.46989, + "grad_norm": 0.9201290803706176, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 46989 + }, + { + "epoch": 0.4699, + "grad_norm": 1.3741744605362116, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 46990 + }, + { + "epoch": 0.46991, + "grad_norm": 1.0846062462048827, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 46991 + }, + { + "epoch": 0.46992, + "grad_norm": 1.2106317232006696, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 46992 + }, + { + "epoch": 0.46993, + "grad_norm": 1.0971965167483846, + "learning_rate": 0.003, + "loss": 4.0848, + "step": 46993 + }, + { + "epoch": 0.46994, + "grad_norm": 0.8996565069100357, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 46994 + }, + { + "epoch": 0.46995, + "grad_norm": 0.952246980632302, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 46995 + }, + { + "epoch": 0.46996, + "grad_norm": 1.1401825949282405, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 46996 + }, + { + "epoch": 0.46997, + "grad_norm": 1.462925119212866, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 46997 + }, + { + "epoch": 0.46998, + "grad_norm": 0.8906688259054713, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 46998 + }, + { + "epoch": 0.46999, + "grad_norm": 0.7354213550312292, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 46999 + }, + { + "epoch": 0.47, + "grad_norm": 0.9442992335956615, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 47000 + }, + { + "epoch": 0.47001, + "grad_norm": 1.3183015401166043, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 47001 + }, + { + "epoch": 0.47002, + "grad_norm": 1.2936734906823268, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 47002 + }, + { + "epoch": 0.47003, + "grad_norm": 0.9364225230595772, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 47003 + }, + { + "epoch": 0.47004, + "grad_norm": 0.9213481857874497, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 47004 + }, + { + "epoch": 0.47005, + "grad_norm": 1.129940964723215, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 47005 + }, + { + "epoch": 0.47006, + "grad_norm": 1.1751536692206321, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 47006 + }, + { + "epoch": 0.47007, + "grad_norm": 0.8795410847840189, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 47007 + }, + { + "epoch": 0.47008, + "grad_norm": 0.8681570847034393, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 47008 + }, + { + "epoch": 0.47009, + "grad_norm": 0.8470588314639754, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 47009 + }, + { + "epoch": 0.4701, + "grad_norm": 0.8929148119350815, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 47010 + }, + { + "epoch": 0.47011, + "grad_norm": 1.0646250266795867, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 47011 + }, + { + "epoch": 0.47012, + "grad_norm": 1.1750156007966328, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 47012 + }, + { + "epoch": 0.47013, + "grad_norm": 0.8727249082926137, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 47013 + }, + { + "epoch": 0.47014, + "grad_norm": 1.0549894198178171, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 47014 + }, + { + "epoch": 0.47015, + "grad_norm": 1.1928192616670934, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 47015 + }, + { + "epoch": 0.47016, + "grad_norm": 1.1458700711636536, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 47016 + }, + { + "epoch": 0.47017, + "grad_norm": 1.0203919159791268, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 47017 + }, + { + "epoch": 0.47018, + "grad_norm": 1.0821173228662573, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 47018 + }, + { + "epoch": 0.47019, + "grad_norm": 1.0466299143272249, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 47019 + }, + { + "epoch": 0.4702, + "grad_norm": 1.113204840598067, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 47020 + }, + { + "epoch": 0.47021, + "grad_norm": 1.0340949076701882, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 47021 + }, + { + "epoch": 0.47022, + "grad_norm": 1.2212233867715727, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 47022 + }, + { + "epoch": 0.47023, + "grad_norm": 1.0273162376300065, + "learning_rate": 0.003, + "loss": 4.041, + "step": 47023 + }, + { + "epoch": 0.47024, + "grad_norm": 1.2781544475254085, + "learning_rate": 0.003, + "loss": 4.059, + "step": 47024 + }, + { + "epoch": 0.47025, + "grad_norm": 1.026876235896194, + "learning_rate": 0.003, + "loss": 4.026, + "step": 47025 + }, + { + "epoch": 0.47026, + "grad_norm": 1.1418101949883621, + "learning_rate": 0.003, + "loss": 4.073, + "step": 47026 + }, + { + "epoch": 0.47027, + "grad_norm": 0.8404137713018313, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 47027 + }, + { + "epoch": 0.47028, + "grad_norm": 0.8311610745127861, + "learning_rate": 0.003, + "loss": 4.015, + "step": 47028 + }, + { + "epoch": 0.47029, + "grad_norm": 1.1166323618803007, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 47029 + }, + { + "epoch": 0.4703, + "grad_norm": 0.9914684332142837, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 47030 + }, + { + "epoch": 0.47031, + "grad_norm": 1.2705318258562877, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 47031 + }, + { + "epoch": 0.47032, + "grad_norm": 1.136181616807093, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 47032 + }, + { + "epoch": 0.47033, + "grad_norm": 1.1466273842086943, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 47033 + }, + { + "epoch": 0.47034, + "grad_norm": 1.0518795929969624, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 47034 + }, + { + "epoch": 0.47035, + "grad_norm": 0.9150498650207703, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 47035 + }, + { + "epoch": 0.47036, + "grad_norm": 0.9485503653184217, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 47036 + }, + { + "epoch": 0.47037, + "grad_norm": 1.2740541611088223, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 47037 + }, + { + "epoch": 0.47038, + "grad_norm": 1.217911351859132, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 47038 + }, + { + "epoch": 0.47039, + "grad_norm": 1.0004419663688844, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 47039 + }, + { + "epoch": 0.4704, + "grad_norm": 1.2221086152991403, + "learning_rate": 0.003, + "loss": 4.056, + "step": 47040 + }, + { + "epoch": 0.47041, + "grad_norm": 0.9945407239560049, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 47041 + }, + { + "epoch": 0.47042, + "grad_norm": 1.2087353362820092, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 47042 + }, + { + "epoch": 0.47043, + "grad_norm": 1.1898051354005315, + "learning_rate": 0.003, + "loss": 4.059, + "step": 47043 + }, + { + "epoch": 0.47044, + "grad_norm": 1.0680051705250575, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 47044 + }, + { + "epoch": 0.47045, + "grad_norm": 0.9661118992610916, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 47045 + }, + { + "epoch": 0.47046, + "grad_norm": 1.0887183053496812, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 47046 + }, + { + "epoch": 0.47047, + "grad_norm": 1.15201448102089, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 47047 + }, + { + "epoch": 0.47048, + "grad_norm": 1.0190744069599553, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 47048 + }, + { + "epoch": 0.47049, + "grad_norm": 1.0673775567497337, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 47049 + }, + { + "epoch": 0.4705, + "grad_norm": 0.8764641133330593, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 47050 + }, + { + "epoch": 0.47051, + "grad_norm": 0.9663566435129618, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 47051 + }, + { + "epoch": 0.47052, + "grad_norm": 1.0661855802428675, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 47052 + }, + { + "epoch": 0.47053, + "grad_norm": 1.0074715639694234, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 47053 + }, + { + "epoch": 0.47054, + "grad_norm": 1.1760664351352672, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 47054 + }, + { + "epoch": 0.47055, + "grad_norm": 1.173354561550787, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 47055 + }, + { + "epoch": 0.47056, + "grad_norm": 0.900153654091279, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 47056 + }, + { + "epoch": 0.47057, + "grad_norm": 1.0689108513953711, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 47057 + }, + { + "epoch": 0.47058, + "grad_norm": 1.4650569988979734, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 47058 + }, + { + "epoch": 0.47059, + "grad_norm": 0.9744796843923446, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 47059 + }, + { + "epoch": 0.4706, + "grad_norm": 0.9389167603387536, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 47060 + }, + { + "epoch": 0.47061, + "grad_norm": 1.0876256239133764, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 47061 + }, + { + "epoch": 0.47062, + "grad_norm": 1.1840596786642232, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 47062 + }, + { + "epoch": 0.47063, + "grad_norm": 0.9154639122716438, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 47063 + }, + { + "epoch": 0.47064, + "grad_norm": 1.0539418729360281, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 47064 + }, + { + "epoch": 0.47065, + "grad_norm": 1.285305082206712, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 47065 + }, + { + "epoch": 0.47066, + "grad_norm": 0.7699118336935876, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 47066 + }, + { + "epoch": 0.47067, + "grad_norm": 0.8221683400974741, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 47067 + }, + { + "epoch": 0.47068, + "grad_norm": 1.0366360790596256, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 47068 + }, + { + "epoch": 0.47069, + "grad_norm": 1.2663410587881823, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 47069 + }, + { + "epoch": 0.4707, + "grad_norm": 0.7756303999181368, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 47070 + }, + { + "epoch": 0.47071, + "grad_norm": 0.7647789422069307, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 47071 + }, + { + "epoch": 0.47072, + "grad_norm": 0.7943164198533986, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 47072 + }, + { + "epoch": 0.47073, + "grad_norm": 0.9378381711788226, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 47073 + }, + { + "epoch": 0.47074, + "grad_norm": 1.1799421064577456, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 47074 + }, + { + "epoch": 0.47075, + "grad_norm": 1.2063919790229782, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 47075 + }, + { + "epoch": 0.47076, + "grad_norm": 1.3017984260306414, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 47076 + }, + { + "epoch": 0.47077, + "grad_norm": 0.9985105278032861, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 47077 + }, + { + "epoch": 0.47078, + "grad_norm": 1.193442767344976, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 47078 + }, + { + "epoch": 0.47079, + "grad_norm": 1.1552028717588652, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 47079 + }, + { + "epoch": 0.4708, + "grad_norm": 1.106362456118197, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 47080 + }, + { + "epoch": 0.47081, + "grad_norm": 1.1654811176191615, + "learning_rate": 0.003, + "loss": 4.021, + "step": 47081 + }, + { + "epoch": 0.47082, + "grad_norm": 0.930719207473284, + "learning_rate": 0.003, + "loss": 4.062, + "step": 47082 + }, + { + "epoch": 0.47083, + "grad_norm": 1.0140394183331511, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 47083 + }, + { + "epoch": 0.47084, + "grad_norm": 1.1066796694950527, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 47084 + }, + { + "epoch": 0.47085, + "grad_norm": 1.2082812920447161, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 47085 + }, + { + "epoch": 0.47086, + "grad_norm": 0.8507270339774522, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 47086 + }, + { + "epoch": 0.47087, + "grad_norm": 0.7687640688060402, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 47087 + }, + { + "epoch": 0.47088, + "grad_norm": 0.8085831725718131, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 47088 + }, + { + "epoch": 0.47089, + "grad_norm": 0.9728978456738948, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 47089 + }, + { + "epoch": 0.4709, + "grad_norm": 1.1959128854211682, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 47090 + }, + { + "epoch": 0.47091, + "grad_norm": 0.9251714158574068, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 47091 + }, + { + "epoch": 0.47092, + "grad_norm": 1.0797546280429908, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 47092 + }, + { + "epoch": 0.47093, + "grad_norm": 1.0298099891272456, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 47093 + }, + { + "epoch": 0.47094, + "grad_norm": 1.1327377466456094, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 47094 + }, + { + "epoch": 0.47095, + "grad_norm": 1.147499328931647, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 47095 + }, + { + "epoch": 0.47096, + "grad_norm": 1.2168696031042727, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 47096 + }, + { + "epoch": 0.47097, + "grad_norm": 0.9831647839787495, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 47097 + }, + { + "epoch": 0.47098, + "grad_norm": 1.1079649184712947, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 47098 + }, + { + "epoch": 0.47099, + "grad_norm": 1.1500153073573252, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 47099 + }, + { + "epoch": 0.471, + "grad_norm": 1.203468458755053, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 47100 + }, + { + "epoch": 0.47101, + "grad_norm": 0.8425993544754998, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 47101 + }, + { + "epoch": 0.47102, + "grad_norm": 0.986731730753535, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 47102 + }, + { + "epoch": 0.47103, + "grad_norm": 1.2534708929480989, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 47103 + }, + { + "epoch": 0.47104, + "grad_norm": 0.994792637461191, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 47104 + }, + { + "epoch": 0.47105, + "grad_norm": 1.3876464159340947, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 47105 + }, + { + "epoch": 0.47106, + "grad_norm": 1.094707843550068, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 47106 + }, + { + "epoch": 0.47107, + "grad_norm": 1.0490457166778766, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 47107 + }, + { + "epoch": 0.47108, + "grad_norm": 1.1650009257919716, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 47108 + }, + { + "epoch": 0.47109, + "grad_norm": 0.8387668356886046, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 47109 + }, + { + "epoch": 0.4711, + "grad_norm": 0.862082135020135, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 47110 + }, + { + "epoch": 0.47111, + "grad_norm": 0.9078889418830123, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 47111 + }, + { + "epoch": 0.47112, + "grad_norm": 1.0957501848467965, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 47112 + }, + { + "epoch": 0.47113, + "grad_norm": 1.259145384991058, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 47113 + }, + { + "epoch": 0.47114, + "grad_norm": 0.7620840208470238, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 47114 + }, + { + "epoch": 0.47115, + "grad_norm": 0.7327558325206038, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 47115 + }, + { + "epoch": 0.47116, + "grad_norm": 0.8620713853364467, + "learning_rate": 0.003, + "loss": 4.014, + "step": 47116 + }, + { + "epoch": 0.47117, + "grad_norm": 0.9865926945235641, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 47117 + }, + { + "epoch": 0.47118, + "grad_norm": 0.929362699878257, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 47118 + }, + { + "epoch": 0.47119, + "grad_norm": 0.8454222241179398, + "learning_rate": 0.003, + "loss": 4.007, + "step": 47119 + }, + { + "epoch": 0.4712, + "grad_norm": 1.087253860157815, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 47120 + }, + { + "epoch": 0.47121, + "grad_norm": 1.4134541196286474, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 47121 + }, + { + "epoch": 0.47122, + "grad_norm": 0.7064182511282521, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 47122 + }, + { + "epoch": 0.47123, + "grad_norm": 0.7053681744746996, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 47123 + }, + { + "epoch": 0.47124, + "grad_norm": 0.9151622651231472, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 47124 + }, + { + "epoch": 0.47125, + "grad_norm": 1.1811781669615053, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 47125 + }, + { + "epoch": 0.47126, + "grad_norm": 1.29514291600615, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 47126 + }, + { + "epoch": 0.47127, + "grad_norm": 0.8920789260194247, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 47127 + }, + { + "epoch": 0.47128, + "grad_norm": 1.008477256757158, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 47128 + }, + { + "epoch": 0.47129, + "grad_norm": 1.4358099374194393, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 47129 + }, + { + "epoch": 0.4713, + "grad_norm": 0.9728865334840863, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 47130 + }, + { + "epoch": 0.47131, + "grad_norm": 0.8932694628691841, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 47131 + }, + { + "epoch": 0.47132, + "grad_norm": 1.0730682411405814, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 47132 + }, + { + "epoch": 0.47133, + "grad_norm": 1.312025854751486, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 47133 + }, + { + "epoch": 0.47134, + "grad_norm": 1.0403371982326939, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 47134 + }, + { + "epoch": 0.47135, + "grad_norm": 1.138968243601966, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 47135 + }, + { + "epoch": 0.47136, + "grad_norm": 1.1234563802663358, + "learning_rate": 0.003, + "loss": 4.025, + "step": 47136 + }, + { + "epoch": 0.47137, + "grad_norm": 1.2046395892024622, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 47137 + }, + { + "epoch": 0.47138, + "grad_norm": 0.9050254135319017, + "learning_rate": 0.003, + "loss": 4.036, + "step": 47138 + }, + { + "epoch": 0.47139, + "grad_norm": 1.010189433897938, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 47139 + }, + { + "epoch": 0.4714, + "grad_norm": 1.2211115831246935, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 47140 + }, + { + "epoch": 0.47141, + "grad_norm": 1.0368250739675824, + "learning_rate": 0.003, + "loss": 4.024, + "step": 47141 + }, + { + "epoch": 0.47142, + "grad_norm": 1.2099386448460396, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 47142 + }, + { + "epoch": 0.47143, + "grad_norm": 0.8821878386752869, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 47143 + }, + { + "epoch": 0.47144, + "grad_norm": 0.924685284685409, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 47144 + }, + { + "epoch": 0.47145, + "grad_norm": 1.0853850885332932, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 47145 + }, + { + "epoch": 0.47146, + "grad_norm": 1.2478276565014779, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 47146 + }, + { + "epoch": 0.47147, + "grad_norm": 0.9777747573057037, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 47147 + }, + { + "epoch": 0.47148, + "grad_norm": 1.0811176660236377, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 47148 + }, + { + "epoch": 0.47149, + "grad_norm": 0.997864353157992, + "learning_rate": 0.003, + "loss": 4.024, + "step": 47149 + }, + { + "epoch": 0.4715, + "grad_norm": 1.2890949065254838, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 47150 + }, + { + "epoch": 0.47151, + "grad_norm": 1.0596553820089356, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 47151 + }, + { + "epoch": 0.47152, + "grad_norm": 1.249293904838947, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 47152 + }, + { + "epoch": 0.47153, + "grad_norm": 1.0822512408951372, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 47153 + }, + { + "epoch": 0.47154, + "grad_norm": 1.0929804150435298, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 47154 + }, + { + "epoch": 0.47155, + "grad_norm": 1.149207413018262, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 47155 + }, + { + "epoch": 0.47156, + "grad_norm": 1.0191974624616011, + "learning_rate": 0.003, + "loss": 4.026, + "step": 47156 + }, + { + "epoch": 0.47157, + "grad_norm": 1.1039719050077954, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 47157 + }, + { + "epoch": 0.47158, + "grad_norm": 1.1434950465261087, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 47158 + }, + { + "epoch": 0.47159, + "grad_norm": 0.9319781319018853, + "learning_rate": 0.003, + "loss": 4.035, + "step": 47159 + }, + { + "epoch": 0.4716, + "grad_norm": 1.2619288634300707, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 47160 + }, + { + "epoch": 0.47161, + "grad_norm": 1.1026491035585462, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 47161 + }, + { + "epoch": 0.47162, + "grad_norm": 0.8597687585773796, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 47162 + }, + { + "epoch": 0.47163, + "grad_norm": 0.8635570683777387, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 47163 + }, + { + "epoch": 0.47164, + "grad_norm": 1.1215038551750864, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 47164 + }, + { + "epoch": 0.47165, + "grad_norm": 1.1778907325512367, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 47165 + }, + { + "epoch": 0.47166, + "grad_norm": 0.9701149876548635, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 47166 + }, + { + "epoch": 0.47167, + "grad_norm": 1.0215060297769556, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 47167 + }, + { + "epoch": 0.47168, + "grad_norm": 1.08413410814241, + "learning_rate": 0.003, + "loss": 4.065, + "step": 47168 + }, + { + "epoch": 0.47169, + "grad_norm": 1.1581510037717535, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 47169 + }, + { + "epoch": 0.4717, + "grad_norm": 0.967459952282485, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 47170 + }, + { + "epoch": 0.47171, + "grad_norm": 1.0096173567673834, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 47171 + }, + { + "epoch": 0.47172, + "grad_norm": 1.1064225674096948, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 47172 + }, + { + "epoch": 0.47173, + "grad_norm": 0.8918599219282749, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 47173 + }, + { + "epoch": 0.47174, + "grad_norm": 0.9267498765958608, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 47174 + }, + { + "epoch": 0.47175, + "grad_norm": 1.0350873007983843, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 47175 + }, + { + "epoch": 0.47176, + "grad_norm": 1.0170181391335449, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 47176 + }, + { + "epoch": 0.47177, + "grad_norm": 1.0767736734353512, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 47177 + }, + { + "epoch": 0.47178, + "grad_norm": 0.9341439891661066, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 47178 + }, + { + "epoch": 0.47179, + "grad_norm": 1.0765033056843882, + "learning_rate": 0.003, + "loss": 4.053, + "step": 47179 + }, + { + "epoch": 0.4718, + "grad_norm": 1.0230874971158845, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 47180 + }, + { + "epoch": 0.47181, + "grad_norm": 1.1318186117094444, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 47181 + }, + { + "epoch": 0.47182, + "grad_norm": 1.1523340897087866, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 47182 + }, + { + "epoch": 0.47183, + "grad_norm": 1.3070651950135905, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 47183 + }, + { + "epoch": 0.47184, + "grad_norm": 0.9851989610004083, + "learning_rate": 0.003, + "loss": 4.032, + "step": 47184 + }, + { + "epoch": 0.47185, + "grad_norm": 0.9567050142175367, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 47185 + }, + { + "epoch": 0.47186, + "grad_norm": 1.1184057508978287, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 47186 + }, + { + "epoch": 0.47187, + "grad_norm": 0.9345653117346293, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 47187 + }, + { + "epoch": 0.47188, + "grad_norm": 1.011050279165983, + "learning_rate": 0.003, + "loss": 4.044, + "step": 47188 + }, + { + "epoch": 0.47189, + "grad_norm": 1.1338357755492847, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 47189 + }, + { + "epoch": 0.4719, + "grad_norm": 0.9561814635876247, + "learning_rate": 0.003, + "loss": 4.07, + "step": 47190 + }, + { + "epoch": 0.47191, + "grad_norm": 1.0978066126638053, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 47191 + }, + { + "epoch": 0.47192, + "grad_norm": 0.891791319278807, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 47192 + }, + { + "epoch": 0.47193, + "grad_norm": 1.1941931719038454, + "learning_rate": 0.003, + "loss": 4.015, + "step": 47193 + }, + { + "epoch": 0.47194, + "grad_norm": 1.1555933349120397, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 47194 + }, + { + "epoch": 0.47195, + "grad_norm": 0.9222600203476744, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 47195 + }, + { + "epoch": 0.47196, + "grad_norm": 1.1157589575217346, + "learning_rate": 0.003, + "loss": 4.066, + "step": 47196 + }, + { + "epoch": 0.47197, + "grad_norm": 1.157142095838771, + "learning_rate": 0.003, + "loss": 4.036, + "step": 47197 + }, + { + "epoch": 0.47198, + "grad_norm": 1.1882499986593373, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 47198 + }, + { + "epoch": 0.47199, + "grad_norm": 0.990530891713764, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 47199 + }, + { + "epoch": 0.472, + "grad_norm": 1.0369495726248672, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 47200 + }, + { + "epoch": 0.47201, + "grad_norm": 1.1489186675285668, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 47201 + }, + { + "epoch": 0.47202, + "grad_norm": 1.0660526165117177, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 47202 + }, + { + "epoch": 0.47203, + "grad_norm": 1.0038323517928918, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 47203 + }, + { + "epoch": 0.47204, + "grad_norm": 1.2092755900810563, + "learning_rate": 0.003, + "loss": 4.037, + "step": 47204 + }, + { + "epoch": 0.47205, + "grad_norm": 1.0607217818084511, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 47205 + }, + { + "epoch": 0.47206, + "grad_norm": 1.1772131346049879, + "learning_rate": 0.003, + "loss": 4.028, + "step": 47206 + }, + { + "epoch": 0.47207, + "grad_norm": 1.0073832104763012, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 47207 + }, + { + "epoch": 0.47208, + "grad_norm": 1.278732083588577, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 47208 + }, + { + "epoch": 0.47209, + "grad_norm": 0.9769118320378171, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 47209 + }, + { + "epoch": 0.4721, + "grad_norm": 0.97020991003699, + "learning_rate": 0.003, + "loss": 4.059, + "step": 47210 + }, + { + "epoch": 0.47211, + "grad_norm": 1.068736141193329, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 47211 + }, + { + "epoch": 0.47212, + "grad_norm": 0.9882817248005289, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 47212 + }, + { + "epoch": 0.47213, + "grad_norm": 1.029083993390518, + "learning_rate": 0.003, + "loss": 4.022, + "step": 47213 + }, + { + "epoch": 0.47214, + "grad_norm": 0.8827933856786863, + "learning_rate": 0.003, + "loss": 4.026, + "step": 47214 + }, + { + "epoch": 0.47215, + "grad_norm": 0.9455711802347037, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 47215 + }, + { + "epoch": 0.47216, + "grad_norm": 0.9872012619500881, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 47216 + }, + { + "epoch": 0.47217, + "grad_norm": 1.0262463612903896, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 47217 + }, + { + "epoch": 0.47218, + "grad_norm": 1.1001953432289473, + "learning_rate": 0.003, + "loss": 4.025, + "step": 47218 + }, + { + "epoch": 0.47219, + "grad_norm": 1.2015075946724465, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 47219 + }, + { + "epoch": 0.4722, + "grad_norm": 0.947792900509673, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 47220 + }, + { + "epoch": 0.47221, + "grad_norm": 1.149798170319063, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 47221 + }, + { + "epoch": 0.47222, + "grad_norm": 1.0646623530408708, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 47222 + }, + { + "epoch": 0.47223, + "grad_norm": 1.0551318443461501, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 47223 + }, + { + "epoch": 0.47224, + "grad_norm": 1.3840511142970493, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 47224 + }, + { + "epoch": 0.47225, + "grad_norm": 1.3449107292849665, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 47225 + }, + { + "epoch": 0.47226, + "grad_norm": 0.7893772414105577, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 47226 + }, + { + "epoch": 0.47227, + "grad_norm": 0.8245328847363401, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 47227 + }, + { + "epoch": 0.47228, + "grad_norm": 1.1153164852141733, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 47228 + }, + { + "epoch": 0.47229, + "grad_norm": 1.0830142637013143, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 47229 + }, + { + "epoch": 0.4723, + "grad_norm": 1.2264549210169395, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 47230 + }, + { + "epoch": 0.47231, + "grad_norm": 1.1415936460904594, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 47231 + }, + { + "epoch": 0.47232, + "grad_norm": 1.1328468924369366, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 47232 + }, + { + "epoch": 0.47233, + "grad_norm": 0.9554197504983228, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 47233 + }, + { + "epoch": 0.47234, + "grad_norm": 0.948490780591899, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 47234 + }, + { + "epoch": 0.47235, + "grad_norm": 0.9988024791918695, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 47235 + }, + { + "epoch": 0.47236, + "grad_norm": 1.431405524688099, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 47236 + }, + { + "epoch": 0.47237, + "grad_norm": 0.9705536487583136, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 47237 + }, + { + "epoch": 0.47238, + "grad_norm": 1.0038641850099863, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 47238 + }, + { + "epoch": 0.47239, + "grad_norm": 1.0732194116825662, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 47239 + }, + { + "epoch": 0.4724, + "grad_norm": 0.9596288115101353, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 47240 + }, + { + "epoch": 0.47241, + "grad_norm": 0.9908411621479905, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 47241 + }, + { + "epoch": 0.47242, + "grad_norm": 1.1244939959149796, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 47242 + }, + { + "epoch": 0.47243, + "grad_norm": 1.0111210632613215, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 47243 + }, + { + "epoch": 0.47244, + "grad_norm": 0.9298551149862476, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 47244 + }, + { + "epoch": 0.47245, + "grad_norm": 1.2069217798214495, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 47245 + }, + { + "epoch": 0.47246, + "grad_norm": 1.138415764734701, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 47246 + }, + { + "epoch": 0.47247, + "grad_norm": 1.4091968446273626, + "learning_rate": 0.003, + "loss": 4.022, + "step": 47247 + }, + { + "epoch": 0.47248, + "grad_norm": 0.7486711157719603, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 47248 + }, + { + "epoch": 0.47249, + "grad_norm": 0.8197378357127573, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 47249 + }, + { + "epoch": 0.4725, + "grad_norm": 0.8973855753644547, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 47250 + }, + { + "epoch": 0.47251, + "grad_norm": 1.0994271838618608, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 47251 + }, + { + "epoch": 0.47252, + "grad_norm": 1.1663049781795152, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 47252 + }, + { + "epoch": 0.47253, + "grad_norm": 0.9798545408061, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 47253 + }, + { + "epoch": 0.47254, + "grad_norm": 1.2158869817670128, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 47254 + }, + { + "epoch": 0.47255, + "grad_norm": 0.8729378142830474, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 47255 + }, + { + "epoch": 0.47256, + "grad_norm": 1.1490970259221822, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 47256 + }, + { + "epoch": 0.47257, + "grad_norm": 1.3246805898710956, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 47257 + }, + { + "epoch": 0.47258, + "grad_norm": 0.8474751427608859, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 47258 + }, + { + "epoch": 0.47259, + "grad_norm": 0.8739765534016928, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 47259 + }, + { + "epoch": 0.4726, + "grad_norm": 1.068900193161689, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 47260 + }, + { + "epoch": 0.47261, + "grad_norm": 1.088144885569565, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 47261 + }, + { + "epoch": 0.47262, + "grad_norm": 1.2821670019581035, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 47262 + }, + { + "epoch": 0.47263, + "grad_norm": 1.0484268481054717, + "learning_rate": 0.003, + "loss": 4.026, + "step": 47263 + }, + { + "epoch": 0.47264, + "grad_norm": 0.9490191648115275, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 47264 + }, + { + "epoch": 0.47265, + "grad_norm": 1.060468647005238, + "learning_rate": 0.003, + "loss": 4.036, + "step": 47265 + }, + { + "epoch": 0.47266, + "grad_norm": 1.1265242876504915, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 47266 + }, + { + "epoch": 0.47267, + "grad_norm": 0.9939965788096977, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 47267 + }, + { + "epoch": 0.47268, + "grad_norm": 1.094912927057855, + "learning_rate": 0.003, + "loss": 4.026, + "step": 47268 + }, + { + "epoch": 0.47269, + "grad_norm": 0.8984673907156586, + "learning_rate": 0.003, + "loss": 4.014, + "step": 47269 + }, + { + "epoch": 0.4727, + "grad_norm": 0.7924325704186926, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 47270 + }, + { + "epoch": 0.47271, + "grad_norm": 0.7490133190870513, + "learning_rate": 0.003, + "loss": 4.036, + "step": 47271 + }, + { + "epoch": 0.47272, + "grad_norm": 0.9252689221397681, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 47272 + }, + { + "epoch": 0.47273, + "grad_norm": 1.1109212698662707, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 47273 + }, + { + "epoch": 0.47274, + "grad_norm": 0.8500454575443239, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 47274 + }, + { + "epoch": 0.47275, + "grad_norm": 0.9160081111472108, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 47275 + }, + { + "epoch": 0.47276, + "grad_norm": 0.9224886223981974, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 47276 + }, + { + "epoch": 0.47277, + "grad_norm": 1.1043012968287163, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 47277 + }, + { + "epoch": 0.47278, + "grad_norm": 1.1159788248165172, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 47278 + }, + { + "epoch": 0.47279, + "grad_norm": 1.1591704896689818, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 47279 + }, + { + "epoch": 0.4728, + "grad_norm": 1.0356852244646548, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 47280 + }, + { + "epoch": 0.47281, + "grad_norm": 1.1765884343181123, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 47281 + }, + { + "epoch": 0.47282, + "grad_norm": 1.2801643195253314, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 47282 + }, + { + "epoch": 0.47283, + "grad_norm": 1.2987967621946974, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 47283 + }, + { + "epoch": 0.47284, + "grad_norm": 0.8791573300587633, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 47284 + }, + { + "epoch": 0.47285, + "grad_norm": 0.8908244686205332, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 47285 + }, + { + "epoch": 0.47286, + "grad_norm": 1.10766229571132, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 47286 + }, + { + "epoch": 0.47287, + "grad_norm": 1.2756658006529078, + "learning_rate": 0.003, + "loss": 4.06, + "step": 47287 + }, + { + "epoch": 0.47288, + "grad_norm": 0.9139559361677165, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 47288 + }, + { + "epoch": 0.47289, + "grad_norm": 0.9997786068254041, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 47289 + }, + { + "epoch": 0.4729, + "grad_norm": 1.4578300546271947, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 47290 + }, + { + "epoch": 0.47291, + "grad_norm": 0.8509916837778841, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 47291 + }, + { + "epoch": 0.47292, + "grad_norm": 0.9380571671469324, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 47292 + }, + { + "epoch": 0.47293, + "grad_norm": 1.197021676020931, + "learning_rate": 0.003, + "loss": 4.015, + "step": 47293 + }, + { + "epoch": 0.47294, + "grad_norm": 1.272092824549448, + "learning_rate": 0.003, + "loss": 4.047, + "step": 47294 + }, + { + "epoch": 0.47295, + "grad_norm": 1.0594484070271744, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 47295 + }, + { + "epoch": 0.47296, + "grad_norm": 0.9280784042791473, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 47296 + }, + { + "epoch": 0.47297, + "grad_norm": 0.9313333452320757, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 47297 + }, + { + "epoch": 0.47298, + "grad_norm": 1.2670253361914436, + "learning_rate": 0.003, + "loss": 4.056, + "step": 47298 + }, + { + "epoch": 0.47299, + "grad_norm": 0.8812129456370199, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 47299 + }, + { + "epoch": 0.473, + "grad_norm": 0.9212146970473368, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 47300 + }, + { + "epoch": 0.47301, + "grad_norm": 0.885036819747353, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 47301 + }, + { + "epoch": 0.47302, + "grad_norm": 1.0078823981199219, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 47302 + }, + { + "epoch": 0.47303, + "grad_norm": 1.080189454787792, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 47303 + }, + { + "epoch": 0.47304, + "grad_norm": 0.991882215975573, + "learning_rate": 0.003, + "loss": 4.046, + "step": 47304 + }, + { + "epoch": 0.47305, + "grad_norm": 1.209757899607912, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 47305 + }, + { + "epoch": 0.47306, + "grad_norm": 1.0295788231740646, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 47306 + }, + { + "epoch": 0.47307, + "grad_norm": 1.2657833141003163, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 47307 + }, + { + "epoch": 0.47308, + "grad_norm": 0.9072148988529763, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 47308 + }, + { + "epoch": 0.47309, + "grad_norm": 0.8401058757224532, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 47309 + }, + { + "epoch": 0.4731, + "grad_norm": 1.1389126979821167, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 47310 + }, + { + "epoch": 0.47311, + "grad_norm": 1.3436886068252687, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 47311 + }, + { + "epoch": 0.47312, + "grad_norm": 0.7817080201083535, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 47312 + }, + { + "epoch": 0.47313, + "grad_norm": 0.7449405197880007, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 47313 + }, + { + "epoch": 0.47314, + "grad_norm": 0.8149973427847396, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 47314 + }, + { + "epoch": 0.47315, + "grad_norm": 0.8011277017855114, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 47315 + }, + { + "epoch": 0.47316, + "grad_norm": 1.1071308458037965, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 47316 + }, + { + "epoch": 0.47317, + "grad_norm": 1.175627337331059, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 47317 + }, + { + "epoch": 0.47318, + "grad_norm": 1.0600635147826465, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 47318 + }, + { + "epoch": 0.47319, + "grad_norm": 1.1880605992052364, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 47319 + }, + { + "epoch": 0.4732, + "grad_norm": 1.0576352648333027, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 47320 + }, + { + "epoch": 0.47321, + "grad_norm": 1.3006643328759324, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 47321 + }, + { + "epoch": 0.47322, + "grad_norm": 0.8815355184882778, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 47322 + }, + { + "epoch": 0.47323, + "grad_norm": 0.9023962817168447, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 47323 + }, + { + "epoch": 0.47324, + "grad_norm": 1.1420814769344947, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 47324 + }, + { + "epoch": 0.47325, + "grad_norm": 1.0549284720152723, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 47325 + }, + { + "epoch": 0.47326, + "grad_norm": 0.9779759019066637, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 47326 + }, + { + "epoch": 0.47327, + "grad_norm": 1.14514391362014, + "learning_rate": 0.003, + "loss": 4.069, + "step": 47327 + }, + { + "epoch": 0.47328, + "grad_norm": 1.1956160411960477, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 47328 + }, + { + "epoch": 0.47329, + "grad_norm": 0.7594018400708763, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 47329 + }, + { + "epoch": 0.4733, + "grad_norm": 0.8142149992170331, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 47330 + }, + { + "epoch": 0.47331, + "grad_norm": 0.9192271563976913, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 47331 + }, + { + "epoch": 0.47332, + "grad_norm": 1.21900475928042, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 47332 + }, + { + "epoch": 0.47333, + "grad_norm": 0.9680787096751817, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 47333 + }, + { + "epoch": 0.47334, + "grad_norm": 1.0460313481293817, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 47334 + }, + { + "epoch": 0.47335, + "grad_norm": 1.0816323192627375, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 47335 + }, + { + "epoch": 0.47336, + "grad_norm": 0.9912838890329745, + "learning_rate": 0.003, + "loss": 4.06, + "step": 47336 + }, + { + "epoch": 0.47337, + "grad_norm": 1.1057081756884248, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 47337 + }, + { + "epoch": 0.47338, + "grad_norm": 1.3525961481365507, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 47338 + }, + { + "epoch": 0.47339, + "grad_norm": 0.881929902681015, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 47339 + }, + { + "epoch": 0.4734, + "grad_norm": 0.9028872189364596, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 47340 + }, + { + "epoch": 0.47341, + "grad_norm": 1.1078839218120955, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 47341 + }, + { + "epoch": 0.47342, + "grad_norm": 1.3778578127086236, + "learning_rate": 0.003, + "loss": 4.051, + "step": 47342 + }, + { + "epoch": 0.47343, + "grad_norm": 1.0466724114421704, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 47343 + }, + { + "epoch": 0.47344, + "grad_norm": 1.0313731677619877, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 47344 + }, + { + "epoch": 0.47345, + "grad_norm": 1.2052460704382009, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 47345 + }, + { + "epoch": 0.47346, + "grad_norm": 0.9918769629789614, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 47346 + }, + { + "epoch": 0.47347, + "grad_norm": 1.2164800931973931, + "learning_rate": 0.003, + "loss": 4.004, + "step": 47347 + }, + { + "epoch": 0.47348, + "grad_norm": 1.0826398404142477, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 47348 + }, + { + "epoch": 0.47349, + "grad_norm": 1.0524331090637162, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 47349 + }, + { + "epoch": 0.4735, + "grad_norm": 1.2360853016040532, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 47350 + }, + { + "epoch": 0.47351, + "grad_norm": 0.940701345628974, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 47351 + }, + { + "epoch": 0.47352, + "grad_norm": 1.069536160080996, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 47352 + }, + { + "epoch": 0.47353, + "grad_norm": 0.9694491812386744, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 47353 + }, + { + "epoch": 0.47354, + "grad_norm": 1.0531329156700393, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 47354 + }, + { + "epoch": 0.47355, + "grad_norm": 1.3021270841008123, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 47355 + }, + { + "epoch": 0.47356, + "grad_norm": 0.8434152807180602, + "learning_rate": 0.003, + "loss": 4.087, + "step": 47356 + }, + { + "epoch": 0.47357, + "grad_norm": 0.8655288588275214, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 47357 + }, + { + "epoch": 0.47358, + "grad_norm": 1.1093138198187262, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 47358 + }, + { + "epoch": 0.47359, + "grad_norm": 1.33379887146566, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 47359 + }, + { + "epoch": 0.4736, + "grad_norm": 0.9126725677188489, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 47360 + }, + { + "epoch": 0.47361, + "grad_norm": 0.9981669335940576, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 47361 + }, + { + "epoch": 0.47362, + "grad_norm": 1.2161848926483272, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 47362 + }, + { + "epoch": 0.47363, + "grad_norm": 1.023112616063265, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 47363 + }, + { + "epoch": 0.47364, + "grad_norm": 1.265827906561684, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 47364 + }, + { + "epoch": 0.47365, + "grad_norm": 0.7875992436381684, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 47365 + }, + { + "epoch": 0.47366, + "grad_norm": 0.7938738801266337, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 47366 + }, + { + "epoch": 0.47367, + "grad_norm": 1.079846314413868, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 47367 + }, + { + "epoch": 0.47368, + "grad_norm": 1.0077134568514445, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 47368 + }, + { + "epoch": 0.47369, + "grad_norm": 1.183546519806548, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 47369 + }, + { + "epoch": 0.4737, + "grad_norm": 0.8782426343704459, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 47370 + }, + { + "epoch": 0.47371, + "grad_norm": 1.0965077484262986, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 47371 + }, + { + "epoch": 0.47372, + "grad_norm": 1.4679505880706896, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 47372 + }, + { + "epoch": 0.47373, + "grad_norm": 0.8642682358023746, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 47373 + }, + { + "epoch": 0.47374, + "grad_norm": 0.9698001731146491, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 47374 + }, + { + "epoch": 0.47375, + "grad_norm": 1.1476126491850118, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 47375 + }, + { + "epoch": 0.47376, + "grad_norm": 1.0451905547393412, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 47376 + }, + { + "epoch": 0.47377, + "grad_norm": 0.9904907713467676, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 47377 + }, + { + "epoch": 0.47378, + "grad_norm": 0.9193761241697809, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 47378 + }, + { + "epoch": 0.47379, + "grad_norm": 1.0515109118727173, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 47379 + }, + { + "epoch": 0.4738, + "grad_norm": 1.1039787767350056, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 47380 + }, + { + "epoch": 0.47381, + "grad_norm": 0.99499252382198, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 47381 + }, + { + "epoch": 0.47382, + "grad_norm": 1.1358215321906993, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 47382 + }, + { + "epoch": 0.47383, + "grad_norm": 0.9841193981205727, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 47383 + }, + { + "epoch": 0.47384, + "grad_norm": 1.2436511662189742, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 47384 + }, + { + "epoch": 0.47385, + "grad_norm": 0.8934014488111485, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 47385 + }, + { + "epoch": 0.47386, + "grad_norm": 1.056078693071865, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 47386 + }, + { + "epoch": 0.47387, + "grad_norm": 1.2030801964926943, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 47387 + }, + { + "epoch": 0.47388, + "grad_norm": 1.1034646442172744, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 47388 + }, + { + "epoch": 0.47389, + "grad_norm": 1.0499138462405047, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 47389 + }, + { + "epoch": 0.4739, + "grad_norm": 1.1053160783650384, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 47390 + }, + { + "epoch": 0.47391, + "grad_norm": 1.0844139804871857, + "learning_rate": 0.003, + "loss": 4.013, + "step": 47391 + }, + { + "epoch": 0.47392, + "grad_norm": 1.3846986080121566, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 47392 + }, + { + "epoch": 0.47393, + "grad_norm": 0.9664655532270297, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 47393 + }, + { + "epoch": 0.47394, + "grad_norm": 1.275847892127107, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 47394 + }, + { + "epoch": 0.47395, + "grad_norm": 1.199133403080432, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 47395 + }, + { + "epoch": 0.47396, + "grad_norm": 1.0152226105074131, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 47396 + }, + { + "epoch": 0.47397, + "grad_norm": 1.2474431764106568, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 47397 + }, + { + "epoch": 0.47398, + "grad_norm": 1.0308484517126522, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 47398 + }, + { + "epoch": 0.47399, + "grad_norm": 1.0305615415338405, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 47399 + }, + { + "epoch": 0.474, + "grad_norm": 1.234483791264893, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 47400 + }, + { + "epoch": 0.47401, + "grad_norm": 0.9015140436555331, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 47401 + }, + { + "epoch": 0.47402, + "grad_norm": 0.7432079941091433, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 47402 + }, + { + "epoch": 0.47403, + "grad_norm": 0.873204592786412, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 47403 + }, + { + "epoch": 0.47404, + "grad_norm": 0.9260592206500522, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 47404 + }, + { + "epoch": 0.47405, + "grad_norm": 1.041571700263793, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 47405 + }, + { + "epoch": 0.47406, + "grad_norm": 0.8580953520588336, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 47406 + }, + { + "epoch": 0.47407, + "grad_norm": 0.8964980524439288, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 47407 + }, + { + "epoch": 0.47408, + "grad_norm": 1.1693268470444926, + "learning_rate": 0.003, + "loss": 4.043, + "step": 47408 + }, + { + "epoch": 0.47409, + "grad_norm": 1.0469552672318252, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 47409 + }, + { + "epoch": 0.4741, + "grad_norm": 1.1154798413743343, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 47410 + }, + { + "epoch": 0.47411, + "grad_norm": 1.1674635115112963, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 47411 + }, + { + "epoch": 0.47412, + "grad_norm": 1.1112163305484455, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 47412 + }, + { + "epoch": 0.47413, + "grad_norm": 1.0833190842685787, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 47413 + }, + { + "epoch": 0.47414, + "grad_norm": 0.9468960997685735, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 47414 + }, + { + "epoch": 0.47415, + "grad_norm": 1.0028226077281972, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 47415 + }, + { + "epoch": 0.47416, + "grad_norm": 1.328115131921231, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 47416 + }, + { + "epoch": 0.47417, + "grad_norm": 0.9913206994414465, + "learning_rate": 0.003, + "loss": 4.029, + "step": 47417 + }, + { + "epoch": 0.47418, + "grad_norm": 0.9980840003787388, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 47418 + }, + { + "epoch": 0.47419, + "grad_norm": 1.0169915358988884, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 47419 + }, + { + "epoch": 0.4742, + "grad_norm": 0.9502892656083647, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 47420 + }, + { + "epoch": 0.47421, + "grad_norm": 1.1319862327573276, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 47421 + }, + { + "epoch": 0.47422, + "grad_norm": 1.0247625943295458, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 47422 + }, + { + "epoch": 0.47423, + "grad_norm": 1.1808417002621912, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 47423 + }, + { + "epoch": 0.47424, + "grad_norm": 0.9975135020418531, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 47424 + }, + { + "epoch": 0.47425, + "grad_norm": 1.1975070745940448, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 47425 + }, + { + "epoch": 0.47426, + "grad_norm": 0.9313572234045896, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 47426 + }, + { + "epoch": 0.47427, + "grad_norm": 0.9133007903885131, + "learning_rate": 0.003, + "loss": 4.042, + "step": 47427 + }, + { + "epoch": 0.47428, + "grad_norm": 0.9649806790585915, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 47428 + }, + { + "epoch": 0.47429, + "grad_norm": 1.1158178664463543, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 47429 + }, + { + "epoch": 0.4743, + "grad_norm": 0.9997085164526459, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 47430 + }, + { + "epoch": 0.47431, + "grad_norm": 1.1913770538158093, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 47431 + }, + { + "epoch": 0.47432, + "grad_norm": 1.0762758133252315, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 47432 + }, + { + "epoch": 0.47433, + "grad_norm": 1.2293403746149298, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 47433 + }, + { + "epoch": 0.47434, + "grad_norm": 1.0270619107113188, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 47434 + }, + { + "epoch": 0.47435, + "grad_norm": 1.1572202694809455, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 47435 + }, + { + "epoch": 0.47436, + "grad_norm": 1.0689410685869618, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 47436 + }, + { + "epoch": 0.47437, + "grad_norm": 1.155611038220423, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 47437 + }, + { + "epoch": 0.47438, + "grad_norm": 0.9398353040633759, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 47438 + }, + { + "epoch": 0.47439, + "grad_norm": 0.9488701918969357, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 47439 + }, + { + "epoch": 0.4744, + "grad_norm": 1.3764953682455798, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 47440 + }, + { + "epoch": 0.47441, + "grad_norm": 1.0546407925281442, + "learning_rate": 0.003, + "loss": 4.032, + "step": 47441 + }, + { + "epoch": 0.47442, + "grad_norm": 1.0213340299245774, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 47442 + }, + { + "epoch": 0.47443, + "grad_norm": 1.0857499564727737, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 47443 + }, + { + "epoch": 0.47444, + "grad_norm": 1.0629781167073418, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 47444 + }, + { + "epoch": 0.47445, + "grad_norm": 1.191486537204772, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 47445 + }, + { + "epoch": 0.47446, + "grad_norm": 0.9425320165879446, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 47446 + }, + { + "epoch": 0.47447, + "grad_norm": 0.9898473843002542, + "learning_rate": 0.003, + "loss": 4.018, + "step": 47447 + }, + { + "epoch": 0.47448, + "grad_norm": 1.230449574670339, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 47448 + }, + { + "epoch": 0.47449, + "grad_norm": 0.8461121096849368, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 47449 + }, + { + "epoch": 0.4745, + "grad_norm": 0.9383707040699585, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 47450 + }, + { + "epoch": 0.47451, + "grad_norm": 1.0663626067649852, + "learning_rate": 0.003, + "loss": 4.01, + "step": 47451 + }, + { + "epoch": 0.47452, + "grad_norm": 1.1548604757075789, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 47452 + }, + { + "epoch": 0.47453, + "grad_norm": 1.1140228707677158, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 47453 + }, + { + "epoch": 0.47454, + "grad_norm": 0.9120330393417602, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 47454 + }, + { + "epoch": 0.47455, + "grad_norm": 0.8463035117229628, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 47455 + }, + { + "epoch": 0.47456, + "grad_norm": 0.8133334143047493, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 47456 + }, + { + "epoch": 0.47457, + "grad_norm": 0.9574870088481443, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 47457 + }, + { + "epoch": 0.47458, + "grad_norm": 1.0524593033707261, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 47458 + }, + { + "epoch": 0.47459, + "grad_norm": 1.0800518666643455, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 47459 + }, + { + "epoch": 0.4746, + "grad_norm": 1.4484700255415282, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 47460 + }, + { + "epoch": 0.47461, + "grad_norm": 0.9475128819558624, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 47461 + }, + { + "epoch": 0.47462, + "grad_norm": 1.204496951018478, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 47462 + }, + { + "epoch": 0.47463, + "grad_norm": 1.0325488859529617, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 47463 + }, + { + "epoch": 0.47464, + "grad_norm": 1.236435798107091, + "learning_rate": 0.003, + "loss": 4.035, + "step": 47464 + }, + { + "epoch": 0.47465, + "grad_norm": 1.035074311199802, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 47465 + }, + { + "epoch": 0.47466, + "grad_norm": 1.1174006167279342, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 47466 + }, + { + "epoch": 0.47467, + "grad_norm": 1.0431938676997825, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 47467 + }, + { + "epoch": 0.47468, + "grad_norm": 1.386348434347027, + "learning_rate": 0.003, + "loss": 4.042, + "step": 47468 + }, + { + "epoch": 0.47469, + "grad_norm": 0.8698313437829478, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 47469 + }, + { + "epoch": 0.4747, + "grad_norm": 0.880782398030305, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 47470 + }, + { + "epoch": 0.47471, + "grad_norm": 0.9979979970724024, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 47471 + }, + { + "epoch": 0.47472, + "grad_norm": 1.0183437942978093, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 47472 + }, + { + "epoch": 0.47473, + "grad_norm": 1.0829597118897007, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 47473 + }, + { + "epoch": 0.47474, + "grad_norm": 1.016143264810431, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 47474 + }, + { + "epoch": 0.47475, + "grad_norm": 0.9504458995172019, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 47475 + }, + { + "epoch": 0.47476, + "grad_norm": 1.0289869432687935, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 47476 + }, + { + "epoch": 0.47477, + "grad_norm": 1.2427217325501423, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 47477 + }, + { + "epoch": 0.47478, + "grad_norm": 1.1211614160875758, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 47478 + }, + { + "epoch": 0.47479, + "grad_norm": 1.1346100757352768, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 47479 + }, + { + "epoch": 0.4748, + "grad_norm": 1.0690527483177585, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 47480 + }, + { + "epoch": 0.47481, + "grad_norm": 1.1073769906775037, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 47481 + }, + { + "epoch": 0.47482, + "grad_norm": 1.027893431358507, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 47482 + }, + { + "epoch": 0.47483, + "grad_norm": 1.160562614101432, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 47483 + }, + { + "epoch": 0.47484, + "grad_norm": 0.9436789017000448, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 47484 + }, + { + "epoch": 0.47485, + "grad_norm": 1.1388296722015763, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 47485 + }, + { + "epoch": 0.47486, + "grad_norm": 1.0176896747099091, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 47486 + }, + { + "epoch": 0.47487, + "grad_norm": 1.1431455472918635, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 47487 + }, + { + "epoch": 0.47488, + "grad_norm": 0.9484033784057869, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 47488 + }, + { + "epoch": 0.47489, + "grad_norm": 1.0048004999562157, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 47489 + }, + { + "epoch": 0.4749, + "grad_norm": 1.1536535424046643, + "learning_rate": 0.003, + "loss": 4.058, + "step": 47490 + }, + { + "epoch": 0.47491, + "grad_norm": 0.9076748904475349, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 47491 + }, + { + "epoch": 0.47492, + "grad_norm": 0.9733571619596972, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 47492 + }, + { + "epoch": 0.47493, + "grad_norm": 1.2053078220735403, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 47493 + }, + { + "epoch": 0.47494, + "grad_norm": 0.9501968935376363, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 47494 + }, + { + "epoch": 0.47495, + "grad_norm": 0.9918731399487182, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 47495 + }, + { + "epoch": 0.47496, + "grad_norm": 1.2655822495948976, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 47496 + }, + { + "epoch": 0.47497, + "grad_norm": 1.1720466191992636, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 47497 + }, + { + "epoch": 0.47498, + "grad_norm": 1.3330834160095075, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 47498 + }, + { + "epoch": 0.47499, + "grad_norm": 1.0220531204276244, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 47499 + }, + { + "epoch": 0.475, + "grad_norm": 0.9434544021465373, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 47500 + }, + { + "epoch": 0.47501, + "grad_norm": 1.3265579536300989, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 47501 + }, + { + "epoch": 0.47502, + "grad_norm": 1.0611902172758767, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 47502 + }, + { + "epoch": 0.47503, + "grad_norm": 1.1133215598866273, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 47503 + }, + { + "epoch": 0.47504, + "grad_norm": 0.8056817177775707, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 47504 + }, + { + "epoch": 0.47505, + "grad_norm": 0.8438231820333787, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 47505 + }, + { + "epoch": 0.47506, + "grad_norm": 1.0665903824129006, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 47506 + }, + { + "epoch": 0.47507, + "grad_norm": 1.1993674423500464, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 47507 + }, + { + "epoch": 0.47508, + "grad_norm": 1.1072818429531777, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 47508 + }, + { + "epoch": 0.47509, + "grad_norm": 1.1208098860035451, + "learning_rate": 0.003, + "loss": 4.029, + "step": 47509 + }, + { + "epoch": 0.4751, + "grad_norm": 1.2055022382935656, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 47510 + }, + { + "epoch": 0.47511, + "grad_norm": 0.8959220792948612, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 47511 + }, + { + "epoch": 0.47512, + "grad_norm": 0.9020623950655223, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 47512 + }, + { + "epoch": 0.47513, + "grad_norm": 1.2023554424285463, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 47513 + }, + { + "epoch": 0.47514, + "grad_norm": 0.9557784997976764, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 47514 + }, + { + "epoch": 0.47515, + "grad_norm": 0.9708337771115206, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 47515 + }, + { + "epoch": 0.47516, + "grad_norm": 1.1564733721057618, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 47516 + }, + { + "epoch": 0.47517, + "grad_norm": 1.0297941432960072, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 47517 + }, + { + "epoch": 0.47518, + "grad_norm": 1.3589371744814362, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 47518 + }, + { + "epoch": 0.47519, + "grad_norm": 0.837192748108817, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 47519 + }, + { + "epoch": 0.4752, + "grad_norm": 0.801596868481655, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 47520 + }, + { + "epoch": 0.47521, + "grad_norm": 0.8439658456065179, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 47521 + }, + { + "epoch": 0.47522, + "grad_norm": 0.9852762788036976, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 47522 + }, + { + "epoch": 0.47523, + "grad_norm": 1.2836598990578691, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 47523 + }, + { + "epoch": 0.47524, + "grad_norm": 0.9842537159237994, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 47524 + }, + { + "epoch": 0.47525, + "grad_norm": 1.4163399785361015, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 47525 + }, + { + "epoch": 0.47526, + "grad_norm": 0.7702727029438302, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 47526 + }, + { + "epoch": 0.47527, + "grad_norm": 0.6136331433210251, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 47527 + }, + { + "epoch": 0.47528, + "grad_norm": 0.8384916870114224, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 47528 + }, + { + "epoch": 0.47529, + "grad_norm": 1.2655711491699675, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 47529 + }, + { + "epoch": 0.4753, + "grad_norm": 1.1786630726708753, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 47530 + }, + { + "epoch": 0.47531, + "grad_norm": 0.955638681149684, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 47531 + }, + { + "epoch": 0.47532, + "grad_norm": 0.9196977993102742, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 47532 + }, + { + "epoch": 0.47533, + "grad_norm": 1.050957534659656, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 47533 + }, + { + "epoch": 0.47534, + "grad_norm": 1.3908795049403841, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 47534 + }, + { + "epoch": 0.47535, + "grad_norm": 0.9347082399695344, + "learning_rate": 0.003, + "loss": 4.034, + "step": 47535 + }, + { + "epoch": 0.47536, + "grad_norm": 0.9792139356760068, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 47536 + }, + { + "epoch": 0.47537, + "grad_norm": 1.3099081711225584, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 47537 + }, + { + "epoch": 0.47538, + "grad_norm": 1.0647907945573936, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 47538 + }, + { + "epoch": 0.47539, + "grad_norm": 1.0020038405831473, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 47539 + }, + { + "epoch": 0.4754, + "grad_norm": 1.3084645537317872, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 47540 + }, + { + "epoch": 0.47541, + "grad_norm": 0.8999993183385981, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 47541 + }, + { + "epoch": 0.47542, + "grad_norm": 0.9009994480162957, + "learning_rate": 0.003, + "loss": 4.024, + "step": 47542 + }, + { + "epoch": 0.47543, + "grad_norm": 0.9077585908276963, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 47543 + }, + { + "epoch": 0.47544, + "grad_norm": 0.9870399050361867, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 47544 + }, + { + "epoch": 0.47545, + "grad_norm": 1.0501449893098158, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 47545 + }, + { + "epoch": 0.47546, + "grad_norm": 1.0251114372397232, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 47546 + }, + { + "epoch": 0.47547, + "grad_norm": 1.2203358869847845, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 47547 + }, + { + "epoch": 0.47548, + "grad_norm": 1.1254140165793112, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 47548 + }, + { + "epoch": 0.47549, + "grad_norm": 1.1260606929911725, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 47549 + }, + { + "epoch": 0.4755, + "grad_norm": 0.9556140073293301, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 47550 + }, + { + "epoch": 0.47551, + "grad_norm": 1.0760436502378805, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 47551 + }, + { + "epoch": 0.47552, + "grad_norm": 1.2409841415625635, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 47552 + }, + { + "epoch": 0.47553, + "grad_norm": 1.0347471639818344, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 47553 + }, + { + "epoch": 0.47554, + "grad_norm": 1.3484830728373016, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 47554 + }, + { + "epoch": 0.47555, + "grad_norm": 1.005026210199678, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 47555 + }, + { + "epoch": 0.47556, + "grad_norm": 1.167193207608081, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 47556 + }, + { + "epoch": 0.47557, + "grad_norm": 1.007264572889355, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 47557 + }, + { + "epoch": 0.47558, + "grad_norm": 0.9699538577214862, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 47558 + }, + { + "epoch": 0.47559, + "grad_norm": 1.037165449476097, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 47559 + }, + { + "epoch": 0.4756, + "grad_norm": 0.9689595912745659, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 47560 + }, + { + "epoch": 0.47561, + "grad_norm": 1.1008771013538732, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 47561 + }, + { + "epoch": 0.47562, + "grad_norm": 1.258699825109875, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 47562 + }, + { + "epoch": 0.47563, + "grad_norm": 0.9679057015637431, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 47563 + }, + { + "epoch": 0.47564, + "grad_norm": 0.9584687802730263, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 47564 + }, + { + "epoch": 0.47565, + "grad_norm": 1.1512811049235128, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 47565 + }, + { + "epoch": 0.47566, + "grad_norm": 1.193574193071406, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 47566 + }, + { + "epoch": 0.47567, + "grad_norm": 1.0755729634237798, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 47567 + }, + { + "epoch": 0.47568, + "grad_norm": 0.9258147191088926, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 47568 + }, + { + "epoch": 0.47569, + "grad_norm": 0.9644999899283914, + "learning_rate": 0.003, + "loss": 4.004, + "step": 47569 + }, + { + "epoch": 0.4757, + "grad_norm": 1.1263082163188611, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 47570 + }, + { + "epoch": 0.47571, + "grad_norm": 1.064755579804766, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 47571 + }, + { + "epoch": 0.47572, + "grad_norm": 1.072542124229266, + "learning_rate": 0.003, + "loss": 4.035, + "step": 47572 + }, + { + "epoch": 0.47573, + "grad_norm": 1.0171625994320255, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 47573 + }, + { + "epoch": 0.47574, + "grad_norm": 1.2636973093660955, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 47574 + }, + { + "epoch": 0.47575, + "grad_norm": 0.9834857009683976, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 47575 + }, + { + "epoch": 0.47576, + "grad_norm": 1.0774891134444822, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 47576 + }, + { + "epoch": 0.47577, + "grad_norm": 1.068750821506429, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 47577 + }, + { + "epoch": 0.47578, + "grad_norm": 1.0180069759830337, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 47578 + }, + { + "epoch": 0.47579, + "grad_norm": 1.1539074347263154, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 47579 + }, + { + "epoch": 0.4758, + "grad_norm": 1.1442505546135395, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 47580 + }, + { + "epoch": 0.47581, + "grad_norm": 0.9206321576454071, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 47581 + }, + { + "epoch": 0.47582, + "grad_norm": 0.8558128323948205, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 47582 + }, + { + "epoch": 0.47583, + "grad_norm": 0.9272939552630292, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 47583 + }, + { + "epoch": 0.47584, + "grad_norm": 1.2057883601432284, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 47584 + }, + { + "epoch": 0.47585, + "grad_norm": 0.9777357813115404, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 47585 + }, + { + "epoch": 0.47586, + "grad_norm": 1.0995741224882254, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 47586 + }, + { + "epoch": 0.47587, + "grad_norm": 1.0757159824627323, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 47587 + }, + { + "epoch": 0.47588, + "grad_norm": 1.1748298792517538, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 47588 + }, + { + "epoch": 0.47589, + "grad_norm": 1.002344879991844, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 47589 + }, + { + "epoch": 0.4759, + "grad_norm": 1.0937876974051202, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 47590 + }, + { + "epoch": 0.47591, + "grad_norm": 1.0040044228502245, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 47591 + }, + { + "epoch": 0.47592, + "grad_norm": 1.065552663214471, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 47592 + }, + { + "epoch": 0.47593, + "grad_norm": 1.0060424606616614, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 47593 + }, + { + "epoch": 0.47594, + "grad_norm": 1.233474448039573, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 47594 + }, + { + "epoch": 0.47595, + "grad_norm": 1.148889822600502, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 47595 + }, + { + "epoch": 0.47596, + "grad_norm": 1.1099635595091841, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 47596 + }, + { + "epoch": 0.47597, + "grad_norm": 0.9757184213999801, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 47597 + }, + { + "epoch": 0.47598, + "grad_norm": 1.241452891691366, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 47598 + }, + { + "epoch": 0.47599, + "grad_norm": 1.1048222403952308, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 47599 + }, + { + "epoch": 0.476, + "grad_norm": 1.61300497356374, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 47600 + }, + { + "epoch": 0.47601, + "grad_norm": 1.0296501141965635, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 47601 + }, + { + "epoch": 0.47602, + "grad_norm": 1.1100935425205891, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 47602 + }, + { + "epoch": 0.47603, + "grad_norm": 1.0109018720056488, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 47603 + }, + { + "epoch": 0.47604, + "grad_norm": 1.1450001487414059, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 47604 + }, + { + "epoch": 0.47605, + "grad_norm": 1.057452693412492, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 47605 + }, + { + "epoch": 0.47606, + "grad_norm": 1.100209407723787, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 47606 + }, + { + "epoch": 0.47607, + "grad_norm": 1.219347928387113, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 47607 + }, + { + "epoch": 0.47608, + "grad_norm": 1.0141347190385603, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 47608 + }, + { + "epoch": 0.47609, + "grad_norm": 1.2808572489374228, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 47609 + }, + { + "epoch": 0.4761, + "grad_norm": 0.9864538390348548, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 47610 + }, + { + "epoch": 0.47611, + "grad_norm": 0.9820972313392743, + "learning_rate": 0.003, + "loss": 4.036, + "step": 47611 + }, + { + "epoch": 0.47612, + "grad_norm": 1.2414320738429825, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 47612 + }, + { + "epoch": 0.47613, + "grad_norm": 1.0111295185690687, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 47613 + }, + { + "epoch": 0.47614, + "grad_norm": 1.1334948413201908, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 47614 + }, + { + "epoch": 0.47615, + "grad_norm": 1.082962934546363, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 47615 + }, + { + "epoch": 0.47616, + "grad_norm": 1.1354037563750177, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 47616 + }, + { + "epoch": 0.47617, + "grad_norm": 1.03416573823058, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 47617 + }, + { + "epoch": 0.47618, + "grad_norm": 0.9386319879662589, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 47618 + }, + { + "epoch": 0.47619, + "grad_norm": 1.0305999881162002, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 47619 + }, + { + "epoch": 0.4762, + "grad_norm": 1.0841155286054316, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 47620 + }, + { + "epoch": 0.47621, + "grad_norm": 1.1793745098291677, + "learning_rate": 0.003, + "loss": 4.041, + "step": 47621 + }, + { + "epoch": 0.47622, + "grad_norm": 0.8007510476893512, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 47622 + }, + { + "epoch": 0.47623, + "grad_norm": 0.7214794191357179, + "learning_rate": 0.003, + "loss": 4.021, + "step": 47623 + }, + { + "epoch": 0.47624, + "grad_norm": 0.8458591393784309, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 47624 + }, + { + "epoch": 0.47625, + "grad_norm": 1.0269021411820023, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 47625 + }, + { + "epoch": 0.47626, + "grad_norm": 1.1690809119447216, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 47626 + }, + { + "epoch": 0.47627, + "grad_norm": 1.1872370162992394, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 47627 + }, + { + "epoch": 0.47628, + "grad_norm": 0.9727874356007005, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 47628 + }, + { + "epoch": 0.47629, + "grad_norm": 1.074242599882503, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 47629 + }, + { + "epoch": 0.4763, + "grad_norm": 0.8676155230354953, + "learning_rate": 0.003, + "loss": 4.039, + "step": 47630 + }, + { + "epoch": 0.47631, + "grad_norm": 0.9146852370595505, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 47631 + }, + { + "epoch": 0.47632, + "grad_norm": 1.2799483316738687, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 47632 + }, + { + "epoch": 0.47633, + "grad_norm": 1.1612198306108839, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 47633 + }, + { + "epoch": 0.47634, + "grad_norm": 1.0134451742796005, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 47634 + }, + { + "epoch": 0.47635, + "grad_norm": 1.1056601188932391, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 47635 + }, + { + "epoch": 0.47636, + "grad_norm": 1.1250743261465428, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 47636 + }, + { + "epoch": 0.47637, + "grad_norm": 1.1165111301538255, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 47637 + }, + { + "epoch": 0.47638, + "grad_norm": 1.0902142617228623, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 47638 + }, + { + "epoch": 0.47639, + "grad_norm": 1.0764494598744592, + "learning_rate": 0.003, + "loss": 4.059, + "step": 47639 + }, + { + "epoch": 0.4764, + "grad_norm": 1.0342525409894554, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 47640 + }, + { + "epoch": 0.47641, + "grad_norm": 0.9931488372058068, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 47641 + }, + { + "epoch": 0.47642, + "grad_norm": 1.371997459169083, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 47642 + }, + { + "epoch": 0.47643, + "grad_norm": 0.9440489234767495, + "learning_rate": 0.003, + "loss": 4.028, + "step": 47643 + }, + { + "epoch": 0.47644, + "grad_norm": 1.1230363142304687, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 47644 + }, + { + "epoch": 0.47645, + "grad_norm": 1.2046118959536363, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 47645 + }, + { + "epoch": 0.47646, + "grad_norm": 0.9670851781210102, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 47646 + }, + { + "epoch": 0.47647, + "grad_norm": 1.1442801320961886, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 47647 + }, + { + "epoch": 0.47648, + "grad_norm": 0.8772513154480014, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 47648 + }, + { + "epoch": 0.47649, + "grad_norm": 0.8959038701221385, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 47649 + }, + { + "epoch": 0.4765, + "grad_norm": 1.1250323100313209, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 47650 + }, + { + "epoch": 0.47651, + "grad_norm": 1.3136889509459897, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 47651 + }, + { + "epoch": 0.47652, + "grad_norm": 1.060472289106147, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 47652 + }, + { + "epoch": 0.47653, + "grad_norm": 0.9248449560086507, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 47653 + }, + { + "epoch": 0.47654, + "grad_norm": 0.9498705998289964, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 47654 + }, + { + "epoch": 0.47655, + "grad_norm": 1.2519278044850894, + "learning_rate": 0.003, + "loss": 3.995, + "step": 47655 + }, + { + "epoch": 0.47656, + "grad_norm": 1.1022446265796597, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 47656 + }, + { + "epoch": 0.47657, + "grad_norm": 1.0252976345996794, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 47657 + }, + { + "epoch": 0.47658, + "grad_norm": 1.0043605163989784, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 47658 + }, + { + "epoch": 0.47659, + "grad_norm": 1.133926802018375, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 47659 + }, + { + "epoch": 0.4766, + "grad_norm": 1.0923709192686437, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 47660 + }, + { + "epoch": 0.47661, + "grad_norm": 1.0597145649263862, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 47661 + }, + { + "epoch": 0.47662, + "grad_norm": 1.0474937046285409, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 47662 + }, + { + "epoch": 0.47663, + "grad_norm": 0.9257995170919784, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 47663 + }, + { + "epoch": 0.47664, + "grad_norm": 1.0831000629235166, + "learning_rate": 0.003, + "loss": 3.994, + "step": 47664 + }, + { + "epoch": 0.47665, + "grad_norm": 1.0985998652843072, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 47665 + }, + { + "epoch": 0.47666, + "grad_norm": 1.1246500703802496, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 47666 + }, + { + "epoch": 0.47667, + "grad_norm": 0.8591388326923863, + "learning_rate": 0.003, + "loss": 4.061, + "step": 47667 + }, + { + "epoch": 0.47668, + "grad_norm": 0.999712395941417, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 47668 + }, + { + "epoch": 0.47669, + "grad_norm": 1.319814921458931, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 47669 + }, + { + "epoch": 0.4767, + "grad_norm": 1.0700538488425624, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 47670 + }, + { + "epoch": 0.47671, + "grad_norm": 1.1042338121491715, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 47671 + }, + { + "epoch": 0.47672, + "grad_norm": 0.9620394823013203, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 47672 + }, + { + "epoch": 0.47673, + "grad_norm": 1.0558055929482584, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 47673 + }, + { + "epoch": 0.47674, + "grad_norm": 1.1453637933492702, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 47674 + }, + { + "epoch": 0.47675, + "grad_norm": 1.1372235246220124, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 47675 + }, + { + "epoch": 0.47676, + "grad_norm": 0.9413839082869083, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 47676 + }, + { + "epoch": 0.47677, + "grad_norm": 0.9200909051003672, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 47677 + }, + { + "epoch": 0.47678, + "grad_norm": 1.0094735900714475, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 47678 + }, + { + "epoch": 0.47679, + "grad_norm": 1.0282207710025941, + "learning_rate": 0.003, + "loss": 4.017, + "step": 47679 + }, + { + "epoch": 0.4768, + "grad_norm": 1.0608496198744075, + "learning_rate": 0.003, + "loss": 4.021, + "step": 47680 + }, + { + "epoch": 0.47681, + "grad_norm": 1.0151281719175758, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 47681 + }, + { + "epoch": 0.47682, + "grad_norm": 1.0583325343581156, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 47682 + }, + { + "epoch": 0.47683, + "grad_norm": 1.1788770831462703, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 47683 + }, + { + "epoch": 0.47684, + "grad_norm": 0.9076418549910813, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 47684 + }, + { + "epoch": 0.47685, + "grad_norm": 0.9915082719363719, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 47685 + }, + { + "epoch": 0.47686, + "grad_norm": 0.9986428625592126, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 47686 + }, + { + "epoch": 0.47687, + "grad_norm": 1.0278710250771654, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 47687 + }, + { + "epoch": 0.47688, + "grad_norm": 1.1034988022343626, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 47688 + }, + { + "epoch": 0.47689, + "grad_norm": 1.2916199548481013, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 47689 + }, + { + "epoch": 0.4769, + "grad_norm": 1.2795011407423653, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 47690 + }, + { + "epoch": 0.47691, + "grad_norm": 0.934865395106171, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 47691 + }, + { + "epoch": 0.47692, + "grad_norm": 1.0318676142345293, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 47692 + }, + { + "epoch": 0.47693, + "grad_norm": 1.1846183008467772, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 47693 + }, + { + "epoch": 0.47694, + "grad_norm": 1.0889126222776324, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 47694 + }, + { + "epoch": 0.47695, + "grad_norm": 1.1040112294217226, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 47695 + }, + { + "epoch": 0.47696, + "grad_norm": 0.9699825065312657, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 47696 + }, + { + "epoch": 0.47697, + "grad_norm": 1.1061454306183993, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 47697 + }, + { + "epoch": 0.47698, + "grad_norm": 1.0824432373490374, + "learning_rate": 0.003, + "loss": 4.039, + "step": 47698 + }, + { + "epoch": 0.47699, + "grad_norm": 1.0867758041492048, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 47699 + }, + { + "epoch": 0.477, + "grad_norm": 0.9162358998184574, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 47700 + }, + { + "epoch": 0.47701, + "grad_norm": 0.9485073387375399, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 47701 + }, + { + "epoch": 0.47702, + "grad_norm": 0.9746956648156374, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 47702 + }, + { + "epoch": 0.47703, + "grad_norm": 0.938942967443415, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 47703 + }, + { + "epoch": 0.47704, + "grad_norm": 0.8880429066641821, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 47704 + }, + { + "epoch": 0.47705, + "grad_norm": 1.033485241488399, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 47705 + }, + { + "epoch": 0.47706, + "grad_norm": 1.4463138979420265, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 47706 + }, + { + "epoch": 0.47707, + "grad_norm": 1.0239228026573461, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 47707 + }, + { + "epoch": 0.47708, + "grad_norm": 0.9969830850392816, + "learning_rate": 0.003, + "loss": 4.034, + "step": 47708 + }, + { + "epoch": 0.47709, + "grad_norm": 1.3346198331324801, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 47709 + }, + { + "epoch": 0.4771, + "grad_norm": 0.8253349939335412, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 47710 + }, + { + "epoch": 0.47711, + "grad_norm": 0.8274186947363804, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 47711 + }, + { + "epoch": 0.47712, + "grad_norm": 0.93448920305871, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 47712 + }, + { + "epoch": 0.47713, + "grad_norm": 0.9313987973801708, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 47713 + }, + { + "epoch": 0.47714, + "grad_norm": 1.1418775569609558, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 47714 + }, + { + "epoch": 0.47715, + "grad_norm": 1.2890691961326743, + "learning_rate": 0.003, + "loss": 4.046, + "step": 47715 + }, + { + "epoch": 0.47716, + "grad_norm": 0.8768884891667245, + "learning_rate": 0.003, + "loss": 4.03, + "step": 47716 + }, + { + "epoch": 0.47717, + "grad_norm": 0.9043362425287939, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 47717 + }, + { + "epoch": 0.47718, + "grad_norm": 0.9391495915269057, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 47718 + }, + { + "epoch": 0.47719, + "grad_norm": 1.317885246921009, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 47719 + }, + { + "epoch": 0.4772, + "grad_norm": 1.136158339136171, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 47720 + }, + { + "epoch": 0.47721, + "grad_norm": 1.0352581654662856, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 47721 + }, + { + "epoch": 0.47722, + "grad_norm": 0.8300140628961163, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 47722 + }, + { + "epoch": 0.47723, + "grad_norm": 0.8306203944490556, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 47723 + }, + { + "epoch": 0.47724, + "grad_norm": 1.0520418090565762, + "learning_rate": 0.003, + "loss": 4.053, + "step": 47724 + }, + { + "epoch": 0.47725, + "grad_norm": 1.2498197578348085, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 47725 + }, + { + "epoch": 0.47726, + "grad_norm": 0.7946969933773622, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 47726 + }, + { + "epoch": 0.47727, + "grad_norm": 0.8917561942630303, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 47727 + }, + { + "epoch": 0.47728, + "grad_norm": 1.2150335338554803, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 47728 + }, + { + "epoch": 0.47729, + "grad_norm": 1.2246562938326153, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 47729 + }, + { + "epoch": 0.4773, + "grad_norm": 0.9725524557404529, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 47730 + }, + { + "epoch": 0.47731, + "grad_norm": 1.2528550912015797, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 47731 + }, + { + "epoch": 0.47732, + "grad_norm": 0.8503586871441607, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 47732 + }, + { + "epoch": 0.47733, + "grad_norm": 1.1606769147436296, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 47733 + }, + { + "epoch": 0.47734, + "grad_norm": 1.3059743088713869, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 47734 + }, + { + "epoch": 0.47735, + "grad_norm": 1.198602085741504, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 47735 + }, + { + "epoch": 0.47736, + "grad_norm": 1.04058237459528, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 47736 + }, + { + "epoch": 0.47737, + "grad_norm": 1.2384472238663187, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 47737 + }, + { + "epoch": 0.47738, + "grad_norm": 0.929749973754888, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 47738 + }, + { + "epoch": 0.47739, + "grad_norm": 0.9675377781109128, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 47739 + }, + { + "epoch": 0.4774, + "grad_norm": 1.0868943272835434, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 47740 + }, + { + "epoch": 0.47741, + "grad_norm": 1.1899867786648963, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 47741 + }, + { + "epoch": 0.47742, + "grad_norm": 1.2498717816184366, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 47742 + }, + { + "epoch": 0.47743, + "grad_norm": 1.1074970918720828, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 47743 + }, + { + "epoch": 0.47744, + "grad_norm": 0.9011422737027758, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 47744 + }, + { + "epoch": 0.47745, + "grad_norm": 0.8642314567578647, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 47745 + }, + { + "epoch": 0.47746, + "grad_norm": 0.9635618575595539, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 47746 + }, + { + "epoch": 0.47747, + "grad_norm": 1.10015950926401, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 47747 + }, + { + "epoch": 0.47748, + "grad_norm": 1.2234324423204574, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 47748 + }, + { + "epoch": 0.47749, + "grad_norm": 0.9660188091945532, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 47749 + }, + { + "epoch": 0.4775, + "grad_norm": 1.0058184811777593, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 47750 + }, + { + "epoch": 0.47751, + "grad_norm": 1.1926741426431013, + "learning_rate": 0.003, + "loss": 4.06, + "step": 47751 + }, + { + "epoch": 0.47752, + "grad_norm": 0.9058923244655958, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 47752 + }, + { + "epoch": 0.47753, + "grad_norm": 1.0287556599928862, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 47753 + }, + { + "epoch": 0.47754, + "grad_norm": 1.1060920529921694, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 47754 + }, + { + "epoch": 0.47755, + "grad_norm": 0.9871446447246457, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 47755 + }, + { + "epoch": 0.47756, + "grad_norm": 1.0798577919436685, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 47756 + }, + { + "epoch": 0.47757, + "grad_norm": 1.0336691311576847, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 47757 + }, + { + "epoch": 0.47758, + "grad_norm": 1.2043555682166056, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 47758 + }, + { + "epoch": 0.47759, + "grad_norm": 0.9237355479129883, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 47759 + }, + { + "epoch": 0.4776, + "grad_norm": 1.0087325384680903, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 47760 + }, + { + "epoch": 0.47761, + "grad_norm": 1.1820374517315755, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 47761 + }, + { + "epoch": 0.47762, + "grad_norm": 1.3814933432259815, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 47762 + }, + { + "epoch": 0.47763, + "grad_norm": 0.8462025421270374, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 47763 + }, + { + "epoch": 0.47764, + "grad_norm": 0.8121478494164066, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 47764 + }, + { + "epoch": 0.47765, + "grad_norm": 1.1232475455768842, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 47765 + }, + { + "epoch": 0.47766, + "grad_norm": 1.1767846044015944, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 47766 + }, + { + "epoch": 0.47767, + "grad_norm": 0.9836678425301802, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 47767 + }, + { + "epoch": 0.47768, + "grad_norm": 1.441937085472143, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 47768 + }, + { + "epoch": 0.47769, + "grad_norm": 0.7197338111276704, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 47769 + }, + { + "epoch": 0.4777, + "grad_norm": 0.6660336797860801, + "learning_rate": 0.003, + "loss": 4.017, + "step": 47770 + }, + { + "epoch": 0.47771, + "grad_norm": 0.9590969841220021, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 47771 + }, + { + "epoch": 0.47772, + "grad_norm": 1.2395233775023347, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 47772 + }, + { + "epoch": 0.47773, + "grad_norm": 1.179015280861268, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 47773 + }, + { + "epoch": 0.47774, + "grad_norm": 0.9608474915471892, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 47774 + }, + { + "epoch": 0.47775, + "grad_norm": 0.8851334292180034, + "learning_rate": 0.003, + "loss": 4.028, + "step": 47775 + }, + { + "epoch": 0.47776, + "grad_norm": 1.0214737646450869, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 47776 + }, + { + "epoch": 0.47777, + "grad_norm": 1.3056878874817601, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 47777 + }, + { + "epoch": 0.47778, + "grad_norm": 0.9653808097151135, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 47778 + }, + { + "epoch": 0.47779, + "grad_norm": 1.1061561028439095, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 47779 + }, + { + "epoch": 0.4778, + "grad_norm": 0.9421770457168077, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 47780 + }, + { + "epoch": 0.47781, + "grad_norm": 1.3853461788337325, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 47781 + }, + { + "epoch": 0.47782, + "grad_norm": 0.9595703658514297, + "learning_rate": 0.003, + "loss": 4.032, + "step": 47782 + }, + { + "epoch": 0.47783, + "grad_norm": 1.1357822175674603, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 47783 + }, + { + "epoch": 0.47784, + "grad_norm": 1.2011904323782576, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 47784 + }, + { + "epoch": 0.47785, + "grad_norm": 1.0050746557095593, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 47785 + }, + { + "epoch": 0.47786, + "grad_norm": 1.062586487758172, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 47786 + }, + { + "epoch": 0.47787, + "grad_norm": 1.1072340921034984, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 47787 + }, + { + "epoch": 0.47788, + "grad_norm": 1.1166295612221822, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 47788 + }, + { + "epoch": 0.47789, + "grad_norm": 1.20269765797181, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 47789 + }, + { + "epoch": 0.4779, + "grad_norm": 1.0353100855214472, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 47790 + }, + { + "epoch": 0.47791, + "grad_norm": 1.296629021760754, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 47791 + }, + { + "epoch": 0.47792, + "grad_norm": 1.019479552312936, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 47792 + }, + { + "epoch": 0.47793, + "grad_norm": 1.352184518610802, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 47793 + }, + { + "epoch": 0.47794, + "grad_norm": 0.8597436242681162, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 47794 + }, + { + "epoch": 0.47795, + "grad_norm": 0.8595213758249477, + "learning_rate": 0.003, + "loss": 4.026, + "step": 47795 + }, + { + "epoch": 0.47796, + "grad_norm": 1.131077087057644, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 47796 + }, + { + "epoch": 0.47797, + "grad_norm": 1.1120801008802186, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 47797 + }, + { + "epoch": 0.47798, + "grad_norm": 1.114045443770663, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 47798 + }, + { + "epoch": 0.47799, + "grad_norm": 1.322351694174341, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 47799 + }, + { + "epoch": 0.478, + "grad_norm": 0.952420352240436, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 47800 + }, + { + "epoch": 0.47801, + "grad_norm": 1.0323051169397475, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 47801 + }, + { + "epoch": 0.47802, + "grad_norm": 1.0959879040107825, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 47802 + }, + { + "epoch": 0.47803, + "grad_norm": 0.8217564894107834, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 47803 + }, + { + "epoch": 0.47804, + "grad_norm": 0.935498948834184, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 47804 + }, + { + "epoch": 0.47805, + "grad_norm": 1.0515468473782266, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 47805 + }, + { + "epoch": 0.47806, + "grad_norm": 1.2040645936121854, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 47806 + }, + { + "epoch": 0.47807, + "grad_norm": 1.024016159627921, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 47807 + }, + { + "epoch": 0.47808, + "grad_norm": 1.1804963877719983, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 47808 + }, + { + "epoch": 0.47809, + "grad_norm": 1.0336892151362105, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 47809 + }, + { + "epoch": 0.4781, + "grad_norm": 1.2891119723839044, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 47810 + }, + { + "epoch": 0.47811, + "grad_norm": 0.8547285739896923, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 47811 + }, + { + "epoch": 0.47812, + "grad_norm": 0.8845106449943704, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 47812 + }, + { + "epoch": 0.47813, + "grad_norm": 1.096136468126035, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 47813 + }, + { + "epoch": 0.47814, + "grad_norm": 1.3698527081717589, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 47814 + }, + { + "epoch": 0.47815, + "grad_norm": 0.8691752294176591, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 47815 + }, + { + "epoch": 0.47816, + "grad_norm": 0.9555972702183364, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 47816 + }, + { + "epoch": 0.47817, + "grad_norm": 1.0345931814001619, + "learning_rate": 0.003, + "loss": 4.034, + "step": 47817 + }, + { + "epoch": 0.47818, + "grad_norm": 1.205121617146146, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 47818 + }, + { + "epoch": 0.47819, + "grad_norm": 1.329502723805605, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 47819 + }, + { + "epoch": 0.4782, + "grad_norm": 0.7315962014440113, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 47820 + }, + { + "epoch": 0.47821, + "grad_norm": 0.7897422966880777, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 47821 + }, + { + "epoch": 0.47822, + "grad_norm": 1.1276072113539457, + "learning_rate": 0.003, + "loss": 4.025, + "step": 47822 + }, + { + "epoch": 0.47823, + "grad_norm": 1.1047159996929317, + "learning_rate": 0.003, + "loss": 4.064, + "step": 47823 + }, + { + "epoch": 0.47824, + "grad_norm": 1.1431031102246163, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 47824 + }, + { + "epoch": 0.47825, + "grad_norm": 0.9660571558542965, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 47825 + }, + { + "epoch": 0.47826, + "grad_norm": 1.148198620502916, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 47826 + }, + { + "epoch": 0.47827, + "grad_norm": 0.9754298864362959, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 47827 + }, + { + "epoch": 0.47828, + "grad_norm": 0.9461833923993113, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 47828 + }, + { + "epoch": 0.47829, + "grad_norm": 1.1970267048349303, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 47829 + }, + { + "epoch": 0.4783, + "grad_norm": 1.0788390210125514, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 47830 + }, + { + "epoch": 0.47831, + "grad_norm": 1.005785768164241, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 47831 + }, + { + "epoch": 0.47832, + "grad_norm": 1.2280977152718715, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 47832 + }, + { + "epoch": 0.47833, + "grad_norm": 0.9811555011533714, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 47833 + }, + { + "epoch": 0.47834, + "grad_norm": 1.2424263513875553, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 47834 + }, + { + "epoch": 0.47835, + "grad_norm": 0.9368594686508566, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 47835 + }, + { + "epoch": 0.47836, + "grad_norm": 0.9409320788539908, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 47836 + }, + { + "epoch": 0.47837, + "grad_norm": 1.0770945533180867, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 47837 + }, + { + "epoch": 0.47838, + "grad_norm": 1.2187844442842528, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 47838 + }, + { + "epoch": 0.47839, + "grad_norm": 1.0469511607502178, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 47839 + }, + { + "epoch": 0.4784, + "grad_norm": 0.9934244358105395, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 47840 + }, + { + "epoch": 0.47841, + "grad_norm": 1.0969139439589253, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 47841 + }, + { + "epoch": 0.47842, + "grad_norm": 1.4330517595808185, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 47842 + }, + { + "epoch": 0.47843, + "grad_norm": 0.9997031857603181, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 47843 + }, + { + "epoch": 0.47844, + "grad_norm": 0.9746264889506555, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 47844 + }, + { + "epoch": 0.47845, + "grad_norm": 1.4251649623198963, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 47845 + }, + { + "epoch": 0.47846, + "grad_norm": 1.080306489523157, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 47846 + }, + { + "epoch": 0.47847, + "grad_norm": 1.0408539187669148, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 47847 + }, + { + "epoch": 0.47848, + "grad_norm": 1.0718323113628496, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 47848 + }, + { + "epoch": 0.47849, + "grad_norm": 1.0778501631958182, + "learning_rate": 0.003, + "loss": 4.027, + "step": 47849 + }, + { + "epoch": 0.4785, + "grad_norm": 1.1287414807187544, + "learning_rate": 0.003, + "loss": 4.016, + "step": 47850 + }, + { + "epoch": 0.47851, + "grad_norm": 1.0042109006806992, + "learning_rate": 0.003, + "loss": 4.021, + "step": 47851 + }, + { + "epoch": 0.47852, + "grad_norm": 0.9709697118089249, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 47852 + }, + { + "epoch": 0.47853, + "grad_norm": 0.9374186796785637, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 47853 + }, + { + "epoch": 0.47854, + "grad_norm": 0.9213064110358554, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 47854 + }, + { + "epoch": 0.47855, + "grad_norm": 0.9195550112183003, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 47855 + }, + { + "epoch": 0.47856, + "grad_norm": 0.9766576754706097, + "learning_rate": 0.003, + "loss": 4.001, + "step": 47856 + }, + { + "epoch": 0.47857, + "grad_norm": 1.1938733533284642, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 47857 + }, + { + "epoch": 0.47858, + "grad_norm": 1.060956221489585, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 47858 + }, + { + "epoch": 0.47859, + "grad_norm": 1.1139135655443355, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 47859 + }, + { + "epoch": 0.4786, + "grad_norm": 1.1090406340782, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 47860 + }, + { + "epoch": 0.47861, + "grad_norm": 1.1708048750845557, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 47861 + }, + { + "epoch": 0.47862, + "grad_norm": 0.9840390419112909, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 47862 + }, + { + "epoch": 0.47863, + "grad_norm": 1.1102309623026156, + "learning_rate": 0.003, + "loss": 4.026, + "step": 47863 + }, + { + "epoch": 0.47864, + "grad_norm": 1.016773575886989, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 47864 + }, + { + "epoch": 0.47865, + "grad_norm": 1.1410374134896801, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 47865 + }, + { + "epoch": 0.47866, + "grad_norm": 1.0875830364163013, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 47866 + }, + { + "epoch": 0.47867, + "grad_norm": 1.1474475363803913, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 47867 + }, + { + "epoch": 0.47868, + "grad_norm": 0.9949673424135796, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 47868 + }, + { + "epoch": 0.47869, + "grad_norm": 1.175623160112082, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 47869 + }, + { + "epoch": 0.4787, + "grad_norm": 1.2773160408455422, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 47870 + }, + { + "epoch": 0.47871, + "grad_norm": 0.952319476632654, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 47871 + }, + { + "epoch": 0.47872, + "grad_norm": 1.1363143470580375, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 47872 + }, + { + "epoch": 0.47873, + "grad_norm": 1.1498787484073452, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 47873 + }, + { + "epoch": 0.47874, + "grad_norm": 1.116457237680033, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 47874 + }, + { + "epoch": 0.47875, + "grad_norm": 1.0844709489320612, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 47875 + }, + { + "epoch": 0.47876, + "grad_norm": 1.1775061153874968, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 47876 + }, + { + "epoch": 0.47877, + "grad_norm": 0.95033409470767, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 47877 + }, + { + "epoch": 0.47878, + "grad_norm": 0.9734116489498763, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 47878 + }, + { + "epoch": 0.47879, + "grad_norm": 1.2388945323493126, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 47879 + }, + { + "epoch": 0.4788, + "grad_norm": 0.9584207657948438, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 47880 + }, + { + "epoch": 0.47881, + "grad_norm": 0.9366439422556018, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 47881 + }, + { + "epoch": 0.47882, + "grad_norm": 1.0051461687009458, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 47882 + }, + { + "epoch": 0.47883, + "grad_norm": 1.2829224811383917, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 47883 + }, + { + "epoch": 0.47884, + "grad_norm": 1.0008369732122606, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 47884 + }, + { + "epoch": 0.47885, + "grad_norm": 1.2238181290749341, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 47885 + }, + { + "epoch": 0.47886, + "grad_norm": 0.8105263839407082, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 47886 + }, + { + "epoch": 0.47887, + "grad_norm": 0.8095686840224446, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 47887 + }, + { + "epoch": 0.47888, + "grad_norm": 0.9212972951900613, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 47888 + }, + { + "epoch": 0.47889, + "grad_norm": 1.1224397503446268, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 47889 + }, + { + "epoch": 0.4789, + "grad_norm": 1.2188994831365898, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 47890 + }, + { + "epoch": 0.47891, + "grad_norm": 1.0267499478915487, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 47891 + }, + { + "epoch": 0.47892, + "grad_norm": 0.9905114559757213, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 47892 + }, + { + "epoch": 0.47893, + "grad_norm": 1.1901153945020335, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 47893 + }, + { + "epoch": 0.47894, + "grad_norm": 0.9726456871294493, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 47894 + }, + { + "epoch": 0.47895, + "grad_norm": 0.8252033797740767, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 47895 + }, + { + "epoch": 0.47896, + "grad_norm": 0.6911662015991156, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 47896 + }, + { + "epoch": 0.47897, + "grad_norm": 0.8829133610637129, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 47897 + }, + { + "epoch": 0.47898, + "grad_norm": 1.2533718424362987, + "learning_rate": 0.003, + "loss": 4.025, + "step": 47898 + }, + { + "epoch": 0.47899, + "grad_norm": 0.9761954536128781, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 47899 + }, + { + "epoch": 0.479, + "grad_norm": 1.0990406576327723, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 47900 + }, + { + "epoch": 0.47901, + "grad_norm": 1.0237305848521439, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 47901 + }, + { + "epoch": 0.47902, + "grad_norm": 1.2323076319971054, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 47902 + }, + { + "epoch": 0.47903, + "grad_norm": 0.8872651907770097, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 47903 + }, + { + "epoch": 0.47904, + "grad_norm": 0.8862022879506793, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 47904 + }, + { + "epoch": 0.47905, + "grad_norm": 1.2689680671421781, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 47905 + }, + { + "epoch": 0.47906, + "grad_norm": 1.1159085812806755, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 47906 + }, + { + "epoch": 0.47907, + "grad_norm": 0.9723011526331676, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 47907 + }, + { + "epoch": 0.47908, + "grad_norm": 0.9737980502198481, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 47908 + }, + { + "epoch": 0.47909, + "grad_norm": 1.1272983023494343, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 47909 + }, + { + "epoch": 0.4791, + "grad_norm": 0.9989048907013961, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 47910 + }, + { + "epoch": 0.47911, + "grad_norm": 1.1786870428629899, + "learning_rate": 0.003, + "loss": 4.017, + "step": 47911 + }, + { + "epoch": 0.47912, + "grad_norm": 1.000632339489031, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 47912 + }, + { + "epoch": 0.47913, + "grad_norm": 1.0587528691114603, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 47913 + }, + { + "epoch": 0.47914, + "grad_norm": 0.9832842652813513, + "learning_rate": 0.003, + "loss": 4.025, + "step": 47914 + }, + { + "epoch": 0.47915, + "grad_norm": 1.2236267820689473, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 47915 + }, + { + "epoch": 0.47916, + "grad_norm": 1.0422313198743458, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 47916 + }, + { + "epoch": 0.47917, + "grad_norm": 1.0190855836327137, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 47917 + }, + { + "epoch": 0.47918, + "grad_norm": 1.056402919159169, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 47918 + }, + { + "epoch": 0.47919, + "grad_norm": 1.118649547237801, + "learning_rate": 0.003, + "loss": 4.003, + "step": 47919 + }, + { + "epoch": 0.4792, + "grad_norm": 0.9219425213170895, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 47920 + }, + { + "epoch": 0.47921, + "grad_norm": 1.0154722691127616, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 47921 + }, + { + "epoch": 0.47922, + "grad_norm": 1.2035674555858227, + "learning_rate": 0.003, + "loss": 4.047, + "step": 47922 + }, + { + "epoch": 0.47923, + "grad_norm": 1.0962364185709432, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 47923 + }, + { + "epoch": 0.47924, + "grad_norm": 0.9232185198136309, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 47924 + }, + { + "epoch": 0.47925, + "grad_norm": 1.0601686500592493, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 47925 + }, + { + "epoch": 0.47926, + "grad_norm": 1.3584213071539424, + "learning_rate": 0.003, + "loss": 4.047, + "step": 47926 + }, + { + "epoch": 0.47927, + "grad_norm": 0.7874068272640211, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 47927 + }, + { + "epoch": 0.47928, + "grad_norm": 0.9294128180255297, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 47928 + }, + { + "epoch": 0.47929, + "grad_norm": 1.0960025531006212, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 47929 + }, + { + "epoch": 0.4793, + "grad_norm": 1.210384000264869, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 47930 + }, + { + "epoch": 0.47931, + "grad_norm": 1.0904004549689863, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 47931 + }, + { + "epoch": 0.47932, + "grad_norm": 1.033207720829563, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 47932 + }, + { + "epoch": 0.47933, + "grad_norm": 1.2654829798940284, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 47933 + }, + { + "epoch": 0.47934, + "grad_norm": 0.7672792883110016, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 47934 + }, + { + "epoch": 0.47935, + "grad_norm": 0.7487867452431597, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 47935 + }, + { + "epoch": 0.47936, + "grad_norm": 0.9642364062659071, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 47936 + }, + { + "epoch": 0.47937, + "grad_norm": 1.3438035962667376, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 47937 + }, + { + "epoch": 0.47938, + "grad_norm": 1.100909214375497, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 47938 + }, + { + "epoch": 0.47939, + "grad_norm": 0.9400072491028754, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 47939 + }, + { + "epoch": 0.4794, + "grad_norm": 0.7773427284339592, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 47940 + }, + { + "epoch": 0.47941, + "grad_norm": 0.8591042001726595, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 47941 + }, + { + "epoch": 0.47942, + "grad_norm": 1.0652139090507484, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 47942 + }, + { + "epoch": 0.47943, + "grad_norm": 1.0025737043964598, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 47943 + }, + { + "epoch": 0.47944, + "grad_norm": 1.0172678259452794, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 47944 + }, + { + "epoch": 0.47945, + "grad_norm": 0.9596258840602212, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 47945 + }, + { + "epoch": 0.47946, + "grad_norm": 1.2826008763490526, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 47946 + }, + { + "epoch": 0.47947, + "grad_norm": 0.9182568976666882, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 47947 + }, + { + "epoch": 0.47948, + "grad_norm": 1.1526650809178143, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 47948 + }, + { + "epoch": 0.47949, + "grad_norm": 1.1176955719846853, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 47949 + }, + { + "epoch": 0.4795, + "grad_norm": 1.135097857948623, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 47950 + }, + { + "epoch": 0.47951, + "grad_norm": 1.302562399899229, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 47951 + }, + { + "epoch": 0.47952, + "grad_norm": 0.9154462445540814, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 47952 + }, + { + "epoch": 0.47953, + "grad_norm": 1.1032496334204014, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 47953 + }, + { + "epoch": 0.47954, + "grad_norm": 1.3781170700803815, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 47954 + }, + { + "epoch": 0.47955, + "grad_norm": 0.9520085804339171, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 47955 + }, + { + "epoch": 0.47956, + "grad_norm": 0.974138306525612, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 47956 + }, + { + "epoch": 0.47957, + "grad_norm": 1.147866941168071, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 47957 + }, + { + "epoch": 0.47958, + "grad_norm": 1.0958923664636722, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 47958 + }, + { + "epoch": 0.47959, + "grad_norm": 0.991788076641618, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 47959 + }, + { + "epoch": 0.4796, + "grad_norm": 1.0251753594586503, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 47960 + }, + { + "epoch": 0.47961, + "grad_norm": 1.109695666218445, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 47961 + }, + { + "epoch": 0.47962, + "grad_norm": 0.9922621276587378, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 47962 + }, + { + "epoch": 0.47963, + "grad_norm": 1.0188410107094819, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 47963 + }, + { + "epoch": 0.47964, + "grad_norm": 1.0199242928335561, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 47964 + }, + { + "epoch": 0.47965, + "grad_norm": 1.0878985768658138, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 47965 + }, + { + "epoch": 0.47966, + "grad_norm": 0.937165390744106, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 47966 + }, + { + "epoch": 0.47967, + "grad_norm": 1.0412675712442148, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 47967 + }, + { + "epoch": 0.47968, + "grad_norm": 1.2189365703033757, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 47968 + }, + { + "epoch": 0.47969, + "grad_norm": 0.9946313763858122, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 47969 + }, + { + "epoch": 0.4797, + "grad_norm": 1.1353809719481638, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 47970 + }, + { + "epoch": 0.47971, + "grad_norm": 0.8487065120900085, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 47971 + }, + { + "epoch": 0.47972, + "grad_norm": 0.833594259447485, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 47972 + }, + { + "epoch": 0.47973, + "grad_norm": 1.0073292307262953, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 47973 + }, + { + "epoch": 0.47974, + "grad_norm": 1.1095196757073627, + "learning_rate": 0.003, + "loss": 4.064, + "step": 47974 + }, + { + "epoch": 0.47975, + "grad_norm": 1.1032404396391102, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 47975 + }, + { + "epoch": 0.47976, + "grad_norm": 1.2339547635428263, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 47976 + }, + { + "epoch": 0.47977, + "grad_norm": 0.9602658162409902, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 47977 + }, + { + "epoch": 0.47978, + "grad_norm": 1.1247134404715253, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 47978 + }, + { + "epoch": 0.47979, + "grad_norm": 1.2526558064765088, + "learning_rate": 0.003, + "loss": 4.048, + "step": 47979 + }, + { + "epoch": 0.4798, + "grad_norm": 1.3185045454130253, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 47980 + }, + { + "epoch": 0.47981, + "grad_norm": 0.9104055713711411, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 47981 + }, + { + "epoch": 0.47982, + "grad_norm": 0.8907697091080247, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 47982 + }, + { + "epoch": 0.47983, + "grad_norm": 1.1243628332569298, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 47983 + }, + { + "epoch": 0.47984, + "grad_norm": 1.3486953783237574, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 47984 + }, + { + "epoch": 0.47985, + "grad_norm": 0.8314334064058512, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 47985 + }, + { + "epoch": 0.47986, + "grad_norm": 0.8380587101459378, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 47986 + }, + { + "epoch": 0.47987, + "grad_norm": 1.0105203500819626, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 47987 + }, + { + "epoch": 0.47988, + "grad_norm": 1.128851054233218, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 47988 + }, + { + "epoch": 0.47989, + "grad_norm": 1.0103480732820598, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 47989 + }, + { + "epoch": 0.4799, + "grad_norm": 1.0483112933794956, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 47990 + }, + { + "epoch": 0.47991, + "grad_norm": 0.9656892574415045, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 47991 + }, + { + "epoch": 0.47992, + "grad_norm": 1.132701749656565, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 47992 + }, + { + "epoch": 0.47993, + "grad_norm": 0.9653925570467108, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 47993 + }, + { + "epoch": 0.47994, + "grad_norm": 1.1539443702274002, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 47994 + }, + { + "epoch": 0.47995, + "grad_norm": 1.0417071027233396, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 47995 + }, + { + "epoch": 0.47996, + "grad_norm": 1.1134763207100886, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 47996 + }, + { + "epoch": 0.47997, + "grad_norm": 1.2600238620416808, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 47997 + }, + { + "epoch": 0.47998, + "grad_norm": 1.0415032790239058, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 47998 + }, + { + "epoch": 0.47999, + "grad_norm": 0.9070657834798633, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 47999 + }, + { + "epoch": 0.48, + "grad_norm": 1.0115350313362617, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 48000 + }, + { + "epoch": 0.48001, + "grad_norm": 1.4613577052019804, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 48001 + }, + { + "epoch": 0.48002, + "grad_norm": 0.8234779761608875, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 48002 + }, + { + "epoch": 0.48003, + "grad_norm": 0.9021464215473844, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 48003 + }, + { + "epoch": 0.48004, + "grad_norm": 0.7683357193341759, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 48004 + }, + { + "epoch": 0.48005, + "grad_norm": 0.8437810462803363, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 48005 + }, + { + "epoch": 0.48006, + "grad_norm": 1.0357142482118065, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 48006 + }, + { + "epoch": 0.48007, + "grad_norm": 1.2665044512389374, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 48007 + }, + { + "epoch": 0.48008, + "grad_norm": 0.9239826230438294, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 48008 + }, + { + "epoch": 0.48009, + "grad_norm": 1.2514027077050895, + "learning_rate": 0.003, + "loss": 4.053, + "step": 48009 + }, + { + "epoch": 0.4801, + "grad_norm": 1.112491189434892, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 48010 + }, + { + "epoch": 0.48011, + "grad_norm": 0.7503509969099602, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 48011 + }, + { + "epoch": 0.48012, + "grad_norm": 0.733098544793545, + "learning_rate": 0.003, + "loss": 4.041, + "step": 48012 + }, + { + "epoch": 0.48013, + "grad_norm": 0.8532105313842906, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 48013 + }, + { + "epoch": 0.48014, + "grad_norm": 1.082016590225533, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 48014 + }, + { + "epoch": 0.48015, + "grad_norm": 1.1178721333831094, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 48015 + }, + { + "epoch": 0.48016, + "grad_norm": 1.0682851800153028, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 48016 + }, + { + "epoch": 0.48017, + "grad_norm": 1.217216200943203, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 48017 + }, + { + "epoch": 0.48018, + "grad_norm": 0.9606011256251199, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 48018 + }, + { + "epoch": 0.48019, + "grad_norm": 1.0100629196828907, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 48019 + }, + { + "epoch": 0.4802, + "grad_norm": 1.0669768388220138, + "learning_rate": 0.003, + "loss": 4.025, + "step": 48020 + }, + { + "epoch": 0.48021, + "grad_norm": 0.9586049154205194, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 48021 + }, + { + "epoch": 0.48022, + "grad_norm": 1.0206871302310145, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 48022 + }, + { + "epoch": 0.48023, + "grad_norm": 1.0416212686862116, + "learning_rate": 0.003, + "loss": 4.04, + "step": 48023 + }, + { + "epoch": 0.48024, + "grad_norm": 1.0790649074212562, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 48024 + }, + { + "epoch": 0.48025, + "grad_norm": 0.9357747705000106, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 48025 + }, + { + "epoch": 0.48026, + "grad_norm": 1.2179572834150882, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 48026 + }, + { + "epoch": 0.48027, + "grad_norm": 1.0516267138451234, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 48027 + }, + { + "epoch": 0.48028, + "grad_norm": 1.6354552119060641, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 48028 + }, + { + "epoch": 0.48029, + "grad_norm": 0.8118484090617865, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 48029 + }, + { + "epoch": 0.4803, + "grad_norm": 1.0196159915953287, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 48030 + }, + { + "epoch": 0.48031, + "grad_norm": 1.3051735163920257, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 48031 + }, + { + "epoch": 0.48032, + "grad_norm": 1.2088603681927694, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 48032 + }, + { + "epoch": 0.48033, + "grad_norm": 1.0429402084502717, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 48033 + }, + { + "epoch": 0.48034, + "grad_norm": 0.9402517279903391, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 48034 + }, + { + "epoch": 0.48035, + "grad_norm": 1.0634583209511927, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 48035 + }, + { + "epoch": 0.48036, + "grad_norm": 1.0998519236410547, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 48036 + }, + { + "epoch": 0.48037, + "grad_norm": 1.0733124282392437, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 48037 + }, + { + "epoch": 0.48038, + "grad_norm": 1.1326928808558316, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 48038 + }, + { + "epoch": 0.48039, + "grad_norm": 1.2465663048808415, + "learning_rate": 0.003, + "loss": 4.023, + "step": 48039 + }, + { + "epoch": 0.4804, + "grad_norm": 0.9993475838432794, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 48040 + }, + { + "epoch": 0.48041, + "grad_norm": 1.124503346604199, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 48041 + }, + { + "epoch": 0.48042, + "grad_norm": 1.5407826831106954, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 48042 + }, + { + "epoch": 0.48043, + "grad_norm": 0.7776549814671965, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 48043 + }, + { + "epoch": 0.48044, + "grad_norm": 0.7796598452779168, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 48044 + }, + { + "epoch": 0.48045, + "grad_norm": 0.9003795160436967, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 48045 + }, + { + "epoch": 0.48046, + "grad_norm": 1.1189062362530144, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 48046 + }, + { + "epoch": 0.48047, + "grad_norm": 1.1342660357760712, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 48047 + }, + { + "epoch": 0.48048, + "grad_norm": 1.0610546578876938, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 48048 + }, + { + "epoch": 0.48049, + "grad_norm": 1.2297145499781768, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 48049 + }, + { + "epoch": 0.4805, + "grad_norm": 0.9085073718817485, + "learning_rate": 0.003, + "loss": 4.046, + "step": 48050 + }, + { + "epoch": 0.48051, + "grad_norm": 1.025372966316617, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 48051 + }, + { + "epoch": 0.48052, + "grad_norm": 1.2080169402958896, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 48052 + }, + { + "epoch": 0.48053, + "grad_norm": 1.1238940930604298, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 48053 + }, + { + "epoch": 0.48054, + "grad_norm": 0.8940580310044928, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 48054 + }, + { + "epoch": 0.48055, + "grad_norm": 0.9026290096470521, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 48055 + }, + { + "epoch": 0.48056, + "grad_norm": 1.0004911519003032, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 48056 + }, + { + "epoch": 0.48057, + "grad_norm": 1.0796970441764377, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 48057 + }, + { + "epoch": 0.48058, + "grad_norm": 1.0198150898639404, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 48058 + }, + { + "epoch": 0.48059, + "grad_norm": 0.9317203427601165, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 48059 + }, + { + "epoch": 0.4806, + "grad_norm": 1.0247862547654472, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 48060 + }, + { + "epoch": 0.48061, + "grad_norm": 1.2702843975626392, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 48061 + }, + { + "epoch": 0.48062, + "grad_norm": 1.3228527267158157, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 48062 + }, + { + "epoch": 0.48063, + "grad_norm": 0.8715155685421243, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 48063 + }, + { + "epoch": 0.48064, + "grad_norm": 0.9820752755486076, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 48064 + }, + { + "epoch": 0.48065, + "grad_norm": 1.382937203194477, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 48065 + }, + { + "epoch": 0.48066, + "grad_norm": 1.2598590434254142, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 48066 + }, + { + "epoch": 0.48067, + "grad_norm": 0.89604306302723, + "learning_rate": 0.003, + "loss": 4.042, + "step": 48067 + }, + { + "epoch": 0.48068, + "grad_norm": 1.0315976664788644, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 48068 + }, + { + "epoch": 0.48069, + "grad_norm": 1.1239160159478352, + "learning_rate": 0.003, + "loss": 4.052, + "step": 48069 + }, + { + "epoch": 0.4807, + "grad_norm": 1.010376603818838, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 48070 + }, + { + "epoch": 0.48071, + "grad_norm": 1.035470696602303, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 48071 + }, + { + "epoch": 0.48072, + "grad_norm": 1.0610708765032715, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 48072 + }, + { + "epoch": 0.48073, + "grad_norm": 1.09026483624074, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 48073 + }, + { + "epoch": 0.48074, + "grad_norm": 1.069715135752776, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 48074 + }, + { + "epoch": 0.48075, + "grad_norm": 0.9771319601889948, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 48075 + }, + { + "epoch": 0.48076, + "grad_norm": 1.2607377080961528, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 48076 + }, + { + "epoch": 0.48077, + "grad_norm": 1.1122783362918054, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 48077 + }, + { + "epoch": 0.48078, + "grad_norm": 1.0659466074719146, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 48078 + }, + { + "epoch": 0.48079, + "grad_norm": 0.9803126268088387, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 48079 + }, + { + "epoch": 0.4808, + "grad_norm": 0.9410613571504947, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 48080 + }, + { + "epoch": 0.48081, + "grad_norm": 0.9977752836977695, + "learning_rate": 0.003, + "loss": 4.022, + "step": 48081 + }, + { + "epoch": 0.48082, + "grad_norm": 1.1414666257406694, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 48082 + }, + { + "epoch": 0.48083, + "grad_norm": 0.7991579825275018, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 48083 + }, + { + "epoch": 0.48084, + "grad_norm": 0.8234808769811452, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 48084 + }, + { + "epoch": 0.48085, + "grad_norm": 0.9822127811019871, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 48085 + }, + { + "epoch": 0.48086, + "grad_norm": 1.1286431322744972, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 48086 + }, + { + "epoch": 0.48087, + "grad_norm": 0.8776247317242473, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 48087 + }, + { + "epoch": 0.48088, + "grad_norm": 1.0397056658911163, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 48088 + }, + { + "epoch": 0.48089, + "grad_norm": 1.1646599287383161, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 48089 + }, + { + "epoch": 0.4809, + "grad_norm": 0.8190473987784893, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 48090 + }, + { + "epoch": 0.48091, + "grad_norm": 1.0651764713505816, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 48091 + }, + { + "epoch": 0.48092, + "grad_norm": 1.3005986661168176, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 48092 + }, + { + "epoch": 0.48093, + "grad_norm": 0.9019729490219416, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 48093 + }, + { + "epoch": 0.48094, + "grad_norm": 0.948149661727636, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 48094 + }, + { + "epoch": 0.48095, + "grad_norm": 1.1210330371679467, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 48095 + }, + { + "epoch": 0.48096, + "grad_norm": 1.2011910653842468, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 48096 + }, + { + "epoch": 0.48097, + "grad_norm": 1.0334624842048081, + "learning_rate": 0.003, + "loss": 3.995, + "step": 48097 + }, + { + "epoch": 0.48098, + "grad_norm": 1.083483274230073, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 48098 + }, + { + "epoch": 0.48099, + "grad_norm": 0.9203074895230198, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 48099 + }, + { + "epoch": 0.481, + "grad_norm": 0.9584588510995097, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 48100 + }, + { + "epoch": 0.48101, + "grad_norm": 1.1943293702906181, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 48101 + }, + { + "epoch": 0.48102, + "grad_norm": 0.8890238190687436, + "learning_rate": 0.003, + "loss": 4.035, + "step": 48102 + }, + { + "epoch": 0.48103, + "grad_norm": 0.8377802565356359, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 48103 + }, + { + "epoch": 0.48104, + "grad_norm": 0.986758066189774, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 48104 + }, + { + "epoch": 0.48105, + "grad_norm": 1.2437909378128562, + "learning_rate": 0.003, + "loss": 4.058, + "step": 48105 + }, + { + "epoch": 0.48106, + "grad_norm": 1.0472109627665327, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 48106 + }, + { + "epoch": 0.48107, + "grad_norm": 1.0318667843583853, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 48107 + }, + { + "epoch": 0.48108, + "grad_norm": 1.0729596261890293, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 48108 + }, + { + "epoch": 0.48109, + "grad_norm": 1.0597575368959355, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 48109 + }, + { + "epoch": 0.4811, + "grad_norm": 0.9424013351093122, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 48110 + }, + { + "epoch": 0.48111, + "grad_norm": 1.0472046929135868, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 48111 + }, + { + "epoch": 0.48112, + "grad_norm": 1.2177298133683867, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 48112 + }, + { + "epoch": 0.48113, + "grad_norm": 1.1011117797977794, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 48113 + }, + { + "epoch": 0.48114, + "grad_norm": 1.1244786683348704, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 48114 + }, + { + "epoch": 0.48115, + "grad_norm": 1.0465515821639024, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 48115 + }, + { + "epoch": 0.48116, + "grad_norm": 1.298601520823733, + "learning_rate": 0.003, + "loss": 4.062, + "step": 48116 + }, + { + "epoch": 0.48117, + "grad_norm": 1.012839214169806, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 48117 + }, + { + "epoch": 0.48118, + "grad_norm": 1.0961589265663636, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 48118 + }, + { + "epoch": 0.48119, + "grad_norm": 1.0577455318545963, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 48119 + }, + { + "epoch": 0.4812, + "grad_norm": 1.2429816284642499, + "learning_rate": 0.003, + "loss": 4.052, + "step": 48120 + }, + { + "epoch": 0.48121, + "grad_norm": 1.1001062132785508, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 48121 + }, + { + "epoch": 0.48122, + "grad_norm": 0.9097958436500221, + "learning_rate": 0.003, + "loss": 4.034, + "step": 48122 + }, + { + "epoch": 0.48123, + "grad_norm": 0.8767616274735556, + "learning_rate": 0.003, + "loss": 3.992, + "step": 48123 + }, + { + "epoch": 0.48124, + "grad_norm": 0.963251234366295, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 48124 + }, + { + "epoch": 0.48125, + "grad_norm": 1.3506012368442577, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 48125 + }, + { + "epoch": 0.48126, + "grad_norm": 0.8181445832115652, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 48126 + }, + { + "epoch": 0.48127, + "grad_norm": 0.870672197483675, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 48127 + }, + { + "epoch": 0.48128, + "grad_norm": 1.1742183557722286, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 48128 + }, + { + "epoch": 0.48129, + "grad_norm": 1.055517643669951, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 48129 + }, + { + "epoch": 0.4813, + "grad_norm": 1.085900081395679, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 48130 + }, + { + "epoch": 0.48131, + "grad_norm": 0.9737572285374739, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 48131 + }, + { + "epoch": 0.48132, + "grad_norm": 1.0831708237102557, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 48132 + }, + { + "epoch": 0.48133, + "grad_norm": 1.2011743664631114, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 48133 + }, + { + "epoch": 0.48134, + "grad_norm": 0.9626988700592471, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 48134 + }, + { + "epoch": 0.48135, + "grad_norm": 0.9377200100180954, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 48135 + }, + { + "epoch": 0.48136, + "grad_norm": 0.9922820759844329, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 48136 + }, + { + "epoch": 0.48137, + "grad_norm": 1.1201291637803035, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 48137 + }, + { + "epoch": 0.48138, + "grad_norm": 1.2411065946327742, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 48138 + }, + { + "epoch": 0.48139, + "grad_norm": 1.0814123835344482, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 48139 + }, + { + "epoch": 0.4814, + "grad_norm": 0.9984853494957081, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 48140 + }, + { + "epoch": 0.48141, + "grad_norm": 1.2129038992578676, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 48141 + }, + { + "epoch": 0.48142, + "grad_norm": 1.1117749762996787, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 48142 + }, + { + "epoch": 0.48143, + "grad_norm": 1.0170206455271051, + "learning_rate": 0.003, + "loss": 4.024, + "step": 48143 + }, + { + "epoch": 0.48144, + "grad_norm": 1.0998252394683279, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 48144 + }, + { + "epoch": 0.48145, + "grad_norm": 1.1935419039303812, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 48145 + }, + { + "epoch": 0.48146, + "grad_norm": 1.089314751369574, + "learning_rate": 0.003, + "loss": 4.02, + "step": 48146 + }, + { + "epoch": 0.48147, + "grad_norm": 0.9813824783695236, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 48147 + }, + { + "epoch": 0.48148, + "grad_norm": 1.0786591525061529, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 48148 + }, + { + "epoch": 0.48149, + "grad_norm": 1.084220823361747, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 48149 + }, + { + "epoch": 0.4815, + "grad_norm": 1.0204383482707193, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 48150 + }, + { + "epoch": 0.48151, + "grad_norm": 1.089237272682469, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 48151 + }, + { + "epoch": 0.48152, + "grad_norm": 0.9643775354940005, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 48152 + }, + { + "epoch": 0.48153, + "grad_norm": 1.108164891462851, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 48153 + }, + { + "epoch": 0.48154, + "grad_norm": 0.9857317284441535, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 48154 + }, + { + "epoch": 0.48155, + "grad_norm": 1.259043109153202, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 48155 + }, + { + "epoch": 0.48156, + "grad_norm": 1.038951859924789, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 48156 + }, + { + "epoch": 0.48157, + "grad_norm": 1.368933981926878, + "learning_rate": 0.003, + "loss": 4.058, + "step": 48157 + }, + { + "epoch": 0.48158, + "grad_norm": 1.019542999019362, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 48158 + }, + { + "epoch": 0.48159, + "grad_norm": 1.2971879337167354, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 48159 + }, + { + "epoch": 0.4816, + "grad_norm": 1.1249277714382417, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 48160 + }, + { + "epoch": 0.48161, + "grad_norm": 1.074788606381812, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 48161 + }, + { + "epoch": 0.48162, + "grad_norm": 1.180972687676037, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 48162 + }, + { + "epoch": 0.48163, + "grad_norm": 1.0746804321571417, + "learning_rate": 0.003, + "loss": 4.033, + "step": 48163 + }, + { + "epoch": 0.48164, + "grad_norm": 0.8872186233486878, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 48164 + }, + { + "epoch": 0.48165, + "grad_norm": 1.0996248458510303, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 48165 + }, + { + "epoch": 0.48166, + "grad_norm": 1.0575164098054541, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 48166 + }, + { + "epoch": 0.48167, + "grad_norm": 1.096474020645899, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 48167 + }, + { + "epoch": 0.48168, + "grad_norm": 0.9355360520710418, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 48168 + }, + { + "epoch": 0.48169, + "grad_norm": 1.1075438521609238, + "learning_rate": 0.003, + "loss": 4.061, + "step": 48169 + }, + { + "epoch": 0.4817, + "grad_norm": 1.2052771716013657, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 48170 + }, + { + "epoch": 0.48171, + "grad_norm": 1.0745616290841842, + "learning_rate": 0.003, + "loss": 4.066, + "step": 48171 + }, + { + "epoch": 0.48172, + "grad_norm": 1.1303696092014501, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 48172 + }, + { + "epoch": 0.48173, + "grad_norm": 1.0496806528890865, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 48173 + }, + { + "epoch": 0.48174, + "grad_norm": 1.142668575933369, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 48174 + }, + { + "epoch": 0.48175, + "grad_norm": 1.0265029858829058, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 48175 + }, + { + "epoch": 0.48176, + "grad_norm": 1.0871412576489197, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 48176 + }, + { + "epoch": 0.48177, + "grad_norm": 0.9231697594788433, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 48177 + }, + { + "epoch": 0.48178, + "grad_norm": 0.9023750304817516, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 48178 + }, + { + "epoch": 0.48179, + "grad_norm": 1.0422114956715962, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 48179 + }, + { + "epoch": 0.4818, + "grad_norm": 0.9913558277153162, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 48180 + }, + { + "epoch": 0.48181, + "grad_norm": 1.3515874916604593, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 48181 + }, + { + "epoch": 0.48182, + "grad_norm": 0.9315072952977629, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 48182 + }, + { + "epoch": 0.48183, + "grad_norm": 1.1235329303706236, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 48183 + }, + { + "epoch": 0.48184, + "grad_norm": 1.1971316074266978, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 48184 + }, + { + "epoch": 0.48185, + "grad_norm": 1.089616617416901, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 48185 + }, + { + "epoch": 0.48186, + "grad_norm": 1.0031941810761527, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 48186 + }, + { + "epoch": 0.48187, + "grad_norm": 1.0424399018500303, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 48187 + }, + { + "epoch": 0.48188, + "grad_norm": 1.113283031401793, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 48188 + }, + { + "epoch": 0.48189, + "grad_norm": 1.2501089257351348, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 48189 + }, + { + "epoch": 0.4819, + "grad_norm": 0.6790954740267193, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 48190 + }, + { + "epoch": 0.48191, + "grad_norm": 0.652723545519816, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 48191 + }, + { + "epoch": 0.48192, + "grad_norm": 0.8796774462584083, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 48192 + }, + { + "epoch": 0.48193, + "grad_norm": 1.097979090002702, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 48193 + }, + { + "epoch": 0.48194, + "grad_norm": 1.0390111589564455, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 48194 + }, + { + "epoch": 0.48195, + "grad_norm": 1.0322956660240938, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 48195 + }, + { + "epoch": 0.48196, + "grad_norm": 1.2081648293265566, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 48196 + }, + { + "epoch": 0.48197, + "grad_norm": 1.0842121552463346, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 48197 + }, + { + "epoch": 0.48198, + "grad_norm": 0.9563327156764225, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 48198 + }, + { + "epoch": 0.48199, + "grad_norm": 1.1912984740199912, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 48199 + }, + { + "epoch": 0.482, + "grad_norm": 1.024137714192127, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 48200 + }, + { + "epoch": 0.48201, + "grad_norm": 1.1877896951375866, + "learning_rate": 0.003, + "loss": 4.013, + "step": 48201 + }, + { + "epoch": 0.48202, + "grad_norm": 1.0651335950096132, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 48202 + }, + { + "epoch": 0.48203, + "grad_norm": 1.1584721695031843, + "learning_rate": 0.003, + "loss": 4.043, + "step": 48203 + }, + { + "epoch": 0.48204, + "grad_norm": 1.278506022486796, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 48204 + }, + { + "epoch": 0.48205, + "grad_norm": 1.1786359446257317, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 48205 + }, + { + "epoch": 0.48206, + "grad_norm": 1.0465503561595024, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 48206 + }, + { + "epoch": 0.48207, + "grad_norm": 1.170614642058885, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 48207 + }, + { + "epoch": 0.48208, + "grad_norm": 1.0142816666265804, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 48208 + }, + { + "epoch": 0.48209, + "grad_norm": 0.8948434590113814, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 48209 + }, + { + "epoch": 0.4821, + "grad_norm": 0.9762124581436916, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 48210 + }, + { + "epoch": 0.48211, + "grad_norm": 1.1160037942497214, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 48211 + }, + { + "epoch": 0.48212, + "grad_norm": 0.9240599944842436, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 48212 + }, + { + "epoch": 0.48213, + "grad_norm": 1.0350624598975486, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 48213 + }, + { + "epoch": 0.48214, + "grad_norm": 1.1443491075958472, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 48214 + }, + { + "epoch": 0.48215, + "grad_norm": 1.0625103423216715, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 48215 + }, + { + "epoch": 0.48216, + "grad_norm": 1.1906143652930767, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 48216 + }, + { + "epoch": 0.48217, + "grad_norm": 1.0243754276541326, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 48217 + }, + { + "epoch": 0.48218, + "grad_norm": 1.2214130957878298, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 48218 + }, + { + "epoch": 0.48219, + "grad_norm": 1.012333810876671, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 48219 + }, + { + "epoch": 0.4822, + "grad_norm": 1.0515460590346892, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 48220 + }, + { + "epoch": 0.48221, + "grad_norm": 0.9643831301112029, + "learning_rate": 0.003, + "loss": 4.003, + "step": 48221 + }, + { + "epoch": 0.48222, + "grad_norm": 0.9691796596175875, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 48222 + }, + { + "epoch": 0.48223, + "grad_norm": 1.099954065456874, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 48223 + }, + { + "epoch": 0.48224, + "grad_norm": 1.0204895168398278, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 48224 + }, + { + "epoch": 0.48225, + "grad_norm": 1.1964523709087587, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 48225 + }, + { + "epoch": 0.48226, + "grad_norm": 1.0628036331144328, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 48226 + }, + { + "epoch": 0.48227, + "grad_norm": 1.0690133915072597, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 48227 + }, + { + "epoch": 0.48228, + "grad_norm": 0.8309334545683851, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 48228 + }, + { + "epoch": 0.48229, + "grad_norm": 0.8811816461227122, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 48229 + }, + { + "epoch": 0.4823, + "grad_norm": 0.9956232710645787, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 48230 + }, + { + "epoch": 0.48231, + "grad_norm": 1.0664123811651425, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 48231 + }, + { + "epoch": 0.48232, + "grad_norm": 1.0377024413561757, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 48232 + }, + { + "epoch": 0.48233, + "grad_norm": 1.165928177968674, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 48233 + }, + { + "epoch": 0.48234, + "grad_norm": 1.3377548151877325, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 48234 + }, + { + "epoch": 0.48235, + "grad_norm": 0.7820667882852774, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 48235 + }, + { + "epoch": 0.48236, + "grad_norm": 0.7657767766272148, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 48236 + }, + { + "epoch": 0.48237, + "grad_norm": 0.915510170205994, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 48237 + }, + { + "epoch": 0.48238, + "grad_norm": 1.2156943215895666, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 48238 + }, + { + "epoch": 0.48239, + "grad_norm": 0.9425513275616775, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 48239 + }, + { + "epoch": 0.4824, + "grad_norm": 1.1563016840873335, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 48240 + }, + { + "epoch": 0.48241, + "grad_norm": 1.228584111636029, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 48241 + }, + { + "epoch": 0.48242, + "grad_norm": 0.953260906519015, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 48242 + }, + { + "epoch": 0.48243, + "grad_norm": 1.1290791779452372, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 48243 + }, + { + "epoch": 0.48244, + "grad_norm": 1.106361835755777, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 48244 + }, + { + "epoch": 0.48245, + "grad_norm": 1.0503391420684722, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 48245 + }, + { + "epoch": 0.48246, + "grad_norm": 1.0392889295770056, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 48246 + }, + { + "epoch": 0.48247, + "grad_norm": 1.0872291604521094, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 48247 + }, + { + "epoch": 0.48248, + "grad_norm": 0.9501802362616726, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 48248 + }, + { + "epoch": 0.48249, + "grad_norm": 0.9111690892516409, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 48249 + }, + { + "epoch": 0.4825, + "grad_norm": 1.006415059371983, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 48250 + }, + { + "epoch": 0.48251, + "grad_norm": 1.2196849520940845, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 48251 + }, + { + "epoch": 0.48252, + "grad_norm": 0.9024965742925021, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 48252 + }, + { + "epoch": 0.48253, + "grad_norm": 1.0479955387023705, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 48253 + }, + { + "epoch": 0.48254, + "grad_norm": 1.1272765657803971, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 48254 + }, + { + "epoch": 0.48255, + "grad_norm": 0.9102393690521984, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 48255 + }, + { + "epoch": 0.48256, + "grad_norm": 0.9111224412128849, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 48256 + }, + { + "epoch": 0.48257, + "grad_norm": 0.8878340396285854, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 48257 + }, + { + "epoch": 0.48258, + "grad_norm": 0.9328411943067622, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 48258 + }, + { + "epoch": 0.48259, + "grad_norm": 0.9848877103732538, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 48259 + }, + { + "epoch": 0.4826, + "grad_norm": 1.125194594395715, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 48260 + }, + { + "epoch": 0.48261, + "grad_norm": 1.0724565802442845, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 48261 + }, + { + "epoch": 0.48262, + "grad_norm": 1.0804931951122236, + "learning_rate": 0.003, + "loss": 4.015, + "step": 48262 + }, + { + "epoch": 0.48263, + "grad_norm": 0.9884936842633164, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 48263 + }, + { + "epoch": 0.48264, + "grad_norm": 1.0092919090673063, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 48264 + }, + { + "epoch": 0.48265, + "grad_norm": 1.020740419799132, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 48265 + }, + { + "epoch": 0.48266, + "grad_norm": 1.3855653089213287, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 48266 + }, + { + "epoch": 0.48267, + "grad_norm": 1.3135392143654978, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 48267 + }, + { + "epoch": 0.48268, + "grad_norm": 0.8908951765509086, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 48268 + }, + { + "epoch": 0.48269, + "grad_norm": 0.8122149757576765, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 48269 + }, + { + "epoch": 0.4827, + "grad_norm": 0.9379961723717687, + "learning_rate": 0.003, + "loss": 3.998, + "step": 48270 + }, + { + "epoch": 0.48271, + "grad_norm": 1.0398827046111714, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 48271 + }, + { + "epoch": 0.48272, + "grad_norm": 1.3538162333010038, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 48272 + }, + { + "epoch": 0.48273, + "grad_norm": 0.962427430318889, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 48273 + }, + { + "epoch": 0.48274, + "grad_norm": 1.0559248552190748, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 48274 + }, + { + "epoch": 0.48275, + "grad_norm": 1.1691332178928489, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 48275 + }, + { + "epoch": 0.48276, + "grad_norm": 0.8955722936812599, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 48276 + }, + { + "epoch": 0.48277, + "grad_norm": 0.7666510238923625, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 48277 + }, + { + "epoch": 0.48278, + "grad_norm": 0.8374021258767866, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 48278 + }, + { + "epoch": 0.48279, + "grad_norm": 1.0882244327472965, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 48279 + }, + { + "epoch": 0.4828, + "grad_norm": 1.0698793760174412, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 48280 + }, + { + "epoch": 0.48281, + "grad_norm": 1.150475837336831, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 48281 + }, + { + "epoch": 0.48282, + "grad_norm": 1.0866770994537582, + "learning_rate": 0.003, + "loss": 4.038, + "step": 48282 + }, + { + "epoch": 0.48283, + "grad_norm": 1.1162567980102118, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 48283 + }, + { + "epoch": 0.48284, + "grad_norm": 1.098103968701357, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 48284 + }, + { + "epoch": 0.48285, + "grad_norm": 1.2688863702007167, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 48285 + }, + { + "epoch": 0.48286, + "grad_norm": 1.0520777793587366, + "learning_rate": 0.003, + "loss": 4.047, + "step": 48286 + }, + { + "epoch": 0.48287, + "grad_norm": 1.2644597186708975, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 48287 + }, + { + "epoch": 0.48288, + "grad_norm": 1.0664946704742349, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 48288 + }, + { + "epoch": 0.48289, + "grad_norm": 1.1783488481472395, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 48289 + }, + { + "epoch": 0.4829, + "grad_norm": 0.968566529034745, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 48290 + }, + { + "epoch": 0.48291, + "grad_norm": 1.1288941693562908, + "learning_rate": 0.003, + "loss": 4.026, + "step": 48291 + }, + { + "epoch": 0.48292, + "grad_norm": 1.1237863872033043, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 48292 + }, + { + "epoch": 0.48293, + "grad_norm": 0.9765344473715281, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 48293 + }, + { + "epoch": 0.48294, + "grad_norm": 1.2380144241361128, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 48294 + }, + { + "epoch": 0.48295, + "grad_norm": 0.9513033315648506, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 48295 + }, + { + "epoch": 0.48296, + "grad_norm": 1.0826555802482394, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 48296 + }, + { + "epoch": 0.48297, + "grad_norm": 1.041460859847744, + "learning_rate": 0.003, + "loss": 4.034, + "step": 48297 + }, + { + "epoch": 0.48298, + "grad_norm": 1.1505270546064323, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 48298 + }, + { + "epoch": 0.48299, + "grad_norm": 0.8381972227150466, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 48299 + }, + { + "epoch": 0.483, + "grad_norm": 0.9132300056462501, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 48300 + }, + { + "epoch": 0.48301, + "grad_norm": 1.101169797879547, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 48301 + }, + { + "epoch": 0.48302, + "grad_norm": 1.108475427942103, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 48302 + }, + { + "epoch": 0.48303, + "grad_norm": 1.4389124071789738, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 48303 + }, + { + "epoch": 0.48304, + "grad_norm": 0.9807852454307531, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 48304 + }, + { + "epoch": 0.48305, + "grad_norm": 1.0988439631818638, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 48305 + }, + { + "epoch": 0.48306, + "grad_norm": 0.9149626065801607, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 48306 + }, + { + "epoch": 0.48307, + "grad_norm": 1.0533677446721013, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 48307 + }, + { + "epoch": 0.48308, + "grad_norm": 1.1245095344421723, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 48308 + }, + { + "epoch": 0.48309, + "grad_norm": 1.05787589652894, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 48309 + }, + { + "epoch": 0.4831, + "grad_norm": 1.3359024678282512, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 48310 + }, + { + "epoch": 0.48311, + "grad_norm": 0.9103754783554693, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 48311 + }, + { + "epoch": 0.48312, + "grad_norm": 1.0351567688986483, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 48312 + }, + { + "epoch": 0.48313, + "grad_norm": 1.1375231281773484, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 48313 + }, + { + "epoch": 0.48314, + "grad_norm": 1.0230925549419834, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 48314 + }, + { + "epoch": 0.48315, + "grad_norm": 0.9894733778208598, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 48315 + }, + { + "epoch": 0.48316, + "grad_norm": 1.150100173307382, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 48316 + }, + { + "epoch": 0.48317, + "grad_norm": 1.219278252383701, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 48317 + }, + { + "epoch": 0.48318, + "grad_norm": 0.9779034114016363, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 48318 + }, + { + "epoch": 0.48319, + "grad_norm": 1.0770782661160168, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 48319 + }, + { + "epoch": 0.4832, + "grad_norm": 1.0045082716051768, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 48320 + }, + { + "epoch": 0.48321, + "grad_norm": 1.3343580710771796, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 48321 + }, + { + "epoch": 0.48322, + "grad_norm": 0.9403310050051494, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 48322 + }, + { + "epoch": 0.48323, + "grad_norm": 0.8685271264231557, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 48323 + }, + { + "epoch": 0.48324, + "grad_norm": 0.8067663544716973, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 48324 + }, + { + "epoch": 0.48325, + "grad_norm": 0.9274776412310454, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 48325 + }, + { + "epoch": 0.48326, + "grad_norm": 1.1342759969797844, + "learning_rate": 0.003, + "loss": 4.058, + "step": 48326 + }, + { + "epoch": 0.48327, + "grad_norm": 1.0409901137604645, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 48327 + }, + { + "epoch": 0.48328, + "grad_norm": 1.1044168527215477, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 48328 + }, + { + "epoch": 0.48329, + "grad_norm": 0.8753762404159225, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 48329 + }, + { + "epoch": 0.4833, + "grad_norm": 0.99091929096044, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 48330 + }, + { + "epoch": 0.48331, + "grad_norm": 1.22291570309036, + "learning_rate": 0.003, + "loss": 4.019, + "step": 48331 + }, + { + "epoch": 0.48332, + "grad_norm": 1.1651051606215796, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 48332 + }, + { + "epoch": 0.48333, + "grad_norm": 1.0461496163433601, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 48333 + }, + { + "epoch": 0.48334, + "grad_norm": 0.9529403246586754, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 48334 + }, + { + "epoch": 0.48335, + "grad_norm": 0.9904738496370075, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 48335 + }, + { + "epoch": 0.48336, + "grad_norm": 1.0949972200206912, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 48336 + }, + { + "epoch": 0.48337, + "grad_norm": 0.9374607445945512, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 48337 + }, + { + "epoch": 0.48338, + "grad_norm": 1.1854296968052236, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 48338 + }, + { + "epoch": 0.48339, + "grad_norm": 0.9765673999558919, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 48339 + }, + { + "epoch": 0.4834, + "grad_norm": 1.0248878911502541, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 48340 + }, + { + "epoch": 0.48341, + "grad_norm": 1.120391492857599, + "learning_rate": 0.003, + "loss": 4.041, + "step": 48341 + }, + { + "epoch": 0.48342, + "grad_norm": 1.0483188861239405, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 48342 + }, + { + "epoch": 0.48343, + "grad_norm": 1.0207741250471227, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 48343 + }, + { + "epoch": 0.48344, + "grad_norm": 1.0697145725219377, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 48344 + }, + { + "epoch": 0.48345, + "grad_norm": 1.3229689586407596, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 48345 + }, + { + "epoch": 0.48346, + "grad_norm": 1.2082858810113137, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 48346 + }, + { + "epoch": 0.48347, + "grad_norm": 0.9378907876144795, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 48347 + }, + { + "epoch": 0.48348, + "grad_norm": 1.0583881037737053, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 48348 + }, + { + "epoch": 0.48349, + "grad_norm": 1.2889727350971365, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 48349 + }, + { + "epoch": 0.4835, + "grad_norm": 0.7716547262800835, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 48350 + }, + { + "epoch": 0.48351, + "grad_norm": 0.9678684344735762, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 48351 + }, + { + "epoch": 0.48352, + "grad_norm": 1.1540841397734778, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 48352 + }, + { + "epoch": 0.48353, + "grad_norm": 0.9717499724974052, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 48353 + }, + { + "epoch": 0.48354, + "grad_norm": 0.9633353850035938, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 48354 + }, + { + "epoch": 0.48355, + "grad_norm": 1.1658392164304585, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 48355 + }, + { + "epoch": 0.48356, + "grad_norm": 1.2236262123961994, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 48356 + }, + { + "epoch": 0.48357, + "grad_norm": 1.0415877085161382, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 48357 + }, + { + "epoch": 0.48358, + "grad_norm": 1.1633386935180337, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 48358 + }, + { + "epoch": 0.48359, + "grad_norm": 1.0526876517519186, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 48359 + }, + { + "epoch": 0.4836, + "grad_norm": 1.1600134386970282, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 48360 + }, + { + "epoch": 0.48361, + "grad_norm": 1.0696708653148488, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 48361 + }, + { + "epoch": 0.48362, + "grad_norm": 1.083674037115344, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 48362 + }, + { + "epoch": 0.48363, + "grad_norm": 0.9756687486321197, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 48363 + }, + { + "epoch": 0.48364, + "grad_norm": 1.131389699123463, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 48364 + }, + { + "epoch": 0.48365, + "grad_norm": 1.1396141227863124, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 48365 + }, + { + "epoch": 0.48366, + "grad_norm": 0.8851398467423565, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 48366 + }, + { + "epoch": 0.48367, + "grad_norm": 1.0293861226550793, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 48367 + }, + { + "epoch": 0.48368, + "grad_norm": 0.9743894343925619, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 48368 + }, + { + "epoch": 0.48369, + "grad_norm": 0.9097539885837804, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 48369 + }, + { + "epoch": 0.4837, + "grad_norm": 1.039491872579355, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 48370 + }, + { + "epoch": 0.48371, + "grad_norm": 1.1701560767376935, + "learning_rate": 0.003, + "loss": 4.029, + "step": 48371 + }, + { + "epoch": 0.48372, + "grad_norm": 0.9922021002255509, + "learning_rate": 0.003, + "loss": 4.046, + "step": 48372 + }, + { + "epoch": 0.48373, + "grad_norm": 1.1203861460165605, + "learning_rate": 0.003, + "loss": 4.026, + "step": 48373 + }, + { + "epoch": 0.48374, + "grad_norm": 1.0427290135888765, + "learning_rate": 0.003, + "loss": 4.035, + "step": 48374 + }, + { + "epoch": 0.48375, + "grad_norm": 1.12393739458992, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 48375 + }, + { + "epoch": 0.48376, + "grad_norm": 1.3012336727095513, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 48376 + }, + { + "epoch": 0.48377, + "grad_norm": 0.8507784098793848, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 48377 + }, + { + "epoch": 0.48378, + "grad_norm": 0.8652093851268721, + "learning_rate": 0.003, + "loss": 4.042, + "step": 48378 + }, + { + "epoch": 0.48379, + "grad_norm": 0.9844633056653785, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 48379 + }, + { + "epoch": 0.4838, + "grad_norm": 1.3121618039605234, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 48380 + }, + { + "epoch": 0.48381, + "grad_norm": 0.8181596514028362, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 48381 + }, + { + "epoch": 0.48382, + "grad_norm": 1.05422707074545, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 48382 + }, + { + "epoch": 0.48383, + "grad_norm": 1.1462026038197792, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 48383 + }, + { + "epoch": 0.48384, + "grad_norm": 0.9047174383311113, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 48384 + }, + { + "epoch": 0.48385, + "grad_norm": 1.0147221142233929, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 48385 + }, + { + "epoch": 0.48386, + "grad_norm": 1.0731660075905027, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 48386 + }, + { + "epoch": 0.48387, + "grad_norm": 1.3505497246781637, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 48387 + }, + { + "epoch": 0.48388, + "grad_norm": 0.8644448182045537, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 48388 + }, + { + "epoch": 0.48389, + "grad_norm": 0.9120097513651043, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 48389 + }, + { + "epoch": 0.4839, + "grad_norm": 0.9918011935550587, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 48390 + }, + { + "epoch": 0.48391, + "grad_norm": 1.1351834346969976, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 48391 + }, + { + "epoch": 0.48392, + "grad_norm": 0.7966127586739642, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 48392 + }, + { + "epoch": 0.48393, + "grad_norm": 0.8168687075897313, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 48393 + }, + { + "epoch": 0.48394, + "grad_norm": 1.0827411078936189, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 48394 + }, + { + "epoch": 0.48395, + "grad_norm": 1.2993154709202321, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 48395 + }, + { + "epoch": 0.48396, + "grad_norm": 1.1264784683851248, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 48396 + }, + { + "epoch": 0.48397, + "grad_norm": 1.1668196494810452, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 48397 + }, + { + "epoch": 0.48398, + "grad_norm": 1.0146517677837499, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 48398 + }, + { + "epoch": 0.48399, + "grad_norm": 1.1144020036250006, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 48399 + }, + { + "epoch": 0.484, + "grad_norm": 1.016933626242556, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 48400 + }, + { + "epoch": 0.48401, + "grad_norm": 1.3400416636934516, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 48401 + }, + { + "epoch": 0.48402, + "grad_norm": 0.8650460610387756, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 48402 + }, + { + "epoch": 0.48403, + "grad_norm": 0.8108054913334486, + "learning_rate": 0.003, + "loss": 4.011, + "step": 48403 + }, + { + "epoch": 0.48404, + "grad_norm": 0.8705689288149109, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 48404 + }, + { + "epoch": 0.48405, + "grad_norm": 0.912438512264318, + "learning_rate": 0.003, + "loss": 4.022, + "step": 48405 + }, + { + "epoch": 0.48406, + "grad_norm": 1.0400193038123429, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 48406 + }, + { + "epoch": 0.48407, + "grad_norm": 1.1210761563787575, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 48407 + }, + { + "epoch": 0.48408, + "grad_norm": 1.1194081335175001, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 48408 + }, + { + "epoch": 0.48409, + "grad_norm": 1.0953818084675653, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 48409 + }, + { + "epoch": 0.4841, + "grad_norm": 1.1224808693744917, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 48410 + }, + { + "epoch": 0.48411, + "grad_norm": 1.0111023328112172, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 48411 + }, + { + "epoch": 0.48412, + "grad_norm": 1.0788705345414524, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 48412 + }, + { + "epoch": 0.48413, + "grad_norm": 1.2509147784159556, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 48413 + }, + { + "epoch": 0.48414, + "grad_norm": 1.1022534499308152, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 48414 + }, + { + "epoch": 0.48415, + "grad_norm": 0.9158451887933631, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 48415 + }, + { + "epoch": 0.48416, + "grad_norm": 0.8496126147349956, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 48416 + }, + { + "epoch": 0.48417, + "grad_norm": 0.9264694080055142, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 48417 + }, + { + "epoch": 0.48418, + "grad_norm": 1.161804780678207, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 48418 + }, + { + "epoch": 0.48419, + "grad_norm": 1.0313951721051418, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 48419 + }, + { + "epoch": 0.4842, + "grad_norm": 0.965688481636051, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 48420 + }, + { + "epoch": 0.48421, + "grad_norm": 1.0461560740285487, + "learning_rate": 0.003, + "loss": 4.037, + "step": 48421 + }, + { + "epoch": 0.48422, + "grad_norm": 1.2600850255459366, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 48422 + }, + { + "epoch": 0.48423, + "grad_norm": 0.9346246582738095, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 48423 + }, + { + "epoch": 0.48424, + "grad_norm": 1.1680037145688777, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 48424 + }, + { + "epoch": 0.48425, + "grad_norm": 1.0676238654340642, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 48425 + }, + { + "epoch": 0.48426, + "grad_norm": 1.33675923736028, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 48426 + }, + { + "epoch": 0.48427, + "grad_norm": 1.2081376164811217, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 48427 + }, + { + "epoch": 0.48428, + "grad_norm": 0.9098850248297536, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 48428 + }, + { + "epoch": 0.48429, + "grad_norm": 0.945194700323021, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 48429 + }, + { + "epoch": 0.4843, + "grad_norm": 1.313361717540014, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 48430 + }, + { + "epoch": 0.48431, + "grad_norm": 1.1661252799076083, + "learning_rate": 0.003, + "loss": 4.032, + "step": 48431 + }, + { + "epoch": 0.48432, + "grad_norm": 0.8972575654854013, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 48432 + }, + { + "epoch": 0.48433, + "grad_norm": 0.9771876160859074, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 48433 + }, + { + "epoch": 0.48434, + "grad_norm": 1.163019953790333, + "learning_rate": 0.003, + "loss": 4.054, + "step": 48434 + }, + { + "epoch": 0.48435, + "grad_norm": 0.9572549960782131, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 48435 + }, + { + "epoch": 0.48436, + "grad_norm": 1.1293404295863347, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 48436 + }, + { + "epoch": 0.48437, + "grad_norm": 1.1122152572867143, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 48437 + }, + { + "epoch": 0.48438, + "grad_norm": 1.1187887139210089, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 48438 + }, + { + "epoch": 0.48439, + "grad_norm": 1.0344852602696908, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 48439 + }, + { + "epoch": 0.4844, + "grad_norm": 0.9977055683220791, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 48440 + }, + { + "epoch": 0.48441, + "grad_norm": 0.9513811968849244, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 48441 + }, + { + "epoch": 0.48442, + "grad_norm": 0.969604388440626, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 48442 + }, + { + "epoch": 0.48443, + "grad_norm": 1.0716946141812838, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 48443 + }, + { + "epoch": 0.48444, + "grad_norm": 1.2025150864469867, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 48444 + }, + { + "epoch": 0.48445, + "grad_norm": 1.292351763445584, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 48445 + }, + { + "epoch": 0.48446, + "grad_norm": 1.148103303422619, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 48446 + }, + { + "epoch": 0.48447, + "grad_norm": 1.0262184640897003, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 48447 + }, + { + "epoch": 0.48448, + "grad_norm": 1.064471864062191, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 48448 + }, + { + "epoch": 0.48449, + "grad_norm": 1.115973432757093, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 48449 + }, + { + "epoch": 0.4845, + "grad_norm": 0.9172223587304211, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 48450 + }, + { + "epoch": 0.48451, + "grad_norm": 1.0722739002029524, + "learning_rate": 0.003, + "loss": 4.036, + "step": 48451 + }, + { + "epoch": 0.48452, + "grad_norm": 1.329383033867218, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 48452 + }, + { + "epoch": 0.48453, + "grad_norm": 0.8782101824868197, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 48453 + }, + { + "epoch": 0.48454, + "grad_norm": 0.9260445043590287, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 48454 + }, + { + "epoch": 0.48455, + "grad_norm": 0.9888005895807578, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 48455 + }, + { + "epoch": 0.48456, + "grad_norm": 1.2292301014457703, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 48456 + }, + { + "epoch": 0.48457, + "grad_norm": 0.9500826416771759, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 48457 + }, + { + "epoch": 0.48458, + "grad_norm": 1.0279082958603631, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 48458 + }, + { + "epoch": 0.48459, + "grad_norm": 1.1584036421588755, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 48459 + }, + { + "epoch": 0.4846, + "grad_norm": 1.0695880299425164, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 48460 + }, + { + "epoch": 0.48461, + "grad_norm": 0.9812315180891542, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 48461 + }, + { + "epoch": 0.48462, + "grad_norm": 0.9922133613252644, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 48462 + }, + { + "epoch": 0.48463, + "grad_norm": 1.0576796271651767, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 48463 + }, + { + "epoch": 0.48464, + "grad_norm": 1.0704406486982176, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 48464 + }, + { + "epoch": 0.48465, + "grad_norm": 1.0736890978821723, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 48465 + }, + { + "epoch": 0.48466, + "grad_norm": 1.0511496501416622, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 48466 + }, + { + "epoch": 0.48467, + "grad_norm": 1.0894949291139016, + "learning_rate": 0.003, + "loss": 4.056, + "step": 48467 + }, + { + "epoch": 0.48468, + "grad_norm": 0.9036558367936038, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 48468 + }, + { + "epoch": 0.48469, + "grad_norm": 1.005988384154109, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 48469 + }, + { + "epoch": 0.4847, + "grad_norm": 1.0204992169307667, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 48470 + }, + { + "epoch": 0.48471, + "grad_norm": 1.0462978464428305, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 48471 + }, + { + "epoch": 0.48472, + "grad_norm": 0.9370705699759342, + "learning_rate": 0.003, + "loss": 4.037, + "step": 48472 + }, + { + "epoch": 0.48473, + "grad_norm": 1.3367646712846326, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 48473 + }, + { + "epoch": 0.48474, + "grad_norm": 1.3004796379027077, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 48474 + }, + { + "epoch": 0.48475, + "grad_norm": 1.0243400689006983, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 48475 + }, + { + "epoch": 0.48476, + "grad_norm": 0.9921184529240418, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 48476 + }, + { + "epoch": 0.48477, + "grad_norm": 0.9492478598361599, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 48477 + }, + { + "epoch": 0.48478, + "grad_norm": 1.0746580819235376, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 48478 + }, + { + "epoch": 0.48479, + "grad_norm": 1.1422404718761519, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 48479 + }, + { + "epoch": 0.4848, + "grad_norm": 1.1965438658340302, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 48480 + }, + { + "epoch": 0.48481, + "grad_norm": 1.1356053365538914, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 48481 + }, + { + "epoch": 0.48482, + "grad_norm": 1.0701416630646308, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 48482 + }, + { + "epoch": 0.48483, + "grad_norm": 1.043862883645345, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 48483 + }, + { + "epoch": 0.48484, + "grad_norm": 1.066193312088051, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 48484 + }, + { + "epoch": 0.48485, + "grad_norm": 1.0220953057138438, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 48485 + }, + { + "epoch": 0.48486, + "grad_norm": 1.1145562034801235, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 48486 + }, + { + "epoch": 0.48487, + "grad_norm": 1.0234218877606056, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 48487 + }, + { + "epoch": 0.48488, + "grad_norm": 0.9577883258087627, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 48488 + }, + { + "epoch": 0.48489, + "grad_norm": 0.9212851564005078, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 48489 + }, + { + "epoch": 0.4849, + "grad_norm": 1.236888650129782, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 48490 + }, + { + "epoch": 0.48491, + "grad_norm": 1.1676947161960063, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 48491 + }, + { + "epoch": 0.48492, + "grad_norm": 0.8830223107629166, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 48492 + }, + { + "epoch": 0.48493, + "grad_norm": 0.8767403903495882, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 48493 + }, + { + "epoch": 0.48494, + "grad_norm": 1.1644271686231018, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 48494 + }, + { + "epoch": 0.48495, + "grad_norm": 1.2892324153895813, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 48495 + }, + { + "epoch": 0.48496, + "grad_norm": 0.8033025582193895, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 48496 + }, + { + "epoch": 0.48497, + "grad_norm": 0.8997618947408793, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 48497 + }, + { + "epoch": 0.48498, + "grad_norm": 1.0933264964517635, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 48498 + }, + { + "epoch": 0.48499, + "grad_norm": 1.2802640583460096, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 48499 + }, + { + "epoch": 0.485, + "grad_norm": 0.8163092225766941, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 48500 + }, + { + "epoch": 0.48501, + "grad_norm": 0.7302435898372522, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 48501 + }, + { + "epoch": 0.48502, + "grad_norm": 0.7747472017000874, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 48502 + }, + { + "epoch": 0.48503, + "grad_norm": 1.0571407750891708, + "learning_rate": 0.003, + "loss": 4.001, + "step": 48503 + }, + { + "epoch": 0.48504, + "grad_norm": 1.3929814663536608, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 48504 + }, + { + "epoch": 0.48505, + "grad_norm": 0.8477442744830077, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 48505 + }, + { + "epoch": 0.48506, + "grad_norm": 0.9351090472492498, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 48506 + }, + { + "epoch": 0.48507, + "grad_norm": 1.0585409202584204, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 48507 + }, + { + "epoch": 0.48508, + "grad_norm": 1.2104134643358546, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 48508 + }, + { + "epoch": 0.48509, + "grad_norm": 0.9438056080040598, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 48509 + }, + { + "epoch": 0.4851, + "grad_norm": 0.8658368408375597, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 48510 + }, + { + "epoch": 0.48511, + "grad_norm": 1.003913116932371, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 48511 + }, + { + "epoch": 0.48512, + "grad_norm": 1.3520566074140319, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 48512 + }, + { + "epoch": 0.48513, + "grad_norm": 0.9467493525552511, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 48513 + }, + { + "epoch": 0.48514, + "grad_norm": 1.4025436529929314, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 48514 + }, + { + "epoch": 0.48515, + "grad_norm": 1.026499459354996, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 48515 + }, + { + "epoch": 0.48516, + "grad_norm": 0.8416792596084919, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 48516 + }, + { + "epoch": 0.48517, + "grad_norm": 0.9623817146513546, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 48517 + }, + { + "epoch": 0.48518, + "grad_norm": 0.9805328772412211, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 48518 + }, + { + "epoch": 0.48519, + "grad_norm": 1.26355289327147, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 48519 + }, + { + "epoch": 0.4852, + "grad_norm": 0.9302835407204311, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 48520 + }, + { + "epoch": 0.48521, + "grad_norm": 0.9950500503361985, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 48521 + }, + { + "epoch": 0.48522, + "grad_norm": 1.4062673095518468, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 48522 + }, + { + "epoch": 0.48523, + "grad_norm": 0.9159575308491094, + "learning_rate": 0.003, + "loss": 4.031, + "step": 48523 + }, + { + "epoch": 0.48524, + "grad_norm": 0.9854247340170105, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 48524 + }, + { + "epoch": 0.48525, + "grad_norm": 1.1815795183648194, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 48525 + }, + { + "epoch": 0.48526, + "grad_norm": 0.9856132072958677, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 48526 + }, + { + "epoch": 0.48527, + "grad_norm": 1.0676062478833876, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 48527 + }, + { + "epoch": 0.48528, + "grad_norm": 1.0011257358015402, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 48528 + }, + { + "epoch": 0.48529, + "grad_norm": 1.1376751276699046, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 48529 + }, + { + "epoch": 0.4853, + "grad_norm": 1.0449953937368226, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 48530 + }, + { + "epoch": 0.48531, + "grad_norm": 1.189588327414242, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 48531 + }, + { + "epoch": 0.48532, + "grad_norm": 0.9156090345670063, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 48532 + }, + { + "epoch": 0.48533, + "grad_norm": 0.9541274296924541, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 48533 + }, + { + "epoch": 0.48534, + "grad_norm": 1.082065570923224, + "learning_rate": 0.003, + "loss": 4.068, + "step": 48534 + }, + { + "epoch": 0.48535, + "grad_norm": 0.9491518037432507, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 48535 + }, + { + "epoch": 0.48536, + "grad_norm": 0.9476643549420088, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 48536 + }, + { + "epoch": 0.48537, + "grad_norm": 1.2364938887765953, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 48537 + }, + { + "epoch": 0.48538, + "grad_norm": 0.9755443604373968, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 48538 + }, + { + "epoch": 0.48539, + "grad_norm": 1.0927025445047205, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 48539 + }, + { + "epoch": 0.4854, + "grad_norm": 1.2577261556059731, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 48540 + }, + { + "epoch": 0.48541, + "grad_norm": 1.168971678649769, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 48541 + }, + { + "epoch": 0.48542, + "grad_norm": 0.8904692340557727, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 48542 + }, + { + "epoch": 0.48543, + "grad_norm": 0.9059207434989744, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 48543 + }, + { + "epoch": 0.48544, + "grad_norm": 1.0609611283211493, + "learning_rate": 0.003, + "loss": 4.063, + "step": 48544 + }, + { + "epoch": 0.48545, + "grad_norm": 1.0017066070521254, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 48545 + }, + { + "epoch": 0.48546, + "grad_norm": 1.0829751804169752, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 48546 + }, + { + "epoch": 0.48547, + "grad_norm": 0.9521717902177007, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 48547 + }, + { + "epoch": 0.48548, + "grad_norm": 1.0577943879869067, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 48548 + }, + { + "epoch": 0.48549, + "grad_norm": 1.2511179952216906, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 48549 + }, + { + "epoch": 0.4855, + "grad_norm": 0.8243555817879111, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 48550 + }, + { + "epoch": 0.48551, + "grad_norm": 0.914885759362607, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 48551 + }, + { + "epoch": 0.48552, + "grad_norm": 1.2775426553671352, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 48552 + }, + { + "epoch": 0.48553, + "grad_norm": 1.1807379715762665, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 48553 + }, + { + "epoch": 0.48554, + "grad_norm": 1.1244621256791305, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 48554 + }, + { + "epoch": 0.48555, + "grad_norm": 1.159380937050873, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 48555 + }, + { + "epoch": 0.48556, + "grad_norm": 0.9520106265120579, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 48556 + }, + { + "epoch": 0.48557, + "grad_norm": 0.987011600999268, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 48557 + }, + { + "epoch": 0.48558, + "grad_norm": 1.297067794989544, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 48558 + }, + { + "epoch": 0.48559, + "grad_norm": 1.171569477044244, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 48559 + }, + { + "epoch": 0.4856, + "grad_norm": 1.1890791438833892, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 48560 + }, + { + "epoch": 0.48561, + "grad_norm": 0.9633799778729242, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 48561 + }, + { + "epoch": 0.48562, + "grad_norm": 1.1464662372030703, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 48562 + }, + { + "epoch": 0.48563, + "grad_norm": 1.1417794561713694, + "learning_rate": 0.003, + "loss": 4.034, + "step": 48563 + }, + { + "epoch": 0.48564, + "grad_norm": 1.0171997506199077, + "learning_rate": 0.003, + "loss": 4.042, + "step": 48564 + }, + { + "epoch": 0.48565, + "grad_norm": 0.8891224570819553, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 48565 + }, + { + "epoch": 0.48566, + "grad_norm": 0.991115745641837, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 48566 + }, + { + "epoch": 0.48567, + "grad_norm": 1.2488216379685986, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 48567 + }, + { + "epoch": 0.48568, + "grad_norm": 0.8924447393058688, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 48568 + }, + { + "epoch": 0.48569, + "grad_norm": 0.9320823408417617, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 48569 + }, + { + "epoch": 0.4857, + "grad_norm": 1.06684544134441, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 48570 + }, + { + "epoch": 0.48571, + "grad_norm": 1.2489029628784312, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 48571 + }, + { + "epoch": 0.48572, + "grad_norm": 0.8355411940888456, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 48572 + }, + { + "epoch": 0.48573, + "grad_norm": 0.9072860910966406, + "learning_rate": 0.003, + "loss": 4.046, + "step": 48573 + }, + { + "epoch": 0.48574, + "grad_norm": 1.189898344202288, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 48574 + }, + { + "epoch": 0.48575, + "grad_norm": 0.984711925524381, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 48575 + }, + { + "epoch": 0.48576, + "grad_norm": 1.0951887043413435, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 48576 + }, + { + "epoch": 0.48577, + "grad_norm": 0.9864441916032878, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 48577 + }, + { + "epoch": 0.48578, + "grad_norm": 0.9788061765651133, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 48578 + }, + { + "epoch": 0.48579, + "grad_norm": 0.9071527772639305, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 48579 + }, + { + "epoch": 0.4858, + "grad_norm": 1.1030065291251876, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 48580 + }, + { + "epoch": 0.48581, + "grad_norm": 1.1165880724404873, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 48581 + }, + { + "epoch": 0.48582, + "grad_norm": 1.0848558611009833, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 48582 + }, + { + "epoch": 0.48583, + "grad_norm": 1.0784171102347089, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 48583 + }, + { + "epoch": 0.48584, + "grad_norm": 1.2081904181379792, + "learning_rate": 0.003, + "loss": 4.022, + "step": 48584 + }, + { + "epoch": 0.48585, + "grad_norm": 1.0241387842411407, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 48585 + }, + { + "epoch": 0.48586, + "grad_norm": 1.007165576704762, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 48586 + }, + { + "epoch": 0.48587, + "grad_norm": 1.0335353659903486, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 48587 + }, + { + "epoch": 0.48588, + "grad_norm": 1.1295380858729576, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 48588 + }, + { + "epoch": 0.48589, + "grad_norm": 1.0164650889882343, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 48589 + }, + { + "epoch": 0.4859, + "grad_norm": 1.1055065642187107, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 48590 + }, + { + "epoch": 0.48591, + "grad_norm": 0.9490002695807603, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 48591 + }, + { + "epoch": 0.48592, + "grad_norm": 1.0809287074693261, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 48592 + }, + { + "epoch": 0.48593, + "grad_norm": 1.0594045053494412, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 48593 + }, + { + "epoch": 0.48594, + "grad_norm": 0.9732387143196267, + "learning_rate": 0.003, + "loss": 4.071, + "step": 48594 + }, + { + "epoch": 0.48595, + "grad_norm": 0.9372820158155107, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 48595 + }, + { + "epoch": 0.48596, + "grad_norm": 1.0113037434405439, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 48596 + }, + { + "epoch": 0.48597, + "grad_norm": 1.0364766402533636, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 48597 + }, + { + "epoch": 0.48598, + "grad_norm": 1.0756732710723689, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 48598 + }, + { + "epoch": 0.48599, + "grad_norm": 1.4258253697056658, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 48599 + }, + { + "epoch": 0.486, + "grad_norm": 0.8428174932465583, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 48600 + }, + { + "epoch": 0.48601, + "grad_norm": 0.9792519374697556, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 48601 + }, + { + "epoch": 0.48602, + "grad_norm": 1.378910594003317, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 48602 + }, + { + "epoch": 0.48603, + "grad_norm": 1.452872657927826, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 48603 + }, + { + "epoch": 0.48604, + "grad_norm": 0.865650407818537, + "learning_rate": 0.003, + "loss": 4.016, + "step": 48604 + }, + { + "epoch": 0.48605, + "grad_norm": 0.9174894553169, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 48605 + }, + { + "epoch": 0.48606, + "grad_norm": 1.0508158372803538, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 48606 + }, + { + "epoch": 0.48607, + "grad_norm": 1.0828654426412674, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 48607 + }, + { + "epoch": 0.48608, + "grad_norm": 1.1758681111076927, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 48608 + }, + { + "epoch": 0.48609, + "grad_norm": 1.0373580317298938, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 48609 + }, + { + "epoch": 0.4861, + "grad_norm": 1.2384220739447511, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 48610 + }, + { + "epoch": 0.48611, + "grad_norm": 0.9722709154041139, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 48611 + }, + { + "epoch": 0.48612, + "grad_norm": 0.985077523564927, + "learning_rate": 0.003, + "loss": 4.022, + "step": 48612 + }, + { + "epoch": 0.48613, + "grad_norm": 1.2717998613714605, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 48613 + }, + { + "epoch": 0.48614, + "grad_norm": 0.9169804200472724, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 48614 + }, + { + "epoch": 0.48615, + "grad_norm": 0.7661469981469134, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 48615 + }, + { + "epoch": 0.48616, + "grad_norm": 0.8444685034848429, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 48616 + }, + { + "epoch": 0.48617, + "grad_norm": 0.9947134690587481, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 48617 + }, + { + "epoch": 0.48618, + "grad_norm": 1.275457731506487, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 48618 + }, + { + "epoch": 0.48619, + "grad_norm": 0.7748224725417763, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 48619 + }, + { + "epoch": 0.4862, + "grad_norm": 0.8406707480649528, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 48620 + }, + { + "epoch": 0.48621, + "grad_norm": 0.9011438447212783, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 48621 + }, + { + "epoch": 0.48622, + "grad_norm": 1.4547149300253264, + "learning_rate": 0.003, + "loss": 4.035, + "step": 48622 + }, + { + "epoch": 0.48623, + "grad_norm": 0.8049191905648401, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 48623 + }, + { + "epoch": 0.48624, + "grad_norm": 0.7431105359212283, + "learning_rate": 0.003, + "loss": 4.006, + "step": 48624 + }, + { + "epoch": 0.48625, + "grad_norm": 1.0288234629450104, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 48625 + }, + { + "epoch": 0.48626, + "grad_norm": 1.4442575206825488, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 48626 + }, + { + "epoch": 0.48627, + "grad_norm": 0.7245736924131655, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 48627 + }, + { + "epoch": 0.48628, + "grad_norm": 0.7889962472945133, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 48628 + }, + { + "epoch": 0.48629, + "grad_norm": 0.999791537445404, + "learning_rate": 0.003, + "loss": 4.026, + "step": 48629 + }, + { + "epoch": 0.4863, + "grad_norm": 1.5392219428152758, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 48630 + }, + { + "epoch": 0.48631, + "grad_norm": 0.7865135100565812, + "learning_rate": 0.003, + "loss": 4.028, + "step": 48631 + }, + { + "epoch": 0.48632, + "grad_norm": 0.8818326103374723, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 48632 + }, + { + "epoch": 0.48633, + "grad_norm": 1.159674167706202, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 48633 + }, + { + "epoch": 0.48634, + "grad_norm": 1.2310517282882505, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 48634 + }, + { + "epoch": 0.48635, + "grad_norm": 1.0098978771924985, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 48635 + }, + { + "epoch": 0.48636, + "grad_norm": 1.1737524729057043, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 48636 + }, + { + "epoch": 0.48637, + "grad_norm": 1.039188651662492, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 48637 + }, + { + "epoch": 0.48638, + "grad_norm": 0.9767551544181111, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 48638 + }, + { + "epoch": 0.48639, + "grad_norm": 1.1019470824366433, + "learning_rate": 0.003, + "loss": 4.076, + "step": 48639 + }, + { + "epoch": 0.4864, + "grad_norm": 1.2151332913812476, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 48640 + }, + { + "epoch": 0.48641, + "grad_norm": 1.319861562678172, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 48641 + }, + { + "epoch": 0.48642, + "grad_norm": 0.9185938222117421, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 48642 + }, + { + "epoch": 0.48643, + "grad_norm": 0.9513679609804169, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 48643 + }, + { + "epoch": 0.48644, + "grad_norm": 1.3575741949253575, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 48644 + }, + { + "epoch": 0.48645, + "grad_norm": 0.8854603755908655, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 48645 + }, + { + "epoch": 0.48646, + "grad_norm": 0.8898256400236546, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 48646 + }, + { + "epoch": 0.48647, + "grad_norm": 0.929093356780397, + "learning_rate": 0.003, + "loss": 4.023, + "step": 48647 + }, + { + "epoch": 0.48648, + "grad_norm": 1.1228179388479538, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 48648 + }, + { + "epoch": 0.48649, + "grad_norm": 1.0903787192466317, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 48649 + }, + { + "epoch": 0.4865, + "grad_norm": 1.1182527700457312, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 48650 + }, + { + "epoch": 0.48651, + "grad_norm": 0.9606809898642642, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 48651 + }, + { + "epoch": 0.48652, + "grad_norm": 0.9159489595642002, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 48652 + }, + { + "epoch": 0.48653, + "grad_norm": 1.0609949409842323, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 48653 + }, + { + "epoch": 0.48654, + "grad_norm": 1.2650145179717296, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 48654 + }, + { + "epoch": 0.48655, + "grad_norm": 0.9884165365842714, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 48655 + }, + { + "epoch": 0.48656, + "grad_norm": 1.0514403530135803, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 48656 + }, + { + "epoch": 0.48657, + "grad_norm": 1.1508125119244708, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 48657 + }, + { + "epoch": 0.48658, + "grad_norm": 1.1291600448309986, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 48658 + }, + { + "epoch": 0.48659, + "grad_norm": 0.8796266229642556, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 48659 + }, + { + "epoch": 0.4866, + "grad_norm": 0.8934475019198171, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 48660 + }, + { + "epoch": 0.48661, + "grad_norm": 1.1292183746949322, + "learning_rate": 0.003, + "loss": 4.029, + "step": 48661 + }, + { + "epoch": 0.48662, + "grad_norm": 1.1933950901468175, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 48662 + }, + { + "epoch": 0.48663, + "grad_norm": 1.00721085145173, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 48663 + }, + { + "epoch": 0.48664, + "grad_norm": 1.1836674225471586, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 48664 + }, + { + "epoch": 0.48665, + "grad_norm": 1.0582868868873387, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 48665 + }, + { + "epoch": 0.48666, + "grad_norm": 1.1193941290083036, + "learning_rate": 0.003, + "loss": 4.035, + "step": 48666 + }, + { + "epoch": 0.48667, + "grad_norm": 1.1443114618459118, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 48667 + }, + { + "epoch": 0.48668, + "grad_norm": 0.9004031696400132, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 48668 + }, + { + "epoch": 0.48669, + "grad_norm": 0.9539425764998913, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 48669 + }, + { + "epoch": 0.4867, + "grad_norm": 0.8210889333470873, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 48670 + }, + { + "epoch": 0.48671, + "grad_norm": 0.9574114622942315, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 48671 + }, + { + "epoch": 0.48672, + "grad_norm": 1.2578769625958823, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 48672 + }, + { + "epoch": 0.48673, + "grad_norm": 0.8550407692530421, + "learning_rate": 0.003, + "loss": 4.047, + "step": 48673 + }, + { + "epoch": 0.48674, + "grad_norm": 1.0271724147739003, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 48674 + }, + { + "epoch": 0.48675, + "grad_norm": 1.2828675309900122, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 48675 + }, + { + "epoch": 0.48676, + "grad_norm": 1.2039781809890182, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 48676 + }, + { + "epoch": 0.48677, + "grad_norm": 1.018622603362046, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 48677 + }, + { + "epoch": 0.48678, + "grad_norm": 0.9796736885166836, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 48678 + }, + { + "epoch": 0.48679, + "grad_norm": 1.2353917247941097, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 48679 + }, + { + "epoch": 0.4868, + "grad_norm": 1.116018975008936, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 48680 + }, + { + "epoch": 0.48681, + "grad_norm": 0.9524157903131095, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 48681 + }, + { + "epoch": 0.48682, + "grad_norm": 0.9786263722457904, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 48682 + }, + { + "epoch": 0.48683, + "grad_norm": 0.9696246007834429, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 48683 + }, + { + "epoch": 0.48684, + "grad_norm": 1.082073793927187, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 48684 + }, + { + "epoch": 0.48685, + "grad_norm": 1.0919349823134787, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 48685 + }, + { + "epoch": 0.48686, + "grad_norm": 0.991437322812774, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 48686 + }, + { + "epoch": 0.48687, + "grad_norm": 1.179097531512117, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 48687 + }, + { + "epoch": 0.48688, + "grad_norm": 1.0138984766246164, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 48688 + }, + { + "epoch": 0.48689, + "grad_norm": 1.0999323192938857, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 48689 + }, + { + "epoch": 0.4869, + "grad_norm": 1.0171880308321417, + "learning_rate": 0.003, + "loss": 4.064, + "step": 48690 + }, + { + "epoch": 0.48691, + "grad_norm": 1.0530913322189452, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 48691 + }, + { + "epoch": 0.48692, + "grad_norm": 1.1989680119997594, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 48692 + }, + { + "epoch": 0.48693, + "grad_norm": 1.0626068626156682, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 48693 + }, + { + "epoch": 0.48694, + "grad_norm": 1.0605049324163558, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 48694 + }, + { + "epoch": 0.48695, + "grad_norm": 1.2178381956748463, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 48695 + }, + { + "epoch": 0.48696, + "grad_norm": 0.7909149088221235, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 48696 + }, + { + "epoch": 0.48697, + "grad_norm": 0.9516857712052734, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 48697 + }, + { + "epoch": 0.48698, + "grad_norm": 1.1121655871910907, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 48698 + }, + { + "epoch": 0.48699, + "grad_norm": 1.097082027174629, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 48699 + }, + { + "epoch": 0.487, + "grad_norm": 1.2645936356383005, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 48700 + }, + { + "epoch": 0.48701, + "grad_norm": 1.1173693849393935, + "learning_rate": 0.003, + "loss": 4.044, + "step": 48701 + }, + { + "epoch": 0.48702, + "grad_norm": 1.0647971517708497, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 48702 + }, + { + "epoch": 0.48703, + "grad_norm": 1.2717766349022575, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 48703 + }, + { + "epoch": 0.48704, + "grad_norm": 0.7642440250043377, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 48704 + }, + { + "epoch": 0.48705, + "grad_norm": 0.7252316986267854, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 48705 + }, + { + "epoch": 0.48706, + "grad_norm": 0.7236502264703852, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 48706 + }, + { + "epoch": 0.48707, + "grad_norm": 0.8833930062017146, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 48707 + }, + { + "epoch": 0.48708, + "grad_norm": 0.896178709770834, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 48708 + }, + { + "epoch": 0.48709, + "grad_norm": 0.967829256641069, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 48709 + }, + { + "epoch": 0.4871, + "grad_norm": 1.0291578067562928, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 48710 + }, + { + "epoch": 0.48711, + "grad_norm": 1.1002538436516296, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 48711 + }, + { + "epoch": 0.48712, + "grad_norm": 1.1349117061317078, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 48712 + }, + { + "epoch": 0.48713, + "grad_norm": 1.286743461945293, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 48713 + }, + { + "epoch": 0.48714, + "grad_norm": 0.8268123821569399, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 48714 + }, + { + "epoch": 0.48715, + "grad_norm": 0.9509625963984838, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 48715 + }, + { + "epoch": 0.48716, + "grad_norm": 1.1245119490782975, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 48716 + }, + { + "epoch": 0.48717, + "grad_norm": 1.296948959863321, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 48717 + }, + { + "epoch": 0.48718, + "grad_norm": 0.8341338134960389, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 48718 + }, + { + "epoch": 0.48719, + "grad_norm": 0.8060217976615196, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 48719 + }, + { + "epoch": 0.4872, + "grad_norm": 0.8738402399353038, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 48720 + }, + { + "epoch": 0.48721, + "grad_norm": 1.0975785613249736, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 48721 + }, + { + "epoch": 0.48722, + "grad_norm": 1.063738019657289, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 48722 + }, + { + "epoch": 0.48723, + "grad_norm": 0.9933369357670037, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 48723 + }, + { + "epoch": 0.48724, + "grad_norm": 1.543085630923439, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 48724 + }, + { + "epoch": 0.48725, + "grad_norm": 0.8459865037358747, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 48725 + }, + { + "epoch": 0.48726, + "grad_norm": 0.81586963867616, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 48726 + }, + { + "epoch": 0.48727, + "grad_norm": 0.8917430282284088, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 48727 + }, + { + "epoch": 0.48728, + "grad_norm": 1.1141816850228552, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 48728 + }, + { + "epoch": 0.48729, + "grad_norm": 1.1483713955283277, + "learning_rate": 0.003, + "loss": 4.033, + "step": 48729 + }, + { + "epoch": 0.4873, + "grad_norm": 1.2163711818277079, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 48730 + }, + { + "epoch": 0.48731, + "grad_norm": 1.0495353740198026, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 48731 + }, + { + "epoch": 0.48732, + "grad_norm": 1.289526188767768, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 48732 + }, + { + "epoch": 0.48733, + "grad_norm": 0.9394227036046623, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 48733 + }, + { + "epoch": 0.48734, + "grad_norm": 0.9479311447639133, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 48734 + }, + { + "epoch": 0.48735, + "grad_norm": 1.0227042453623651, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 48735 + }, + { + "epoch": 0.48736, + "grad_norm": 1.3676740214098082, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 48736 + }, + { + "epoch": 0.48737, + "grad_norm": 0.8168725282596888, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 48737 + }, + { + "epoch": 0.48738, + "grad_norm": 1.0088369022024966, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 48738 + }, + { + "epoch": 0.48739, + "grad_norm": 1.0401912138679084, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 48739 + }, + { + "epoch": 0.4874, + "grad_norm": 1.037488974322982, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 48740 + }, + { + "epoch": 0.48741, + "grad_norm": 1.0049257606486337, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 48741 + }, + { + "epoch": 0.48742, + "grad_norm": 1.1713222186978183, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 48742 + }, + { + "epoch": 0.48743, + "grad_norm": 0.9807709309004681, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 48743 + }, + { + "epoch": 0.48744, + "grad_norm": 0.956809784206664, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 48744 + }, + { + "epoch": 0.48745, + "grad_norm": 1.2075549355895407, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 48745 + }, + { + "epoch": 0.48746, + "grad_norm": 1.3032499968955402, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 48746 + }, + { + "epoch": 0.48747, + "grad_norm": 1.0342549677361506, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 48747 + }, + { + "epoch": 0.48748, + "grad_norm": 1.3937003315383474, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 48748 + }, + { + "epoch": 0.48749, + "grad_norm": 1.0975516684428872, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 48749 + }, + { + "epoch": 0.4875, + "grad_norm": 1.0142222299438344, + "learning_rate": 0.003, + "loss": 4.023, + "step": 48750 + }, + { + "epoch": 0.48751, + "grad_norm": 1.0296433984065676, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 48751 + }, + { + "epoch": 0.48752, + "grad_norm": 1.2395405578904146, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 48752 + }, + { + "epoch": 0.48753, + "grad_norm": 0.8880379178709402, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 48753 + }, + { + "epoch": 0.48754, + "grad_norm": 0.9548948874377013, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 48754 + }, + { + "epoch": 0.48755, + "grad_norm": 1.0949569261535341, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 48755 + }, + { + "epoch": 0.48756, + "grad_norm": 1.362644599993793, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 48756 + }, + { + "epoch": 0.48757, + "grad_norm": 0.9605451985484251, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 48757 + }, + { + "epoch": 0.48758, + "grad_norm": 1.2189209066022992, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 48758 + }, + { + "epoch": 0.48759, + "grad_norm": 1.1431452693036759, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 48759 + }, + { + "epoch": 0.4876, + "grad_norm": 0.9396502783191782, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 48760 + }, + { + "epoch": 0.48761, + "grad_norm": 1.1514214261548972, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 48761 + }, + { + "epoch": 0.48762, + "grad_norm": 1.1494760179351486, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 48762 + }, + { + "epoch": 0.48763, + "grad_norm": 1.0218610084937545, + "learning_rate": 0.003, + "loss": 4.056, + "step": 48763 + }, + { + "epoch": 0.48764, + "grad_norm": 0.8011113929401039, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 48764 + }, + { + "epoch": 0.48765, + "grad_norm": 0.9378324013383054, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 48765 + }, + { + "epoch": 0.48766, + "grad_norm": 1.0890262069408285, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 48766 + }, + { + "epoch": 0.48767, + "grad_norm": 0.9930996311330192, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 48767 + }, + { + "epoch": 0.48768, + "grad_norm": 1.0759647472197686, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 48768 + }, + { + "epoch": 0.48769, + "grad_norm": 1.5290479521043716, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 48769 + }, + { + "epoch": 0.4877, + "grad_norm": 0.81300583361832, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 48770 + }, + { + "epoch": 0.48771, + "grad_norm": 0.9099530810522456, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 48771 + }, + { + "epoch": 0.48772, + "grad_norm": 0.9572990287407916, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 48772 + }, + { + "epoch": 0.48773, + "grad_norm": 1.2532049332707262, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 48773 + }, + { + "epoch": 0.48774, + "grad_norm": 0.8743018685897441, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 48774 + }, + { + "epoch": 0.48775, + "grad_norm": 0.8964438345653586, + "learning_rate": 0.003, + "loss": 4.038, + "step": 48775 + }, + { + "epoch": 0.48776, + "grad_norm": 1.0717186052580678, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 48776 + }, + { + "epoch": 0.48777, + "grad_norm": 1.3264776326732517, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 48777 + }, + { + "epoch": 0.48778, + "grad_norm": 1.0126723830779023, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 48778 + }, + { + "epoch": 0.48779, + "grad_norm": 1.0658767656256762, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 48779 + }, + { + "epoch": 0.4878, + "grad_norm": 1.1650101555291728, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 48780 + }, + { + "epoch": 0.48781, + "grad_norm": 1.13419749212359, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 48781 + }, + { + "epoch": 0.48782, + "grad_norm": 0.9808961702592359, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 48782 + }, + { + "epoch": 0.48783, + "grad_norm": 1.086896395575693, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 48783 + }, + { + "epoch": 0.48784, + "grad_norm": 1.0686797340448153, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 48784 + }, + { + "epoch": 0.48785, + "grad_norm": 0.9478762928436612, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 48785 + }, + { + "epoch": 0.48786, + "grad_norm": 0.8699277237858866, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 48786 + }, + { + "epoch": 0.48787, + "grad_norm": 0.8863057434470983, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 48787 + }, + { + "epoch": 0.48788, + "grad_norm": 0.9575411024798068, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 48788 + }, + { + "epoch": 0.48789, + "grad_norm": 1.3161580693271953, + "learning_rate": 0.003, + "loss": 4.032, + "step": 48789 + }, + { + "epoch": 0.4879, + "grad_norm": 1.1227519906844492, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 48790 + }, + { + "epoch": 0.48791, + "grad_norm": 0.8938429397484041, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 48791 + }, + { + "epoch": 0.48792, + "grad_norm": 0.966317035087971, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 48792 + }, + { + "epoch": 0.48793, + "grad_norm": 1.1050568897360655, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 48793 + }, + { + "epoch": 0.48794, + "grad_norm": 1.1373151500264307, + "learning_rate": 0.003, + "loss": 4.042, + "step": 48794 + }, + { + "epoch": 0.48795, + "grad_norm": 1.0572093812797203, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 48795 + }, + { + "epoch": 0.48796, + "grad_norm": 1.0996773142053966, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 48796 + }, + { + "epoch": 0.48797, + "grad_norm": 0.9895179547600627, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 48797 + }, + { + "epoch": 0.48798, + "grad_norm": 1.2760183786548231, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 48798 + }, + { + "epoch": 0.48799, + "grad_norm": 0.6928290840421234, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 48799 + }, + { + "epoch": 0.488, + "grad_norm": 0.7302272509784645, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 48800 + }, + { + "epoch": 0.48801, + "grad_norm": 0.8807575759750463, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 48801 + }, + { + "epoch": 0.48802, + "grad_norm": 1.129658329910449, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 48802 + }, + { + "epoch": 0.48803, + "grad_norm": 1.0493945893705476, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 48803 + }, + { + "epoch": 0.48804, + "grad_norm": 1.0857031259490342, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 48804 + }, + { + "epoch": 0.48805, + "grad_norm": 1.0504590895447012, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 48805 + }, + { + "epoch": 0.48806, + "grad_norm": 1.0897022060764883, + "learning_rate": 0.003, + "loss": 4.067, + "step": 48806 + }, + { + "epoch": 0.48807, + "grad_norm": 1.1775288259543162, + "learning_rate": 0.003, + "loss": 4.04, + "step": 48807 + }, + { + "epoch": 0.48808, + "grad_norm": 1.0027489731084316, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 48808 + }, + { + "epoch": 0.48809, + "grad_norm": 0.9157132251109132, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 48809 + }, + { + "epoch": 0.4881, + "grad_norm": 0.9980002040256547, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 48810 + }, + { + "epoch": 0.48811, + "grad_norm": 1.2861108630056426, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 48811 + }, + { + "epoch": 0.48812, + "grad_norm": 1.186937306519685, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 48812 + }, + { + "epoch": 0.48813, + "grad_norm": 1.0110461834948956, + "learning_rate": 0.003, + "loss": 4.01, + "step": 48813 + }, + { + "epoch": 0.48814, + "grad_norm": 1.1225179288079874, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 48814 + }, + { + "epoch": 0.48815, + "grad_norm": 1.2230604550304127, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 48815 + }, + { + "epoch": 0.48816, + "grad_norm": 0.8826305485981293, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 48816 + }, + { + "epoch": 0.48817, + "grad_norm": 1.0012099890523014, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 48817 + }, + { + "epoch": 0.48818, + "grad_norm": 1.0987182318695994, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 48818 + }, + { + "epoch": 0.48819, + "grad_norm": 1.072332769346498, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 48819 + }, + { + "epoch": 0.4882, + "grad_norm": 1.268845758864542, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 48820 + }, + { + "epoch": 0.48821, + "grad_norm": 1.0378781103159436, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 48821 + }, + { + "epoch": 0.48822, + "grad_norm": 0.9408946979926168, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 48822 + }, + { + "epoch": 0.48823, + "grad_norm": 0.9757117414716336, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 48823 + }, + { + "epoch": 0.48824, + "grad_norm": 1.1081564024838686, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 48824 + }, + { + "epoch": 0.48825, + "grad_norm": 1.0249767204369313, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 48825 + }, + { + "epoch": 0.48826, + "grad_norm": 1.3775590079343225, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 48826 + }, + { + "epoch": 0.48827, + "grad_norm": 0.770160288878485, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 48827 + }, + { + "epoch": 0.48828, + "grad_norm": 0.7762460315797292, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 48828 + }, + { + "epoch": 0.48829, + "grad_norm": 0.7790560319235531, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 48829 + }, + { + "epoch": 0.4883, + "grad_norm": 1.0100479716796433, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 48830 + }, + { + "epoch": 0.48831, + "grad_norm": 1.5829680341290477, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 48831 + }, + { + "epoch": 0.48832, + "grad_norm": 0.8358045417792873, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 48832 + }, + { + "epoch": 0.48833, + "grad_norm": 0.9558245998978707, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 48833 + }, + { + "epoch": 0.48834, + "grad_norm": 1.1442522081256787, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 48834 + }, + { + "epoch": 0.48835, + "grad_norm": 0.9343015568362754, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 48835 + }, + { + "epoch": 0.48836, + "grad_norm": 1.165816220497808, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 48836 + }, + { + "epoch": 0.48837, + "grad_norm": 1.24097461354687, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 48837 + }, + { + "epoch": 0.48838, + "grad_norm": 1.2056699332409408, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 48838 + }, + { + "epoch": 0.48839, + "grad_norm": 0.8274812547593152, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 48839 + }, + { + "epoch": 0.4884, + "grad_norm": 0.7917172603175179, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 48840 + }, + { + "epoch": 0.48841, + "grad_norm": 0.914715743802581, + "learning_rate": 0.003, + "loss": 4.043, + "step": 48841 + }, + { + "epoch": 0.48842, + "grad_norm": 1.2469361849729443, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 48842 + }, + { + "epoch": 0.48843, + "grad_norm": 0.9107675825700473, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 48843 + }, + { + "epoch": 0.48844, + "grad_norm": 0.9219837687749836, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 48844 + }, + { + "epoch": 0.48845, + "grad_norm": 1.1218132916353696, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 48845 + }, + { + "epoch": 0.48846, + "grad_norm": 1.24659396449277, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 48846 + }, + { + "epoch": 0.48847, + "grad_norm": 0.8331189484052829, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 48847 + }, + { + "epoch": 0.48848, + "grad_norm": 0.8952095080140442, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 48848 + }, + { + "epoch": 0.48849, + "grad_norm": 1.1221413369297555, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 48849 + }, + { + "epoch": 0.4885, + "grad_norm": 1.1156172244380704, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 48850 + }, + { + "epoch": 0.48851, + "grad_norm": 1.1477326315380991, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 48851 + }, + { + "epoch": 0.48852, + "grad_norm": 1.1785900253668997, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 48852 + }, + { + "epoch": 0.48853, + "grad_norm": 1.1695349993006785, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 48853 + }, + { + "epoch": 0.48854, + "grad_norm": 1.0192208220476962, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 48854 + }, + { + "epoch": 0.48855, + "grad_norm": 1.0046536863140856, + "learning_rate": 0.003, + "loss": 4.031, + "step": 48855 + }, + { + "epoch": 0.48856, + "grad_norm": 1.1659077375450841, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 48856 + }, + { + "epoch": 0.48857, + "grad_norm": 0.8496017370960572, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 48857 + }, + { + "epoch": 0.48858, + "grad_norm": 0.9651729187732349, + "learning_rate": 0.003, + "loss": 4.059, + "step": 48858 + }, + { + "epoch": 0.48859, + "grad_norm": 1.3795286701912488, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 48859 + }, + { + "epoch": 0.4886, + "grad_norm": 1.1131161984235778, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 48860 + }, + { + "epoch": 0.48861, + "grad_norm": 1.1905543925745632, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 48861 + }, + { + "epoch": 0.48862, + "grad_norm": 0.9609213921282552, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 48862 + }, + { + "epoch": 0.48863, + "grad_norm": 1.025230687860591, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 48863 + }, + { + "epoch": 0.48864, + "grad_norm": 1.2949490053683532, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 48864 + }, + { + "epoch": 0.48865, + "grad_norm": 0.8671443822632676, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 48865 + }, + { + "epoch": 0.48866, + "grad_norm": 0.9944464120277341, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 48866 + }, + { + "epoch": 0.48867, + "grad_norm": 1.3841186641784877, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 48867 + }, + { + "epoch": 0.48868, + "grad_norm": 0.8558617991162019, + "learning_rate": 0.003, + "loss": 4.039, + "step": 48868 + }, + { + "epoch": 0.48869, + "grad_norm": 0.9153581226475287, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 48869 + }, + { + "epoch": 0.4887, + "grad_norm": 1.0311060989745315, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 48870 + }, + { + "epoch": 0.48871, + "grad_norm": 1.1334295601964384, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 48871 + }, + { + "epoch": 0.48872, + "grad_norm": 1.2983788920073718, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 48872 + }, + { + "epoch": 0.48873, + "grad_norm": 1.0364057946125662, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 48873 + }, + { + "epoch": 0.48874, + "grad_norm": 1.1398255113421947, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 48874 + }, + { + "epoch": 0.48875, + "grad_norm": 1.1386877165465117, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 48875 + }, + { + "epoch": 0.48876, + "grad_norm": 1.0372000284039762, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 48876 + }, + { + "epoch": 0.48877, + "grad_norm": 1.0766671856191252, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 48877 + }, + { + "epoch": 0.48878, + "grad_norm": 0.9141835699079321, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 48878 + }, + { + "epoch": 0.48879, + "grad_norm": 0.7668574771464343, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 48879 + }, + { + "epoch": 0.4888, + "grad_norm": 0.828140136109984, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 48880 + }, + { + "epoch": 0.48881, + "grad_norm": 1.0059631025246782, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 48881 + }, + { + "epoch": 0.48882, + "grad_norm": 1.1791299902844294, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 48882 + }, + { + "epoch": 0.48883, + "grad_norm": 0.9213396849149986, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 48883 + }, + { + "epoch": 0.48884, + "grad_norm": 1.3752248779879257, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 48884 + }, + { + "epoch": 0.48885, + "grad_norm": 1.3298156262367964, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 48885 + }, + { + "epoch": 0.48886, + "grad_norm": 0.8377750647453698, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 48886 + }, + { + "epoch": 0.48887, + "grad_norm": 0.774657607658734, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 48887 + }, + { + "epoch": 0.48888, + "grad_norm": 0.8000266687020058, + "learning_rate": 0.003, + "loss": 4.028, + "step": 48888 + }, + { + "epoch": 0.48889, + "grad_norm": 0.937407974676957, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 48889 + }, + { + "epoch": 0.4889, + "grad_norm": 1.1308539255831702, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 48890 + }, + { + "epoch": 0.48891, + "grad_norm": 1.1858583364429487, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 48891 + }, + { + "epoch": 0.48892, + "grad_norm": 1.1935282425407068, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 48892 + }, + { + "epoch": 0.48893, + "grad_norm": 0.8720694485672434, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 48893 + }, + { + "epoch": 0.48894, + "grad_norm": 1.0051592434305892, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 48894 + }, + { + "epoch": 0.48895, + "grad_norm": 1.3586445823064317, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 48895 + }, + { + "epoch": 0.48896, + "grad_norm": 1.064790538339109, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 48896 + }, + { + "epoch": 0.48897, + "grad_norm": 1.0340811775851095, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 48897 + }, + { + "epoch": 0.48898, + "grad_norm": 1.1620728621161438, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 48898 + }, + { + "epoch": 0.48899, + "grad_norm": 1.0019259403065108, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 48899 + }, + { + "epoch": 0.489, + "grad_norm": 0.962638302556279, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 48900 + }, + { + "epoch": 0.48901, + "grad_norm": 1.0548826461674017, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 48901 + }, + { + "epoch": 0.48902, + "grad_norm": 1.1640854950229915, + "learning_rate": 0.003, + "loss": 4.0868, + "step": 48902 + }, + { + "epoch": 0.48903, + "grad_norm": 0.969413718248838, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 48903 + }, + { + "epoch": 0.48904, + "grad_norm": 1.1502716771360881, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 48904 + }, + { + "epoch": 0.48905, + "grad_norm": 1.087565859363513, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 48905 + }, + { + "epoch": 0.48906, + "grad_norm": 1.0287708516049037, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 48906 + }, + { + "epoch": 0.48907, + "grad_norm": 1.2347578284904295, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 48907 + }, + { + "epoch": 0.48908, + "grad_norm": 1.023719964009332, + "learning_rate": 0.003, + "loss": 4.02, + "step": 48908 + }, + { + "epoch": 0.48909, + "grad_norm": 1.2563750070760815, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 48909 + }, + { + "epoch": 0.4891, + "grad_norm": 1.0383777076448344, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 48910 + }, + { + "epoch": 0.48911, + "grad_norm": 1.1633419632212196, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 48911 + }, + { + "epoch": 0.48912, + "grad_norm": 0.9907545662829795, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 48912 + }, + { + "epoch": 0.48913, + "grad_norm": 1.3206127472250344, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 48913 + }, + { + "epoch": 0.48914, + "grad_norm": 0.9773397586127495, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 48914 + }, + { + "epoch": 0.48915, + "grad_norm": 1.1812901846619428, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 48915 + }, + { + "epoch": 0.48916, + "grad_norm": 1.2349252696128876, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 48916 + }, + { + "epoch": 0.48917, + "grad_norm": 0.9406786102863931, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 48917 + }, + { + "epoch": 0.48918, + "grad_norm": 1.1722799065840999, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 48918 + }, + { + "epoch": 0.48919, + "grad_norm": 1.1358176272126153, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 48919 + }, + { + "epoch": 0.4892, + "grad_norm": 0.8786116511977459, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 48920 + }, + { + "epoch": 0.48921, + "grad_norm": 0.8795569403844965, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 48921 + }, + { + "epoch": 0.48922, + "grad_norm": 0.908865915980031, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 48922 + }, + { + "epoch": 0.48923, + "grad_norm": 0.9211936171143349, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 48923 + }, + { + "epoch": 0.48924, + "grad_norm": 1.044180646645428, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 48924 + }, + { + "epoch": 0.48925, + "grad_norm": 1.0168549469600145, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 48925 + }, + { + "epoch": 0.48926, + "grad_norm": 1.2659324731288581, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 48926 + }, + { + "epoch": 0.48927, + "grad_norm": 0.846506726515703, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 48927 + }, + { + "epoch": 0.48928, + "grad_norm": 0.8664422134977217, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 48928 + }, + { + "epoch": 0.48929, + "grad_norm": 1.3739110414383966, + "learning_rate": 0.003, + "loss": 4.043, + "step": 48929 + }, + { + "epoch": 0.4893, + "grad_norm": 1.3588429938885762, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 48930 + }, + { + "epoch": 0.48931, + "grad_norm": 0.9053722112166304, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 48931 + }, + { + "epoch": 0.48932, + "grad_norm": 1.0473534072357562, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 48932 + }, + { + "epoch": 0.48933, + "grad_norm": 1.4888485493714538, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 48933 + }, + { + "epoch": 0.48934, + "grad_norm": 0.9787341922917073, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 48934 + }, + { + "epoch": 0.48935, + "grad_norm": 0.9511065896086544, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 48935 + }, + { + "epoch": 0.48936, + "grad_norm": 0.857232926803757, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 48936 + }, + { + "epoch": 0.48937, + "grad_norm": 1.00849818132864, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 48937 + }, + { + "epoch": 0.48938, + "grad_norm": 1.3504757603067916, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 48938 + }, + { + "epoch": 0.48939, + "grad_norm": 0.7950110828762996, + "learning_rate": 0.003, + "loss": 4.013, + "step": 48939 + }, + { + "epoch": 0.4894, + "grad_norm": 0.7864148427288531, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 48940 + }, + { + "epoch": 0.48941, + "grad_norm": 0.8251268341268788, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 48941 + }, + { + "epoch": 0.48942, + "grad_norm": 0.9334189441430917, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 48942 + }, + { + "epoch": 0.48943, + "grad_norm": 1.1924571536188513, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 48943 + }, + { + "epoch": 0.48944, + "grad_norm": 1.0115732780504112, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 48944 + }, + { + "epoch": 0.48945, + "grad_norm": 1.2504411237314363, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 48945 + }, + { + "epoch": 0.48946, + "grad_norm": 0.9576531008107402, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 48946 + }, + { + "epoch": 0.48947, + "grad_norm": 1.2220879823231325, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 48947 + }, + { + "epoch": 0.48948, + "grad_norm": 1.2108208867260628, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 48948 + }, + { + "epoch": 0.48949, + "grad_norm": 0.8954562161790792, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 48949 + }, + { + "epoch": 0.4895, + "grad_norm": 0.8640544267343468, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 48950 + }, + { + "epoch": 0.48951, + "grad_norm": 1.3405799825071867, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 48951 + }, + { + "epoch": 0.48952, + "grad_norm": 1.1601538075271784, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 48952 + }, + { + "epoch": 0.48953, + "grad_norm": 0.8541347995626516, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 48953 + }, + { + "epoch": 0.48954, + "grad_norm": 0.8861498394972743, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 48954 + }, + { + "epoch": 0.48955, + "grad_norm": 0.881647774898461, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 48955 + }, + { + "epoch": 0.48956, + "grad_norm": 1.0703894699565555, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 48956 + }, + { + "epoch": 0.48957, + "grad_norm": 1.1727361509604766, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 48957 + }, + { + "epoch": 0.48958, + "grad_norm": 1.0762356158625845, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 48958 + }, + { + "epoch": 0.48959, + "grad_norm": 1.1675591682614983, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 48959 + }, + { + "epoch": 0.4896, + "grad_norm": 1.019531655875176, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 48960 + }, + { + "epoch": 0.48961, + "grad_norm": 1.0593720908635869, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 48961 + }, + { + "epoch": 0.48962, + "grad_norm": 1.0655244558828805, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 48962 + }, + { + "epoch": 0.48963, + "grad_norm": 1.1415895230897337, + "learning_rate": 0.003, + "loss": 4.013, + "step": 48963 + }, + { + "epoch": 0.48964, + "grad_norm": 1.1767680391198312, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 48964 + }, + { + "epoch": 0.48965, + "grad_norm": 0.8407712611588645, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 48965 + }, + { + "epoch": 0.48966, + "grad_norm": 0.9542499677437966, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 48966 + }, + { + "epoch": 0.48967, + "grad_norm": 1.174609801480071, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 48967 + }, + { + "epoch": 0.48968, + "grad_norm": 1.2345545313854758, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 48968 + }, + { + "epoch": 0.48969, + "grad_norm": 1.139351758757994, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 48969 + }, + { + "epoch": 0.4897, + "grad_norm": 1.0455459166120873, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 48970 + }, + { + "epoch": 0.48971, + "grad_norm": 1.1515715952883347, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 48971 + }, + { + "epoch": 0.48972, + "grad_norm": 1.0211097276050727, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 48972 + }, + { + "epoch": 0.48973, + "grad_norm": 1.2106767852365827, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 48973 + }, + { + "epoch": 0.48974, + "grad_norm": 0.9833037065543626, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 48974 + }, + { + "epoch": 0.48975, + "grad_norm": 1.0544627200665737, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 48975 + }, + { + "epoch": 0.48976, + "grad_norm": 1.0092369377841794, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 48976 + }, + { + "epoch": 0.48977, + "grad_norm": 1.2131699812288046, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 48977 + }, + { + "epoch": 0.48978, + "grad_norm": 1.0130668979998936, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 48978 + }, + { + "epoch": 0.48979, + "grad_norm": 1.163093815727833, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 48979 + }, + { + "epoch": 0.4898, + "grad_norm": 1.1329037646223334, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 48980 + }, + { + "epoch": 0.48981, + "grad_norm": 1.1264873115387373, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 48981 + }, + { + "epoch": 0.48982, + "grad_norm": 1.1534429707915368, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 48982 + }, + { + "epoch": 0.48983, + "grad_norm": 0.9616864254747549, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 48983 + }, + { + "epoch": 0.48984, + "grad_norm": 1.5050465716781918, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 48984 + }, + { + "epoch": 0.48985, + "grad_norm": 1.089220519846415, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 48985 + }, + { + "epoch": 0.48986, + "grad_norm": 1.1662724716538597, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 48986 + }, + { + "epoch": 0.48987, + "grad_norm": 1.096105800777478, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 48987 + }, + { + "epoch": 0.48988, + "grad_norm": 1.0414181291373559, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 48988 + }, + { + "epoch": 0.48989, + "grad_norm": 1.1154116021249671, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 48989 + }, + { + "epoch": 0.4899, + "grad_norm": 1.001678518676386, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 48990 + }, + { + "epoch": 0.48991, + "grad_norm": 1.3149263246063043, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 48991 + }, + { + "epoch": 0.48992, + "grad_norm": 1.0012055683148409, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 48992 + }, + { + "epoch": 0.48993, + "grad_norm": 0.9570199297431635, + "learning_rate": 0.003, + "loss": 4.021, + "step": 48993 + }, + { + "epoch": 0.48994, + "grad_norm": 1.1704576534830071, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 48994 + }, + { + "epoch": 0.48995, + "grad_norm": 1.2398397639302046, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 48995 + }, + { + "epoch": 0.48996, + "grad_norm": 0.8938182421651276, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 48996 + }, + { + "epoch": 0.48997, + "grad_norm": 1.0008389350695186, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 48997 + }, + { + "epoch": 0.48998, + "grad_norm": 1.2233446986290664, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 48998 + }, + { + "epoch": 0.48999, + "grad_norm": 1.0895926491469843, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 48999 + }, + { + "epoch": 0.49, + "grad_norm": 1.2050675216157563, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 49000 + }, + { + "epoch": 0.49001, + "grad_norm": 1.010359902059954, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 49001 + }, + { + "epoch": 0.49002, + "grad_norm": 0.9454317801777694, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 49002 + }, + { + "epoch": 0.49003, + "grad_norm": 0.9565395838939322, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 49003 + }, + { + "epoch": 0.49004, + "grad_norm": 1.084489276360542, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 49004 + }, + { + "epoch": 0.49005, + "grad_norm": 1.1297210830454603, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 49005 + }, + { + "epoch": 0.49006, + "grad_norm": 0.8775465555891675, + "learning_rate": 0.003, + "loss": 4.046, + "step": 49006 + }, + { + "epoch": 0.49007, + "grad_norm": 0.9242999598062525, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 49007 + }, + { + "epoch": 0.49008, + "grad_norm": 1.2191103028250476, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 49008 + }, + { + "epoch": 0.49009, + "grad_norm": 1.1843254512626489, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 49009 + }, + { + "epoch": 0.4901, + "grad_norm": 1.0062087395937018, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 49010 + }, + { + "epoch": 0.49011, + "grad_norm": 1.0826632306626234, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 49011 + }, + { + "epoch": 0.49012, + "grad_norm": 1.3206736053214931, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 49012 + }, + { + "epoch": 0.49013, + "grad_norm": 0.987056403505762, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 49013 + }, + { + "epoch": 0.49014, + "grad_norm": 1.3129028301483638, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 49014 + }, + { + "epoch": 0.49015, + "grad_norm": 0.8954531701380999, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 49015 + }, + { + "epoch": 0.49016, + "grad_norm": 1.084226057794335, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 49016 + }, + { + "epoch": 0.49017, + "grad_norm": 1.1083583047193748, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 49017 + }, + { + "epoch": 0.49018, + "grad_norm": 1.1047488737731213, + "learning_rate": 0.003, + "loss": 4.039, + "step": 49018 + }, + { + "epoch": 0.49019, + "grad_norm": 0.9196966752677854, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 49019 + }, + { + "epoch": 0.4902, + "grad_norm": 1.049676062815256, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 49020 + }, + { + "epoch": 0.49021, + "grad_norm": 1.3773881992192725, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 49021 + }, + { + "epoch": 0.49022, + "grad_norm": 0.8119823862741572, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 49022 + }, + { + "epoch": 0.49023, + "grad_norm": 0.7387187409401781, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 49023 + }, + { + "epoch": 0.49024, + "grad_norm": 0.721501924833064, + "learning_rate": 0.003, + "loss": 4.018, + "step": 49024 + }, + { + "epoch": 0.49025, + "grad_norm": 0.7948549379139022, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 49025 + }, + { + "epoch": 0.49026, + "grad_norm": 1.1024997676167547, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 49026 + }, + { + "epoch": 0.49027, + "grad_norm": 1.1413224174625267, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 49027 + }, + { + "epoch": 0.49028, + "grad_norm": 1.1854531883493356, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 49028 + }, + { + "epoch": 0.49029, + "grad_norm": 1.1986092567665143, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 49029 + }, + { + "epoch": 0.4903, + "grad_norm": 0.8909202906155237, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 49030 + }, + { + "epoch": 0.49031, + "grad_norm": 1.0539218331496847, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 49031 + }, + { + "epoch": 0.49032, + "grad_norm": 1.388615414398309, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 49032 + }, + { + "epoch": 0.49033, + "grad_norm": 0.8147819575710591, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 49033 + }, + { + "epoch": 0.49034, + "grad_norm": 0.7600069963702779, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 49034 + }, + { + "epoch": 0.49035, + "grad_norm": 0.8027573704778115, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 49035 + }, + { + "epoch": 0.49036, + "grad_norm": 0.8726531830499856, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 49036 + }, + { + "epoch": 0.49037, + "grad_norm": 1.170906419335551, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 49037 + }, + { + "epoch": 0.49038, + "grad_norm": 1.4743693300383, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 49038 + }, + { + "epoch": 0.49039, + "grad_norm": 0.898383472703651, + "learning_rate": 0.003, + "loss": 4.011, + "step": 49039 + }, + { + "epoch": 0.4904, + "grad_norm": 1.037504401541361, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 49040 + }, + { + "epoch": 0.49041, + "grad_norm": 1.280392694922106, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 49041 + }, + { + "epoch": 0.49042, + "grad_norm": 1.0727522816442963, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 49042 + }, + { + "epoch": 0.49043, + "grad_norm": 1.0863259143345412, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 49043 + }, + { + "epoch": 0.49044, + "grad_norm": 1.1737836254662817, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 49044 + }, + { + "epoch": 0.49045, + "grad_norm": 0.8413109863605902, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 49045 + }, + { + "epoch": 0.49046, + "grad_norm": 0.7623233188374375, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 49046 + }, + { + "epoch": 0.49047, + "grad_norm": 1.2503937794430844, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 49047 + }, + { + "epoch": 0.49048, + "grad_norm": 1.2405381132625337, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 49048 + }, + { + "epoch": 0.49049, + "grad_norm": 0.9363385446689454, + "learning_rate": 0.003, + "loss": 4.035, + "step": 49049 + }, + { + "epoch": 0.4905, + "grad_norm": 1.048763341987416, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 49050 + }, + { + "epoch": 0.49051, + "grad_norm": 0.8747438063215898, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 49051 + }, + { + "epoch": 0.49052, + "grad_norm": 0.9873519743455379, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 49052 + }, + { + "epoch": 0.49053, + "grad_norm": 1.3861199020239205, + "learning_rate": 0.003, + "loss": 4.053, + "step": 49053 + }, + { + "epoch": 0.49054, + "grad_norm": 0.7223590484608226, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 49054 + }, + { + "epoch": 0.49055, + "grad_norm": 0.8918973058819941, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 49055 + }, + { + "epoch": 0.49056, + "grad_norm": 1.0455863858034462, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 49056 + }, + { + "epoch": 0.49057, + "grad_norm": 1.0716607782313772, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 49057 + }, + { + "epoch": 0.49058, + "grad_norm": 1.045750081736579, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 49058 + }, + { + "epoch": 0.49059, + "grad_norm": 1.3581055508104074, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 49059 + }, + { + "epoch": 0.4906, + "grad_norm": 0.9615328467359114, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 49060 + }, + { + "epoch": 0.49061, + "grad_norm": 1.0470008765382466, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 49061 + }, + { + "epoch": 0.49062, + "grad_norm": 1.1394986574156876, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 49062 + }, + { + "epoch": 0.49063, + "grad_norm": 1.1694281892102, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 49063 + }, + { + "epoch": 0.49064, + "grad_norm": 1.101725154522915, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 49064 + }, + { + "epoch": 0.49065, + "grad_norm": 0.9066333384825974, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 49065 + }, + { + "epoch": 0.49066, + "grad_norm": 0.864130484001445, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 49066 + }, + { + "epoch": 0.49067, + "grad_norm": 1.0404982108469454, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 49067 + }, + { + "epoch": 0.49068, + "grad_norm": 1.3464351467051927, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 49068 + }, + { + "epoch": 0.49069, + "grad_norm": 0.820509831710882, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 49069 + }, + { + "epoch": 0.4907, + "grad_norm": 0.8611504070787629, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 49070 + }, + { + "epoch": 0.49071, + "grad_norm": 1.0431768799641106, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 49071 + }, + { + "epoch": 0.49072, + "grad_norm": 1.453422840234712, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 49072 + }, + { + "epoch": 0.49073, + "grad_norm": 0.8626623854088812, + "learning_rate": 0.003, + "loss": 4.003, + "step": 49073 + }, + { + "epoch": 0.49074, + "grad_norm": 1.0499084814754533, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 49074 + }, + { + "epoch": 0.49075, + "grad_norm": 1.3356555506850878, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 49075 + }, + { + "epoch": 0.49076, + "grad_norm": 0.897042234471013, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 49076 + }, + { + "epoch": 0.49077, + "grad_norm": 0.8409967773390916, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 49077 + }, + { + "epoch": 0.49078, + "grad_norm": 0.9333216840386088, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 49078 + }, + { + "epoch": 0.49079, + "grad_norm": 1.0745767193547335, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 49079 + }, + { + "epoch": 0.4908, + "grad_norm": 1.0573080645753583, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 49080 + }, + { + "epoch": 0.49081, + "grad_norm": 0.9764914911611122, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 49081 + }, + { + "epoch": 0.49082, + "grad_norm": 1.0811113817365472, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 49082 + }, + { + "epoch": 0.49083, + "grad_norm": 0.9694487727567008, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 49083 + }, + { + "epoch": 0.49084, + "grad_norm": 1.0972225889044418, + "learning_rate": 0.003, + "loss": 4.044, + "step": 49084 + }, + { + "epoch": 0.49085, + "grad_norm": 1.3358479618714871, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 49085 + }, + { + "epoch": 0.49086, + "grad_norm": 1.2333854905156965, + "learning_rate": 0.003, + "loss": 4.027, + "step": 49086 + }, + { + "epoch": 0.49087, + "grad_norm": 1.0633964287352808, + "learning_rate": 0.003, + "loss": 4.026, + "step": 49087 + }, + { + "epoch": 0.49088, + "grad_norm": 1.0830712231411377, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 49088 + }, + { + "epoch": 0.49089, + "grad_norm": 1.0204300196082956, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 49089 + }, + { + "epoch": 0.4909, + "grad_norm": 0.9725299250583778, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 49090 + }, + { + "epoch": 0.49091, + "grad_norm": 1.1357268510386134, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 49091 + }, + { + "epoch": 0.49092, + "grad_norm": 1.018504171769619, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 49092 + }, + { + "epoch": 0.49093, + "grad_norm": 0.8759842508970341, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 49093 + }, + { + "epoch": 0.49094, + "grad_norm": 1.1439140852695773, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 49094 + }, + { + "epoch": 0.49095, + "grad_norm": 1.5428940150257533, + "learning_rate": 0.003, + "loss": 4.013, + "step": 49095 + }, + { + "epoch": 0.49096, + "grad_norm": 0.8781282869853977, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 49096 + }, + { + "epoch": 0.49097, + "grad_norm": 1.1156257515568644, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 49097 + }, + { + "epoch": 0.49098, + "grad_norm": 1.3286070631197733, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 49098 + }, + { + "epoch": 0.49099, + "grad_norm": 0.890745219176396, + "learning_rate": 0.003, + "loss": 4.049, + "step": 49099 + }, + { + "epoch": 0.491, + "grad_norm": 0.955798847607149, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 49100 + }, + { + "epoch": 0.49101, + "grad_norm": 1.0217691880761464, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 49101 + }, + { + "epoch": 0.49102, + "grad_norm": 1.2074300971804643, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 49102 + }, + { + "epoch": 0.49103, + "grad_norm": 0.9609022300835338, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 49103 + }, + { + "epoch": 0.49104, + "grad_norm": 1.1884504154616782, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 49104 + }, + { + "epoch": 0.49105, + "grad_norm": 1.1053495245734606, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 49105 + }, + { + "epoch": 0.49106, + "grad_norm": 1.0749315318557429, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 49106 + }, + { + "epoch": 0.49107, + "grad_norm": 0.9969514952410988, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 49107 + }, + { + "epoch": 0.49108, + "grad_norm": 1.1000783212003193, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 49108 + }, + { + "epoch": 0.49109, + "grad_norm": 1.237192537763126, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 49109 + }, + { + "epoch": 0.4911, + "grad_norm": 0.9099693904485768, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 49110 + }, + { + "epoch": 0.49111, + "grad_norm": 0.8477125605750591, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 49111 + }, + { + "epoch": 0.49112, + "grad_norm": 0.9686736897778037, + "learning_rate": 0.003, + "loss": 4.043, + "step": 49112 + }, + { + "epoch": 0.49113, + "grad_norm": 1.195777075427517, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 49113 + }, + { + "epoch": 0.49114, + "grad_norm": 0.9680490711335262, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 49114 + }, + { + "epoch": 0.49115, + "grad_norm": 1.1501928889562814, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 49115 + }, + { + "epoch": 0.49116, + "grad_norm": 1.1492757241507683, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 49116 + }, + { + "epoch": 0.49117, + "grad_norm": 1.009290356108231, + "learning_rate": 0.003, + "loss": 4.022, + "step": 49117 + }, + { + "epoch": 0.49118, + "grad_norm": 0.9747533583971518, + "learning_rate": 0.003, + "loss": 4.019, + "step": 49118 + }, + { + "epoch": 0.49119, + "grad_norm": 1.1490026122626578, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 49119 + }, + { + "epoch": 0.4912, + "grad_norm": 0.9666636848482159, + "learning_rate": 0.003, + "loss": 4.03, + "step": 49120 + }, + { + "epoch": 0.49121, + "grad_norm": 1.0994037048646275, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 49121 + }, + { + "epoch": 0.49122, + "grad_norm": 1.0806566083646143, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 49122 + }, + { + "epoch": 0.49123, + "grad_norm": 1.103336864022119, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 49123 + }, + { + "epoch": 0.49124, + "grad_norm": 1.1037309172338943, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 49124 + }, + { + "epoch": 0.49125, + "grad_norm": 0.9764883970630656, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 49125 + }, + { + "epoch": 0.49126, + "grad_norm": 1.1148771461371472, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 49126 + }, + { + "epoch": 0.49127, + "grad_norm": 1.0203485810874544, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 49127 + }, + { + "epoch": 0.49128, + "grad_norm": 1.0005556820126076, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 49128 + }, + { + "epoch": 0.49129, + "grad_norm": 1.0672243168030326, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 49129 + }, + { + "epoch": 0.4913, + "grad_norm": 1.4243146006221177, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 49130 + }, + { + "epoch": 0.49131, + "grad_norm": 1.0791817503931707, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 49131 + }, + { + "epoch": 0.49132, + "grad_norm": 1.0846324284220183, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 49132 + }, + { + "epoch": 0.49133, + "grad_norm": 1.1727979076021304, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 49133 + }, + { + "epoch": 0.49134, + "grad_norm": 0.9190040054442621, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 49134 + }, + { + "epoch": 0.49135, + "grad_norm": 0.8829467785245331, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 49135 + }, + { + "epoch": 0.49136, + "grad_norm": 1.0497531207108801, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 49136 + }, + { + "epoch": 0.49137, + "grad_norm": 1.2551247243958725, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 49137 + }, + { + "epoch": 0.49138, + "grad_norm": 1.0758661570824284, + "learning_rate": 0.003, + "loss": 4.045, + "step": 49138 + }, + { + "epoch": 0.49139, + "grad_norm": 1.056097825620937, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 49139 + }, + { + "epoch": 0.4914, + "grad_norm": 1.038570815791298, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 49140 + }, + { + "epoch": 0.49141, + "grad_norm": 1.062830681677059, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 49141 + }, + { + "epoch": 0.49142, + "grad_norm": 1.1765600993719088, + "learning_rate": 0.003, + "loss": 4.006, + "step": 49142 + }, + { + "epoch": 0.49143, + "grad_norm": 1.2041441259972077, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 49143 + }, + { + "epoch": 0.49144, + "grad_norm": 1.0159126232145932, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 49144 + }, + { + "epoch": 0.49145, + "grad_norm": 1.1274638805735593, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 49145 + }, + { + "epoch": 0.49146, + "grad_norm": 0.9668408366781147, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 49146 + }, + { + "epoch": 0.49147, + "grad_norm": 0.8849132719167117, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 49147 + }, + { + "epoch": 0.49148, + "grad_norm": 1.0994834832405787, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 49148 + }, + { + "epoch": 0.49149, + "grad_norm": 1.3425845829477312, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 49149 + }, + { + "epoch": 0.4915, + "grad_norm": 0.8269836851462565, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 49150 + }, + { + "epoch": 0.49151, + "grad_norm": 0.8165145326419523, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 49151 + }, + { + "epoch": 0.49152, + "grad_norm": 1.0924680823927266, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 49152 + }, + { + "epoch": 0.49153, + "grad_norm": 1.225488467125904, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 49153 + }, + { + "epoch": 0.49154, + "grad_norm": 0.8727332023503293, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 49154 + }, + { + "epoch": 0.49155, + "grad_norm": 0.949101788999269, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 49155 + }, + { + "epoch": 0.49156, + "grad_norm": 1.0621891426449788, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 49156 + }, + { + "epoch": 0.49157, + "grad_norm": 1.0663445477620883, + "learning_rate": 0.003, + "loss": 4.042, + "step": 49157 + }, + { + "epoch": 0.49158, + "grad_norm": 1.045534681841705, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 49158 + }, + { + "epoch": 0.49159, + "grad_norm": 1.1812810529040045, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 49159 + }, + { + "epoch": 0.4916, + "grad_norm": 1.1378478091094584, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 49160 + }, + { + "epoch": 0.49161, + "grad_norm": 1.0779249241582314, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 49161 + }, + { + "epoch": 0.49162, + "grad_norm": 1.1221266675205201, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 49162 + }, + { + "epoch": 0.49163, + "grad_norm": 0.9663001392063153, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 49163 + }, + { + "epoch": 0.49164, + "grad_norm": 1.0619155078158844, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 49164 + }, + { + "epoch": 0.49165, + "grad_norm": 0.9530282347532888, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 49165 + }, + { + "epoch": 0.49166, + "grad_norm": 1.0217006739446934, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 49166 + }, + { + "epoch": 0.49167, + "grad_norm": 1.0773608948968636, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 49167 + }, + { + "epoch": 0.49168, + "grad_norm": 0.9341049173028858, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 49168 + }, + { + "epoch": 0.49169, + "grad_norm": 0.9118984941760417, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 49169 + }, + { + "epoch": 0.4917, + "grad_norm": 1.0813575383243177, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 49170 + }, + { + "epoch": 0.49171, + "grad_norm": 0.9577212008575938, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 49171 + }, + { + "epoch": 0.49172, + "grad_norm": 1.266094529750386, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 49172 + }, + { + "epoch": 0.49173, + "grad_norm": 1.07979277968552, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 49173 + }, + { + "epoch": 0.49174, + "grad_norm": 0.9901047186851188, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 49174 + }, + { + "epoch": 0.49175, + "grad_norm": 0.9982711953619293, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 49175 + }, + { + "epoch": 0.49176, + "grad_norm": 1.0152739734439185, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 49176 + }, + { + "epoch": 0.49177, + "grad_norm": 1.158535346334141, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 49177 + }, + { + "epoch": 0.49178, + "grad_norm": 1.08851846170299, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 49178 + }, + { + "epoch": 0.49179, + "grad_norm": 1.0600796891575706, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 49179 + }, + { + "epoch": 0.4918, + "grad_norm": 1.0693748633512947, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 49180 + }, + { + "epoch": 0.49181, + "grad_norm": 1.0264502985054877, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 49181 + }, + { + "epoch": 0.49182, + "grad_norm": 1.146252557262954, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 49182 + }, + { + "epoch": 0.49183, + "grad_norm": 0.9482529397369714, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 49183 + }, + { + "epoch": 0.49184, + "grad_norm": 0.9904523971994915, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 49184 + }, + { + "epoch": 0.49185, + "grad_norm": 1.4063875974135307, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 49185 + }, + { + "epoch": 0.49186, + "grad_norm": 1.12745866615605, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 49186 + }, + { + "epoch": 0.49187, + "grad_norm": 1.068743033827619, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 49187 + }, + { + "epoch": 0.49188, + "grad_norm": 1.1160192686433512, + "learning_rate": 0.003, + "loss": 4.042, + "step": 49188 + }, + { + "epoch": 0.49189, + "grad_norm": 0.9621050342503317, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 49189 + }, + { + "epoch": 0.4919, + "grad_norm": 1.2926906724891352, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 49190 + }, + { + "epoch": 0.49191, + "grad_norm": 1.0730178410433133, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 49191 + }, + { + "epoch": 0.49192, + "grad_norm": 1.2926611970669193, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 49192 + }, + { + "epoch": 0.49193, + "grad_norm": 0.8125934073722505, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 49193 + }, + { + "epoch": 0.49194, + "grad_norm": 0.8878150875665828, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 49194 + }, + { + "epoch": 0.49195, + "grad_norm": 1.1196088268049038, + "learning_rate": 0.003, + "loss": 4.024, + "step": 49195 + }, + { + "epoch": 0.49196, + "grad_norm": 1.2398656722781476, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 49196 + }, + { + "epoch": 0.49197, + "grad_norm": 1.0642316870280042, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 49197 + }, + { + "epoch": 0.49198, + "grad_norm": 0.9362911474277427, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 49198 + }, + { + "epoch": 0.49199, + "grad_norm": 1.1430421764186018, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 49199 + }, + { + "epoch": 0.492, + "grad_norm": 1.5049820720298317, + "learning_rate": 0.003, + "loss": 4.001, + "step": 49200 + }, + { + "epoch": 0.49201, + "grad_norm": 0.6575975551630873, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 49201 + }, + { + "epoch": 0.49202, + "grad_norm": 0.7636010128158577, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 49202 + }, + { + "epoch": 0.49203, + "grad_norm": 0.9927494681840622, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 49203 + }, + { + "epoch": 0.49204, + "grad_norm": 1.2028733076734675, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 49204 + }, + { + "epoch": 0.49205, + "grad_norm": 1.081078012051541, + "learning_rate": 0.003, + "loss": 4.032, + "step": 49205 + }, + { + "epoch": 0.49206, + "grad_norm": 1.141850956485316, + "learning_rate": 0.003, + "loss": 4.044, + "step": 49206 + }, + { + "epoch": 0.49207, + "grad_norm": 0.9076351314577369, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 49207 + }, + { + "epoch": 0.49208, + "grad_norm": 1.1918598953650363, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 49208 + }, + { + "epoch": 0.49209, + "grad_norm": 1.1484706328050547, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 49209 + }, + { + "epoch": 0.4921, + "grad_norm": 0.8195714849785308, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 49210 + }, + { + "epoch": 0.49211, + "grad_norm": 0.7475366893858983, + "learning_rate": 0.003, + "loss": 4.044, + "step": 49211 + }, + { + "epoch": 0.49212, + "grad_norm": 0.9403470087963496, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 49212 + }, + { + "epoch": 0.49213, + "grad_norm": 1.248047008340962, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 49213 + }, + { + "epoch": 0.49214, + "grad_norm": 1.1531169313561498, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 49214 + }, + { + "epoch": 0.49215, + "grad_norm": 1.339518259193491, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 49215 + }, + { + "epoch": 0.49216, + "grad_norm": 1.0099925891479649, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 49216 + }, + { + "epoch": 0.49217, + "grad_norm": 1.199683245122753, + "learning_rate": 0.003, + "loss": 4.022, + "step": 49217 + }, + { + "epoch": 0.49218, + "grad_norm": 0.9914024207260711, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 49218 + }, + { + "epoch": 0.49219, + "grad_norm": 1.0886264861979953, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 49219 + }, + { + "epoch": 0.4922, + "grad_norm": 1.0958728140656266, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 49220 + }, + { + "epoch": 0.49221, + "grad_norm": 1.0086754686758403, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 49221 + }, + { + "epoch": 0.49222, + "grad_norm": 0.9740704888928666, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 49222 + }, + { + "epoch": 0.49223, + "grad_norm": 1.0762385972520319, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 49223 + }, + { + "epoch": 0.49224, + "grad_norm": 0.8897500698024731, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 49224 + }, + { + "epoch": 0.49225, + "grad_norm": 0.9759168474056341, + "learning_rate": 0.003, + "loss": 4.015, + "step": 49225 + }, + { + "epoch": 0.49226, + "grad_norm": 1.0420361252087942, + "learning_rate": 0.003, + "loss": 4.043, + "step": 49226 + }, + { + "epoch": 0.49227, + "grad_norm": 1.1424943322046095, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 49227 + }, + { + "epoch": 0.49228, + "grad_norm": 1.1210406879912203, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 49228 + }, + { + "epoch": 0.49229, + "grad_norm": 1.2299216157427302, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 49229 + }, + { + "epoch": 0.4923, + "grad_norm": 1.078513913659395, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 49230 + }, + { + "epoch": 0.49231, + "grad_norm": 1.3804616557703921, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 49231 + }, + { + "epoch": 0.49232, + "grad_norm": 0.9912648793491445, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 49232 + }, + { + "epoch": 0.49233, + "grad_norm": 1.1874995680934082, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 49233 + }, + { + "epoch": 0.49234, + "grad_norm": 1.1840871528532297, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 49234 + }, + { + "epoch": 0.49235, + "grad_norm": 1.1636785814423045, + "learning_rate": 0.003, + "loss": 4.044, + "step": 49235 + }, + { + "epoch": 0.49236, + "grad_norm": 0.895007757956601, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 49236 + }, + { + "epoch": 0.49237, + "grad_norm": 0.8799178600049452, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 49237 + }, + { + "epoch": 0.49238, + "grad_norm": 0.9798887218582555, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 49238 + }, + { + "epoch": 0.49239, + "grad_norm": 1.14819313552999, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 49239 + }, + { + "epoch": 0.4924, + "grad_norm": 1.0844763652653895, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 49240 + }, + { + "epoch": 0.49241, + "grad_norm": 0.993852092228944, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 49241 + }, + { + "epoch": 0.49242, + "grad_norm": 1.228539643367483, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 49242 + }, + { + "epoch": 0.49243, + "grad_norm": 1.0152500559848598, + "learning_rate": 0.003, + "loss": 4.003, + "step": 49243 + }, + { + "epoch": 0.49244, + "grad_norm": 1.1344375692337376, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 49244 + }, + { + "epoch": 0.49245, + "grad_norm": 0.957181817897976, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 49245 + }, + { + "epoch": 0.49246, + "grad_norm": 1.2132456475235527, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 49246 + }, + { + "epoch": 0.49247, + "grad_norm": 1.0516842883641941, + "learning_rate": 0.003, + "loss": 4.03, + "step": 49247 + }, + { + "epoch": 0.49248, + "grad_norm": 1.0417543287503592, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 49248 + }, + { + "epoch": 0.49249, + "grad_norm": 0.9862029914012487, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 49249 + }, + { + "epoch": 0.4925, + "grad_norm": 1.173079139867489, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 49250 + }, + { + "epoch": 0.49251, + "grad_norm": 1.0390757337281529, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 49251 + }, + { + "epoch": 0.49252, + "grad_norm": 1.0925855857517053, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 49252 + }, + { + "epoch": 0.49253, + "grad_norm": 0.820700990962515, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 49253 + }, + { + "epoch": 0.49254, + "grad_norm": 1.1679758153560398, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 49254 + }, + { + "epoch": 0.49255, + "grad_norm": 1.046235005005969, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 49255 + }, + { + "epoch": 0.49256, + "grad_norm": 1.090229677803213, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 49256 + }, + { + "epoch": 0.49257, + "grad_norm": 1.1992294865873896, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 49257 + }, + { + "epoch": 0.49258, + "grad_norm": 0.9647110090712416, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 49258 + }, + { + "epoch": 0.49259, + "grad_norm": 1.3309587256861244, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 49259 + }, + { + "epoch": 0.4926, + "grad_norm": 0.8629737985646972, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 49260 + }, + { + "epoch": 0.49261, + "grad_norm": 0.9809763859441851, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 49261 + }, + { + "epoch": 0.49262, + "grad_norm": 1.204519241863506, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 49262 + }, + { + "epoch": 0.49263, + "grad_norm": 0.8537798563209814, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 49263 + }, + { + "epoch": 0.49264, + "grad_norm": 0.9344407203880835, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 49264 + }, + { + "epoch": 0.49265, + "grad_norm": 1.0693146915800675, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 49265 + }, + { + "epoch": 0.49266, + "grad_norm": 1.1624605114924762, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 49266 + }, + { + "epoch": 0.49267, + "grad_norm": 1.050795285617536, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 49267 + }, + { + "epoch": 0.49268, + "grad_norm": 1.158004083658773, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 49268 + }, + { + "epoch": 0.49269, + "grad_norm": 1.1670701010187334, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 49269 + }, + { + "epoch": 0.4927, + "grad_norm": 1.0199947232631053, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 49270 + }, + { + "epoch": 0.49271, + "grad_norm": 1.2797229446648986, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 49271 + }, + { + "epoch": 0.49272, + "grad_norm": 0.9666823122134364, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 49272 + }, + { + "epoch": 0.49273, + "grad_norm": 1.0234817354437151, + "learning_rate": 0.003, + "loss": 4.026, + "step": 49273 + }, + { + "epoch": 0.49274, + "grad_norm": 1.0612267350618425, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 49274 + }, + { + "epoch": 0.49275, + "grad_norm": 1.035412345588068, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 49275 + }, + { + "epoch": 0.49276, + "grad_norm": 1.2764975074248934, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 49276 + }, + { + "epoch": 0.49277, + "grad_norm": 0.8895862868420857, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 49277 + }, + { + "epoch": 0.49278, + "grad_norm": 0.879033201862357, + "learning_rate": 0.003, + "loss": 4.044, + "step": 49278 + }, + { + "epoch": 0.49279, + "grad_norm": 0.9211270741422579, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 49279 + }, + { + "epoch": 0.4928, + "grad_norm": 1.0215303736347894, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 49280 + }, + { + "epoch": 0.49281, + "grad_norm": 1.070471446997014, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 49281 + }, + { + "epoch": 0.49282, + "grad_norm": 1.0386971750937941, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 49282 + }, + { + "epoch": 0.49283, + "grad_norm": 1.1046010855608246, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 49283 + }, + { + "epoch": 0.49284, + "grad_norm": 1.1473377951693873, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 49284 + }, + { + "epoch": 0.49285, + "grad_norm": 0.9940383266568799, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 49285 + }, + { + "epoch": 0.49286, + "grad_norm": 1.2240921921006527, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 49286 + }, + { + "epoch": 0.49287, + "grad_norm": 1.0789739922587394, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 49287 + }, + { + "epoch": 0.49288, + "grad_norm": 1.2299689459236443, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 49288 + }, + { + "epoch": 0.49289, + "grad_norm": 1.1405280445505064, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 49289 + }, + { + "epoch": 0.4929, + "grad_norm": 0.9846670600528359, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 49290 + }, + { + "epoch": 0.49291, + "grad_norm": 1.214297535986238, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 49291 + }, + { + "epoch": 0.49292, + "grad_norm": 1.1276148149160758, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 49292 + }, + { + "epoch": 0.49293, + "grad_norm": 0.9772872946458645, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 49293 + }, + { + "epoch": 0.49294, + "grad_norm": 1.062550852075605, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 49294 + }, + { + "epoch": 0.49295, + "grad_norm": 0.932036214660298, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 49295 + }, + { + "epoch": 0.49296, + "grad_norm": 1.1244879025344807, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 49296 + }, + { + "epoch": 0.49297, + "grad_norm": 0.9633193832740404, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 49297 + }, + { + "epoch": 0.49298, + "grad_norm": 1.0363948064311863, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 49298 + }, + { + "epoch": 0.49299, + "grad_norm": 1.1444944254884297, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 49299 + }, + { + "epoch": 0.493, + "grad_norm": 1.0720290584272223, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 49300 + }, + { + "epoch": 0.49301, + "grad_norm": 1.2946784080689964, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 49301 + }, + { + "epoch": 0.49302, + "grad_norm": 0.8949785082170475, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 49302 + }, + { + "epoch": 0.49303, + "grad_norm": 0.9992686550108063, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 49303 + }, + { + "epoch": 0.49304, + "grad_norm": 1.1928051175305958, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 49304 + }, + { + "epoch": 0.49305, + "grad_norm": 1.0717810822625744, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 49305 + }, + { + "epoch": 0.49306, + "grad_norm": 0.915616500497949, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 49306 + }, + { + "epoch": 0.49307, + "grad_norm": 1.0897039173417684, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 49307 + }, + { + "epoch": 0.49308, + "grad_norm": 1.1718039193134406, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 49308 + }, + { + "epoch": 0.49309, + "grad_norm": 0.8510904967260997, + "learning_rate": 0.003, + "loss": 4.031, + "step": 49309 + }, + { + "epoch": 0.4931, + "grad_norm": 0.9473273010854314, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 49310 + }, + { + "epoch": 0.49311, + "grad_norm": 0.9953368524853462, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 49311 + }, + { + "epoch": 0.49312, + "grad_norm": 1.2051571028746038, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 49312 + }, + { + "epoch": 0.49313, + "grad_norm": 0.9377162515335703, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 49313 + }, + { + "epoch": 0.49314, + "grad_norm": 1.3424556392890636, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 49314 + }, + { + "epoch": 0.49315, + "grad_norm": 0.651621095009993, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 49315 + }, + { + "epoch": 0.49316, + "grad_norm": 0.7814772950782236, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 49316 + }, + { + "epoch": 0.49317, + "grad_norm": 0.8318862433280095, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 49317 + }, + { + "epoch": 0.49318, + "grad_norm": 0.8844941233588448, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 49318 + }, + { + "epoch": 0.49319, + "grad_norm": 1.0211468738720593, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 49319 + }, + { + "epoch": 0.4932, + "grad_norm": 1.2198889979044552, + "learning_rate": 0.003, + "loss": 4.002, + "step": 49320 + }, + { + "epoch": 0.49321, + "grad_norm": 0.9731566529020359, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 49321 + }, + { + "epoch": 0.49322, + "grad_norm": 1.1219417629711734, + "learning_rate": 0.003, + "loss": 4.019, + "step": 49322 + }, + { + "epoch": 0.49323, + "grad_norm": 1.202579946678168, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 49323 + }, + { + "epoch": 0.49324, + "grad_norm": 1.0546107228496897, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 49324 + }, + { + "epoch": 0.49325, + "grad_norm": 1.002341354180026, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 49325 + }, + { + "epoch": 0.49326, + "grad_norm": 1.196721104622245, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 49326 + }, + { + "epoch": 0.49327, + "grad_norm": 1.0728430375487057, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 49327 + }, + { + "epoch": 0.49328, + "grad_norm": 1.1506576566935751, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 49328 + }, + { + "epoch": 0.49329, + "grad_norm": 1.027476002425296, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 49329 + }, + { + "epoch": 0.4933, + "grad_norm": 1.1103566336244561, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 49330 + }, + { + "epoch": 0.49331, + "grad_norm": 1.0633559374026362, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 49331 + }, + { + "epoch": 0.49332, + "grad_norm": 1.0226777168607795, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 49332 + }, + { + "epoch": 0.49333, + "grad_norm": 1.151311377199092, + "learning_rate": 0.003, + "loss": 4.05, + "step": 49333 + }, + { + "epoch": 0.49334, + "grad_norm": 0.8572606511609909, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 49334 + }, + { + "epoch": 0.49335, + "grad_norm": 0.8133581236449123, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 49335 + }, + { + "epoch": 0.49336, + "grad_norm": 0.7949201133889732, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 49336 + }, + { + "epoch": 0.49337, + "grad_norm": 0.8697985597284199, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 49337 + }, + { + "epoch": 0.49338, + "grad_norm": 1.0181576742215044, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 49338 + }, + { + "epoch": 0.49339, + "grad_norm": 1.0835374340692527, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 49339 + }, + { + "epoch": 0.4934, + "grad_norm": 1.202264100770691, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 49340 + }, + { + "epoch": 0.49341, + "grad_norm": 1.2071411816216686, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 49341 + }, + { + "epoch": 0.49342, + "grad_norm": 0.8824264451362509, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 49342 + }, + { + "epoch": 0.49343, + "grad_norm": 0.882001481541728, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 49343 + }, + { + "epoch": 0.49344, + "grad_norm": 1.0985617109471837, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 49344 + }, + { + "epoch": 0.49345, + "grad_norm": 1.1806400815192477, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 49345 + }, + { + "epoch": 0.49346, + "grad_norm": 0.93761254754691, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 49346 + }, + { + "epoch": 0.49347, + "grad_norm": 1.1038632461339402, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 49347 + }, + { + "epoch": 0.49348, + "grad_norm": 1.339144554384445, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 49348 + }, + { + "epoch": 0.49349, + "grad_norm": 0.9608674876128022, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 49349 + }, + { + "epoch": 0.4935, + "grad_norm": 1.0927349905582555, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 49350 + }, + { + "epoch": 0.49351, + "grad_norm": 1.2421517336110521, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 49351 + }, + { + "epoch": 0.49352, + "grad_norm": 0.9194831184671909, + "learning_rate": 0.003, + "loss": 4.066, + "step": 49352 + }, + { + "epoch": 0.49353, + "grad_norm": 0.999813643825235, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 49353 + }, + { + "epoch": 0.49354, + "grad_norm": 1.176185225109736, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 49354 + }, + { + "epoch": 0.49355, + "grad_norm": 0.965430266432237, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 49355 + }, + { + "epoch": 0.49356, + "grad_norm": 1.171791038770554, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 49356 + }, + { + "epoch": 0.49357, + "grad_norm": 0.9957867949987222, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 49357 + }, + { + "epoch": 0.49358, + "grad_norm": 1.1184201440680324, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 49358 + }, + { + "epoch": 0.49359, + "grad_norm": 1.111818058473322, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 49359 + }, + { + "epoch": 0.4936, + "grad_norm": 0.9782511248171073, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 49360 + }, + { + "epoch": 0.49361, + "grad_norm": 1.194077313130977, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 49361 + }, + { + "epoch": 0.49362, + "grad_norm": 0.9906308039317462, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 49362 + }, + { + "epoch": 0.49363, + "grad_norm": 1.2526969744502057, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 49363 + }, + { + "epoch": 0.49364, + "grad_norm": 1.0834502452903136, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 49364 + }, + { + "epoch": 0.49365, + "grad_norm": 0.9792655225490066, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 49365 + }, + { + "epoch": 0.49366, + "grad_norm": 1.0121326292223043, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 49366 + }, + { + "epoch": 0.49367, + "grad_norm": 1.0403272404931645, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 49367 + }, + { + "epoch": 0.49368, + "grad_norm": 1.0750261323760657, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 49368 + }, + { + "epoch": 0.49369, + "grad_norm": 0.9051498123196164, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 49369 + }, + { + "epoch": 0.4937, + "grad_norm": 1.0107482712939364, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 49370 + }, + { + "epoch": 0.49371, + "grad_norm": 1.1632158378420416, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 49371 + }, + { + "epoch": 0.49372, + "grad_norm": 1.035061935283036, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 49372 + }, + { + "epoch": 0.49373, + "grad_norm": 1.1808689396095642, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 49373 + }, + { + "epoch": 0.49374, + "grad_norm": 1.0654202223967473, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 49374 + }, + { + "epoch": 0.49375, + "grad_norm": 1.2069662708943074, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 49375 + }, + { + "epoch": 0.49376, + "grad_norm": 1.0632252252970518, + "learning_rate": 0.003, + "loss": 4.031, + "step": 49376 + }, + { + "epoch": 0.49377, + "grad_norm": 1.0125758249286798, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 49377 + }, + { + "epoch": 0.49378, + "grad_norm": 1.1525204326766518, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 49378 + }, + { + "epoch": 0.49379, + "grad_norm": 0.9527353736581791, + "learning_rate": 0.003, + "loss": 4.033, + "step": 49379 + }, + { + "epoch": 0.4938, + "grad_norm": 1.1028470054599722, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 49380 + }, + { + "epoch": 0.49381, + "grad_norm": 1.0976042324695057, + "learning_rate": 0.003, + "loss": 4.068, + "step": 49381 + }, + { + "epoch": 0.49382, + "grad_norm": 0.9198826346234367, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 49382 + }, + { + "epoch": 0.49383, + "grad_norm": 0.9138030579004658, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 49383 + }, + { + "epoch": 0.49384, + "grad_norm": 1.0019538521210423, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 49384 + }, + { + "epoch": 0.49385, + "grad_norm": 1.260753507337035, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 49385 + }, + { + "epoch": 0.49386, + "grad_norm": 1.1621582599500047, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 49386 + }, + { + "epoch": 0.49387, + "grad_norm": 1.045160421065518, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 49387 + }, + { + "epoch": 0.49388, + "grad_norm": 1.1832157570466886, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 49388 + }, + { + "epoch": 0.49389, + "grad_norm": 1.0115879963245318, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 49389 + }, + { + "epoch": 0.4939, + "grad_norm": 1.091822236963356, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 49390 + }, + { + "epoch": 0.49391, + "grad_norm": 1.1248707733762031, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 49391 + }, + { + "epoch": 0.49392, + "grad_norm": 1.28056076184448, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 49392 + }, + { + "epoch": 0.49393, + "grad_norm": 0.7410122489743448, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 49393 + }, + { + "epoch": 0.49394, + "grad_norm": 0.6918819913706387, + "learning_rate": 0.003, + "loss": 4.031, + "step": 49394 + }, + { + "epoch": 0.49395, + "grad_norm": 0.8038810838020398, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 49395 + }, + { + "epoch": 0.49396, + "grad_norm": 0.9842402719852733, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 49396 + }, + { + "epoch": 0.49397, + "grad_norm": 1.3488057835445648, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 49397 + }, + { + "epoch": 0.49398, + "grad_norm": 1.015103673359849, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 49398 + }, + { + "epoch": 0.49399, + "grad_norm": 1.098721782146029, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 49399 + }, + { + "epoch": 0.494, + "grad_norm": 0.924921840028469, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 49400 + }, + { + "epoch": 0.49401, + "grad_norm": 0.899597989899219, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 49401 + }, + { + "epoch": 0.49402, + "grad_norm": 1.0367993408866807, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 49402 + }, + { + "epoch": 0.49403, + "grad_norm": 1.090243144502289, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 49403 + }, + { + "epoch": 0.49404, + "grad_norm": 0.9262868803173476, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 49404 + }, + { + "epoch": 0.49405, + "grad_norm": 0.888648854687553, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 49405 + }, + { + "epoch": 0.49406, + "grad_norm": 0.9191730978342209, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 49406 + }, + { + "epoch": 0.49407, + "grad_norm": 1.000181846167293, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 49407 + }, + { + "epoch": 0.49408, + "grad_norm": 1.5420325620270734, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 49408 + }, + { + "epoch": 0.49409, + "grad_norm": 1.196699994467587, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 49409 + }, + { + "epoch": 0.4941, + "grad_norm": 0.9837886077269481, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 49410 + }, + { + "epoch": 0.49411, + "grad_norm": 1.1526302435204372, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 49411 + }, + { + "epoch": 0.49412, + "grad_norm": 1.3189575776110616, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 49412 + }, + { + "epoch": 0.49413, + "grad_norm": 1.0518690889433993, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 49413 + }, + { + "epoch": 0.49414, + "grad_norm": 0.9703035654743843, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 49414 + }, + { + "epoch": 0.49415, + "grad_norm": 0.8925926453187825, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 49415 + }, + { + "epoch": 0.49416, + "grad_norm": 0.9856122950675201, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 49416 + }, + { + "epoch": 0.49417, + "grad_norm": 1.207681284500811, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 49417 + }, + { + "epoch": 0.49418, + "grad_norm": 0.9454287367204349, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 49418 + }, + { + "epoch": 0.49419, + "grad_norm": 0.9961634798722736, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 49419 + }, + { + "epoch": 0.4942, + "grad_norm": 1.012932943289938, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 49420 + }, + { + "epoch": 0.49421, + "grad_norm": 1.060109376028338, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 49421 + }, + { + "epoch": 0.49422, + "grad_norm": 1.3453877257081266, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 49422 + }, + { + "epoch": 0.49423, + "grad_norm": 1.1962707057310775, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 49423 + }, + { + "epoch": 0.49424, + "grad_norm": 0.8836234898739973, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 49424 + }, + { + "epoch": 0.49425, + "grad_norm": 0.9709510802887479, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 49425 + }, + { + "epoch": 0.49426, + "grad_norm": 1.1773995684158263, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 49426 + }, + { + "epoch": 0.49427, + "grad_norm": 1.0574810576157168, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 49427 + }, + { + "epoch": 0.49428, + "grad_norm": 1.1805497017689492, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 49428 + }, + { + "epoch": 0.49429, + "grad_norm": 1.0117355086293867, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 49429 + }, + { + "epoch": 0.4943, + "grad_norm": 1.2436241885835306, + "learning_rate": 0.003, + "loss": 4.025, + "step": 49430 + }, + { + "epoch": 0.49431, + "grad_norm": 0.8247131140833968, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 49431 + }, + { + "epoch": 0.49432, + "grad_norm": 0.9343700199170635, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 49432 + }, + { + "epoch": 0.49433, + "grad_norm": 1.0844894451713791, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 49433 + }, + { + "epoch": 0.49434, + "grad_norm": 1.0661547899326673, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 49434 + }, + { + "epoch": 0.49435, + "grad_norm": 1.2516355021340908, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 49435 + }, + { + "epoch": 0.49436, + "grad_norm": 0.8417064329119782, + "learning_rate": 0.003, + "loss": 4.055, + "step": 49436 + }, + { + "epoch": 0.49437, + "grad_norm": 1.0564100620571546, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 49437 + }, + { + "epoch": 0.49438, + "grad_norm": 1.2676907624419946, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 49438 + }, + { + "epoch": 0.49439, + "grad_norm": 1.3097766311003944, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 49439 + }, + { + "epoch": 0.4944, + "grad_norm": 0.8948128109771962, + "learning_rate": 0.003, + "loss": 4.061, + "step": 49440 + }, + { + "epoch": 0.49441, + "grad_norm": 0.9163607701462889, + "learning_rate": 0.003, + "loss": 3.999, + "step": 49441 + }, + { + "epoch": 0.49442, + "grad_norm": 1.1938095154409327, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 49442 + }, + { + "epoch": 0.49443, + "grad_norm": 1.222765971316597, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 49443 + }, + { + "epoch": 0.49444, + "grad_norm": 0.9842913928932754, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 49444 + }, + { + "epoch": 0.49445, + "grad_norm": 1.0145109923906108, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 49445 + }, + { + "epoch": 0.49446, + "grad_norm": 1.0910683916176505, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 49446 + }, + { + "epoch": 0.49447, + "grad_norm": 1.1579739858967926, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 49447 + }, + { + "epoch": 0.49448, + "grad_norm": 0.9272982603279489, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 49448 + }, + { + "epoch": 0.49449, + "grad_norm": 0.8595727704963307, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 49449 + }, + { + "epoch": 0.4945, + "grad_norm": 0.8942408948591483, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 49450 + }, + { + "epoch": 0.49451, + "grad_norm": 0.9869823516197072, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 49451 + }, + { + "epoch": 0.49452, + "grad_norm": 1.1978031425462732, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 49452 + }, + { + "epoch": 0.49453, + "grad_norm": 0.9711972571062094, + "learning_rate": 0.003, + "loss": 4.042, + "step": 49453 + }, + { + "epoch": 0.49454, + "grad_norm": 1.0308404687309909, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 49454 + }, + { + "epoch": 0.49455, + "grad_norm": 1.0913561902168465, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 49455 + }, + { + "epoch": 0.49456, + "grad_norm": 1.2337750720998866, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 49456 + }, + { + "epoch": 0.49457, + "grad_norm": 0.9474631667222174, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 49457 + }, + { + "epoch": 0.49458, + "grad_norm": 1.1962770570228312, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 49458 + }, + { + "epoch": 0.49459, + "grad_norm": 1.0244473635094762, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 49459 + }, + { + "epoch": 0.4946, + "grad_norm": 1.1082417552097674, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 49460 + }, + { + "epoch": 0.49461, + "grad_norm": 1.002465549592426, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 49461 + }, + { + "epoch": 0.49462, + "grad_norm": 1.3019888192366114, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 49462 + }, + { + "epoch": 0.49463, + "grad_norm": 0.9971756643393889, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 49463 + }, + { + "epoch": 0.49464, + "grad_norm": 1.0234227299871703, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 49464 + }, + { + "epoch": 0.49465, + "grad_norm": 1.1762152148699727, + "learning_rate": 0.003, + "loss": 3.995, + "step": 49465 + }, + { + "epoch": 0.49466, + "grad_norm": 1.2240234033785529, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 49466 + }, + { + "epoch": 0.49467, + "grad_norm": 0.9589697521371644, + "learning_rate": 0.003, + "loss": 4.045, + "step": 49467 + }, + { + "epoch": 0.49468, + "grad_norm": 1.0568004587374078, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 49468 + }, + { + "epoch": 0.49469, + "grad_norm": 1.0734338282509588, + "learning_rate": 0.003, + "loss": 4.067, + "step": 49469 + }, + { + "epoch": 0.4947, + "grad_norm": 0.9336533155564415, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 49470 + }, + { + "epoch": 0.49471, + "grad_norm": 1.0725592372646684, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 49471 + }, + { + "epoch": 0.49472, + "grad_norm": 1.162835832389635, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 49472 + }, + { + "epoch": 0.49473, + "grad_norm": 1.0704240269854866, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 49473 + }, + { + "epoch": 0.49474, + "grad_norm": 1.056071275969261, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 49474 + }, + { + "epoch": 0.49475, + "grad_norm": 1.2068859041427205, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 49475 + }, + { + "epoch": 0.49476, + "grad_norm": 0.7701442125535043, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 49476 + }, + { + "epoch": 0.49477, + "grad_norm": 0.6294685528071784, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 49477 + }, + { + "epoch": 0.49478, + "grad_norm": 0.6845921327661119, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 49478 + }, + { + "epoch": 0.49479, + "grad_norm": 0.7781549210789336, + "learning_rate": 0.003, + "loss": 4.029, + "step": 49479 + }, + { + "epoch": 0.4948, + "grad_norm": 0.8867968616564282, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 49480 + }, + { + "epoch": 0.49481, + "grad_norm": 1.1130414542821558, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 49481 + }, + { + "epoch": 0.49482, + "grad_norm": 1.2291172626326785, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 49482 + }, + { + "epoch": 0.49483, + "grad_norm": 0.9335112816237269, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 49483 + }, + { + "epoch": 0.49484, + "grad_norm": 0.8757275413677973, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 49484 + }, + { + "epoch": 0.49485, + "grad_norm": 1.0813978449413915, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 49485 + }, + { + "epoch": 0.49486, + "grad_norm": 1.317680486594207, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 49486 + }, + { + "epoch": 0.49487, + "grad_norm": 0.9027242477698153, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 49487 + }, + { + "epoch": 0.49488, + "grad_norm": 1.0106702531073506, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 49488 + }, + { + "epoch": 0.49489, + "grad_norm": 1.3460270760536845, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 49489 + }, + { + "epoch": 0.4949, + "grad_norm": 0.9149149550213526, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 49490 + }, + { + "epoch": 0.49491, + "grad_norm": 1.015406009237547, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 49491 + }, + { + "epoch": 0.49492, + "grad_norm": 1.2271908784394783, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 49492 + }, + { + "epoch": 0.49493, + "grad_norm": 1.0727259818587462, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 49493 + }, + { + "epoch": 0.49494, + "grad_norm": 1.2243582040176928, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 49494 + }, + { + "epoch": 0.49495, + "grad_norm": 1.0461598635725828, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 49495 + }, + { + "epoch": 0.49496, + "grad_norm": 1.0756682875796937, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 49496 + }, + { + "epoch": 0.49497, + "grad_norm": 1.1975310247521778, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 49497 + }, + { + "epoch": 0.49498, + "grad_norm": 1.010355438327365, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 49498 + }, + { + "epoch": 0.49499, + "grad_norm": 1.2404387893916948, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 49499 + }, + { + "epoch": 0.495, + "grad_norm": 1.055798061074413, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 49500 + }, + { + "epoch": 0.49501, + "grad_norm": 1.0249259443445473, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 49501 + }, + { + "epoch": 0.49502, + "grad_norm": 1.0239619024989002, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 49502 + }, + { + "epoch": 0.49503, + "grad_norm": 1.304880726169324, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 49503 + }, + { + "epoch": 0.49504, + "grad_norm": 0.9807342757054749, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 49504 + }, + { + "epoch": 0.49505, + "grad_norm": 1.1966405218983913, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 49505 + }, + { + "epoch": 0.49506, + "grad_norm": 1.1859539821513339, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 49506 + }, + { + "epoch": 0.49507, + "grad_norm": 0.9765252638369759, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 49507 + }, + { + "epoch": 0.49508, + "grad_norm": 0.9944149464518177, + "learning_rate": 0.003, + "loss": 4.029, + "step": 49508 + }, + { + "epoch": 0.49509, + "grad_norm": 1.4503468286485055, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 49509 + }, + { + "epoch": 0.4951, + "grad_norm": 1.0295878609013402, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 49510 + }, + { + "epoch": 0.49511, + "grad_norm": 1.010227874190945, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 49511 + }, + { + "epoch": 0.49512, + "grad_norm": 1.3598419780260649, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 49512 + }, + { + "epoch": 0.49513, + "grad_norm": 0.9349335007807497, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 49513 + }, + { + "epoch": 0.49514, + "grad_norm": 1.1298279488382874, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 49514 + }, + { + "epoch": 0.49515, + "grad_norm": 1.1422605086629, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 49515 + }, + { + "epoch": 0.49516, + "grad_norm": 1.290576906444309, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 49516 + }, + { + "epoch": 0.49517, + "grad_norm": 1.2138762609574654, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 49517 + }, + { + "epoch": 0.49518, + "grad_norm": 0.94553346301336, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 49518 + }, + { + "epoch": 0.49519, + "grad_norm": 1.0718356387812458, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 49519 + }, + { + "epoch": 0.4952, + "grad_norm": 1.067358620832765, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 49520 + }, + { + "epoch": 0.49521, + "grad_norm": 1.0365559606445258, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 49521 + }, + { + "epoch": 0.49522, + "grad_norm": 0.9316684034402541, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 49522 + }, + { + "epoch": 0.49523, + "grad_norm": 1.009987261057333, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 49523 + }, + { + "epoch": 0.49524, + "grad_norm": 1.133782000633094, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 49524 + }, + { + "epoch": 0.49525, + "grad_norm": 0.9318867109675432, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 49525 + }, + { + "epoch": 0.49526, + "grad_norm": 0.974450208176286, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 49526 + }, + { + "epoch": 0.49527, + "grad_norm": 0.9896391046579439, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 49527 + }, + { + "epoch": 0.49528, + "grad_norm": 1.0396271580516807, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 49528 + }, + { + "epoch": 0.49529, + "grad_norm": 1.1140504102557929, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 49529 + }, + { + "epoch": 0.4953, + "grad_norm": 1.0565148686588128, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 49530 + }, + { + "epoch": 0.49531, + "grad_norm": 1.0531919822072857, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 49531 + }, + { + "epoch": 0.49532, + "grad_norm": 1.3080264638141288, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 49532 + }, + { + "epoch": 0.49533, + "grad_norm": 0.8081937009277861, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 49533 + }, + { + "epoch": 0.49534, + "grad_norm": 0.9390162526182518, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 49534 + }, + { + "epoch": 0.49535, + "grad_norm": 1.1849116042382732, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 49535 + }, + { + "epoch": 0.49536, + "grad_norm": 0.9642467570534508, + "learning_rate": 0.003, + "loss": 4.048, + "step": 49536 + }, + { + "epoch": 0.49537, + "grad_norm": 1.191505007690368, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 49537 + }, + { + "epoch": 0.49538, + "grad_norm": 1.0083073883978004, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 49538 + }, + { + "epoch": 0.49539, + "grad_norm": 0.9862223379194267, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 49539 + }, + { + "epoch": 0.4954, + "grad_norm": 1.1205960772455756, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 49540 + }, + { + "epoch": 0.49541, + "grad_norm": 0.9799587366430164, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 49541 + }, + { + "epoch": 0.49542, + "grad_norm": 1.019695580386879, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 49542 + }, + { + "epoch": 0.49543, + "grad_norm": 1.053050433359199, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 49543 + }, + { + "epoch": 0.49544, + "grad_norm": 1.092160662263048, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 49544 + }, + { + "epoch": 0.49545, + "grad_norm": 0.9300353402811721, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 49545 + }, + { + "epoch": 0.49546, + "grad_norm": 0.9965796727858397, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 49546 + }, + { + "epoch": 0.49547, + "grad_norm": 1.1107012320585476, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 49547 + }, + { + "epoch": 0.49548, + "grad_norm": 0.9283186034836278, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 49548 + }, + { + "epoch": 0.49549, + "grad_norm": 1.1303485852312072, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 49549 + }, + { + "epoch": 0.4955, + "grad_norm": 1.2967063986591483, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 49550 + }, + { + "epoch": 0.49551, + "grad_norm": 0.9549175081421215, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 49551 + }, + { + "epoch": 0.49552, + "grad_norm": 1.1190361684354104, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 49552 + }, + { + "epoch": 0.49553, + "grad_norm": 0.8998103831807202, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 49553 + }, + { + "epoch": 0.49554, + "grad_norm": 0.7620457133571493, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 49554 + }, + { + "epoch": 0.49555, + "grad_norm": 1.3327059801817267, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 49555 + }, + { + "epoch": 0.49556, + "grad_norm": 1.3378602320668023, + "learning_rate": 0.003, + "loss": 4.054, + "step": 49556 + }, + { + "epoch": 0.49557, + "grad_norm": 0.8051602454233808, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 49557 + }, + { + "epoch": 0.49558, + "grad_norm": 0.8107879940965942, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 49558 + }, + { + "epoch": 0.49559, + "grad_norm": 0.9546115731357979, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 49559 + }, + { + "epoch": 0.4956, + "grad_norm": 1.1678400768370365, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 49560 + }, + { + "epoch": 0.49561, + "grad_norm": 0.999233064977031, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 49561 + }, + { + "epoch": 0.49562, + "grad_norm": 1.2365492013669344, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 49562 + }, + { + "epoch": 0.49563, + "grad_norm": 0.9164198155348434, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 49563 + }, + { + "epoch": 0.49564, + "grad_norm": 1.1479449478720096, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 49564 + }, + { + "epoch": 0.49565, + "grad_norm": 1.2073586612073681, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 49565 + }, + { + "epoch": 0.49566, + "grad_norm": 1.283723439062403, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 49566 + }, + { + "epoch": 0.49567, + "grad_norm": 1.1315054649719571, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 49567 + }, + { + "epoch": 0.49568, + "grad_norm": 1.1205680021370015, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 49568 + }, + { + "epoch": 0.49569, + "grad_norm": 1.0584779914555045, + "learning_rate": 0.003, + "loss": 4.052, + "step": 49569 + }, + { + "epoch": 0.4957, + "grad_norm": 1.0986381930116536, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 49570 + }, + { + "epoch": 0.49571, + "grad_norm": 1.1237931946615838, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 49571 + }, + { + "epoch": 0.49572, + "grad_norm": 0.955778823529564, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 49572 + }, + { + "epoch": 0.49573, + "grad_norm": 0.8702368521760232, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 49573 + }, + { + "epoch": 0.49574, + "grad_norm": 0.9104216719406693, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 49574 + }, + { + "epoch": 0.49575, + "grad_norm": 1.2423077951910875, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 49575 + }, + { + "epoch": 0.49576, + "grad_norm": 0.9321848937756503, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 49576 + }, + { + "epoch": 0.49577, + "grad_norm": 0.8891420427219295, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 49577 + }, + { + "epoch": 0.49578, + "grad_norm": 1.0257717406127635, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 49578 + }, + { + "epoch": 0.49579, + "grad_norm": 1.1726116882285889, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 49579 + }, + { + "epoch": 0.4958, + "grad_norm": 0.9774447876142305, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 49580 + }, + { + "epoch": 0.49581, + "grad_norm": 1.1415120469263282, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 49581 + }, + { + "epoch": 0.49582, + "grad_norm": 1.139463306563303, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 49582 + }, + { + "epoch": 0.49583, + "grad_norm": 0.9928311761998326, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 49583 + }, + { + "epoch": 0.49584, + "grad_norm": 1.145390326713825, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 49584 + }, + { + "epoch": 0.49585, + "grad_norm": 1.2375678175437888, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 49585 + }, + { + "epoch": 0.49586, + "grad_norm": 0.9292072619807983, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 49586 + }, + { + "epoch": 0.49587, + "grad_norm": 0.9772969365386606, + "learning_rate": 0.003, + "loss": 4.018, + "step": 49587 + }, + { + "epoch": 0.49588, + "grad_norm": 1.1075394657913584, + "learning_rate": 0.003, + "loss": 4.036, + "step": 49588 + }, + { + "epoch": 0.49589, + "grad_norm": 0.918585570495671, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 49589 + }, + { + "epoch": 0.4959, + "grad_norm": 1.0292227991426854, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 49590 + }, + { + "epoch": 0.49591, + "grad_norm": 1.1252786419957637, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 49591 + }, + { + "epoch": 0.49592, + "grad_norm": 1.1590869178599903, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 49592 + }, + { + "epoch": 0.49593, + "grad_norm": 1.1169142632440283, + "learning_rate": 0.003, + "loss": 4.004, + "step": 49593 + }, + { + "epoch": 0.49594, + "grad_norm": 1.011975164605672, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 49594 + }, + { + "epoch": 0.49595, + "grad_norm": 1.1148379966368647, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 49595 + }, + { + "epoch": 0.49596, + "grad_norm": 1.0168162903409415, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 49596 + }, + { + "epoch": 0.49597, + "grad_norm": 1.1341972533364857, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 49597 + }, + { + "epoch": 0.49598, + "grad_norm": 0.8571174465177315, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 49598 + }, + { + "epoch": 0.49599, + "grad_norm": 0.9481758571999349, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 49599 + }, + { + "epoch": 0.496, + "grad_norm": 1.2775084016132159, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 49600 + }, + { + "epoch": 0.49601, + "grad_norm": 0.8954558906677722, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 49601 + }, + { + "epoch": 0.49602, + "grad_norm": 0.8164928411558803, + "learning_rate": 0.003, + "loss": 4.036, + "step": 49602 + }, + { + "epoch": 0.49603, + "grad_norm": 1.0037327997932877, + "learning_rate": 0.003, + "loss": 4.032, + "step": 49603 + }, + { + "epoch": 0.49604, + "grad_norm": 1.3653177769151172, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 49604 + }, + { + "epoch": 0.49605, + "grad_norm": 0.7864548580694585, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 49605 + }, + { + "epoch": 0.49606, + "grad_norm": 0.9087476665300214, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 49606 + }, + { + "epoch": 0.49607, + "grad_norm": 1.0711546693837193, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 49607 + }, + { + "epoch": 0.49608, + "grad_norm": 1.140957962988931, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 49608 + }, + { + "epoch": 0.49609, + "grad_norm": 0.9364092014893411, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 49609 + }, + { + "epoch": 0.4961, + "grad_norm": 1.0441807095435367, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 49610 + }, + { + "epoch": 0.49611, + "grad_norm": 1.3144457173873276, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 49611 + }, + { + "epoch": 0.49612, + "grad_norm": 1.1417633842662513, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 49612 + }, + { + "epoch": 0.49613, + "grad_norm": 0.9501490684839682, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 49613 + }, + { + "epoch": 0.49614, + "grad_norm": 0.9064873123022475, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 49614 + }, + { + "epoch": 0.49615, + "grad_norm": 1.013459745284911, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 49615 + }, + { + "epoch": 0.49616, + "grad_norm": 1.1714205443890753, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 49616 + }, + { + "epoch": 0.49617, + "grad_norm": 1.0204923807211437, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 49617 + }, + { + "epoch": 0.49618, + "grad_norm": 1.1070163447691581, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 49618 + }, + { + "epoch": 0.49619, + "grad_norm": 1.0576856900263205, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 49619 + }, + { + "epoch": 0.4962, + "grad_norm": 1.1398347658369918, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 49620 + }, + { + "epoch": 0.49621, + "grad_norm": 1.2075212334677832, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 49621 + }, + { + "epoch": 0.49622, + "grad_norm": 1.045262920906611, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 49622 + }, + { + "epoch": 0.49623, + "grad_norm": 1.0551742126316062, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 49623 + }, + { + "epoch": 0.49624, + "grad_norm": 1.1433855613034554, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 49624 + }, + { + "epoch": 0.49625, + "grad_norm": 0.9152233041318805, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 49625 + }, + { + "epoch": 0.49626, + "grad_norm": 0.9809403427900282, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 49626 + }, + { + "epoch": 0.49627, + "grad_norm": 1.2199212535789568, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 49627 + }, + { + "epoch": 0.49628, + "grad_norm": 0.855703598316456, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 49628 + }, + { + "epoch": 0.49629, + "grad_norm": 0.8995123796919021, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 49629 + }, + { + "epoch": 0.4963, + "grad_norm": 1.31005316482914, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 49630 + }, + { + "epoch": 0.49631, + "grad_norm": 0.9484271299812407, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 49631 + }, + { + "epoch": 0.49632, + "grad_norm": 1.0186623172989488, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 49632 + }, + { + "epoch": 0.49633, + "grad_norm": 1.2036768472990536, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 49633 + }, + { + "epoch": 0.49634, + "grad_norm": 0.9987028271020612, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 49634 + }, + { + "epoch": 0.49635, + "grad_norm": 1.1912797661813614, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 49635 + }, + { + "epoch": 0.49636, + "grad_norm": 1.0132869200404306, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 49636 + }, + { + "epoch": 0.49637, + "grad_norm": 1.1061691798602857, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 49637 + }, + { + "epoch": 0.49638, + "grad_norm": 1.0465323705070315, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 49638 + }, + { + "epoch": 0.49639, + "grad_norm": 1.033240033767283, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 49639 + }, + { + "epoch": 0.4964, + "grad_norm": 1.1178500566564304, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 49640 + }, + { + "epoch": 0.49641, + "grad_norm": 0.999639351263027, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 49641 + }, + { + "epoch": 0.49642, + "grad_norm": 1.1400377879065564, + "learning_rate": 0.003, + "loss": 4.013, + "step": 49642 + }, + { + "epoch": 0.49643, + "grad_norm": 0.9596990018349181, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 49643 + }, + { + "epoch": 0.49644, + "grad_norm": 1.1479311958656024, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 49644 + }, + { + "epoch": 0.49645, + "grad_norm": 0.9088926436126541, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 49645 + }, + { + "epoch": 0.49646, + "grad_norm": 0.9678005592538285, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 49646 + }, + { + "epoch": 0.49647, + "grad_norm": 1.2293739612693708, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 49647 + }, + { + "epoch": 0.49648, + "grad_norm": 0.8647578414093389, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 49648 + }, + { + "epoch": 0.49649, + "grad_norm": 0.9497416768566925, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 49649 + }, + { + "epoch": 0.4965, + "grad_norm": 1.1286464383197203, + "learning_rate": 0.003, + "loss": 4.039, + "step": 49650 + }, + { + "epoch": 0.49651, + "grad_norm": 1.129678174555683, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 49651 + }, + { + "epoch": 0.49652, + "grad_norm": 1.2314981169643469, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 49652 + }, + { + "epoch": 0.49653, + "grad_norm": 1.1823247027779908, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 49653 + }, + { + "epoch": 0.49654, + "grad_norm": 1.13507996853683, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 49654 + }, + { + "epoch": 0.49655, + "grad_norm": 0.9826778965390449, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 49655 + }, + { + "epoch": 0.49656, + "grad_norm": 1.193407185455883, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 49656 + }, + { + "epoch": 0.49657, + "grad_norm": 1.036394099396497, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 49657 + }, + { + "epoch": 0.49658, + "grad_norm": 1.1826206780381407, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 49658 + }, + { + "epoch": 0.49659, + "grad_norm": 1.0163488034428096, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 49659 + }, + { + "epoch": 0.4966, + "grad_norm": 1.251771391958464, + "learning_rate": 0.003, + "loss": 4.031, + "step": 49660 + }, + { + "epoch": 0.49661, + "grad_norm": 0.9593836811033258, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 49661 + }, + { + "epoch": 0.49662, + "grad_norm": 1.149274477342746, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 49662 + }, + { + "epoch": 0.49663, + "grad_norm": 1.0083107617694802, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 49663 + }, + { + "epoch": 0.49664, + "grad_norm": 1.2134787762305608, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 49664 + }, + { + "epoch": 0.49665, + "grad_norm": 0.9385616650326712, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 49665 + }, + { + "epoch": 0.49666, + "grad_norm": 0.9549648423020882, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 49666 + }, + { + "epoch": 0.49667, + "grad_norm": 1.0557562767607898, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 49667 + }, + { + "epoch": 0.49668, + "grad_norm": 1.0596355376460014, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 49668 + }, + { + "epoch": 0.49669, + "grad_norm": 0.954407334814445, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 49669 + }, + { + "epoch": 0.4967, + "grad_norm": 1.0616504205244386, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 49670 + }, + { + "epoch": 0.49671, + "grad_norm": 1.1120172832885438, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 49671 + }, + { + "epoch": 0.49672, + "grad_norm": 0.9857960964704445, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 49672 + }, + { + "epoch": 0.49673, + "grad_norm": 1.0831523339076998, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 49673 + }, + { + "epoch": 0.49674, + "grad_norm": 1.0425876974263828, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 49674 + }, + { + "epoch": 0.49675, + "grad_norm": 0.9593616611978678, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 49675 + }, + { + "epoch": 0.49676, + "grad_norm": 1.1866588486210148, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 49676 + }, + { + "epoch": 0.49677, + "grad_norm": 0.8488206901324277, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 49677 + }, + { + "epoch": 0.49678, + "grad_norm": 0.9158832381798342, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 49678 + }, + { + "epoch": 0.49679, + "grad_norm": 1.0864772438809527, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 49679 + }, + { + "epoch": 0.4968, + "grad_norm": 0.9415995559281816, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 49680 + }, + { + "epoch": 0.49681, + "grad_norm": 0.955438048937139, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 49681 + }, + { + "epoch": 0.49682, + "grad_norm": 1.2739520673535325, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 49682 + }, + { + "epoch": 0.49683, + "grad_norm": 1.0196859600086465, + "learning_rate": 0.003, + "loss": 4.058, + "step": 49683 + }, + { + "epoch": 0.49684, + "grad_norm": 1.2365143435962522, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 49684 + }, + { + "epoch": 0.49685, + "grad_norm": 1.1344250197399472, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 49685 + }, + { + "epoch": 0.49686, + "grad_norm": 1.3822276725566183, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 49686 + }, + { + "epoch": 0.49687, + "grad_norm": 0.813575169339493, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 49687 + }, + { + "epoch": 0.49688, + "grad_norm": 0.8411152129059359, + "learning_rate": 0.003, + "loss": 4.022, + "step": 49688 + }, + { + "epoch": 0.49689, + "grad_norm": 1.1660308906536834, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 49689 + }, + { + "epoch": 0.4969, + "grad_norm": 1.2717455717301576, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 49690 + }, + { + "epoch": 0.49691, + "grad_norm": 0.8917646694268575, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 49691 + }, + { + "epoch": 0.49692, + "grad_norm": 1.0004700743983261, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 49692 + }, + { + "epoch": 0.49693, + "grad_norm": 1.1615141252203949, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 49693 + }, + { + "epoch": 0.49694, + "grad_norm": 1.1500089718313324, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 49694 + }, + { + "epoch": 0.49695, + "grad_norm": 1.1378093162498393, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 49695 + }, + { + "epoch": 0.49696, + "grad_norm": 1.2016833740596216, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 49696 + }, + { + "epoch": 0.49697, + "grad_norm": 0.9330759724202541, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 49697 + }, + { + "epoch": 0.49698, + "grad_norm": 0.953972332744415, + "learning_rate": 0.003, + "loss": 4.016, + "step": 49698 + }, + { + "epoch": 0.49699, + "grad_norm": 1.0516267576035578, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 49699 + }, + { + "epoch": 0.497, + "grad_norm": 1.0440727204146403, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 49700 + }, + { + "epoch": 0.49701, + "grad_norm": 0.985076635891579, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 49701 + }, + { + "epoch": 0.49702, + "grad_norm": 1.043479178444512, + "learning_rate": 0.003, + "loss": 4.026, + "step": 49702 + }, + { + "epoch": 0.49703, + "grad_norm": 1.0265904715905696, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 49703 + }, + { + "epoch": 0.49704, + "grad_norm": 1.2235033726265707, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 49704 + }, + { + "epoch": 0.49705, + "grad_norm": 0.9515689104208639, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 49705 + }, + { + "epoch": 0.49706, + "grad_norm": 1.0434096616812412, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 49706 + }, + { + "epoch": 0.49707, + "grad_norm": 1.4796914537056836, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 49707 + }, + { + "epoch": 0.49708, + "grad_norm": 0.675322818631768, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 49708 + }, + { + "epoch": 0.49709, + "grad_norm": 0.8556011020303496, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 49709 + }, + { + "epoch": 0.4971, + "grad_norm": 1.0487057507224808, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 49710 + }, + { + "epoch": 0.49711, + "grad_norm": 0.8348218619353129, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 49711 + }, + { + "epoch": 0.49712, + "grad_norm": 1.1353304026472708, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 49712 + }, + { + "epoch": 0.49713, + "grad_norm": 1.0340752002741593, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 49713 + }, + { + "epoch": 0.49714, + "grad_norm": 0.9567675546136585, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 49714 + }, + { + "epoch": 0.49715, + "grad_norm": 1.1655967038529385, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 49715 + }, + { + "epoch": 0.49716, + "grad_norm": 1.0800190949975166, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 49716 + }, + { + "epoch": 0.49717, + "grad_norm": 1.1653381222512522, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 49717 + }, + { + "epoch": 0.49718, + "grad_norm": 1.098654707607607, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 49718 + }, + { + "epoch": 0.49719, + "grad_norm": 1.2552077336357488, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 49719 + }, + { + "epoch": 0.4972, + "grad_norm": 1.0463647746034523, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 49720 + }, + { + "epoch": 0.49721, + "grad_norm": 1.0376422686007905, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 49721 + }, + { + "epoch": 0.49722, + "grad_norm": 1.0395067323100542, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 49722 + }, + { + "epoch": 0.49723, + "grad_norm": 1.0536700121611433, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 49723 + }, + { + "epoch": 0.49724, + "grad_norm": 1.2316020053815517, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 49724 + }, + { + "epoch": 0.49725, + "grad_norm": 0.8683072830643318, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 49725 + }, + { + "epoch": 0.49726, + "grad_norm": 0.8759839593491109, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 49726 + }, + { + "epoch": 0.49727, + "grad_norm": 0.9914429848837595, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 49727 + }, + { + "epoch": 0.49728, + "grad_norm": 1.118984283218648, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 49728 + }, + { + "epoch": 0.49729, + "grad_norm": 1.1166062581204537, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 49729 + }, + { + "epoch": 0.4973, + "grad_norm": 0.8597835323637608, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 49730 + }, + { + "epoch": 0.49731, + "grad_norm": 0.8163823981785892, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 49731 + }, + { + "epoch": 0.49732, + "grad_norm": 1.0889692276398935, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 49732 + }, + { + "epoch": 0.49733, + "grad_norm": 1.419713208340765, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 49733 + }, + { + "epoch": 0.49734, + "grad_norm": 0.8021935876878334, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 49734 + }, + { + "epoch": 0.49735, + "grad_norm": 0.7705713646292169, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 49735 + }, + { + "epoch": 0.49736, + "grad_norm": 0.8463272017951418, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 49736 + }, + { + "epoch": 0.49737, + "grad_norm": 1.2111947059955828, + "learning_rate": 0.003, + "loss": 4.033, + "step": 49737 + }, + { + "epoch": 0.49738, + "grad_norm": 1.236650248208658, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 49738 + }, + { + "epoch": 0.49739, + "grad_norm": 0.9723345054462602, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 49739 + }, + { + "epoch": 0.4974, + "grad_norm": 0.960646287710632, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 49740 + }, + { + "epoch": 0.49741, + "grad_norm": 1.1227905098686304, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 49741 + }, + { + "epoch": 0.49742, + "grad_norm": 0.9047514710392711, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 49742 + }, + { + "epoch": 0.49743, + "grad_norm": 0.8904906618221012, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 49743 + }, + { + "epoch": 0.49744, + "grad_norm": 1.1346041548891685, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 49744 + }, + { + "epoch": 0.49745, + "grad_norm": 0.8331176781517297, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 49745 + }, + { + "epoch": 0.49746, + "grad_norm": 1.1189784952891084, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 49746 + }, + { + "epoch": 0.49747, + "grad_norm": 1.4355070209741538, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 49747 + }, + { + "epoch": 0.49748, + "grad_norm": 0.674265608045974, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 49748 + }, + { + "epoch": 0.49749, + "grad_norm": 0.9067919736840768, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 49749 + }, + { + "epoch": 0.4975, + "grad_norm": 1.3497552522313585, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 49750 + }, + { + "epoch": 0.49751, + "grad_norm": 0.9578832398874906, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 49751 + }, + { + "epoch": 0.49752, + "grad_norm": 1.1675601250308305, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 49752 + }, + { + "epoch": 0.49753, + "grad_norm": 0.9614830208054452, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 49753 + }, + { + "epoch": 0.49754, + "grad_norm": 0.9736266925150258, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 49754 + }, + { + "epoch": 0.49755, + "grad_norm": 1.1535778714881404, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 49755 + }, + { + "epoch": 0.49756, + "grad_norm": 0.9485482011872374, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 49756 + }, + { + "epoch": 0.49757, + "grad_norm": 1.1752046889149548, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 49757 + }, + { + "epoch": 0.49758, + "grad_norm": 1.1199785171664471, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 49758 + }, + { + "epoch": 0.49759, + "grad_norm": 0.9410111286599888, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 49759 + }, + { + "epoch": 0.4976, + "grad_norm": 1.1355337619814925, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 49760 + }, + { + "epoch": 0.49761, + "grad_norm": 1.0868807199955577, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 49761 + }, + { + "epoch": 0.49762, + "grad_norm": 1.0245004673822014, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 49762 + }, + { + "epoch": 0.49763, + "grad_norm": 1.0552259034513378, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 49763 + }, + { + "epoch": 0.49764, + "grad_norm": 1.083535108722862, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 49764 + }, + { + "epoch": 0.49765, + "grad_norm": 1.0673284595580645, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 49765 + }, + { + "epoch": 0.49766, + "grad_norm": 0.9640570526390049, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 49766 + }, + { + "epoch": 0.49767, + "grad_norm": 1.071424533731562, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 49767 + }, + { + "epoch": 0.49768, + "grad_norm": 1.1089824945887128, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 49768 + }, + { + "epoch": 0.49769, + "grad_norm": 0.9836503246742354, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 49769 + }, + { + "epoch": 0.4977, + "grad_norm": 1.0264458088654553, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 49770 + }, + { + "epoch": 0.49771, + "grad_norm": 1.3203535342417734, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 49771 + }, + { + "epoch": 0.49772, + "grad_norm": 0.924577579001054, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 49772 + }, + { + "epoch": 0.49773, + "grad_norm": 1.1925830229967336, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 49773 + }, + { + "epoch": 0.49774, + "grad_norm": 1.2089857071087644, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 49774 + }, + { + "epoch": 0.49775, + "grad_norm": 1.0763651404975845, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 49775 + }, + { + "epoch": 0.49776, + "grad_norm": 1.1114048240479646, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 49776 + }, + { + "epoch": 0.49777, + "grad_norm": 0.932565429785076, + "learning_rate": 0.003, + "loss": 4.064, + "step": 49777 + }, + { + "epoch": 0.49778, + "grad_norm": 1.0950154750812298, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 49778 + }, + { + "epoch": 0.49779, + "grad_norm": 1.05373042410669, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 49779 + }, + { + "epoch": 0.4978, + "grad_norm": 0.9995961790177914, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 49780 + }, + { + "epoch": 0.49781, + "grad_norm": 1.0973194219132412, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 49781 + }, + { + "epoch": 0.49782, + "grad_norm": 1.1464140827351676, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 49782 + }, + { + "epoch": 0.49783, + "grad_norm": 1.0003609966372582, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 49783 + }, + { + "epoch": 0.49784, + "grad_norm": 1.0823952459626511, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 49784 + }, + { + "epoch": 0.49785, + "grad_norm": 1.192946274718134, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 49785 + }, + { + "epoch": 0.49786, + "grad_norm": 1.323620320660141, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 49786 + }, + { + "epoch": 0.49787, + "grad_norm": 0.8928728255918457, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 49787 + }, + { + "epoch": 0.49788, + "grad_norm": 0.9944036953527924, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 49788 + }, + { + "epoch": 0.49789, + "grad_norm": 1.2316035214822558, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 49789 + }, + { + "epoch": 0.4979, + "grad_norm": 1.111924162093948, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 49790 + }, + { + "epoch": 0.49791, + "grad_norm": 1.0108601114566125, + "learning_rate": 0.003, + "loss": 4.019, + "step": 49791 + }, + { + "epoch": 0.49792, + "grad_norm": 1.0916531797508344, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 49792 + }, + { + "epoch": 0.49793, + "grad_norm": 1.0852298478039928, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 49793 + }, + { + "epoch": 0.49794, + "grad_norm": 1.0977968172140697, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 49794 + }, + { + "epoch": 0.49795, + "grad_norm": 1.051661607142881, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 49795 + }, + { + "epoch": 0.49796, + "grad_norm": 1.0767331360823305, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 49796 + }, + { + "epoch": 0.49797, + "grad_norm": 0.9886669013563881, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 49797 + }, + { + "epoch": 0.49798, + "grad_norm": 1.0443897733048757, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 49798 + }, + { + "epoch": 0.49799, + "grad_norm": 1.2238093884917867, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 49799 + }, + { + "epoch": 0.498, + "grad_norm": 0.8616885223103984, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 49800 + }, + { + "epoch": 0.49801, + "grad_norm": 0.9642223300125562, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 49801 + }, + { + "epoch": 0.49802, + "grad_norm": 1.4509370338532819, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 49802 + }, + { + "epoch": 0.49803, + "grad_norm": 1.0095442168789643, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 49803 + }, + { + "epoch": 0.49804, + "grad_norm": 1.2872319083137678, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 49804 + }, + { + "epoch": 0.49805, + "grad_norm": 1.0545525056413123, + "learning_rate": 0.003, + "loss": 4.046, + "step": 49805 + }, + { + "epoch": 0.49806, + "grad_norm": 1.0371847187963528, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 49806 + }, + { + "epoch": 0.49807, + "grad_norm": 1.1266779392092368, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 49807 + }, + { + "epoch": 0.49808, + "grad_norm": 0.9578514078482576, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 49808 + }, + { + "epoch": 0.49809, + "grad_norm": 1.019209450448921, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 49809 + }, + { + "epoch": 0.4981, + "grad_norm": 1.1688788752766925, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 49810 + }, + { + "epoch": 0.49811, + "grad_norm": 0.9763184644876561, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 49811 + }, + { + "epoch": 0.49812, + "grad_norm": 0.9454490350735182, + "learning_rate": 0.003, + "loss": 4.059, + "step": 49812 + }, + { + "epoch": 0.49813, + "grad_norm": 1.1089361948897158, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 49813 + }, + { + "epoch": 0.49814, + "grad_norm": 1.0092373493349642, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 49814 + }, + { + "epoch": 0.49815, + "grad_norm": 1.2001142578201203, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 49815 + }, + { + "epoch": 0.49816, + "grad_norm": 1.2593828814482275, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 49816 + }, + { + "epoch": 0.49817, + "grad_norm": 1.0366674349812668, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 49817 + }, + { + "epoch": 0.49818, + "grad_norm": 1.088316734188058, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 49818 + }, + { + "epoch": 0.49819, + "grad_norm": 1.1248286839790265, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 49819 + }, + { + "epoch": 0.4982, + "grad_norm": 1.1639081844165793, + "learning_rate": 0.003, + "loss": 4.007, + "step": 49820 + }, + { + "epoch": 0.49821, + "grad_norm": 1.0298580099266252, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 49821 + }, + { + "epoch": 0.49822, + "grad_norm": 1.020227359237908, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 49822 + }, + { + "epoch": 0.49823, + "grad_norm": 0.8922852269089055, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 49823 + }, + { + "epoch": 0.49824, + "grad_norm": 0.8898325819928065, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 49824 + }, + { + "epoch": 0.49825, + "grad_norm": 1.0100461863457249, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 49825 + }, + { + "epoch": 0.49826, + "grad_norm": 1.2166850629870216, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 49826 + }, + { + "epoch": 0.49827, + "grad_norm": 0.7718139581043718, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 49827 + }, + { + "epoch": 0.49828, + "grad_norm": 0.8881071185132858, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 49828 + }, + { + "epoch": 0.49829, + "grad_norm": 1.0110524511567511, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 49829 + }, + { + "epoch": 0.4983, + "grad_norm": 1.3245079117384246, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 49830 + }, + { + "epoch": 0.49831, + "grad_norm": 0.725664071524078, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 49831 + }, + { + "epoch": 0.49832, + "grad_norm": 0.8117564428518309, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 49832 + }, + { + "epoch": 0.49833, + "grad_norm": 0.8871580712194491, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 49833 + }, + { + "epoch": 0.49834, + "grad_norm": 1.1817386335535822, + "learning_rate": 0.003, + "loss": 4.047, + "step": 49834 + }, + { + "epoch": 0.49835, + "grad_norm": 1.1057509521183397, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 49835 + }, + { + "epoch": 0.49836, + "grad_norm": 0.9641912592780104, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 49836 + }, + { + "epoch": 0.49837, + "grad_norm": 1.1934809833655438, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 49837 + }, + { + "epoch": 0.49838, + "grad_norm": 1.3405822374234768, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 49838 + }, + { + "epoch": 0.49839, + "grad_norm": 0.8772458939914964, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 49839 + }, + { + "epoch": 0.4984, + "grad_norm": 0.7657741725618057, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 49840 + }, + { + "epoch": 0.49841, + "grad_norm": 0.8773323350627292, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 49841 + }, + { + "epoch": 0.49842, + "grad_norm": 1.268025302833992, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 49842 + }, + { + "epoch": 0.49843, + "grad_norm": 1.1001180216089148, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 49843 + }, + { + "epoch": 0.49844, + "grad_norm": 0.9622998462883092, + "learning_rate": 0.003, + "loss": 4.044, + "step": 49844 + }, + { + "epoch": 0.49845, + "grad_norm": 1.0250411158029613, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 49845 + }, + { + "epoch": 0.49846, + "grad_norm": 1.0682877986294204, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 49846 + }, + { + "epoch": 0.49847, + "grad_norm": 0.9307703753320157, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 49847 + }, + { + "epoch": 0.49848, + "grad_norm": 0.937666303623847, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 49848 + }, + { + "epoch": 0.49849, + "grad_norm": 1.324564448665457, + "learning_rate": 0.003, + "loss": 4.031, + "step": 49849 + }, + { + "epoch": 0.4985, + "grad_norm": 0.9424041481471734, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 49850 + }, + { + "epoch": 0.49851, + "grad_norm": 1.13129852282287, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 49851 + }, + { + "epoch": 0.49852, + "grad_norm": 1.037784337948939, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 49852 + }, + { + "epoch": 0.49853, + "grad_norm": 1.2775822302555795, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 49853 + }, + { + "epoch": 0.49854, + "grad_norm": 0.9095171330715371, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 49854 + }, + { + "epoch": 0.49855, + "grad_norm": 0.9673072238411989, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 49855 + }, + { + "epoch": 0.49856, + "grad_norm": 1.2168798218331598, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 49856 + }, + { + "epoch": 0.49857, + "grad_norm": 1.0168625659662525, + "learning_rate": 0.003, + "loss": 4.014, + "step": 49857 + }, + { + "epoch": 0.49858, + "grad_norm": 1.221440985700905, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 49858 + }, + { + "epoch": 0.49859, + "grad_norm": 1.0651215820537019, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 49859 + }, + { + "epoch": 0.4986, + "grad_norm": 1.0822937351808783, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 49860 + }, + { + "epoch": 0.49861, + "grad_norm": 1.1238826861310338, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 49861 + }, + { + "epoch": 0.49862, + "grad_norm": 1.0663528094951085, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 49862 + }, + { + "epoch": 0.49863, + "grad_norm": 1.1572054299298085, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 49863 + }, + { + "epoch": 0.49864, + "grad_norm": 0.885945675435199, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 49864 + }, + { + "epoch": 0.49865, + "grad_norm": 0.8325282548527716, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 49865 + }, + { + "epoch": 0.49866, + "grad_norm": 0.8472941878931058, + "learning_rate": 0.003, + "loss": 4.036, + "step": 49866 + }, + { + "epoch": 0.49867, + "grad_norm": 0.9111483267439591, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 49867 + }, + { + "epoch": 0.49868, + "grad_norm": 1.0403730744104671, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 49868 + }, + { + "epoch": 0.49869, + "grad_norm": 1.1171782049227637, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 49869 + }, + { + "epoch": 0.4987, + "grad_norm": 1.018856192164235, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 49870 + }, + { + "epoch": 0.49871, + "grad_norm": 1.351305214636442, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 49871 + }, + { + "epoch": 0.49872, + "grad_norm": 0.9132157279546456, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 49872 + }, + { + "epoch": 0.49873, + "grad_norm": 0.7531104509053733, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 49873 + }, + { + "epoch": 0.49874, + "grad_norm": 0.9009189237870333, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 49874 + }, + { + "epoch": 0.49875, + "grad_norm": 1.1572657197067115, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 49875 + }, + { + "epoch": 0.49876, + "grad_norm": 1.1169924709871852, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 49876 + }, + { + "epoch": 0.49877, + "grad_norm": 1.0622397780988921, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 49877 + }, + { + "epoch": 0.49878, + "grad_norm": 1.153979647277428, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 49878 + }, + { + "epoch": 0.49879, + "grad_norm": 1.0185402109571382, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 49879 + }, + { + "epoch": 0.4988, + "grad_norm": 1.0685886950979635, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 49880 + }, + { + "epoch": 0.49881, + "grad_norm": 1.1548332555063467, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 49881 + }, + { + "epoch": 0.49882, + "grad_norm": 0.9831432792759929, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 49882 + }, + { + "epoch": 0.49883, + "grad_norm": 1.0629154928674764, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 49883 + }, + { + "epoch": 0.49884, + "grad_norm": 1.1633897199874657, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 49884 + }, + { + "epoch": 0.49885, + "grad_norm": 1.0016614434833242, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 49885 + }, + { + "epoch": 0.49886, + "grad_norm": 1.2036193325179045, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 49886 + }, + { + "epoch": 0.49887, + "grad_norm": 1.109760794748179, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 49887 + }, + { + "epoch": 0.49888, + "grad_norm": 1.008668172677349, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 49888 + }, + { + "epoch": 0.49889, + "grad_norm": 1.1187946538938187, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 49889 + }, + { + "epoch": 0.4989, + "grad_norm": 1.1505349840862265, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 49890 + }, + { + "epoch": 0.49891, + "grad_norm": 1.1218462310152704, + "learning_rate": 0.003, + "loss": 4.047, + "step": 49891 + }, + { + "epoch": 0.49892, + "grad_norm": 1.0383780850642463, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 49892 + }, + { + "epoch": 0.49893, + "grad_norm": 0.8511856418473984, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 49893 + }, + { + "epoch": 0.49894, + "grad_norm": 1.0484895088747894, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 49894 + }, + { + "epoch": 0.49895, + "grad_norm": 1.2373116236453237, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 49895 + }, + { + "epoch": 0.49896, + "grad_norm": 1.025516202947874, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 49896 + }, + { + "epoch": 0.49897, + "grad_norm": 1.1843453048202797, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 49897 + }, + { + "epoch": 0.49898, + "grad_norm": 0.947185534757458, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 49898 + }, + { + "epoch": 0.49899, + "grad_norm": 0.9348920872094152, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 49899 + }, + { + "epoch": 0.499, + "grad_norm": 1.0860257519572598, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 49900 + }, + { + "epoch": 0.49901, + "grad_norm": 1.0289882652444484, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 49901 + }, + { + "epoch": 0.49902, + "grad_norm": 1.1178099258406593, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 49902 + }, + { + "epoch": 0.49903, + "grad_norm": 1.159428866219153, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 49903 + }, + { + "epoch": 0.49904, + "grad_norm": 1.259459381194775, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 49904 + }, + { + "epoch": 0.49905, + "grad_norm": 1.0216639817713908, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 49905 + }, + { + "epoch": 0.49906, + "grad_norm": 1.0661716729261295, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 49906 + }, + { + "epoch": 0.49907, + "grad_norm": 0.997339291413772, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 49907 + }, + { + "epoch": 0.49908, + "grad_norm": 1.0611065750544488, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 49908 + }, + { + "epoch": 0.49909, + "grad_norm": 1.246311874271977, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 49909 + }, + { + "epoch": 0.4991, + "grad_norm": 0.9252483420657944, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 49910 + }, + { + "epoch": 0.49911, + "grad_norm": 0.8354642312266808, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 49911 + }, + { + "epoch": 0.49912, + "grad_norm": 0.9425052234116781, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 49912 + }, + { + "epoch": 0.49913, + "grad_norm": 1.2624138705058472, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 49913 + }, + { + "epoch": 0.49914, + "grad_norm": 0.7987656698121401, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 49914 + }, + { + "epoch": 0.49915, + "grad_norm": 0.8302142222103096, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 49915 + }, + { + "epoch": 0.49916, + "grad_norm": 1.047268569917543, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 49916 + }, + { + "epoch": 0.49917, + "grad_norm": 1.111524851491636, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 49917 + }, + { + "epoch": 0.49918, + "grad_norm": 0.986511004084511, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 49918 + }, + { + "epoch": 0.49919, + "grad_norm": 1.3191901640022552, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 49919 + }, + { + "epoch": 0.4992, + "grad_norm": 1.0239062561044872, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 49920 + }, + { + "epoch": 0.49921, + "grad_norm": 1.1775722986309665, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 49921 + }, + { + "epoch": 0.49922, + "grad_norm": 1.0088038345666128, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 49922 + }, + { + "epoch": 0.49923, + "grad_norm": 1.0991521313032866, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 49923 + }, + { + "epoch": 0.49924, + "grad_norm": 1.0266258834005433, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 49924 + }, + { + "epoch": 0.49925, + "grad_norm": 1.0704028879488423, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 49925 + }, + { + "epoch": 0.49926, + "grad_norm": 0.8637049874328842, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 49926 + }, + { + "epoch": 0.49927, + "grad_norm": 0.7167898626231981, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 49927 + }, + { + "epoch": 0.49928, + "grad_norm": 0.793392573621242, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 49928 + }, + { + "epoch": 0.49929, + "grad_norm": 1.108693522426604, + "learning_rate": 0.003, + "loss": 4.019, + "step": 49929 + }, + { + "epoch": 0.4993, + "grad_norm": 1.063342844091234, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 49930 + }, + { + "epoch": 0.49931, + "grad_norm": 1.175245926331794, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 49931 + }, + { + "epoch": 0.49932, + "grad_norm": 1.0582214496753184, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 49932 + }, + { + "epoch": 0.49933, + "grad_norm": 0.9769642048425827, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 49933 + }, + { + "epoch": 0.49934, + "grad_norm": 1.3888962274527548, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 49934 + }, + { + "epoch": 0.49935, + "grad_norm": 1.080335045779711, + "learning_rate": 0.003, + "loss": 4.024, + "step": 49935 + }, + { + "epoch": 0.49936, + "grad_norm": 0.911039004047956, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 49936 + }, + { + "epoch": 0.49937, + "grad_norm": 0.9232414583193357, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 49937 + }, + { + "epoch": 0.49938, + "grad_norm": 1.0489555929428585, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 49938 + }, + { + "epoch": 0.49939, + "grad_norm": 1.0726462018431653, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 49939 + }, + { + "epoch": 0.4994, + "grad_norm": 1.1512611046064911, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 49940 + }, + { + "epoch": 0.49941, + "grad_norm": 0.9377917564369648, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 49941 + }, + { + "epoch": 0.49942, + "grad_norm": 1.0260445744235156, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 49942 + }, + { + "epoch": 0.49943, + "grad_norm": 1.1166292978513477, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 49943 + }, + { + "epoch": 0.49944, + "grad_norm": 1.0354842638222204, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 49944 + }, + { + "epoch": 0.49945, + "grad_norm": 1.0602012664624658, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 49945 + }, + { + "epoch": 0.49946, + "grad_norm": 1.0464064195131066, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 49946 + }, + { + "epoch": 0.49947, + "grad_norm": 1.116717673654169, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 49947 + }, + { + "epoch": 0.49948, + "grad_norm": 1.1162938827372186, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 49948 + }, + { + "epoch": 0.49949, + "grad_norm": 1.281755713173269, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 49949 + }, + { + "epoch": 0.4995, + "grad_norm": 1.1806905758388953, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 49950 + }, + { + "epoch": 0.49951, + "grad_norm": 1.0936269189790995, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 49951 + }, + { + "epoch": 0.49952, + "grad_norm": 1.0643851664019575, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 49952 + }, + { + "epoch": 0.49953, + "grad_norm": 1.0271624551211576, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 49953 + }, + { + "epoch": 0.49954, + "grad_norm": 1.1507282418200697, + "learning_rate": 0.003, + "loss": 4.047, + "step": 49954 + }, + { + "epoch": 0.49955, + "grad_norm": 1.0084647187640343, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 49955 + }, + { + "epoch": 0.49956, + "grad_norm": 1.1087797380338067, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 49956 + }, + { + "epoch": 0.49957, + "grad_norm": 0.916812302947403, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 49957 + }, + { + "epoch": 0.49958, + "grad_norm": 1.0214369663665466, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 49958 + }, + { + "epoch": 0.49959, + "grad_norm": 1.178684200023995, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 49959 + }, + { + "epoch": 0.4996, + "grad_norm": 1.1216141548492227, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 49960 + }, + { + "epoch": 0.49961, + "grad_norm": 0.9181442031459335, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 49961 + }, + { + "epoch": 0.49962, + "grad_norm": 0.9140072451596907, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 49962 + }, + { + "epoch": 0.49963, + "grad_norm": 1.045723491052052, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 49963 + }, + { + "epoch": 0.49964, + "grad_norm": 1.1932489071969055, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 49964 + }, + { + "epoch": 0.49965, + "grad_norm": 0.9785473211671507, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 49965 + }, + { + "epoch": 0.49966, + "grad_norm": 1.3026536838020686, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 49966 + }, + { + "epoch": 0.49967, + "grad_norm": 1.0524126920013899, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 49967 + }, + { + "epoch": 0.49968, + "grad_norm": 0.9185957162662413, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 49968 + }, + { + "epoch": 0.49969, + "grad_norm": 0.7999381813225376, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 49969 + }, + { + "epoch": 0.4997, + "grad_norm": 0.8385303278755036, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 49970 + }, + { + "epoch": 0.49971, + "grad_norm": 0.9006952920979011, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 49971 + }, + { + "epoch": 0.49972, + "grad_norm": 1.2088395477470708, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 49972 + }, + { + "epoch": 0.49973, + "grad_norm": 1.04536147821133, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 49973 + }, + { + "epoch": 0.49974, + "grad_norm": 1.1016252186445545, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 49974 + }, + { + "epoch": 0.49975, + "grad_norm": 1.0370184355416896, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 49975 + }, + { + "epoch": 0.49976, + "grad_norm": 1.0280675551359106, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 49976 + }, + { + "epoch": 0.49977, + "grad_norm": 0.9804404880592086, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 49977 + }, + { + "epoch": 0.49978, + "grad_norm": 0.9532307837868296, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 49978 + }, + { + "epoch": 0.49979, + "grad_norm": 1.0342350293347984, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 49979 + }, + { + "epoch": 0.4998, + "grad_norm": 1.0932755381894395, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 49980 + }, + { + "epoch": 0.49981, + "grad_norm": 1.1132902396667095, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 49981 + }, + { + "epoch": 0.49982, + "grad_norm": 1.3449726127851558, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 49982 + }, + { + "epoch": 0.49983, + "grad_norm": 0.9322224825839271, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 49983 + }, + { + "epoch": 0.49984, + "grad_norm": 1.1615108027064691, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 49984 + }, + { + "epoch": 0.49985, + "grad_norm": 1.2410150569472442, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 49985 + }, + { + "epoch": 0.49986, + "grad_norm": 0.9812230884629507, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 49986 + }, + { + "epoch": 0.49987, + "grad_norm": 1.3408843761921778, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 49987 + }, + { + "epoch": 0.49988, + "grad_norm": 0.881618288449289, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 49988 + }, + { + "epoch": 0.49989, + "grad_norm": 0.8195012697787643, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 49989 + }, + { + "epoch": 0.4999, + "grad_norm": 0.8725418672867923, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 49990 + }, + { + "epoch": 0.49991, + "grad_norm": 0.8552496199447442, + "learning_rate": 0.003, + "loss": 4.03, + "step": 49991 + }, + { + "epoch": 0.49992, + "grad_norm": 0.8912456280427871, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 49992 + }, + { + "epoch": 0.49993, + "grad_norm": 0.9832785914523948, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 49993 + }, + { + "epoch": 0.49994, + "grad_norm": 1.2929786374362713, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 49994 + }, + { + "epoch": 0.49995, + "grad_norm": 1.1129112018347949, + "learning_rate": 0.003, + "loss": 4.014, + "step": 49995 + }, + { + "epoch": 0.49996, + "grad_norm": 1.0920136945647398, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 49996 + }, + { + "epoch": 0.49997, + "grad_norm": 1.0161749471587214, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 49997 + }, + { + "epoch": 0.49998, + "grad_norm": 1.1498554199628819, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 49998 + }, + { + "epoch": 0.49999, + "grad_norm": 1.1660135547200403, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 49999 + }, + { + "epoch": 0.5, + "grad_norm": 1.123061768915059, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 50000 + }, + { + "epoch": 0.50001, + "grad_norm": 1.014395397616107, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 50001 + }, + { + "epoch": 0.50002, + "grad_norm": 1.289212998770998, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 50002 + }, + { + "epoch": 0.50003, + "grad_norm": 0.808960472365385, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 50003 + }, + { + "epoch": 0.50004, + "grad_norm": 0.7617022921054099, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 50004 + }, + { + "epoch": 0.50005, + "grad_norm": 0.8435344920256282, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 50005 + }, + { + "epoch": 0.50006, + "grad_norm": 0.9049576813465953, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 50006 + }, + { + "epoch": 0.50007, + "grad_norm": 0.9074925813490613, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 50007 + }, + { + "epoch": 0.50008, + "grad_norm": 1.0850928653206293, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 50008 + }, + { + "epoch": 0.50009, + "grad_norm": 1.1305324489492876, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 50009 + }, + { + "epoch": 0.5001, + "grad_norm": 0.8746158113509895, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 50010 + }, + { + "epoch": 0.50011, + "grad_norm": 0.8406936651825199, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 50011 + }, + { + "epoch": 0.50012, + "grad_norm": 0.9879817983712039, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 50012 + }, + { + "epoch": 0.50013, + "grad_norm": 1.2071848925437154, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 50013 + }, + { + "epoch": 0.50014, + "grad_norm": 1.1668182650861183, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 50014 + }, + { + "epoch": 0.50015, + "grad_norm": 1.5237222186690305, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 50015 + }, + { + "epoch": 0.50016, + "grad_norm": 0.8705220925519459, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 50016 + }, + { + "epoch": 0.50017, + "grad_norm": 0.8181475619171567, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 50017 + }, + { + "epoch": 0.50018, + "grad_norm": 0.9243459765976857, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 50018 + }, + { + "epoch": 0.50019, + "grad_norm": 0.8704675644406998, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 50019 + }, + { + "epoch": 0.5002, + "grad_norm": 1.109964702051581, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 50020 + }, + { + "epoch": 0.50021, + "grad_norm": 1.1995541099782998, + "learning_rate": 0.003, + "loss": 4.059, + "step": 50021 + }, + { + "epoch": 0.50022, + "grad_norm": 1.00173700007151, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 50022 + }, + { + "epoch": 0.50023, + "grad_norm": 1.2625184491762516, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 50023 + }, + { + "epoch": 0.50024, + "grad_norm": 1.0548418685369336, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 50024 + }, + { + "epoch": 0.50025, + "grad_norm": 1.0708367319479883, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 50025 + }, + { + "epoch": 0.50026, + "grad_norm": 1.1726457158828578, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 50026 + }, + { + "epoch": 0.50027, + "grad_norm": 1.1165569986936117, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 50027 + }, + { + "epoch": 0.50028, + "grad_norm": 1.1052226064404274, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 50028 + }, + { + "epoch": 0.50029, + "grad_norm": 0.9460241639947375, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 50029 + }, + { + "epoch": 0.5003, + "grad_norm": 1.159849248209572, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 50030 + }, + { + "epoch": 0.50031, + "grad_norm": 1.2873243190214392, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 50031 + }, + { + "epoch": 0.50032, + "grad_norm": 1.1485948242821313, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 50032 + }, + { + "epoch": 0.50033, + "grad_norm": 0.9195838363492852, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 50033 + }, + { + "epoch": 0.50034, + "grad_norm": 0.9824748404633746, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 50034 + }, + { + "epoch": 0.50035, + "grad_norm": 1.0634086360125539, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 50035 + }, + { + "epoch": 0.50036, + "grad_norm": 0.9737983084534348, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 50036 + }, + { + "epoch": 0.50037, + "grad_norm": 1.016822584813816, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 50037 + }, + { + "epoch": 0.50038, + "grad_norm": 1.3886647568987638, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 50038 + }, + { + "epoch": 0.50039, + "grad_norm": 0.8260733338558308, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 50039 + }, + { + "epoch": 0.5004, + "grad_norm": 0.810548142974595, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 50040 + }, + { + "epoch": 0.50041, + "grad_norm": 0.7789630817462775, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 50041 + }, + { + "epoch": 0.50042, + "grad_norm": 0.8852772208508248, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 50042 + }, + { + "epoch": 0.50043, + "grad_norm": 1.044713665309575, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 50043 + }, + { + "epoch": 0.50044, + "grad_norm": 1.1884067114973562, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 50044 + }, + { + "epoch": 0.50045, + "grad_norm": 0.9907465738345709, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 50045 + }, + { + "epoch": 0.50046, + "grad_norm": 1.1142704101563419, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 50046 + }, + { + "epoch": 0.50047, + "grad_norm": 1.1859091025375206, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 50047 + }, + { + "epoch": 0.50048, + "grad_norm": 0.9733464694011238, + "learning_rate": 0.003, + "loss": 4.048, + "step": 50048 + }, + { + "epoch": 0.50049, + "grad_norm": 1.0357687069436883, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 50049 + }, + { + "epoch": 0.5005, + "grad_norm": 0.9947687431206036, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 50050 + }, + { + "epoch": 0.50051, + "grad_norm": 1.1645259754880992, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 50051 + }, + { + "epoch": 0.50052, + "grad_norm": 0.9317348632539469, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 50052 + }, + { + "epoch": 0.50053, + "grad_norm": 0.9320489937227608, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 50053 + }, + { + "epoch": 0.50054, + "grad_norm": 0.9866819905751499, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 50054 + }, + { + "epoch": 0.50055, + "grad_norm": 1.2281864215759906, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 50055 + }, + { + "epoch": 0.50056, + "grad_norm": 1.0082384106767395, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 50056 + }, + { + "epoch": 0.50057, + "grad_norm": 1.3527052535116943, + "learning_rate": 0.003, + "loss": 4.045, + "step": 50057 + }, + { + "epoch": 0.50058, + "grad_norm": 0.9111149522598846, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 50058 + }, + { + "epoch": 0.50059, + "grad_norm": 0.9363378298013842, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 50059 + }, + { + "epoch": 0.5006, + "grad_norm": 1.0348811909149345, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 50060 + }, + { + "epoch": 0.50061, + "grad_norm": 1.1565544252105233, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 50061 + }, + { + "epoch": 0.50062, + "grad_norm": 1.0543258379000473, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 50062 + }, + { + "epoch": 0.50063, + "grad_norm": 1.3242237270504063, + "learning_rate": 0.003, + "loss": 4.053, + "step": 50063 + }, + { + "epoch": 0.50064, + "grad_norm": 1.0482237925387312, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 50064 + }, + { + "epoch": 0.50065, + "grad_norm": 1.164614288721052, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 50065 + }, + { + "epoch": 0.50066, + "grad_norm": 1.1664728954915087, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 50066 + }, + { + "epoch": 0.50067, + "grad_norm": 0.983046714617495, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 50067 + }, + { + "epoch": 0.50068, + "grad_norm": 1.1505593533181298, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 50068 + }, + { + "epoch": 0.50069, + "grad_norm": 1.1650642689373008, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 50069 + }, + { + "epoch": 0.5007, + "grad_norm": 1.1008644056753414, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 50070 + }, + { + "epoch": 0.50071, + "grad_norm": 0.9410959584519409, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 50071 + }, + { + "epoch": 0.50072, + "grad_norm": 0.9956494441238493, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 50072 + }, + { + "epoch": 0.50073, + "grad_norm": 1.086141031572696, + "learning_rate": 0.003, + "loss": 4.038, + "step": 50073 + }, + { + "epoch": 0.50074, + "grad_norm": 0.9855613089027042, + "learning_rate": 0.003, + "loss": 4.01, + "step": 50074 + }, + { + "epoch": 0.50075, + "grad_norm": 1.2399687242218416, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 50075 + }, + { + "epoch": 0.50076, + "grad_norm": 0.9581055360814698, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 50076 + }, + { + "epoch": 0.50077, + "grad_norm": 1.0340480878945095, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 50077 + }, + { + "epoch": 0.50078, + "grad_norm": 0.8874906180901839, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 50078 + }, + { + "epoch": 0.50079, + "grad_norm": 0.7846938396928025, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 50079 + }, + { + "epoch": 0.5008, + "grad_norm": 0.7058293205522811, + "learning_rate": 0.003, + "loss": 4.051, + "step": 50080 + }, + { + "epoch": 0.50081, + "grad_norm": 0.9502387333473312, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 50081 + }, + { + "epoch": 0.50082, + "grad_norm": 1.4763430821648762, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 50082 + }, + { + "epoch": 0.50083, + "grad_norm": 0.916106404463864, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 50083 + }, + { + "epoch": 0.50084, + "grad_norm": 1.0160945074747305, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 50084 + }, + { + "epoch": 0.50085, + "grad_norm": 1.1549294282978322, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 50085 + }, + { + "epoch": 0.50086, + "grad_norm": 0.9698377974581989, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 50086 + }, + { + "epoch": 0.50087, + "grad_norm": 1.2322869749507754, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 50087 + }, + { + "epoch": 0.50088, + "grad_norm": 0.909191250703076, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 50088 + }, + { + "epoch": 0.50089, + "grad_norm": 0.9149817689608375, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 50089 + }, + { + "epoch": 0.5009, + "grad_norm": 1.1606012331171507, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 50090 + }, + { + "epoch": 0.50091, + "grad_norm": 1.23525936116233, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 50091 + }, + { + "epoch": 0.50092, + "grad_norm": 0.8721212221432126, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 50092 + }, + { + "epoch": 0.50093, + "grad_norm": 0.7815471618654777, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 50093 + }, + { + "epoch": 0.50094, + "grad_norm": 0.8673805750366611, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 50094 + }, + { + "epoch": 0.50095, + "grad_norm": 1.2425711841996079, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 50095 + }, + { + "epoch": 0.50096, + "grad_norm": 1.0482252357971638, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 50096 + }, + { + "epoch": 0.50097, + "grad_norm": 1.2115132536395519, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 50097 + }, + { + "epoch": 0.50098, + "grad_norm": 0.9335022677277532, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 50098 + }, + { + "epoch": 0.50099, + "grad_norm": 0.8790276032255149, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 50099 + }, + { + "epoch": 0.501, + "grad_norm": 0.9960829436010483, + "learning_rate": 0.003, + "loss": 4.05, + "step": 50100 + }, + { + "epoch": 0.50101, + "grad_norm": 1.2957236530610905, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 50101 + }, + { + "epoch": 0.50102, + "grad_norm": 0.9092491063543603, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 50102 + }, + { + "epoch": 0.50103, + "grad_norm": 1.049126318707036, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 50103 + }, + { + "epoch": 0.50104, + "grad_norm": 1.3490186347175785, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 50104 + }, + { + "epoch": 0.50105, + "grad_norm": 1.0579544806644063, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 50105 + }, + { + "epoch": 0.50106, + "grad_norm": 1.2084902046964747, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 50106 + }, + { + "epoch": 0.50107, + "grad_norm": 0.9324045837493757, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 50107 + }, + { + "epoch": 0.50108, + "grad_norm": 0.880650573689024, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 50108 + }, + { + "epoch": 0.50109, + "grad_norm": 1.0302384809154457, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 50109 + }, + { + "epoch": 0.5011, + "grad_norm": 1.2584956287885674, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 50110 + }, + { + "epoch": 0.50111, + "grad_norm": 1.1571484620558632, + "learning_rate": 0.003, + "loss": 4.028, + "step": 50111 + }, + { + "epoch": 0.50112, + "grad_norm": 1.003352225032447, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 50112 + }, + { + "epoch": 0.50113, + "grad_norm": 1.1339210428440667, + "learning_rate": 0.003, + "loss": 4.027, + "step": 50113 + }, + { + "epoch": 0.50114, + "grad_norm": 1.0603399511260305, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 50114 + }, + { + "epoch": 0.50115, + "grad_norm": 1.024911151065019, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 50115 + }, + { + "epoch": 0.50116, + "grad_norm": 1.0699384049765908, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 50116 + }, + { + "epoch": 0.50117, + "grad_norm": 1.0073073692791865, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 50117 + }, + { + "epoch": 0.50118, + "grad_norm": 1.19314958800045, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 50118 + }, + { + "epoch": 0.50119, + "grad_norm": 1.0603178595923282, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 50119 + }, + { + "epoch": 0.5012, + "grad_norm": 1.0378939403431535, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 50120 + }, + { + "epoch": 0.50121, + "grad_norm": 1.0663742794427937, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 50121 + }, + { + "epoch": 0.50122, + "grad_norm": 1.203709572555392, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 50122 + }, + { + "epoch": 0.50123, + "grad_norm": 1.0280737669957694, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 50123 + }, + { + "epoch": 0.50124, + "grad_norm": 1.1186887681203186, + "learning_rate": 0.003, + "loss": 4.016, + "step": 50124 + }, + { + "epoch": 0.50125, + "grad_norm": 1.1265279240199364, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 50125 + }, + { + "epoch": 0.50126, + "grad_norm": 1.3425039500923577, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 50126 + }, + { + "epoch": 0.50127, + "grad_norm": 0.7625159249567651, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 50127 + }, + { + "epoch": 0.50128, + "grad_norm": 0.7190710973450077, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 50128 + }, + { + "epoch": 0.50129, + "grad_norm": 0.8220113677817221, + "learning_rate": 0.003, + "loss": 4.05, + "step": 50129 + }, + { + "epoch": 0.5013, + "grad_norm": 0.9707783846796687, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 50130 + }, + { + "epoch": 0.50131, + "grad_norm": 0.9834324333392536, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 50131 + }, + { + "epoch": 0.50132, + "grad_norm": 1.211284126735191, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 50132 + }, + { + "epoch": 0.50133, + "grad_norm": 1.2079204793041527, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 50133 + }, + { + "epoch": 0.50134, + "grad_norm": 1.0568264168469526, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 50134 + }, + { + "epoch": 0.50135, + "grad_norm": 1.0604774851422525, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 50135 + }, + { + "epoch": 0.50136, + "grad_norm": 1.195575169836388, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 50136 + }, + { + "epoch": 0.50137, + "grad_norm": 1.0989499334552408, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 50137 + }, + { + "epoch": 0.50138, + "grad_norm": 1.0405385011934456, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 50138 + }, + { + "epoch": 0.50139, + "grad_norm": 1.0394074154513373, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 50139 + }, + { + "epoch": 0.5014, + "grad_norm": 0.9731701237133769, + "learning_rate": 0.003, + "loss": 4.049, + "step": 50140 + }, + { + "epoch": 0.50141, + "grad_norm": 1.1176598796190889, + "learning_rate": 0.003, + "loss": 4.018, + "step": 50141 + }, + { + "epoch": 0.50142, + "grad_norm": 0.9002095965888246, + "learning_rate": 0.003, + "loss": 4.006, + "step": 50142 + }, + { + "epoch": 0.50143, + "grad_norm": 0.9640525486481546, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 50143 + }, + { + "epoch": 0.50144, + "grad_norm": 1.0868148794631014, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 50144 + }, + { + "epoch": 0.50145, + "grad_norm": 1.123763633343236, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 50145 + }, + { + "epoch": 0.50146, + "grad_norm": 1.1800375561120369, + "learning_rate": 0.003, + "loss": 4.052, + "step": 50146 + }, + { + "epoch": 0.50147, + "grad_norm": 0.9545823260751812, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 50147 + }, + { + "epoch": 0.50148, + "grad_norm": 0.9043540468728254, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 50148 + }, + { + "epoch": 0.50149, + "grad_norm": 1.1412135643325794, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 50149 + }, + { + "epoch": 0.5015, + "grad_norm": 1.1404762099896641, + "learning_rate": 0.003, + "loss": 4.066, + "step": 50150 + }, + { + "epoch": 0.50151, + "grad_norm": 0.9927789719876497, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 50151 + }, + { + "epoch": 0.50152, + "grad_norm": 1.1084859921528543, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 50152 + }, + { + "epoch": 0.50153, + "grad_norm": 0.8918448891198453, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 50153 + }, + { + "epoch": 0.50154, + "grad_norm": 0.8542093691692103, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 50154 + }, + { + "epoch": 0.50155, + "grad_norm": 0.976403681228889, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 50155 + }, + { + "epoch": 0.50156, + "grad_norm": 1.2857349442204493, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 50156 + }, + { + "epoch": 0.50157, + "grad_norm": 1.0938811510737203, + "learning_rate": 0.003, + "loss": 4.045, + "step": 50157 + }, + { + "epoch": 0.50158, + "grad_norm": 1.1418711217984205, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 50158 + }, + { + "epoch": 0.50159, + "grad_norm": 1.0103647939715579, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 50159 + }, + { + "epoch": 0.5016, + "grad_norm": 1.1259223480838005, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 50160 + }, + { + "epoch": 0.50161, + "grad_norm": 0.9739881411026228, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 50161 + }, + { + "epoch": 0.50162, + "grad_norm": 1.1629650274755003, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 50162 + }, + { + "epoch": 0.50163, + "grad_norm": 1.127891824275566, + "learning_rate": 0.003, + "loss": 4.032, + "step": 50163 + }, + { + "epoch": 0.50164, + "grad_norm": 1.126129927882213, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 50164 + }, + { + "epoch": 0.50165, + "grad_norm": 1.0004933669430138, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 50165 + }, + { + "epoch": 0.50166, + "grad_norm": 1.1821274272644777, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 50166 + }, + { + "epoch": 0.50167, + "grad_norm": 0.8851042849576547, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 50167 + }, + { + "epoch": 0.50168, + "grad_norm": 0.8401817082291443, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 50168 + }, + { + "epoch": 0.50169, + "grad_norm": 0.843174332493622, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 50169 + }, + { + "epoch": 0.5017, + "grad_norm": 1.082175439070082, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 50170 + }, + { + "epoch": 0.50171, + "grad_norm": 1.2427376620104786, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 50171 + }, + { + "epoch": 0.50172, + "grad_norm": 1.0248861215132483, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 50172 + }, + { + "epoch": 0.50173, + "grad_norm": 1.4092047220774648, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 50173 + }, + { + "epoch": 0.50174, + "grad_norm": 0.8240921161711119, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 50174 + }, + { + "epoch": 0.50175, + "grad_norm": 0.9134201128773416, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 50175 + }, + { + "epoch": 0.50176, + "grad_norm": 1.2821400787057942, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 50176 + }, + { + "epoch": 0.50177, + "grad_norm": 1.0163892150128075, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 50177 + }, + { + "epoch": 0.50178, + "grad_norm": 1.2306950097640794, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 50178 + }, + { + "epoch": 0.50179, + "grad_norm": 0.9117201331123717, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 50179 + }, + { + "epoch": 0.5018, + "grad_norm": 1.3363576484170887, + "learning_rate": 0.003, + "loss": 4.055, + "step": 50180 + }, + { + "epoch": 0.50181, + "grad_norm": 0.9268654246558381, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 50181 + }, + { + "epoch": 0.50182, + "grad_norm": 0.8849670243110852, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 50182 + }, + { + "epoch": 0.50183, + "grad_norm": 1.3070366575183012, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 50183 + }, + { + "epoch": 0.50184, + "grad_norm": 1.3192214730565215, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 50184 + }, + { + "epoch": 0.50185, + "grad_norm": 0.7696272838708361, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 50185 + }, + { + "epoch": 0.50186, + "grad_norm": 0.8247852851597617, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 50186 + }, + { + "epoch": 0.50187, + "grad_norm": 1.0015695575013628, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 50187 + }, + { + "epoch": 0.50188, + "grad_norm": 1.0640966613887388, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 50188 + }, + { + "epoch": 0.50189, + "grad_norm": 1.0096748561176614, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 50189 + }, + { + "epoch": 0.5019, + "grad_norm": 1.0440238987241448, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 50190 + }, + { + "epoch": 0.50191, + "grad_norm": 0.9314022820131388, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 50191 + }, + { + "epoch": 0.50192, + "grad_norm": 1.0440775662549198, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 50192 + }, + { + "epoch": 0.50193, + "grad_norm": 1.269310228803169, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 50193 + }, + { + "epoch": 0.50194, + "grad_norm": 0.9424152160789311, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 50194 + }, + { + "epoch": 0.50195, + "grad_norm": 1.244293197144148, + "learning_rate": 0.003, + "loss": 3.989, + "step": 50195 + }, + { + "epoch": 0.50196, + "grad_norm": 1.0667050615733935, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 50196 + }, + { + "epoch": 0.50197, + "grad_norm": 1.0196977719935507, + "learning_rate": 0.003, + "loss": 4.034, + "step": 50197 + }, + { + "epoch": 0.50198, + "grad_norm": 1.2480500836772526, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 50198 + }, + { + "epoch": 0.50199, + "grad_norm": 0.9941492804267417, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 50199 + }, + { + "epoch": 0.502, + "grad_norm": 1.1295483087581384, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 50200 + }, + { + "epoch": 0.50201, + "grad_norm": 1.0389622837872097, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 50201 + }, + { + "epoch": 0.50202, + "grad_norm": 1.0528731447381197, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 50202 + }, + { + "epoch": 0.50203, + "grad_norm": 1.0684499360927893, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 50203 + }, + { + "epoch": 0.50204, + "grad_norm": 1.103520426636223, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 50204 + }, + { + "epoch": 0.50205, + "grad_norm": 0.8874388818675072, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 50205 + }, + { + "epoch": 0.50206, + "grad_norm": 0.9383803296189656, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 50206 + }, + { + "epoch": 0.50207, + "grad_norm": 0.935434226703032, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 50207 + }, + { + "epoch": 0.50208, + "grad_norm": 0.9703116298942223, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 50208 + }, + { + "epoch": 0.50209, + "grad_norm": 1.2149749736942843, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 50209 + }, + { + "epoch": 0.5021, + "grad_norm": 1.1506140352275698, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 50210 + }, + { + "epoch": 0.50211, + "grad_norm": 0.9841689107720846, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 50211 + }, + { + "epoch": 0.50212, + "grad_norm": 1.0841216290619302, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 50212 + }, + { + "epoch": 0.50213, + "grad_norm": 1.0756402108072016, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 50213 + }, + { + "epoch": 0.50214, + "grad_norm": 1.1318429667381362, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 50214 + }, + { + "epoch": 0.50215, + "grad_norm": 1.0346383963339392, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 50215 + }, + { + "epoch": 0.50216, + "grad_norm": 1.0127917334374956, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 50216 + }, + { + "epoch": 0.50217, + "grad_norm": 0.9181436551345117, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 50217 + }, + { + "epoch": 0.50218, + "grad_norm": 1.0386519745234424, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 50218 + }, + { + "epoch": 0.50219, + "grad_norm": 1.1868010784280276, + "learning_rate": 0.003, + "loss": 4.022, + "step": 50219 + }, + { + "epoch": 0.5022, + "grad_norm": 1.0640240740456843, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 50220 + }, + { + "epoch": 0.50221, + "grad_norm": 1.060105123400291, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 50221 + }, + { + "epoch": 0.50222, + "grad_norm": 0.9344889297622943, + "learning_rate": 0.003, + "loss": 4.025, + "step": 50222 + }, + { + "epoch": 0.50223, + "grad_norm": 0.9059592759722487, + "learning_rate": 0.003, + "loss": 4.047, + "step": 50223 + }, + { + "epoch": 0.50224, + "grad_norm": 1.0120079860143003, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 50224 + }, + { + "epoch": 0.50225, + "grad_norm": 1.1175916817733031, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 50225 + }, + { + "epoch": 0.50226, + "grad_norm": 1.0469002744461182, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 50226 + }, + { + "epoch": 0.50227, + "grad_norm": 1.2938813151169857, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 50227 + }, + { + "epoch": 0.50228, + "grad_norm": 1.020669671128128, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 50228 + }, + { + "epoch": 0.50229, + "grad_norm": 1.0986090742348706, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 50229 + }, + { + "epoch": 0.5023, + "grad_norm": 1.1491995555717012, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 50230 + }, + { + "epoch": 0.50231, + "grad_norm": 1.1764158953847075, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 50231 + }, + { + "epoch": 0.50232, + "grad_norm": 0.998024338540344, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 50232 + }, + { + "epoch": 0.50233, + "grad_norm": 1.111388033123073, + "learning_rate": 0.003, + "loss": 4.036, + "step": 50233 + }, + { + "epoch": 0.50234, + "grad_norm": 0.898572032959165, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 50234 + }, + { + "epoch": 0.50235, + "grad_norm": 0.949436508951406, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 50235 + }, + { + "epoch": 0.50236, + "grad_norm": 1.034444639444726, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 50236 + }, + { + "epoch": 0.50237, + "grad_norm": 1.0591485982697224, + "learning_rate": 0.003, + "loss": 4.043, + "step": 50237 + }, + { + "epoch": 0.50238, + "grad_norm": 1.03586688730548, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 50238 + }, + { + "epoch": 0.50239, + "grad_norm": 1.0987991790275908, + "learning_rate": 0.003, + "loss": 4.029, + "step": 50239 + }, + { + "epoch": 0.5024, + "grad_norm": 0.9179916161297982, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 50240 + }, + { + "epoch": 0.50241, + "grad_norm": 0.9424926542628699, + "learning_rate": 0.003, + "loss": 4.027, + "step": 50241 + }, + { + "epoch": 0.50242, + "grad_norm": 1.1819222268543461, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 50242 + }, + { + "epoch": 0.50243, + "grad_norm": 0.8913987716700188, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 50243 + }, + { + "epoch": 0.50244, + "grad_norm": 1.074986127062411, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 50244 + }, + { + "epoch": 0.50245, + "grad_norm": 0.9858037356774284, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 50245 + }, + { + "epoch": 0.50246, + "grad_norm": 1.0128743473249218, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 50246 + }, + { + "epoch": 0.50247, + "grad_norm": 1.0884563581243325, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 50247 + }, + { + "epoch": 0.50248, + "grad_norm": 0.9531058591159101, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 50248 + }, + { + "epoch": 0.50249, + "grad_norm": 1.1351515064108502, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 50249 + }, + { + "epoch": 0.5025, + "grad_norm": 0.9880700122541527, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 50250 + }, + { + "epoch": 0.50251, + "grad_norm": 1.0400350731592185, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 50251 + }, + { + "epoch": 0.50252, + "grad_norm": 1.1433321399231766, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 50252 + }, + { + "epoch": 0.50253, + "grad_norm": 1.3351663247535597, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 50253 + }, + { + "epoch": 0.50254, + "grad_norm": 0.8679224443358103, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 50254 + }, + { + "epoch": 0.50255, + "grad_norm": 0.9174917012096133, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 50255 + }, + { + "epoch": 0.50256, + "grad_norm": 0.9237688843361467, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 50256 + }, + { + "epoch": 0.50257, + "grad_norm": 1.097442867437807, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 50257 + }, + { + "epoch": 0.50258, + "grad_norm": 1.1972612685373292, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 50258 + }, + { + "epoch": 0.50259, + "grad_norm": 1.015770019022727, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 50259 + }, + { + "epoch": 0.5026, + "grad_norm": 1.0088186601083495, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 50260 + }, + { + "epoch": 0.50261, + "grad_norm": 1.1234500978792443, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 50261 + }, + { + "epoch": 0.50262, + "grad_norm": 1.1515434007584648, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 50262 + }, + { + "epoch": 0.50263, + "grad_norm": 1.1827726596742287, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 50263 + }, + { + "epoch": 0.50264, + "grad_norm": 0.9417892971999522, + "learning_rate": 0.003, + "loss": 4.02, + "step": 50264 + }, + { + "epoch": 0.50265, + "grad_norm": 1.0155538938378466, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 50265 + }, + { + "epoch": 0.50266, + "grad_norm": 1.2220566635724628, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 50266 + }, + { + "epoch": 0.50267, + "grad_norm": 1.023964061103494, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 50267 + }, + { + "epoch": 0.50268, + "grad_norm": 0.9925857513149936, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 50268 + }, + { + "epoch": 0.50269, + "grad_norm": 1.1428591176282241, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 50269 + }, + { + "epoch": 0.5027, + "grad_norm": 1.1192883284772766, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 50270 + }, + { + "epoch": 0.50271, + "grad_norm": 0.9949704037411808, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 50271 + }, + { + "epoch": 0.50272, + "grad_norm": 1.0338943403268188, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 50272 + }, + { + "epoch": 0.50273, + "grad_norm": 0.8991571877275368, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 50273 + }, + { + "epoch": 0.50274, + "grad_norm": 0.929346226940194, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 50274 + }, + { + "epoch": 0.50275, + "grad_norm": 1.0433560204996828, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 50275 + }, + { + "epoch": 0.50276, + "grad_norm": 0.9439445307285721, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 50276 + }, + { + "epoch": 0.50277, + "grad_norm": 0.967457868421743, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 50277 + }, + { + "epoch": 0.50278, + "grad_norm": 1.4180110088510929, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 50278 + }, + { + "epoch": 0.50279, + "grad_norm": 1.0460332413756284, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 50279 + }, + { + "epoch": 0.5028, + "grad_norm": 1.283777392948356, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 50280 + }, + { + "epoch": 0.50281, + "grad_norm": 0.7761359937778632, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 50281 + }, + { + "epoch": 0.50282, + "grad_norm": 0.7388192038085473, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 50282 + }, + { + "epoch": 0.50283, + "grad_norm": 0.8983199056791616, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 50283 + }, + { + "epoch": 0.50284, + "grad_norm": 1.0358076651947667, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 50284 + }, + { + "epoch": 0.50285, + "grad_norm": 1.0000952718069915, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 50285 + }, + { + "epoch": 0.50286, + "grad_norm": 1.0528909615019384, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 50286 + }, + { + "epoch": 0.50287, + "grad_norm": 1.338525462586192, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 50287 + }, + { + "epoch": 0.50288, + "grad_norm": 0.9898857952831495, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 50288 + }, + { + "epoch": 0.50289, + "grad_norm": 1.0442775775185629, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 50289 + }, + { + "epoch": 0.5029, + "grad_norm": 1.1081693947229163, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 50290 + }, + { + "epoch": 0.50291, + "grad_norm": 0.972789304767873, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 50291 + }, + { + "epoch": 0.50292, + "grad_norm": 1.0732094094430276, + "learning_rate": 0.003, + "loss": 4.092, + "step": 50292 + }, + { + "epoch": 0.50293, + "grad_norm": 1.0004404978295836, + "learning_rate": 0.003, + "loss": 4.027, + "step": 50293 + }, + { + "epoch": 0.50294, + "grad_norm": 1.0619288956635176, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 50294 + }, + { + "epoch": 0.50295, + "grad_norm": 0.9050402146675681, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 50295 + }, + { + "epoch": 0.50296, + "grad_norm": 1.0457738813215107, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 50296 + }, + { + "epoch": 0.50297, + "grad_norm": 1.369249758091998, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 50297 + }, + { + "epoch": 0.50298, + "grad_norm": 1.1035320936574844, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 50298 + }, + { + "epoch": 0.50299, + "grad_norm": 0.8206601063607403, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 50299 + }, + { + "epoch": 0.503, + "grad_norm": 0.7629915899050412, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 50300 + }, + { + "epoch": 0.50301, + "grad_norm": 0.8530046969284408, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 50301 + }, + { + "epoch": 0.50302, + "grad_norm": 1.0525806793327162, + "learning_rate": 0.003, + "loss": 4.01, + "step": 50302 + }, + { + "epoch": 0.50303, + "grad_norm": 1.1044020753038937, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 50303 + }, + { + "epoch": 0.50304, + "grad_norm": 0.9385338808246075, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 50304 + }, + { + "epoch": 0.50305, + "grad_norm": 1.1724061017125145, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 50305 + }, + { + "epoch": 0.50306, + "grad_norm": 1.0609438752543996, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 50306 + }, + { + "epoch": 0.50307, + "grad_norm": 1.0193719879805478, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 50307 + }, + { + "epoch": 0.50308, + "grad_norm": 1.2112252556641065, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 50308 + }, + { + "epoch": 0.50309, + "grad_norm": 1.0141124565661996, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 50309 + }, + { + "epoch": 0.5031, + "grad_norm": 1.1715716814480726, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 50310 + }, + { + "epoch": 0.50311, + "grad_norm": 1.2041139303465098, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 50311 + }, + { + "epoch": 0.50312, + "grad_norm": 0.9277812256006833, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 50312 + }, + { + "epoch": 0.50313, + "grad_norm": 0.9371572169115374, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 50313 + }, + { + "epoch": 0.50314, + "grad_norm": 1.0342986973110007, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 50314 + }, + { + "epoch": 0.50315, + "grad_norm": 1.2310197379576036, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 50315 + }, + { + "epoch": 0.50316, + "grad_norm": 1.0593532842292672, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 50316 + }, + { + "epoch": 0.50317, + "grad_norm": 1.1022889080433798, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 50317 + }, + { + "epoch": 0.50318, + "grad_norm": 1.1214523008420652, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 50318 + }, + { + "epoch": 0.50319, + "grad_norm": 1.033768654256001, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 50319 + }, + { + "epoch": 0.5032, + "grad_norm": 1.2557211660762604, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 50320 + }, + { + "epoch": 0.50321, + "grad_norm": 1.1154739872092756, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 50321 + }, + { + "epoch": 0.50322, + "grad_norm": 1.1113766023461786, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 50322 + }, + { + "epoch": 0.50323, + "grad_norm": 1.042592188669435, + "learning_rate": 0.003, + "loss": 4.027, + "step": 50323 + }, + { + "epoch": 0.50324, + "grad_norm": 1.2266165147232737, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 50324 + }, + { + "epoch": 0.50325, + "grad_norm": 1.0054388874117466, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 50325 + }, + { + "epoch": 0.50326, + "grad_norm": 0.9618617826940907, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 50326 + }, + { + "epoch": 0.50327, + "grad_norm": 0.867334876742346, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 50327 + }, + { + "epoch": 0.50328, + "grad_norm": 0.8473211491772966, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 50328 + }, + { + "epoch": 0.50329, + "grad_norm": 1.0119880657913511, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 50329 + }, + { + "epoch": 0.5033, + "grad_norm": 1.1365755673182443, + "learning_rate": 0.003, + "loss": 4.041, + "step": 50330 + }, + { + "epoch": 0.50331, + "grad_norm": 1.0024566032748863, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 50331 + }, + { + "epoch": 0.50332, + "grad_norm": 1.0191882149864335, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 50332 + }, + { + "epoch": 0.50333, + "grad_norm": 1.2284261009386166, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 50333 + }, + { + "epoch": 0.50334, + "grad_norm": 1.240478176099343, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 50334 + }, + { + "epoch": 0.50335, + "grad_norm": 1.102383726338929, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 50335 + }, + { + "epoch": 0.50336, + "grad_norm": 1.0727976475706626, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 50336 + }, + { + "epoch": 0.50337, + "grad_norm": 1.1296443490034418, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 50337 + }, + { + "epoch": 0.50338, + "grad_norm": 0.9577955982121399, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 50338 + }, + { + "epoch": 0.50339, + "grad_norm": 1.021774907909643, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 50339 + }, + { + "epoch": 0.5034, + "grad_norm": 1.0633907341789546, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 50340 + }, + { + "epoch": 0.50341, + "grad_norm": 1.0759793846774135, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 50341 + }, + { + "epoch": 0.50342, + "grad_norm": 1.0055102751336116, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 50342 + }, + { + "epoch": 0.50343, + "grad_norm": 1.1794121742449692, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 50343 + }, + { + "epoch": 0.50344, + "grad_norm": 1.0986608395142703, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 50344 + }, + { + "epoch": 0.50345, + "grad_norm": 1.2117312909365947, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 50345 + }, + { + "epoch": 0.50346, + "grad_norm": 0.9047580414310709, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 50346 + }, + { + "epoch": 0.50347, + "grad_norm": 1.0056977624821495, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 50347 + }, + { + "epoch": 0.50348, + "grad_norm": 1.1644074483943454, + "learning_rate": 0.003, + "loss": 4.021, + "step": 50348 + }, + { + "epoch": 0.50349, + "grad_norm": 1.036994243815829, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 50349 + }, + { + "epoch": 0.5035, + "grad_norm": 0.8961744772709065, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 50350 + }, + { + "epoch": 0.50351, + "grad_norm": 1.0145517914573887, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 50351 + }, + { + "epoch": 0.50352, + "grad_norm": 1.1307665338346002, + "learning_rate": 0.003, + "loss": 4.027, + "step": 50352 + }, + { + "epoch": 0.50353, + "grad_norm": 1.0332136509629837, + "learning_rate": 0.003, + "loss": 4.037, + "step": 50353 + }, + { + "epoch": 0.50354, + "grad_norm": 1.3415417312817361, + "learning_rate": 0.003, + "loss": 4.037, + "step": 50354 + }, + { + "epoch": 0.50355, + "grad_norm": 1.099394387106615, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 50355 + }, + { + "epoch": 0.50356, + "grad_norm": 0.9591004769266407, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 50356 + }, + { + "epoch": 0.50357, + "grad_norm": 1.1483018514565522, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 50357 + }, + { + "epoch": 0.50358, + "grad_norm": 1.2936985185941314, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 50358 + }, + { + "epoch": 0.50359, + "grad_norm": 0.9056811250842343, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 50359 + }, + { + "epoch": 0.5036, + "grad_norm": 0.8251796415394618, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 50360 + }, + { + "epoch": 0.50361, + "grad_norm": 0.9252724995330214, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 50361 + }, + { + "epoch": 0.50362, + "grad_norm": 1.065396699231751, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 50362 + }, + { + "epoch": 0.50363, + "grad_norm": 1.1594883891678829, + "learning_rate": 0.003, + "loss": 4.042, + "step": 50363 + }, + { + "epoch": 0.50364, + "grad_norm": 0.8532765867205593, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 50364 + }, + { + "epoch": 0.50365, + "grad_norm": 0.8872259043354338, + "learning_rate": 0.003, + "loss": 4.01, + "step": 50365 + }, + { + "epoch": 0.50366, + "grad_norm": 0.9974417504493576, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 50366 + }, + { + "epoch": 0.50367, + "grad_norm": 1.1213896719924263, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 50367 + }, + { + "epoch": 0.50368, + "grad_norm": 0.8862853658265007, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 50368 + }, + { + "epoch": 0.50369, + "grad_norm": 0.9059797507940134, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 50369 + }, + { + "epoch": 0.5037, + "grad_norm": 1.0446004049485083, + "learning_rate": 0.003, + "loss": 4.044, + "step": 50370 + }, + { + "epoch": 0.50371, + "grad_norm": 1.4946819896949062, + "learning_rate": 0.003, + "loss": 4.034, + "step": 50371 + }, + { + "epoch": 0.50372, + "grad_norm": 0.6699455093193906, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 50372 + }, + { + "epoch": 0.50373, + "grad_norm": 0.7771904079377671, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 50373 + }, + { + "epoch": 0.50374, + "grad_norm": 0.897814170307552, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 50374 + }, + { + "epoch": 0.50375, + "grad_norm": 1.0146620525802608, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 50375 + }, + { + "epoch": 0.50376, + "grad_norm": 1.104148366377113, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 50376 + }, + { + "epoch": 0.50377, + "grad_norm": 1.1292216272096918, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 50377 + }, + { + "epoch": 0.50378, + "grad_norm": 1.1190908707695566, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 50378 + }, + { + "epoch": 0.50379, + "grad_norm": 1.0645210968277572, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 50379 + }, + { + "epoch": 0.5038, + "grad_norm": 1.139903864168234, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 50380 + }, + { + "epoch": 0.50381, + "grad_norm": 1.2840067838339613, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 50381 + }, + { + "epoch": 0.50382, + "grad_norm": 0.9764702984186221, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 50382 + }, + { + "epoch": 0.50383, + "grad_norm": 1.0761294811453466, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 50383 + }, + { + "epoch": 0.50384, + "grad_norm": 1.296556390649517, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 50384 + }, + { + "epoch": 0.50385, + "grad_norm": 0.9858305140198402, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 50385 + }, + { + "epoch": 0.50386, + "grad_norm": 1.045955395595887, + "learning_rate": 0.003, + "loss": 4.03, + "step": 50386 + }, + { + "epoch": 0.50387, + "grad_norm": 1.2770786253890565, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 50387 + }, + { + "epoch": 0.50388, + "grad_norm": 1.0421601746950253, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 50388 + }, + { + "epoch": 0.50389, + "grad_norm": 1.0425991171656268, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 50389 + }, + { + "epoch": 0.5039, + "grad_norm": 1.148483903580569, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 50390 + }, + { + "epoch": 0.50391, + "grad_norm": 0.9818385784646284, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 50391 + }, + { + "epoch": 0.50392, + "grad_norm": 1.1673619138989206, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 50392 + }, + { + "epoch": 0.50393, + "grad_norm": 1.0778831490466856, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 50393 + }, + { + "epoch": 0.50394, + "grad_norm": 1.010942716264703, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 50394 + }, + { + "epoch": 0.50395, + "grad_norm": 0.9630880195823325, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 50395 + }, + { + "epoch": 0.50396, + "grad_norm": 0.8905532242483121, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 50396 + }, + { + "epoch": 0.50397, + "grad_norm": 0.907467505353368, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 50397 + }, + { + "epoch": 0.50398, + "grad_norm": 1.2680867490839414, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 50398 + }, + { + "epoch": 0.50399, + "grad_norm": 0.9052666301591458, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 50399 + }, + { + "epoch": 0.504, + "grad_norm": 0.9500356696645883, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 50400 + }, + { + "epoch": 0.50401, + "grad_norm": 1.1078295899920125, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 50401 + }, + { + "epoch": 0.50402, + "grad_norm": 1.0428651627041958, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 50402 + }, + { + "epoch": 0.50403, + "grad_norm": 1.175297436184729, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 50403 + }, + { + "epoch": 0.50404, + "grad_norm": 1.1862176848117607, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 50404 + }, + { + "epoch": 0.50405, + "grad_norm": 1.1308595644973696, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 50405 + }, + { + "epoch": 0.50406, + "grad_norm": 1.1416324017406503, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 50406 + }, + { + "epoch": 0.50407, + "grad_norm": 1.1793699536712683, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 50407 + }, + { + "epoch": 0.50408, + "grad_norm": 1.1319582415539982, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 50408 + }, + { + "epoch": 0.50409, + "grad_norm": 1.0407185145989029, + "learning_rate": 0.003, + "loss": 4.01, + "step": 50409 + }, + { + "epoch": 0.5041, + "grad_norm": 1.057882149662993, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 50410 + }, + { + "epoch": 0.50411, + "grad_norm": 1.3538617029865585, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 50411 + }, + { + "epoch": 0.50412, + "grad_norm": 0.9664496889567741, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 50412 + }, + { + "epoch": 0.50413, + "grad_norm": 1.185599859160511, + "learning_rate": 0.003, + "loss": 4.034, + "step": 50413 + }, + { + "epoch": 0.50414, + "grad_norm": 0.9949601450038279, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 50414 + }, + { + "epoch": 0.50415, + "grad_norm": 1.0840630525554593, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 50415 + }, + { + "epoch": 0.50416, + "grad_norm": 0.8988979877185991, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 50416 + }, + { + "epoch": 0.50417, + "grad_norm": 0.9007192875671, + "learning_rate": 0.003, + "loss": 4.029, + "step": 50417 + }, + { + "epoch": 0.50418, + "grad_norm": 1.1097028117505503, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 50418 + }, + { + "epoch": 0.50419, + "grad_norm": 1.195320419314715, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 50419 + }, + { + "epoch": 0.5042, + "grad_norm": 0.8547197960029156, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 50420 + }, + { + "epoch": 0.50421, + "grad_norm": 0.7898017514389338, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 50421 + }, + { + "epoch": 0.50422, + "grad_norm": 1.0019967224909478, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 50422 + }, + { + "epoch": 0.50423, + "grad_norm": 1.0338126908101086, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 50423 + }, + { + "epoch": 0.50424, + "grad_norm": 0.9680496674280802, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 50424 + }, + { + "epoch": 0.50425, + "grad_norm": 1.2226060270788102, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 50425 + }, + { + "epoch": 0.50426, + "grad_norm": 0.9960757870112679, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 50426 + }, + { + "epoch": 0.50427, + "grad_norm": 1.230758589083851, + "learning_rate": 0.003, + "loss": 4.011, + "step": 50427 + }, + { + "epoch": 0.50428, + "grad_norm": 0.9352506312653793, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 50428 + }, + { + "epoch": 0.50429, + "grad_norm": 1.1206376764974588, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 50429 + }, + { + "epoch": 0.5043, + "grad_norm": 1.2462603374015246, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 50430 + }, + { + "epoch": 0.50431, + "grad_norm": 1.0726408129884306, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 50431 + }, + { + "epoch": 0.50432, + "grad_norm": 1.1027019587648912, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 50432 + }, + { + "epoch": 0.50433, + "grad_norm": 0.9014084249286315, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 50433 + }, + { + "epoch": 0.50434, + "grad_norm": 0.8988823654068725, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 50434 + }, + { + "epoch": 0.50435, + "grad_norm": 1.0335534012978336, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 50435 + }, + { + "epoch": 0.50436, + "grad_norm": 1.166933156150317, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 50436 + }, + { + "epoch": 0.50437, + "grad_norm": 1.092788083965125, + "learning_rate": 0.003, + "loss": 4.049, + "step": 50437 + }, + { + "epoch": 0.50438, + "grad_norm": 1.005915024106274, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 50438 + }, + { + "epoch": 0.50439, + "grad_norm": 1.1715974705620986, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 50439 + }, + { + "epoch": 0.5044, + "grad_norm": 1.1405243345751057, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 50440 + }, + { + "epoch": 0.50441, + "grad_norm": 1.0402533406746162, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 50441 + }, + { + "epoch": 0.50442, + "grad_norm": 1.0036207619014974, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 50442 + }, + { + "epoch": 0.50443, + "grad_norm": 1.1456256967807104, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 50443 + }, + { + "epoch": 0.50444, + "grad_norm": 1.1319438927114243, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 50444 + }, + { + "epoch": 0.50445, + "grad_norm": 0.9803858530878821, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 50445 + }, + { + "epoch": 0.50446, + "grad_norm": 1.0449353353585584, + "learning_rate": 0.003, + "loss": 4.041, + "step": 50446 + }, + { + "epoch": 0.50447, + "grad_norm": 1.01993710033826, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 50447 + }, + { + "epoch": 0.50448, + "grad_norm": 0.8561782057590603, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 50448 + }, + { + "epoch": 0.50449, + "grad_norm": 0.7973731846110789, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 50449 + }, + { + "epoch": 0.5045, + "grad_norm": 0.9664485707683789, + "learning_rate": 0.003, + "loss": 4.024, + "step": 50450 + }, + { + "epoch": 0.50451, + "grad_norm": 1.1852554319361697, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 50451 + }, + { + "epoch": 0.50452, + "grad_norm": 0.8952213485373146, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 50452 + }, + { + "epoch": 0.50453, + "grad_norm": 1.0119854091571703, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 50453 + }, + { + "epoch": 0.50454, + "grad_norm": 1.217511291498417, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 50454 + }, + { + "epoch": 0.50455, + "grad_norm": 1.1919449466350813, + "learning_rate": 0.003, + "loss": 4.06, + "step": 50455 + }, + { + "epoch": 0.50456, + "grad_norm": 1.3800304286482725, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 50456 + }, + { + "epoch": 0.50457, + "grad_norm": 0.9544078648439964, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 50457 + }, + { + "epoch": 0.50458, + "grad_norm": 1.1022697909184709, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 50458 + }, + { + "epoch": 0.50459, + "grad_norm": 1.1926881465325276, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 50459 + }, + { + "epoch": 0.5046, + "grad_norm": 1.1093190423359427, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 50460 + }, + { + "epoch": 0.50461, + "grad_norm": 1.0638867194790183, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 50461 + }, + { + "epoch": 0.50462, + "grad_norm": 1.0712010567265333, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 50462 + }, + { + "epoch": 0.50463, + "grad_norm": 0.8811553762238161, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 50463 + }, + { + "epoch": 0.50464, + "grad_norm": 1.0674020616621636, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 50464 + }, + { + "epoch": 0.50465, + "grad_norm": 1.4794584942680868, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 50465 + }, + { + "epoch": 0.50466, + "grad_norm": 0.9036936400065215, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 50466 + }, + { + "epoch": 0.50467, + "grad_norm": 1.0859959001247743, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 50467 + }, + { + "epoch": 0.50468, + "grad_norm": 1.2981017045155028, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 50468 + }, + { + "epoch": 0.50469, + "grad_norm": 1.291608715475292, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 50469 + }, + { + "epoch": 0.5047, + "grad_norm": 0.8991250869720648, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 50470 + }, + { + "epoch": 0.50471, + "grad_norm": 0.7600657768194872, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 50471 + }, + { + "epoch": 0.50472, + "grad_norm": 0.904467975468352, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 50472 + }, + { + "epoch": 0.50473, + "grad_norm": 1.0248811597920346, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 50473 + }, + { + "epoch": 0.50474, + "grad_norm": 1.1387278656639341, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 50474 + }, + { + "epoch": 0.50475, + "grad_norm": 1.2877320130523622, + "learning_rate": 0.003, + "loss": 4.028, + "step": 50475 + }, + { + "epoch": 0.50476, + "grad_norm": 0.8801836166594346, + "learning_rate": 0.003, + "loss": 4.029, + "step": 50476 + }, + { + "epoch": 0.50477, + "grad_norm": 0.942314587039255, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 50477 + }, + { + "epoch": 0.50478, + "grad_norm": 1.1704378194615905, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 50478 + }, + { + "epoch": 0.50479, + "grad_norm": 0.9763224196305514, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 50479 + }, + { + "epoch": 0.5048, + "grad_norm": 0.8699424523630298, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 50480 + }, + { + "epoch": 0.50481, + "grad_norm": 0.9178324889680477, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 50481 + }, + { + "epoch": 0.50482, + "grad_norm": 1.1387986967666535, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 50482 + }, + { + "epoch": 0.50483, + "grad_norm": 1.2109603790437742, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 50483 + }, + { + "epoch": 0.50484, + "grad_norm": 0.9110350613892901, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 50484 + }, + { + "epoch": 0.50485, + "grad_norm": 0.9470560794723877, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 50485 + }, + { + "epoch": 0.50486, + "grad_norm": 1.1091685101660442, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 50486 + }, + { + "epoch": 0.50487, + "grad_norm": 1.0602015073103306, + "learning_rate": 0.003, + "loss": 3.998, + "step": 50487 + }, + { + "epoch": 0.50488, + "grad_norm": 1.1639835791143538, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 50488 + }, + { + "epoch": 0.50489, + "grad_norm": 0.9867477386188459, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 50489 + }, + { + "epoch": 0.5049, + "grad_norm": 1.308483132859364, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 50490 + }, + { + "epoch": 0.50491, + "grad_norm": 0.7951255935741249, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 50491 + }, + { + "epoch": 0.50492, + "grad_norm": 0.7572826810899829, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 50492 + }, + { + "epoch": 0.50493, + "grad_norm": 0.7685710181159602, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 50493 + }, + { + "epoch": 0.50494, + "grad_norm": 0.9097268589974117, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 50494 + }, + { + "epoch": 0.50495, + "grad_norm": 1.142750295561488, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 50495 + }, + { + "epoch": 0.50496, + "grad_norm": 1.0356152015408846, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 50496 + }, + { + "epoch": 0.50497, + "grad_norm": 1.2259853271318877, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 50497 + }, + { + "epoch": 0.50498, + "grad_norm": 0.965554838435964, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 50498 + }, + { + "epoch": 0.50499, + "grad_norm": 1.0933375427015557, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 50499 + }, + { + "epoch": 0.505, + "grad_norm": 1.0666633621234907, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 50500 + }, + { + "epoch": 0.50501, + "grad_norm": 1.0560258924407073, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 50501 + }, + { + "epoch": 0.50502, + "grad_norm": 1.1162393381557654, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 50502 + }, + { + "epoch": 0.50503, + "grad_norm": 0.9162204459814526, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 50503 + }, + { + "epoch": 0.50504, + "grad_norm": 1.1488582045353664, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 50504 + }, + { + "epoch": 0.50505, + "grad_norm": 1.0950837546091217, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 50505 + }, + { + "epoch": 0.50506, + "grad_norm": 1.2683401971703239, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 50506 + }, + { + "epoch": 0.50507, + "grad_norm": 1.1563964326903475, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 50507 + }, + { + "epoch": 0.50508, + "grad_norm": 1.0177810370329845, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 50508 + }, + { + "epoch": 0.50509, + "grad_norm": 1.148022730270005, + "learning_rate": 0.003, + "loss": 4.031, + "step": 50509 + }, + { + "epoch": 0.5051, + "grad_norm": 1.2308358573230571, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 50510 + }, + { + "epoch": 0.50511, + "grad_norm": 1.127429307410341, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 50511 + }, + { + "epoch": 0.50512, + "grad_norm": 1.1079209236104346, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 50512 + }, + { + "epoch": 0.50513, + "grad_norm": 0.8737079677464575, + "learning_rate": 0.003, + "loss": 4.054, + "step": 50513 + }, + { + "epoch": 0.50514, + "grad_norm": 0.860853634981808, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 50514 + }, + { + "epoch": 0.50515, + "grad_norm": 1.0581018182503004, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 50515 + }, + { + "epoch": 0.50516, + "grad_norm": 1.0718508716119501, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 50516 + }, + { + "epoch": 0.50517, + "grad_norm": 0.9647953943224966, + "learning_rate": 0.003, + "loss": 4.029, + "step": 50517 + }, + { + "epoch": 0.50518, + "grad_norm": 1.1339662742240912, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 50518 + }, + { + "epoch": 0.50519, + "grad_norm": 0.954957006094006, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 50519 + }, + { + "epoch": 0.5052, + "grad_norm": 1.0399552924881488, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 50520 + }, + { + "epoch": 0.50521, + "grad_norm": 1.323117697887, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 50521 + }, + { + "epoch": 0.50522, + "grad_norm": 0.9306571124499157, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 50522 + }, + { + "epoch": 0.50523, + "grad_norm": 0.8794384503883659, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 50523 + }, + { + "epoch": 0.50524, + "grad_norm": 0.9200824615956046, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 50524 + }, + { + "epoch": 0.50525, + "grad_norm": 1.204232659179393, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 50525 + }, + { + "epoch": 0.50526, + "grad_norm": 1.0419757982092124, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 50526 + }, + { + "epoch": 0.50527, + "grad_norm": 1.009295492074071, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 50527 + }, + { + "epoch": 0.50528, + "grad_norm": 0.9783278153626275, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 50528 + }, + { + "epoch": 0.50529, + "grad_norm": 1.0335053730356658, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 50529 + }, + { + "epoch": 0.5053, + "grad_norm": 1.2526770920320336, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 50530 + }, + { + "epoch": 0.50531, + "grad_norm": 1.1015586101379415, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 50531 + }, + { + "epoch": 0.50532, + "grad_norm": 1.2520495372087723, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 50532 + }, + { + "epoch": 0.50533, + "grad_norm": 0.943716517827168, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 50533 + }, + { + "epoch": 0.50534, + "grad_norm": 0.9678333852349105, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 50534 + }, + { + "epoch": 0.50535, + "grad_norm": 1.2927647879828437, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 50535 + }, + { + "epoch": 0.50536, + "grad_norm": 1.0994276115001427, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 50536 + }, + { + "epoch": 0.50537, + "grad_norm": 1.2192102766553905, + "learning_rate": 0.003, + "loss": 4.066, + "step": 50537 + }, + { + "epoch": 0.50538, + "grad_norm": 0.9512175546107897, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 50538 + }, + { + "epoch": 0.50539, + "grad_norm": 0.9772910725271574, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 50539 + }, + { + "epoch": 0.5054, + "grad_norm": 1.160073354846429, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 50540 + }, + { + "epoch": 0.50541, + "grad_norm": 0.8927754510679118, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 50541 + }, + { + "epoch": 0.50542, + "grad_norm": 0.9933910786795146, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 50542 + }, + { + "epoch": 0.50543, + "grad_norm": 1.0810604610777212, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 50543 + }, + { + "epoch": 0.50544, + "grad_norm": 0.7925201956170449, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 50544 + }, + { + "epoch": 0.50545, + "grad_norm": 0.833917631907308, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 50545 + }, + { + "epoch": 0.50546, + "grad_norm": 0.8727713914940692, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 50546 + }, + { + "epoch": 0.50547, + "grad_norm": 0.8456304035738912, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 50547 + }, + { + "epoch": 0.50548, + "grad_norm": 0.9534381049026167, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 50548 + }, + { + "epoch": 0.50549, + "grad_norm": 0.9688301900548676, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 50549 + }, + { + "epoch": 0.5055, + "grad_norm": 1.2999011743973183, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 50550 + }, + { + "epoch": 0.50551, + "grad_norm": 1.2898911267247362, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 50551 + }, + { + "epoch": 0.50552, + "grad_norm": 0.8781368817222378, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 50552 + }, + { + "epoch": 0.50553, + "grad_norm": 1.0034095180237599, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 50553 + }, + { + "epoch": 0.50554, + "grad_norm": 1.4432938239965587, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 50554 + }, + { + "epoch": 0.50555, + "grad_norm": 0.986613793936823, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 50555 + }, + { + "epoch": 0.50556, + "grad_norm": 1.2370456125097606, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 50556 + }, + { + "epoch": 0.50557, + "grad_norm": 0.8726958119877302, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 50557 + }, + { + "epoch": 0.50558, + "grad_norm": 0.8418398812913107, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 50558 + }, + { + "epoch": 0.50559, + "grad_norm": 0.9702134313246407, + "learning_rate": 0.003, + "loss": 4.044, + "step": 50559 + }, + { + "epoch": 0.5056, + "grad_norm": 1.1238436056761643, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 50560 + }, + { + "epoch": 0.50561, + "grad_norm": 1.146358405592178, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 50561 + }, + { + "epoch": 0.50562, + "grad_norm": 1.0781866729318557, + "learning_rate": 0.003, + "loss": 4.054, + "step": 50562 + }, + { + "epoch": 0.50563, + "grad_norm": 1.047104813305847, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 50563 + }, + { + "epoch": 0.50564, + "grad_norm": 1.1392583758588266, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 50564 + }, + { + "epoch": 0.50565, + "grad_norm": 1.0364654217504827, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 50565 + }, + { + "epoch": 0.50566, + "grad_norm": 1.1195034348458126, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 50566 + }, + { + "epoch": 0.50567, + "grad_norm": 1.0891721347211487, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 50567 + }, + { + "epoch": 0.50568, + "grad_norm": 0.9620148326706875, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 50568 + }, + { + "epoch": 0.50569, + "grad_norm": 1.1617413223620274, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 50569 + }, + { + "epoch": 0.5057, + "grad_norm": 1.225213886555211, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 50570 + }, + { + "epoch": 0.50571, + "grad_norm": 0.9508068598511186, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 50571 + }, + { + "epoch": 0.50572, + "grad_norm": 1.0429486679290791, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 50572 + }, + { + "epoch": 0.50573, + "grad_norm": 0.9922376877673219, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 50573 + }, + { + "epoch": 0.50574, + "grad_norm": 0.9439314523643507, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 50574 + }, + { + "epoch": 0.50575, + "grad_norm": 1.055846792399555, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 50575 + }, + { + "epoch": 0.50576, + "grad_norm": 1.2716611199512706, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 50576 + }, + { + "epoch": 0.50577, + "grad_norm": 1.0434756990468768, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 50577 + }, + { + "epoch": 0.50578, + "grad_norm": 1.314193070928745, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 50578 + }, + { + "epoch": 0.50579, + "grad_norm": 1.0010142386280465, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 50579 + }, + { + "epoch": 0.5058, + "grad_norm": 1.1312101126451615, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 50580 + }, + { + "epoch": 0.50581, + "grad_norm": 1.1501721938452076, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 50581 + }, + { + "epoch": 0.50582, + "grad_norm": 1.1015428112631995, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 50582 + }, + { + "epoch": 0.50583, + "grad_norm": 0.8917145851842146, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 50583 + }, + { + "epoch": 0.50584, + "grad_norm": 0.8298464459986912, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 50584 + }, + { + "epoch": 0.50585, + "grad_norm": 0.8166794635437623, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 50585 + }, + { + "epoch": 0.50586, + "grad_norm": 0.8554735760935064, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 50586 + }, + { + "epoch": 0.50587, + "grad_norm": 0.8491946252847836, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 50587 + }, + { + "epoch": 0.50588, + "grad_norm": 0.9251457916168797, + "learning_rate": 0.003, + "loss": 4.014, + "step": 50588 + }, + { + "epoch": 0.50589, + "grad_norm": 1.1760231904936587, + "learning_rate": 0.003, + "loss": 4.06, + "step": 50589 + }, + { + "epoch": 0.5059, + "grad_norm": 1.15100627601197, + "learning_rate": 0.003, + "loss": 4.028, + "step": 50590 + }, + { + "epoch": 0.50591, + "grad_norm": 1.1539405758036356, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 50591 + }, + { + "epoch": 0.50592, + "grad_norm": 1.110753477797431, + "learning_rate": 0.003, + "loss": 4.047, + "step": 50592 + }, + { + "epoch": 0.50593, + "grad_norm": 0.9958514608697467, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 50593 + }, + { + "epoch": 0.50594, + "grad_norm": 1.1098967252089382, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 50594 + }, + { + "epoch": 0.50595, + "grad_norm": 1.00264543559662, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 50595 + }, + { + "epoch": 0.50596, + "grad_norm": 1.1323439746150166, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 50596 + }, + { + "epoch": 0.50597, + "grad_norm": 0.9940006574087431, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 50597 + }, + { + "epoch": 0.50598, + "grad_norm": 0.9949209595612508, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 50598 + }, + { + "epoch": 0.50599, + "grad_norm": 1.0885101334361096, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 50599 + }, + { + "epoch": 0.506, + "grad_norm": 1.0422566923291352, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 50600 + }, + { + "epoch": 0.50601, + "grad_norm": 1.024772089950975, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 50601 + }, + { + "epoch": 0.50602, + "grad_norm": 1.297905818477589, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 50602 + }, + { + "epoch": 0.50603, + "grad_norm": 1.0168459182389145, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 50603 + }, + { + "epoch": 0.50604, + "grad_norm": 0.8869843546493946, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 50604 + }, + { + "epoch": 0.50605, + "grad_norm": 0.7625566998570661, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 50605 + }, + { + "epoch": 0.50606, + "grad_norm": 0.9021165047288137, + "learning_rate": 0.003, + "loss": 4.004, + "step": 50606 + }, + { + "epoch": 0.50607, + "grad_norm": 1.4115469142342667, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 50607 + }, + { + "epoch": 0.50608, + "grad_norm": 1.10760431261444, + "learning_rate": 0.003, + "loss": 4.017, + "step": 50608 + }, + { + "epoch": 0.50609, + "grad_norm": 0.8529681620352085, + "learning_rate": 0.003, + "loss": 4.023, + "step": 50609 + }, + { + "epoch": 0.5061, + "grad_norm": 0.8016998745714133, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 50610 + }, + { + "epoch": 0.50611, + "grad_norm": 0.9290352423501278, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 50611 + }, + { + "epoch": 0.50612, + "grad_norm": 1.3365611393013455, + "learning_rate": 0.003, + "loss": 4.029, + "step": 50612 + }, + { + "epoch": 0.50613, + "grad_norm": 1.0468401807718353, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 50613 + }, + { + "epoch": 0.50614, + "grad_norm": 1.3092943801318428, + "learning_rate": 0.003, + "loss": 4.028, + "step": 50614 + }, + { + "epoch": 0.50615, + "grad_norm": 0.9045805700462187, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 50615 + }, + { + "epoch": 0.50616, + "grad_norm": 0.8331589181257646, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 50616 + }, + { + "epoch": 0.50617, + "grad_norm": 1.0153055001913665, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 50617 + }, + { + "epoch": 0.50618, + "grad_norm": 1.3491646441127545, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 50618 + }, + { + "epoch": 0.50619, + "grad_norm": 0.8429013969147121, + "learning_rate": 0.003, + "loss": 4.03, + "step": 50619 + }, + { + "epoch": 0.5062, + "grad_norm": 1.036621804499526, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 50620 + }, + { + "epoch": 0.50621, + "grad_norm": 1.4824761911034199, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 50621 + }, + { + "epoch": 0.50622, + "grad_norm": 0.9398798792366683, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 50622 + }, + { + "epoch": 0.50623, + "grad_norm": 1.0781767888793228, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 50623 + }, + { + "epoch": 0.50624, + "grad_norm": 1.092590370440129, + "learning_rate": 0.003, + "loss": 4.018, + "step": 50624 + }, + { + "epoch": 0.50625, + "grad_norm": 1.2546420386347914, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 50625 + }, + { + "epoch": 0.50626, + "grad_norm": 0.8786259257821275, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 50626 + }, + { + "epoch": 0.50627, + "grad_norm": 0.8580783750193436, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 50627 + }, + { + "epoch": 0.50628, + "grad_norm": 1.0005728438793735, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 50628 + }, + { + "epoch": 0.50629, + "grad_norm": 1.2207652995893385, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 50629 + }, + { + "epoch": 0.5063, + "grad_norm": 0.8946425739214021, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 50630 + }, + { + "epoch": 0.50631, + "grad_norm": 1.0105434911267448, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 50631 + }, + { + "epoch": 0.50632, + "grad_norm": 1.175261573323661, + "learning_rate": 0.003, + "loss": 4.055, + "step": 50632 + }, + { + "epoch": 0.50633, + "grad_norm": 1.0402907045946566, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 50633 + }, + { + "epoch": 0.50634, + "grad_norm": 1.2406303619385703, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 50634 + }, + { + "epoch": 0.50635, + "grad_norm": 0.7643596778664941, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 50635 + }, + { + "epoch": 0.50636, + "grad_norm": 0.7444987161049496, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 50636 + }, + { + "epoch": 0.50637, + "grad_norm": 0.9523609761943425, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 50637 + }, + { + "epoch": 0.50638, + "grad_norm": 1.3016520617374616, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 50638 + }, + { + "epoch": 0.50639, + "grad_norm": 1.0129246830786707, + "learning_rate": 0.003, + "loss": 4.031, + "step": 50639 + }, + { + "epoch": 0.5064, + "grad_norm": 1.2842318262787786, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 50640 + }, + { + "epoch": 0.50641, + "grad_norm": 0.9065776716147923, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 50641 + }, + { + "epoch": 0.50642, + "grad_norm": 0.9269303064604923, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 50642 + }, + { + "epoch": 0.50643, + "grad_norm": 1.2518413895776144, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 50643 + }, + { + "epoch": 0.50644, + "grad_norm": 1.2808070884974603, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 50644 + }, + { + "epoch": 0.50645, + "grad_norm": 0.8430935131858084, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 50645 + }, + { + "epoch": 0.50646, + "grad_norm": 0.8032688960558214, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 50646 + }, + { + "epoch": 0.50647, + "grad_norm": 1.0799756383635046, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 50647 + }, + { + "epoch": 0.50648, + "grad_norm": 1.3386366057282741, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 50648 + }, + { + "epoch": 0.50649, + "grad_norm": 0.8524743771652107, + "learning_rate": 0.003, + "loss": 4.031, + "step": 50649 + }, + { + "epoch": 0.5065, + "grad_norm": 0.9237960482392638, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 50650 + }, + { + "epoch": 0.50651, + "grad_norm": 1.2948834366302113, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 50651 + }, + { + "epoch": 0.50652, + "grad_norm": 0.9866406269396619, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 50652 + }, + { + "epoch": 0.50653, + "grad_norm": 1.0726991758798745, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 50653 + }, + { + "epoch": 0.50654, + "grad_norm": 1.0283687517539055, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 50654 + }, + { + "epoch": 0.50655, + "grad_norm": 1.0306113581853176, + "learning_rate": 0.003, + "loss": 3.992, + "step": 50655 + }, + { + "epoch": 0.50656, + "grad_norm": 1.0661856772234122, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 50656 + }, + { + "epoch": 0.50657, + "grad_norm": 0.9685385441661073, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 50657 + }, + { + "epoch": 0.50658, + "grad_norm": 1.0388026329669737, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 50658 + }, + { + "epoch": 0.50659, + "grad_norm": 0.9912825276566947, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 50659 + }, + { + "epoch": 0.5066, + "grad_norm": 1.1531840391599004, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 50660 + }, + { + "epoch": 0.50661, + "grad_norm": 1.1442385649076936, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 50661 + }, + { + "epoch": 0.50662, + "grad_norm": 1.2132572475709151, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 50662 + }, + { + "epoch": 0.50663, + "grad_norm": 1.2157541314045996, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 50663 + }, + { + "epoch": 0.50664, + "grad_norm": 0.8740317833370072, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 50664 + }, + { + "epoch": 0.50665, + "grad_norm": 1.0344386861992354, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 50665 + }, + { + "epoch": 0.50666, + "grad_norm": 1.1025970723291532, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 50666 + }, + { + "epoch": 0.50667, + "grad_norm": 0.9784204998213792, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 50667 + }, + { + "epoch": 0.50668, + "grad_norm": 1.238365276375865, + "learning_rate": 0.003, + "loss": 4.023, + "step": 50668 + }, + { + "epoch": 0.50669, + "grad_norm": 0.9821296223895705, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 50669 + }, + { + "epoch": 0.5067, + "grad_norm": 0.996452489229408, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 50670 + }, + { + "epoch": 0.50671, + "grad_norm": 1.075210274928592, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 50671 + }, + { + "epoch": 0.50672, + "grad_norm": 1.0022359765977025, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 50672 + }, + { + "epoch": 0.50673, + "grad_norm": 1.3232239523961449, + "learning_rate": 0.003, + "loss": 4.0929, + "step": 50673 + }, + { + "epoch": 0.50674, + "grad_norm": 0.8993846973777808, + "learning_rate": 0.003, + "loss": 4.043, + "step": 50674 + }, + { + "epoch": 0.50675, + "grad_norm": 0.8992367357983029, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 50675 + }, + { + "epoch": 0.50676, + "grad_norm": 1.2605275288149205, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 50676 + }, + { + "epoch": 0.50677, + "grad_norm": 1.177882547202925, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 50677 + }, + { + "epoch": 0.50678, + "grad_norm": 0.90103389324252, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 50678 + }, + { + "epoch": 0.50679, + "grad_norm": 0.9249978937157622, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 50679 + }, + { + "epoch": 0.5068, + "grad_norm": 1.089695556420538, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 50680 + }, + { + "epoch": 0.50681, + "grad_norm": 1.0055041198185672, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 50681 + }, + { + "epoch": 0.50682, + "grad_norm": 1.1150112993140533, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 50682 + }, + { + "epoch": 0.50683, + "grad_norm": 0.9540158913654515, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 50683 + }, + { + "epoch": 0.50684, + "grad_norm": 0.9784112477462678, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 50684 + }, + { + "epoch": 0.50685, + "grad_norm": 1.027054500466185, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 50685 + }, + { + "epoch": 0.50686, + "grad_norm": 0.9291408236608507, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 50686 + }, + { + "epoch": 0.50687, + "grad_norm": 0.9379767415830479, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 50687 + }, + { + "epoch": 0.50688, + "grad_norm": 1.1674587186738596, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 50688 + }, + { + "epoch": 0.50689, + "grad_norm": 0.9422925085808325, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 50689 + }, + { + "epoch": 0.5069, + "grad_norm": 0.9951032469197987, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 50690 + }, + { + "epoch": 0.50691, + "grad_norm": 1.1721001418961776, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 50691 + }, + { + "epoch": 0.50692, + "grad_norm": 0.8999540398812095, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 50692 + }, + { + "epoch": 0.50693, + "grad_norm": 0.9900510665731594, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 50693 + }, + { + "epoch": 0.50694, + "grad_norm": 1.3272803731189502, + "learning_rate": 0.003, + "loss": 4.028, + "step": 50694 + }, + { + "epoch": 0.50695, + "grad_norm": 0.9560355313433591, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 50695 + }, + { + "epoch": 0.50696, + "grad_norm": 1.2238329933322283, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 50696 + }, + { + "epoch": 0.50697, + "grad_norm": 1.2118787827028077, + "learning_rate": 0.003, + "loss": 4.014, + "step": 50697 + }, + { + "epoch": 0.50698, + "grad_norm": 1.1384055458001912, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 50698 + }, + { + "epoch": 0.50699, + "grad_norm": 1.007716417520016, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 50699 + }, + { + "epoch": 0.507, + "grad_norm": 1.0182366698899306, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 50700 + }, + { + "epoch": 0.50701, + "grad_norm": 1.1747022432575307, + "learning_rate": 0.003, + "loss": 4.052, + "step": 50701 + }, + { + "epoch": 0.50702, + "grad_norm": 0.8388528622742787, + "learning_rate": 0.003, + "loss": 4.044, + "step": 50702 + }, + { + "epoch": 0.50703, + "grad_norm": 0.9516609261904854, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 50703 + }, + { + "epoch": 0.50704, + "grad_norm": 0.9587454212354066, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 50704 + }, + { + "epoch": 0.50705, + "grad_norm": 0.9822304490030781, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 50705 + }, + { + "epoch": 0.50706, + "grad_norm": 0.9875228729049559, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 50706 + }, + { + "epoch": 0.50707, + "grad_norm": 1.1849375632346948, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 50707 + }, + { + "epoch": 0.50708, + "grad_norm": 0.9462100319965031, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 50708 + }, + { + "epoch": 0.50709, + "grad_norm": 1.0435838725734528, + "learning_rate": 0.003, + "loss": 4.0908, + "step": 50709 + }, + { + "epoch": 0.5071, + "grad_norm": 1.135313416444869, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 50710 + }, + { + "epoch": 0.50711, + "grad_norm": 1.2103558606317382, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 50711 + }, + { + "epoch": 0.50712, + "grad_norm": 1.1303521540419044, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 50712 + }, + { + "epoch": 0.50713, + "grad_norm": 1.1323561977868704, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 50713 + }, + { + "epoch": 0.50714, + "grad_norm": 0.9064119911656023, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 50714 + }, + { + "epoch": 0.50715, + "grad_norm": 1.0204749087748852, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 50715 + }, + { + "epoch": 0.50716, + "grad_norm": 1.1366679273927054, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 50716 + }, + { + "epoch": 0.50717, + "grad_norm": 1.2004866092721869, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 50717 + }, + { + "epoch": 0.50718, + "grad_norm": 1.3214337156606282, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 50718 + }, + { + "epoch": 0.50719, + "grad_norm": 1.0416021890793015, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 50719 + }, + { + "epoch": 0.5072, + "grad_norm": 1.1586599281746608, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 50720 + }, + { + "epoch": 0.50721, + "grad_norm": 1.1306674184826395, + "learning_rate": 0.003, + "loss": 4.045, + "step": 50721 + }, + { + "epoch": 0.50722, + "grad_norm": 0.9752714537890732, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 50722 + }, + { + "epoch": 0.50723, + "grad_norm": 1.065113917072155, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 50723 + }, + { + "epoch": 0.50724, + "grad_norm": 1.1068087459814804, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 50724 + }, + { + "epoch": 0.50725, + "grad_norm": 1.0955975945584604, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 50725 + }, + { + "epoch": 0.50726, + "grad_norm": 1.106470490662196, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 50726 + }, + { + "epoch": 0.50727, + "grad_norm": 0.9579227608134824, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 50727 + }, + { + "epoch": 0.50728, + "grad_norm": 0.9274206371487114, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 50728 + }, + { + "epoch": 0.50729, + "grad_norm": 0.9353245931323108, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 50729 + }, + { + "epoch": 0.5073, + "grad_norm": 1.1174423199664905, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 50730 + }, + { + "epoch": 0.50731, + "grad_norm": 1.1116506227337202, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 50731 + }, + { + "epoch": 0.50732, + "grad_norm": 0.8598101962081709, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 50732 + }, + { + "epoch": 0.50733, + "grad_norm": 0.8094615199802181, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 50733 + }, + { + "epoch": 0.50734, + "grad_norm": 0.8808168475901543, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 50734 + }, + { + "epoch": 0.50735, + "grad_norm": 1.1419578674362774, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 50735 + }, + { + "epoch": 0.50736, + "grad_norm": 1.1130408390180517, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 50736 + }, + { + "epoch": 0.50737, + "grad_norm": 0.9012011431932898, + "learning_rate": 0.003, + "loss": 4.035, + "step": 50737 + }, + { + "epoch": 0.50738, + "grad_norm": 0.9950444817431198, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 50738 + }, + { + "epoch": 0.50739, + "grad_norm": 1.0519599458319477, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 50739 + }, + { + "epoch": 0.5074, + "grad_norm": 1.1829754955225453, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 50740 + }, + { + "epoch": 0.50741, + "grad_norm": 1.2154221749947793, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 50741 + }, + { + "epoch": 0.50742, + "grad_norm": 1.107336170073259, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 50742 + }, + { + "epoch": 0.50743, + "grad_norm": 0.994583097218631, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 50743 + }, + { + "epoch": 0.50744, + "grad_norm": 1.1877778138038566, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 50744 + }, + { + "epoch": 0.50745, + "grad_norm": 0.8579264199168228, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 50745 + }, + { + "epoch": 0.50746, + "grad_norm": 1.0423039448025693, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 50746 + }, + { + "epoch": 0.50747, + "grad_norm": 1.2696910081162749, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 50747 + }, + { + "epoch": 0.50748, + "grad_norm": 0.760265308420248, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 50748 + }, + { + "epoch": 0.50749, + "grad_norm": 0.9503108718501894, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 50749 + }, + { + "epoch": 0.5075, + "grad_norm": 1.2717159622994114, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 50750 + }, + { + "epoch": 0.50751, + "grad_norm": 0.8473050344546443, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 50751 + }, + { + "epoch": 0.50752, + "grad_norm": 1.0291063001072034, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 50752 + }, + { + "epoch": 0.50753, + "grad_norm": 1.6368286450921825, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 50753 + }, + { + "epoch": 0.50754, + "grad_norm": 1.055330182364674, + "learning_rate": 0.003, + "loss": 4.027, + "step": 50754 + }, + { + "epoch": 0.50755, + "grad_norm": 1.0012694023668145, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 50755 + }, + { + "epoch": 0.50756, + "grad_norm": 1.1396692420714614, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 50756 + }, + { + "epoch": 0.50757, + "grad_norm": 1.1395353757145894, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 50757 + }, + { + "epoch": 0.50758, + "grad_norm": 1.1662081918516498, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 50758 + }, + { + "epoch": 0.50759, + "grad_norm": 1.1145415274921564, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 50759 + }, + { + "epoch": 0.5076, + "grad_norm": 0.9563189763074288, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 50760 + }, + { + "epoch": 0.50761, + "grad_norm": 0.9688278308585538, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 50761 + }, + { + "epoch": 0.50762, + "grad_norm": 0.9940440331501476, + "learning_rate": 0.003, + "loss": 4.058, + "step": 50762 + }, + { + "epoch": 0.50763, + "grad_norm": 1.0471808925083501, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 50763 + }, + { + "epoch": 0.50764, + "grad_norm": 1.0695012393404717, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 50764 + }, + { + "epoch": 0.50765, + "grad_norm": 1.1827471476778928, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 50765 + }, + { + "epoch": 0.50766, + "grad_norm": 1.096679366904232, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 50766 + }, + { + "epoch": 0.50767, + "grad_norm": 1.0796263826410402, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 50767 + }, + { + "epoch": 0.50768, + "grad_norm": 1.2746768929073453, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 50768 + }, + { + "epoch": 0.50769, + "grad_norm": 1.0416829110458592, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 50769 + }, + { + "epoch": 0.5077, + "grad_norm": 1.0875998318126374, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 50770 + }, + { + "epoch": 0.50771, + "grad_norm": 0.9315207062590201, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 50771 + }, + { + "epoch": 0.50772, + "grad_norm": 0.882796560485162, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 50772 + }, + { + "epoch": 0.50773, + "grad_norm": 0.98128179228831, + "learning_rate": 0.003, + "loss": 4.037, + "step": 50773 + }, + { + "epoch": 0.50774, + "grad_norm": 1.2736903538286757, + "learning_rate": 0.003, + "loss": 4.047, + "step": 50774 + }, + { + "epoch": 0.50775, + "grad_norm": 1.1810352201973422, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 50775 + }, + { + "epoch": 0.50776, + "grad_norm": 0.98058656796798, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 50776 + }, + { + "epoch": 0.50777, + "grad_norm": 1.110581595835357, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 50777 + }, + { + "epoch": 0.50778, + "grad_norm": 1.260915687571586, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 50778 + }, + { + "epoch": 0.50779, + "grad_norm": 0.926822826395, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 50779 + }, + { + "epoch": 0.5078, + "grad_norm": 0.9857734785260185, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 50780 + }, + { + "epoch": 0.50781, + "grad_norm": 0.8901884520715182, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 50781 + }, + { + "epoch": 0.50782, + "grad_norm": 0.9362988728720161, + "learning_rate": 0.003, + "loss": 4.038, + "step": 50782 + }, + { + "epoch": 0.50783, + "grad_norm": 0.9382748201600734, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 50783 + }, + { + "epoch": 0.50784, + "grad_norm": 0.9960448612538488, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 50784 + }, + { + "epoch": 0.50785, + "grad_norm": 1.0223871808266458, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 50785 + }, + { + "epoch": 0.50786, + "grad_norm": 1.1496699240635329, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 50786 + }, + { + "epoch": 0.50787, + "grad_norm": 1.1148638507545274, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 50787 + }, + { + "epoch": 0.50788, + "grad_norm": 1.309192000774339, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 50788 + }, + { + "epoch": 0.50789, + "grad_norm": 0.8930268391172596, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 50789 + }, + { + "epoch": 0.5079, + "grad_norm": 0.9168085794700639, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 50790 + }, + { + "epoch": 0.50791, + "grad_norm": 1.0700042762362914, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 50791 + }, + { + "epoch": 0.50792, + "grad_norm": 1.3621305916574589, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 50792 + }, + { + "epoch": 0.50793, + "grad_norm": 0.7633372274788218, + "learning_rate": 0.003, + "loss": 4.015, + "step": 50793 + }, + { + "epoch": 0.50794, + "grad_norm": 0.831643969760896, + "learning_rate": 0.003, + "loss": 4.008, + "step": 50794 + }, + { + "epoch": 0.50795, + "grad_norm": 0.9815915190088659, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 50795 + }, + { + "epoch": 0.50796, + "grad_norm": 1.550142787660606, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 50796 + }, + { + "epoch": 0.50797, + "grad_norm": 0.6242266089814553, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 50797 + }, + { + "epoch": 0.50798, + "grad_norm": 0.7978297077492295, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 50798 + }, + { + "epoch": 0.50799, + "grad_norm": 1.0716766378644724, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 50799 + }, + { + "epoch": 0.508, + "grad_norm": 1.2661975808835733, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 50800 + }, + { + "epoch": 0.50801, + "grad_norm": 0.8230848947301784, + "learning_rate": 0.003, + "loss": 4.049, + "step": 50801 + }, + { + "epoch": 0.50802, + "grad_norm": 0.8101683448800386, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 50802 + }, + { + "epoch": 0.50803, + "grad_norm": 0.8875974363772142, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 50803 + }, + { + "epoch": 0.50804, + "grad_norm": 1.0005794817275944, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 50804 + }, + { + "epoch": 0.50805, + "grad_norm": 1.107414933782095, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 50805 + }, + { + "epoch": 0.50806, + "grad_norm": 1.032932530209, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 50806 + }, + { + "epoch": 0.50807, + "grad_norm": 1.3345884065211953, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 50807 + }, + { + "epoch": 0.50808, + "grad_norm": 0.9275268446568755, + "learning_rate": 0.003, + "loss": 4.055, + "step": 50808 + }, + { + "epoch": 0.50809, + "grad_norm": 1.0765206454835659, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 50809 + }, + { + "epoch": 0.5081, + "grad_norm": 1.270749461942428, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 50810 + }, + { + "epoch": 0.50811, + "grad_norm": 0.963820990947602, + "learning_rate": 0.003, + "loss": 4.015, + "step": 50811 + }, + { + "epoch": 0.50812, + "grad_norm": 1.045490005043709, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 50812 + }, + { + "epoch": 0.50813, + "grad_norm": 1.296704006310593, + "learning_rate": 0.003, + "loss": 4.046, + "step": 50813 + }, + { + "epoch": 0.50814, + "grad_norm": 0.8059067357135348, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 50814 + }, + { + "epoch": 0.50815, + "grad_norm": 0.8353253081291591, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 50815 + }, + { + "epoch": 0.50816, + "grad_norm": 0.9906552743917868, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 50816 + }, + { + "epoch": 0.50817, + "grad_norm": 1.0627315270726478, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 50817 + }, + { + "epoch": 0.50818, + "grad_norm": 1.0357337736297048, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 50818 + }, + { + "epoch": 0.50819, + "grad_norm": 1.2367650841428333, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 50819 + }, + { + "epoch": 0.5082, + "grad_norm": 0.9507081595607314, + "learning_rate": 0.003, + "loss": 4.021, + "step": 50820 + }, + { + "epoch": 0.50821, + "grad_norm": 1.256094233088149, + "learning_rate": 0.003, + "loss": 4.038, + "step": 50821 + }, + { + "epoch": 0.50822, + "grad_norm": 1.3110985591146256, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 50822 + }, + { + "epoch": 0.50823, + "grad_norm": 0.8564968665670416, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 50823 + }, + { + "epoch": 0.50824, + "grad_norm": 1.0478630985804602, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 50824 + }, + { + "epoch": 0.50825, + "grad_norm": 1.287135014634666, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 50825 + }, + { + "epoch": 0.50826, + "grad_norm": 0.952987369658974, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 50826 + }, + { + "epoch": 0.50827, + "grad_norm": 0.9359623198319867, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 50827 + }, + { + "epoch": 0.50828, + "grad_norm": 0.9692148929858284, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 50828 + }, + { + "epoch": 0.50829, + "grad_norm": 1.101336328688301, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 50829 + }, + { + "epoch": 0.5083, + "grad_norm": 0.7668052621308992, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 50830 + }, + { + "epoch": 0.50831, + "grad_norm": 0.9872587537164815, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 50831 + }, + { + "epoch": 0.50832, + "grad_norm": 1.5743097090681912, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 50832 + }, + { + "epoch": 0.50833, + "grad_norm": 0.6301717523214184, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 50833 + }, + { + "epoch": 0.50834, + "grad_norm": 0.7503115130296564, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 50834 + }, + { + "epoch": 0.50835, + "grad_norm": 0.9756712337557755, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 50835 + }, + { + "epoch": 0.50836, + "grad_norm": 1.3454590261042372, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 50836 + }, + { + "epoch": 0.50837, + "grad_norm": 1.0933383118313706, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 50837 + }, + { + "epoch": 0.50838, + "grad_norm": 0.9747648688227892, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 50838 + }, + { + "epoch": 0.50839, + "grad_norm": 0.9805004464708592, + "learning_rate": 0.003, + "loss": 4.014, + "step": 50839 + }, + { + "epoch": 0.5084, + "grad_norm": 1.3050687476377822, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 50840 + }, + { + "epoch": 0.50841, + "grad_norm": 0.8226402031028061, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 50841 + }, + { + "epoch": 0.50842, + "grad_norm": 0.875010617714495, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 50842 + }, + { + "epoch": 0.50843, + "grad_norm": 0.9449869951821404, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 50843 + }, + { + "epoch": 0.50844, + "grad_norm": 1.0146657866387747, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 50844 + }, + { + "epoch": 0.50845, + "grad_norm": 1.0858118487070627, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 50845 + }, + { + "epoch": 0.50846, + "grad_norm": 0.993917350309296, + "learning_rate": 0.003, + "loss": 4.011, + "step": 50846 + }, + { + "epoch": 0.50847, + "grad_norm": 1.12580556652038, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 50847 + }, + { + "epoch": 0.50848, + "grad_norm": 1.1064486649031842, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 50848 + }, + { + "epoch": 0.50849, + "grad_norm": 1.0280731007941426, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 50849 + }, + { + "epoch": 0.5085, + "grad_norm": 1.1162586225490052, + "learning_rate": 0.003, + "loss": 4.013, + "step": 50850 + }, + { + "epoch": 0.50851, + "grad_norm": 1.0978599793155888, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 50851 + }, + { + "epoch": 0.50852, + "grad_norm": 0.8391387304814995, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 50852 + }, + { + "epoch": 0.50853, + "grad_norm": 0.9673802510434403, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 50853 + }, + { + "epoch": 0.50854, + "grad_norm": 1.0702588839624116, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 50854 + }, + { + "epoch": 0.50855, + "grad_norm": 1.1546824647988685, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 50855 + }, + { + "epoch": 0.50856, + "grad_norm": 0.9993310856463392, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 50856 + }, + { + "epoch": 0.50857, + "grad_norm": 1.20289159895422, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 50857 + }, + { + "epoch": 0.50858, + "grad_norm": 1.3437839873850428, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 50858 + }, + { + "epoch": 0.50859, + "grad_norm": 1.1385346316064977, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 50859 + }, + { + "epoch": 0.5086, + "grad_norm": 1.07702521478228, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 50860 + }, + { + "epoch": 0.50861, + "grad_norm": 1.118200032546152, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 50861 + }, + { + "epoch": 0.50862, + "grad_norm": 1.1238457758236884, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 50862 + }, + { + "epoch": 0.50863, + "grad_norm": 0.9602897668485096, + "learning_rate": 0.003, + "loss": 4.027, + "step": 50863 + }, + { + "epoch": 0.50864, + "grad_norm": 1.0944028440569424, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 50864 + }, + { + "epoch": 0.50865, + "grad_norm": 1.062691720035527, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 50865 + }, + { + "epoch": 0.50866, + "grad_norm": 1.0645599594127055, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 50866 + }, + { + "epoch": 0.50867, + "grad_norm": 0.9502856921243356, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 50867 + }, + { + "epoch": 0.50868, + "grad_norm": 0.9587800791084288, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 50868 + }, + { + "epoch": 0.50869, + "grad_norm": 1.1827620355793118, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 50869 + }, + { + "epoch": 0.5087, + "grad_norm": 1.0374144475576697, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 50870 + }, + { + "epoch": 0.50871, + "grad_norm": 1.0567230260914187, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 50871 + }, + { + "epoch": 0.50872, + "grad_norm": 1.0798253465213206, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 50872 + }, + { + "epoch": 0.50873, + "grad_norm": 1.0453441306162008, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 50873 + }, + { + "epoch": 0.50874, + "grad_norm": 1.1020848783272157, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 50874 + }, + { + "epoch": 0.50875, + "grad_norm": 1.018790908582407, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 50875 + }, + { + "epoch": 0.50876, + "grad_norm": 1.197627353351876, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 50876 + }, + { + "epoch": 0.50877, + "grad_norm": 0.9207123561504924, + "learning_rate": 0.003, + "loss": 4.032, + "step": 50877 + }, + { + "epoch": 0.50878, + "grad_norm": 0.9191323627483747, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 50878 + }, + { + "epoch": 0.50879, + "grad_norm": 1.0042890847450863, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 50879 + }, + { + "epoch": 0.5088, + "grad_norm": 1.2391579283059038, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 50880 + }, + { + "epoch": 0.50881, + "grad_norm": 1.1094115685010675, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 50881 + }, + { + "epoch": 0.50882, + "grad_norm": 1.1337905144938625, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 50882 + }, + { + "epoch": 0.50883, + "grad_norm": 0.9652974731860895, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 50883 + }, + { + "epoch": 0.50884, + "grad_norm": 1.0646462800997973, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 50884 + }, + { + "epoch": 0.50885, + "grad_norm": 1.1776439251955229, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 50885 + }, + { + "epoch": 0.50886, + "grad_norm": 1.0963278094156148, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 50886 + }, + { + "epoch": 0.50887, + "grad_norm": 1.016126409366749, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 50887 + }, + { + "epoch": 0.50888, + "grad_norm": 1.2007616171119118, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 50888 + }, + { + "epoch": 0.50889, + "grad_norm": 1.0105902124491093, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 50889 + }, + { + "epoch": 0.5089, + "grad_norm": 1.2786768632970413, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 50890 + }, + { + "epoch": 0.50891, + "grad_norm": 0.8256146319493582, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 50891 + }, + { + "epoch": 0.50892, + "grad_norm": 0.7226191939967103, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 50892 + }, + { + "epoch": 0.50893, + "grad_norm": 0.858670265952443, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 50893 + }, + { + "epoch": 0.50894, + "grad_norm": 0.8886880572416408, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 50894 + }, + { + "epoch": 0.50895, + "grad_norm": 1.113809245565595, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 50895 + }, + { + "epoch": 0.50896, + "grad_norm": 1.2734335721531058, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 50896 + }, + { + "epoch": 0.50897, + "grad_norm": 1.1462516717707174, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 50897 + }, + { + "epoch": 0.50898, + "grad_norm": 1.0008805962783427, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 50898 + }, + { + "epoch": 0.50899, + "grad_norm": 1.2184569369974456, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 50899 + }, + { + "epoch": 0.509, + "grad_norm": 1.0849130774246114, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 50900 + }, + { + "epoch": 0.50901, + "grad_norm": 1.0160227308161618, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 50901 + }, + { + "epoch": 0.50902, + "grad_norm": 0.9500186683839599, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 50902 + }, + { + "epoch": 0.50903, + "grad_norm": 0.8758909374038267, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 50903 + }, + { + "epoch": 0.50904, + "grad_norm": 0.908739090382015, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 50904 + }, + { + "epoch": 0.50905, + "grad_norm": 0.9938600755620762, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 50905 + }, + { + "epoch": 0.50906, + "grad_norm": 1.070928884862748, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 50906 + }, + { + "epoch": 0.50907, + "grad_norm": 0.9774540766012811, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 50907 + }, + { + "epoch": 0.50908, + "grad_norm": 1.1357838393643274, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 50908 + }, + { + "epoch": 0.50909, + "grad_norm": 1.1430339587632565, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 50909 + }, + { + "epoch": 0.5091, + "grad_norm": 1.0080006907536085, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 50910 + }, + { + "epoch": 0.50911, + "grad_norm": 1.0818688329464792, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 50911 + }, + { + "epoch": 0.50912, + "grad_norm": 1.281588824087215, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 50912 + }, + { + "epoch": 0.50913, + "grad_norm": 0.922543474627782, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 50913 + }, + { + "epoch": 0.50914, + "grad_norm": 1.0040073178388076, + "learning_rate": 0.003, + "loss": 4.043, + "step": 50914 + }, + { + "epoch": 0.50915, + "grad_norm": 1.1090372082403024, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 50915 + }, + { + "epoch": 0.50916, + "grad_norm": 1.019043885239996, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 50916 + }, + { + "epoch": 0.50917, + "grad_norm": 0.9721886007251364, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 50917 + }, + { + "epoch": 0.50918, + "grad_norm": 1.1233770606437528, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 50918 + }, + { + "epoch": 0.50919, + "grad_norm": 1.217935460123561, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 50919 + }, + { + "epoch": 0.5092, + "grad_norm": 0.8264342654733495, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 50920 + }, + { + "epoch": 0.50921, + "grad_norm": 0.8160753832725494, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 50921 + }, + { + "epoch": 0.50922, + "grad_norm": 0.9499911343674083, + "learning_rate": 0.003, + "loss": 4.029, + "step": 50922 + }, + { + "epoch": 0.50923, + "grad_norm": 1.032061675141883, + "learning_rate": 0.003, + "loss": 4.066, + "step": 50923 + }, + { + "epoch": 0.50924, + "grad_norm": 1.2329926294143903, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 50924 + }, + { + "epoch": 0.50925, + "grad_norm": 1.384809118491773, + "learning_rate": 0.003, + "loss": 4.031, + "step": 50925 + }, + { + "epoch": 0.50926, + "grad_norm": 0.8446938458078708, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 50926 + }, + { + "epoch": 0.50927, + "grad_norm": 0.7346847195122553, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 50927 + }, + { + "epoch": 0.50928, + "grad_norm": 0.9176762506444948, + "learning_rate": 0.003, + "loss": 4.052, + "step": 50928 + }, + { + "epoch": 0.50929, + "grad_norm": 1.3909610703508537, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 50929 + }, + { + "epoch": 0.5093, + "grad_norm": 0.9825601321808193, + "learning_rate": 0.003, + "loss": 4.04, + "step": 50930 + }, + { + "epoch": 0.50931, + "grad_norm": 1.2791806149782723, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 50931 + }, + { + "epoch": 0.50932, + "grad_norm": 0.9636304221929521, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 50932 + }, + { + "epoch": 0.50933, + "grad_norm": 0.8111517835485068, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 50933 + }, + { + "epoch": 0.50934, + "grad_norm": 0.8405842489578675, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 50934 + }, + { + "epoch": 0.50935, + "grad_norm": 0.9653521612324485, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 50935 + }, + { + "epoch": 0.50936, + "grad_norm": 1.0861410140112011, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 50936 + }, + { + "epoch": 0.50937, + "grad_norm": 0.9965993735139247, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 50937 + }, + { + "epoch": 0.50938, + "grad_norm": 1.2114068110648422, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 50938 + }, + { + "epoch": 0.50939, + "grad_norm": 1.0860284211370728, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 50939 + }, + { + "epoch": 0.5094, + "grad_norm": 1.2550426859433532, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 50940 + }, + { + "epoch": 0.50941, + "grad_norm": 0.9180343289122953, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 50941 + }, + { + "epoch": 0.50942, + "grad_norm": 0.8038828312338612, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 50942 + }, + { + "epoch": 0.50943, + "grad_norm": 0.8043266287516665, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 50943 + }, + { + "epoch": 0.50944, + "grad_norm": 0.884047086927419, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 50944 + }, + { + "epoch": 0.50945, + "grad_norm": 1.0129808229856818, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 50945 + }, + { + "epoch": 0.50946, + "grad_norm": 1.2460068046104984, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 50946 + }, + { + "epoch": 0.50947, + "grad_norm": 1.049416120008444, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 50947 + }, + { + "epoch": 0.50948, + "grad_norm": 1.052463014548424, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 50948 + }, + { + "epoch": 0.50949, + "grad_norm": 1.0327150429055831, + "learning_rate": 0.003, + "loss": 4.043, + "step": 50949 + }, + { + "epoch": 0.5095, + "grad_norm": 1.1241153175490979, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 50950 + }, + { + "epoch": 0.50951, + "grad_norm": 1.0930068077514967, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 50951 + }, + { + "epoch": 0.50952, + "grad_norm": 0.9260763418165958, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 50952 + }, + { + "epoch": 0.50953, + "grad_norm": 0.8023832747645115, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 50953 + }, + { + "epoch": 0.50954, + "grad_norm": 0.8273857044746695, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 50954 + }, + { + "epoch": 0.50955, + "grad_norm": 0.9049235075730795, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 50955 + }, + { + "epoch": 0.50956, + "grad_norm": 0.9284686731962366, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 50956 + }, + { + "epoch": 0.50957, + "grad_norm": 1.031560273025396, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 50957 + }, + { + "epoch": 0.50958, + "grad_norm": 1.2739813561983604, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 50958 + }, + { + "epoch": 0.50959, + "grad_norm": 0.985091416220845, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 50959 + }, + { + "epoch": 0.5096, + "grad_norm": 1.1944074970019467, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 50960 + }, + { + "epoch": 0.50961, + "grad_norm": 1.1841946954307114, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 50961 + }, + { + "epoch": 0.50962, + "grad_norm": 1.0576597084872588, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 50962 + }, + { + "epoch": 0.50963, + "grad_norm": 1.1396383154714942, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 50963 + }, + { + "epoch": 0.50964, + "grad_norm": 0.9860716620273997, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 50964 + }, + { + "epoch": 0.50965, + "grad_norm": 1.2241988539482007, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 50965 + }, + { + "epoch": 0.50966, + "grad_norm": 1.12222883641216, + "learning_rate": 0.003, + "loss": 4.047, + "step": 50966 + }, + { + "epoch": 0.50967, + "grad_norm": 0.9004161301068699, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 50967 + }, + { + "epoch": 0.50968, + "grad_norm": 0.9044834004006748, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 50968 + }, + { + "epoch": 0.50969, + "grad_norm": 1.0950427730655192, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 50969 + }, + { + "epoch": 0.5097, + "grad_norm": 1.0446563195861085, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 50970 + }, + { + "epoch": 0.50971, + "grad_norm": 1.0018905213191776, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 50971 + }, + { + "epoch": 0.50972, + "grad_norm": 1.2545068600704001, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 50972 + }, + { + "epoch": 0.50973, + "grad_norm": 1.0095565532460382, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 50973 + }, + { + "epoch": 0.50974, + "grad_norm": 1.1442284978945205, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 50974 + }, + { + "epoch": 0.50975, + "grad_norm": 0.9510999111076139, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 50975 + }, + { + "epoch": 0.50976, + "grad_norm": 1.3194392977816656, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 50976 + }, + { + "epoch": 0.50977, + "grad_norm": 1.1028533163226146, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 50977 + }, + { + "epoch": 0.50978, + "grad_norm": 1.1438416586935216, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 50978 + }, + { + "epoch": 0.50979, + "grad_norm": 1.0629319112984035, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 50979 + }, + { + "epoch": 0.5098, + "grad_norm": 1.0180100855144552, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 50980 + }, + { + "epoch": 0.50981, + "grad_norm": 1.056219331321115, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 50981 + }, + { + "epoch": 0.50982, + "grad_norm": 1.2209648153813728, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 50982 + }, + { + "epoch": 0.50983, + "grad_norm": 0.9222914834678758, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 50983 + }, + { + "epoch": 0.50984, + "grad_norm": 1.1737037303498428, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 50984 + }, + { + "epoch": 0.50985, + "grad_norm": 0.994727611035997, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 50985 + }, + { + "epoch": 0.50986, + "grad_norm": 1.0914841526883698, + "learning_rate": 0.003, + "loss": 4.028, + "step": 50986 + }, + { + "epoch": 0.50987, + "grad_norm": 1.0827077007675978, + "learning_rate": 0.003, + "loss": 4.034, + "step": 50987 + }, + { + "epoch": 0.50988, + "grad_norm": 1.1191788307501394, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 50988 + }, + { + "epoch": 0.50989, + "grad_norm": 1.0295824264768456, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 50989 + }, + { + "epoch": 0.5099, + "grad_norm": 1.1103033960897501, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 50990 + }, + { + "epoch": 0.50991, + "grad_norm": 0.9491207008193986, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 50991 + }, + { + "epoch": 0.50992, + "grad_norm": 1.0661966926181468, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 50992 + }, + { + "epoch": 0.50993, + "grad_norm": 1.2955267601395308, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 50993 + }, + { + "epoch": 0.50994, + "grad_norm": 1.0539812687316261, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 50994 + }, + { + "epoch": 0.50995, + "grad_norm": 1.1448029510483377, + "learning_rate": 0.003, + "loss": 4.016, + "step": 50995 + }, + { + "epoch": 0.50996, + "grad_norm": 1.0023560479434843, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 50996 + }, + { + "epoch": 0.50997, + "grad_norm": 1.153967411475945, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 50997 + }, + { + "epoch": 0.50998, + "grad_norm": 1.0047348373233709, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 50998 + }, + { + "epoch": 0.50999, + "grad_norm": 1.0323127533243264, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 50999 + }, + { + "epoch": 0.51, + "grad_norm": 1.0624711216492115, + "learning_rate": 0.003, + "loss": 4.033, + "step": 51000 + }, + { + "epoch": 0.51001, + "grad_norm": 1.0919496422946053, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 51001 + }, + { + "epoch": 0.51002, + "grad_norm": 1.1066082003659843, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 51002 + }, + { + "epoch": 0.51003, + "grad_norm": 0.9154743309007551, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 51003 + }, + { + "epoch": 0.51004, + "grad_norm": 0.9244771510097893, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 51004 + }, + { + "epoch": 0.51005, + "grad_norm": 0.9699250934095232, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 51005 + }, + { + "epoch": 0.51006, + "grad_norm": 1.3254613602351615, + "learning_rate": 0.003, + "loss": 4.015, + "step": 51006 + }, + { + "epoch": 0.51007, + "grad_norm": 1.1432220242220303, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 51007 + }, + { + "epoch": 0.51008, + "grad_norm": 0.9221948503113004, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 51008 + }, + { + "epoch": 0.51009, + "grad_norm": 0.8902792411972812, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 51009 + }, + { + "epoch": 0.5101, + "grad_norm": 1.1175777533503917, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 51010 + }, + { + "epoch": 0.51011, + "grad_norm": 1.0221923653010494, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 51011 + }, + { + "epoch": 0.51012, + "grad_norm": 1.0336145813960849, + "learning_rate": 0.003, + "loss": 4.036, + "step": 51012 + }, + { + "epoch": 0.51013, + "grad_norm": 1.0554842949431829, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 51013 + }, + { + "epoch": 0.51014, + "grad_norm": 1.1398686440679975, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 51014 + }, + { + "epoch": 0.51015, + "grad_norm": 0.9899611350212034, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 51015 + }, + { + "epoch": 0.51016, + "grad_norm": 0.9756458255624163, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 51016 + }, + { + "epoch": 0.51017, + "grad_norm": 1.1853005997848014, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 51017 + }, + { + "epoch": 0.51018, + "grad_norm": 0.9266145489453641, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 51018 + }, + { + "epoch": 0.51019, + "grad_norm": 0.8527304287141395, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 51019 + }, + { + "epoch": 0.5102, + "grad_norm": 0.9065489811861597, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 51020 + }, + { + "epoch": 0.51021, + "grad_norm": 1.1859389343658233, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 51021 + }, + { + "epoch": 0.51022, + "grad_norm": 1.0579408618458186, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 51022 + }, + { + "epoch": 0.51023, + "grad_norm": 1.1916434754225251, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 51023 + }, + { + "epoch": 0.51024, + "grad_norm": 1.0321444186056352, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 51024 + }, + { + "epoch": 0.51025, + "grad_norm": 1.0237702048613346, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 51025 + }, + { + "epoch": 0.51026, + "grad_norm": 0.9454487848063313, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 51026 + }, + { + "epoch": 0.51027, + "grad_norm": 1.1239986849479333, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 51027 + }, + { + "epoch": 0.51028, + "grad_norm": 0.9968924072947377, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 51028 + }, + { + "epoch": 0.51029, + "grad_norm": 1.285069729336539, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 51029 + }, + { + "epoch": 0.5103, + "grad_norm": 0.8710439358219687, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 51030 + }, + { + "epoch": 0.51031, + "grad_norm": 0.9653745412416551, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 51031 + }, + { + "epoch": 0.51032, + "grad_norm": 1.1013694576190571, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 51032 + }, + { + "epoch": 0.51033, + "grad_norm": 1.3167843437422222, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 51033 + }, + { + "epoch": 0.51034, + "grad_norm": 0.8192167099585168, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 51034 + }, + { + "epoch": 0.51035, + "grad_norm": 0.8801954202602148, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 51035 + }, + { + "epoch": 0.51036, + "grad_norm": 1.1958174537779358, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 51036 + }, + { + "epoch": 0.51037, + "grad_norm": 1.066657583715853, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 51037 + }, + { + "epoch": 0.51038, + "grad_norm": 1.2160707221103826, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 51038 + }, + { + "epoch": 0.51039, + "grad_norm": 0.8806517335121765, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 51039 + }, + { + "epoch": 0.5104, + "grad_norm": 0.8811757697569901, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 51040 + }, + { + "epoch": 0.51041, + "grad_norm": 0.9379360571684101, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 51041 + }, + { + "epoch": 0.51042, + "grad_norm": 1.1932557338501666, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 51042 + }, + { + "epoch": 0.51043, + "grad_norm": 0.9673233137196688, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 51043 + }, + { + "epoch": 0.51044, + "grad_norm": 1.0384211696641976, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 51044 + }, + { + "epoch": 0.51045, + "grad_norm": 1.1428024879087977, + "learning_rate": 0.003, + "loss": 4.03, + "step": 51045 + }, + { + "epoch": 0.51046, + "grad_norm": 1.1605506974705218, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 51046 + }, + { + "epoch": 0.51047, + "grad_norm": 0.9838426198242718, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 51047 + }, + { + "epoch": 0.51048, + "grad_norm": 1.3510475061117844, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 51048 + }, + { + "epoch": 0.51049, + "grad_norm": 0.9673509645669657, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 51049 + }, + { + "epoch": 0.5105, + "grad_norm": 1.065981085098261, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 51050 + }, + { + "epoch": 0.51051, + "grad_norm": 1.1644058930279613, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 51051 + }, + { + "epoch": 0.51052, + "grad_norm": 1.0718271129842105, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 51052 + }, + { + "epoch": 0.51053, + "grad_norm": 0.9088868759099541, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 51053 + }, + { + "epoch": 0.51054, + "grad_norm": 1.0462518414732327, + "learning_rate": 0.003, + "loss": 4.018, + "step": 51054 + }, + { + "epoch": 0.51055, + "grad_norm": 1.1467387008501955, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 51055 + }, + { + "epoch": 0.51056, + "grad_norm": 1.356159918756247, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 51056 + }, + { + "epoch": 0.51057, + "grad_norm": 0.8891127863406791, + "learning_rate": 0.003, + "loss": 4.028, + "step": 51057 + }, + { + "epoch": 0.51058, + "grad_norm": 0.8863968831896331, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 51058 + }, + { + "epoch": 0.51059, + "grad_norm": 1.0131938195349237, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 51059 + }, + { + "epoch": 0.5106, + "grad_norm": 1.204551278395701, + "learning_rate": 0.003, + "loss": 4.035, + "step": 51060 + }, + { + "epoch": 0.51061, + "grad_norm": 1.109816280089387, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 51061 + }, + { + "epoch": 0.51062, + "grad_norm": 1.1657327416995196, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 51062 + }, + { + "epoch": 0.51063, + "grad_norm": 0.9149251398746682, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 51063 + }, + { + "epoch": 0.51064, + "grad_norm": 1.1135929319326316, + "learning_rate": 0.003, + "loss": 4.018, + "step": 51064 + }, + { + "epoch": 0.51065, + "grad_norm": 1.0766544752063922, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 51065 + }, + { + "epoch": 0.51066, + "grad_norm": 0.9290814315419781, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 51066 + }, + { + "epoch": 0.51067, + "grad_norm": 1.1428512171939025, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 51067 + }, + { + "epoch": 0.51068, + "grad_norm": 1.109690154182945, + "learning_rate": 0.003, + "loss": 4.014, + "step": 51068 + }, + { + "epoch": 0.51069, + "grad_norm": 1.0045053084525448, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 51069 + }, + { + "epoch": 0.5107, + "grad_norm": 1.1239207252696146, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 51070 + }, + { + "epoch": 0.51071, + "grad_norm": 1.2848321831091258, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 51071 + }, + { + "epoch": 0.51072, + "grad_norm": 0.8824895237775626, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 51072 + }, + { + "epoch": 0.51073, + "grad_norm": 0.980722024426876, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 51073 + }, + { + "epoch": 0.51074, + "grad_norm": 1.2686806734634453, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 51074 + }, + { + "epoch": 0.51075, + "grad_norm": 0.8447363978926213, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 51075 + }, + { + "epoch": 0.51076, + "grad_norm": 1.0077249037164027, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 51076 + }, + { + "epoch": 0.51077, + "grad_norm": 1.2834424366460206, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 51077 + }, + { + "epoch": 0.51078, + "grad_norm": 1.1405422561382632, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 51078 + }, + { + "epoch": 0.51079, + "grad_norm": 1.1548714086646372, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 51079 + }, + { + "epoch": 0.5108, + "grad_norm": 1.127475981876333, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 51080 + }, + { + "epoch": 0.51081, + "grad_norm": 0.9362952160560696, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 51081 + }, + { + "epoch": 0.51082, + "grad_norm": 1.0081768054251892, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 51082 + }, + { + "epoch": 0.51083, + "grad_norm": 1.1109821658574344, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 51083 + }, + { + "epoch": 0.51084, + "grad_norm": 1.057799077574275, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 51084 + }, + { + "epoch": 0.51085, + "grad_norm": 1.0380892341437158, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 51085 + }, + { + "epoch": 0.51086, + "grad_norm": 0.8905017337371668, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 51086 + }, + { + "epoch": 0.51087, + "grad_norm": 0.9147521125127781, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 51087 + }, + { + "epoch": 0.51088, + "grad_norm": 1.2201425086597446, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 51088 + }, + { + "epoch": 0.51089, + "grad_norm": 0.9821567587833065, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 51089 + }, + { + "epoch": 0.5109, + "grad_norm": 1.218046168805865, + "learning_rate": 0.003, + "loss": 4.01, + "step": 51090 + }, + { + "epoch": 0.51091, + "grad_norm": 1.0097921801245846, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 51091 + }, + { + "epoch": 0.51092, + "grad_norm": 1.4871744449094504, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 51092 + }, + { + "epoch": 0.51093, + "grad_norm": 0.6571676099311771, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 51093 + }, + { + "epoch": 0.51094, + "grad_norm": 0.8095418028705201, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 51094 + }, + { + "epoch": 0.51095, + "grad_norm": 1.1818738488906542, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 51095 + }, + { + "epoch": 0.51096, + "grad_norm": 0.8922447059324082, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 51096 + }, + { + "epoch": 0.51097, + "grad_norm": 0.9872078487506909, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 51097 + }, + { + "epoch": 0.51098, + "grad_norm": 1.1369257342908834, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 51098 + }, + { + "epoch": 0.51099, + "grad_norm": 0.9921758182201071, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 51099 + }, + { + "epoch": 0.511, + "grad_norm": 1.2238780154352202, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 51100 + }, + { + "epoch": 0.51101, + "grad_norm": 0.9274927765844205, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 51101 + }, + { + "epoch": 0.51102, + "grad_norm": 0.8699714356917674, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 51102 + }, + { + "epoch": 0.51103, + "grad_norm": 0.9399482926015219, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 51103 + }, + { + "epoch": 0.51104, + "grad_norm": 1.0580884183449337, + "learning_rate": 0.003, + "loss": 4.051, + "step": 51104 + }, + { + "epoch": 0.51105, + "grad_norm": 1.07827612009929, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 51105 + }, + { + "epoch": 0.51106, + "grad_norm": 1.0165281624245783, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 51106 + }, + { + "epoch": 0.51107, + "grad_norm": 0.9342661687313454, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 51107 + }, + { + "epoch": 0.51108, + "grad_norm": 1.1643577338434057, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 51108 + }, + { + "epoch": 0.51109, + "grad_norm": 1.057501913810358, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 51109 + }, + { + "epoch": 0.5111, + "grad_norm": 1.159394233240555, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 51110 + }, + { + "epoch": 0.51111, + "grad_norm": 1.2324445755786804, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 51111 + }, + { + "epoch": 0.51112, + "grad_norm": 1.007376417781725, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 51112 + }, + { + "epoch": 0.51113, + "grad_norm": 1.248631833560887, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 51113 + }, + { + "epoch": 0.51114, + "grad_norm": 1.0282913570679038, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 51114 + }, + { + "epoch": 0.51115, + "grad_norm": 1.1519764835841655, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 51115 + }, + { + "epoch": 0.51116, + "grad_norm": 1.031853084216603, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 51116 + }, + { + "epoch": 0.51117, + "grad_norm": 1.2610598194696, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 51117 + }, + { + "epoch": 0.51118, + "grad_norm": 0.939210322918759, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 51118 + }, + { + "epoch": 0.51119, + "grad_norm": 1.0546895857834508, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 51119 + }, + { + "epoch": 0.5112, + "grad_norm": 1.0474463144776325, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 51120 + }, + { + "epoch": 0.51121, + "grad_norm": 1.0879741779119758, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 51121 + }, + { + "epoch": 0.51122, + "grad_norm": 1.0615793104715825, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 51122 + }, + { + "epoch": 0.51123, + "grad_norm": 1.2487601752929367, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 51123 + }, + { + "epoch": 0.51124, + "grad_norm": 1.0625276137240092, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 51124 + }, + { + "epoch": 0.51125, + "grad_norm": 0.9519202412870161, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 51125 + }, + { + "epoch": 0.51126, + "grad_norm": 0.7516302529462112, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 51126 + }, + { + "epoch": 0.51127, + "grad_norm": 0.7741473263480605, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 51127 + }, + { + "epoch": 0.51128, + "grad_norm": 0.9181367570976842, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 51128 + }, + { + "epoch": 0.51129, + "grad_norm": 1.2252198151509932, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 51129 + }, + { + "epoch": 0.5113, + "grad_norm": 0.975200284882854, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 51130 + }, + { + "epoch": 0.51131, + "grad_norm": 1.0985421296791522, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 51131 + }, + { + "epoch": 0.51132, + "grad_norm": 1.0201811020921236, + "learning_rate": 0.003, + "loss": 4.057, + "step": 51132 + }, + { + "epoch": 0.51133, + "grad_norm": 1.0014138235357746, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 51133 + }, + { + "epoch": 0.51134, + "grad_norm": 1.1796484430296161, + "learning_rate": 0.003, + "loss": 4.054, + "step": 51134 + }, + { + "epoch": 0.51135, + "grad_norm": 0.967234435922614, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 51135 + }, + { + "epoch": 0.51136, + "grad_norm": 1.0596384718795362, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 51136 + }, + { + "epoch": 0.51137, + "grad_norm": 1.0280013404722455, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 51137 + }, + { + "epoch": 0.51138, + "grad_norm": 1.3997369941346007, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 51138 + }, + { + "epoch": 0.51139, + "grad_norm": 1.0102860279164898, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 51139 + }, + { + "epoch": 0.5114, + "grad_norm": 1.3153892269135568, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 51140 + }, + { + "epoch": 0.51141, + "grad_norm": 1.0394337451745848, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 51141 + }, + { + "epoch": 0.51142, + "grad_norm": 1.2560178102479016, + "learning_rate": 0.003, + "loss": 4.039, + "step": 51142 + }, + { + "epoch": 0.51143, + "grad_norm": 0.9915970380287967, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 51143 + }, + { + "epoch": 0.51144, + "grad_norm": 1.2419591539283368, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 51144 + }, + { + "epoch": 0.51145, + "grad_norm": 0.8845089032737249, + "learning_rate": 0.003, + "loss": 4.024, + "step": 51145 + }, + { + "epoch": 0.51146, + "grad_norm": 1.1215821104189876, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 51146 + }, + { + "epoch": 0.51147, + "grad_norm": 1.1143232538208387, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 51147 + }, + { + "epoch": 0.51148, + "grad_norm": 0.8776039889294931, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 51148 + }, + { + "epoch": 0.51149, + "grad_norm": 1.0541749920031431, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 51149 + }, + { + "epoch": 0.5115, + "grad_norm": 1.291055015359996, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 51150 + }, + { + "epoch": 0.51151, + "grad_norm": 0.8153833626629493, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 51151 + }, + { + "epoch": 0.51152, + "grad_norm": 0.7732179425685396, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 51152 + }, + { + "epoch": 0.51153, + "grad_norm": 0.8290526261206187, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 51153 + }, + { + "epoch": 0.51154, + "grad_norm": 0.9059445968060754, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 51154 + }, + { + "epoch": 0.51155, + "grad_norm": 1.1014537961787298, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 51155 + }, + { + "epoch": 0.51156, + "grad_norm": 1.0574475198032092, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 51156 + }, + { + "epoch": 0.51157, + "grad_norm": 1.1962318503837344, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 51157 + }, + { + "epoch": 0.51158, + "grad_norm": 1.2468171541202862, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 51158 + }, + { + "epoch": 0.51159, + "grad_norm": 0.8827154088143657, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 51159 + }, + { + "epoch": 0.5116, + "grad_norm": 0.9977779357479114, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 51160 + }, + { + "epoch": 0.51161, + "grad_norm": 1.474917805256161, + "learning_rate": 0.003, + "loss": 4.051, + "step": 51161 + }, + { + "epoch": 0.51162, + "grad_norm": 0.8529151488278298, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 51162 + }, + { + "epoch": 0.51163, + "grad_norm": 1.0367870610282912, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 51163 + }, + { + "epoch": 0.51164, + "grad_norm": 1.0073909004083863, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 51164 + }, + { + "epoch": 0.51165, + "grad_norm": 1.0643383898981449, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 51165 + }, + { + "epoch": 0.51166, + "grad_norm": 1.1739247549700371, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 51166 + }, + { + "epoch": 0.51167, + "grad_norm": 1.0668111994227938, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 51167 + }, + { + "epoch": 0.51168, + "grad_norm": 1.1798422474327752, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 51168 + }, + { + "epoch": 0.51169, + "grad_norm": 1.3056820970476453, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 51169 + }, + { + "epoch": 0.5117, + "grad_norm": 0.7264399796117385, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 51170 + }, + { + "epoch": 0.51171, + "grad_norm": 0.7848758573130457, + "learning_rate": 0.003, + "loss": 4.027, + "step": 51171 + }, + { + "epoch": 0.51172, + "grad_norm": 0.791709280718169, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 51172 + }, + { + "epoch": 0.51173, + "grad_norm": 0.8946492720136471, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 51173 + }, + { + "epoch": 0.51174, + "grad_norm": 1.0002465766183006, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 51174 + }, + { + "epoch": 0.51175, + "grad_norm": 1.373112850096026, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 51175 + }, + { + "epoch": 0.51176, + "grad_norm": 0.900682715172846, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 51176 + }, + { + "epoch": 0.51177, + "grad_norm": 0.8937486519119029, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 51177 + }, + { + "epoch": 0.51178, + "grad_norm": 1.0900403035151691, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 51178 + }, + { + "epoch": 0.51179, + "grad_norm": 1.4576621999497819, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 51179 + }, + { + "epoch": 0.5118, + "grad_norm": 0.9355766873755534, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 51180 + }, + { + "epoch": 0.51181, + "grad_norm": 0.7389419494065396, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 51181 + }, + { + "epoch": 0.51182, + "grad_norm": 0.7749590973618546, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 51182 + }, + { + "epoch": 0.51183, + "grad_norm": 0.7622237112891876, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 51183 + }, + { + "epoch": 0.51184, + "grad_norm": 0.7899617349580036, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 51184 + }, + { + "epoch": 0.51185, + "grad_norm": 0.9327128019774624, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 51185 + }, + { + "epoch": 0.51186, + "grad_norm": 1.244743460352384, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 51186 + }, + { + "epoch": 0.51187, + "grad_norm": 1.0313597495397278, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 51187 + }, + { + "epoch": 0.51188, + "grad_norm": 1.2978122185997722, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 51188 + }, + { + "epoch": 0.51189, + "grad_norm": 1.001209389429322, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 51189 + }, + { + "epoch": 0.5119, + "grad_norm": 1.1697741742306065, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 51190 + }, + { + "epoch": 0.51191, + "grad_norm": 0.9619804335746419, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 51191 + }, + { + "epoch": 0.51192, + "grad_norm": 1.0971866232648622, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 51192 + }, + { + "epoch": 0.51193, + "grad_norm": 1.142300534477441, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 51193 + }, + { + "epoch": 0.51194, + "grad_norm": 0.8870332788872657, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 51194 + }, + { + "epoch": 0.51195, + "grad_norm": 0.9650537133087047, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 51195 + }, + { + "epoch": 0.51196, + "grad_norm": 1.3940254177498912, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 51196 + }, + { + "epoch": 0.51197, + "grad_norm": 0.8856723745867212, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 51197 + }, + { + "epoch": 0.51198, + "grad_norm": 0.9723532549108284, + "learning_rate": 0.003, + "loss": 4.056, + "step": 51198 + }, + { + "epoch": 0.51199, + "grad_norm": 1.1672924801407545, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 51199 + }, + { + "epoch": 0.512, + "grad_norm": 1.0912278460384341, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 51200 + }, + { + "epoch": 0.51201, + "grad_norm": 1.3982822444247731, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 51201 + }, + { + "epoch": 0.51202, + "grad_norm": 0.8134071037628531, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 51202 + }, + { + "epoch": 0.51203, + "grad_norm": 0.8705513016239304, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 51203 + }, + { + "epoch": 0.51204, + "grad_norm": 1.009348936621592, + "learning_rate": 0.003, + "loss": 4.031, + "step": 51204 + }, + { + "epoch": 0.51205, + "grad_norm": 1.075613233252376, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 51205 + }, + { + "epoch": 0.51206, + "grad_norm": 1.02583941206166, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 51206 + }, + { + "epoch": 0.51207, + "grad_norm": 1.309792434020884, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 51207 + }, + { + "epoch": 0.51208, + "grad_norm": 1.020673294310824, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 51208 + }, + { + "epoch": 0.51209, + "grad_norm": 0.9867080174736577, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 51209 + }, + { + "epoch": 0.5121, + "grad_norm": 1.1013140264943189, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 51210 + }, + { + "epoch": 0.51211, + "grad_norm": 0.9226930828164911, + "learning_rate": 0.003, + "loss": 4.034, + "step": 51211 + }, + { + "epoch": 0.51212, + "grad_norm": 0.8187441615123877, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 51212 + }, + { + "epoch": 0.51213, + "grad_norm": 0.8985434817934791, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 51213 + }, + { + "epoch": 0.51214, + "grad_norm": 1.0736515900690322, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 51214 + }, + { + "epoch": 0.51215, + "grad_norm": 1.1165150743208412, + "learning_rate": 0.003, + "loss": 4.004, + "step": 51215 + }, + { + "epoch": 0.51216, + "grad_norm": 0.9340923160256088, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 51216 + }, + { + "epoch": 0.51217, + "grad_norm": 1.1580113006113204, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 51217 + }, + { + "epoch": 0.51218, + "grad_norm": 0.879671575951385, + "learning_rate": 0.003, + "loss": 4.056, + "step": 51218 + }, + { + "epoch": 0.51219, + "grad_norm": 0.8922597608380936, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 51219 + }, + { + "epoch": 0.5122, + "grad_norm": 0.9063657007711231, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 51220 + }, + { + "epoch": 0.51221, + "grad_norm": 1.0890557039269397, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 51221 + }, + { + "epoch": 0.51222, + "grad_norm": 0.9898558008092111, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 51222 + }, + { + "epoch": 0.51223, + "grad_norm": 1.142347987594563, + "learning_rate": 0.003, + "loss": 4.038, + "step": 51223 + }, + { + "epoch": 0.51224, + "grad_norm": 1.1164400914916381, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 51224 + }, + { + "epoch": 0.51225, + "grad_norm": 1.1072082632399463, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 51225 + }, + { + "epoch": 0.51226, + "grad_norm": 1.0468062303829428, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 51226 + }, + { + "epoch": 0.51227, + "grad_norm": 1.2105159611713072, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 51227 + }, + { + "epoch": 0.51228, + "grad_norm": 1.0299778304639764, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 51228 + }, + { + "epoch": 0.51229, + "grad_norm": 1.375185699957797, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 51229 + }, + { + "epoch": 0.5123, + "grad_norm": 1.0118089095734901, + "learning_rate": 0.003, + "loss": 4.032, + "step": 51230 + }, + { + "epoch": 0.51231, + "grad_norm": 1.3175287109410263, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 51231 + }, + { + "epoch": 0.51232, + "grad_norm": 0.8988263559194923, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 51232 + }, + { + "epoch": 0.51233, + "grad_norm": 0.9467363035149058, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 51233 + }, + { + "epoch": 0.51234, + "grad_norm": 1.3750489488511073, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 51234 + }, + { + "epoch": 0.51235, + "grad_norm": 0.9515009262971785, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 51235 + }, + { + "epoch": 0.51236, + "grad_norm": 0.8287607764342684, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 51236 + }, + { + "epoch": 0.51237, + "grad_norm": 0.8177391043506279, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 51237 + }, + { + "epoch": 0.51238, + "grad_norm": 0.9454174727062187, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 51238 + }, + { + "epoch": 0.51239, + "grad_norm": 1.0902053502395528, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 51239 + }, + { + "epoch": 0.5124, + "grad_norm": 0.8961800222656263, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 51240 + }, + { + "epoch": 0.51241, + "grad_norm": 0.9077556136910776, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 51241 + }, + { + "epoch": 0.51242, + "grad_norm": 1.0203913717035078, + "learning_rate": 0.003, + "loss": 4.03, + "step": 51242 + }, + { + "epoch": 0.51243, + "grad_norm": 1.3622976449937279, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 51243 + }, + { + "epoch": 0.51244, + "grad_norm": 0.868457357509739, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 51244 + }, + { + "epoch": 0.51245, + "grad_norm": 0.9040875296040656, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 51245 + }, + { + "epoch": 0.51246, + "grad_norm": 1.017392217163094, + "learning_rate": 0.003, + "loss": 4.059, + "step": 51246 + }, + { + "epoch": 0.51247, + "grad_norm": 1.1852752955923735, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 51247 + }, + { + "epoch": 0.51248, + "grad_norm": 1.0232295726262985, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 51248 + }, + { + "epoch": 0.51249, + "grad_norm": 1.1626374272250324, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 51249 + }, + { + "epoch": 0.5125, + "grad_norm": 1.1822339173820544, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 51250 + }, + { + "epoch": 0.51251, + "grad_norm": 1.0957756948569919, + "learning_rate": 0.003, + "loss": 4.029, + "step": 51251 + }, + { + "epoch": 0.51252, + "grad_norm": 1.0553757690201708, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 51252 + }, + { + "epoch": 0.51253, + "grad_norm": 0.9365620451890788, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 51253 + }, + { + "epoch": 0.51254, + "grad_norm": 0.8958040009281827, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 51254 + }, + { + "epoch": 0.51255, + "grad_norm": 0.9551377289311495, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 51255 + }, + { + "epoch": 0.51256, + "grad_norm": 1.2604440583914225, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 51256 + }, + { + "epoch": 0.51257, + "grad_norm": 0.8639576205413453, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 51257 + }, + { + "epoch": 0.51258, + "grad_norm": 0.974242479059517, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 51258 + }, + { + "epoch": 0.51259, + "grad_norm": 1.0701527626184681, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 51259 + }, + { + "epoch": 0.5126, + "grad_norm": 1.171320225336169, + "learning_rate": 0.003, + "loss": 4.038, + "step": 51260 + }, + { + "epoch": 0.51261, + "grad_norm": 1.1161455493024486, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 51261 + }, + { + "epoch": 0.51262, + "grad_norm": 1.1602920155549663, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 51262 + }, + { + "epoch": 0.51263, + "grad_norm": 0.9144037451258439, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 51263 + }, + { + "epoch": 0.51264, + "grad_norm": 1.040920462067782, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 51264 + }, + { + "epoch": 0.51265, + "grad_norm": 0.9109264224455762, + "learning_rate": 0.003, + "loss": 4.016, + "step": 51265 + }, + { + "epoch": 0.51266, + "grad_norm": 0.8318435618350495, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 51266 + }, + { + "epoch": 0.51267, + "grad_norm": 0.991222280623672, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 51267 + }, + { + "epoch": 0.51268, + "grad_norm": 1.2694261084877914, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 51268 + }, + { + "epoch": 0.51269, + "grad_norm": 1.086243236918449, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 51269 + }, + { + "epoch": 0.5127, + "grad_norm": 1.1798823554442315, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 51270 + }, + { + "epoch": 0.51271, + "grad_norm": 1.3813519159016694, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 51271 + }, + { + "epoch": 0.51272, + "grad_norm": 0.9995806191893103, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 51272 + }, + { + "epoch": 0.51273, + "grad_norm": 0.9547838615417499, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 51273 + }, + { + "epoch": 0.51274, + "grad_norm": 1.0685152433015919, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 51274 + }, + { + "epoch": 0.51275, + "grad_norm": 1.258571153431766, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 51275 + }, + { + "epoch": 0.51276, + "grad_norm": 1.2186720536648392, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 51276 + }, + { + "epoch": 0.51277, + "grad_norm": 0.9490463215354153, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 51277 + }, + { + "epoch": 0.51278, + "grad_norm": 0.9677058022454775, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 51278 + }, + { + "epoch": 0.51279, + "grad_norm": 1.1470651668863048, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 51279 + }, + { + "epoch": 0.5128, + "grad_norm": 1.1522561926924375, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 51280 + }, + { + "epoch": 0.51281, + "grad_norm": 0.8105939676440749, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 51281 + }, + { + "epoch": 0.51282, + "grad_norm": 0.7805706781647708, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 51282 + }, + { + "epoch": 0.51283, + "grad_norm": 0.9576183930981526, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 51283 + }, + { + "epoch": 0.51284, + "grad_norm": 1.4133156033323588, + "learning_rate": 0.003, + "loss": 4.02, + "step": 51284 + }, + { + "epoch": 0.51285, + "grad_norm": 0.7817187376490233, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 51285 + }, + { + "epoch": 0.51286, + "grad_norm": 0.8292760551051139, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 51286 + }, + { + "epoch": 0.51287, + "grad_norm": 0.9283314039118196, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 51287 + }, + { + "epoch": 0.51288, + "grad_norm": 1.0400703962749844, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 51288 + }, + { + "epoch": 0.51289, + "grad_norm": 1.2027936251616869, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 51289 + }, + { + "epoch": 0.5129, + "grad_norm": 0.9475260688199129, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 51290 + }, + { + "epoch": 0.51291, + "grad_norm": 1.2885637574956905, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 51291 + }, + { + "epoch": 0.51292, + "grad_norm": 1.189103281411861, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 51292 + }, + { + "epoch": 0.51293, + "grad_norm": 1.272662355809198, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 51293 + }, + { + "epoch": 0.51294, + "grad_norm": 0.7446004252676462, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 51294 + }, + { + "epoch": 0.51295, + "grad_norm": 0.6190145705385977, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 51295 + }, + { + "epoch": 0.51296, + "grad_norm": 0.6610174587888917, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 51296 + }, + { + "epoch": 0.51297, + "grad_norm": 0.7496981466405394, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 51297 + }, + { + "epoch": 0.51298, + "grad_norm": 0.9058919453342096, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 51298 + }, + { + "epoch": 0.51299, + "grad_norm": 1.085764524637936, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 51299 + }, + { + "epoch": 0.513, + "grad_norm": 1.1856351853176568, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 51300 + }, + { + "epoch": 0.51301, + "grad_norm": 1.0441812802612604, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 51301 + }, + { + "epoch": 0.51302, + "grad_norm": 1.1001685442002123, + "learning_rate": 0.003, + "loss": 3.999, + "step": 51302 + }, + { + "epoch": 0.51303, + "grad_norm": 1.081876192018517, + "learning_rate": 0.003, + "loss": 4.028, + "step": 51303 + }, + { + "epoch": 0.51304, + "grad_norm": 1.0310843465088912, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 51304 + }, + { + "epoch": 0.51305, + "grad_norm": 1.2335888738326746, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 51305 + }, + { + "epoch": 0.51306, + "grad_norm": 1.0365003703021896, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 51306 + }, + { + "epoch": 0.51307, + "grad_norm": 1.1955345960845192, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 51307 + }, + { + "epoch": 0.51308, + "grad_norm": 1.0898839727831295, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 51308 + }, + { + "epoch": 0.51309, + "grad_norm": 0.9782278489686548, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 51309 + }, + { + "epoch": 0.5131, + "grad_norm": 1.0815178499551144, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 51310 + }, + { + "epoch": 0.51311, + "grad_norm": 1.2067660958803446, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 51311 + }, + { + "epoch": 0.51312, + "grad_norm": 0.9426385449635607, + "learning_rate": 0.003, + "loss": 4.013, + "step": 51312 + }, + { + "epoch": 0.51313, + "grad_norm": 0.9874147824638538, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 51313 + }, + { + "epoch": 0.51314, + "grad_norm": 1.0569888331654507, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 51314 + }, + { + "epoch": 0.51315, + "grad_norm": 1.2006100781605764, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 51315 + }, + { + "epoch": 0.51316, + "grad_norm": 1.1305297280101634, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 51316 + }, + { + "epoch": 0.51317, + "grad_norm": 0.9342631525104512, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 51317 + }, + { + "epoch": 0.51318, + "grad_norm": 1.1279405448630937, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 51318 + }, + { + "epoch": 0.51319, + "grad_norm": 1.096078955373031, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 51319 + }, + { + "epoch": 0.5132, + "grad_norm": 1.04618146860444, + "learning_rate": 0.003, + "loss": 4.007, + "step": 51320 + }, + { + "epoch": 0.51321, + "grad_norm": 1.1410659895255988, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 51321 + }, + { + "epoch": 0.51322, + "grad_norm": 0.9437738425329555, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 51322 + }, + { + "epoch": 0.51323, + "grad_norm": 0.9254157456894728, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 51323 + }, + { + "epoch": 0.51324, + "grad_norm": 0.8494735795556224, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 51324 + }, + { + "epoch": 0.51325, + "grad_norm": 0.8709643759738445, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 51325 + }, + { + "epoch": 0.51326, + "grad_norm": 1.0287766575134745, + "learning_rate": 0.003, + "loss": 4.061, + "step": 51326 + }, + { + "epoch": 0.51327, + "grad_norm": 1.2298298920537831, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 51327 + }, + { + "epoch": 0.51328, + "grad_norm": 1.0517316683796252, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 51328 + }, + { + "epoch": 0.51329, + "grad_norm": 1.117149804142827, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 51329 + }, + { + "epoch": 0.5133, + "grad_norm": 1.0438936728439132, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 51330 + }, + { + "epoch": 0.51331, + "grad_norm": 0.9308797353394964, + "learning_rate": 0.003, + "loss": 4.04, + "step": 51331 + }, + { + "epoch": 0.51332, + "grad_norm": 0.9825362075944698, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 51332 + }, + { + "epoch": 0.51333, + "grad_norm": 1.0024379152255183, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 51333 + }, + { + "epoch": 0.51334, + "grad_norm": 1.1028544660096278, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 51334 + }, + { + "epoch": 0.51335, + "grad_norm": 0.965634919345543, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 51335 + }, + { + "epoch": 0.51336, + "grad_norm": 1.156362846846959, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 51336 + }, + { + "epoch": 0.51337, + "grad_norm": 1.148179474307434, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 51337 + }, + { + "epoch": 0.51338, + "grad_norm": 1.4493155292776176, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 51338 + }, + { + "epoch": 0.51339, + "grad_norm": 0.8209069408147409, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 51339 + }, + { + "epoch": 0.5134, + "grad_norm": 0.803500374040574, + "learning_rate": 0.003, + "loss": 4.016, + "step": 51340 + }, + { + "epoch": 0.51341, + "grad_norm": 1.0593805289030458, + "learning_rate": 0.003, + "loss": 4.035, + "step": 51341 + }, + { + "epoch": 0.51342, + "grad_norm": 1.2422446745870517, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 51342 + }, + { + "epoch": 0.51343, + "grad_norm": 1.057497733596926, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 51343 + }, + { + "epoch": 0.51344, + "grad_norm": 1.2872381129362767, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 51344 + }, + { + "epoch": 0.51345, + "grad_norm": 0.936953277638291, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 51345 + }, + { + "epoch": 0.51346, + "grad_norm": 1.108743796138456, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 51346 + }, + { + "epoch": 0.51347, + "grad_norm": 1.2996790150894237, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 51347 + }, + { + "epoch": 0.51348, + "grad_norm": 0.7723971558894652, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 51348 + }, + { + "epoch": 0.51349, + "grad_norm": 0.8099383536277929, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 51349 + }, + { + "epoch": 0.5135, + "grad_norm": 0.8420762328471824, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 51350 + }, + { + "epoch": 0.51351, + "grad_norm": 1.0199464063403656, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 51351 + }, + { + "epoch": 0.51352, + "grad_norm": 1.0331486257251339, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 51352 + }, + { + "epoch": 0.51353, + "grad_norm": 1.1914923533170116, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 51353 + }, + { + "epoch": 0.51354, + "grad_norm": 1.0887905625154795, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 51354 + }, + { + "epoch": 0.51355, + "grad_norm": 1.1585378602223337, + "learning_rate": 0.003, + "loss": 4.036, + "step": 51355 + }, + { + "epoch": 0.51356, + "grad_norm": 1.012688622054203, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 51356 + }, + { + "epoch": 0.51357, + "grad_norm": 1.0722183973664305, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 51357 + }, + { + "epoch": 0.51358, + "grad_norm": 0.9352538816917569, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 51358 + }, + { + "epoch": 0.51359, + "grad_norm": 0.9061296163311089, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 51359 + }, + { + "epoch": 0.5136, + "grad_norm": 1.1829064959976718, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 51360 + }, + { + "epoch": 0.51361, + "grad_norm": 1.4318018782509565, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 51361 + }, + { + "epoch": 0.51362, + "grad_norm": 0.7263303423138374, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 51362 + }, + { + "epoch": 0.51363, + "grad_norm": 0.8043589652502995, + "learning_rate": 0.003, + "loss": 4.055, + "step": 51363 + }, + { + "epoch": 0.51364, + "grad_norm": 0.8510677035901394, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 51364 + }, + { + "epoch": 0.51365, + "grad_norm": 1.1387041413939438, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 51365 + }, + { + "epoch": 0.51366, + "grad_norm": 1.113302872830735, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 51366 + }, + { + "epoch": 0.51367, + "grad_norm": 1.1744303171365582, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 51367 + }, + { + "epoch": 0.51368, + "grad_norm": 0.9738759834335075, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 51368 + }, + { + "epoch": 0.51369, + "grad_norm": 0.9624304707481578, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 51369 + }, + { + "epoch": 0.5137, + "grad_norm": 0.9221782259626864, + "learning_rate": 0.003, + "loss": 4.027, + "step": 51370 + }, + { + "epoch": 0.51371, + "grad_norm": 1.0531384025025978, + "learning_rate": 0.003, + "loss": 4.023, + "step": 51371 + }, + { + "epoch": 0.51372, + "grad_norm": 1.0670126755866354, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 51372 + }, + { + "epoch": 0.51373, + "grad_norm": 1.1721209550642133, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 51373 + }, + { + "epoch": 0.51374, + "grad_norm": 1.0865649337517826, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 51374 + }, + { + "epoch": 0.51375, + "grad_norm": 1.284096805311452, + "learning_rate": 0.003, + "loss": 4.066, + "step": 51375 + }, + { + "epoch": 0.51376, + "grad_norm": 1.237550711926837, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 51376 + }, + { + "epoch": 0.51377, + "grad_norm": 0.9447865637619789, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 51377 + }, + { + "epoch": 0.51378, + "grad_norm": 0.9242901619651769, + "learning_rate": 0.003, + "loss": 4.004, + "step": 51378 + }, + { + "epoch": 0.51379, + "grad_norm": 1.2031418010642634, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 51379 + }, + { + "epoch": 0.5138, + "grad_norm": 0.966136423626314, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 51380 + }, + { + "epoch": 0.51381, + "grad_norm": 1.134070299465344, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 51381 + }, + { + "epoch": 0.51382, + "grad_norm": 1.0494441796576461, + "learning_rate": 0.003, + "loss": 4.045, + "step": 51382 + }, + { + "epoch": 0.51383, + "grad_norm": 1.1811200600838971, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 51383 + }, + { + "epoch": 0.51384, + "grad_norm": 1.0443553077470777, + "learning_rate": 0.003, + "loss": 4.021, + "step": 51384 + }, + { + "epoch": 0.51385, + "grad_norm": 0.9840892057620657, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 51385 + }, + { + "epoch": 0.51386, + "grad_norm": 1.0760277106920522, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 51386 + }, + { + "epoch": 0.51387, + "grad_norm": 1.282310680392927, + "learning_rate": 0.003, + "loss": 4.017, + "step": 51387 + }, + { + "epoch": 0.51388, + "grad_norm": 0.8539885597107109, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 51388 + }, + { + "epoch": 0.51389, + "grad_norm": 0.8370814478732407, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 51389 + }, + { + "epoch": 0.5139, + "grad_norm": 0.8430178041704247, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 51390 + }, + { + "epoch": 0.51391, + "grad_norm": 1.0747229556386297, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 51391 + }, + { + "epoch": 0.51392, + "grad_norm": 1.2562080108739557, + "learning_rate": 0.003, + "loss": 4.032, + "step": 51392 + }, + { + "epoch": 0.51393, + "grad_norm": 0.8627984801625412, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 51393 + }, + { + "epoch": 0.51394, + "grad_norm": 1.1310937164828432, + "learning_rate": 0.003, + "loss": 4.045, + "step": 51394 + }, + { + "epoch": 0.51395, + "grad_norm": 1.3181857976311937, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 51395 + }, + { + "epoch": 0.51396, + "grad_norm": 1.020699130740774, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 51396 + }, + { + "epoch": 0.51397, + "grad_norm": 1.185950139159914, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 51397 + }, + { + "epoch": 0.51398, + "grad_norm": 0.8892142222114839, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 51398 + }, + { + "epoch": 0.51399, + "grad_norm": 0.9616691941537333, + "learning_rate": 0.003, + "loss": 4.012, + "step": 51399 + }, + { + "epoch": 0.514, + "grad_norm": 0.9631850503510115, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 51400 + }, + { + "epoch": 0.51401, + "grad_norm": 1.031663313764615, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 51401 + }, + { + "epoch": 0.51402, + "grad_norm": 1.1531370032077366, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 51402 + }, + { + "epoch": 0.51403, + "grad_norm": 1.121252914585084, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 51403 + }, + { + "epoch": 0.51404, + "grad_norm": 0.8702812436357571, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 51404 + }, + { + "epoch": 0.51405, + "grad_norm": 1.1105965044599027, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 51405 + }, + { + "epoch": 0.51406, + "grad_norm": 1.4170255729513959, + "learning_rate": 0.003, + "loss": 4.057, + "step": 51406 + }, + { + "epoch": 0.51407, + "grad_norm": 0.9185405399137603, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 51407 + }, + { + "epoch": 0.51408, + "grad_norm": 0.9704357719092551, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 51408 + }, + { + "epoch": 0.51409, + "grad_norm": 1.0584440024313353, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 51409 + }, + { + "epoch": 0.5141, + "grad_norm": 1.2031186997360872, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 51410 + }, + { + "epoch": 0.51411, + "grad_norm": 1.4159919614336833, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 51411 + }, + { + "epoch": 0.51412, + "grad_norm": 0.9054190105701402, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 51412 + }, + { + "epoch": 0.51413, + "grad_norm": 0.7958967585265715, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 51413 + }, + { + "epoch": 0.51414, + "grad_norm": 1.0231086866335057, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 51414 + }, + { + "epoch": 0.51415, + "grad_norm": 1.0541767728455933, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 51415 + }, + { + "epoch": 0.51416, + "grad_norm": 0.9584170897860338, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 51416 + }, + { + "epoch": 0.51417, + "grad_norm": 1.106670550839486, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 51417 + }, + { + "epoch": 0.51418, + "grad_norm": 1.1858625888028997, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 51418 + }, + { + "epoch": 0.51419, + "grad_norm": 1.0245055048319927, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 51419 + }, + { + "epoch": 0.5142, + "grad_norm": 1.1539677575229355, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 51420 + }, + { + "epoch": 0.51421, + "grad_norm": 1.1023527128717387, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 51421 + }, + { + "epoch": 0.51422, + "grad_norm": 0.991516904860207, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 51422 + }, + { + "epoch": 0.51423, + "grad_norm": 1.008466232558089, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 51423 + }, + { + "epoch": 0.51424, + "grad_norm": 1.3708446999837878, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 51424 + }, + { + "epoch": 0.51425, + "grad_norm": 1.0631787619147295, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 51425 + }, + { + "epoch": 0.51426, + "grad_norm": 0.9154781036206405, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 51426 + }, + { + "epoch": 0.51427, + "grad_norm": 0.9124748538202052, + "learning_rate": 0.003, + "loss": 4.031, + "step": 51427 + }, + { + "epoch": 0.51428, + "grad_norm": 1.035614064338118, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 51428 + }, + { + "epoch": 0.51429, + "grad_norm": 1.4695364121028565, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 51429 + }, + { + "epoch": 0.5143, + "grad_norm": 0.8162754112281586, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 51430 + }, + { + "epoch": 0.51431, + "grad_norm": 0.8974267201531105, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 51431 + }, + { + "epoch": 0.51432, + "grad_norm": 0.9847397075890737, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 51432 + }, + { + "epoch": 0.51433, + "grad_norm": 1.328749387727288, + "learning_rate": 0.003, + "loss": 4.029, + "step": 51433 + }, + { + "epoch": 0.51434, + "grad_norm": 0.9405087168871755, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 51434 + }, + { + "epoch": 0.51435, + "grad_norm": 1.0124005698145193, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 51435 + }, + { + "epoch": 0.51436, + "grad_norm": 1.192040381614661, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 51436 + }, + { + "epoch": 0.51437, + "grad_norm": 0.8214514617110878, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 51437 + }, + { + "epoch": 0.51438, + "grad_norm": 0.9259579296899094, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 51438 + }, + { + "epoch": 0.51439, + "grad_norm": 1.231729066098146, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 51439 + }, + { + "epoch": 0.5144, + "grad_norm": 0.8652048354180008, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 51440 + }, + { + "epoch": 0.51441, + "grad_norm": 0.9892365289461552, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 51441 + }, + { + "epoch": 0.51442, + "grad_norm": 1.0550234476264333, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 51442 + }, + { + "epoch": 0.51443, + "grad_norm": 1.4539381879085889, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 51443 + }, + { + "epoch": 0.51444, + "grad_norm": 0.8689168301654819, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 51444 + }, + { + "epoch": 0.51445, + "grad_norm": 0.8393668137072144, + "learning_rate": 0.003, + "loss": 4.001, + "step": 51445 + }, + { + "epoch": 0.51446, + "grad_norm": 0.8770228401639695, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 51446 + }, + { + "epoch": 0.51447, + "grad_norm": 0.9963917930109917, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 51447 + }, + { + "epoch": 0.51448, + "grad_norm": 1.1670054266674947, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 51448 + }, + { + "epoch": 0.51449, + "grad_norm": 0.9771770519673177, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 51449 + }, + { + "epoch": 0.5145, + "grad_norm": 1.1974368796916108, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 51450 + }, + { + "epoch": 0.51451, + "grad_norm": 0.9557508692653208, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 51451 + }, + { + "epoch": 0.51452, + "grad_norm": 1.2786904562579022, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 51452 + }, + { + "epoch": 0.51453, + "grad_norm": 1.2107496405650504, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 51453 + }, + { + "epoch": 0.51454, + "grad_norm": 1.2226813317724878, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 51454 + }, + { + "epoch": 0.51455, + "grad_norm": 0.9270157393759206, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 51455 + }, + { + "epoch": 0.51456, + "grad_norm": 1.0576574202445779, + "learning_rate": 0.003, + "loss": 4.028, + "step": 51456 + }, + { + "epoch": 0.51457, + "grad_norm": 1.2440855850363834, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 51457 + }, + { + "epoch": 0.51458, + "grad_norm": 0.8927901224519714, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 51458 + }, + { + "epoch": 0.51459, + "grad_norm": 0.8698365917042554, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 51459 + }, + { + "epoch": 0.5146, + "grad_norm": 1.0933802469701013, + "learning_rate": 0.003, + "loss": 4.047, + "step": 51460 + }, + { + "epoch": 0.51461, + "grad_norm": 1.3113830936866395, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 51461 + }, + { + "epoch": 0.51462, + "grad_norm": 0.925993427320199, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 51462 + }, + { + "epoch": 0.51463, + "grad_norm": 1.0275470599767242, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 51463 + }, + { + "epoch": 0.51464, + "grad_norm": 1.0172291718333737, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 51464 + }, + { + "epoch": 0.51465, + "grad_norm": 1.0510361437648454, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 51465 + }, + { + "epoch": 0.51466, + "grad_norm": 1.1102130191564579, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 51466 + }, + { + "epoch": 0.51467, + "grad_norm": 1.0119575836366215, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 51467 + }, + { + "epoch": 0.51468, + "grad_norm": 1.1109591257703386, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 51468 + }, + { + "epoch": 0.51469, + "grad_norm": 1.0429238669650696, + "learning_rate": 0.003, + "loss": 4.04, + "step": 51469 + }, + { + "epoch": 0.5147, + "grad_norm": 1.1073816761093724, + "learning_rate": 0.003, + "loss": 4.005, + "step": 51470 + }, + { + "epoch": 0.51471, + "grad_norm": 1.0792742820142613, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 51471 + }, + { + "epoch": 0.51472, + "grad_norm": 1.137305334692618, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 51472 + }, + { + "epoch": 0.51473, + "grad_norm": 1.005061722393366, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 51473 + }, + { + "epoch": 0.51474, + "grad_norm": 1.0938172745418107, + "learning_rate": 0.003, + "loss": 4.051, + "step": 51474 + }, + { + "epoch": 0.51475, + "grad_norm": 1.0780198945886923, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 51475 + }, + { + "epoch": 0.51476, + "grad_norm": 1.1204442459059114, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 51476 + }, + { + "epoch": 0.51477, + "grad_norm": 0.9895902490016271, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 51477 + }, + { + "epoch": 0.51478, + "grad_norm": 1.0834133361125868, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 51478 + }, + { + "epoch": 0.51479, + "grad_norm": 0.9452723833523808, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 51479 + }, + { + "epoch": 0.5148, + "grad_norm": 1.0892827246465806, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 51480 + }, + { + "epoch": 0.51481, + "grad_norm": 1.1368773220874109, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 51481 + }, + { + "epoch": 0.51482, + "grad_norm": 1.1168932932071178, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 51482 + }, + { + "epoch": 0.51483, + "grad_norm": 0.8849571627305829, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 51483 + }, + { + "epoch": 0.51484, + "grad_norm": 1.017823321354005, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 51484 + }, + { + "epoch": 0.51485, + "grad_norm": 1.5326533939076104, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 51485 + }, + { + "epoch": 0.51486, + "grad_norm": 0.887728160725026, + "learning_rate": 0.003, + "loss": 4.056, + "step": 51486 + }, + { + "epoch": 0.51487, + "grad_norm": 1.0541982891474344, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 51487 + }, + { + "epoch": 0.51488, + "grad_norm": 1.140065777447464, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 51488 + }, + { + "epoch": 0.51489, + "grad_norm": 1.09393584309307, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 51489 + }, + { + "epoch": 0.5149, + "grad_norm": 1.0521634952633723, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 51490 + }, + { + "epoch": 0.51491, + "grad_norm": 0.8066380161400588, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 51491 + }, + { + "epoch": 0.51492, + "grad_norm": 0.9444120207120371, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 51492 + }, + { + "epoch": 0.51493, + "grad_norm": 1.1811903007489837, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 51493 + }, + { + "epoch": 0.51494, + "grad_norm": 0.8318823984003537, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 51494 + }, + { + "epoch": 0.51495, + "grad_norm": 0.904273560917048, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 51495 + }, + { + "epoch": 0.51496, + "grad_norm": 0.8689230401151985, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 51496 + }, + { + "epoch": 0.51497, + "grad_norm": 1.0215888214582574, + "learning_rate": 0.003, + "loss": 4.019, + "step": 51497 + }, + { + "epoch": 0.51498, + "grad_norm": 1.0818350466276494, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 51498 + }, + { + "epoch": 0.51499, + "grad_norm": 1.3417827915698994, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 51499 + }, + { + "epoch": 0.515, + "grad_norm": 1.164838614963451, + "learning_rate": 0.003, + "loss": 4.031, + "step": 51500 + }, + { + "epoch": 0.51501, + "grad_norm": 1.052434252570172, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 51501 + }, + { + "epoch": 0.51502, + "grad_norm": 1.222418747018295, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 51502 + }, + { + "epoch": 0.51503, + "grad_norm": 0.9107401331312918, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 51503 + }, + { + "epoch": 0.51504, + "grad_norm": 0.9728458639537738, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 51504 + }, + { + "epoch": 0.51505, + "grad_norm": 1.276696511117269, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 51505 + }, + { + "epoch": 0.51506, + "grad_norm": 1.0938825030231345, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 51506 + }, + { + "epoch": 0.51507, + "grad_norm": 1.0480296806950664, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 51507 + }, + { + "epoch": 0.51508, + "grad_norm": 1.0686448495296104, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 51508 + }, + { + "epoch": 0.51509, + "grad_norm": 0.9353006818562576, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 51509 + }, + { + "epoch": 0.5151, + "grad_norm": 0.8611257657927736, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 51510 + }, + { + "epoch": 0.51511, + "grad_norm": 0.9572523578490404, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 51511 + }, + { + "epoch": 0.51512, + "grad_norm": 0.9816955790097769, + "learning_rate": 0.003, + "loss": 4.042, + "step": 51512 + }, + { + "epoch": 0.51513, + "grad_norm": 1.1131734637424888, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 51513 + }, + { + "epoch": 0.51514, + "grad_norm": 1.0455500746400208, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 51514 + }, + { + "epoch": 0.51515, + "grad_norm": 0.9520091168950013, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 51515 + }, + { + "epoch": 0.51516, + "grad_norm": 1.091290040872617, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 51516 + }, + { + "epoch": 0.51517, + "grad_norm": 1.3686588838735747, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 51517 + }, + { + "epoch": 0.51518, + "grad_norm": 0.934411465414524, + "learning_rate": 0.003, + "loss": 4.032, + "step": 51518 + }, + { + "epoch": 0.51519, + "grad_norm": 1.1166628812507247, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 51519 + }, + { + "epoch": 0.5152, + "grad_norm": 1.1200519748811884, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 51520 + }, + { + "epoch": 0.51521, + "grad_norm": 1.273142407312981, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 51521 + }, + { + "epoch": 0.51522, + "grad_norm": 0.7911666949835862, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 51522 + }, + { + "epoch": 0.51523, + "grad_norm": 0.7771985706736073, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 51523 + }, + { + "epoch": 0.51524, + "grad_norm": 0.9048356479217339, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 51524 + }, + { + "epoch": 0.51525, + "grad_norm": 1.140461135961179, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 51525 + }, + { + "epoch": 0.51526, + "grad_norm": 1.1397210738693448, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 51526 + }, + { + "epoch": 0.51527, + "grad_norm": 1.0405299322749038, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 51527 + }, + { + "epoch": 0.51528, + "grad_norm": 1.0366021848098657, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 51528 + }, + { + "epoch": 0.51529, + "grad_norm": 0.9868808255023277, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 51529 + }, + { + "epoch": 0.5153, + "grad_norm": 0.9707293859461845, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 51530 + }, + { + "epoch": 0.51531, + "grad_norm": 0.95547769931421, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 51531 + }, + { + "epoch": 0.51532, + "grad_norm": 0.9509652670919662, + "learning_rate": 0.003, + "loss": 4.053, + "step": 51532 + }, + { + "epoch": 0.51533, + "grad_norm": 1.0242569801658559, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 51533 + }, + { + "epoch": 0.51534, + "grad_norm": 1.2069849331760922, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 51534 + }, + { + "epoch": 0.51535, + "grad_norm": 1.0172442825563368, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 51535 + }, + { + "epoch": 0.51536, + "grad_norm": 1.330618395733716, + "learning_rate": 0.003, + "loss": 4.041, + "step": 51536 + }, + { + "epoch": 0.51537, + "grad_norm": 1.3458836417690125, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 51537 + }, + { + "epoch": 0.51538, + "grad_norm": 0.9780725153089403, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 51538 + }, + { + "epoch": 0.51539, + "grad_norm": 1.0095241692678099, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 51539 + }, + { + "epoch": 0.5154, + "grad_norm": 1.193975993683191, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 51540 + }, + { + "epoch": 0.51541, + "grad_norm": 0.9909728475865967, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 51541 + }, + { + "epoch": 0.51542, + "grad_norm": 1.1059162582462194, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 51542 + }, + { + "epoch": 0.51543, + "grad_norm": 1.482054882414508, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 51543 + }, + { + "epoch": 0.51544, + "grad_norm": 0.7626147427258974, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 51544 + }, + { + "epoch": 0.51545, + "grad_norm": 0.748195065234232, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 51545 + }, + { + "epoch": 0.51546, + "grad_norm": 0.8872523010392074, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 51546 + }, + { + "epoch": 0.51547, + "grad_norm": 1.0779181198959171, + "learning_rate": 0.003, + "loss": 4.02, + "step": 51547 + }, + { + "epoch": 0.51548, + "grad_norm": 1.2175391661509156, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 51548 + }, + { + "epoch": 0.51549, + "grad_norm": 1.115551887389898, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 51549 + }, + { + "epoch": 0.5155, + "grad_norm": 1.0807861780931465, + "learning_rate": 0.003, + "loss": 4.035, + "step": 51550 + }, + { + "epoch": 0.51551, + "grad_norm": 1.0237753122840392, + "learning_rate": 0.003, + "loss": 4.053, + "step": 51551 + }, + { + "epoch": 0.51552, + "grad_norm": 0.9779901031584957, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 51552 + }, + { + "epoch": 0.51553, + "grad_norm": 0.955840390930506, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 51553 + }, + { + "epoch": 0.51554, + "grad_norm": 0.9320301946080591, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 51554 + }, + { + "epoch": 0.51555, + "grad_norm": 0.8868637687447719, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 51555 + }, + { + "epoch": 0.51556, + "grad_norm": 0.9613933538830982, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 51556 + }, + { + "epoch": 0.51557, + "grad_norm": 1.0669003901350218, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 51557 + }, + { + "epoch": 0.51558, + "grad_norm": 1.2983881606567387, + "learning_rate": 0.003, + "loss": 4.023, + "step": 51558 + }, + { + "epoch": 0.51559, + "grad_norm": 0.927361278776475, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 51559 + }, + { + "epoch": 0.5156, + "grad_norm": 1.1470192071944763, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 51560 + }, + { + "epoch": 0.51561, + "grad_norm": 1.1678729342650274, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 51561 + }, + { + "epoch": 0.51562, + "grad_norm": 1.0226628578816597, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 51562 + }, + { + "epoch": 0.51563, + "grad_norm": 0.9849095046848171, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 51563 + }, + { + "epoch": 0.51564, + "grad_norm": 1.2110211828766995, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 51564 + }, + { + "epoch": 0.51565, + "grad_norm": 1.002943190980155, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 51565 + }, + { + "epoch": 0.51566, + "grad_norm": 1.0489446652130903, + "learning_rate": 0.003, + "loss": 4.029, + "step": 51566 + }, + { + "epoch": 0.51567, + "grad_norm": 1.0509329531605522, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 51567 + }, + { + "epoch": 0.51568, + "grad_norm": 1.1970616230025646, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 51568 + }, + { + "epoch": 0.51569, + "grad_norm": 0.948245834593323, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 51569 + }, + { + "epoch": 0.5157, + "grad_norm": 1.0056550848921952, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 51570 + }, + { + "epoch": 0.51571, + "grad_norm": 1.1590637163485893, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 51571 + }, + { + "epoch": 0.51572, + "grad_norm": 0.8831618225244259, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 51572 + }, + { + "epoch": 0.51573, + "grad_norm": 0.8952615313616643, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 51573 + }, + { + "epoch": 0.51574, + "grad_norm": 0.9840964816487159, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 51574 + }, + { + "epoch": 0.51575, + "grad_norm": 1.2441438953043356, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 51575 + }, + { + "epoch": 0.51576, + "grad_norm": 0.8249444137493995, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 51576 + }, + { + "epoch": 0.51577, + "grad_norm": 0.8793574550692221, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 51577 + }, + { + "epoch": 0.51578, + "grad_norm": 0.8970801788877097, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 51578 + }, + { + "epoch": 0.51579, + "grad_norm": 1.068507829178549, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 51579 + }, + { + "epoch": 0.5158, + "grad_norm": 1.0293012099494103, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 51580 + }, + { + "epoch": 0.51581, + "grad_norm": 1.0870561408248183, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 51581 + }, + { + "epoch": 0.51582, + "grad_norm": 1.098888593812415, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 51582 + }, + { + "epoch": 0.51583, + "grad_norm": 0.9293933286343344, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 51583 + }, + { + "epoch": 0.51584, + "grad_norm": 0.8780227299660499, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 51584 + }, + { + "epoch": 0.51585, + "grad_norm": 0.877700005875705, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 51585 + }, + { + "epoch": 0.51586, + "grad_norm": 1.1866570966474674, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 51586 + }, + { + "epoch": 0.51587, + "grad_norm": 1.2366279744218767, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 51587 + }, + { + "epoch": 0.51588, + "grad_norm": 0.9702860854310715, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 51588 + }, + { + "epoch": 0.51589, + "grad_norm": 1.0179555305208858, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 51589 + }, + { + "epoch": 0.5159, + "grad_norm": 1.29613699335009, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 51590 + }, + { + "epoch": 0.51591, + "grad_norm": 0.968119099062326, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 51591 + }, + { + "epoch": 0.51592, + "grad_norm": 1.1220789720959057, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 51592 + }, + { + "epoch": 0.51593, + "grad_norm": 1.1326703903950996, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 51593 + }, + { + "epoch": 0.51594, + "grad_norm": 1.2472243783483028, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 51594 + }, + { + "epoch": 0.51595, + "grad_norm": 0.9217050443973756, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 51595 + }, + { + "epoch": 0.51596, + "grad_norm": 0.9507830416947715, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 51596 + }, + { + "epoch": 0.51597, + "grad_norm": 1.2072977343476383, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 51597 + }, + { + "epoch": 0.51598, + "grad_norm": 1.0449975853394398, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 51598 + }, + { + "epoch": 0.51599, + "grad_norm": 0.9383094433473961, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 51599 + }, + { + "epoch": 0.516, + "grad_norm": 0.812535354590491, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 51600 + }, + { + "epoch": 0.51601, + "grad_norm": 0.9382384432203107, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 51601 + }, + { + "epoch": 0.51602, + "grad_norm": 1.0524067908979204, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 51602 + }, + { + "epoch": 0.51603, + "grad_norm": 0.9879777286437557, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 51603 + }, + { + "epoch": 0.51604, + "grad_norm": 0.9561888161870227, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 51604 + }, + { + "epoch": 0.51605, + "grad_norm": 0.9719658488511908, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 51605 + }, + { + "epoch": 0.51606, + "grad_norm": 0.9705877778727898, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 51606 + }, + { + "epoch": 0.51607, + "grad_norm": 0.9834681823072107, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 51607 + }, + { + "epoch": 0.51608, + "grad_norm": 1.3694923422699588, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 51608 + }, + { + "epoch": 0.51609, + "grad_norm": 1.11096499422685, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 51609 + }, + { + "epoch": 0.5161, + "grad_norm": 1.2991617848384494, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 51610 + }, + { + "epoch": 0.51611, + "grad_norm": 1.113154841486853, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 51611 + }, + { + "epoch": 0.51612, + "grad_norm": 0.9980007653423615, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 51612 + }, + { + "epoch": 0.51613, + "grad_norm": 1.0459842137350166, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 51613 + }, + { + "epoch": 0.51614, + "grad_norm": 1.0170502021555328, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 51614 + }, + { + "epoch": 0.51615, + "grad_norm": 1.0734830270023983, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 51615 + }, + { + "epoch": 0.51616, + "grad_norm": 1.1362339213769501, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 51616 + }, + { + "epoch": 0.51617, + "grad_norm": 1.3953793511851462, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 51617 + }, + { + "epoch": 0.51618, + "grad_norm": 0.8205283055722703, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 51618 + }, + { + "epoch": 0.51619, + "grad_norm": 0.8524638702259989, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 51619 + }, + { + "epoch": 0.5162, + "grad_norm": 1.0289203940152383, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 51620 + }, + { + "epoch": 0.51621, + "grad_norm": 1.2433717369133717, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 51621 + }, + { + "epoch": 0.51622, + "grad_norm": 1.0258132884516602, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 51622 + }, + { + "epoch": 0.51623, + "grad_norm": 1.0741589545324677, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 51623 + }, + { + "epoch": 0.51624, + "grad_norm": 1.238429802939064, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 51624 + }, + { + "epoch": 0.51625, + "grad_norm": 0.8646894993328726, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 51625 + }, + { + "epoch": 0.51626, + "grad_norm": 0.8290216538258104, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 51626 + }, + { + "epoch": 0.51627, + "grad_norm": 0.9494383158022244, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 51627 + }, + { + "epoch": 0.51628, + "grad_norm": 1.1818506972522154, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 51628 + }, + { + "epoch": 0.51629, + "grad_norm": 0.9532386195341525, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 51629 + }, + { + "epoch": 0.5163, + "grad_norm": 1.1607369527023925, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 51630 + }, + { + "epoch": 0.51631, + "grad_norm": 1.1103201499675805, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 51631 + }, + { + "epoch": 0.51632, + "grad_norm": 0.9882139431146166, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 51632 + }, + { + "epoch": 0.51633, + "grad_norm": 1.1286077860986399, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 51633 + }, + { + "epoch": 0.51634, + "grad_norm": 1.017191329237471, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 51634 + }, + { + "epoch": 0.51635, + "grad_norm": 1.1017746999845603, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 51635 + }, + { + "epoch": 0.51636, + "grad_norm": 1.1108482003178233, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 51636 + }, + { + "epoch": 0.51637, + "grad_norm": 1.0907543348028461, + "learning_rate": 0.003, + "loss": 4.02, + "step": 51637 + }, + { + "epoch": 0.51638, + "grad_norm": 1.0500005392469434, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 51638 + }, + { + "epoch": 0.51639, + "grad_norm": 1.0787598385209205, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 51639 + }, + { + "epoch": 0.5164, + "grad_norm": 1.0594557749051714, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 51640 + }, + { + "epoch": 0.51641, + "grad_norm": 1.044164899976895, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 51641 + }, + { + "epoch": 0.51642, + "grad_norm": 1.0639723866689008, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 51642 + }, + { + "epoch": 0.51643, + "grad_norm": 1.0444564361342028, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 51643 + }, + { + "epoch": 0.51644, + "grad_norm": 1.0304194020998498, + "learning_rate": 0.003, + "loss": 4.049, + "step": 51644 + }, + { + "epoch": 0.51645, + "grad_norm": 1.1810393943853972, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 51645 + }, + { + "epoch": 0.51646, + "grad_norm": 1.0404468880812108, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 51646 + }, + { + "epoch": 0.51647, + "grad_norm": 1.1570874872629302, + "learning_rate": 0.003, + "loss": 4.055, + "step": 51647 + }, + { + "epoch": 0.51648, + "grad_norm": 1.3057935796702642, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 51648 + }, + { + "epoch": 0.51649, + "grad_norm": 1.157151534413518, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 51649 + }, + { + "epoch": 0.5165, + "grad_norm": 1.007218707303275, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 51650 + }, + { + "epoch": 0.51651, + "grad_norm": 1.0175353308963413, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 51651 + }, + { + "epoch": 0.51652, + "grad_norm": 1.0328481805241614, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 51652 + }, + { + "epoch": 0.51653, + "grad_norm": 1.1412600926282335, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 51653 + }, + { + "epoch": 0.51654, + "grad_norm": 1.0546087308913357, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 51654 + }, + { + "epoch": 0.51655, + "grad_norm": 0.9488977098884929, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 51655 + }, + { + "epoch": 0.51656, + "grad_norm": 0.9077711276720088, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 51656 + }, + { + "epoch": 0.51657, + "grad_norm": 1.11448652835538, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 51657 + }, + { + "epoch": 0.51658, + "grad_norm": 1.1817132405260684, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 51658 + }, + { + "epoch": 0.51659, + "grad_norm": 0.9806067936502127, + "learning_rate": 0.003, + "loss": 3.983, + "step": 51659 + }, + { + "epoch": 0.5166, + "grad_norm": 0.95002166786699, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 51660 + }, + { + "epoch": 0.51661, + "grad_norm": 1.0331450330105119, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 51661 + }, + { + "epoch": 0.51662, + "grad_norm": 1.0730758296760288, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 51662 + }, + { + "epoch": 0.51663, + "grad_norm": 0.9466052775359631, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 51663 + }, + { + "epoch": 0.51664, + "grad_norm": 0.9702316956599976, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 51664 + }, + { + "epoch": 0.51665, + "grad_norm": 1.1518309482620772, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 51665 + }, + { + "epoch": 0.51666, + "grad_norm": 1.0418129421071058, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 51666 + }, + { + "epoch": 0.51667, + "grad_norm": 0.9231576338660378, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 51667 + }, + { + "epoch": 0.51668, + "grad_norm": 0.9403825010868126, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 51668 + }, + { + "epoch": 0.51669, + "grad_norm": 0.959001523166352, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 51669 + }, + { + "epoch": 0.5167, + "grad_norm": 0.9219055138651561, + "learning_rate": 0.003, + "loss": 4.035, + "step": 51670 + }, + { + "epoch": 0.51671, + "grad_norm": 1.019381243977837, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 51671 + }, + { + "epoch": 0.51672, + "grad_norm": 1.1418068424940682, + "learning_rate": 0.003, + "loss": 4.015, + "step": 51672 + }, + { + "epoch": 0.51673, + "grad_norm": 0.9314869495938014, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 51673 + }, + { + "epoch": 0.51674, + "grad_norm": 1.0705271323781753, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 51674 + }, + { + "epoch": 0.51675, + "grad_norm": 1.0627343167640095, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 51675 + }, + { + "epoch": 0.51676, + "grad_norm": 1.1595153747662121, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 51676 + }, + { + "epoch": 0.51677, + "grad_norm": 1.5664994489876773, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 51677 + }, + { + "epoch": 0.51678, + "grad_norm": 0.7951895647842374, + "learning_rate": 0.003, + "loss": 4.028, + "step": 51678 + }, + { + "epoch": 0.51679, + "grad_norm": 0.877697912402043, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 51679 + }, + { + "epoch": 0.5168, + "grad_norm": 1.0052984729995584, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 51680 + }, + { + "epoch": 0.51681, + "grad_norm": 1.318397258552825, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 51681 + }, + { + "epoch": 0.51682, + "grad_norm": 0.8740511768102861, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 51682 + }, + { + "epoch": 0.51683, + "grad_norm": 0.942645506265315, + "learning_rate": 0.003, + "loss": 4.017, + "step": 51683 + }, + { + "epoch": 0.51684, + "grad_norm": 1.1963059558103195, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 51684 + }, + { + "epoch": 0.51685, + "grad_norm": 1.0030351302446165, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 51685 + }, + { + "epoch": 0.51686, + "grad_norm": 1.072897036024652, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 51686 + }, + { + "epoch": 0.51687, + "grad_norm": 1.0020078770639789, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 51687 + }, + { + "epoch": 0.51688, + "grad_norm": 1.2757171583957692, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 51688 + }, + { + "epoch": 0.51689, + "grad_norm": 0.9039913679628295, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 51689 + }, + { + "epoch": 0.5169, + "grad_norm": 0.9472232825660687, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 51690 + }, + { + "epoch": 0.51691, + "grad_norm": 1.2603428350868944, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 51691 + }, + { + "epoch": 0.51692, + "grad_norm": 1.0472190018604546, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 51692 + }, + { + "epoch": 0.51693, + "grad_norm": 1.0739435461449847, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 51693 + }, + { + "epoch": 0.51694, + "grad_norm": 1.0519923359927554, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 51694 + }, + { + "epoch": 0.51695, + "grad_norm": 1.0263576266736705, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 51695 + }, + { + "epoch": 0.51696, + "grad_norm": 1.0529710747920502, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 51696 + }, + { + "epoch": 0.51697, + "grad_norm": 1.0685250813140497, + "learning_rate": 0.003, + "loss": 4.052, + "step": 51697 + }, + { + "epoch": 0.51698, + "grad_norm": 1.1075606071346331, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 51698 + }, + { + "epoch": 0.51699, + "grad_norm": 1.00513272418208, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 51699 + }, + { + "epoch": 0.517, + "grad_norm": 1.165365507022737, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 51700 + }, + { + "epoch": 0.51701, + "grad_norm": 0.973016714174497, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 51701 + }, + { + "epoch": 0.51702, + "grad_norm": 0.9763941742023237, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 51702 + }, + { + "epoch": 0.51703, + "grad_norm": 1.0984465278855065, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 51703 + }, + { + "epoch": 0.51704, + "grad_norm": 1.174712977283597, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 51704 + }, + { + "epoch": 0.51705, + "grad_norm": 0.9416451748331609, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 51705 + }, + { + "epoch": 0.51706, + "grad_norm": 1.0953331656075174, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 51706 + }, + { + "epoch": 0.51707, + "grad_norm": 1.0505902088003443, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 51707 + }, + { + "epoch": 0.51708, + "grad_norm": 1.1592345502271784, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 51708 + }, + { + "epoch": 0.51709, + "grad_norm": 0.8919441782002577, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 51709 + }, + { + "epoch": 0.5171, + "grad_norm": 0.779838678343517, + "learning_rate": 0.003, + "loss": 4.025, + "step": 51710 + }, + { + "epoch": 0.51711, + "grad_norm": 0.6956342482337761, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 51711 + }, + { + "epoch": 0.51712, + "grad_norm": 0.6720497418069151, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 51712 + }, + { + "epoch": 0.51713, + "grad_norm": 0.7984105854300034, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 51713 + }, + { + "epoch": 0.51714, + "grad_norm": 1.0050256498010655, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 51714 + }, + { + "epoch": 0.51715, + "grad_norm": 1.2760930832774164, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 51715 + }, + { + "epoch": 0.51716, + "grad_norm": 1.035763492069404, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 51716 + }, + { + "epoch": 0.51717, + "grad_norm": 1.216792705123771, + "learning_rate": 0.003, + "loss": 4.01, + "step": 51717 + }, + { + "epoch": 0.51718, + "grad_norm": 0.966068278962887, + "learning_rate": 0.003, + "loss": 4.04, + "step": 51718 + }, + { + "epoch": 0.51719, + "grad_norm": 0.9358912920051854, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 51719 + }, + { + "epoch": 0.5172, + "grad_norm": 0.9008104023645127, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 51720 + }, + { + "epoch": 0.51721, + "grad_norm": 0.9771134089908111, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 51721 + }, + { + "epoch": 0.51722, + "grad_norm": 1.3207655244308132, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 51722 + }, + { + "epoch": 0.51723, + "grad_norm": 0.9912463833177149, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 51723 + }, + { + "epoch": 0.51724, + "grad_norm": 1.189905679861531, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 51724 + }, + { + "epoch": 0.51725, + "grad_norm": 1.0464857063312862, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 51725 + }, + { + "epoch": 0.51726, + "grad_norm": 1.2521323048544024, + "learning_rate": 0.003, + "loss": 4.031, + "step": 51726 + }, + { + "epoch": 0.51727, + "grad_norm": 0.8861074020866543, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 51727 + }, + { + "epoch": 0.51728, + "grad_norm": 0.97449326077515, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 51728 + }, + { + "epoch": 0.51729, + "grad_norm": 1.1157805071231313, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 51729 + }, + { + "epoch": 0.5173, + "grad_norm": 0.9062751157441629, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 51730 + }, + { + "epoch": 0.51731, + "grad_norm": 0.9874293824179525, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 51731 + }, + { + "epoch": 0.51732, + "grad_norm": 1.0154150624897347, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 51732 + }, + { + "epoch": 0.51733, + "grad_norm": 1.396405287108666, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 51733 + }, + { + "epoch": 0.51734, + "grad_norm": 0.8022753984525197, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 51734 + }, + { + "epoch": 0.51735, + "grad_norm": 0.8720439216411503, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 51735 + }, + { + "epoch": 0.51736, + "grad_norm": 0.9417340664415976, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 51736 + }, + { + "epoch": 0.51737, + "grad_norm": 1.337262600870529, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 51737 + }, + { + "epoch": 0.51738, + "grad_norm": 1.0483695122417729, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 51738 + }, + { + "epoch": 0.51739, + "grad_norm": 1.139758202717597, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 51739 + }, + { + "epoch": 0.5174, + "grad_norm": 1.3425726147650991, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 51740 + }, + { + "epoch": 0.51741, + "grad_norm": 1.013457812162115, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 51741 + }, + { + "epoch": 0.51742, + "grad_norm": 1.2282643054151912, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 51742 + }, + { + "epoch": 0.51743, + "grad_norm": 0.8491925542119102, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 51743 + }, + { + "epoch": 0.51744, + "grad_norm": 0.8059305365220486, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 51744 + }, + { + "epoch": 0.51745, + "grad_norm": 0.8450330510165478, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 51745 + }, + { + "epoch": 0.51746, + "grad_norm": 0.9277817329614083, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 51746 + }, + { + "epoch": 0.51747, + "grad_norm": 1.0448446764744086, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 51747 + }, + { + "epoch": 0.51748, + "grad_norm": 1.2148069274185678, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 51748 + }, + { + "epoch": 0.51749, + "grad_norm": 0.9455926645194747, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 51749 + }, + { + "epoch": 0.5175, + "grad_norm": 1.1362199562362916, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 51750 + }, + { + "epoch": 0.51751, + "grad_norm": 1.1168982534659222, + "learning_rate": 0.003, + "loss": 4.039, + "step": 51751 + }, + { + "epoch": 0.51752, + "grad_norm": 0.8712109115631769, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 51752 + }, + { + "epoch": 0.51753, + "grad_norm": 1.1001789352627604, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 51753 + }, + { + "epoch": 0.51754, + "grad_norm": 1.3043023904263957, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 51754 + }, + { + "epoch": 0.51755, + "grad_norm": 0.9709247101330257, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 51755 + }, + { + "epoch": 0.51756, + "grad_norm": 1.0079967484645573, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 51756 + }, + { + "epoch": 0.51757, + "grad_norm": 1.1726808752097015, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 51757 + }, + { + "epoch": 0.51758, + "grad_norm": 0.933054354474804, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 51758 + }, + { + "epoch": 0.51759, + "grad_norm": 1.1433508028173336, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 51759 + }, + { + "epoch": 0.5176, + "grad_norm": 1.1974719573135748, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 51760 + }, + { + "epoch": 0.51761, + "grad_norm": 0.9165167231155443, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 51761 + }, + { + "epoch": 0.51762, + "grad_norm": 0.8231431623812482, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 51762 + }, + { + "epoch": 0.51763, + "grad_norm": 0.8489517604577144, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 51763 + }, + { + "epoch": 0.51764, + "grad_norm": 1.1406143667590596, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 51764 + }, + { + "epoch": 0.51765, + "grad_norm": 1.1627512999319574, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 51765 + }, + { + "epoch": 0.51766, + "grad_norm": 0.9806809781412712, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 51766 + }, + { + "epoch": 0.51767, + "grad_norm": 0.9615712748306764, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 51767 + }, + { + "epoch": 0.51768, + "grad_norm": 1.1861711096570575, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 51768 + }, + { + "epoch": 0.51769, + "grad_norm": 0.9866407417613056, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 51769 + }, + { + "epoch": 0.5177, + "grad_norm": 1.0788459004495732, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 51770 + }, + { + "epoch": 0.51771, + "grad_norm": 1.1178063311279898, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 51771 + }, + { + "epoch": 0.51772, + "grad_norm": 1.070804238192607, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 51772 + }, + { + "epoch": 0.51773, + "grad_norm": 0.9451627056585133, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 51773 + }, + { + "epoch": 0.51774, + "grad_norm": 1.0009116422281203, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 51774 + }, + { + "epoch": 0.51775, + "grad_norm": 1.3563033792866703, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 51775 + }, + { + "epoch": 0.51776, + "grad_norm": 0.9748488919872952, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 51776 + }, + { + "epoch": 0.51777, + "grad_norm": 1.092897574829363, + "learning_rate": 0.003, + "loss": 4.02, + "step": 51777 + }, + { + "epoch": 0.51778, + "grad_norm": 1.117975176623707, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 51778 + }, + { + "epoch": 0.51779, + "grad_norm": 0.9762018015916699, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 51779 + }, + { + "epoch": 0.5178, + "grad_norm": 0.945598930700229, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 51780 + }, + { + "epoch": 0.51781, + "grad_norm": 1.0755745200607059, + "learning_rate": 0.003, + "loss": 4.032, + "step": 51781 + }, + { + "epoch": 0.51782, + "grad_norm": 0.9802153528702412, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 51782 + }, + { + "epoch": 0.51783, + "grad_norm": 1.067663682646381, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 51783 + }, + { + "epoch": 0.51784, + "grad_norm": 1.1891554251325345, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 51784 + }, + { + "epoch": 0.51785, + "grad_norm": 0.9152474758075608, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 51785 + }, + { + "epoch": 0.51786, + "grad_norm": 0.8655195168208802, + "learning_rate": 0.003, + "loss": 4.031, + "step": 51786 + }, + { + "epoch": 0.51787, + "grad_norm": 0.9322534310292361, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 51787 + }, + { + "epoch": 0.51788, + "grad_norm": 1.1873864780030288, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 51788 + }, + { + "epoch": 0.51789, + "grad_norm": 1.0856334702974337, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 51789 + }, + { + "epoch": 0.5179, + "grad_norm": 1.2773158980412491, + "learning_rate": 0.003, + "loss": 4.061, + "step": 51790 + }, + { + "epoch": 0.51791, + "grad_norm": 1.1431295796694159, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 51791 + }, + { + "epoch": 0.51792, + "grad_norm": 1.0246547628406908, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 51792 + }, + { + "epoch": 0.51793, + "grad_norm": 1.0224199953555326, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 51793 + }, + { + "epoch": 0.51794, + "grad_norm": 1.0924446945040316, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 51794 + }, + { + "epoch": 0.51795, + "grad_norm": 1.0288874301895365, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 51795 + }, + { + "epoch": 0.51796, + "grad_norm": 1.0826317725643162, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 51796 + }, + { + "epoch": 0.51797, + "grad_norm": 1.0318571781987704, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 51797 + }, + { + "epoch": 0.51798, + "grad_norm": 1.1717149349351719, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 51798 + }, + { + "epoch": 0.51799, + "grad_norm": 0.7445147417271167, + "learning_rate": 0.003, + "loss": 4.038, + "step": 51799 + }, + { + "epoch": 0.518, + "grad_norm": 0.8231068218941745, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 51800 + }, + { + "epoch": 0.51801, + "grad_norm": 1.0465565818226115, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 51801 + }, + { + "epoch": 0.51802, + "grad_norm": 1.1598029583410518, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 51802 + }, + { + "epoch": 0.51803, + "grad_norm": 1.051326890591547, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 51803 + }, + { + "epoch": 0.51804, + "grad_norm": 1.046910256638884, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 51804 + }, + { + "epoch": 0.51805, + "grad_norm": 1.064804968289778, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 51805 + }, + { + "epoch": 0.51806, + "grad_norm": 1.0418099471792057, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 51806 + }, + { + "epoch": 0.51807, + "grad_norm": 0.8497859625384753, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 51807 + }, + { + "epoch": 0.51808, + "grad_norm": 0.9754611867356686, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 51808 + }, + { + "epoch": 0.51809, + "grad_norm": 1.0484731441016295, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 51809 + }, + { + "epoch": 0.5181, + "grad_norm": 0.9358067901324362, + "learning_rate": 0.003, + "loss": 4.026, + "step": 51810 + }, + { + "epoch": 0.51811, + "grad_norm": 1.1397412185025249, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 51811 + }, + { + "epoch": 0.51812, + "grad_norm": 1.0783671079383934, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 51812 + }, + { + "epoch": 0.51813, + "grad_norm": 1.0232097388570787, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 51813 + }, + { + "epoch": 0.51814, + "grad_norm": 1.2237207387491325, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 51814 + }, + { + "epoch": 0.51815, + "grad_norm": 1.1696340668280747, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 51815 + }, + { + "epoch": 0.51816, + "grad_norm": 1.149793905859327, + "learning_rate": 0.003, + "loss": 4.029, + "step": 51816 + }, + { + "epoch": 0.51817, + "grad_norm": 0.907106212792136, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 51817 + }, + { + "epoch": 0.51818, + "grad_norm": 1.052651354593618, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 51818 + }, + { + "epoch": 0.51819, + "grad_norm": 1.2364781051856453, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 51819 + }, + { + "epoch": 0.5182, + "grad_norm": 0.8211149953468962, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 51820 + }, + { + "epoch": 0.51821, + "grad_norm": 0.8394299310787472, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 51821 + }, + { + "epoch": 0.51822, + "grad_norm": 1.0859030671303085, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 51822 + }, + { + "epoch": 0.51823, + "grad_norm": 1.161193872759313, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 51823 + }, + { + "epoch": 0.51824, + "grad_norm": 0.9348138872795648, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 51824 + }, + { + "epoch": 0.51825, + "grad_norm": 0.9358569054908817, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 51825 + }, + { + "epoch": 0.51826, + "grad_norm": 1.0037280165168492, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 51826 + }, + { + "epoch": 0.51827, + "grad_norm": 1.096402665776661, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 51827 + }, + { + "epoch": 0.51828, + "grad_norm": 1.2309121843872932, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 51828 + }, + { + "epoch": 0.51829, + "grad_norm": 0.9728247887100577, + "learning_rate": 0.003, + "loss": 4.017, + "step": 51829 + }, + { + "epoch": 0.5183, + "grad_norm": 1.073313809014844, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 51830 + }, + { + "epoch": 0.51831, + "grad_norm": 0.8914789841920242, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 51831 + }, + { + "epoch": 0.51832, + "grad_norm": 0.9002164189888135, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 51832 + }, + { + "epoch": 0.51833, + "grad_norm": 0.9962058430072394, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 51833 + }, + { + "epoch": 0.51834, + "grad_norm": 1.243821276945901, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 51834 + }, + { + "epoch": 0.51835, + "grad_norm": 0.9598317629806365, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 51835 + }, + { + "epoch": 0.51836, + "grad_norm": 1.179673296843085, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 51836 + }, + { + "epoch": 0.51837, + "grad_norm": 0.9587318758513439, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 51837 + }, + { + "epoch": 0.51838, + "grad_norm": 1.0804589437052643, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 51838 + }, + { + "epoch": 0.51839, + "grad_norm": 1.2254947860304124, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 51839 + }, + { + "epoch": 0.5184, + "grad_norm": 1.04461933511887, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 51840 + }, + { + "epoch": 0.51841, + "grad_norm": 1.1622981431447243, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 51841 + }, + { + "epoch": 0.51842, + "grad_norm": 1.0354074714240349, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 51842 + }, + { + "epoch": 0.51843, + "grad_norm": 1.2294426571562616, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 51843 + }, + { + "epoch": 0.51844, + "grad_norm": 0.947766234475969, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 51844 + }, + { + "epoch": 0.51845, + "grad_norm": 1.0544607823182919, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 51845 + }, + { + "epoch": 0.51846, + "grad_norm": 0.9503145394895945, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 51846 + }, + { + "epoch": 0.51847, + "grad_norm": 0.8605190280385491, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 51847 + }, + { + "epoch": 0.51848, + "grad_norm": 0.9153268854906712, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 51848 + }, + { + "epoch": 0.51849, + "grad_norm": 0.8770810677157003, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 51849 + }, + { + "epoch": 0.5185, + "grad_norm": 0.9533170672349192, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 51850 + }, + { + "epoch": 0.51851, + "grad_norm": 1.0569710338164517, + "learning_rate": 0.003, + "loss": 4.036, + "step": 51851 + }, + { + "epoch": 0.51852, + "grad_norm": 0.9924938345799903, + "learning_rate": 0.003, + "loss": 4.012, + "step": 51852 + }, + { + "epoch": 0.51853, + "grad_norm": 1.115991832500027, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 51853 + }, + { + "epoch": 0.51854, + "grad_norm": 1.1603016390720944, + "learning_rate": 0.003, + "loss": 4.042, + "step": 51854 + }, + { + "epoch": 0.51855, + "grad_norm": 1.0520952347505772, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 51855 + }, + { + "epoch": 0.51856, + "grad_norm": 1.0677523608269177, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 51856 + }, + { + "epoch": 0.51857, + "grad_norm": 1.2338032170683726, + "learning_rate": 0.003, + "loss": 4.024, + "step": 51857 + }, + { + "epoch": 0.51858, + "grad_norm": 1.0918470071114015, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 51858 + }, + { + "epoch": 0.51859, + "grad_norm": 1.055281246545636, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 51859 + }, + { + "epoch": 0.5186, + "grad_norm": 0.9391743483660959, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 51860 + }, + { + "epoch": 0.51861, + "grad_norm": 0.9290847782067093, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 51861 + }, + { + "epoch": 0.51862, + "grad_norm": 0.9977170328965032, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 51862 + }, + { + "epoch": 0.51863, + "grad_norm": 0.9899045863242918, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 51863 + }, + { + "epoch": 0.51864, + "grad_norm": 1.1397632735535472, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 51864 + }, + { + "epoch": 0.51865, + "grad_norm": 1.2556302113262405, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 51865 + }, + { + "epoch": 0.51866, + "grad_norm": 0.9558039020586742, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 51866 + }, + { + "epoch": 0.51867, + "grad_norm": 0.870650692209168, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 51867 + }, + { + "epoch": 0.51868, + "grad_norm": 1.010775234040795, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 51868 + }, + { + "epoch": 0.51869, + "grad_norm": 1.2364473143771264, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 51869 + }, + { + "epoch": 0.5187, + "grad_norm": 1.1328623816199268, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 51870 + }, + { + "epoch": 0.51871, + "grad_norm": 0.9821261031446041, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 51871 + }, + { + "epoch": 0.51872, + "grad_norm": 1.15261713155942, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 51872 + }, + { + "epoch": 0.51873, + "grad_norm": 1.3086166504415906, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 51873 + }, + { + "epoch": 0.51874, + "grad_norm": 0.9429464575513864, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 51874 + }, + { + "epoch": 0.51875, + "grad_norm": 1.0288533218007345, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 51875 + }, + { + "epoch": 0.51876, + "grad_norm": 1.0910808419408684, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 51876 + }, + { + "epoch": 0.51877, + "grad_norm": 1.2814656382832157, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 51877 + }, + { + "epoch": 0.51878, + "grad_norm": 0.7977706398606739, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 51878 + }, + { + "epoch": 0.51879, + "grad_norm": 0.8226641805854364, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 51879 + }, + { + "epoch": 0.5188, + "grad_norm": 0.9125545557883947, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 51880 + }, + { + "epoch": 0.51881, + "grad_norm": 0.9637603078118472, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 51881 + }, + { + "epoch": 0.51882, + "grad_norm": 1.3268616783146705, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 51882 + }, + { + "epoch": 0.51883, + "grad_norm": 0.873653263207897, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 51883 + }, + { + "epoch": 0.51884, + "grad_norm": 0.90786271140163, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 51884 + }, + { + "epoch": 0.51885, + "grad_norm": 1.0641501676584098, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 51885 + }, + { + "epoch": 0.51886, + "grad_norm": 1.3195076251784146, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 51886 + }, + { + "epoch": 0.51887, + "grad_norm": 1.0809393770889335, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 51887 + }, + { + "epoch": 0.51888, + "grad_norm": 1.1883530470599566, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 51888 + }, + { + "epoch": 0.51889, + "grad_norm": 1.0764608320349847, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 51889 + }, + { + "epoch": 0.5189, + "grad_norm": 1.1345883835494521, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 51890 + }, + { + "epoch": 0.51891, + "grad_norm": 0.94621465992375, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 51891 + }, + { + "epoch": 0.51892, + "grad_norm": 1.036885200031019, + "learning_rate": 0.003, + "loss": 4.02, + "step": 51892 + }, + { + "epoch": 0.51893, + "grad_norm": 1.3398363075360784, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 51893 + }, + { + "epoch": 0.51894, + "grad_norm": 1.0074962345590637, + "learning_rate": 0.003, + "loss": 4.03, + "step": 51894 + }, + { + "epoch": 0.51895, + "grad_norm": 0.9264586679178114, + "learning_rate": 0.003, + "loss": 4.046, + "step": 51895 + }, + { + "epoch": 0.51896, + "grad_norm": 0.9314619078846731, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 51896 + }, + { + "epoch": 0.51897, + "grad_norm": 1.0759409401832756, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 51897 + }, + { + "epoch": 0.51898, + "grad_norm": 0.9129524804398016, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 51898 + }, + { + "epoch": 0.51899, + "grad_norm": 1.0516109892662326, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 51899 + }, + { + "epoch": 0.519, + "grad_norm": 1.1232279278074428, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 51900 + }, + { + "epoch": 0.51901, + "grad_norm": 0.9209424248942074, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 51901 + }, + { + "epoch": 0.51902, + "grad_norm": 1.0097343105009209, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 51902 + }, + { + "epoch": 0.51903, + "grad_norm": 1.3423007893743175, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 51903 + }, + { + "epoch": 0.51904, + "grad_norm": 0.8279020528299075, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 51904 + }, + { + "epoch": 0.51905, + "grad_norm": 0.7879567305108078, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 51905 + }, + { + "epoch": 0.51906, + "grad_norm": 0.8542848237332619, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 51906 + }, + { + "epoch": 0.51907, + "grad_norm": 1.1270425086829094, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 51907 + }, + { + "epoch": 0.51908, + "grad_norm": 1.1023515195999631, + "learning_rate": 0.003, + "loss": 4.025, + "step": 51908 + }, + { + "epoch": 0.51909, + "grad_norm": 1.309261028454588, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 51909 + }, + { + "epoch": 0.5191, + "grad_norm": 1.057218048767696, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 51910 + }, + { + "epoch": 0.51911, + "grad_norm": 1.320701410732809, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 51911 + }, + { + "epoch": 0.51912, + "grad_norm": 0.7712901409199288, + "learning_rate": 0.003, + "loss": 4.006, + "step": 51912 + }, + { + "epoch": 0.51913, + "grad_norm": 0.8322509997177235, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 51913 + }, + { + "epoch": 0.51914, + "grad_norm": 1.2631641957316735, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 51914 + }, + { + "epoch": 0.51915, + "grad_norm": 1.2055388116122887, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 51915 + }, + { + "epoch": 0.51916, + "grad_norm": 0.9146562920977701, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 51916 + }, + { + "epoch": 0.51917, + "grad_norm": 1.158768694868664, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 51917 + }, + { + "epoch": 0.51918, + "grad_norm": 0.9776754513453182, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 51918 + }, + { + "epoch": 0.51919, + "grad_norm": 1.148435237482733, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 51919 + }, + { + "epoch": 0.5192, + "grad_norm": 1.0924273736272554, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 51920 + }, + { + "epoch": 0.51921, + "grad_norm": 0.9219290047202707, + "learning_rate": 0.003, + "loss": 4.005, + "step": 51921 + }, + { + "epoch": 0.51922, + "grad_norm": 0.826667635832606, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 51922 + }, + { + "epoch": 0.51923, + "grad_norm": 0.9548034728968453, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 51923 + }, + { + "epoch": 0.51924, + "grad_norm": 1.0082532151577204, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 51924 + }, + { + "epoch": 0.51925, + "grad_norm": 1.4090058314533387, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 51925 + }, + { + "epoch": 0.51926, + "grad_norm": 1.1549169954921392, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 51926 + }, + { + "epoch": 0.51927, + "grad_norm": 0.8650938038915789, + "learning_rate": 0.003, + "loss": 4.026, + "step": 51927 + }, + { + "epoch": 0.51928, + "grad_norm": 0.8482142005490416, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 51928 + }, + { + "epoch": 0.51929, + "grad_norm": 0.9309208751760651, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 51929 + }, + { + "epoch": 0.5193, + "grad_norm": 1.1294319058308162, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 51930 + }, + { + "epoch": 0.51931, + "grad_norm": 1.1217270689618466, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 51931 + }, + { + "epoch": 0.51932, + "grad_norm": 1.1840449490585014, + "learning_rate": 0.003, + "loss": 4.021, + "step": 51932 + }, + { + "epoch": 0.51933, + "grad_norm": 1.112604849979629, + "learning_rate": 0.003, + "loss": 4.035, + "step": 51933 + }, + { + "epoch": 0.51934, + "grad_norm": 1.2723605413664436, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 51934 + }, + { + "epoch": 0.51935, + "grad_norm": 0.8175502758284993, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 51935 + }, + { + "epoch": 0.51936, + "grad_norm": 0.7372475335806274, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 51936 + }, + { + "epoch": 0.51937, + "grad_norm": 0.8160344602950068, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 51937 + }, + { + "epoch": 0.51938, + "grad_norm": 1.0965201319684497, + "learning_rate": 0.003, + "loss": 4.005, + "step": 51938 + }, + { + "epoch": 0.51939, + "grad_norm": 1.4747930134000216, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 51939 + }, + { + "epoch": 0.5194, + "grad_norm": 0.7853760742221403, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 51940 + }, + { + "epoch": 0.51941, + "grad_norm": 0.7480904194188703, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 51941 + }, + { + "epoch": 0.51942, + "grad_norm": 0.8445419042253685, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 51942 + }, + { + "epoch": 0.51943, + "grad_norm": 0.9568868706815602, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 51943 + }, + { + "epoch": 0.51944, + "grad_norm": 0.9519713995062681, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 51944 + }, + { + "epoch": 0.51945, + "grad_norm": 1.1031785691064242, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 51945 + }, + { + "epoch": 0.51946, + "grad_norm": 1.1015550628725153, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 51946 + }, + { + "epoch": 0.51947, + "grad_norm": 0.9937627650587841, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 51947 + }, + { + "epoch": 0.51948, + "grad_norm": 1.3576371122286912, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 51948 + }, + { + "epoch": 0.51949, + "grad_norm": 1.039441595722718, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 51949 + }, + { + "epoch": 0.5195, + "grad_norm": 1.0563299300286741, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 51950 + }, + { + "epoch": 0.51951, + "grad_norm": 0.9582820627042377, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 51951 + }, + { + "epoch": 0.51952, + "grad_norm": 1.1486054693610612, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 51952 + }, + { + "epoch": 0.51953, + "grad_norm": 1.1752094291930097, + "learning_rate": 0.003, + "loss": 4.05, + "step": 51953 + }, + { + "epoch": 0.51954, + "grad_norm": 1.1044537632595246, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 51954 + }, + { + "epoch": 0.51955, + "grad_norm": 0.9763160864901482, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 51955 + }, + { + "epoch": 0.51956, + "grad_norm": 1.228515138869417, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 51956 + }, + { + "epoch": 0.51957, + "grad_norm": 0.9943506621748649, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 51957 + }, + { + "epoch": 0.51958, + "grad_norm": 1.0500513512476968, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 51958 + }, + { + "epoch": 0.51959, + "grad_norm": 0.9345930955324471, + "learning_rate": 0.003, + "loss": 4.028, + "step": 51959 + }, + { + "epoch": 0.5196, + "grad_norm": 1.2310128715193158, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 51960 + }, + { + "epoch": 0.51961, + "grad_norm": 1.1044938233189705, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 51961 + }, + { + "epoch": 0.51962, + "grad_norm": 1.0702528900593598, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 51962 + }, + { + "epoch": 0.51963, + "grad_norm": 1.084265433161052, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 51963 + }, + { + "epoch": 0.51964, + "grad_norm": 0.9398599977896287, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 51964 + }, + { + "epoch": 0.51965, + "grad_norm": 1.0168572638992475, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 51965 + }, + { + "epoch": 0.51966, + "grad_norm": 1.1058867265426275, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 51966 + }, + { + "epoch": 0.51967, + "grad_norm": 0.940950655427101, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 51967 + }, + { + "epoch": 0.51968, + "grad_norm": 1.4052963411566486, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 51968 + }, + { + "epoch": 0.51969, + "grad_norm": 0.9887037647693936, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 51969 + }, + { + "epoch": 0.5197, + "grad_norm": 1.070136388308676, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 51970 + }, + { + "epoch": 0.51971, + "grad_norm": 1.0328558347572152, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 51971 + }, + { + "epoch": 0.51972, + "grad_norm": 1.2758664496312229, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 51972 + }, + { + "epoch": 0.51973, + "grad_norm": 0.9511330756203464, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 51973 + }, + { + "epoch": 0.51974, + "grad_norm": 0.9126596323507774, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 51974 + }, + { + "epoch": 0.51975, + "grad_norm": 1.1022064940775638, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 51975 + }, + { + "epoch": 0.51976, + "grad_norm": 1.2295148463296626, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 51976 + }, + { + "epoch": 0.51977, + "grad_norm": 0.9382804086321455, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 51977 + }, + { + "epoch": 0.51978, + "grad_norm": 1.1018974507421262, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 51978 + }, + { + "epoch": 0.51979, + "grad_norm": 1.0463208455165391, + "learning_rate": 0.003, + "loss": 4.017, + "step": 51979 + }, + { + "epoch": 0.5198, + "grad_norm": 1.2736882492719177, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 51980 + }, + { + "epoch": 0.51981, + "grad_norm": 1.1111565744000564, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 51981 + }, + { + "epoch": 0.51982, + "grad_norm": 1.0659437680002295, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 51982 + }, + { + "epoch": 0.51983, + "grad_norm": 1.0788872900272766, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 51983 + }, + { + "epoch": 0.51984, + "grad_norm": 1.2037134004218608, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 51984 + }, + { + "epoch": 0.51985, + "grad_norm": 1.0542081903306966, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 51985 + }, + { + "epoch": 0.51986, + "grad_norm": 1.0213722045970088, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 51986 + }, + { + "epoch": 0.51987, + "grad_norm": 0.9836596709892343, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 51987 + }, + { + "epoch": 0.51988, + "grad_norm": 1.0747844255608996, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 51988 + }, + { + "epoch": 0.51989, + "grad_norm": 0.9757139002830024, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 51989 + }, + { + "epoch": 0.5199, + "grad_norm": 1.0455716314993684, + "learning_rate": 0.003, + "loss": 4.039, + "step": 51990 + }, + { + "epoch": 0.51991, + "grad_norm": 0.9324083003905798, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 51991 + }, + { + "epoch": 0.51992, + "grad_norm": 0.9684063687972836, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 51992 + }, + { + "epoch": 0.51993, + "grad_norm": 0.9486115946079992, + "learning_rate": 0.003, + "loss": 4.039, + "step": 51993 + }, + { + "epoch": 0.51994, + "grad_norm": 1.088527991930432, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 51994 + }, + { + "epoch": 0.51995, + "grad_norm": 1.1984031846501584, + "learning_rate": 0.003, + "loss": 4.054, + "step": 51995 + }, + { + "epoch": 0.51996, + "grad_norm": 1.0013512318335285, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 51996 + }, + { + "epoch": 0.51997, + "grad_norm": 0.98833563582233, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 51997 + }, + { + "epoch": 0.51998, + "grad_norm": 1.2538989936940703, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 51998 + }, + { + "epoch": 0.51999, + "grad_norm": 1.005867958587519, + "learning_rate": 0.003, + "loss": 4.048, + "step": 51999 + }, + { + "epoch": 0.52, + "grad_norm": 1.041885430358958, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 52000 + }, + { + "epoch": 0.52001, + "grad_norm": 1.0787798876323946, + "learning_rate": 0.003, + "loss": 4.025, + "step": 52001 + }, + { + "epoch": 0.52002, + "grad_norm": 1.1165835836344444, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 52002 + }, + { + "epoch": 0.52003, + "grad_norm": 0.9499349493237674, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 52003 + }, + { + "epoch": 0.52004, + "grad_norm": 0.8153958935802487, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 52004 + }, + { + "epoch": 0.52005, + "grad_norm": 0.93963403223258, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 52005 + }, + { + "epoch": 0.52006, + "grad_norm": 1.211644079587557, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 52006 + }, + { + "epoch": 0.52007, + "grad_norm": 1.1861850093866064, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 52007 + }, + { + "epoch": 0.52008, + "grad_norm": 1.0190040550070996, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 52008 + }, + { + "epoch": 0.52009, + "grad_norm": 1.0840159495560366, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 52009 + }, + { + "epoch": 0.5201, + "grad_norm": 1.0811849227231805, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 52010 + }, + { + "epoch": 0.52011, + "grad_norm": 1.2068250188792011, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 52011 + }, + { + "epoch": 0.52012, + "grad_norm": 0.9915448989685625, + "learning_rate": 0.003, + "loss": 4.027, + "step": 52012 + }, + { + "epoch": 0.52013, + "grad_norm": 1.1408076155943623, + "learning_rate": 0.003, + "loss": 4.049, + "step": 52013 + }, + { + "epoch": 0.52014, + "grad_norm": 1.0445452362575238, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 52014 + }, + { + "epoch": 0.52015, + "grad_norm": 1.0308092940776483, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 52015 + }, + { + "epoch": 0.52016, + "grad_norm": 1.1176737401125623, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 52016 + }, + { + "epoch": 0.52017, + "grad_norm": 0.9629517679474403, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 52017 + }, + { + "epoch": 0.52018, + "grad_norm": 0.9876167015870487, + "learning_rate": 0.003, + "loss": 4.053, + "step": 52018 + }, + { + "epoch": 0.52019, + "grad_norm": 0.9640962190584401, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 52019 + }, + { + "epoch": 0.5202, + "grad_norm": 1.1236054365947288, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 52020 + }, + { + "epoch": 0.52021, + "grad_norm": 1.0959507232914691, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 52021 + }, + { + "epoch": 0.52022, + "grad_norm": 1.193301611681895, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 52022 + }, + { + "epoch": 0.52023, + "grad_norm": 0.837379271507973, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 52023 + }, + { + "epoch": 0.52024, + "grad_norm": 0.8890873235611029, + "learning_rate": 0.003, + "loss": 4.027, + "step": 52024 + }, + { + "epoch": 0.52025, + "grad_norm": 1.0025099824843355, + "learning_rate": 0.003, + "loss": 4.031, + "step": 52025 + }, + { + "epoch": 0.52026, + "grad_norm": 1.061840229313926, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 52026 + }, + { + "epoch": 0.52027, + "grad_norm": 1.2860296482274145, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 52027 + }, + { + "epoch": 0.52028, + "grad_norm": 0.8913439387248397, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 52028 + }, + { + "epoch": 0.52029, + "grad_norm": 0.9973467916234565, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 52029 + }, + { + "epoch": 0.5203, + "grad_norm": 1.234742299154602, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 52030 + }, + { + "epoch": 0.52031, + "grad_norm": 1.0364980262744248, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 52031 + }, + { + "epoch": 0.52032, + "grad_norm": 1.2556776317167617, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 52032 + }, + { + "epoch": 0.52033, + "grad_norm": 0.9674078053131613, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 52033 + }, + { + "epoch": 0.52034, + "grad_norm": 1.0918964352534068, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 52034 + }, + { + "epoch": 0.52035, + "grad_norm": 1.2971501858688121, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 52035 + }, + { + "epoch": 0.52036, + "grad_norm": 1.0898768671764296, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 52036 + }, + { + "epoch": 0.52037, + "grad_norm": 0.8344082095548762, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 52037 + }, + { + "epoch": 0.52038, + "grad_norm": 0.9061404922951821, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 52038 + }, + { + "epoch": 0.52039, + "grad_norm": 1.2365583667335118, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 52039 + }, + { + "epoch": 0.5204, + "grad_norm": 1.173518653633505, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 52040 + }, + { + "epoch": 0.52041, + "grad_norm": 0.9645215244708597, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 52041 + }, + { + "epoch": 0.52042, + "grad_norm": 0.9862766486982523, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 52042 + }, + { + "epoch": 0.52043, + "grad_norm": 1.1624406416285813, + "learning_rate": 0.003, + "loss": 4.013, + "step": 52043 + }, + { + "epoch": 0.52044, + "grad_norm": 0.9957107778128651, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 52044 + }, + { + "epoch": 0.52045, + "grad_norm": 1.1427901596984877, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 52045 + }, + { + "epoch": 0.52046, + "grad_norm": 0.9459678847083886, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 52046 + }, + { + "epoch": 0.52047, + "grad_norm": 0.9699077424413225, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 52047 + }, + { + "epoch": 0.52048, + "grad_norm": 0.8924789358498006, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 52048 + }, + { + "epoch": 0.52049, + "grad_norm": 0.9807960389157632, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 52049 + }, + { + "epoch": 0.5205, + "grad_norm": 1.258548696946895, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 52050 + }, + { + "epoch": 0.52051, + "grad_norm": 1.0388430232199204, + "learning_rate": 0.003, + "loss": 4.004, + "step": 52051 + }, + { + "epoch": 0.52052, + "grad_norm": 1.0941025907990969, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 52052 + }, + { + "epoch": 0.52053, + "grad_norm": 0.8819705227294519, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 52053 + }, + { + "epoch": 0.52054, + "grad_norm": 0.8710895601041982, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 52054 + }, + { + "epoch": 0.52055, + "grad_norm": 1.0235442799161298, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 52055 + }, + { + "epoch": 0.52056, + "grad_norm": 1.4722425220101427, + "learning_rate": 0.003, + "loss": 4.038, + "step": 52056 + }, + { + "epoch": 0.52057, + "grad_norm": 0.6636263417995756, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 52057 + }, + { + "epoch": 0.52058, + "grad_norm": 0.7855935569806477, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 52058 + }, + { + "epoch": 0.52059, + "grad_norm": 1.0057881029732203, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 52059 + }, + { + "epoch": 0.5206, + "grad_norm": 1.4600554988224541, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 52060 + }, + { + "epoch": 0.52061, + "grad_norm": 0.7530374744719568, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 52061 + }, + { + "epoch": 0.52062, + "grad_norm": 0.8602506136614027, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 52062 + }, + { + "epoch": 0.52063, + "grad_norm": 0.9932971985673114, + "learning_rate": 0.003, + "loss": 4.024, + "step": 52063 + }, + { + "epoch": 0.52064, + "grad_norm": 1.30615287547847, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 52064 + }, + { + "epoch": 0.52065, + "grad_norm": 0.9884955481245016, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 52065 + }, + { + "epoch": 0.52066, + "grad_norm": 1.302763572861002, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 52066 + }, + { + "epoch": 0.52067, + "grad_norm": 0.9174812235467189, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 52067 + }, + { + "epoch": 0.52068, + "grad_norm": 0.9493675743442528, + "learning_rate": 0.003, + "loss": 4.042, + "step": 52068 + }, + { + "epoch": 0.52069, + "grad_norm": 1.205592631831318, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 52069 + }, + { + "epoch": 0.5207, + "grad_norm": 1.0087356198214907, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 52070 + }, + { + "epoch": 0.52071, + "grad_norm": 1.0082865084679418, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 52071 + }, + { + "epoch": 0.52072, + "grad_norm": 1.261336111655982, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 52072 + }, + { + "epoch": 0.52073, + "grad_norm": 0.7549448750018329, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 52073 + }, + { + "epoch": 0.52074, + "grad_norm": 0.7080517254106767, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 52074 + }, + { + "epoch": 0.52075, + "grad_norm": 1.1795392754187497, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 52075 + }, + { + "epoch": 0.52076, + "grad_norm": 1.2174504106838016, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 52076 + }, + { + "epoch": 0.52077, + "grad_norm": 0.9197339425477279, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 52077 + }, + { + "epoch": 0.52078, + "grad_norm": 1.212164520310179, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 52078 + }, + { + "epoch": 0.52079, + "grad_norm": 1.1326281034104217, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 52079 + }, + { + "epoch": 0.5208, + "grad_norm": 0.8255443227286761, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 52080 + }, + { + "epoch": 0.52081, + "grad_norm": 0.7864991279797119, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 52081 + }, + { + "epoch": 0.52082, + "grad_norm": 0.8969701554499162, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 52082 + }, + { + "epoch": 0.52083, + "grad_norm": 1.0863222352509831, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 52083 + }, + { + "epoch": 0.52084, + "grad_norm": 1.243622154907313, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 52084 + }, + { + "epoch": 0.52085, + "grad_norm": 0.8898245984017807, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 52085 + }, + { + "epoch": 0.52086, + "grad_norm": 0.9832969127404546, + "learning_rate": 0.003, + "loss": 4.042, + "step": 52086 + }, + { + "epoch": 0.52087, + "grad_norm": 1.0768314451888066, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 52087 + }, + { + "epoch": 0.52088, + "grad_norm": 1.1530715583213604, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 52088 + }, + { + "epoch": 0.52089, + "grad_norm": 1.2528049450709817, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 52089 + }, + { + "epoch": 0.5209, + "grad_norm": 0.7839758579368546, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 52090 + }, + { + "epoch": 0.52091, + "grad_norm": 0.6692069156332521, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 52091 + }, + { + "epoch": 0.52092, + "grad_norm": 0.7610153856322475, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 52092 + }, + { + "epoch": 0.52093, + "grad_norm": 0.8583869815460297, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 52093 + }, + { + "epoch": 0.52094, + "grad_norm": 0.9384027659241393, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 52094 + }, + { + "epoch": 0.52095, + "grad_norm": 0.9540943367507685, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 52095 + }, + { + "epoch": 0.52096, + "grad_norm": 1.03840467930665, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 52096 + }, + { + "epoch": 0.52097, + "grad_norm": 1.2216783027526013, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 52097 + }, + { + "epoch": 0.52098, + "grad_norm": 0.8687512863486025, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 52098 + }, + { + "epoch": 0.52099, + "grad_norm": 1.1479067892280224, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 52099 + }, + { + "epoch": 0.521, + "grad_norm": 1.5245436273983204, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 52100 + }, + { + "epoch": 0.52101, + "grad_norm": 0.8164277793271707, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 52101 + }, + { + "epoch": 0.52102, + "grad_norm": 0.8325318314860113, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 52102 + }, + { + "epoch": 0.52103, + "grad_norm": 0.8428634355866514, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 52103 + }, + { + "epoch": 0.52104, + "grad_norm": 0.9535046991465608, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 52104 + }, + { + "epoch": 0.52105, + "grad_norm": 1.1988450629699257, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 52105 + }, + { + "epoch": 0.52106, + "grad_norm": 1.1557682770001634, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 52106 + }, + { + "epoch": 0.52107, + "grad_norm": 1.0020384011721928, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 52107 + }, + { + "epoch": 0.52108, + "grad_norm": 1.3691723890608145, + "learning_rate": 0.003, + "loss": 4.006, + "step": 52108 + }, + { + "epoch": 0.52109, + "grad_norm": 0.9285770935177609, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 52109 + }, + { + "epoch": 0.5211, + "grad_norm": 0.822361837584587, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 52110 + }, + { + "epoch": 0.52111, + "grad_norm": 0.9492731776405248, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 52111 + }, + { + "epoch": 0.52112, + "grad_norm": 1.2832965839866495, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 52112 + }, + { + "epoch": 0.52113, + "grad_norm": 0.9443866665880528, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 52113 + }, + { + "epoch": 0.52114, + "grad_norm": 1.2879201607961441, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 52114 + }, + { + "epoch": 0.52115, + "grad_norm": 1.2044243699105606, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 52115 + }, + { + "epoch": 0.52116, + "grad_norm": 0.7910372844558313, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 52116 + }, + { + "epoch": 0.52117, + "grad_norm": 0.8831753494002006, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 52117 + }, + { + "epoch": 0.52118, + "grad_norm": 0.9828800216739475, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 52118 + }, + { + "epoch": 0.52119, + "grad_norm": 1.046019969567199, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 52119 + }, + { + "epoch": 0.5212, + "grad_norm": 0.9931633473979419, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 52120 + }, + { + "epoch": 0.52121, + "grad_norm": 1.1687779115860788, + "learning_rate": 0.003, + "loss": 4.043, + "step": 52121 + }, + { + "epoch": 0.52122, + "grad_norm": 1.0788045938909565, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 52122 + }, + { + "epoch": 0.52123, + "grad_norm": 1.325748509754312, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 52123 + }, + { + "epoch": 0.52124, + "grad_norm": 0.8933183442418346, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 52124 + }, + { + "epoch": 0.52125, + "grad_norm": 0.9425269893097992, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 52125 + }, + { + "epoch": 0.52126, + "grad_norm": 1.1691097677785218, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 52126 + }, + { + "epoch": 0.52127, + "grad_norm": 1.0945436737622944, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 52127 + }, + { + "epoch": 0.52128, + "grad_norm": 1.1084657329003025, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 52128 + }, + { + "epoch": 0.52129, + "grad_norm": 1.1146828227750476, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 52129 + }, + { + "epoch": 0.5213, + "grad_norm": 1.1064836828004503, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 52130 + }, + { + "epoch": 0.52131, + "grad_norm": 1.2731599975703007, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 52131 + }, + { + "epoch": 0.52132, + "grad_norm": 0.9622571516876671, + "learning_rate": 0.003, + "loss": 4.04, + "step": 52132 + }, + { + "epoch": 0.52133, + "grad_norm": 1.0194248518871794, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 52133 + }, + { + "epoch": 0.52134, + "grad_norm": 1.2237089282764027, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 52134 + }, + { + "epoch": 0.52135, + "grad_norm": 1.028870304955406, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 52135 + }, + { + "epoch": 0.52136, + "grad_norm": 1.1747511652762888, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 52136 + }, + { + "epoch": 0.52137, + "grad_norm": 0.9692671254440108, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 52137 + }, + { + "epoch": 0.52138, + "grad_norm": 1.3010373171154064, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 52138 + }, + { + "epoch": 0.52139, + "grad_norm": 1.003455079145991, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 52139 + }, + { + "epoch": 0.5214, + "grad_norm": 1.3168762489876786, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 52140 + }, + { + "epoch": 0.52141, + "grad_norm": 1.0343303055012, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 52141 + }, + { + "epoch": 0.52142, + "grad_norm": 1.1302274411559483, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 52142 + }, + { + "epoch": 0.52143, + "grad_norm": 0.8890716618317072, + "learning_rate": 0.003, + "loss": 4.043, + "step": 52143 + }, + { + "epoch": 0.52144, + "grad_norm": 0.8349555140457668, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 52144 + }, + { + "epoch": 0.52145, + "grad_norm": 1.1191199453960816, + "learning_rate": 0.003, + "loss": 4.02, + "step": 52145 + }, + { + "epoch": 0.52146, + "grad_norm": 1.0930806572294887, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 52146 + }, + { + "epoch": 0.52147, + "grad_norm": 0.9402476993609543, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 52147 + }, + { + "epoch": 0.52148, + "grad_norm": 1.0575341272588825, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 52148 + }, + { + "epoch": 0.52149, + "grad_norm": 1.279114509309662, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 52149 + }, + { + "epoch": 0.5215, + "grad_norm": 0.9131336657221312, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 52150 + }, + { + "epoch": 0.52151, + "grad_norm": 0.9741563001654061, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 52151 + }, + { + "epoch": 0.52152, + "grad_norm": 1.2054865046667844, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 52152 + }, + { + "epoch": 0.52153, + "grad_norm": 0.9564212495011023, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 52153 + }, + { + "epoch": 0.52154, + "grad_norm": 1.0057502270467225, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 52154 + }, + { + "epoch": 0.52155, + "grad_norm": 1.184803685378276, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 52155 + }, + { + "epoch": 0.52156, + "grad_norm": 1.0494555188213206, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 52156 + }, + { + "epoch": 0.52157, + "grad_norm": 0.8719456363903458, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 52157 + }, + { + "epoch": 0.52158, + "grad_norm": 0.9834805314878122, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 52158 + }, + { + "epoch": 0.52159, + "grad_norm": 1.333469397089722, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 52159 + }, + { + "epoch": 0.5216, + "grad_norm": 0.8260636689084703, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 52160 + }, + { + "epoch": 0.52161, + "grad_norm": 0.7162960519319782, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 52161 + }, + { + "epoch": 0.52162, + "grad_norm": 0.8735900454199218, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 52162 + }, + { + "epoch": 0.52163, + "grad_norm": 1.1797109055205577, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 52163 + }, + { + "epoch": 0.52164, + "grad_norm": 0.9226365603509864, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 52164 + }, + { + "epoch": 0.52165, + "grad_norm": 1.2174960466062876, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 52165 + }, + { + "epoch": 0.52166, + "grad_norm": 1.0259461499255018, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 52166 + }, + { + "epoch": 0.52167, + "grad_norm": 0.8824812307094739, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 52167 + }, + { + "epoch": 0.52168, + "grad_norm": 1.1301226426293085, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 52168 + }, + { + "epoch": 0.52169, + "grad_norm": 1.3021884563356723, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 52169 + }, + { + "epoch": 0.5217, + "grad_norm": 1.1471146247366013, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 52170 + }, + { + "epoch": 0.52171, + "grad_norm": 0.9759166865801852, + "learning_rate": 0.003, + "loss": 4.045, + "step": 52171 + }, + { + "epoch": 0.52172, + "grad_norm": 0.9677145697074555, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 52172 + }, + { + "epoch": 0.52173, + "grad_norm": 1.205427501050405, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 52173 + }, + { + "epoch": 0.52174, + "grad_norm": 0.8734862778505417, + "learning_rate": 0.003, + "loss": 4.023, + "step": 52174 + }, + { + "epoch": 0.52175, + "grad_norm": 0.9669095099694065, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 52175 + }, + { + "epoch": 0.52176, + "grad_norm": 1.0738312541904655, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 52176 + }, + { + "epoch": 0.52177, + "grad_norm": 1.0309041973962503, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 52177 + }, + { + "epoch": 0.52178, + "grad_norm": 1.214198865282597, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 52178 + }, + { + "epoch": 0.52179, + "grad_norm": 1.0585698763304452, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 52179 + }, + { + "epoch": 0.5218, + "grad_norm": 1.2918093774564563, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 52180 + }, + { + "epoch": 0.52181, + "grad_norm": 0.8966879174531708, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 52181 + }, + { + "epoch": 0.52182, + "grad_norm": 0.9607143053455051, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 52182 + }, + { + "epoch": 0.52183, + "grad_norm": 1.0306274874753978, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 52183 + }, + { + "epoch": 0.52184, + "grad_norm": 1.209973757623207, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 52184 + }, + { + "epoch": 0.52185, + "grad_norm": 0.9496220488879981, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 52185 + }, + { + "epoch": 0.52186, + "grad_norm": 0.9825095871732268, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 52186 + }, + { + "epoch": 0.52187, + "grad_norm": 1.169623871793569, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 52187 + }, + { + "epoch": 0.52188, + "grad_norm": 0.9523731092985774, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 52188 + }, + { + "epoch": 0.52189, + "grad_norm": 1.073947252591763, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 52189 + }, + { + "epoch": 0.5219, + "grad_norm": 1.0783284142508494, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 52190 + }, + { + "epoch": 0.52191, + "grad_norm": 1.0777191256098488, + "learning_rate": 0.003, + "loss": 4.01, + "step": 52191 + }, + { + "epoch": 0.52192, + "grad_norm": 1.1658344351957617, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 52192 + }, + { + "epoch": 0.52193, + "grad_norm": 1.0127136148158673, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 52193 + }, + { + "epoch": 0.52194, + "grad_norm": 1.2666854922770234, + "learning_rate": 0.003, + "loss": 4.049, + "step": 52194 + }, + { + "epoch": 0.52195, + "grad_norm": 0.9881332566514824, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 52195 + }, + { + "epoch": 0.52196, + "grad_norm": 0.951716339551118, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 52196 + }, + { + "epoch": 0.52197, + "grad_norm": 1.1560706633659166, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 52197 + }, + { + "epoch": 0.52198, + "grad_norm": 1.189995242806972, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 52198 + }, + { + "epoch": 0.52199, + "grad_norm": 1.087813114618236, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 52199 + }, + { + "epoch": 0.522, + "grad_norm": 1.1550669160695748, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 52200 + }, + { + "epoch": 0.52201, + "grad_norm": 1.0225568153729, + "learning_rate": 0.003, + "loss": 4.025, + "step": 52201 + }, + { + "epoch": 0.52202, + "grad_norm": 1.2178063443255118, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 52202 + }, + { + "epoch": 0.52203, + "grad_norm": 1.1622605503882022, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 52203 + }, + { + "epoch": 0.52204, + "grad_norm": 0.9326223200043307, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 52204 + }, + { + "epoch": 0.52205, + "grad_norm": 0.8921122720905647, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 52205 + }, + { + "epoch": 0.52206, + "grad_norm": 0.8667398066236802, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 52206 + }, + { + "epoch": 0.52207, + "grad_norm": 1.0195822037615827, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 52207 + }, + { + "epoch": 0.52208, + "grad_norm": 1.146396734226052, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 52208 + }, + { + "epoch": 0.52209, + "grad_norm": 1.2283013556978897, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 52209 + }, + { + "epoch": 0.5221, + "grad_norm": 1.050524183498062, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 52210 + }, + { + "epoch": 0.52211, + "grad_norm": 1.113248199757799, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 52211 + }, + { + "epoch": 0.52212, + "grad_norm": 0.891156154528138, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 52212 + }, + { + "epoch": 0.52213, + "grad_norm": 0.9418688792950823, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 52213 + }, + { + "epoch": 0.52214, + "grad_norm": 1.2129715339413354, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 52214 + }, + { + "epoch": 0.52215, + "grad_norm": 0.9627775779913266, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 52215 + }, + { + "epoch": 0.52216, + "grad_norm": 1.0360758262524197, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 52216 + }, + { + "epoch": 0.52217, + "grad_norm": 1.0877427679298943, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 52217 + }, + { + "epoch": 0.52218, + "grad_norm": 0.9367432810179933, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 52218 + }, + { + "epoch": 0.52219, + "grad_norm": 0.9327018722311368, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 52219 + }, + { + "epoch": 0.5222, + "grad_norm": 1.0382032913108938, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 52220 + }, + { + "epoch": 0.52221, + "grad_norm": 1.125423340489113, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 52221 + }, + { + "epoch": 0.52222, + "grad_norm": 1.0539087588533698, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 52222 + }, + { + "epoch": 0.52223, + "grad_norm": 0.9765764421435473, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 52223 + }, + { + "epoch": 0.52224, + "grad_norm": 0.9259774595091818, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 52224 + }, + { + "epoch": 0.52225, + "grad_norm": 1.1281196746117172, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 52225 + }, + { + "epoch": 0.52226, + "grad_norm": 1.1192425032580113, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 52226 + }, + { + "epoch": 0.52227, + "grad_norm": 0.9230119769852513, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 52227 + }, + { + "epoch": 0.52228, + "grad_norm": 0.9855236648670591, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 52228 + }, + { + "epoch": 0.52229, + "grad_norm": 1.0809669255976522, + "learning_rate": 0.003, + "loss": 4.02, + "step": 52229 + }, + { + "epoch": 0.5223, + "grad_norm": 1.003387529135544, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 52230 + }, + { + "epoch": 0.52231, + "grad_norm": 1.1709370498974387, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 52231 + }, + { + "epoch": 0.52232, + "grad_norm": 1.0782283612151864, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 52232 + }, + { + "epoch": 0.52233, + "grad_norm": 1.2741135510671815, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 52233 + }, + { + "epoch": 0.52234, + "grad_norm": 1.2817967253976867, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 52234 + }, + { + "epoch": 0.52235, + "grad_norm": 0.8891702452719, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 52235 + }, + { + "epoch": 0.52236, + "grad_norm": 0.953365426632585, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 52236 + }, + { + "epoch": 0.52237, + "grad_norm": 1.2604833080108437, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 52237 + }, + { + "epoch": 0.52238, + "grad_norm": 1.0719181956944879, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 52238 + }, + { + "epoch": 0.52239, + "grad_norm": 0.9611024141979102, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 52239 + }, + { + "epoch": 0.5224, + "grad_norm": 1.0204118163652058, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 52240 + }, + { + "epoch": 0.52241, + "grad_norm": 1.3716441846128218, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 52241 + }, + { + "epoch": 0.52242, + "grad_norm": 0.9735369474745258, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 52242 + }, + { + "epoch": 0.52243, + "grad_norm": 1.1181870262047355, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 52243 + }, + { + "epoch": 0.52244, + "grad_norm": 0.9725456759998846, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 52244 + }, + { + "epoch": 0.52245, + "grad_norm": 1.0577999163685285, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 52245 + }, + { + "epoch": 0.52246, + "grad_norm": 1.2094581020288635, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 52246 + }, + { + "epoch": 0.52247, + "grad_norm": 1.1326136764765307, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 52247 + }, + { + "epoch": 0.52248, + "grad_norm": 1.2369080538142507, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 52248 + }, + { + "epoch": 0.52249, + "grad_norm": 0.8267965984099152, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 52249 + }, + { + "epoch": 0.5225, + "grad_norm": 0.7531630700978699, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 52250 + }, + { + "epoch": 0.52251, + "grad_norm": 0.8663148675130232, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 52251 + }, + { + "epoch": 0.52252, + "grad_norm": 1.0499753178874687, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 52252 + }, + { + "epoch": 0.52253, + "grad_norm": 1.0329078216505838, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 52253 + }, + { + "epoch": 0.52254, + "grad_norm": 1.1482247927581126, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 52254 + }, + { + "epoch": 0.52255, + "grad_norm": 0.9998151719208506, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 52255 + }, + { + "epoch": 0.52256, + "grad_norm": 0.9677883109260156, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 52256 + }, + { + "epoch": 0.52257, + "grad_norm": 1.047563698747162, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 52257 + }, + { + "epoch": 0.52258, + "grad_norm": 1.1698413310171027, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 52258 + }, + { + "epoch": 0.52259, + "grad_norm": 1.0041209079212836, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 52259 + }, + { + "epoch": 0.5226, + "grad_norm": 1.1179787136467925, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 52260 + }, + { + "epoch": 0.52261, + "grad_norm": 1.155262764331175, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 52261 + }, + { + "epoch": 0.52262, + "grad_norm": 1.1249193853645065, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 52262 + }, + { + "epoch": 0.52263, + "grad_norm": 1.191520284423632, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 52263 + }, + { + "epoch": 0.52264, + "grad_norm": 1.1039558704707306, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 52264 + }, + { + "epoch": 0.52265, + "grad_norm": 1.2686708715634143, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 52265 + }, + { + "epoch": 0.52266, + "grad_norm": 1.0511865210192528, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 52266 + }, + { + "epoch": 0.52267, + "grad_norm": 0.9901028641117217, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 52267 + }, + { + "epoch": 0.52268, + "grad_norm": 1.066527127857423, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 52268 + }, + { + "epoch": 0.52269, + "grad_norm": 1.3333285284042118, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 52269 + }, + { + "epoch": 0.5227, + "grad_norm": 0.8974556642934826, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 52270 + }, + { + "epoch": 0.52271, + "grad_norm": 0.9526263742578471, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 52271 + }, + { + "epoch": 0.52272, + "grad_norm": 0.9663597306564657, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 52272 + }, + { + "epoch": 0.52273, + "grad_norm": 1.1194191837344736, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 52273 + }, + { + "epoch": 0.52274, + "grad_norm": 1.0554735108967996, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 52274 + }, + { + "epoch": 0.52275, + "grad_norm": 1.1717015880680481, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 52275 + }, + { + "epoch": 0.52276, + "grad_norm": 1.0021819727960743, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 52276 + }, + { + "epoch": 0.52277, + "grad_norm": 1.0208498822384902, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 52277 + }, + { + "epoch": 0.52278, + "grad_norm": 1.0205870915504114, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 52278 + }, + { + "epoch": 0.52279, + "grad_norm": 1.0620879948721988, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 52279 + }, + { + "epoch": 0.5228, + "grad_norm": 0.9557075644289242, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 52280 + }, + { + "epoch": 0.52281, + "grad_norm": 0.8852989216716698, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 52281 + }, + { + "epoch": 0.52282, + "grad_norm": 0.9468705672417022, + "learning_rate": 0.003, + "loss": 4.019, + "step": 52282 + }, + { + "epoch": 0.52283, + "grad_norm": 1.0024290534460996, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 52283 + }, + { + "epoch": 0.52284, + "grad_norm": 1.2049913701324422, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 52284 + }, + { + "epoch": 0.52285, + "grad_norm": 1.3526630733948732, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 52285 + }, + { + "epoch": 0.52286, + "grad_norm": 1.064128087691948, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 52286 + }, + { + "epoch": 0.52287, + "grad_norm": 1.207291155449006, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 52287 + }, + { + "epoch": 0.52288, + "grad_norm": 0.7162439343216626, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 52288 + }, + { + "epoch": 0.52289, + "grad_norm": 0.6505916142649759, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 52289 + }, + { + "epoch": 0.5229, + "grad_norm": 0.6891995227603676, + "learning_rate": 0.003, + "loss": 4.02, + "step": 52290 + }, + { + "epoch": 0.52291, + "grad_norm": 0.8209445729752938, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 52291 + }, + { + "epoch": 0.52292, + "grad_norm": 0.9271706761924211, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 52292 + }, + { + "epoch": 0.52293, + "grad_norm": 1.044137650758453, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 52293 + }, + { + "epoch": 0.52294, + "grad_norm": 1.2036489605138143, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 52294 + }, + { + "epoch": 0.52295, + "grad_norm": 1.0522878651121204, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 52295 + }, + { + "epoch": 0.52296, + "grad_norm": 1.1419248006634777, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 52296 + }, + { + "epoch": 0.52297, + "grad_norm": 0.9919824834430824, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 52297 + }, + { + "epoch": 0.52298, + "grad_norm": 1.1128138255310154, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 52298 + }, + { + "epoch": 0.52299, + "grad_norm": 1.1004263528950746, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 52299 + }, + { + "epoch": 0.523, + "grad_norm": 1.0229604262458383, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 52300 + }, + { + "epoch": 0.52301, + "grad_norm": 1.1781762255054746, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 52301 + }, + { + "epoch": 0.52302, + "grad_norm": 0.9201338982164774, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 52302 + }, + { + "epoch": 0.52303, + "grad_norm": 0.9940258010672359, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 52303 + }, + { + "epoch": 0.52304, + "grad_norm": 1.0751640883231801, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 52304 + }, + { + "epoch": 0.52305, + "grad_norm": 1.0629167006194156, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 52305 + }, + { + "epoch": 0.52306, + "grad_norm": 1.1043234970412432, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 52306 + }, + { + "epoch": 0.52307, + "grad_norm": 1.0730546259962825, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 52307 + }, + { + "epoch": 0.52308, + "grad_norm": 1.1503313645364355, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 52308 + }, + { + "epoch": 0.52309, + "grad_norm": 1.095687126893752, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 52309 + }, + { + "epoch": 0.5231, + "grad_norm": 1.0027792585816326, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 52310 + }, + { + "epoch": 0.52311, + "grad_norm": 1.088204501742905, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 52311 + }, + { + "epoch": 0.52312, + "grad_norm": 1.0568049979992382, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 52312 + }, + { + "epoch": 0.52313, + "grad_norm": 1.15714587288564, + "learning_rate": 0.003, + "loss": 4.044, + "step": 52313 + }, + { + "epoch": 0.52314, + "grad_norm": 1.1323852371738163, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 52314 + }, + { + "epoch": 0.52315, + "grad_norm": 0.9490156686360213, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 52315 + }, + { + "epoch": 0.52316, + "grad_norm": 0.832356464639864, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 52316 + }, + { + "epoch": 0.52317, + "grad_norm": 0.9134694038843963, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 52317 + }, + { + "epoch": 0.52318, + "grad_norm": 0.9900768305824664, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 52318 + }, + { + "epoch": 0.52319, + "grad_norm": 1.1188118162383502, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 52319 + }, + { + "epoch": 0.5232, + "grad_norm": 1.0396513914066705, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 52320 + }, + { + "epoch": 0.52321, + "grad_norm": 1.033382546787875, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 52321 + }, + { + "epoch": 0.52322, + "grad_norm": 1.1349724417671387, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 52322 + }, + { + "epoch": 0.52323, + "grad_norm": 1.0279388022353346, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 52323 + }, + { + "epoch": 0.52324, + "grad_norm": 1.2844096912500877, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 52324 + }, + { + "epoch": 0.52325, + "grad_norm": 1.1350098617402855, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 52325 + }, + { + "epoch": 0.52326, + "grad_norm": 1.348240523708591, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 52326 + }, + { + "epoch": 0.52327, + "grad_norm": 1.0944169062142062, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 52327 + }, + { + "epoch": 0.52328, + "grad_norm": 1.0771328603841577, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 52328 + }, + { + "epoch": 0.52329, + "grad_norm": 1.0156204348755287, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 52329 + }, + { + "epoch": 0.5233, + "grad_norm": 1.1089225852831017, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 52330 + }, + { + "epoch": 0.52331, + "grad_norm": 1.0311165433127836, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 52331 + }, + { + "epoch": 0.52332, + "grad_norm": 1.0260112135025865, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 52332 + }, + { + "epoch": 0.52333, + "grad_norm": 0.9264097370961265, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 52333 + }, + { + "epoch": 0.52334, + "grad_norm": 0.9539457609319819, + "learning_rate": 0.003, + "loss": 3.983, + "step": 52334 + }, + { + "epoch": 0.52335, + "grad_norm": 0.9904458075932637, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 52335 + }, + { + "epoch": 0.52336, + "grad_norm": 0.9537969676057234, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 52336 + }, + { + "epoch": 0.52337, + "grad_norm": 1.0228530632479014, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 52337 + }, + { + "epoch": 0.52338, + "grad_norm": 1.0508258652756262, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 52338 + }, + { + "epoch": 0.52339, + "grad_norm": 1.0611825590262005, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 52339 + }, + { + "epoch": 0.5234, + "grad_norm": 1.0393915855491798, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 52340 + }, + { + "epoch": 0.52341, + "grad_norm": 1.390153515234499, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 52341 + }, + { + "epoch": 0.52342, + "grad_norm": 0.8177527018967949, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 52342 + }, + { + "epoch": 0.52343, + "grad_norm": 0.8265571705609343, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 52343 + }, + { + "epoch": 0.52344, + "grad_norm": 0.9445934345563592, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 52344 + }, + { + "epoch": 0.52345, + "grad_norm": 1.1661586278227933, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 52345 + }, + { + "epoch": 0.52346, + "grad_norm": 1.1211013721347378, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 52346 + }, + { + "epoch": 0.52347, + "grad_norm": 0.9797602743379211, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 52347 + }, + { + "epoch": 0.52348, + "grad_norm": 1.1306603577496692, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 52348 + }, + { + "epoch": 0.52349, + "grad_norm": 1.1860181099628688, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 52349 + }, + { + "epoch": 0.5235, + "grad_norm": 0.8922739191206474, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 52350 + }, + { + "epoch": 0.52351, + "grad_norm": 0.8653865920460869, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 52351 + }, + { + "epoch": 0.52352, + "grad_norm": 1.380697291115728, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 52352 + }, + { + "epoch": 0.52353, + "grad_norm": 0.9989380665347121, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 52353 + }, + { + "epoch": 0.52354, + "grad_norm": 0.8316625851602649, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 52354 + }, + { + "epoch": 0.52355, + "grad_norm": 0.7893362758679089, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 52355 + }, + { + "epoch": 0.52356, + "grad_norm": 0.7584663459141703, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 52356 + }, + { + "epoch": 0.52357, + "grad_norm": 0.9647938492863215, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 52357 + }, + { + "epoch": 0.52358, + "grad_norm": 1.2234745741816728, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 52358 + }, + { + "epoch": 0.52359, + "grad_norm": 0.787230137252084, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 52359 + }, + { + "epoch": 0.5236, + "grad_norm": 0.9478334355729843, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 52360 + }, + { + "epoch": 0.52361, + "grad_norm": 1.104728896908862, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 52361 + }, + { + "epoch": 0.52362, + "grad_norm": 1.1125665595074248, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 52362 + }, + { + "epoch": 0.52363, + "grad_norm": 1.3773329895837187, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 52363 + }, + { + "epoch": 0.52364, + "grad_norm": 0.8066245664924699, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 52364 + }, + { + "epoch": 0.52365, + "grad_norm": 0.9198856377466827, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 52365 + }, + { + "epoch": 0.52366, + "grad_norm": 1.084833051194992, + "learning_rate": 0.003, + "loss": 4.006, + "step": 52366 + }, + { + "epoch": 0.52367, + "grad_norm": 1.1891152743344444, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 52367 + }, + { + "epoch": 0.52368, + "grad_norm": 1.0282570646977138, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 52368 + }, + { + "epoch": 0.52369, + "grad_norm": 1.2438485200454423, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 52369 + }, + { + "epoch": 0.5237, + "grad_norm": 0.8581841588529537, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 52370 + }, + { + "epoch": 0.52371, + "grad_norm": 0.8645031452033938, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 52371 + }, + { + "epoch": 0.52372, + "grad_norm": 0.9209620012568581, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 52372 + }, + { + "epoch": 0.52373, + "grad_norm": 0.9921462261658698, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 52373 + }, + { + "epoch": 0.52374, + "grad_norm": 1.2077883119739679, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 52374 + }, + { + "epoch": 0.52375, + "grad_norm": 1.1593780470447521, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 52375 + }, + { + "epoch": 0.52376, + "grad_norm": 0.9957374709422974, + "learning_rate": 0.003, + "loss": 4.012, + "step": 52376 + }, + { + "epoch": 0.52377, + "grad_norm": 0.9156744068311098, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 52377 + }, + { + "epoch": 0.52378, + "grad_norm": 1.0178440522267296, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 52378 + }, + { + "epoch": 0.52379, + "grad_norm": 1.1788024575532976, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 52379 + }, + { + "epoch": 0.5238, + "grad_norm": 0.9580880318175817, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 52380 + }, + { + "epoch": 0.52381, + "grad_norm": 0.9634616814083555, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 52381 + }, + { + "epoch": 0.52382, + "grad_norm": 1.1384430568938022, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 52382 + }, + { + "epoch": 0.52383, + "grad_norm": 0.9499289279576667, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 52383 + }, + { + "epoch": 0.52384, + "grad_norm": 1.1785682141939489, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 52384 + }, + { + "epoch": 0.52385, + "grad_norm": 1.0057438013522968, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 52385 + }, + { + "epoch": 0.52386, + "grad_norm": 1.4957852189002199, + "learning_rate": 0.003, + "loss": 4.066, + "step": 52386 + }, + { + "epoch": 0.52387, + "grad_norm": 0.8570806432394967, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 52387 + }, + { + "epoch": 0.52388, + "grad_norm": 0.9206554981045971, + "learning_rate": 0.003, + "loss": 4.042, + "step": 52388 + }, + { + "epoch": 0.52389, + "grad_norm": 1.073600612561025, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 52389 + }, + { + "epoch": 0.5239, + "grad_norm": 1.097201907900249, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 52390 + }, + { + "epoch": 0.52391, + "grad_norm": 1.3100751941569213, + "learning_rate": 0.003, + "loss": 4.048, + "step": 52391 + }, + { + "epoch": 0.52392, + "grad_norm": 0.9082122962011044, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 52392 + }, + { + "epoch": 0.52393, + "grad_norm": 0.9545541893298833, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 52393 + }, + { + "epoch": 0.52394, + "grad_norm": 1.0912867477343935, + "learning_rate": 0.003, + "loss": 4.041, + "step": 52394 + }, + { + "epoch": 0.52395, + "grad_norm": 1.0852861401405924, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 52395 + }, + { + "epoch": 0.52396, + "grad_norm": 0.95517000833825, + "learning_rate": 0.003, + "loss": 4.055, + "step": 52396 + }, + { + "epoch": 0.52397, + "grad_norm": 0.9705888498504351, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 52397 + }, + { + "epoch": 0.52398, + "grad_norm": 1.1623415468372087, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 52398 + }, + { + "epoch": 0.52399, + "grad_norm": 0.9054383595311576, + "learning_rate": 0.003, + "loss": 4.044, + "step": 52399 + }, + { + "epoch": 0.524, + "grad_norm": 1.007789197182216, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 52400 + }, + { + "epoch": 0.52401, + "grad_norm": 0.9089618819041269, + "learning_rate": 0.003, + "loss": 3.999, + "step": 52401 + }, + { + "epoch": 0.52402, + "grad_norm": 1.005959714443638, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 52402 + }, + { + "epoch": 0.52403, + "grad_norm": 1.442842308986881, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 52403 + }, + { + "epoch": 0.52404, + "grad_norm": 0.7792218772643495, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 52404 + }, + { + "epoch": 0.52405, + "grad_norm": 0.7642098534803548, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 52405 + }, + { + "epoch": 0.52406, + "grad_norm": 0.804659900074884, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 52406 + }, + { + "epoch": 0.52407, + "grad_norm": 1.0612803176392331, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 52407 + }, + { + "epoch": 0.52408, + "grad_norm": 1.3344868333511064, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 52408 + }, + { + "epoch": 0.52409, + "grad_norm": 0.844589980355875, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 52409 + }, + { + "epoch": 0.5241, + "grad_norm": 0.9354094637573717, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 52410 + }, + { + "epoch": 0.52411, + "grad_norm": 1.2163718292939165, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 52411 + }, + { + "epoch": 0.52412, + "grad_norm": 1.2210862051167453, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 52412 + }, + { + "epoch": 0.52413, + "grad_norm": 1.1045660431651805, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 52413 + }, + { + "epoch": 0.52414, + "grad_norm": 1.1058945953916821, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 52414 + }, + { + "epoch": 0.52415, + "grad_norm": 0.8068202121729603, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 52415 + }, + { + "epoch": 0.52416, + "grad_norm": 0.9013814017003748, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 52416 + }, + { + "epoch": 0.52417, + "grad_norm": 1.0122274831528444, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 52417 + }, + { + "epoch": 0.52418, + "grad_norm": 1.041788849677362, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 52418 + }, + { + "epoch": 0.52419, + "grad_norm": 0.9968524564818757, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 52419 + }, + { + "epoch": 0.5242, + "grad_norm": 0.9679033175428607, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 52420 + }, + { + "epoch": 0.52421, + "grad_norm": 1.1415096087961432, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 52421 + }, + { + "epoch": 0.52422, + "grad_norm": 0.9221375162831938, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 52422 + }, + { + "epoch": 0.52423, + "grad_norm": 1.0019994941592467, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 52423 + }, + { + "epoch": 0.52424, + "grad_norm": 1.08912056056056, + "learning_rate": 0.003, + "loss": 4.055, + "step": 52424 + }, + { + "epoch": 0.52425, + "grad_norm": 1.0370051853778204, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 52425 + }, + { + "epoch": 0.52426, + "grad_norm": 1.2026628043898597, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 52426 + }, + { + "epoch": 0.52427, + "grad_norm": 1.1166380678585122, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 52427 + }, + { + "epoch": 0.52428, + "grad_norm": 1.1323493931984374, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 52428 + }, + { + "epoch": 0.52429, + "grad_norm": 1.145357515270356, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 52429 + }, + { + "epoch": 0.5243, + "grad_norm": 1.1097889777955434, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 52430 + }, + { + "epoch": 0.52431, + "grad_norm": 1.1078946087062993, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 52431 + }, + { + "epoch": 0.52432, + "grad_norm": 0.9767353968715876, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 52432 + }, + { + "epoch": 0.52433, + "grad_norm": 1.1108539287049477, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 52433 + }, + { + "epoch": 0.52434, + "grad_norm": 1.0474252172422347, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 52434 + }, + { + "epoch": 0.52435, + "grad_norm": 0.8918945217525651, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 52435 + }, + { + "epoch": 0.52436, + "grad_norm": 1.008415809250134, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 52436 + }, + { + "epoch": 0.52437, + "grad_norm": 1.1736444008868598, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 52437 + }, + { + "epoch": 0.52438, + "grad_norm": 0.8763340292347702, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 52438 + }, + { + "epoch": 0.52439, + "grad_norm": 0.9163819695234846, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 52439 + }, + { + "epoch": 0.5244, + "grad_norm": 1.0345615120539258, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 52440 + }, + { + "epoch": 0.52441, + "grad_norm": 1.1415918423434734, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 52441 + }, + { + "epoch": 0.52442, + "grad_norm": 1.0053699066270887, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 52442 + }, + { + "epoch": 0.52443, + "grad_norm": 1.0161393972362547, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 52443 + }, + { + "epoch": 0.52444, + "grad_norm": 0.8749374743871404, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 52444 + }, + { + "epoch": 0.52445, + "grad_norm": 0.8789631524379139, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 52445 + }, + { + "epoch": 0.52446, + "grad_norm": 0.9639377805386189, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 52446 + }, + { + "epoch": 0.52447, + "grad_norm": 0.9012401921274058, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 52447 + }, + { + "epoch": 0.52448, + "grad_norm": 0.9127924392815597, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 52448 + }, + { + "epoch": 0.52449, + "grad_norm": 0.8118896923461364, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 52449 + }, + { + "epoch": 0.5245, + "grad_norm": 0.9123215421528775, + "learning_rate": 0.003, + "loss": 4.058, + "step": 52450 + }, + { + "epoch": 0.52451, + "grad_norm": 1.0580929867780768, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 52451 + }, + { + "epoch": 0.52452, + "grad_norm": 1.220135986605782, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 52452 + }, + { + "epoch": 0.52453, + "grad_norm": 1.3185607705849283, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 52453 + }, + { + "epoch": 0.52454, + "grad_norm": 1.0800343024495265, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 52454 + }, + { + "epoch": 0.52455, + "grad_norm": 1.368034293931605, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 52455 + }, + { + "epoch": 0.52456, + "grad_norm": 0.846090205175908, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 52456 + }, + { + "epoch": 0.52457, + "grad_norm": 0.8025475269428833, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 52457 + }, + { + "epoch": 0.52458, + "grad_norm": 1.0001345543820046, + "learning_rate": 0.003, + "loss": 4.047, + "step": 52458 + }, + { + "epoch": 0.52459, + "grad_norm": 1.3334573332888204, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 52459 + }, + { + "epoch": 0.5246, + "grad_norm": 0.7949065194327032, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 52460 + }, + { + "epoch": 0.52461, + "grad_norm": 1.043867707642945, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 52461 + }, + { + "epoch": 0.52462, + "grad_norm": 1.471642676531113, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 52462 + }, + { + "epoch": 0.52463, + "grad_norm": 0.9594399209110362, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 52463 + }, + { + "epoch": 0.52464, + "grad_norm": 0.8836389336495679, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 52464 + }, + { + "epoch": 0.52465, + "grad_norm": 0.84058286047285, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 52465 + }, + { + "epoch": 0.52466, + "grad_norm": 0.8687393374378624, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 52466 + }, + { + "epoch": 0.52467, + "grad_norm": 1.173362927039854, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 52467 + }, + { + "epoch": 0.52468, + "grad_norm": 0.9380438457691752, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 52468 + }, + { + "epoch": 0.52469, + "grad_norm": 1.0417370199427145, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 52469 + }, + { + "epoch": 0.5247, + "grad_norm": 1.365854474338192, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 52470 + }, + { + "epoch": 0.52471, + "grad_norm": 0.7156534027180398, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 52471 + }, + { + "epoch": 0.52472, + "grad_norm": 0.7090980565365801, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 52472 + }, + { + "epoch": 0.52473, + "grad_norm": 0.7484274671895101, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 52473 + }, + { + "epoch": 0.52474, + "grad_norm": 0.7637589579206635, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 52474 + }, + { + "epoch": 0.52475, + "grad_norm": 0.8416778244303171, + "learning_rate": 0.003, + "loss": 3.999, + "step": 52475 + }, + { + "epoch": 0.52476, + "grad_norm": 1.0870696249060519, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 52476 + }, + { + "epoch": 0.52477, + "grad_norm": 1.2320071767518315, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 52477 + }, + { + "epoch": 0.52478, + "grad_norm": 0.9916686454937206, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 52478 + }, + { + "epoch": 0.52479, + "grad_norm": 1.26135519094472, + "learning_rate": 0.003, + "loss": 4.051, + "step": 52479 + }, + { + "epoch": 0.5248, + "grad_norm": 0.9297520851096857, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 52480 + }, + { + "epoch": 0.52481, + "grad_norm": 0.9004722748208805, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 52481 + }, + { + "epoch": 0.52482, + "grad_norm": 0.8733071189356224, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 52482 + }, + { + "epoch": 0.52483, + "grad_norm": 1.1307976439342375, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 52483 + }, + { + "epoch": 0.52484, + "grad_norm": 1.390731037830067, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 52484 + }, + { + "epoch": 0.52485, + "grad_norm": 0.873069195303397, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 52485 + }, + { + "epoch": 0.52486, + "grad_norm": 1.0307760146467866, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 52486 + }, + { + "epoch": 0.52487, + "grad_norm": 0.9732939288424978, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 52487 + }, + { + "epoch": 0.52488, + "grad_norm": 1.1329361507545082, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 52488 + }, + { + "epoch": 0.52489, + "grad_norm": 1.0966299139478437, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 52489 + }, + { + "epoch": 0.5249, + "grad_norm": 0.86758370994375, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 52490 + }, + { + "epoch": 0.52491, + "grad_norm": 1.1023559003457555, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 52491 + }, + { + "epoch": 0.52492, + "grad_norm": 1.308365183552296, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 52492 + }, + { + "epoch": 0.52493, + "grad_norm": 0.79036867683864, + "learning_rate": 0.003, + "loss": 4.03, + "step": 52493 + }, + { + "epoch": 0.52494, + "grad_norm": 0.6711861857453396, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 52494 + }, + { + "epoch": 0.52495, + "grad_norm": 0.8242515208013989, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 52495 + }, + { + "epoch": 0.52496, + "grad_norm": 0.926845118065039, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 52496 + }, + { + "epoch": 0.52497, + "grad_norm": 1.159562949406579, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 52497 + }, + { + "epoch": 0.52498, + "grad_norm": 1.080287170008281, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 52498 + }, + { + "epoch": 0.52499, + "grad_norm": 1.0904511296458923, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 52499 + }, + { + "epoch": 0.525, + "grad_norm": 0.9997616662938122, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 52500 + }, + { + "epoch": 0.52501, + "grad_norm": 1.2506498873484522, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 52501 + }, + { + "epoch": 0.52502, + "grad_norm": 1.0357567452060772, + "learning_rate": 0.003, + "loss": 4.0, + "step": 52502 + }, + { + "epoch": 0.52503, + "grad_norm": 1.5559749093267572, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 52503 + }, + { + "epoch": 0.52504, + "grad_norm": 0.9742070088475253, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 52504 + }, + { + "epoch": 0.52505, + "grad_norm": 1.0155707094658717, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 52505 + }, + { + "epoch": 0.52506, + "grad_norm": 1.2292683965810645, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 52506 + }, + { + "epoch": 0.52507, + "grad_norm": 1.1471209788348027, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 52507 + }, + { + "epoch": 0.52508, + "grad_norm": 1.0784335762397355, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 52508 + }, + { + "epoch": 0.52509, + "grad_norm": 1.0366417991898749, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 52509 + }, + { + "epoch": 0.5251, + "grad_norm": 1.1501801138159797, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 52510 + }, + { + "epoch": 0.52511, + "grad_norm": 1.142894943186683, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 52511 + }, + { + "epoch": 0.52512, + "grad_norm": 0.9783556423033312, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 52512 + }, + { + "epoch": 0.52513, + "grad_norm": 1.0222243400069484, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 52513 + }, + { + "epoch": 0.52514, + "grad_norm": 0.9736021648820722, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 52514 + }, + { + "epoch": 0.52515, + "grad_norm": 1.0487887091875752, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 52515 + }, + { + "epoch": 0.52516, + "grad_norm": 1.147032025786009, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 52516 + }, + { + "epoch": 0.52517, + "grad_norm": 1.1568111143656827, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 52517 + }, + { + "epoch": 0.52518, + "grad_norm": 1.237226133792044, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 52518 + }, + { + "epoch": 0.52519, + "grad_norm": 0.9175835857666784, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 52519 + }, + { + "epoch": 0.5252, + "grad_norm": 0.8538948620375869, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 52520 + }, + { + "epoch": 0.52521, + "grad_norm": 1.0657046204701772, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 52521 + }, + { + "epoch": 0.52522, + "grad_norm": 1.357039766242647, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 52522 + }, + { + "epoch": 0.52523, + "grad_norm": 0.8843459904990155, + "learning_rate": 0.003, + "loss": 4.033, + "step": 52523 + }, + { + "epoch": 0.52524, + "grad_norm": 1.1226068319202336, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 52524 + }, + { + "epoch": 0.52525, + "grad_norm": 1.2009595606352452, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 52525 + }, + { + "epoch": 0.52526, + "grad_norm": 1.0099879437774044, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 52526 + }, + { + "epoch": 0.52527, + "grad_norm": 1.0786237492560211, + "learning_rate": 0.003, + "loss": 4.022, + "step": 52527 + }, + { + "epoch": 0.52528, + "grad_norm": 0.9370817684322561, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 52528 + }, + { + "epoch": 0.52529, + "grad_norm": 0.9277816432829454, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 52529 + }, + { + "epoch": 0.5253, + "grad_norm": 0.9531538199152814, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 52530 + }, + { + "epoch": 0.52531, + "grad_norm": 1.055521703440711, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 52531 + }, + { + "epoch": 0.52532, + "grad_norm": 1.102332088445154, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 52532 + }, + { + "epoch": 0.52533, + "grad_norm": 1.1884750751216635, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 52533 + }, + { + "epoch": 0.52534, + "grad_norm": 0.894805288939292, + "learning_rate": 0.003, + "loss": 4.029, + "step": 52534 + }, + { + "epoch": 0.52535, + "grad_norm": 1.029963234666276, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 52535 + }, + { + "epoch": 0.52536, + "grad_norm": 1.134577395719947, + "learning_rate": 0.003, + "loss": 3.998, + "step": 52536 + }, + { + "epoch": 0.52537, + "grad_norm": 1.071853162635812, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 52537 + }, + { + "epoch": 0.52538, + "grad_norm": 1.0618297947797914, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 52538 + }, + { + "epoch": 0.52539, + "grad_norm": 1.0983218740063725, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 52539 + }, + { + "epoch": 0.5254, + "grad_norm": 1.0946855915520348, + "learning_rate": 0.003, + "loss": 4.011, + "step": 52540 + }, + { + "epoch": 0.52541, + "grad_norm": 0.9442814881538496, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 52541 + }, + { + "epoch": 0.52542, + "grad_norm": 0.8775790573184876, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 52542 + }, + { + "epoch": 0.52543, + "grad_norm": 0.9277784673463343, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 52543 + }, + { + "epoch": 0.52544, + "grad_norm": 1.1578278224182335, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 52544 + }, + { + "epoch": 0.52545, + "grad_norm": 1.0479431367695946, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 52545 + }, + { + "epoch": 0.52546, + "grad_norm": 1.0324263721171647, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 52546 + }, + { + "epoch": 0.52547, + "grad_norm": 1.194021592238172, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 52547 + }, + { + "epoch": 0.52548, + "grad_norm": 1.190174533314591, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 52548 + }, + { + "epoch": 0.52549, + "grad_norm": 0.9910346408803747, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 52549 + }, + { + "epoch": 0.5255, + "grad_norm": 1.4018611665554752, + "learning_rate": 0.003, + "loss": 4.035, + "step": 52550 + }, + { + "epoch": 0.52551, + "grad_norm": 1.0198270348293956, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 52551 + }, + { + "epoch": 0.52552, + "grad_norm": 1.1773443693460952, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 52552 + }, + { + "epoch": 0.52553, + "grad_norm": 1.0991038418943946, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 52553 + }, + { + "epoch": 0.52554, + "grad_norm": 1.1160335587069325, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 52554 + }, + { + "epoch": 0.52555, + "grad_norm": 1.0119938099334507, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 52555 + }, + { + "epoch": 0.52556, + "grad_norm": 1.0388261437278488, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 52556 + }, + { + "epoch": 0.52557, + "grad_norm": 1.1181877837871455, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 52557 + }, + { + "epoch": 0.52558, + "grad_norm": 0.9869883773127396, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 52558 + }, + { + "epoch": 0.52559, + "grad_norm": 1.273779559522332, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 52559 + }, + { + "epoch": 0.5256, + "grad_norm": 1.0252822494869755, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 52560 + }, + { + "epoch": 0.52561, + "grad_norm": 1.0884295418487908, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 52561 + }, + { + "epoch": 0.52562, + "grad_norm": 1.2373415292599879, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 52562 + }, + { + "epoch": 0.52563, + "grad_norm": 0.9992886476463909, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 52563 + }, + { + "epoch": 0.52564, + "grad_norm": 1.1574114431588143, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 52564 + }, + { + "epoch": 0.52565, + "grad_norm": 1.0634316052003414, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 52565 + }, + { + "epoch": 0.52566, + "grad_norm": 1.1428240450398517, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 52566 + }, + { + "epoch": 0.52567, + "grad_norm": 0.8929501430393797, + "learning_rate": 0.003, + "loss": 4.033, + "step": 52567 + }, + { + "epoch": 0.52568, + "grad_norm": 0.9045484791558852, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 52568 + }, + { + "epoch": 0.52569, + "grad_norm": 0.8923509981014536, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 52569 + }, + { + "epoch": 0.5257, + "grad_norm": 0.9485814366933596, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 52570 + }, + { + "epoch": 0.52571, + "grad_norm": 1.0030358208310461, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 52571 + }, + { + "epoch": 0.52572, + "grad_norm": 1.003678224563355, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 52572 + }, + { + "epoch": 0.52573, + "grad_norm": 1.2323369666855706, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 52573 + }, + { + "epoch": 0.52574, + "grad_norm": 0.9230068143581699, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 52574 + }, + { + "epoch": 0.52575, + "grad_norm": 1.092527272782081, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 52575 + }, + { + "epoch": 0.52576, + "grad_norm": 1.2165241897241952, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 52576 + }, + { + "epoch": 0.52577, + "grad_norm": 1.1832438093899924, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 52577 + }, + { + "epoch": 0.52578, + "grad_norm": 0.8868396021529031, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 52578 + }, + { + "epoch": 0.52579, + "grad_norm": 1.0090843755583345, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 52579 + }, + { + "epoch": 0.5258, + "grad_norm": 1.2149644819592764, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 52580 + }, + { + "epoch": 0.52581, + "grad_norm": 1.186416959187845, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 52581 + }, + { + "epoch": 0.52582, + "grad_norm": 1.0362757767112982, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 52582 + }, + { + "epoch": 0.52583, + "grad_norm": 1.014160079046229, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 52583 + }, + { + "epoch": 0.52584, + "grad_norm": 1.1640670945640377, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 52584 + }, + { + "epoch": 0.52585, + "grad_norm": 1.1170918737976478, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 52585 + }, + { + "epoch": 0.52586, + "grad_norm": 0.9977877077614411, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 52586 + }, + { + "epoch": 0.52587, + "grad_norm": 0.9191715364567755, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 52587 + }, + { + "epoch": 0.52588, + "grad_norm": 1.022754208344892, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 52588 + }, + { + "epoch": 0.52589, + "grad_norm": 1.059898472605055, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 52589 + }, + { + "epoch": 0.5259, + "grad_norm": 0.893726235913755, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 52590 + }, + { + "epoch": 0.52591, + "grad_norm": 1.0473134182826997, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 52591 + }, + { + "epoch": 0.52592, + "grad_norm": 1.151816450501264, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 52592 + }, + { + "epoch": 0.52593, + "grad_norm": 0.9239488365933693, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 52593 + }, + { + "epoch": 0.52594, + "grad_norm": 1.18557180121636, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 52594 + }, + { + "epoch": 0.52595, + "grad_norm": 1.1728504144814862, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 52595 + }, + { + "epoch": 0.52596, + "grad_norm": 1.0413496114013736, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 52596 + }, + { + "epoch": 0.52597, + "grad_norm": 1.2001662742436812, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 52597 + }, + { + "epoch": 0.52598, + "grad_norm": 0.8348086822451128, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 52598 + }, + { + "epoch": 0.52599, + "grad_norm": 0.9021291681886907, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 52599 + }, + { + "epoch": 0.526, + "grad_norm": 1.3127513413274075, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 52600 + }, + { + "epoch": 0.52601, + "grad_norm": 0.8753245961050167, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 52601 + }, + { + "epoch": 0.52602, + "grad_norm": 1.0605712373923368, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 52602 + }, + { + "epoch": 0.52603, + "grad_norm": 1.2497277737268788, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 52603 + }, + { + "epoch": 0.52604, + "grad_norm": 0.833672829517651, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 52604 + }, + { + "epoch": 0.52605, + "grad_norm": 0.7288505901421469, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 52605 + }, + { + "epoch": 0.52606, + "grad_norm": 0.9338888126272255, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 52606 + }, + { + "epoch": 0.52607, + "grad_norm": 1.569682448957753, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 52607 + }, + { + "epoch": 0.52608, + "grad_norm": 0.7019951630525988, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 52608 + }, + { + "epoch": 0.52609, + "grad_norm": 0.7968666048452134, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 52609 + }, + { + "epoch": 0.5261, + "grad_norm": 0.9463283713941105, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 52610 + }, + { + "epoch": 0.52611, + "grad_norm": 1.0613056120664812, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 52611 + }, + { + "epoch": 0.52612, + "grad_norm": 1.0327773580940913, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 52612 + }, + { + "epoch": 0.52613, + "grad_norm": 1.075914144595017, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 52613 + }, + { + "epoch": 0.52614, + "grad_norm": 1.0138166373276751, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 52614 + }, + { + "epoch": 0.52615, + "grad_norm": 1.2896839993391356, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 52615 + }, + { + "epoch": 0.52616, + "grad_norm": 1.143746291221174, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 52616 + }, + { + "epoch": 0.52617, + "grad_norm": 0.9636797417273418, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 52617 + }, + { + "epoch": 0.52618, + "grad_norm": 1.1236173243706047, + "learning_rate": 0.003, + "loss": 3.996, + "step": 52618 + }, + { + "epoch": 0.52619, + "grad_norm": 0.982556291205582, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 52619 + }, + { + "epoch": 0.5262, + "grad_norm": 1.0662783681987693, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 52620 + }, + { + "epoch": 0.52621, + "grad_norm": 0.9479143989474866, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 52621 + }, + { + "epoch": 0.52622, + "grad_norm": 0.9024219910159066, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 52622 + }, + { + "epoch": 0.52623, + "grad_norm": 0.9925651039031086, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 52623 + }, + { + "epoch": 0.52624, + "grad_norm": 1.055373296169983, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 52624 + }, + { + "epoch": 0.52625, + "grad_norm": 1.0239044479182586, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 52625 + }, + { + "epoch": 0.52626, + "grad_norm": 1.0212040650288763, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 52626 + }, + { + "epoch": 0.52627, + "grad_norm": 1.187452111510696, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 52627 + }, + { + "epoch": 0.52628, + "grad_norm": 1.1574038373836912, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 52628 + }, + { + "epoch": 0.52629, + "grad_norm": 1.1625191882542822, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 52629 + }, + { + "epoch": 0.5263, + "grad_norm": 1.112037666803513, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 52630 + }, + { + "epoch": 0.52631, + "grad_norm": 1.0240273880976778, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 52631 + }, + { + "epoch": 0.52632, + "grad_norm": 0.9090334504580065, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 52632 + }, + { + "epoch": 0.52633, + "grad_norm": 1.118568425097285, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 52633 + }, + { + "epoch": 0.52634, + "grad_norm": 1.0885502202677457, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 52634 + }, + { + "epoch": 0.52635, + "grad_norm": 1.0359079963337734, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 52635 + }, + { + "epoch": 0.52636, + "grad_norm": 1.1618741337677865, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 52636 + }, + { + "epoch": 0.52637, + "grad_norm": 1.0959006286419688, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 52637 + }, + { + "epoch": 0.52638, + "grad_norm": 1.094608725102716, + "learning_rate": 0.003, + "loss": 4.025, + "step": 52638 + }, + { + "epoch": 0.52639, + "grad_norm": 1.0992087673162017, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 52639 + }, + { + "epoch": 0.5264, + "grad_norm": 1.0058909814268504, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 52640 + }, + { + "epoch": 0.52641, + "grad_norm": 1.1220782840190195, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 52641 + }, + { + "epoch": 0.52642, + "grad_norm": 0.9808354644622744, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 52642 + }, + { + "epoch": 0.52643, + "grad_norm": 1.1636793977997184, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 52643 + }, + { + "epoch": 0.52644, + "grad_norm": 0.8752294908120767, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 52644 + }, + { + "epoch": 0.52645, + "grad_norm": 0.9574367044754705, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 52645 + }, + { + "epoch": 0.52646, + "grad_norm": 1.0827236644909026, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 52646 + }, + { + "epoch": 0.52647, + "grad_norm": 1.1567360504937332, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 52647 + }, + { + "epoch": 0.52648, + "grad_norm": 0.9838155957515304, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 52648 + }, + { + "epoch": 0.52649, + "grad_norm": 0.8415753940609916, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 52649 + }, + { + "epoch": 0.5265, + "grad_norm": 0.9139430937466405, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 52650 + }, + { + "epoch": 0.52651, + "grad_norm": 1.067564747911286, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 52651 + }, + { + "epoch": 0.52652, + "grad_norm": 1.1752148433209069, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 52652 + }, + { + "epoch": 0.52653, + "grad_norm": 1.0417603521951324, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 52653 + }, + { + "epoch": 0.52654, + "grad_norm": 1.1715743656344364, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 52654 + }, + { + "epoch": 0.52655, + "grad_norm": 0.851602264683132, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 52655 + }, + { + "epoch": 0.52656, + "grad_norm": 0.9111106188185474, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 52656 + }, + { + "epoch": 0.52657, + "grad_norm": 1.1944839570702068, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 52657 + }, + { + "epoch": 0.52658, + "grad_norm": 1.3009729832709525, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 52658 + }, + { + "epoch": 0.52659, + "grad_norm": 0.7844340893132276, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 52659 + }, + { + "epoch": 0.5266, + "grad_norm": 0.7313279304355884, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 52660 + }, + { + "epoch": 0.52661, + "grad_norm": 0.9099285122696502, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 52661 + }, + { + "epoch": 0.52662, + "grad_norm": 1.2948939493382108, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 52662 + }, + { + "epoch": 0.52663, + "grad_norm": 1.0918644447761214, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 52663 + }, + { + "epoch": 0.52664, + "grad_norm": 1.0654144679840347, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 52664 + }, + { + "epoch": 0.52665, + "grad_norm": 1.183352614296464, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 52665 + }, + { + "epoch": 0.52666, + "grad_norm": 0.9902140214979002, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 52666 + }, + { + "epoch": 0.52667, + "grad_norm": 1.073712145782955, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 52667 + }, + { + "epoch": 0.52668, + "grad_norm": 1.1946050566509883, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 52668 + }, + { + "epoch": 0.52669, + "grad_norm": 0.8414295090756921, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 52669 + }, + { + "epoch": 0.5267, + "grad_norm": 0.8645779568939851, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 52670 + }, + { + "epoch": 0.52671, + "grad_norm": 0.9609039895874248, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 52671 + }, + { + "epoch": 0.52672, + "grad_norm": 0.9879966312549758, + "learning_rate": 0.003, + "loss": 4.043, + "step": 52672 + }, + { + "epoch": 0.52673, + "grad_norm": 1.013009233225428, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 52673 + }, + { + "epoch": 0.52674, + "grad_norm": 1.1980523767360483, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 52674 + }, + { + "epoch": 0.52675, + "grad_norm": 1.0139949836192166, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 52675 + }, + { + "epoch": 0.52676, + "grad_norm": 1.2243047394422386, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 52676 + }, + { + "epoch": 0.52677, + "grad_norm": 0.9811762506666396, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 52677 + }, + { + "epoch": 0.52678, + "grad_norm": 1.1751209213512974, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 52678 + }, + { + "epoch": 0.52679, + "grad_norm": 0.9949343273897733, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 52679 + }, + { + "epoch": 0.5268, + "grad_norm": 1.1873419517277892, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 52680 + }, + { + "epoch": 0.52681, + "grad_norm": 1.1519808298787961, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 52681 + }, + { + "epoch": 0.52682, + "grad_norm": 1.043597618183643, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 52682 + }, + { + "epoch": 0.52683, + "grad_norm": 1.1731464450799736, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 52683 + }, + { + "epoch": 0.52684, + "grad_norm": 1.1856832655228093, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 52684 + }, + { + "epoch": 0.52685, + "grad_norm": 1.1240632080450346, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 52685 + }, + { + "epoch": 0.52686, + "grad_norm": 0.8926824493368849, + "learning_rate": 0.003, + "loss": 4.018, + "step": 52686 + }, + { + "epoch": 0.52687, + "grad_norm": 1.1124185970548086, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 52687 + }, + { + "epoch": 0.52688, + "grad_norm": 1.2479829996132603, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 52688 + }, + { + "epoch": 0.52689, + "grad_norm": 0.8744604150374252, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 52689 + }, + { + "epoch": 0.5269, + "grad_norm": 0.8544189374980988, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 52690 + }, + { + "epoch": 0.52691, + "grad_norm": 0.9311404295884712, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 52691 + }, + { + "epoch": 0.52692, + "grad_norm": 1.2616337571749414, + "learning_rate": 0.003, + "loss": 4.051, + "step": 52692 + }, + { + "epoch": 0.52693, + "grad_norm": 1.1596613712540493, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 52693 + }, + { + "epoch": 0.52694, + "grad_norm": 1.1253388528179382, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 52694 + }, + { + "epoch": 0.52695, + "grad_norm": 0.9385055785656922, + "learning_rate": 0.003, + "loss": 4.013, + "step": 52695 + }, + { + "epoch": 0.52696, + "grad_norm": 0.9762376684430527, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 52696 + }, + { + "epoch": 0.52697, + "grad_norm": 1.0591572519573595, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 52697 + }, + { + "epoch": 0.52698, + "grad_norm": 0.9178469256511481, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 52698 + }, + { + "epoch": 0.52699, + "grad_norm": 0.9754700603266812, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 52699 + }, + { + "epoch": 0.527, + "grad_norm": 1.1642072662538294, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 52700 + }, + { + "epoch": 0.52701, + "grad_norm": 0.875141865782222, + "learning_rate": 0.003, + "loss": 4.017, + "step": 52701 + }, + { + "epoch": 0.52702, + "grad_norm": 0.8236803896310309, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 52702 + }, + { + "epoch": 0.52703, + "grad_norm": 0.8032719353298884, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 52703 + }, + { + "epoch": 0.52704, + "grad_norm": 0.9244418302626874, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 52704 + }, + { + "epoch": 0.52705, + "grad_norm": 1.1196828747084873, + "learning_rate": 0.003, + "loss": 4.049, + "step": 52705 + }, + { + "epoch": 0.52706, + "grad_norm": 0.9284647618007597, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 52706 + }, + { + "epoch": 0.52707, + "grad_norm": 0.9445372707926716, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 52707 + }, + { + "epoch": 0.52708, + "grad_norm": 1.1391539362343353, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 52708 + }, + { + "epoch": 0.52709, + "grad_norm": 1.0329725759647188, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 52709 + }, + { + "epoch": 0.5271, + "grad_norm": 0.9666583700367839, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 52710 + }, + { + "epoch": 0.52711, + "grad_norm": 1.1737009304738477, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 52711 + }, + { + "epoch": 0.52712, + "grad_norm": 1.0447329462875987, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 52712 + }, + { + "epoch": 0.52713, + "grad_norm": 1.2175412514051474, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 52713 + }, + { + "epoch": 0.52714, + "grad_norm": 1.0099477786997095, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 52714 + }, + { + "epoch": 0.52715, + "grad_norm": 1.1491632139127397, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 52715 + }, + { + "epoch": 0.52716, + "grad_norm": 1.2023982457080384, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 52716 + }, + { + "epoch": 0.52717, + "grad_norm": 1.195884483483095, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 52717 + }, + { + "epoch": 0.52718, + "grad_norm": 0.9069224602029137, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 52718 + }, + { + "epoch": 0.52719, + "grad_norm": 0.8674205986447421, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 52719 + }, + { + "epoch": 0.5272, + "grad_norm": 0.9925208760010906, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 52720 + }, + { + "epoch": 0.52721, + "grad_norm": 1.2643395473969947, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 52721 + }, + { + "epoch": 0.52722, + "grad_norm": 0.8211627984087913, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 52722 + }, + { + "epoch": 0.52723, + "grad_norm": 0.7831727730350965, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 52723 + }, + { + "epoch": 0.52724, + "grad_norm": 0.7790791629062757, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 52724 + }, + { + "epoch": 0.52725, + "grad_norm": 1.0371745123586864, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 52725 + }, + { + "epoch": 0.52726, + "grad_norm": 1.3060180537819575, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 52726 + }, + { + "epoch": 0.52727, + "grad_norm": 0.7953480446181512, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 52727 + }, + { + "epoch": 0.52728, + "grad_norm": 1.0271735763586758, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 52728 + }, + { + "epoch": 0.52729, + "grad_norm": 1.2394062972361388, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 52729 + }, + { + "epoch": 0.5273, + "grad_norm": 1.105872718480929, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 52730 + }, + { + "epoch": 0.52731, + "grad_norm": 1.0103729544326556, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 52731 + }, + { + "epoch": 0.52732, + "grad_norm": 1.1745105881129656, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 52732 + }, + { + "epoch": 0.52733, + "grad_norm": 1.1213502656293777, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 52733 + }, + { + "epoch": 0.52734, + "grad_norm": 0.8798614718594796, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 52734 + }, + { + "epoch": 0.52735, + "grad_norm": 0.9286527179248392, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 52735 + }, + { + "epoch": 0.52736, + "grad_norm": 0.905498391716495, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 52736 + }, + { + "epoch": 0.52737, + "grad_norm": 0.9238120343937318, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 52737 + }, + { + "epoch": 0.52738, + "grad_norm": 1.0126848295834563, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 52738 + }, + { + "epoch": 0.52739, + "grad_norm": 1.1348854257074317, + "learning_rate": 0.003, + "loss": 4.028, + "step": 52739 + }, + { + "epoch": 0.5274, + "grad_norm": 0.830368083493417, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 52740 + }, + { + "epoch": 0.52741, + "grad_norm": 0.8629917716089989, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 52741 + }, + { + "epoch": 0.52742, + "grad_norm": 1.0906642070465542, + "learning_rate": 0.003, + "loss": 4.032, + "step": 52742 + }, + { + "epoch": 0.52743, + "grad_norm": 1.314737949818752, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 52743 + }, + { + "epoch": 0.52744, + "grad_norm": 1.0780122620000545, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 52744 + }, + { + "epoch": 0.52745, + "grad_norm": 1.0834359230992328, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 52745 + }, + { + "epoch": 0.52746, + "grad_norm": 1.055412241057581, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 52746 + }, + { + "epoch": 0.52747, + "grad_norm": 1.2109196511338773, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 52747 + }, + { + "epoch": 0.52748, + "grad_norm": 0.9049913081337502, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 52748 + }, + { + "epoch": 0.52749, + "grad_norm": 0.9471816809154874, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 52749 + }, + { + "epoch": 0.5275, + "grad_norm": 1.123387263522978, + "learning_rate": 0.003, + "loss": 4.01, + "step": 52750 + }, + { + "epoch": 0.52751, + "grad_norm": 1.1404282502604104, + "learning_rate": 0.003, + "loss": 4.063, + "step": 52751 + }, + { + "epoch": 0.52752, + "grad_norm": 1.0500217793049371, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 52752 + }, + { + "epoch": 0.52753, + "grad_norm": 1.0552934314888955, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 52753 + }, + { + "epoch": 0.52754, + "grad_norm": 0.7685916684305916, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 52754 + }, + { + "epoch": 0.52755, + "grad_norm": 0.7290631342011514, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 52755 + }, + { + "epoch": 0.52756, + "grad_norm": 0.8262324798477416, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 52756 + }, + { + "epoch": 0.52757, + "grad_norm": 1.1363104643302981, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 52757 + }, + { + "epoch": 0.52758, + "grad_norm": 1.1549678878089893, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 52758 + }, + { + "epoch": 0.52759, + "grad_norm": 0.9874642559994419, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 52759 + }, + { + "epoch": 0.5276, + "grad_norm": 1.0911743003385368, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 52760 + }, + { + "epoch": 0.52761, + "grad_norm": 1.0176595813482578, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 52761 + }, + { + "epoch": 0.52762, + "grad_norm": 1.5041078136602894, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 52762 + }, + { + "epoch": 0.52763, + "grad_norm": 0.7718001905637569, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 52763 + }, + { + "epoch": 0.52764, + "grad_norm": 0.9216097847745474, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 52764 + }, + { + "epoch": 0.52765, + "grad_norm": 1.141835146630671, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 52765 + }, + { + "epoch": 0.52766, + "grad_norm": 1.090129550121002, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 52766 + }, + { + "epoch": 0.52767, + "grad_norm": 1.2516526344135126, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 52767 + }, + { + "epoch": 0.52768, + "grad_norm": 1.09116645613223, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 52768 + }, + { + "epoch": 0.52769, + "grad_norm": 1.288138407484221, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 52769 + }, + { + "epoch": 0.5277, + "grad_norm": 1.0741137948079484, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 52770 + }, + { + "epoch": 0.52771, + "grad_norm": 1.0789252646972012, + "learning_rate": 0.003, + "loss": 4.054, + "step": 52771 + }, + { + "epoch": 0.52772, + "grad_norm": 1.0150791899777143, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 52772 + }, + { + "epoch": 0.52773, + "grad_norm": 1.0416722747232052, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 52773 + }, + { + "epoch": 0.52774, + "grad_norm": 1.0916022104041714, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 52774 + }, + { + "epoch": 0.52775, + "grad_norm": 1.0710119843711097, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 52775 + }, + { + "epoch": 0.52776, + "grad_norm": 1.0311000697712862, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 52776 + }, + { + "epoch": 0.52777, + "grad_norm": 1.1565767397599493, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 52777 + }, + { + "epoch": 0.52778, + "grad_norm": 1.0283154553298446, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 52778 + }, + { + "epoch": 0.52779, + "grad_norm": 1.2063700096130372, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 52779 + }, + { + "epoch": 0.5278, + "grad_norm": 0.9979291320965972, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 52780 + }, + { + "epoch": 0.52781, + "grad_norm": 1.2710096864556606, + "learning_rate": 0.003, + "loss": 4.032, + "step": 52781 + }, + { + "epoch": 0.52782, + "grad_norm": 1.0089705687339787, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 52782 + }, + { + "epoch": 0.52783, + "grad_norm": 1.2245580457117695, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 52783 + }, + { + "epoch": 0.52784, + "grad_norm": 0.8488585059517069, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 52784 + }, + { + "epoch": 0.52785, + "grad_norm": 0.8507725070360592, + "learning_rate": 0.003, + "loss": 4.027, + "step": 52785 + }, + { + "epoch": 0.52786, + "grad_norm": 0.9010404987411038, + "learning_rate": 0.003, + "loss": 4.046, + "step": 52786 + }, + { + "epoch": 0.52787, + "grad_norm": 1.0556738245555832, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 52787 + }, + { + "epoch": 0.52788, + "grad_norm": 1.0536965563462428, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 52788 + }, + { + "epoch": 0.52789, + "grad_norm": 1.0199377944370454, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 52789 + }, + { + "epoch": 0.5279, + "grad_norm": 1.2729830372891011, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 52790 + }, + { + "epoch": 0.52791, + "grad_norm": 0.819076350337126, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 52791 + }, + { + "epoch": 0.52792, + "grad_norm": 0.797654096162618, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 52792 + }, + { + "epoch": 0.52793, + "grad_norm": 1.1708980998436438, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 52793 + }, + { + "epoch": 0.52794, + "grad_norm": 1.333093015265016, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 52794 + }, + { + "epoch": 0.52795, + "grad_norm": 0.9091871907415421, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 52795 + }, + { + "epoch": 0.52796, + "grad_norm": 0.955083840683352, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 52796 + }, + { + "epoch": 0.52797, + "grad_norm": 1.2545288255865097, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 52797 + }, + { + "epoch": 0.52798, + "grad_norm": 0.88817242965733, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 52798 + }, + { + "epoch": 0.52799, + "grad_norm": 0.8587986949148896, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 52799 + }, + { + "epoch": 0.528, + "grad_norm": 0.9988781315842002, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 52800 + }, + { + "epoch": 0.52801, + "grad_norm": 1.1849383444943549, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 52801 + }, + { + "epoch": 0.52802, + "grad_norm": 1.1795024080922654, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 52802 + }, + { + "epoch": 0.52803, + "grad_norm": 1.2768115062972047, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 52803 + }, + { + "epoch": 0.52804, + "grad_norm": 1.002508596947601, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 52804 + }, + { + "epoch": 0.52805, + "grad_norm": 1.1804328107128585, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 52805 + }, + { + "epoch": 0.52806, + "grad_norm": 1.041759368804115, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 52806 + }, + { + "epoch": 0.52807, + "grad_norm": 1.1648383237061786, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 52807 + }, + { + "epoch": 0.52808, + "grad_norm": 1.0616047504162287, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 52808 + }, + { + "epoch": 0.52809, + "grad_norm": 0.8889288628890779, + "learning_rate": 0.003, + "loss": 4.018, + "step": 52809 + }, + { + "epoch": 0.5281, + "grad_norm": 0.8428362719200096, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 52810 + }, + { + "epoch": 0.52811, + "grad_norm": 0.9396145720099419, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 52811 + }, + { + "epoch": 0.52812, + "grad_norm": 1.1268701876447407, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 52812 + }, + { + "epoch": 0.52813, + "grad_norm": 1.0123407464584464, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 52813 + }, + { + "epoch": 0.52814, + "grad_norm": 1.3782712686156642, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 52814 + }, + { + "epoch": 0.52815, + "grad_norm": 0.9321890172233657, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 52815 + }, + { + "epoch": 0.52816, + "grad_norm": 1.0984675569652431, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 52816 + }, + { + "epoch": 0.52817, + "grad_norm": 1.034302154700964, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 52817 + }, + { + "epoch": 0.52818, + "grad_norm": 1.1924149589273556, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 52818 + }, + { + "epoch": 0.52819, + "grad_norm": 1.1286983302407578, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 52819 + }, + { + "epoch": 0.5282, + "grad_norm": 1.0123216562899708, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 52820 + }, + { + "epoch": 0.52821, + "grad_norm": 1.0035888955406664, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 52821 + }, + { + "epoch": 0.52822, + "grad_norm": 1.1527962867409338, + "learning_rate": 0.003, + "loss": 4.045, + "step": 52822 + }, + { + "epoch": 0.52823, + "grad_norm": 0.9768850560494474, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 52823 + }, + { + "epoch": 0.52824, + "grad_norm": 1.2608811181379826, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 52824 + }, + { + "epoch": 0.52825, + "grad_norm": 0.8391671071184037, + "learning_rate": 0.003, + "loss": 4.038, + "step": 52825 + }, + { + "epoch": 0.52826, + "grad_norm": 0.8200376520766215, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 52826 + }, + { + "epoch": 0.52827, + "grad_norm": 1.024896106805008, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 52827 + }, + { + "epoch": 0.52828, + "grad_norm": 1.0773580951091697, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 52828 + }, + { + "epoch": 0.52829, + "grad_norm": 0.870568495155201, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 52829 + }, + { + "epoch": 0.5283, + "grad_norm": 1.019503298532924, + "learning_rate": 0.003, + "loss": 4.041, + "step": 52830 + }, + { + "epoch": 0.52831, + "grad_norm": 1.0004345896084874, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 52831 + }, + { + "epoch": 0.52832, + "grad_norm": 1.2723323312052066, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 52832 + }, + { + "epoch": 0.52833, + "grad_norm": 1.0166717337302666, + "learning_rate": 0.003, + "loss": 4.023, + "step": 52833 + }, + { + "epoch": 0.52834, + "grad_norm": 1.3090738590495992, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 52834 + }, + { + "epoch": 0.52835, + "grad_norm": 0.752452017349035, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 52835 + }, + { + "epoch": 0.52836, + "grad_norm": 0.6933689951956731, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 52836 + }, + { + "epoch": 0.52837, + "grad_norm": 0.8554075168512548, + "learning_rate": 0.003, + "loss": 4.023, + "step": 52837 + }, + { + "epoch": 0.52838, + "grad_norm": 1.2566196488435062, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 52838 + }, + { + "epoch": 0.52839, + "grad_norm": 1.1666248582404963, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 52839 + }, + { + "epoch": 0.5284, + "grad_norm": 0.7383207747107001, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 52840 + }, + { + "epoch": 0.52841, + "grad_norm": 0.8490970454352476, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 52841 + }, + { + "epoch": 0.52842, + "grad_norm": 1.1037789915568506, + "learning_rate": 0.003, + "loss": 4.042, + "step": 52842 + }, + { + "epoch": 0.52843, + "grad_norm": 0.8834868420456302, + "learning_rate": 0.003, + "loss": 4.032, + "step": 52843 + }, + { + "epoch": 0.52844, + "grad_norm": 0.9977643754995473, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 52844 + }, + { + "epoch": 0.52845, + "grad_norm": 1.2116458673389643, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 52845 + }, + { + "epoch": 0.52846, + "grad_norm": 0.9264384518468258, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 52846 + }, + { + "epoch": 0.52847, + "grad_norm": 1.1398685494368646, + "learning_rate": 0.003, + "loss": 4.036, + "step": 52847 + }, + { + "epoch": 0.52848, + "grad_norm": 1.3227826046301494, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 52848 + }, + { + "epoch": 0.52849, + "grad_norm": 0.9107803287810127, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 52849 + }, + { + "epoch": 0.5285, + "grad_norm": 0.9056382348282175, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 52850 + }, + { + "epoch": 0.52851, + "grad_norm": 0.9923772684038795, + "learning_rate": 0.003, + "loss": 4.027, + "step": 52851 + }, + { + "epoch": 0.52852, + "grad_norm": 0.9960700347397472, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 52852 + }, + { + "epoch": 0.52853, + "grad_norm": 1.065531285966168, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 52853 + }, + { + "epoch": 0.52854, + "grad_norm": 0.9517459551311022, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 52854 + }, + { + "epoch": 0.52855, + "grad_norm": 1.0808405713384073, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 52855 + }, + { + "epoch": 0.52856, + "grad_norm": 1.0841529784554282, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 52856 + }, + { + "epoch": 0.52857, + "grad_norm": 1.0077103223158803, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 52857 + }, + { + "epoch": 0.52858, + "grad_norm": 1.1025436246653337, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 52858 + }, + { + "epoch": 0.52859, + "grad_norm": 1.2629587431479856, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 52859 + }, + { + "epoch": 0.5286, + "grad_norm": 1.164543708031367, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 52860 + }, + { + "epoch": 0.52861, + "grad_norm": 1.2599560249837307, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 52861 + }, + { + "epoch": 0.52862, + "grad_norm": 0.7881595824864257, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 52862 + }, + { + "epoch": 0.52863, + "grad_norm": 0.8951007183420628, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 52863 + }, + { + "epoch": 0.52864, + "grad_norm": 1.4474528427835336, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 52864 + }, + { + "epoch": 0.52865, + "grad_norm": 1.135186626275248, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 52865 + }, + { + "epoch": 0.52866, + "grad_norm": 1.0457758667288743, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 52866 + }, + { + "epoch": 0.52867, + "grad_norm": 1.1460733453318612, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 52867 + }, + { + "epoch": 0.52868, + "grad_norm": 0.9811224843328563, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 52868 + }, + { + "epoch": 0.52869, + "grad_norm": 1.13899842372293, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 52869 + }, + { + "epoch": 0.5287, + "grad_norm": 0.9540460212432307, + "learning_rate": 0.003, + "loss": 4.003, + "step": 52870 + }, + { + "epoch": 0.52871, + "grad_norm": 0.9731050008065587, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 52871 + }, + { + "epoch": 0.52872, + "grad_norm": 1.1583152939729378, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 52872 + }, + { + "epoch": 0.52873, + "grad_norm": 1.0851518637110487, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 52873 + }, + { + "epoch": 0.52874, + "grad_norm": 0.9531909050396543, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 52874 + }, + { + "epoch": 0.52875, + "grad_norm": 0.9841987510833801, + "learning_rate": 0.003, + "loss": 4.029, + "step": 52875 + }, + { + "epoch": 0.52876, + "grad_norm": 1.0723591381189939, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 52876 + }, + { + "epoch": 0.52877, + "grad_norm": 1.037630175046916, + "learning_rate": 0.003, + "loss": 4.008, + "step": 52877 + }, + { + "epoch": 0.52878, + "grad_norm": 1.0077272751377093, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 52878 + }, + { + "epoch": 0.52879, + "grad_norm": 1.0718414452544367, + "learning_rate": 0.003, + "loss": 4.041, + "step": 52879 + }, + { + "epoch": 0.5288, + "grad_norm": 1.216023348463032, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 52880 + }, + { + "epoch": 0.52881, + "grad_norm": 0.975781600186803, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 52881 + }, + { + "epoch": 0.52882, + "grad_norm": 1.337159655839678, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 52882 + }, + { + "epoch": 0.52883, + "grad_norm": 0.8979836244434212, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 52883 + }, + { + "epoch": 0.52884, + "grad_norm": 0.7251655251628458, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 52884 + }, + { + "epoch": 0.52885, + "grad_norm": 0.7835417963389124, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 52885 + }, + { + "epoch": 0.52886, + "grad_norm": 0.8171500413197258, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 52886 + }, + { + "epoch": 0.52887, + "grad_norm": 1.0143564208295013, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 52887 + }, + { + "epoch": 0.52888, + "grad_norm": 1.2916493138456657, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 52888 + }, + { + "epoch": 0.52889, + "grad_norm": 0.8126596282089772, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 52889 + }, + { + "epoch": 0.5289, + "grad_norm": 1.0685019600873547, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 52890 + }, + { + "epoch": 0.52891, + "grad_norm": 1.242501796482303, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 52891 + }, + { + "epoch": 0.52892, + "grad_norm": 0.9676001433750866, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 52892 + }, + { + "epoch": 0.52893, + "grad_norm": 0.9198666689629015, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 52893 + }, + { + "epoch": 0.52894, + "grad_norm": 1.1366946320806837, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 52894 + }, + { + "epoch": 0.52895, + "grad_norm": 1.1532932014263675, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 52895 + }, + { + "epoch": 0.52896, + "grad_norm": 0.9893681423657655, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 52896 + }, + { + "epoch": 0.52897, + "grad_norm": 0.9957199841560957, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 52897 + }, + { + "epoch": 0.52898, + "grad_norm": 1.1908705351757227, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 52898 + }, + { + "epoch": 0.52899, + "grad_norm": 0.9748568905269477, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 52899 + }, + { + "epoch": 0.529, + "grad_norm": 1.1750592923717216, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 52900 + }, + { + "epoch": 0.52901, + "grad_norm": 0.9499212524822602, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 52901 + }, + { + "epoch": 0.52902, + "grad_norm": 1.102347153719496, + "learning_rate": 0.003, + "loss": 4.046, + "step": 52902 + }, + { + "epoch": 0.52903, + "grad_norm": 1.171139846689073, + "learning_rate": 0.003, + "loss": 4.036, + "step": 52903 + }, + { + "epoch": 0.52904, + "grad_norm": 0.9734802754422426, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 52904 + }, + { + "epoch": 0.52905, + "grad_norm": 1.000783163378802, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 52905 + }, + { + "epoch": 0.52906, + "grad_norm": 1.1520631469187768, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 52906 + }, + { + "epoch": 0.52907, + "grad_norm": 1.0300854576397653, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 52907 + }, + { + "epoch": 0.52908, + "grad_norm": 1.1302276887274596, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 52908 + }, + { + "epoch": 0.52909, + "grad_norm": 1.1593931415731142, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 52909 + }, + { + "epoch": 0.5291, + "grad_norm": 1.2885918441001605, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 52910 + }, + { + "epoch": 0.52911, + "grad_norm": 0.9635021828468441, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 52911 + }, + { + "epoch": 0.52912, + "grad_norm": 1.1055012546111318, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 52912 + }, + { + "epoch": 0.52913, + "grad_norm": 1.2373379243941833, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 52913 + }, + { + "epoch": 0.52914, + "grad_norm": 1.1849755239558135, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 52914 + }, + { + "epoch": 0.52915, + "grad_norm": 0.9158965730096069, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 52915 + }, + { + "epoch": 0.52916, + "grad_norm": 1.0274984334831903, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 52916 + }, + { + "epoch": 0.52917, + "grad_norm": 1.0217206772551717, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 52917 + }, + { + "epoch": 0.52918, + "grad_norm": 1.0404579760421169, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 52918 + }, + { + "epoch": 0.52919, + "grad_norm": 1.0460909109069871, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 52919 + }, + { + "epoch": 0.5292, + "grad_norm": 1.0498916626554893, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 52920 + }, + { + "epoch": 0.52921, + "grad_norm": 0.953267532044876, + "learning_rate": 0.003, + "loss": 4.019, + "step": 52921 + }, + { + "epoch": 0.52922, + "grad_norm": 0.966146545794737, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 52922 + }, + { + "epoch": 0.52923, + "grad_norm": 1.1630500301704059, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 52923 + }, + { + "epoch": 0.52924, + "grad_norm": 0.9018355244392726, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 52924 + }, + { + "epoch": 0.52925, + "grad_norm": 1.0712103243595865, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 52925 + }, + { + "epoch": 0.52926, + "grad_norm": 1.1861191604314305, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 52926 + }, + { + "epoch": 0.52927, + "grad_norm": 1.1076589572787745, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 52927 + }, + { + "epoch": 0.52928, + "grad_norm": 1.1842459773088652, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 52928 + }, + { + "epoch": 0.52929, + "grad_norm": 1.0750648016488373, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 52929 + }, + { + "epoch": 0.5293, + "grad_norm": 0.9918290447303982, + "learning_rate": 0.003, + "loss": 4.03, + "step": 52930 + }, + { + "epoch": 0.52931, + "grad_norm": 1.328844467843608, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 52931 + }, + { + "epoch": 0.52932, + "grad_norm": 0.9961127379091788, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 52932 + }, + { + "epoch": 0.52933, + "grad_norm": 1.0882805284516157, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 52933 + }, + { + "epoch": 0.52934, + "grad_norm": 0.9596433293716248, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 52934 + }, + { + "epoch": 0.52935, + "grad_norm": 1.0134204763949692, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 52935 + }, + { + "epoch": 0.52936, + "grad_norm": 1.2547494483824178, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 52936 + }, + { + "epoch": 0.52937, + "grad_norm": 1.026675460875538, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 52937 + }, + { + "epoch": 0.52938, + "grad_norm": 1.2610853193558802, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 52938 + }, + { + "epoch": 0.52939, + "grad_norm": 0.871669027168624, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 52939 + }, + { + "epoch": 0.5294, + "grad_norm": 0.9428543679396348, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 52940 + }, + { + "epoch": 0.52941, + "grad_norm": 1.218516514144633, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 52941 + }, + { + "epoch": 0.52942, + "grad_norm": 1.027794538519091, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 52942 + }, + { + "epoch": 0.52943, + "grad_norm": 1.1365230841252072, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 52943 + }, + { + "epoch": 0.52944, + "grad_norm": 0.9747907862005628, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 52944 + }, + { + "epoch": 0.52945, + "grad_norm": 1.0804635751180756, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 52945 + }, + { + "epoch": 0.52946, + "grad_norm": 1.068205469317339, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 52946 + }, + { + "epoch": 0.52947, + "grad_norm": 1.1822710927189974, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 52947 + }, + { + "epoch": 0.52948, + "grad_norm": 0.8904511055114063, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 52948 + }, + { + "epoch": 0.52949, + "grad_norm": 0.9711886670274702, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 52949 + }, + { + "epoch": 0.5295, + "grad_norm": 1.1173451843298072, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 52950 + }, + { + "epoch": 0.52951, + "grad_norm": 1.0303221931812894, + "learning_rate": 0.003, + "loss": 4.024, + "step": 52951 + }, + { + "epoch": 0.52952, + "grad_norm": 1.0245330807250257, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 52952 + }, + { + "epoch": 0.52953, + "grad_norm": 1.2229175125252034, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 52953 + }, + { + "epoch": 0.52954, + "grad_norm": 1.1155643472975527, + "learning_rate": 0.003, + "loss": 4.023, + "step": 52954 + }, + { + "epoch": 0.52955, + "grad_norm": 1.0841325826071813, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 52955 + }, + { + "epoch": 0.52956, + "grad_norm": 1.0914741762766955, + "learning_rate": 0.003, + "loss": 4.043, + "step": 52956 + }, + { + "epoch": 0.52957, + "grad_norm": 1.043151496823044, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 52957 + }, + { + "epoch": 0.52958, + "grad_norm": 1.0905604069982462, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 52958 + }, + { + "epoch": 0.52959, + "grad_norm": 1.0399909414055477, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 52959 + }, + { + "epoch": 0.5296, + "grad_norm": 1.0593645691422189, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 52960 + }, + { + "epoch": 0.52961, + "grad_norm": 1.0277894248390036, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 52961 + }, + { + "epoch": 0.52962, + "grad_norm": 0.9137904605256356, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 52962 + }, + { + "epoch": 0.52963, + "grad_norm": 0.9310558370134988, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 52963 + }, + { + "epoch": 0.52964, + "grad_norm": 1.1569877517415927, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 52964 + }, + { + "epoch": 0.52965, + "grad_norm": 1.1090868006715455, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 52965 + }, + { + "epoch": 0.52966, + "grad_norm": 1.0039601368388933, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 52966 + }, + { + "epoch": 0.52967, + "grad_norm": 1.2729896814216521, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 52967 + }, + { + "epoch": 0.52968, + "grad_norm": 0.7381994984496126, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 52968 + }, + { + "epoch": 0.52969, + "grad_norm": 0.8332438475204156, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 52969 + }, + { + "epoch": 0.5297, + "grad_norm": 0.8620863195809036, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 52970 + }, + { + "epoch": 0.52971, + "grad_norm": 0.9565543539411129, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 52971 + }, + { + "epoch": 0.52972, + "grad_norm": 1.19254301943154, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 52972 + }, + { + "epoch": 0.52973, + "grad_norm": 0.9076436835174148, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 52973 + }, + { + "epoch": 0.52974, + "grad_norm": 1.412803578773922, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 52974 + }, + { + "epoch": 0.52975, + "grad_norm": 1.0817883613086618, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 52975 + }, + { + "epoch": 0.52976, + "grad_norm": 1.1483427222517832, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 52976 + }, + { + "epoch": 0.52977, + "grad_norm": 0.8402929951924524, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 52977 + }, + { + "epoch": 0.52978, + "grad_norm": 0.8774551603903687, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 52978 + }, + { + "epoch": 0.52979, + "grad_norm": 1.0038092952538211, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 52979 + }, + { + "epoch": 0.5298, + "grad_norm": 1.1624934429941747, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 52980 + }, + { + "epoch": 0.52981, + "grad_norm": 1.100743448845744, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 52981 + }, + { + "epoch": 0.52982, + "grad_norm": 1.174167974808692, + "learning_rate": 0.003, + "loss": 4.032, + "step": 52982 + }, + { + "epoch": 0.52983, + "grad_norm": 0.9489636557916324, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 52983 + }, + { + "epoch": 0.52984, + "grad_norm": 1.0487722767537948, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 52984 + }, + { + "epoch": 0.52985, + "grad_norm": 1.0648634149330656, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 52985 + }, + { + "epoch": 0.52986, + "grad_norm": 1.044681054924733, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 52986 + }, + { + "epoch": 0.52987, + "grad_norm": 1.0691009454928906, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 52987 + }, + { + "epoch": 0.52988, + "grad_norm": 1.068907596289623, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 52988 + }, + { + "epoch": 0.52989, + "grad_norm": 1.0045333600280049, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 52989 + }, + { + "epoch": 0.5299, + "grad_norm": 1.1952702921623308, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 52990 + }, + { + "epoch": 0.52991, + "grad_norm": 0.9818918210768753, + "learning_rate": 0.003, + "loss": 4.038, + "step": 52991 + }, + { + "epoch": 0.52992, + "grad_norm": 1.2040463423148886, + "learning_rate": 0.003, + "loss": 4.038, + "step": 52992 + }, + { + "epoch": 0.52993, + "grad_norm": 0.9828844926849072, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 52993 + }, + { + "epoch": 0.52994, + "grad_norm": 1.0978057354762074, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 52994 + }, + { + "epoch": 0.52995, + "grad_norm": 1.063035126067899, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 52995 + }, + { + "epoch": 0.52996, + "grad_norm": 1.0940085819144016, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 52996 + }, + { + "epoch": 0.52997, + "grad_norm": 1.1055801756357742, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 52997 + }, + { + "epoch": 0.52998, + "grad_norm": 0.8855424664200459, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 52998 + }, + { + "epoch": 0.52999, + "grad_norm": 1.1446122123792726, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 52999 + }, + { + "epoch": 0.53, + "grad_norm": 1.1710299341822175, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 53000 + } + ], + "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": 2.101113071337472e+18, + "train_batch_size": 1024, + "trial_name": null, + "trial_params": null +}